ML20205F580
| ML20205F580 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 08/15/1986 |
| From: | Mcbride M LEBOEUF, LAMB, LEIBY & MACRAE |
| To: | Bright G, Kelley J, Kline J Atomic Safety and Licensing Board Panel |
| References | |
| CON-#386-335 LRP, NUDOCS 8608190251 | |
| Download: ML20205F580 (116) | |
Text
. T agexteD countsruh"6 '
g LEBOEU F, LAM B, LElBY & MAC RAE A PARTNER 5 NIP INCLUDING PROFESSIONAL CORPORATIONS 1333 N Ew H AM PS HIR E AVENU E, N.W.
WASHINGTON, DC 20036
'86 ale 15 P4 :58
'202) 457 7500 TcLex: 440a74 rACsiwiLc: aoa 4s7-7s" New york, Ny 0FFICE OF Sluit iesosvoN, MA 00CKEIg r,ggtf,{,cor, c, CALT LANC CITY, UT A
G H, N C ALBANY, NY C AN FR ANCISCO, CA N EWAR K, NJ LECOEUF, LAMB. LElBY & MACRAC (UK)
EDISON, NJ LONDON, ENG LAN D August 15, 1986 BY HAND James L. Kelley, Esq.
Mr. Glenn O. Bright Mr. Jerry R. Kline Administrative Judges Nuclear Regulatory Commission Washington, D.C.
20555 Re:
Inquiry Into Three Mile Island Unit 2 Leak Rate Data Falsification, Docket No. LRP
Dear Sirs:
Enclosed are six copies of each of the prefiled statements of most of the Numerous Employees who are parties to this proceeding.
We have enclosed two sets of the statements for each Board member for the convenience of the Board.
In accordance with the provision for an extension of time granted by the Board.
during the July 31, 1986 preparing conference, we will file in the near future, and as soon as they are available, the prefiled statements of Messrs. Booher, Guthrie, Hoyt, Mehler, Scheimann, and Zewe.
We were unable to file the prefiled testimony of these few parties because of logistical difficulties.
We appreciate the Board's indulgence in allowing us a modest extension of time to file the remaining prefiled testimony.
We have been served with a letter from the Board to Mr.
James Seelinger, who is our client, informing him that he has been subpoenaed and that he is to file his prefiled testimony (if any) on or before September 8.
We hereby request that he have at least until. September 29, 1986 to do so.
There.are several reasons for this request, some of which were already communicated I
to the Board.
One, Mr. Seelinger's entire family was involved in a serious automobile accident which will limit his available time for the preparation of his testimony.
Two, Mr. Seelinger is 86o8190251 860815 PDR ADOCK 05000320 PDR r
i
\\
r m_, a s
y Pysaiding Board August 15, 1986 Pcgs Two extremely preoccupied with his duties for his current employer.
Three, the undersigned, whom Mr. Seelinger has relied upon as counsel for all of his involvement in leak rate matters, has been summoned for jury duty in the Superior Court for the District of Columbia between August 27 - September 9, 1986.
Given the Board's inclination to hear from supervisory personnel (such as Mr. Seelinger) late in the proceedings, this extension should produce no hardship to any party.
The alternative might be that Mr. Seelinger would not be able to prepare any prefiled testimony.
Very truly yours, Michael F. McBride cc (w/ encl. ) :
All Persons on Service List e
F w
.s
.t RELM LD COhKWMM, 00CMETED UNITED STATES OF AMERICA USNRC NUCLEAR REGULATORY COMMISSION 16 NE 15 P4 58 BEFORE THE PRESIDING BOARD OFFICE OF EEL:'1TA H Y 00CMETING & SERVICE BRANCH
)
In the Matter of
)
)
Docket No. LRP INQUIRY INTO THREE MILE ISLAND
)
UNIT 2 LEAK RATE DATA FALSIFICATION
)
)
CERTIFICATE OF SERVICE I hereby certify that I have served this 15th day of August, 1986, a copy of the foregoing documents by first-class mail, postage prepaid and properly addressed upon the following persons:
Administrative Judge Administrative Judge James L. Kelley, Chairman Glenn O. Bright Atomic Safety and Atomic Safety and Licensing Board Panel Licensing Board Panel U.S. Nuclear Regulatory U.S. Nuclear Regulatory Commission Commission Washington, D.C.
20555 Washington, D.C.
20555 (Hand Delivery)
(Hand Delivery)
Administrative Judge Jack R. Goldberg, Esq.
l Jerry R. Kline Office of the Executive Legal l
Atomic Safety and Director l
Licensing Board Panel U.S. Nuclear Regulatory l
U.S. Nuclear Regulatory Commission l
Commission Washington, D.C.
20555 Washington, D.C.
20555 (Hand Delivery)
(Hand Delivery) l l
\\%
t Docketing and Service Branch (3)
_ Ernest L.
Blake, Jr., Esq.
U.S. Nuclear Regulatory Shaw, Pittman, Potts & Trowbridge Commission 1800 M Street, N.W.
Washington, D.C.
20555 Washington, D.C.
20036 (Hand Delivery)
James B.
Burns, Esq.
Michael W. Maupin, Esq.
Isham, Lincoln & Beale Hunton & Williams Three First National Plaza P.O. Box 1535 Suite 5200 Richmond, VA 23212 Chicago, IL 60602 Ms. Marjorie M. Aamodt Ms. MarjoriG M. Aamodt 200 North Chunch Street P.O Box 6S2 Parkesburg, PA 19365 Lake Placid, NY 12946 n w - W. Wa.a s A =
J aines W. Moe41er i
i m
om---m.-seser n
-g.
mm,ww a - - +
=m-W---
- %e'a+d 8=
8"
4 NEtATED WdNt.M c
DOCKETED UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION.,
16 NE 15 P5:01 BEFORE THE PRESIDING BOARD --- ---
CFFICE OF SECMTARY 00CKETING & SEPylCf.
BRAtiCH
)
In the Matter of
)
)
INQUIRV INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
-)
PREPARED STATEMENT OF WILLIAM T.
CONAWAY, II My name is William T.
Conaway, II.
I reside in Wrightsville, Pennsylvania.
I am currently employed by GPU Nuclear as the Radioactive Waste Support Manager at Three Mile Island Unit 2.
I no longer hold an NRC license.
I began working for Metropolitan Edison Company in 1973.
I was an auxiliary operator at Unit 1 until I was promoted to a Control. Room Operator (CRO) at Unit 2 in 1975 and later licensed.
In 1978, I was promoted to Shift Foreman.
My Shift Supervisor was Joe Chwastyk.
I supervised CRO's Ted Illjes and John Kidwell.
At the time of the accident at Unit 2, I was a Shift Foreman.
As a Shift Foreman, I was to supervise the people on my shift.
In the control room, I was responsible for the shift foremen's log and for the general paperwork that came through there.
I also spent a lot of time out of the control room and
i i
in the plant, although I cannot state precisely just how much of my time was-spent in the plant.
As a shift foreman, I did not in general directly supervise the auxiliary operators (AO's).
Usually the control room operators gave instructions to the AO's.
I do not recall attending very many shift foremen meetings.
They were held at Unit I regularly, but I did not always know when they were being held.
In any event, it was not mandatory to attend if you were not on shift.
I do not recall any discussion of leak rate testing or invalid leak rate tests at any shift foremen meeting.
I understand that I was required, as a shift foreman,-to ensure that all surveillance testing, including leak rate testing, was conducted in compliance with NRC regulations.
Although I approved the results of leak rate tests, I do not recall specifically exercising direct supervisory control over the tests.
In fact, there were many times that surveillance tests were begun or completed without my specific knowledge at that time.
I expected the CRO's on my shift to comply with the plant procedures for all testing including leak rate testing, and-owing to the confidence I had in my CRO's I did not look over their shoulders while they ran the test.
We used the leak rate test to demonstrate that the plant was operating within the limiting conditions for operation for reactor coolant system leakage.
However, we verified that the plant was operating within the limiting conditions for -
9 s.
operation by reviewing the plant parameters.
We did not nave a lot of faith in the leak rate test itself.
The Operations Department in general had little confidence in the validity of the leak rate test.
A lot of the tests were not valid.
For example, we got a lot of high negative leak rates.
Often the leak rate did not agree with the actual leakage we determined by reviewing other plant parameters.
We reviewed and relied on the makeup tank level to determine RCS leakage.
In retrospect, we relied too heavily on the makeup tank level.
We also used the containment radiation monitor and the containment sump inventory and discharge to ensure that there was no safety concern that must be addressed.
i I am not as familiar today with the technical specifications and administrative procedures governing the leak rate test as I was in 1978-79.
It has been a very long time since I was required to have an understanding of these requirements.
However, I feel they did not prohibit the addition of water to the system during a leak rate test.
Nor did they prohibit feed and bleed operations.
We were primarily concerned with the safe operation of Unit 2 in 1978-79.
We were not as concerned with the administrative requirement of demonstrating that the plant was operating within the limiting conditions for operation for RCS leakage.
We regarded the actual plant safety as more important than the leak rate tests, the pieces of paper, that were often invalid.
-3
_,r-
3 s.
I do not recall any specific pressure to throw away leak rates above 1 gpm or to avoid entering the action statement, even when Unit 1 was shut down for refueling before the accident.
Nor do I recall feeling pressure from management to get a good leak test, that is, one showing less than 1 gpm of unidentified leakage, even when a lot of time had passed since the last good test.
We attempted to avoid adding water to the RCS during-leak rate tetts.
But we did if it was necessary to change the boron concentration in the RCS or to maintain the MUT level.
We used a calculation sheet to determine just how much water to add to the system for boron concentration changes.
We would then send the calculation sheet to the Ops Engineer for his review.
Usually, we logged water additions but we did not log hydrogen additions.
I did not generally question the CRO's on my shift about operator changes to the system during a leak rate test when I approved the test.
However, if I had known that water had been added during a test yet had not been accounted for, I would not have approved the' test.
I was not aware of any practice to add water during a leak rate test and not account for it in order to affect the test result.
Nor did anyone ever discuss with me unlogged water additions.
We added hydrogen to the MUT to maintain a certain pressure band.
There were two valve operators to add hydrogen located near the MUT strip chart.
Usually, the CRO at the panel opened o
6-these to add hydrogen.
For a period of time, the valve operators did not work, and a bottle of hydrogen in the auxiliary building was used to add hydrogen.
In that case, the CRO had to request an AO to actually add the hydrogen.
I do not recall discussing with anyone prior to the accident hydrogen additions to the MUT to affect leak rate tests.
I was told that there are several tests run by my shift involving allegations of hydrogen additions to affect the results.
But I was never aware of the use of hydrogen additior.s to influence leak rate tests.
Moreover, I am informed that the technical experts disagree in their overall analysis of those tests.
I am informed that Joe Chwastyk indicated that he learned of the effect of hydrogen additions on leak rate tests prior to the accident and instructed the shift not to add hydrogen to the MUT during the tests.
I am informed that he did so after a CRO brought it to his attention and he conducted an experiment to investigate it.
However, I do not know that Joe had any knowledge of the effect prior to the accident or that he instructed us not to add hydrogen during the tests.
Joe could have conducted an experiment but I just do not recall any.
I am informed that I approved the February 19, 1979 experimental leak rate test and that Joe recorded the results of that test in his notes.
However, I have no recollection of being involved in that test, and I do not recall any discussion by the CRO's of the effect of hydrogen additions on leak rate _ _.
(
tests.
Nobody ever brought it to my attention as far as I know, with the exception of Marty Cooper, after the accident.
I never approved the use of hydrogen additions to influence leak rate tests and I never myself used hydrogen for that purpose.
Apparently, I approved a test on February 19, 1979, recorded the results in my log, and later crossed the results out.
I do not recall crossing the results out but I assume I did so because the entry was incorrect.
I do not recall that it had anything to do with the experimental test Joe ran on that day.
I do not recall any knowledge of a problem with one of the MUT level transmitters in 1978-79.
I understand now that there was a problem but this understanding is based on knowledge I have acquired since the accident.
I am informed that there were two work requests in December 1978 that I signed that are for a defective level transmitter.
But I do not recall signing them.
Nor do I recall discussing a bad level transmitter with-the CRO's on my shift or instructing them not to use it.
I was not aware of any practice to switch the bad level transmitter to the computer during a leak rate test in order to get a good test, that is, a test below 1 gpm.
In fact, I do not understand how switching the level transmitter could influence l
l the leak rate test.
I never overheard any discussion among the I
Unit 2 operators regarding the switching of the level l
l transmitter to influence the test.
I am informed that there l
l l --
t A
were several tests run by my shift-in which the bad level transmitter was used.
However, I was not aware of.any practice to use that transmitter to get good-leak rate tests.
In conclusion, I did not participate in or have knowledge of any leak rate test falsification, and I ask the Presiding Board to exonerate me of any alleged involvement in such practices.
I have been in the nuclear industry for thirteen years and would like to stay here.
I would also like to put the issue of TMI-2 leak rate testing behind me.
At the same time, I have learned a lot from the leak rate test investigations and have come to appreciate the value of-conservatism in the interpretation of and compliance with procedures and regulatory requirements.
I will continue to apply these lessons in the future.
4 i
1 i
I e w :. :,esrutetANGE 4
00LMETED UNITED STATES OF AMERICA USNRC NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARD 16 @E 5 P5 :01 0FFICE OF SECF.EiARY 00CMETING & SERVICf.
BRANCH
)
In the-Matter of
)
)
INQUIRY INTO THREE MILE. ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF THEODORE F.
ILLJES My name is Theodore F.
Illjes.
I live in Hershey, Pennsylvania.
I began working for Metropolitan Edison Company in 1971, and am still employed at TMI-2 by GPU Nuclear Corporation.
I became a Control Room Operator (CRO) at Three Mile Island Unit 2 in 1976 and at the present time I am a Shift Supervisor.
Between. March 1978 and March 1979 I was assigned to "B"
shift.
My shift supervisor was Joseph Chwastyk, my foreman was William Conaway, and my fellow CRO's were Charles Mell and John Kidwell.
I graduated from high school in 1961.
I attended Colorado State University for two years.
I worked briefly for the Forest Service, then spent seven years and three months.in the Navy Nuclear Submarine Program.
In 1971, I was hired by Met Ed.
I first went to work at the Crawford Coal Station in Middletown, Pennsylvania.
I spent a few months there, then i
i
___-e,..
9 went to Three Mile Island.
I began at Unit 1 as an auxiliary operator, and moved to Unit 2 in CRO training in 1975.
I was a CRO in Unit 2 until late 1979, when I became a radioactive waste shift foreman.
Six months later, I went back to the Control Room as a shift foreman, and later was promoted to the position of shift supervisor.
CRO's generally ran the plant under the direct supervision of a shift foreman.
The CRO's on my shift and I would usually take turns performing the various responsibilities required.
One of us would log in and control the panel while the other would do the switching and tagging, run surveillances, and do paperwork.
During some shifts we had a third CRO or a trainee who would help out.
Performing leak rate tests was part of the CROs' responsibilities.
At least one leak rate test was usually, but not always, run every shift.
I do not remember, now, the specific procedures required to conduct _ leak rate tests.
I do recall that I did not rely on the leak rate test as the only indicator of leakage.
It would appear, now, that the leak rate test procedures I followed were not always in strict compliance with the requirements imposed on us, in that 1 usually discarded test results showing unidentified leakage greater than one gpm.
These tests were discarded if the test was invalid, once the leak was identified and corrected or a satisfactory test result obtained. -During 1978 and 1979, I thought it sufficient to
'f.-
L obtain at least one leak rate test result below one gpm during a 72-hour period.
Once that result was obtained, the 72-hour clock would start again from that point.
If we had not gotten a satisfactory leak rate test result during a seventy-two hour period, I would have shut the plant down, or recommended that it be done.
I always avoided adding water during leak rate tests, if possible, because that could affect the results.
