ML20209G254

Transcript of Kp Bryan Statement Re Leak Rate Data Falsification

ML20209G254
Person / Time
Site:	Three Mile Island
Issue date:	09/11/1986
From:	Bryan K AFFILIATION NOT ASSIGNED
To:
Shared Package
ML20209G260	List: ML20209G254 ML20209G267 ML20209G282 ML20209G289
References
CON-#386-656 LRP, NUDOCS 8609120408
Download: ML20209G254 (6)

Text

,

a d

UNITED STATES OF AMERICA DOCKETED NUCLEAR REGULATORY COMMISSION USNRC BEFORE THE PRESIDING BOARD R249 0FFICE OF SEUciWY 00CKEIlhG h SiiVICL BR;,NCH

)

In the Matter of )

)

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

FALSIFICATION )

)

PREPARED STATEMENT OF KENNETH P. BRYAN My name is Kenneth Paul Bryan. I currently reside in Augusta, Georgia. I am a self-employed consultant.

I began employment with Metropolitan Edison in 1967 at the Crawford Station. In 1969, I started at TMI Unit 1, and I obtained my license as a control room operator in 1974. In 1976, I became a shift foreman at Unit 1. In 1978, I became a dual-licensed shift supervisor. I left TMI in 1980 and was employed by Quadrex, Inc. for approximately a year and a half. ,

I then became self-employed. By my own choice, I am no longer licensed to operate a nuclear power plant.

Through the time of the March 1979 accident, I was a shift supervisor assigned to "F" shift. The shift foreman in Unit 2 was Carl Guthrie and the control room operators were Hugh McGovern and Earl Hemmila. Leonard Germer was a trainee. As 8609120408 860911 PDR ADOCK 05000320 T eoa OSO3

4 shift supervisor, I oversaw the operation of Units 1 and 2. My duties included keeping track of all surveillances and seeing that they were performed. The time I spent in each Unit changed on a daily basis depending upon plant conditions; however, more of my time was spent in Unit 2 until March 1979.

I then spent more time in Unit 1 supervising refueling.

Whenever I worked day shift, I attended plan of the day (POD) meetings. At these meetings, we discussed any major problems with operations. I recall that we discussed personnel and administrative concerns at monthly shift supervisors' meetings.

Unit 2's technical specifications required that a successful leak rate test be performed every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Company policy required us to perform one every shift. By obtaining a test with unidentified leakage less than 1 gpm, we extended the 72-hour time clock. If the computer printout indicated that the unidentified leakage was high or excessively low, it was my understanding that the operators would evaluate it based on plant parameters including makeup tank levels, radiation levels, and so forth. If nothing indicated why there was a change from previous leak rates, the test performer would discard that printout and initiate another. It was my understanding that we had to enter the action statement if we obtained a leak rate higher than 1 gpm that we could not invalidate. I did believe that if I had looked at everything else and did not think that I had obtained a valid leak rate, I -

did not have to start the time clock.

I never specifically instructed my shift how to interpret leak rates because this subject was part of their training on technical specifications. I do not recall receiving instructions from any of my superiors as to the way in which I should construe the technical specification governing the leak rate test.

On my shift, the control room operator would decide when and how often to run a leak rate test. My permission was not required. Often, leak rate tests were not performed on day shift, because the computer was being used for other functions. Each control room operator could decide whether to discard the leak rate printouts. There was nothing recretive about this process.

Unit 2's leak rate numbers varied quite widely, and more leak rates were determined invalid in Unit 2 than in Unit 1. I attributed the inconsistency of the numbers to the fact that Unit 2 was not yet as fine-tuned as Unit 1. As I recall, tests with unidentifed leakage less than 1 gpm were getting harder to obtain as we approached the time of the accident.

The start time for leak rate tests was recorded on the computer printout. I do not remember being aware that the e control room operators also had to record the start time in their logbook. My only explanation for our not recording th-e 4

I s

4 start time is that we must have overlooked it. I also can no longer recall whether water additions made during a leak rate were required to be recorded in the logbook. As far as I can remember, water additions were recorded as a general practice.

i I thought that water could be added during a leak rate test if it were properly accounted for. I would not have accepted a i

test if I knew an operator had not accounted for a water addition, but I am unaware of the existence of any tests where this occurred. I am also unaware of any operators who " jogged" f water during a leak rate test to deliberately affect the result. I could understand, however, if some operators occassionally failed to communicate with each other, causing an

! unaccounted-for water addition.

l I remember being aware that the addition of hydrogen affected makeup tank level. I could not explain how this phenomenon worked, but I was aware that hydrogen caused a change in level. I did not first learn of this by experimenting with hydrogen during a leak rate test; I observed its effect on the makeup tank during normal operations. I l .

I concluded that hydrogen would affect the leak rate test, t

because it changed the makeup tank level. I do not recall having any specific discussions on the subject with supervisory personnel or control room operators. I was unaware of anyone I

who deliberately added hydrogen during leak rate test to affect the test's results.

i t

, ~4-E __ _ . . . . ._ _ _ ._._..____ _._ _ _..._ ____. __ _ __

=

i s I believed that Exception and Deficiency sheets did not have to be applied to leak rates determined invalid due to computer error, variance with plant parameters, or whatever.

For purposes of the "E and D" procedure, the leak rate was 4

treated differently than other surveillances, because the other surveillances were not computer dependent. I did not consider a computer printout to be a surveillance until it was approved 4

and a cover sheet was attached for filing.

Because I was used to seeing negative leak rates on a fairly frequent basis from my Unit 1 days, I permitted the filing of negative leak rate tests as valid. I thought there was no reason why the instruments could not select a number on the other side of zero. My shift made a judgment call on

< whether to file a particular negative number; there was no hard and fast rule on what negative number would be accepted.

I have no recollection of reviewing the November 1, 1978 i LER, or of discussing it with my shift or with my superiors. I I also have no recollection of reading the October 20, 1978 operations memorandum. Neither document changed my interpretation of the technical specifications for leak rate tests. I have some vague recollection of the March 16, 1979 TCN because I remember that we were attempting to account for l the density change in water. I do not remember if it improved our ability to obtain satisfactory leak rate tests.

I do not know of anyone who deliberately falsified a leak l

, - - , _ . _ . _ - _ . _ . , _ - . _ . . - .-__.,--r, . . . , - _ , . . , _ _ , . . .,,._,v._. ..-_ _ _ _ - - - - - . . _ . . , _ _ , _ - . . - _ , . , - . - _ _ - - _ . _, . m,.,__m - _ _ , _ . - - -

rate tast. I do not know of any supervisory personnel who would have directed anyone to do something like that. I knew that operators were experiencing problems obtaining satisfactory leak rate tests, but we were trying to find out what was wrong with the test. I remember that we continuously looked for leaks, we looked at instrumentation; in fact, we looked at anything we could to rectify the problem.

_. _ ~ .. . . - _ - _ - - _ _