If water had to be added for some reason during the hour-long test we would note it in the CRO log and on the computer sheet.
Sometimes, inadvertently, this was not done.
I was not aware of anyone ever deliberately adding water to manipulate the leak rate test.
I am certain that I did not know before the accident that the leak rate test could be affected by adding hydrogen to the makeup tank.
I do not recall ever discussing this matter with my shift supervisor, Mr. Chwastyk.
I also do not recall participating in an experiment to determine the effect of adding hydrogen during a leak rate test.
I believe that when Mr. Chwastyk performed his experiment he did not involve me or all of the members of my shift.
I am aware that my shift foreman, William Conaway, has no recollection of this event.
My fellow CRO, Charles Mell, also does not recall witnessing a hydrogen experiment or being told to avoid adding hydrogen during tests.
I am certain, however, that I did not add hydrogen to manipulate leak rate tests.
If it was added while I was performing a test it was presumably because the system 1
~1 needed it.
I believe we logged hydrogen additions when we were not too busy with other matters, but a lack of communication between the auxiliary operators and the control room could have resulted in a failure to log the additions.
My failure to understand the makeup tank's problems is shown by the fact that I signed only one test where I received an alleged benefit because of an error in the system in calculating a water addition.
In my opinion, I have been unfairly accused of improper conduct, despite the almost total lack of evidence to support the accusations.
I did not have the motivation or inclination in 1978 or 1979 to manipulate leak rate tests.
I have also been accused of performing leak rate tests knowing that the result would be meaningless, because the test relied on a faulty transmitter.
I do not recall a problem with a fluctuating transmitter, and I do not recall discussions about the use of this transmitter as a way to falsify tests.
I think it would have been obvious to me that you could change the result of a leak rate test by switching transmitters during the test, but I never did this.
I could, however, have missed the fact that an oscillating transmitter was reading to the computer, and failed to see that the numbers on the computer printout did not match the strip chart.
In summary, the allegations made about the performance of leak rate tests at TMI-2 do not fairly apply to me, except that I did discard leak rate tests with over 1 gpm of unidentified _
(
_(
leakage.
We discarded those tests because we did not believe the results, which were inconsistent with our perception of actual leakage.
We did not discard those tests because we were trying to deceive the NRC, and we would not have operated TMI-2 if we' thought we were doing so in an unsafe manner.
My license is very important to me.
I believe that my diligent performance of my duties, particularly since the TMI-2 accident, should lead to the conclusion that I have learned valuable lessons from this experience.
I am aware of the absolute necessity for strict compliance with the technical specifications and plant procedures.
I believe that I should be permitted to retain my license without sanctions of any kind.
I can assure the Board that if I am exonerated, I will give the Commission and my employer no cause to regret that decision.
. s-L:. i Q) Vfn.f' s #,M&&
9 4
UNITED STATES OF AMERICA U NRC NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARD 15 NE 15 P5:01 0FFICE OF SECFiTAHY 00CKETimi A SERVICE BRANCH
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF CHARLES F. MELL My name is Charles F. Mell.
I live in Carleton, Michigan.
I am self-employed as a nuclear consultant.
In 1976, I began my employment with Metropolitan Edison Company as an auxiliary operator.
I received my license as a control room operator in the summer of 1979.
I was a control room operator in training at the time of the accident in March 1979.
I left Metropolitan Edison in December 1981.
In January of 1979, I was assigned to "B"
shift.
I was on shift with John Kidwell and Theodore Illjes.
William Conaway was the foreman and Joseph Chwastyk the supervisor.
Because I was in training, all the tasks I performed were supervised by a licensed operator.
I learned how to perform a leak rate on the job.
No one gave me specific training other than the directions I received from the shiftmates who were supervising me.
There was no set
i h
time on our shift for performing a leak rate, but it was one of the things that we tried to do on each shift.
I thought that each shift performed leak rates in the same basic way.
We had 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> to obtain a leak rate test with unidentified leakage less than one gallon per minute.
I did not believe that I had to enter the Action Statement, if I obtained a leak rate over 1 gpm, as long as no more than 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> had gone by since the last one.
I thought the 72-hour time period was the factor that controlled when we might have to enter the Action Statement.
A s f ar as I can remember, everyone on my shift interpreted the procedure the same way.
If I did not obtain a good leak rate, I would run another leak rate test.
I discarded tests that were higher than one gpm when I obtained a satisfactory test.
This was a common practice that I learned on shift.
I did not have to show the foreman the leak rate before throwing it away.
I do not recall that I ever knew the reason for the practice of discarding leak rates, but I do remember believing that the plant parameters did not show excessive leakage.
I viewed running the leak rate surveillance as less important than some of the other tasks assigned to me.
My shift rotated the responsibility for running the leak rate, so I might have run it every third day.
I recall occasions when we did not obtain a satisfactory leak rate on my shift.
?s 4
I do not recall ~whether we ever entered the Action Statement.
I remember that it became harder to get a good leak rate closer to the accident, because we had a valve that was leaking.
I do not think that I knew which valve it was that was leaking.
I never deliberately added water during a leak rate without recording it.
Sometimes, I did add water if the system needed it, but I always tried to record it.
My shift occasionally had a problem communicating to the operator on the panel, who was adding water, that a leak rate test was in progress.
We tried to put up a sign on the batch controller indicating that a leak rate test was underway, but others in the control room still could have been unaware that a test was being performed.
In 1978 and 1979, I was unaware that anyone purposely added water knowing it would be under-recorded, because of an effect on a makeup tank level transmitter.
I was also unaware that hydrogen could affect the leak rate test or that anyone was deliberately trying to do this.
I remember that we added hydrogen only in response to a directive from the chemistry department for corrosion control.
I saw no problem with adding hydrogen during a loak rate.
The addition of hydrogen should have been logged by the control room operator and the auxiliary operator, who made the actual addition.
I do not recall the performance of an experiment with hydrogen on my shift.
I do not remember Joseph Chwastyk
- 1. -.
1, cautioning us not to add hydrogen during a leak rate; in my opinion, if he had noticed a problem with hydrogen, he would have kept it close to his chest until he had so fully checked it out that he could explain it.
I recall that I first discussed the hydrogen phenomenon with my crew after the Hartman allegations were made known, and we were all surprised by Hartman's claims.
I remember that I sometimes filed negative leak rates.
I thought it was normal to obtain negatives because of slight perturbations with pressure and temperature differences in the system.
I have been confronted with the fact that the files contain a negative 8 gpm leak rate test conducted by me.
I do not remember this test, but its existence does suggest that a leak rate might have been accepted without sufficient consideration of whether it was genuinely valid or invalid.
I also remember that for a time, we had a problem with the level transmitters.
One was worse than the other.
I recall' trying to use the least erratic one on the computer when running a leak rate.
Switching a transmitter during a test I
would have thrown off the calculatC.h.
As far as I know, no l
one on my shift ever deliberedil s'.tched transmitters to get l-a good leak rate.
I remember hearing that the NRC had found a bad leak rate i
once, which created some fuss.
I have no recollection of the November 1, 1978 Licensee Event Report, even though it appears l
that I signed the routing sheet.
I remember that we rounded i i
s k
off leak rates at one time; but I never knew much about it and just did what I was told for that period.
I cannot recall receiving any specific' training on the meaning of the 72-hour-requirement either beforeaor after the LER was issued.
I decided to become a party to these proceedings because I
=would like to continue as a nuclear consultant; and I may wish to study for an NRC license in the future.
I have a unique advantage in this business because I have been involved in implementing changes that resulted from the TMI-2 accident.
I 4
can bring a depth to that process that someone who has not been through this cannot approach.
I hope that the NRC finds that 7
my past. conduct does not warrant my removal from the industry, c
1 F
-l m
I:
r l
l
---r
pp*M a
Q rn:J \\ *3 o
DOCKETED UNITED STATES OF AMERICA USNRC NUCLEAR REGULATORY COMMISSION 36 ME 15 PS:d1 BEFORE THE PRESIDING BOARD 0FFICE OF SECEtit,R f 00CMETING & SEPVjCf.
BRANCH
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF JOSEPH J. CHWASTYK My name is Joseph J.
Chwastyk.
I live in Enola, Pennsylvania.
I am employed by Coldwell Banker MGM Realty, Inc.
I was in the Navy from 1962 to 1967.
In 1968, I joined Metropolitan Edison at the Crawford Generating Station.
I went to Unit 1 in 1969 as a Control Room Operator.
In 1972 or 1973, I became a shift foreman and later in that year a shift supervisor.
In 1977 or 1978, I became a dual-licensed shift supervisor at Unit 2.
After the accident, I became Unit 2's Supervisor of Operations.
That position evolved into Manager of Plant Operations, Unit 2.
I left GPU Nuclear in 1983.
To qualify for a license at Unit 2, I went through a training program.
Basically, I learned the Unit's systems, then took an NRC exam.
I think all this occurred during Unit 2's start-up.
I probably ran leak rates as part of my training, but I do not recall any specific training on the subject.
A Initially, most of my time was spent in Unit 1.
That changed as we advanced in the start-up program to the point where I spent about 60% of my time in Unit 2.
My operators in Unit 2 were Charles Mell, Theodore Illjes, and John Kidwell.
William Conaway was the shift foreman.
The crew was designated as "B"
shift.
As shift supervisor, I was responsible for ensuring compliance with the technical specifications.
I was responsible for what happened on shift and for the safety of the reactor. I exchanged information with the other shift supervisors either through turnover notes or through Plan of the Day (POD) meetings.
POD's were a useful way of exchanging information if you were on day shift.
Because I worked day shift only two out of six weeks, I often relied on my turnover notes to communicate information to the other shifts.
In Unit 2, a leak rate showing unidentified leakage under 1 gallon per minute had to be obtained once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
In Unit 1, we were asked to run leak rates more frequently than was required; therefore, we carried this policy over to Unit 2.
This was done to make sure we stayed within technical specifications by always having a satisfactory test that was not more than 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> old.
We started out, I think, by doing a leak rate every day, but as we developed. problems with the computer, we changed this to running one per shift.
In Unit 2, we had problems getting the computer to do an accurate calculation.
I think that we all recognized that we l l
r
.A t-had a problem with the leak rate calculation.
I recall that this was discussed at the shift supervisors' level.
Unit 2 leak rate results were much more erratic than those in Unit 1.
I could not understand how we could have such a good computer program in Unit 1 and yet have so many problems in Unit 2.
I saw numbers as high as positive 5 or 6 gpm and as low as negative 5 or 6 gpm.
I complained to the computer programmers about this problem and they were trying to work it out.
On my shift, if we obtained a leak rate over 1 gpm, we would run another leak rate; if the next one was satisfactory, we would throw out the first one.
Because we viewed leak rates higher than 1 gpm as computer calculation errors, I did not tell my operators to attach Exception and Deficiency sheets to them.
I was not aware of anyone in Unit 1 or Unit 2 who felt that the E & D procedure applied to leak rates.
I normally did not review leak rate data.
The shift foreman reviewed it as part of his duties, and the foreman could decide whether to discard a leak rate.
I cannot recall whether the operator could decide to discard a leak rate without the foreman's approval.
My permission was not needed.
I know that we got some negative leak rates in Unit 2.
I do not know how it was decided whether they were accepted.
I think it depended on the shift foreman and the control room operator.
They determined how negative a one they would accept, depending upon how realistic it seemed.
There were reasons why one could obtain a negative leak rate other than L _
(-
.A e
erroneously accounting for a water addition, including instrument errors, a spike in the plant, or a slight change in power.
All of our logging practices were carried over from Unit 1.
I think that we logged the completion, only, of a satisfactory leak rate test.
This practice was not instituted to hide bad leak rates from the NRC, because we did not have a
. problem with leak rates in Unit 1.
I recall that I was familiar with the November 1, 1978 Licensee Event Report before these investigations started.
As far as I knew, it did not change anyone's interpretation of the technical specifications.
I have no recollection of talking to NRC Inspector Donald Haverkamp about leak rates. I do recall receiving a briefing about leak rates over 1 gpm that had been found.
I also recall walking into the shift supervisors' office when James Seelinger was announcing to someone over the telephone that he was going to file an LER.
I did not hear the other person's voice, and I do not know who Jim was talking to.
I remember Jim stating that he wanted to declare a leak rate valid and write up an LER.
I interjected that we should l
be trying to invalidate the leak rate because of our computer I
problems, but he indicated that my point had been discussed and i
rejected.
At that point I left the shift supervisor's office.
I read the United States Attorney's Statement of Facts, which stated that a shift supervisor would testify that the telephone call was made by Mr. Seelinger to either TMI's,
L
i
.A I
A Station Superintendent (Gary Miller) and/or the Vice President for Generation (Jack Herbein), or both.
At the time I read this, I assumed the shift supervisor referred to was someone who was in the office when the call was made.
I do not know who Mr. Seelinger was talking to, or who placed the call.
During an interview with United States Attorney Queen's assistants, I specifically told them this.
When they asked about who it might have been, I agreed that it could have been Mr. Miller or Mr. Herbein, or both, but I had no way of knowing.
To me, the lesson of the LER was that if a leak rate was determined invalid, make sure to throw it away.
I did not not think the LER had anything to do with interpretation of the technical specifications.
I think that the LER was filed because it was not possible to validate or invalidate the bad leak rates, given the amount of time that passed between running the tests and when they were found by Mr. Haverkamp.
It was brought to my attention in Unit 2 that if you added hydrogen to the makeup tank while you were performing a leak rate test, you would more likely get a satisfactory leak rate.
Because I did not believe what I was hearing, I conducted my own test and hydrogen did seem to affect the leak rate test.
I do not recall definitely who was present during the test.
I remember talking to an I & C supervisor immediately after drawing my conclusion.
I also recall making a statement to "B"
shift CRO's and a shift foreman to the effect that they were to ensure they did not add hydrogen while doing a leak rate test.,
i o-I do not recall hydrogen being a topic at a shift supervisors' meeting.
I was unaware of anyone who intentionally added hydrogen-knowing it would affect the leak rate test.
I was also unaware of anyone who deliberately added water to the makeup tank and failed to record it on the leak rate sheet, or added water to take advantage of an error in the system.
In my opinion, no undue pressure was put on anyone to get a good leak rate test.
perhaps some felt pressure because leak rates were a matter of concern to everyone.
I frequently asked my shift whether they
'had gotten a good leak rate, or whether they had run another one.
It now appears that there were practices that some find i
questionable associated with leak testing at TMI-2.
I would like to emphasize that in 1978-79, I did not think my shift was doing anything wrong.
I felt that I understood the leak rate technical specification requirements, and I thought my shift was following them.
I became a party to these proceedings because I wanted to clear my name of my alleged involvement in any improprieties as a result otf my employment at TMI, including leak rate testing.
I have always tried to testify truthfully because I believe that is my obligation.
However, so many years have passed that it is difficult to recall specific events.
I did my job as well as I could, believing that what I was doing was proper. -
e DOCKETED USNRC UNITED STATES OF AMERICA
. NUCLEAR REGULATORY COMMISSION 16 NE 15 P5 :01 BEFORE THE PRESIDING BOARD OFFICE 0F SECRETARY, 00CKEitNG & SERVICF.
BRANC"
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
i PREPARED STATEMENT OF CHARLES D. ADAMS My name is Charles D. Adams.
I live in Annville, pennsylvania and am currently employed by Gr0 Nuclear Corporation as a -special project assistant working on the TMI-1 Probablistic Risk Assessment.
I served in the United States Navy for 8 1/2 years as an electronics technician.
Upon my discharge from the Navy in 1972, I went to work for Carolina Power and Light Company at their nuclear facility in Southport, North Carolina.
At Carolina Power I was a control room operator and obtained my senior reactor license.
I began employment with Metropolitan Edison Company on October 6, 1975, and during 1978 and 1979, I was a shift foreman in Unit 2.
My shift was C shift and the shift supervisor was Brian Mehler.
The control room operators under my supervision were Joseph Congdon and Martin Cooper.
At some point during 1978, Mark Phillippe was assigned to my shift as a trainee.
.a.
's Generally, my duties as shift foreman were to supervise the control room operators and auxiliary operators, to oversee the operation of the plant and perform various administrative duties.
During the course of a shift, I might spend 7 or'8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> in the Unit 2 control room, or maybe only an hour or two with the rest of the time out in the plant.
I was generally familiar, however, with any unusual problems or plant evolutions that might exist during the course of the shift.
At the beginning of each shift I would direct the operators to perform the evolutions and perform the surveillances that were scheduled, and I would review the CRO log at the end of the shift.
One of the surveillances that the operators performed on a shiftly basis was the leak rate test.
The technical specifications required that a test be performed at least once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, but as a matter of practice we attempted to get one every shift.
If the computer printout reflected unidentified leakage in excess of one gallon per minute, the operator would start another leak rate test.
There were times when.three or four tests might be performed during the course of our shift.
Usually, we would retain the leak rates over one gallon per minute until we obtained one that was within the technical specification limits, and then discard those that were outside of the limits.
It was my belief in 1978 and 1979 that the leak rate test program was not accurate in that it could not give you the.
.. - - ~ _
a precise unidentified leakage, unless the plant was in a stable condition.
However, at that time the plant was new, and when operating at various power levels, there were oscillations in the secondary side which caused the primary side of the plant to oscillate.
This resulted in an unstable condition, and it was extremely difficult to get an exact, believable figure from the computer.
This was proven to me by the fact that you could run two leak rates, back to back, with no power changes or operator-caused changes and the test results would vary.
Although I did not believe the leak rate tests to be accurate, I never felt that we were operating the plant in violation of the technical specifications or in an unsafe condition.
I was generally aware of plant conditions, and if a leak rate was on my desk or handed to me that reflected a drastic change from previous tests, I would try to determine the cause.
On many occasions we would dispatch auxiliary operators to tour the plant and search for leaks.
If leaks were discovered, they were quantified, if possible, and were accounted for in the next leak rate test which was performed.
We constantly monitored other plant parameters such as radiation monitoring instruments in the reactor building, reactor building cooling fans condensate leakage detectors, and the reactor building sump.
If these instruments gave no indication of excessive leakage, I believed that the problem was in the leak rate test program.
. {
b I have become aware over the last six years that some operators have stated that they purposely added hydrogen to the makeup tank during the course of leak rates in-order to affect the results of the test.
Specifically, it has been alleged that the addition of hydrogen during the last few minutes of a test would affect the level transmitter and perhaps enhance the results of the test.
One of the control room operators under my supervision has testified that he conducted an experiment with me to determine whether the addition of hydrogen would in fact affect the level transmitter, and marked the makeup tank strip chart at the time the experiment was conducted.
I have absolutely no recollection of such an event ever occurring, or knowing in 1978 or 1979 that hydrogen additions might affect the level transmitters.
It is possible that I was involved in such an experiment and was aware of the so-called hydrogen phenomenon; however, I really do not recall it.
If I was aware that hydrogen was being added during leak rate tests, I am not certain that I would have stopped the practice since I do not recall any prohibition against hydrogen additions, and may not have believed that such additions would have any effect on the test results.
I have also been informed over the course of many interviews that if water were added to the system during a leak rate test, and properly accounted for in.the computer, that as a result of level instrumentation problems in the makeup tank, the level transmitters would indicate more water than the -
c amount that was added.
I do recall that water was constantly being added to the system, particularly as the date of the accident approached, due to a leak in one of the safety valves.
However, I do not recall any problem with the level indicators that would affect the results of a leak rate test if water were added.
I was on duty on October 18, 1978, when certain events occurred which triggered LER 78-62.
I have been exhaustively interviewed about these events and have very little recall about this particular day.
I remember there was a lot of activity in the control room at that time, and it was unusual because the shift supervisor and the supervisor of operations were actually performing leak rates.
I have reviewed the LER that was generated, and am certain that I reviewed it when it was issued, but I have no specific recollection of it.
I also have no recollection of going into the Action Statement on that date.
I am aware that Edwin Stier, in his assessment of my involvement in leak rate testing, has alleged that I first contended that the plant was in the Action Statement on my shift.
This is absolutely untrue and an unfair assessment.
I told Mr. Stier's investigators repeatedly that although it was possible we were in the Action St).ement, I could never recall it.
The only reason I raised the possibility was because I remembered a shift supervisor and the supervisor of operations actually performing leak rates, which was unusual, and because
-S-
a Q
of documents that I have seen in the course of these investigations which stated that we were in the Action Statement.
I would like to make it clear that I did not know of anyone on my shift or any other shift who deliberately falsified leak rates at Three Mile Island.
The events that are the subject of this investigation occurred almost eight years ago, and because of the time that has passed and the numerous interrogations that I have been subjected to, I believe it is understandable j
that I have very little recollection of these matter.
- However, of one thing I am certain.
During the almost 11 years that I have been employed by Metropolitan Edison Company and GPU
~
Nuclear, I have always attempted to perform my duties in a proper manner with the safety of the plant being paramount.
Mr. Stier's assessment that there is substantial evidence to support a conclusion that I participated in or knowingly tolerated manipulation of leak rate tests is completely unjustified and unfounded.
Including my Navy service, I have been involved in the nuclear industry for almost 20 year and it is my intent and desire to remain in the industry.
It is for l
this reason that I asked to become a party in these proceedings. _ __
a in V
secuso cmm-e-v DOCNgTED UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARD 6 RE 15 P5 :0 1
0FFICE OF SECFj TARY 00CKETir1G A SE?VICE BRANCH
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF JAMES R. FLOYD My name is James R.
Floyd.
I reside in Elizabethtown, Pennsylvania.
I am currently employed by the Harrisburg Steam Works, Ltd. in Harrisburg, Pennsylvania as plant manager.
I have left the nuclear business and I have no present intention of seeking employment in the industry or applying to the NRC for a license.
I hold a Bachelor's Degree in Chemical Engineering from Columbia University.
I was employed by Metropolitan Edison Company or GPU Nuclear Corporation between September 1965 and April 1983.
In 1968, I was assigned to Three Mile Island as a Nuclear Engineer.
In approximately 1971, I became Supervisor of Operations in Unit 1, which position I held until September 1975.
At that time, I became Supervisor of Operations in Unit 2 until after the TMI-2 accident, at which time I became a Senior Engineer.
During 1978 and through the time of the 1979 accident, I was Supervisor of Operations for Three Mile Island Station Unit 2.
My duties included supervising the activities of members of the Operations Department, and scheduling Operations surveillances.
It was also my duty to ensure'that the required number of licensed personnel were manning the control room at all times.
I delegated the actual running of Unit 2 to the shift supervisors.
I provided any support they needed.
I expected them to discuss any problems they had with me, including problems with leak rate procedures.
All leak rate practices at Unit 2 originated in Unit 1.
For example, early on in Unit 1, John Herbein, as the Superintendent, instructed us to run the leak rate test once a l
shift.
We carried this habit over from Unit 1 to Unit 2.
Mr.
Herbein's direction was an administrative control, not a legal requirement.
He instituted this so that if a leak should i
develop, we would be aware of it on an 8-hour basis as opposed to a 72-hour basis.
Similarly, throwing out leak rates was not by direction but something that grew out of Operations and was carried from Unit 1 to Unit 2 by me, the supervisors, and the foremen.
I am now aware that the control room operators believed that I thought l
it permissible to throw invalid leak rates away.
I, however, have no specific recollection of discussing with any CRO l
whether leak rates should be discarded.
It was my opinion that blatantly bad leak rates (for example, excessively negative
\\
2-
r-O%
ones) had no connection with reality; consequently, it was permissible to discard them.
The shift supervisor had the ultimate responsibility for throwing out leak rates.
I never knew that the NRC was unaware that we were discarding leak rates.
I now believe that discarding leak rates was inappropriate, but I did not have that opinion at the time, for the reasons I have stated.
I did not enforce the application of Exceptions and Deficiencies to leak rates.
I believe that this practice developed in Unit 1 because of the frequency with which we were running the procedure.
Because we were running the leak rate more than the technical specifications required, we did not use an "E and D" sheet for a bad leak rate.
The erratic test results that we first obtained in the early period of Unit l's operation could also have influenced our decision to treat leak rate tests differently than other surveillances.
Until October of 1978, it was the general opinion that we had to get one valid leak rate of less than one gallon per minute unidentified leakage into the record every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> to comply with the Unit's technical specifications.
After October 20, 1978, if there w3s a bad leak rate and an operator could not convince himself that it was invalid, steps were to be taken to shut the Unit down.
He would make this decision on the basis of his judgment, not simply on the numbers coming out of the computer.
Resolution of a leak rate higher than 1 gpm was determined by the procedure; for example, was the reactor q
at steady state, were the. power levels changing, did an operator make an error?
Any of these things would invalidate the leak rate test and hence there was no need to enter the action statement.
I issued a memorandum in October of 1978 to explain the change in interpretation of Operations personnel.
My memorandum was issued two days after we started rounding off.
I recall first learning of rounding off from William Fels.
He was re-programming the computer console.
I asked why and he stated that he was doing so on orders of James Seelinger.
I recall that Mr. Seelinger conducted a meeting in the shift supervisors' office in connection with rounding off and the clarification of the action statement.
I do not remember who was there.
I did not agree with the decision to round off and I did not agree with the change in the 72-hour requirement, but I went along with both decisions.
The change in our interpretation of the 72-hour requirement was certainly a safe approach.
I probably learned of the decision to file a Licensee Event Report in connection with this matter when I was handed some leak rates to analyze for Mr. Seelinger.
I thought that the content of my Operations Memorandum was being carried out by the Operations staff.
I was probably lax in not following this up.
Prior to the accident, I had thought that the shift foremen would have less trouble with leak rates if they paid more attention to plant leaks.
I instructed them in my October 1978 memorandum to keep a list of leaks.
I was I
-e L5%
remiss in not pursuing this.
Obviously, I put too much confidence in the list of leaks solving leak rate problems.
Prior to the time that Hartman's allegations were publicized in 1980, I had no reason to believe that the procedure for calculating leak rates at TMI-2 was not capable of producing accurate results.
I assumed that if the test were run properly, the result would be an acceptable measure of unidentified leakage.
Sometime later, I took the time to analyze the procedure in detail and soon concluded that there were so many inaccuracies and errors built into the computer program that-the test results were essentially meaningless.
I know of no basis of asserting, as United States Attorney Queen did in his Statement of Facts, that "the Supervisor of Operations had no faith in the reliability of the test,"
or that I believed in 1978-79 that " leak rate tests being filed for NRC review were
' worthless *."
Those conclusions were reached by me only after the Hartman allegations were made, not during the time that TMI-2 was operational.
In my view, there was not an atmosphere at Unit 2 that would cause or pressure an operator to cheat on a leak rate test.
I know of no instances where operators were directed by management to manipulate results.
I knew the leak rate procedure produced conflicting results.
Because I could recall difficulties with erratic leak rates when Unit 1 began commercial operation, I expected
-S-l L.
T problems at Unit 2.
Apparently others, lacking this background, devised inappropriate methods to cope with leak rate problems.
I did not believe that hydrogen should affect the makeup tank.
If someone would have told me ot the phenomenon prior to the accident, I would have tried to explain that hydrogen should not alter the level of the makeup tank.
I thought it allowable at any time to add hydrogen to the makeup tank if the makeup tank needed hydrogen, whether there was a leak rate in progress or not.
I would not have considered it appropriate to add hydrogen to deliberately tamper with a test, of course.
After the accident I did become aware that water additions to the makeup tank had not been recorded properly.
Two or three cases were brought to my attention by an NRC I &E inspector.
After I questioned the operators, whose names I no longer recall, I was satisfied that these occurrences were unintentional errors resulting from failed communication.
The start and stop time of leak rates should have been recorded in the control room operators' log book.
If those times were not recorded, it was an error of omission on the part of the control room operator.
I also should have been aware of this omission because I was required to review the log book once a week.
Although I was on the distribution list for the morning reports, I did not scrutinize them.
I perceived the morning report as a document used by my superiors to acquaint..
t o-b I
themselves with plant' status on a daily basis.
I encouraged the use of turnovers from shift to shift.
Turnovers were not required, but they were an efficient way to get the job done.
Each level (control room operator, foreman, shift supervisor) had.his own page to prevent things from falling through the cracks.
Shift supervisors' notes were attached to the morning report'because John Herbein had started the practice at Unit 1.
I was only peripherally involved with the November 1, 1978 LER.
I have no recollection of being involved in any discussions that may have occurred between James Seelinger and Donald Haverkamp concerning our interpretation of when to enter the action statement.
I have no recollection of seeing the March 16, 1979 TCN.
I did not sign it.
Two reasons for my lack of knowledge of events in March 1979 were that I had recently suffered a broken ankle, which kept me from work much of the time, and in late March I went to my annual week of training at Babcock & Wilcox in Lynchburg, Virginia.
In conclusion, I decided to appear before the Presiding Board for two reasons.
First, I have information to contribute concerning the events of 1978 and 1979.
Second, I want to put this entire matter behind me.
Consequently, I will cooperate in any feasible way to ensure the final resolution of this subject..
o ED pofiHESPunuq 00CKETED USNRC UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARD OFFICE 0F SKRt1ARY 00CKETlHG A SERVICL BRANCH
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF WALTER J. MARSHALL, JR.
My name is Walter J. Marshall.
I live in Elizabethtown, Pennsylvania.
I am presently employed as an operations engineer at TMI-2.
I maintain a senior reactor operator's license.
I began employment with Metropolitan Edison in February 1977.
I obtained a control room operator's license in September of that year and a senior reactor operator's license in June or July of 1978.
I obtained a license because I was interested in doing so, and because my employer needed licensed personnel available in case of union strikes or other emergencies.
During 1978 and 1979, I was one of two operations engineers assigned to the Supervisor of Operations, James Floyd, We split our responsibilities, which included writing and reviewing procedures, establishing operations surveillance
==
(
o-programs, operating the condensate polishers, the mechanical draft cooling towers and the natural draft cooling towers, plus performing anything else with whic.: Mr. Floyd needed assistance.
I reported to Mr. Floyd at least once a day to l
.tell him about the work I had accomplished and to receive additional assignments.
From September 1978 until sometime in February 1979, I spent about eight hours a day working on the condensate polishers in the basement of the turbine building.
I spent so much time there because we were having difficulty operating the polishers and regenerating the resin beds.
This problem affected TMI-2's ability to run at 100% power.
During this same period, I spent approximately 10% of my time in the control room.
I brought the shift foremen up-to-date on work I had performed on the condensate polishers and I requested personnel, as needed, to help me.
Once a week, for about 15 or 20 minutes, I reviewed the shift foremen's logs for compliance with the company's administrative procedures.
I did not review the logs to discover what evolutions had been performed or if they were conducted in compliance with the procedures.
Occasionally, I stood watch as a shift foreman if someone was sick.
When I did this, a shift supervisor was usually in the control room with me.
I had to become familiar with the requirements of the technical specifications governing the leak rate test in order to pass my license examination.
I did not have to put that -
o k
knowledge into practice.
To the best of my recollection, I never performed a leak rate test and I never approved one.
I am fairly certain that I never saw an operator discard a leak rate test, and I had no idea that operators were discarding tests on a routine basis.
The only problem that I can recall with leak rate tests is one that I associated with the issuance of LER 78-62.
I had come into the control room early in the morning at the end of my shift.
I overheard a discussion about whether the operators should be on a four-hour time clock.
I think that shift supervisor, Joseph Chwastyk, was there and James Floyd may have been present too.
I do not remember anything else about this episode and I do not remember being assigned additional responsibility for the LER.
My understanding is that a computer program error was the reason for the LER's issuance, and that the program error was corrected.
In conclusion, I believe it was apparent to everyone with whom I worked that I was not expected to perform or interpret leak rate tests.
This fact has also become apparent to everyone who has interviewed me on the subject.
I would be exceedingly grateful to the Presiding Board if it would acknowledge that I was not involved with leak rate testing. L
e WMN E
w
-5 00l*ETED UNITED STATES OF AMERICA USNRC NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOAR M E 15 P5 :61 0FFICE OF SECCEA!W DOCKETING \\ SE*V'1 BRAUC'l
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF BERNARD G. SMITH My name is Bernard G. Smith.
I am 50 years old and I live in Hummelstown, Pennsylvania.
I am employed by GPU Nuclear Corporation as Radioactive Waste Operations Manager at Unit 2.
f I no longer hold a license to operate a nuclear power plant.
Prior to January of 1983, I was a shift supervisor for approximately four years, including during the period from September of 1978 through March of 1979.
Before I became a shift supervisor, I was a shift foreman at Unit 1.
I started that position in January of 1968.
I have been an employee of Metropolitan Edison and GPU Nuclear for a total of 27 years.
During 1978 and 1979, I reported to James Floyd who was supervisor of operations.
Kenneth Hoyt, a shift foreman, reported to me in Unit 2.
The control room operators assigned to my shift were Harold Hartman and Raymond Bocher.
A trainee, John Blessing, was also assigned to my Unit 2 shift.
~
O L
The shift supervisors had an office in each unit.
Generally, we started our shifts in Unit 1.
After the first hour, we spent most of our time in whichever unit needed our attention; for example, during Unit 2's hot functional testing, I spent a large part of my time there.
Regardless of where we shift supervisors were actually working, our crews told us about the progress of work at both units.
The shift supervisors had meetings once a month.
At these meetings, we discussed plant problems and personnel problems for both units.
I kept the minutes of our meetings and distributed them to those who were responsible for completing the action items contained in the minutes.
I also attended each unit's plan of-the-day meetings.
These meetings were not designed to discuss problems; they were there so that we could structure the work that needed to be done for that day.
If we had a problem with leak rates on a particular day, we would have a separate meeting in the shift supervisors
- office.
Generally, I did not become involved in leak rate testing.
The operators reported to me only if they could not obtain a leak rate during the shift.
I cannot recall ever having run a leak rate test at either unit.
I have been shown that I approved only two tests at Unit 2.
Usually, of course, leak rate tests were reviewed and approved by the shift foreman.
I have difficulty recalling the exact requirements of the technical specifications concerning the amount of leakage permitted.
I remember that at Unit 2 you needed to show you
w t
were under one gallon per minute in unidentified leakage every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
One valid leak rate during this period satisfied the technical specification; on the other hand, if a valid leak rate over one gallon per minute was obtained, you had to enter the action statement.
I do not recall ever entering the action statement for leak rates at Unit 2.
I did not require my shift to log anything but the completion time of a leak rate and, probably, the result.
I do not recall recognizing any requirement that the start time for leak rates be logged.
I think that the operators used the log book to keep track of leak rates.
I knew that leak rate tests were discarded, and I probably discarded some myself.
This practice began in Unit 1.
There was never an ulterior motive for doing this.
I believed that if a control room operator considered a test invalid, he did not have to document the test.
Exceptions and Deficiencies were not used; possibly for the same reason, or because the leak rate test had different paperwork than the other surveillances, because it was performed on the computer.
This was also a carry-over from the way we operated Unit 1.
I do not think that I even considered whether I should apply administrative procedure 1010 to our leak rate practices.
The interpretation of whether a leak rate test was invalid was left up to the shift foreman, or shift supervisor on occasion.
The control room operator could consult either of them; however, he primarily consulted the shift foreman.
The control room operator could also declare a leak rate test invalid, if he saw a reason to do so from his plant parameters.
If the plant appeared to be in steady state, we would claim a leak rate over one gpm invalid.
This practice started at the beginning of operations in Unit 1.
We did not interpret leak rates any differently in Unit 2 than we did in Unit 1.
If my shift obtained consistently high leak rates, we i
dispatched auxiliary operators to search for leaks.
If leaks were discovered, they would be repaired or measured.
I always tried to fix leakage or to see that it was contained safely inside the plant.
If leakage were identified and measured, it would be added to a new leak rate.
The old one showing results higher than one gpm would be discarded.
I am sure that if I could not have corrected a leakage problem, I would have entered the action statement.
I wish to stress that I thought that I was operating the plant safely even if I was relying on my eyes, rather than the leak rate test printout, to estimate leakage.
This practice was not designed to deceive the NRC or harm the public.
This
(
practice evolved from the way I learned to cope with leakage at Unit 1.
I simply did not comprehend at the time that my concern for the plant's safe operation should not have been given priority over the strict interpretation of the technical specifications that I should have employed, and I would now employ.
1 i
a i
The operators with whom I worked complained to me about the computer problems with the leak rate.
This was no surprise because the computer program produced inconsistent results from day one in Unit 2.
It was evinmon knowledge that peopic were spending many months trying to correct the computer program.
I tolerated invalid leak rate tests because I was confident that the computer problem would eventually be solved.
During the many interviews which I have given, I learned about an NRC inspection that resulted in the issuance of a Licensee Event Report (LER).
I do not think that I was on shift when the incident leading to the report occurred, and I do not recall being part of any decision-making process triggered by this incidant.
I do not mean to suggest that I never discussed the interpretation of the technical specifications with my superiors and my colleagues.
I am sure that we discussed the interpretation of leak rates many times.
I, however, do not remember any formal meetings seeking my opinions on leak rate testing or instructions telling me to abandon my interpretation of the 72-hour rule for leak rate testing.
I have no recollection of reading LER 78-62, although my name is on the sign-off sheet attached to it.
I do not know l
how I interpreted the LER when I did read it.
I do know that I i
never changed the way I did business in response to the LER.
I do not recall reading the Operations Memorandum issued on October 20, 1978, but I do remember that at one point we were l
instructed to round off leak rates by eliminating the decimal <
o l
readings.
I have been told that I directed the rounding off to stop, because there is an entry in a computer log (made by someone else) stating that it ended "by Order of Bernie Smith."
I am sure that I had no role in this, other than to communicate a directive from my superiors that was issued while I was on duty.
I have no knowledge of operators manipulating leak rates by adding hydrogen.
I remember that in 1978 some people believed hydrogen could affect leak rate tests.
I discounted this theory because I thought that adding hydrogen would increase pressure and make the leak rate appear worse.
I would not have prohibited the addition of hydrogen during a leak rate, where the operator knew the system needed hydrogen, because I did not consider it a chemical; therefore, I am sure I did not tell my operators to refrain from adding hydrogen during a leak rate.
I never directed anyone to add hydrogen to deliberately manipulate a leak rate test.
I also have no knowledge of anyone adding water and not recording it or adding water to achieve some type of bonus l
during a leak rate test.
I am sure that the operators occasionally made mistakes and did not inform each other of the addition of water during a leak rate.
That would have j
invalidated the test.
I was unaware of any other way in which an operator could have falsified a leak rate test.
I did not even remember, for example, that there was a level transmitter problem until'th'e post-accident investigations began.
I did I
l l l
t
(
not hear about or learn of anyone discussing the switching of level transmitters as a way to tamper with leak rate tests.
Although Harold Hartman was on my shift, I was totally unaware that he was falsifying leak rate tests.
I still have difficulty believing that he would do such a thing, because I considered him a competent operator.
He constantly brought concerns to me and I would try to correct them depending upon the circumstances.
I cannot remember any safety concern that he brought to me that I did not take some action on.
I am not claiming that he was easy to work with, however.
Even though he had the ability to complete his assignments well, his over-reaction to problems detrimentally affected his performance.
I often had to speak to him about his inability to handle pressure and I know that at least one other shift supervisor became exasperated with him.
I will always regret that I did not realize the effect that the everyday pressures of operation would have on someone such as Mr. Hartman.
I wish Mr. Hartman could have understood that it was my responsibility, not his, if he could not obtain a satisfactory leak rate test.
I decided to participate in this proceeding to clear my name and to remove the impediment which is blocking my progress in the nuclear industry.
Over the years, I have gained a trememdous amount of knowledge about this industry and I have become extremely proficient in the clean-up, removal and shipment of radioactive waste.
My goal is to have these skills 1 !
l
. = __
t s
evaluated for their own worth, without the Hartman allegations marring my reputation.
I hope that the Board helps me to attain.my goal by clearing me of any alleged involvement in leak ~ rate falsification.
i 9
9 i
1
]
I
'I i
4 k
4 l
I t
.j i
e 4.
im.
r
-*== =
--.ws.-w.,-.e-ev-.-
,w...
w,.-
,wy---rr-- -, - - -
-we, e w--w.
--w-,
,.y--4---re==-4 weee,e-v w--,-r, wee---m.,m-g,.
-p.-.--=.s,g y-----
w--
anaiw gw.hynuam J
l DOLKETED l
USNRC UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION 16 AU315 p5:04 BEFORE THE PRESIDING BOARD 0FFICE or sicet ;gg y DOCKF T6Nt: s c r o s o,nr
- M """'
B
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF JOHN A.
BRUMMER My name is John Allen Brummer.
I live in Reading, Pennsylvania.
I am thirty-four years old and am presently employed at Metropolitan Edison Company as an engineer.
I no longer hold a license to operate a nuclear plant, but I may be interested in seeking a license in the future.
Prior to my current position I was start-up and test manager for Oyster Creek Nuclear Power Station.
I was in that position for two years, from August 1,
1981 to August 1,
1983.
Prior to August 1981, I was at Three Mile Island as lead instrumentation engineer.
I was physically at Three Mile Island, on the day shift, since approximately March 18, 1974.
My primary area of responsibility was at Unit 2.
From about January 1979 through March 1979, the time of the accident, I worked with Ivan Porter and primarily reported to George Kunder.
At the same time, I was in the license training
)
J program in training for the NRC SRO license.
I completed my license examination just before the TMI-2 accident, and I obtained my license in July of 1979.
I maintained my license until I left Unit 2 for Oyster Creek.
From 1974 until I left Unit 2 in 1981, I was a Plant Operations Review Committee (PORC) member.
PORC served as the review committee for technical specification procedures as well as operating procedures.
With PORC, I approved and reviewed leak rate test procedures.
I no longer remember whether I performed a leak rate test; if I did, it was not on a routine basis, but rather as part of my training under the supervision of a control room operator.
For my license examination, I was required to know all of the procedures and the leak rate procedure was included among those.
As far as I know, there are no leak rates on file in which I was involved.
Accordingly, I ask the Board to include me on the list of people as to whom there,is no evidence of leak rate test falsification.
em.AniD puttnr.srOf5MENON f
000KETED UNITED STATES OF AMERICA USNRC NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARD OFFICE 0F SECRETAR(
00CKEltNG A SE6VICf.
CIWS
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF ADAM W. MILLER My name is Adam W.
Miller.
I currently reside in Annville, Pennsylvania, and am employed by GPU Nuclear Corpora-tion as Manager of Plant Operations at TMI-2.
I currently hold a senior reactor operator's license.
I began my career with Metropolitan Edison Company as an Auxiliary Operator in Unit 1 in 1973.
I was promoted to Control Room Operator in August 1975, and Shift Foreman at Unit 2 in August 1978.
I was a Shift Foreman until the fall of 1979, when I became an Administrator of Nuclear Technical Training in the Training Department.
In the Spring of 1981 I was transferred back to the Operations Department as a technical analyst in the position of assistant to the Manager of Plant Operations.
I remained in that position until the Summer of 1983, when I was promoted to my present position.
ll 0
i During the latter part of 1978 and 1979, I was Shift Foreman of "D"
shift in Unit 2.
My shift supervisor was Gregory Hitz, and the Control Room Operators under my supervision were Dennis Olson, Lynn Wright and Mark Coleman.
As foreman, I was generally familiar with the activities of the CRO's on shift, although there were some shifts where I was outside the control room part of the time.
As part of our duties, we were required to monitor reactor coolant system leakage.
One of the methods of doing this was to' perform a leak rate test, and the technical specifications required this to be done at least once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
However, to the best of my recollection we ran these tests on a shiftly basis.
Because we were doing this on a more frequent basis than the technical specification requirements, it developed that those tests which exceeded the limit of one gallon per minute in unidentified leakage were not considered to be surveillances.
I do not recall when this understanding started, but I think it existed for years.
To the best of my recollection, this philosophy was not applied to any-surveillances other than leak rates, which were considered more or less a routine thing.
There were other indicators which were utilized besides the leak rate test to determine leakage from the reactor coolant system.
These were the particulate channel instruments for reactor building atmospheric sampling, the leak detector channel in the reactor building air coolers, and the.
1 reactor building sump indicator.
These instruments would not give a number for unidentified leakage, but I believed that they certainly would indicate any change in reactor coolant system leakage within the reactor building.
I formed an opinion when I was a Control Room Operator that the leak rate test was not very accurate.
I have no recollection, however, of ever attempting to ascertain in my own mind what the precise degree of tolerance in the test was.
If a test came out exceeding one gallon per minute in unidentified leakage, it was discarded and another test was performed.
If the second test was under one gallon per minute, it was approved and retained.
I believed at the time that if we had a one gallon per minute leak inside the reactor building, the radiation monitors would have activated.
Therefore, I did not believe that we exceeded the technical specification limit for unidentified leakage.
I have been informed that some Control Room Operators have said that they added hydrogen to the makeup tank during the performance of a leak rate in an effort to influence the test results.
I did not have any knowledge of this practice by anyone on my shift or on any other shift until a year or two after the accident.
In fact, I was not aware of anyone on my shift doing anything in an attempt to manipulate the tests.
I also understand that there have been allegations that some operators deliberately added water to the system during a leak rate in order to influence the tests. -
a
\\
)
Apparently, these additions were logged and accounted for in the test.
I had absolutely no knowledge that this practice was going on, if it was.
To the best of my recollection, I required my operators to attach a Data Sheet 4 to tests when water was added.
I have reviewed several of the tests which I approved where water was added, but no data sheets were attached, and I have no explanation for this.
I believe, however, that I would have been concerned with the fact that no Data Sheet was attached when I reviewed the tests in 1978 and 1979.
There have been allegations by the NRC that some of the operators on my shift added water during leak rate tests and did not account for the additions in the test or log them.
I have reviewed the strip charts for several of these tests, and I believe that most of them do not reflect any water additions.
Mr. Stier apparently agrees with me.
I have also reviewed several strip charts on leak rates performed by my shift when there were apparent problems i
with one of the level transmitters.
My review indicates several tests where the stable level transmitter was switched V
to the computer during the test, which would give a more accurate leak rate.
On other tests, the unstable transmitter went to the computer, and the stable one to the strip chart.
I do not recall knowing that one of the transmitters was out of service at the time, but I am certain I would have known and informed the CRO's at the time if a level transmitter was i
]
9:
I declared out of service. 'According to Mr. Russell of the NRC, there are log entries for periods of time when the transmitters were calibrated, repaired, and returned to service.
Therefore, it is possible that the operators were unawarc that a problem with the transmitter had recurred.
At any rate, when I
. reviewed leak rate tests, I do not recall inspecting the strip chart to-make certain that the proper level transmitter was used.
I would agree in retrospect that perhaps I should have done a better job as shift foreman in reviewing leak rates and attempting to solve the apparent problems we were having with the tests.
However, leak rates were only a small part of the things that had to be done when we came on shift.
Routinely, I would review all of my paperwork at the end of the shift--usually the last half hour.
I would have leak rate tests, shift and dailies, various sets of readings, and other documents to review.
Thus, the review done was necessarily very brief.
I had absolutely no knowledge at the time of anyone taking advantage of the system by adding water, adding hydrogen, or switching level transmitters with the intention of manipulating the leak rates.
Furthermore, I can recall no discussions with anyone, including other shift foreman, of such manipulation.
The leak rate issue has had a significant impact on my attitude towards strict procedural compliance and the need to.
s' cl resolve problems completely to achieve compliance.
The
. performance ' standards.of my company and the industry have risen significantly since the time of.'the TMI-2 accident,-and I
~
believe my.own pertormance has risen with these standards.
.I have always tried to do as good a job as I could.
I would like.
the opportunity to continue to do so in the nuclear industry.
s t
l l.
l L
l- -
g g goggtS M M h
r, DOCKEIED UNITED STATES OF AMERICA-USNRC NUCLEAR REGULATORY COMMISSION If AUS 5 PS:03 BEFORE THE PRESIDING BOARD
$@CK OF SF re_ r.,...
"**ntilHG & ff}Qy{-
BRANcn
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF JOSEPH R. CONGDON My name is Joseph Raymond Congdon.
I currently reside in Elizabethtown, Pennsylvania.
I am currently employed by GPU Nuclear Corporation as a Shift Foreman.
I served in the United States Navy from 1966 to 1973, attaining the rank of Petty Officer First Class.
I began employment with Metropolitan Edison Company in 1974 as an Auxiliary Operator in Unit 1.
I qualified for and received my Reactor Operator's license in 1977, and was assigned as a Control Room Operator (CRO).
During 1978 and through the 1979 accidenc I was a CRO j
assigned to "C"
shift.
The shift supervisor was Brian Mehler, the shift foreman was Charles Adams, and the other control room l
l operators were Martin Cooper and Mark Phillippe.
As a CRO, my general duties consisted of controlling plant parameters and reactor power, operating the controls, overseeing the general operation of the plant, running required l
l l
i
5 i
surveillances, carrying out switching and tagging operations for maintenance and equipment and directing the work of the Auxiliary Operators.
The technical specifications in Unit 2 required that a leak rate surveillance be performed at least once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
As a company policy, however, we attempted to run a surveillance at least once per shift.
There were times when we ran them more than once a shift, maybe twice or even more.
I believe this frequency of performing leak rate tests was carried over from our on-the-job training in Unit 1.
I was aware that if the unidentified leakage exceeded 1 gallon per minute, we were required to enter the action statement, that is, correct the problem within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, or commence shutting the plant down.
I thought this meant that if we did not successfully complete a test within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, that we were required to go into the action statement.
I did not have a lot of faith in the accuracy of the test, considering the size of the system, the inaccuracies built into the instruments, and the problems with maintaining the plant in a stable condition, but it was the only tool we had to calibrate an integrated number of what the overall plant leakage was.
Because I did not have a lot of faith in the test, I tended to read the results against the backdrop of my overall plant experience.
The leak rate test program provided that water could be added to the system during a leak rate test if the operator accounted for it.
However, it was preferable not to add it, so l t
on our shift we tried not to make any water additions during the test.
We used to put signs up on the makeup tank level recorder to alert the other operators that a leak rate was in progress.
I was aware in 1978 and 1979 of the administrative procedure which required the completion of Exception and Deficiency reports (E & D's) on surveillances, and I understand that this procedure applied to the leak rate test.
I do not recall even using this with leak rate tests which I believed to be inaccurate or unreliable, but cannot explain why, without addressing a specific situation.
We were also required to log leak rate tests, and we did so, when they were satisfactory.
We did not log on the beginning and ending times of the tests, and I believe my understanding at the time was that this was not a requirement on surveillances that were run more frequently than once a week.
On my shift, we did not retain leak rate tests where the unidentified leakago was greater than 1 gpm.
Typically, I would discard those tests where there was obviously an error, such as an inadvertent addition of water, or a power change, or some other reason the plant was not stable during the performance of the test.
If there was no apparent reason to believe the test was not as accurate as it could be, I would consult with my shift foreman, and run another test.
It is my recollection that we would retain out-of-spec tests until we -
I got a satisfactory result, then either the shift foreman or I would discard them.
Sometime during 1978 or 1979, I became aware that hydrogen additions sometimes would have an effect on the makeup tank level indication.
It might make it go up or make it go down, and sometimes it had no effect on it.
At one point, we experimented, and marked the strip chart when the addition was made to note the experiment, but I cannot recall whether it had any effect on the result.
I do not recall who ran this test with me, but my best recollection was that it was Charles Adams, my shift foreman.
I cannot be certain that Adams was aware of the hydrogen phenomenon, but I believe he was because we operated very close together, and I had no reason to conceal it from him, since I did not think I was doing anything wrong.
I do not have any recollection of any conversations with Adams or Brian Mehler, my shift supervisor, about hydrogen.
I did not believe at the time that the additi,on of hydrogen during a leak rate was wrong.
I never tried to hide it.
In fact, I believe that I logged those additions every time I made them.
Although we had difficulties in obtaining leak rates, I never felt we were operating the plant in an unsafe condition.
There was plenty of capability built into tne plant to accommodate leakage up to several hundred gallons, and there were other parameters you could use to observe the plant to determine whether it was operating safely.
We were very observant of all these parameters because we did not take our responsibilities lightly. -
p u po w r-4 9' N.
Y ETED UNITED STATES OF AMERICA 00ypgp NUCLEAR REGULATORY COMMISSION j6 NE 15 P5:00 BEFORE THE PRESIDING BOARD 0FF!CE 0r Sgcpi73p <
00CKEItNG A slievir.E BRAffCti In the Matter of
)
s
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF HUGH A. MCGOVERN My name is Hugh Albert McGovern, Jr.
I reside in.
Middletown, Pennsylvania.
I am currently employed by GPU Nuclear Corporation as a Plant Operations Manager for TMI Unit 2.
I served in the United States Navy Nuclear Power Program for six years as a Nuclear Electronics Technician prior to becoming employed by Metropolitan Edison on October 11, 1976 as an Auxiliary Operator A.
I remained as an Auxiliary Operator A until January of 1978, at which time I began a position as a Control Room Operator Trainee at TMI-2.
I became licensed Control Rocm Operator in November or December of a
1978 and remained in that position until October or November of 1980, when I became a Shift Foreman.
I was a Shift Foreman for two years.
In 1982, I became a Shift Supervisor, and served in that capacitt u~ntil August 1984, when I became a Plant
_ NE v
Operations Manager at TMI-2.
This position requires that I maintain a Senior Reactor Operator's license, which I currently do.
In January 1979, I was assigned to the "F"
- Shift, together with Earl Hemmila and Leonard Germer as Control Room Operators (Germer was a CRO Trainee), Carl Guthrie as Shift Foreman and Ken Bryan as Shift Supervisor.
Immediately prior to that time, I was assigned to the "A"
shift with Ed Frederick and Craig Faust as Control Room Operators, Fred Scheimann as Shift Foreman and Bill Zewe as Shift Supervisor.
As a Control Room Operator, my general duties consisted of controlling plant parameters and reactor power, manipulating the controls, running required surveillances, overseeing the general operation of the plant, carrying out switching and tagging applications for maintenance and equipment, and directing the work of the Auxiliary Operators.
I was aware, during 1978-1979, that the one gallon per minute (gpm) limit for unidentified leakage was a limiting condition for plant operation.
I was also aware that certain precautionary measures (that is, reduce the leakage to within limits or begin plant shutdown) were required to be taken if unidentified reactor coolant leakage exceeded one gpm fo.r several hours.
Despite certain inaccuracies in the computer program and instruments used to calculate reactor coolant system leakage, I nevertheless believe that the plant could be operated safely with the information we had.
I was confident. _ __
V that by paying close attention to the makeup tank level and the pressurizer, I could have a reasonably good idea of what was going on in the primary plant.
I was aware, during 1978-1979, of certain TMI-2 rules regarding logging.
All surveillances mandated by the technical specifications (such as leak rate tests) had to be logged.
More specifically, we were required to log the starting time and completion time of any surveillance test.
I did not always comply with that requirement during 1978-1979, however, in large part due to the fact that we were a lot more lax administratively at that time, than we are these days.
I can state, however, that no one ever instructed me to refrain from logging the start and' completion times of a leak rate test so as to avoid revealing how many tests were started.
During 1978-1979, I interpreted the frequency rule regarding leak rate tests to require that only one good leak rate test every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> be obtained.
This was the prevailing interpretation of that requirement when I started work in the Unit 2 control room.
As a new CRO, I did not question this practice.
I do not recall seeing Licensee Event Report (LER) 78-62 at the time it was issued.
It was not until after the accident that I became aware of the lesson of that LER, namely, that we had been misinterpreting the 72-hour rule.
I know now, but did not know prior to the accident, that regardless of how often one runs leak rate tests, the TMI-2 operating rules, -.
s g'
require that.the action statement be entered whenever unidentified leakage in excess of one gpm is found.
I am also more aware now that important safety considerations underlie the 1 gpm limit for unidentified leakage.
I was familiar with the requirement that exceptions and deficiencies be filed with unsatisfactory surveillance tests.
I do not recall ever filing an exception or deficiency with an unsatisfactory leak rate test, although I now realize that I should have done so.
While I knew that exceptions and deficiencies were applicable to technical specification surveillances, I did not regard the inputting I did to the computer as performing a technical specification surveillance.
Rather, I viewed the computer input work I did in connection with leak rate tests as merely fulfilling an administrative
" paper" requirement.
The standard frequency that leak rate tests were run at TMI-2 was two or three times per shift, although this could vary depending upon what had occurred during the previous shift.
Namely, if the previous shift had obtained a satisfactory leak rate test result, we might conduct only one leak rate test on the next shift, or perhaps none.
During 1978-1979, I was aware of several inaccuracies in the instrumentation used to conduct leak rate tests.
One such inaccuracy was in the computer program used to calculate leak rates.
Specifically, that computer program yielded inaccurate results because it did not make the proper._ _ -
1 v-temperature compensation for leakage to the drain tank.
This inaccuracy was a key reason that I did not believe that the leak rate test was a valid indicator of actual leakage.
The actual parameters indicated by makeup tank level and pressurizer level often conflicted with the results of the computer leak rate test.
During 1978 and early 1979, I was not aware of the phenomenon by which water added to the makeup tank during a
leak rate test could enhance the result.
Nor was I aware of the phenomenon by which-the amount of a water addition as registered on the strip chart exceeded the actual amount of water added and logged by the operator.
It was not until after the accident that I was informed that adding water to the makeup. tank during a leak rate test could produce an enhanced result.
The reason for this unawareness was that the existence of this phenomenon regarding water additions was inconsistent with what I had learned as a reactor operator in the Navy.
While I was a control room operator at TMI-2, I never made either unrecorded or under-recorded water additions to the makeup tank during a leak rate test for the purpose of manipulating the test.
Nor did I ever make a " jogged" water I
l addition to the makeup tank to influence the leak rate test result.
I have no personal knowledge that any other operator manipulated leak rate tests through water additions during that i
time period.
I can also state that to the best of my l
f knowledge, Earl Hemmila never made a " jogged" water addition to the makeup tank in an attempt to alter a leak rate test. --.
O During 1978 and early 1979, my shift's approach with respect to water additions was basically as follows.
We would add water routinely to the makeup tank before leak rate tests to better hold the plant steady while the test was being conducted.
We also added water to the makeup tank whenever it approached its low limit, which sometimes occurred during a leak rate test.
We were aware, however, that water additions during leak rate tests should be avoided.
In any event, my shift was aware of the requirement that all water additions to the makeup tank be logged and accounted for, and we complied with that requirement.
Neither I, nor to my knowledge the other CRO's that I worked with, were aware that hydrogen additions to the makeup tank could affect a leak rate test.
I personally never added hydrogen to the makeup tank so as to alter a leak rate test.
Nor did I ever observe anyone else add hydrogen to the makeup tank during a leak rate test for the purpose of influencing the test result.
As far as I knew at the time, the reason one added hydrogen to the makeup tank was to scavenge oxygen (which is a corrosion agent) from the reactor coolant system.
It was always my practice to log hydrogen additions to the makeup tank.
I was generally aware of the fact that at various times prior to the accident, there were problems with both makeup tank level transmitters.
I cannot specifically recall seeing an out-of-service tag placed on the level transmitter selector switch prior to the accident. _ _
O As a general rule, I tried to switch the accurate level transmitter to the computer when running a leak rate test, provided that the other-level transmitter (which would give us our level indication in the Control Room) was not so inaccurate that we could not rely on it.
'I was able to determine which level transmitter was functioning properly by examining the strip chart level recorder.
Specifically, the good level transmitter was the one that exhibited the fewest random oscillations.
There was a time during 1978 and 1979 when we ran some leak rate tests knowing that there were inaccuracies with the makeup tank level instrumentation.
We performed those tests under such circumstances because we were compelled by the 72-hour administrative requirement to run leak rate tests.
We recognized that the equipment was malfunctioning, but it was the only instrumentation available to us at the time.
In addition, there were people working on the problem, so it seemed to me that the problem was only temporary.
I remember a period of time when one makeup tank level indicator was being worked on virtually daily.
I can state unequivocally that I never deliberately switched level transmitters during a leak rate test so as to obtain a satisfactory result.
There would be nothing to be gained by doing that.
The deliberate use of an inaccurate level transmitter could just as easily lead to an unsatisfactory leak rate test result as a satisfactory one.. _.
r i
9 As a general rule, leak rate test results for.which unidentified leakage exceeded one gpm were discarded, basically because it was a common practice in the control room to do so.
I was never informed of a requirement to retain the results of such unsatisfactory tests.
We also discarded leak rate test results more negative than 0.1 gpm.
I personally remember either discarding leak rate test results exceeding 1 gpm, or referring such tests to my shift foreman.
I also recall-discarding more leak rate tests than those I kept.
However, I do not recall that Carl Guthrie ever told me to discard leak rate tests exceeding 1 gpm.
I do recall an incident in which the NRC found some unsatisfactory lea < rate test results that had not been discarded and brought this to the shift foreman's attention.
However, no one ever instructed me to discard unsatisfactory leak rate tests so that the NRC would not see them.
Our standard procedure for measuring identified leakage entailed the shift foreman from time to time sending an auxiliary operator with a graduated cylinder out in the plant to measure leakage.
To the best of my recollection, we made a consistent practice of quantifying identified leakage in tnis way, and of accounting for it in the unidentified leak rate calculation.
I never felt significant pressure from my shift foreman or shift supervisors to obtain a satisfactory leak rate test result.
Nor do I' recall receiving any pressure from those -
4 i
in the managerial levels above shift supervisor to obtain a satisfactory leak rate test result.
Occasionally, someone in the control room would say something such as "run another one, we're approaching the clock", but I don't remember any arm twisting or anything like that.
I do not recall it becoming more difficult to obtain an acceptable leak rate test result as the day'of the accident approached.
However, there was a lot of. leakage from the pressurizer valves just before the accident.
As I stated before, however, we did not regard the leak rate tests as an accurate indicator of actual leakage.
We viewed the obligation to perform leak rate tests as a perfunctory administrative requirement.
Given that we took little stock in those tests, why would we go to extremes to manipulate them?
My job is important to me.
I believe that my experience with leak rate tests and legal proceedings about them and about the TMI-2 accident has taught me to be much more scrupulous about procedures.
As plant Operations Manager at TMI-2, I am very interested in helping the Company complete the cleanup that is well underway.
I believe that I was, and am, a safe operator and that this experience has taught me to be a better operator, and I sincerely hope that the Board will exonerate me from involvement in leak rate test falsification because I did not manipulate leak rate tests.
m g
M D
UNITED STATES OF AMERICA 00CMETED USNRC NUCLEAR REGULATORY COMMISSION j6 AU 15 PS:UO BEFORE THE PRESIDING BOAR OFFICE OF SECEETARY 00CKETIMi s s!Mir::
BUUC:t
)
In the. Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF EDWARD R.
FREDERICK My name is Edward R. Frederick.
I live in Middletown, Pennsylvania.
I am employed at Three Mile Island Unit 2, in the training department.
I am an Instructor V in the corporate training division.
I hold a senior reactor operator's license.
From March of 1968 until November of 1973 I served in the United States Navy as an electrical operator on nuclear submarines.
I started with Metropolitan Edison on November 28, 1973 as an auxiliary operator.
I became a control room operator on October 6, 1975 and obtained my control room operator's license for Unit 2 on October 19, 1977.
I transferred to the training department on July 1, 1979.
At Unit 2, I was assigned to "A"
shift with Craig Faust as the other operator.
We reported to shift foreman Frederick i
Scheimann, and shift supervisor William Zewe.
We usually dealt directly with Mr. Scheimann.
Although Mr. Zewe was in charge
of the shift, he was less visible because he had to divide his time between Units 1 and 2.
I could contact him at any time, of course, if I needed to do so.
From the number of reports on leak rate tests at TMI-2 that are before the Board, it may conclude that the performance of leak rate tests was the major responsibility of the operators.
That, however, was not true; leak rate tests were a small part of our responsibilities.
There were thirty or forty pages of other surveillances that had to be run on a shiftly basis.
We had to take hourly readings whether required by technical specifications or by the company to ensure that we-were maintaining machinery performance.
We had to report test results (including those brought in by other departments) to the shift foreman.
We had to conduct radiation monitor checks and, when needed, check source verifications on our instruments.
We also had to ensure that we were performing all the operations needed to run the plant safely and efficiently.
The technical specifications required that unidentified leakage be checked once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
If unidentified leakage exceeded one gallon per minute, we were required to reduce the leakage to within limits, within several hours, or proceed to shut the plant down.
Before I started a leak rate test, which was the way we measured the amount of unidentified leakage, I would check power level and generator output.
I did this to verify whether the plant was at steady state.
If, by monitoring these conditions, I noticed that there was a degree t
of instability in the plant, I would not run a leak rate test.
If I started a leak rate and detected a change in temperature or power, I would stop the leak rate and start over again.
I have been shown one or two tests where the plant was unstable and I handed in a leak rate; nevertheless, I was conscious of this and did try to watch for steady state conditions.
Once I obtained a leak rate test, it was my practice to visually observe the plant instruments in order to compare their status with the leak rate test result.
While I found my methods to be generally effective in allowing me to determine unidentified system leakage, I cannot state that I regarded the leak rate test procedure as 100%
accurate.
I felt that it was necessary for me to take precautions in setting up the test in order to improve the test's accuracy.
Because I took precautions before I started a leak rate test, I did not find them to be a problem.
I recall that Craig Faust and I had totally different styles for running a leak rate test.
To the best of my recollection, Mr. Faust ran the test as soon as he came on shift, and he assumed responsibility for the surveillance mote often than I did.
I believe he got into the habit of starting a leak rate right away so that he could begin the shiftly readings and start the paper work.
I tended to wait until I had been on shift for a period of time and I had checked out the plant.
We did not assist each other in performing the leak rate; however, we would review them together if needed.
While -
a we did not perform the test as a team, we did try to inform each other if.a test had been started so that neither of us would change an essential plant parameter.
It seemed to me that the leak rate procedure, including the computer program and the hand calculations that had to be added to the program, were constantly changing during 1978 and 1979.
I attributed this to efforts to make the procedure more accurate.
I recall, for example, that the reading for the reactor coolant drain tank was changed so that it would feed directly into the program rather than rely on manual entry.
Also, we briefly tried to round off numbers, and a calculation was made to account for the change in temperature from where the reactor coolant system went to the drain tank.
I do not believe that the leak rates I saw were highly variable in nature, if I had been careful in establishing the right conditions for the test.
I have no recollection of running the test repeatedly in the course of one shift, unless I had made a mistake.
I also have no recollection of Craig Faust running the tests over and over again, or of the tests becoming much more difficult to obtain in February and March of 1979.
About the only problem I can recall from these last two months was that the instability caused by the manual recirculation of the pressurizer would sometimes make me refrain from attempting a leak rate test.
The foreman rarely specified the time he wanted the leak rate test run as he might do for the less routine i
e s
A surveillances.
I do not recall that leak rates were run on every shift, around the clock.
For whatever reason, I remember most frequently running the leak rate tests at night, particularly during the eleven to seven shift.
I have been shown they were perfortaed less f requently on the swing shif t from three to eleven.
I do not remember running leak rates on day shift, and according to William T.
Russell's own analysis, none of the tests signed by me was run on the day shift.
As I remember, there were several ways that we could find out what the last leak rate had been and whether a test had been performed.
One was to look on the status board for the most recent identified leakage number.
That would have the date and time with it, because that was the last time that someone had looked for identified leakage.
The other way was to look in the surveillance outgoing basket to see if a leak rate had been performed.
This was kept on the shift foreman's desk.
We could also learn about leak rate tests during shift turnover.
I did not log leak rate tests, because I considered them a routine surveillance.
Other surveillances that we did on a rcutine basis were not logged either; for example, we did not log the hourly readings required by the technical specifications.
None of the many auditors reviewing my log (including the shift supervisor, supervisor of operations, quality assurance or the NRC) ever brought to my attention the fact that I should be logging leak rate tests.
i !
i
<a 4
I placed all leak rate tests that I completed on the shift foreman's desk.
I gave him all of the results regardless of the amount of unidentified leakage shown, because it was his job to interpret technical specifications.
I do not remember the trend of the results that I gave to him, although I do remember not having much difficulty obtaining results less than 1 gpm.
I cannot remember entering the action statement, but I do remember conducting searches to identify leakage.
When I discussed the tests with the shift foreman, I would try to give
~
him my opinion as to whether the test was accurate or inaccurate.
If he was not in the room, I did not like to leave it in his basket, but if I had to do so, I tried to write a note about the test so that I could discuss it with him later.
I am unaware if my shift foreman discarded the leak rate results.
I presume that he did, because of what I have learned through these investigations.
I do not remember giving him a leak rate test which I thought required entry into the action statement; however, I do remember giving him some that were inaccurate.
It is these that I theorize he discarded.
I do not remember even thinking whether Craig Faust was throwing j
away leak rate tests in 1978 or 1979.
I do not know if Craig followed the same practice I did; that is, handing all leak i
rate tests to the shift foreman.
I have absolutely no memory of whether or not I was aware prior to the accident that it was a common practice to discard leak rates.
i -
i I do not recall whether I filed an Exception and Deficiency sheet with any leak rate test, although I think I understood that the procedure applied to leak rate tests.
I know I filled out many E and D sheets while I was on duty at Unit 2; however, I cannot recall doing so for any particular leak rate test.
For the records that I have seen that I submitted, there was not need to complete and E and D sheet because none of the results required it.
We should have used E and D's if the computer was out of service, or if one of the inputs was out of service.
I learned after the accident that hydrogen had a suspected effect on leak rate tests.
I cannot recall observing hydrogen cause a change in makeup tank level during operation.
I do not recall discussing the effect of hydrgoen on leak rate tests with anyone before tha accident.
I was also unaware that the addition of water could cause a level change greater than what should have been expected by properly accounting for its addition.
As a general rule, I tried not to change anything or add anything during a leak rate test.
I have no recollection of a problem with either or both of the makeup tank level transmitters.
I do not know whether I declared either one out of service.
It is hard to recall particulars about these instruments because I have declared hundreds of pieces of equipment out of service.
I would not have used a level transmitter marked out of service; however, it is possible that I might not have realized one was more i
l l i
6 erratic than the other on any given day.
I never deliberately switched these instruments to falsify a leak rate test and~I have never heard of anyone else doing this.
I cannot contribute anything to what is already known about the issuance of a Licensee Event Report and an operations memorandum in 1978.
I do not remember this time frame.
I have learned that I was apparently on shift when a leak rate with over 1 gpm unidentified leakage was obtained on October 16, 1978 and that this test was mentioned, specifically, in the LER.
I do not remember being told about the October 16th test during this period and I do not remeber being questioned about why my shift did not enter the action statement.
I also do not recall receiving any lectures to stop leaving invalid leak rates in plain view in the control room.
I decided to participate in these hearings no that I could continue to maintain a license, and seek exoneration to prevent a continuance of the career stagnation I have experienced.as a result of these outstanding issues.
I became more anxious to clear my name after reading William T. Russell's opinion of my credibility.
He suspected the accuracy of my answers on several points:
the ease of obtaining satisfactory leak rate test results, the discarding of unsatisfactory test results, the frequency of running leak rate tests, and my interpretation of-the requirements of the technical specifications.
Mr.
Russell was particularly annoyed that my testimony differed, on occasion, from that of Craig Faust.
I would like to respond to his accusations. -
4 t
Mr. Russell strongly suggested that I was dishonest because I'did not recall running leak rate tests on every shift.
To the contrary, it would be dishonest for me not to adhere to my own best recollection that I primarily ran leak rate tests during the evening, on the eleven to seven-shift and, less frequently, on the three to eleven shift.
Mr. Russell's records show that I signed eleven tests from September to March.
Nine out of the eleven were performed in the evening after 6:00 p.m.
Only two tests were performed after 3:00 p.m.
and before 6:00 p.m.,
and as I have stated previously, none was performed on the day shift.
My memory also accords with Harold Hartman's July 26, 1983 statement to the NRC that the "mid-shift was the accepted time to run leak rates."
I also drew Mr. Russell's suspicion because I did not recall having a problem running leak rate tests and I did not recall discarding invalid ones.
I have repeatedly testified that I tried to properly set optimum conditions for a leak rate test before I initiated it.
The evidence shows that I obtained relatively few invalid leak rates and also that I obtained-more valid ones than many of my colleagues who allegedly. resorted to
" falsification".
Perhaps I do not recall anyone discarding leak rates because I actually gave few inaccurate tests to my foreman.
Finally, Mr. Russell faults me for stating that I knew I had to enter the action statement if I obtained a valid leak rate over 1 gpm.
He states that the majority of other.
a 9-
)
operators from my shift and Craig Faust' thought that they could ignore a leak rate over 1 gpm as long as they obtained one under 1 gpm within the 72-hour period.
While I do not wish to dispute the testimony of the other operators, it is my opinion that Mr. Russell erroneously interpreted Craig Faust's opinion of the requirements of the technical specifications.
Mr. Faust testified that he submitted leak rate tests which he could not invalidate to the shift foreman, Frederick Scheimann, so that he could determine whether the technical specifications had been violated.
Mr. Faust presumed that Mr. Scheimann had a reason for not entering the action statement for those tests; therefore, it does not appear that Mr. Faust disregarded leak rates he could not invalidate, as long as he was within the 72-hour period.
It is my belief that Mr. Russell's assessment of my credibility is both strained and unfair.
In my opinion, he deliberately found something negative to say about me, because I firmly maintained that I did not falsify leak rates or delve into the leak rate falsification discussions that he feels were so rampant at Unit 2.
I hope that the Board discounts Mr.
Russell's conclusions about this subject and finds, as he had to, that I was a good operator with a good record.
Should the Board exonerate me, I can assure the Commission that I will conscientiously comply with the requirements of my license..-.
c' 4b
[.
@TED CORRENMC(
v DOCKETED UNITED STATES OF AMERICA USNRC NUCLEAR REGULATORY COMMISSION j6 NE 15 P5:d0 BEFORE THE PRESIDING-BOARD 0FFICE OF SECRETAH f 00CKETING 1 s! vv:r:C DRA!!CH In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF GEORGE A. KUNDER My name is George A.
Kunder.
I reside in Middletown, Pennsylvania.
I am currently employed by GPU Nuclear Corpora-tion as Manager of the Unit 2 Safety Review Group, a position I have held since 1982.
Prior to assuming my present position, I was employed by Metropolitan Edison, with whom I began employment in 1968.
I became TMI-1 Superintendent-Technical Support in the latter part of 1977.
I was transferred to the similar position in Unit 2 in December 1978.
As TMI-2 Superintendent-Technical Support, I was responsible for supervision of the technical aspects of plant engineering.
I had the additional duty of coordinating the planning for refueling outages, and I was the Chairman.of the Unit 2 Plant Operations Review Committee, known as PORC.
I was not licensed on Unit 2, and I had.no direct role in the operation of the Unit and no responsibility for operational
e 0
i
~D decisions.
My relationship to the Operations Department was advisory.
My contact with operations personnel was limited to occasional meetings, such as those conducted by PORC, and intermittent conversations when specific engineering problems were brought to my attention.
In fact, usually I learned about operational problems through people on my staff who performed the engineering tasks assigned to us.
During the time that I was assigned to Unit 2, I never ran a leak rate test, I never signed a leak rate test, and I never reviewed a leak rate test.
I have no recollection of ever being present in the control room while an operator was running a test.
It is obvious, I think, that I had no role in or responsibility for performing or recording leak rates.
That was the job of operations personnel.
As a technical engineering group, my department did not even review leak rates.
We also had no part to play in developing or implementing leak rate practices at the time.
I am also not aware of any role in the preparation of the leak rate surveillance procedure itself.
My only memory of leak rate testing in Unit 2 is that somewhere in the 1979 time frame, I became aware that there was some question whether leak rate test results accurately reflected plant conditions.
The concern, I believe, was that the calculation used to determine leakage was in error, and that erroneously high readings were being obtained.
At one point my department was requested to _
y take a look at the leak rate procedure or calculations to determine if a problem actually existed, and if so, to resolve it.
The Plant Operations Review Committee, of which I became chairman when I came to Unit 2, was responsible for investigating instances of noncompliance with technical specifications that were brought to its attention.
By virtue of the many inquiries that preceded this one, I am aware that in late 1978 or early 1979, PORC reviewed a Licensee Event Report (LER 78-62) that involved a question of interpreting the Unit 2 leak rate technical specifications.
I should note at the outset that I do not have any memory whatsoever of this LER that I can relate to events or discussions that might have taken place in 1978 and 1979.
I was not employed in Unit 2 at the time the LER was issued, and I.have no memory of PORC's disposition of it after I assumed the chairmanship of the Committee.
All I can do to assist the Board in its assessment of the place of the LER in this proceeding is to provide an overview of PORC's role in Unit 2 and its relationship to other departments and management.
PORC was responsible for review of various procedures, tests, and system changes and modifications in Unit 2.
Its specific charges were set forth in Technical Specification 6.5.1.6.
The list is fairly lengthy.
One of PORC's charges investigating violations of technical specifications, was including preparing and forwarding reports covering evaluations D
and recommendations to prevent recurrence to the Station Superintendent and the Generation Review Committee.
As I recall, PORC usually discharged this responsibility by assigning to managerial personnel in the appropriate department the task of following up an LER.
LER 78-62 was typical in that it required that instruction be given to affected personnel and documentation be provided to show that they had received or reviewed the instruction.
While PORC kept records of assignments and close-outs on corrective actions for LER's, it was not required independently to confirm the completion of assignments.
I think it is important to point out that PORC's role was advisory in nature.
PORC was an interdisciplinary committee composed of persons holding certain positions established by the technical specifications.
PORC had no authority either to implement or to enforce changes.
As its name implies, it was a review committee that in matters concerning LER's coordinated the dissemination of action items.
Once PORC chose the appropriate individual to follow through, it expected that individual to fulfill his responsibility.
PORC usually requested that some kind of documentation be provided for our records, so that it could demonstrate, to the extent practicable, that the action item had been completed.
The sign-off sheet on LER 78-62 was typical of one kind of documentation, but in the instance of a repair, or the performance of a mechanical or technical procedure, a notation
-4
'e that the repair orLprocedure had been completed might suffice.
Documentation provided by.the managerial or supervisory personnel was considered a reasonable demonstration that the required action had been completed.
PORC normally did not undertake any-independent steps to verify the implementation of action items, although it would sometimes make recommendations for additional action or follow-up if it found that the documented actions were insufficient.
That being said, I should also point out that in a situation where I understood that the plant was operating in violation of technical specifications, I would have felt it incumbent on me to immediately advise the Unit Superintendent of the problem.
Of course, the nature of my job was such that, almost of necessity, the violation would have to have been brought to my attention by someone else.
Mr. Stier, in his report, has found no evidence that I was involved in, or condoned, improper leak rate practices.
He has also concluded that I was involved in January 1979 in a decision to continue operating TMI-2 in violation of the technical specifications.
Because I believe the latter conclusion is outside the scope of these hearings, I have not addressed it in this statement, even though I disagree with Mr. Stier.
I would like to take this opportunity to clear up an inaccurate conclusion drawn by the NRC staff.
This inaccuracy, which is of part'icular concern to me, appears in NUREG-0680, 1
i g-Supp. No.
5, the NRC staff report on management integrity prepared as part of the restart proceedings for TMI-1.
In that report, the staff noted that it had reviewed the Statement of Facts prepared by the prosecuting attorney in the 1983 criminal action against Metropolitan Edison and accepted the Statement "as fact."
The report proceeded to name me, as Superintendent of Technical Support for Unit 2, as a person " involved or implicated in the indictment" of Metropolitan Edison.
This identification of me is repeated later in Section 13.2.2 and Table 13.2 of the report.
I believe that this conclusion is based on a factual misunderstanding, and this proceeding provides an appropriate moment to put the error to rest.
First, I know I was not involved nor implicated in leak rate falsification nor the indictment.
Further, I reviewed the Statement of Facts myself and found that it contains nothing to suggest that I was " involved or implicated" in indictable offenses.
Section I of the Statement of Facts merely discusses the NRC staff's view, at that time, of the regulatory and operational requirements of Unit 2.
Section II is a review of Unit 2 operations from licensure until October 18, 1978, and Section III discusses specific events on October 18, 1978 all well in advance of the December 1, 1978 date on which I assumed any responsibility for Unit 2.
Thus, while there are repeated references to "the Superintendent of Technical Support" in those sections they could not possibly be references to me.
Section IV of the Statement of Facts, titled W
" Events After October 18, 1978," contains only one reference to a specific event (which occurred on October 19, 1978) -- again, many weeks before I moved to Unit 2.
The remainder of Section IV discusses the perceived failings of certain operators and supervisory operational personnel, none of which could be attributable to me in my role as a provider of technical support, because I was in Unit 1 as Superintendent of Technical Support, not in Unit 2.
For these reasons, I believe that the staff's identification of me, as opposed to my predecessor in the position of Unit 2 Superintendent of Technical Support, was an unintended error.
I attempted to explain this to the
-Commission in a letter dated August 13, 1984, a copy of which is attached to this testimony.
The Commission responded in the form of a letter from NRR, which asserted that because the Metropolitan Edison indictment covered the period October 1978 through March 1979, and I was part of the Unit 2 " management structure" during part of that period, it was not incorrect to state that I was " involved or implicated in the indictment."
I believe that assertion is completely unwarranted by the Statement of Facts, and I respectfully request that the Board l
include express recognition of the Staff's mistake in the findings adopted at the conclusion of this proceeding. - _. _
s-O O
Q.
~
GEORGE A. KUNDER 1906 Light Avenue Middletown, PA 17057 August 13, 1984 i
Hon. Nunzio J. Palladino Chairman U. S. Nuclear Regulation Commission Washington, D.C.
20555 4
Dear Chairman Palladino:
I am writing to you to protest the reference to me contained in NUREG-0680, Supp. No. 5, issued by the Office of Nuclear Reactor Regulation in July, 1984.
In 55.2.1, the staff refers to the prosecuting attorney's Statement of Facts, and in 55.2.2, the staff states that it " accepts, as fact, the. prosecuting attorney's Statement of Facts read into the record as part of the trial settlement, that five Met-Ed management' personnel were involved or implicated in the indictment."
The staff goes on to identify me as one of those five individuals.
That identification is repeated in S13.2.2 and in Table 13.2.
Staff's identification of me, which has no basis in fact, is extremely damaging to my professional and personal reputation.
I have been provided with a copy of the Statement of Facts referred to by the staff.
My name is not included in the document.
Apparently, the staff relied upon the identi-fication of certain position titles in the Statement of Facts to conclude that I was involved.
That conclusion is unwarranted and incorrect.
I I assumed the position of Superindendent of Technical Support at TMI-2 on December 1, 1978.
Prior to that time, I was not assigned to TMI-2, nor was I involved in its operation.
The incidents referred to in the Statement of Facts involving the Superintendent of Technical Support took place prior to December 1, 1978.
There is no basis for the staff's assumption that I was involved.
l
n e
d Hon. Nunzio J.
Palladino August 13, 1984 Page 2 I would also like to point out that at no time prior to March 28, 1979, was I licensed to operate TMI-2.
I had no involvement with the performance of leak rate tests at TMI-2. I am not aware of any statement by any person suggesting that I was associated with any questionable practices concerning leak rate tests at TMI-2.
On September 28, 1983, I was interviewed by the Commission's Office of Investigations concerning my knowledge of leak rate tests performed at TMI-2 in 1978 and 1979.
I have not been furnished with an official transcript or summary of that interview, but my attorney took extensive notes.
I reviewed those notes and found nothing in them to support the staff's finding that I was " involved or implicated" in the indictment of Met-Ed.
In summary, I believe that the staff's accusation of me in NUREG-0680, Supp. No. 5 is completely improper.
In order to protect my reputation and preserve my credentials in the nuclear industry, I ask that the Commission take whatever action is necessary to correct the record publicly and to withdraw the staff's findings concerning me.
Very truly yours, George A.
Kunder GAK/bjk cc:
Commissioner Roberts Commissioner Asselstine Commissioner Bernthal Commissioner Zech Mr. Harold R.
Denton Guy H. Cunningham, Esquire Mr.
P.
R.
Clark
o WTED COW L5fC 00CKETE0 USNRC UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSi@f AUS 15 P5:00 BEFORE THE PRESIDING BOARb~
OFFICE 0F SECHEUK(
00CKETING A MHV!CI.
BRANCH
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF CRAIG C.
FAUST My name is Craig C.
Faust.
I live in Middletown, Pennsylvania.
I am a senior reactor operator at Three Mile Island Unit 2.
I am also employed in the training department as an Instructor IV.
I was in the United States Navy from November of 1966 until December of 1973.
I started at Three Mile Island as an auxiliary operator on December 5, 1973.
I obtained a control room operator's license in Unit 2, effective October 20, 1977.
I remained in this position until I transferred to the training department in March of 1983.
I sLbsequently received my senior reactor operator's license.
My shiftmates at Unit 2 were Edward Frederick (control room operator), Frederick Scheimann (shift foreman) and William Zewe (shift supervisor).
Hugh McGovern was a trainee on our shift until he received his license and a new shift assignment.
Through this period, we were designated as "A"
shift.
i
My shift was organized in the same way most other shifts were.
One control room operator had the panel, so he ran the plant and kept the log; the other took readings, performed surveillances, and did switching and tagging.
The foreman directed our work, and was the shift's liason with higher supervisory and management levels.
As I understood the requirements of the leak rate test, no more than seventy-two hours were to elapse without a leak rate of less than 1 gpm.
The seventy-two hour clock started running again when we got a leak rate that met the 1 gpm standard.
If I obtained a leak rate over 1 gpm that I could invalidate, I discarded it.
It was my impression that it was a management decision to discard bad leak rates; however, I do not remember any specific directives.
We performed the test on a shiftly basis, because that was the company requirement.
I do not remember having trouble getting leak rates when we first went critical in 1978.
I recall that leak rates became progressively harder to obtain.
I associate this starting with the period around Christmas of 1978.
I might have performed two leak rate tests per shift as things became harder, and progressed to three per shift.
I am sure that I complained to my shiftmates about leak rates, but I did not lodge a formal protest.
To resolve the problem, I continued to try to identify leakage.
I recall attributing my difficulties with the tests to the leakage we were experiencinq from the code safety valves.
I I do not recall analyzing or questioning whether the leak rate test was accurate.
From the fact that people were working to improve it, I guess I knew it was not perfect.
I regarded the leak rate test as a device used to satisfy the technical specifications.
As a general rule, once I successfully performed the leak rate, I was satisfied with it.
As far as the technique of running a leak rate, I just made an entry to call up the computer program.
There was no hard and fast rule concerning who did the leak rate, other than that the switching and tagging operator probably performed it more often.
During the day shift, we usually did not run leak rates, because of the many other activities that had to be performed on the computer.
There was no set time on my shift for performing a leak rate, although we did try to tell each other if we were starting one to prevent additions being made to the makeup tank during the leak rate.
Closer to the accident, it became harder to control the addition of water and I remember discarding leak rates because a water addition had not been entered into the computer.
During this period, it was hard to find a time to run a leak rate without having to add water or to cycle the pressurizer.
If I could determine why a leak rate was bad, I could throw it away without obtaining approval.
If I could not explain why a leak rate was bad, I gave it to the shift foreman.
I presumed the foreman justified why it was bad, because my shift never entered the action statement, and I observed the foreman 2y
-looking at the makeup tank, and at other instruments in the plant.
He also sent us out on leak searches.
If I found leakage, I did not go back and enter it on the bad leak rate, because it was already out of the computer.
I would put another leak rate test in.
The first one would be discarded.
I did not log the start time for leak rate surveillances, or the fact of their completion in the log book.
I remember thinking that only non-routine things went into the log book.
The leak rate document we handed in was right there if anyone wanted to see it, and it contained the start and finish time.
No one ever told me to keep leak rates out of the log book.
I also did not file Exception and Deficiency sheets with tests I had decided were invalid.
I real.ize now it would have been a good idea, but I cannot recall even considering it at the time.
I do recall running shifts and being unable to obtain a valid leak rate.
As part of my turnover, I would tell the relieving shift that I did not get a leak rate.
If my shift were the relief, we would also be told if the outgoing shift had not obtained a leak rate.
If I obtained a small negative leak rate, slightly less than zero gpm, I would accept it.
I did so on the basis that it was acceptable within the tolerances of our instrumentation.
I thought that water additions, temperature changes, and plant swings all could cause negative leak rates.
-4
a
\\
I do not recall having a problem calculating the amount of water I was required to add to the system and record in the log.
1 recall there were three ways to figure out how much water had been added.
One of the ways was looking at the strip chart, the other one was looking at the batch controller, and the third was through communication with the other control room operator.
I most frequently figured out the amount of water added by observing the makeup tank level indicator.
I normally used the bleed tank for water additions but I also used t)-
demineralized water supply because it gave the quickest response for getting the rods back in.
I never made an unrecorded water addition to manipulate a leak rate test.
4 During operation, I never knew that the addition of water, even if properly recorded, could improve the chances of obtaining a leak rate under 1 gpm.
The way I made required hydrogen additions was to ask the auxiliary operator to do it during his rounds and to call me when he added it.
These additions could easily have been made' during a leak rate test, because the auxiliary operator would not have known that the test had been started in the control room.
I would not have thought it wrong to file a leak rate test with a hydrogen addition, if hydrogen had been needed by the system.
I did hear about the effect of hydrogen on the makeup tank prior to the accident, although I cannot say it was common knowledge.
I never heard anyone say they were deliberately -
I adding hydrogen to get good leak rates.
I remember a discussion, although I do not remember with whom, where I said you just cannot do that -- use hydrogen to affect the leak rate
-- because of the way the makeup tank instrument is set up.
I figured that hydrogen could only disturb the system momentarily.
I never tried to see what would happen if I added hydrogen during a leak rate.
I never felt pressure to get a good leak rate.
I simply told the next shift that I had not gotten one.
Not being able to get one for 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> only meant we would eventually have to shut the plant down.
We had shut down enough in the past, so that this prospect did not frighten me.
I remember that one of the instruments measuring makeup tank level was more erratic than the other one, although I did not know why its trace-out fluctuated more than the other one's did.
I suppose that if the more erratic one was not marked out of service, it could easily have been used during leak rate tests.
I certainly did not use it to manipulate a leak rate test.
I cannot actually remember the November 1, 1978 Licensee Event Report although I am aware of it because of later investigations.
I do not remember being part of the company's inquiry into the incident or contributing facts leading to the LER's preparation.
I do not recall receiving any instructions from management in connection with this document.
As I understand the LER now, it meant that the Action Statement had
)
to be entered if a leak rate test in excess of 1 gpm was obtained and was a test that we could not invalidate.
I thought I had this interpretation the entire time I was at Unit 2.
I also do not recall the October 20, 1978 Operations Memorandum.
I do recall going to a rounded-off whole digit for a time and I remember that it did not last very long.
I also recall the change that was made in late March of 1979 to permit us to account for the temperature change in the reactor coolant drain tank.
I accepted this as an attempt to make the leak rate test better.
In conclusion, I wish to emphasize that I never manipulated a leak rate test.
I have been shown that several tests I filed were invalid, and I attribute these to my lack of in-depth analysis.
What seems apparent now was not as obvious to me while I was running the plant.
I wish to assure all concerned that I always tried to perform the leak rate test properly.
This does not mean that I analyzed its accuracy; however, it does mean that I tried to ensure that I complied with the procedural requirements for performing the test.
I did my best during this period and I think the plant records and numerous investigations into my conduct prove this to be true.
I do not think that anyone can doubt my commitment to my responsibilities as a licensee, after what I have been through in the last.seven years.
I also trust that no one doubts my dedication to guaranteeing the safe operation of any nuclear power plant that I would be permitted to operate.
[
ge.u W
~
00LKETED UNITED STATES OF AMERICA USNRC NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARda6 RE 15 P5 50 0FFICE OF SKRt.IARY 00CKtilNG & Sl8VICL BRAi!CH
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF MARK S. COLEMAN My name is Mark S. Coleman.
I live in Middletown, Pennsylvania.
I am currently employed by GPU Nuclear Corporation as a training instructor at TMI-2.
I served in the United State Navy for six years as a machinist's mate.
I began my employment with Metropolitan Edison Company as an auxiliary operator in Unit 1 in January 1974.
In 1976, I was assigned as a control room operator in Unit 2, and subsequently received my reactor operator's license.
I held this position until August 1983, when I assumed my present position.
During 1978 and through the 1979 accident I was assigned to I
"D" shift.
The shift. supervisor was Gregory Hitz, ti.e shift foreman was Adam Miller, and the other control room operators assigned to this shift were Dennis Olson and Lynn Wright.
e
~k There were usually two or three control room operators on each shift.
Generally, the work was divided up between the operators, with one operator assigned to the panel controlling the plant and the'other operator or operators assigned to switching and tagging duties and taking readings.
The operator assigned to the panel controlled the reactivity and power level and ran the more complex surveillances.
He was also the operator who made entries into the CRO log.
Usually, the panel operator would be thrs one to add water to the system to maintain plant conditions and make hydrogen additions to the system, but it was possible that any one of the operators could perform these functions without the others knowing about it.
Leak rate tests were usually performed by the operator who was assigned to do the readings or the switching and tagging.
We usually ran leak rate tests on every shift, and it was the practice to run these tects again if we obtained one that reflected unidentified leakage of more than one gallon per minute.
I do not recall anyone giving me directions to run leak rate tests on a shiftly basis; it was just accepted procedure.
It was my practice to discard leak rate tests where the unidentified leakage exceeded one gallon per minute, and then to run another test.
.I do not recall exactly when this began, but that was the way I did it when I started as a control room operator.
I recall that one time I ran a leak rate test and the results were'in excess of one gallon per minute.
I signed b* f the test and put it on my shift foreman's desk.
A short time later, three people came out of the shift supervisor's office and I was told by one of them, I don't remember who, that they did not want to see leak rates that exceeded the technical specifications.
After that incident, I began throwing away those leak rates that showed unidentified leakage over one gallon per minute.
During the numerous investigations regarding leak rates at TMI-2 since the accident, I never attempted to conceal the fact that I made hydrogen and water additions several time during the performance of leak rate tests.
When I was making these additions, I never thought I was falsifying leak rate tests.
I felt at that time that I was operating within the technical specifications and procedures, which did not specifically prohibit the addition of hydrogen or water, if properly accounted for during a test.
I always accounted for the additions I made.
We continually had problems with the computer and it was extremely difficult to get 3 test result that came out less than one gallon per minute in unidentified leakage.
I knew that work was underway on the program to correct the error, and so I thought the problem we faced was just temporary.
When I made additions to the system, I was only trying to do my job the best way I knew how.
When I was first interviewed by the NRC in April 1980, I informed the investigators that on some occasions I added hydrogen to the makeup tank during the performance of a leak :
g I
rate test in order to get a good result.
I first found out about this phenomenon when a control room operator from another shift, I believe it was Harold Hartman, told me about it.
I experimented myself and determined that sometimes if you added hydrogen, usually toward the end of the test, it could affect the makeup tank level indicator.
Initially, it was possible to make a hydrogen addition from the control room, and any of the CRO's could do this.
At some point, however, the hydrogen valve in the control room became inoperable and it was necessary to call an auxiliary operator and request hydrogen additions manually from a remote station outside of the control room.
I believed that a hydrogen addition during the performance of a leak rate test had to be made at the end of the test if it were to have any effect on the level transmitter, and it became extremely difficult to control the timing of these additions when one had to call an auxiliary operator to make them.
There came a time during the operation of TMI-2, when I became aware that the water additions sometimes had the sam 2 effect on the level transmitter as did hydrogen additions.
If water were added toward the end of a test, for a short period of time the level indicator would reflect a higher level in the makeup tank.
I do not know whether I discovered this myself or whether a CRO from another shift told me about it.
I have reviewed several makeup tank strip charts which indicate that for some periods of time one of the level
-4
s transmitters was not operating properly.
I have no recollection of any problems with the level transmitters back in 1978 and 1979.
However, any instrumentation can go bad and require service, so it would not be unusual for a level transmitter to go out of service.
I understand that some have alleged that if a level transmitter was not operating properly, an operator could switch it to the computer during the performance of a leak rate test and perhaps improve his chances of obtaining a satisfactory leak rate.
I have never done this, and have no recollection of knowing of anyone else who did it.
It was always my belief that you would want an instrument that was operating properly on the computer when performing a leak rate test.
Because of fluctuations in the plant instrumentation and the operation of the computer, it was possible to get a test that showed negative unidentified leakage, provided the negative number was small, I considered such results acceptable and in compliance with the technical specifications.
I was not aware of any of my fellow CRO's on my shift or any other shift, who were aware of the water or hydrogen phenomenon or who were cheating on tests, other than Harold Hartman, who discussed hydrogen additions with me.
I do vaguely recall bringing up the subject of hydrogen with Dennis Olson on one occasion, but he walked away from me without discussing it.
I do not know if my immediate supervisor, Adam Miller, was aware of how I was performing leak rate tests.
I t.--
do not believe, however, that he knew about it because I think he would have talked to me and reported it to Gregory Hitz, the shift supervisor.
I realize today that what I did was probably wrong.
I would like to stay in the nuclear industry and have the option of reacquiring an operator's license from the Commission.
I feel that I have learned many lessons during the TMI-2 leak rate investigations.
In my present position as training instructor at TMI-2, I refer to my leak rate experience in order to impress the trainees with the importance of strictly adhering to the requirements of the technical specifications and procedures in operating a nuclear power plant, and to urge them to raise questions if there is something they do not fully understand, t _
fNRC UNITED b S OF AMERICA NUCLEAR REGULATORY COMMISSION j6 AUG 15 h5:60 BEFORE THE PRESIDING BOARD 0FFICE OF SECRiiARY 00CKETING ', 0FPVmE Entet
)
In the Matter of
)
)
INQUIRV INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF LEONARD P. GERMER My name is Leonard Paul Germer.
I reside in Bath, Maine.
I am currently employed by Maine Yankee as an auxiliary operator.
I s'erved in the United States Navy for a period of six years.
After my service in the Navy, I was employed by a textile firm, Martin & Belvish, for five to six years.
After my employment at Martin & Belvish, I was employed by Electric Bolt for a year.
I then worked at a pharmaceutical veterinary supply company for approximately a year.
I began employment-with Metropolitan Edison Company in 1977 as an Auxiliary Operator at Unit 2.
I began training for the Control Room Operator ("CRO") position in October or November of 1978 and continued that training until the time of the accident.
I left the employment of Metropolitan Edison Company in August of 1979, and immediately began my work at Maine Yankee.
I
r When I started my training, I was assigned to the "E"
shift.
Bernie Smith was the shift supervisor for that shift, Ken Bryan.Jas the shift foreman, and the other-control room operators were Ray Bocher and Harold Hartman.
Sometime around January 1, 1979, I switched to the "F"
shift.
Ken Bryan became my shift supervisor, Carl Guthrie was the foreman, Earl Hemmila and Hugh McGovern were the control room operators.
As a CRO trainee, I was permitted to perform leak rate tests and plant evolutions, but only under the supervision of a control room operator.
I also studied the materials given to me by the training department.
As I was taught by the Control Room Operators, the technical specifications required that only one acceptable leak rate test be obtained every seventy-two hours.
A " bad" leak rate test was one that showed unidentified leakage greater than one gallon.
I was aware of the technical specification that required that the plant be in a steady state condition when leak rate tests were run, and which stated that water should not be added during a leak rate test.
However, I also was aware that water additions during a test could be accounted for by entering them into the computer.
In cases in which there was a plant transient or an inadvertent water addition, I would discard the leak rate test as unacceptable.
I also discarded leak rate tests that were run properly, but yielded an unacceptable result.
I had two basic reasons fo'r doing so.
First, it was my understanding l'
that the relevant technical specification required that only one acceptable leak rate test be obtained every seventy-two hours.
Second, it was common knowledge among the CRO's that I was working for that it was permissible to throw away unacceptable leak rate tests.
I do not recall having a problem obtaining leak rates, and I do not remember that it became harder to obtain leak rates prior to the accident.
At the NRC investigative interview on March 28, 1985, my recollection was refreshed to enable me to recall the events leading up to the filing of an LER with the NRC.
With that assistance, I recalled that an NRC inspector named Don Havercamp found a leak rate test that was inconsistent with the Unit 2 technical specifications, and that this discovery culminated in the filing of the LER.
Otherwise, I have no independent recollection of that incident.
I did not perform leak rate tests in a different manner in response to the filing of the LER than before the LER.
I also do not recall any training session or particular instruction that was prompted by the filing of the LER.
I do not recall too much about logging practices at Unit 2, As a trainee, I do not think I would have been very involved with logging.
I do remember that the operators logged completed leak rates to keep track of the 72-hour clock.
I do not recall anyone ever advising me to refrain from logging unacceptable leak rate tests.
I do not recall any instances in which unrecorded or under-recorded water additions were made to a
the makeup tank during a leak rate test.
Further, I am not aware of any instances in which either Hal Hartman, Ray Booher, Earl Hemmila, or Hugh McGovern manipulated leak rate tests in any manner.
During the NRC investigative interview on March 28, 1985, I was informed of several leak rate tests I had been involved with that the NRC deemed to be questionable for various reasons.
My testimony was then, and is now, that I have no knowledge that those tests were intentionally manipulated by anyone.
During the time I was employed by Metropolitan Edison Company, I was not aware of the phenomenon by which hydrogen added to the makeup tank could improve leak rate test results.
Shortly before the NRC investigative interview, this phenomenon was explained to me based on the technical analyses that were done and that are before the Board, although I am still not convinced that it exists.
As far as I know, the reason one adds hydrogen to the makeup tank is to maintain the desired water chemistry in the reaction coolant system.
I was not cognizant, during my employment at Unit 2, that the addition of hydrogen during a leak rate rest could be deemed a prohibited chemical addition pursuant to the governing rules of Unit 2.
I have no personal recollection of instances in which Unit 2 operators made hydrogen additions to manipulate leak rate tests.
I have no personal knowledge of a level transmitter having exhibited instability during my employment at Unit 2.
I never changed level transmitters during a leak rate test for F}
N
-the purpose of manipulating the leak rate test.
Nor have I heard of any instances of in which others manipulated leak rate tests in that fashion.
I basically had faith in what the RCS computer surveillance program told me about actual plant leakage.
No one ever put any pressure.on me to obtain an acceptable leak rate test result.
In conclusion, if any leak rate tests that I was involved in are " questionable", they are not " questionable" because of any deliberate action by me.
I l
S I
O L
W~
UNITED STATES OF AMERICA DOCKETED USNRC NUCLEAR REGULATORY COMMISSION
]
E 15 PS:d0 BEFORE THE PRESIDING BOARD 0FFICE OF SECRETAid 00CMETING A SFFCCE BRANCH
)
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP UNIT 2 LEAK RATE DATA
)
FALSIFICATION
)
)
PREPARED STATEMENT OF GREGORY R. HITZ My name is Gregory R.
Hitz.
I reside in Elizabethtown, Pennsylvania.
I have been employed by General Physics Corporation since April 1982.
I was in the Air Force from 1964 through 1969.
I began employment with Metropolitan Edison at the Crawford Station in 1969.
In October of 1969, I moved to TMI Unit 1.
I was an auxiliary operator until early 1970.
I became a TMI-l control room operator in 1974 and a shift foreman in 1975.
I became a shift supervisor in 1977 and was, of necessity, licensed on both Units.
I left Three Mile Island in 1981 and was employed by Instrument and Engineering Services Corporation until I went with General Physics.
While I was a control room operator at Unit 1, I performed leak rate tests as part of my responsibilities.
I do not remember attending a specific training class on the leak rate
t 1
x technical specifications, although I may have.
I also do not remember receiving instructions on what to do with an invalid leak rate, when I studied for my Unit 2 license.
When I became a shift supervisor, part of my training included spending time with the other shift supervisors in Unit 2.
I watched how they did business, I took notes and asked questions about how they would handle different situations.
They gave me a feel for how to run a shift.
After I received my Unit 2 license, I was-put in charge of "D"
shift.
My crew included Adam Miller, foreman, and Mark Coleman, Lynn Wright, and Dennis Olson, operators.
I believe my assignment to shift supervisor occurred while Unit 2 was in its hot functional testing phase.
I spent 95% of my on-the-job time at Unit 2 during this period supervising hot functional testing, low power physics testing, and then the start-up of Unit 2.
I answered to Jim Floyd, the supervisor of operations.
We used the chain of command very well.
Mr. Floyd dealt with me and let me deal with my shift.
Mr. Floyd relied on the shift supervisors to perform the paperwork part of the job.
His primary concern was the safety of the plant.
My duties included operating both units, coordinating maintenance, health physics, chemistry operations and security.
With the other shift supervisors, I attended plan of the day meetings as well as monthly meetings.
The supervisor of operations fo'r each unit, each unit superintendent and the,.. _ _ - - _.,__
I i
site superintendent would attend these monthly meetings.
We talked about personnel problems and we tried to get things running more smoothly.
As part of my other duties I had to see that surveillances, including leak rate tests, were performed, and that the plant was operating within the specified limits for leakage.
I had the ultimate decision as to whether a leak rate test was valid when a test was questioned and brought to my attention.
In Unit 2, we needed to obtain one valid leak rate test every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
If we could not determine why a leak rate test was invalid, we had to enter the action statement.
My shift would decide whether a test was valid or invalid by looking at plant parameters.
Some of the parameters evaluated included makeup tank level, makeup tank valve position, pressurizer level, and reactor coolant drain tank level.
The operators would also examine whether they had made any unrecorded additions, or other errors.
If we obtained a later valid leak rate, we would-discard the invalid leak rate.
I do not know how the discarding of invalid leak rates developed originally.
Any licensed operator could discard an invalid leak rate.
No one had to ask for my or the foreman's permission.
Perhaps this practice started because there were so many variables present in the leak rate test that were not present in other surveillances.
By variables I mean plant swings, temperatures, levels and pressures. >
We tried to perform a leak rate test once a shift.
I never really analyzed why we performed it more than the procedure required.
My shift foreman generally approved the leak rate, but if he were out in the plant, I would review and sign it.
I probably would have noted that fact in my shift turnover notes.
I did observe that the Unit 2 leak rate test results were quite erratic, as compared to the results of leak rate tests peformed at Unit 1.
Unit 2 was not yet a stable plant; tuning on the heater drain system, the condensate system and tuning within the ICS had not yet been completed.
The plant swings on the Unit 2 side were more frequent.
This all affected the leak rate results; you had no idea where on an oscillation the computer was going to take a picture.
I remember that we had many negative numbers, as well as positive numbers.
There was no consistency.
Although we knew that we were having problems with leak rates, I do not recall specifically discussing leak rates with other shift supervisors in either plan of the day meetings or our regular monthly meetings.
I did, of course, discuss leak rate problems with my shift.
'I thought that the problem would solve itself once the plant tuning was completed.
I have no knowledge of any operator willfully cheating on leak rate tests.
I never pressured anyone to get a leak rate; I do not believe that anyone else in management did, either.
The NRC Staff contends that start times of leak rates should have been recorded in the control room operators' log.
1-No one omitted logging start times' as a way of hiding the number of bad leak rate tests.
I suppose it developed because we handed in the computer sheet as the official record of the leak rate, and the start and stop time was right on that sheet.
I do not know why Exceptions and Deficiencies were not applied to leak rates.
I do not think it ever occurred to me to apply them.
I know that the computer had to be told about water additions made during a leak rate.
I recall, on occasion, when I was an operator in Unit 1 that I forgot to inform my shift that I had a leak rate going.
I had run the leak rate again because the other operator had added water, due to my mistake.
I cannot imagine anyone doing this deliberately and handing in the leak rate.
I attribute unaccounted-for water additions to honest mistakes.
It was the better practice not to do anything to disturb the system or to add anything during a leak rate test.
If water had been added and accounted for, I would have assumed that the system required the addition.
I was unaware that anyone would have deliberately added water knowing that it would affect the makeup tank level.
I knew that when the hydrogen valve was opened in Unit 1 and hydrogen was put in the makeup tank you could see a change.
I do not remember if it went up or down or how it did it, but you would see the movement in the level indication.
I suppose the same thing could have occurred at Unit 2, though I
- l
,----y-m
,y
r do not remember watching that gauge.
I never made a correlation between this and leak rates.
I do not know of anyone who deliberately added hydrogen to affect a leak rate.
I did permit my shift to file negative leak rates.
I attributed the negative leak rates to plant swings.
If the final leak rate data were taken on a low side of a swing, this affected leak rate calculations.
I recall that my shift had a range of acceptable negative leak rates.
If we had a negative
.0 or.05 or 1.0, I would accept that but not a negative 9.0 or 10.0. I threw the high negatives away.
I do not remember there having been an issue about the interpretation of the technical specifications for leak rates.
My interpretation remained the same.
It is obvious that I saw the November 1, 1978 Licensee Event Report at some point, because my signature is on it.
I cannot remember it, and I cannot remember any other related documents or conversations.
In 1978-79, I did not believe that I did anything wrong with respect to leak rate teste.
Today, I would not throw them away, or let others throw them away, but it was not clear to me then that any NRC requirement was violated by doing so.
During the time that I was an operator at TMI, I sincerely tried to do my best.
I was a good operator, but I am human and I made mistakes; however, I believed I learned something from each of them.
I want more than anything for the entire nuclear industry to benefit from the exhaustive examination of our conduct and the lessons we have learned.
Personally, I have L i
r
\\
absorbed the need for-strict compliance with all the rules governing plant operations, and I better understand the need for thorough discussion of problem areas prior to taking actions.
However, I hope those who will pass judgment on us will factor in the differences between what is acceptable in 1986 and what was acceptable in 1979, because there is.a huge learning curve in these years.
I believe there are many more lessons learned and if I am permitted to remain a part of this industry, I will emphasize these to all with whom I come in contact.
I want others to learn from our mistakes and what better way to drive the point home than to hear it from someone who has suffered the consequences of his errors in judgment.
l l
l f
l l
/