ML20209G267
| ML20209G267 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 09/11/1986 |
| From: | Cooper M SOUTHERN CALIFORNIA EDISON CO. |
| To: | |
| Shared Package | |
| ML20209G260 | List: |
| References | |
| LRP, NUDOCS 8609120410 | |
| Download: ML20209G267 (10) | |
Text
1 5
o l
k UNITED STATES OF AMERICA 00CMETED NUCLEAR REGULATORY COMMISSION USNRC BEFORE THE PRESIDING BOARD 36 SP 11 R2:19 0FFfCE 0r SECetiAPy 00ChEIno freca.f~
)
b u. t. >
In the Matter of
)
)
INQUIRY INTO THREE MILE ISLAND
)
Docket No. LRP i
UNIT 2 LEAK RATE DATA
)
l FALSIFICATION
)
)
PREPARED STATEMENT OF MARTIN V. COOPER My name is Martin V. Cooper.
I reside in Oceanside, California.
I am employed by Southern California Edison Company at its San Onof re Nuclear Station Units 2 and 3, as a Shift Superintendent.
I became employed at San Onofre in 1982.
San Onofre is located near San Clemente, California.
I am a high school graduate and attended St. Francis College in Brooklyn, New York for one and one-half years (1967-69).
In June 1969, I joined the Nuclear Navy and after completing my training was assigned to the USS Ethan Allen.
I served in the Navy for six years.
While in the Navy, I was a reactor operator, electronic technician.
Upon leaving the Navy, I was employed by Stone &
Webster Engineering Corporation in New York City as an engineering aide.
Subsequently, I became employed by 8609120410 860911 PDR ADOCK 05000320 PDR
O Metropolitan Edison Company (Met Ed) at Three Mile Island Nuclear Generating Station Unit 2.
My first position was as an auxiliary operator (AO); I was promoted to control room operator (CRO) in 1977 and obtained my reactor operator's license in 1978.
In 1980, I got my senior reactor operator's license and became a shift foreman and, later, shift supervisor.
I left the employ of Met Ed in 1982.
I chose not to become a party to this proceeding because I am fully occupied in my present employment and did not wish to become re-involved in Three Mile Island matters.
I have testified on Three Mile Island leak rate matters in the past and I did not feel that another look at that issue would be productive, as far as I was concerned.
I felt that these matters were historical in nature and not relevant to the more recent performance of my duties at San Onofre.
During 1977-79, I was assigned to "C" shift.
My shift supervisor was Marshall Beards while I was in training for licensing and after receiving my license my shift supervisor was Brian Mehler; my shift foreman was Charles Adams; Adam Miller and Joseph Congdon trained me, and after I licensed, Joseph Congdon and I trained Mark Phillippe while Adam Miller was promoted to shift foreman.
Auxiliary Operators were also assigned to our shift; the AO's answered to the CRO's.
During the day shift, the relief crew was usually available as well, and might assist the duty shift in carrying out its h
assignments.
Occasionally, the personnel on my shift would vary if someone was sick.
Also, as I recall, Brian Mehler was assigned to Unit 1 during some portions of the Unit 1 refueling outage prior to the TMI-2 accident.
During the time period relevant to this proceeding, that is, 1978-79, my duties as a CRO included conducting the actual operation of the plant and performing switching and tagging operations and necessary surveillance testing.
I understand that the scope of this proceeding is to determine the facts about reactor coolant system leak rate testing practices at TMI-2.
On my shift, leak rate tests were performed soon after we began our shift.
The test was performed on the computer at the command of the operator.
Normally, one control room operator had overall charge of operations and kept the CRO log.
Another CRO performed switching and tagging operations and performed surveillances.
If we had a third CRO, he might have performed the surveillance operations.
The leak rate test could have been done by any of the CRO's or the shift foreman.
The leak rate test was performed by punching "RCSL" into the computer, specifying the duration (1 to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />), and allowing the computer to conduct the test.
Normal practice was to conduct the test over a one-hour period.
Sometimes tests were cancelled without generating a printout from the computer if plant changes or inadvertent operator-caused changes i
k occurred while the test was in progress.
Our shift typically performed leak rate tests early in the shift because we tried to obtain a successful leak rate test every shift and we had a problem getting tests to come out within technical specificati'on limits.
Our practice, if a test exceeded technical specification limits, was to run another test.
If the second test's result was within limits, I recall that our shift would throw away the first test and keep the second.
Our shift threw away leak rate tests with negative as well as positive unidentified leakage depicted on the computer printout.
Our rationale for keeping the second test, and discarding the first (in the above example), was that plant parameters, as indicated on strip charts in the control room, including pressurizer level, makeup tank level, and other instrumentation, indicated that the plant was stable and not leaking.
Typically, I could see that the trend since the last shift was such that I did not believe that there was significant unidentified leakage.
Also, we did not have a lot of faith in the leak rate calculation because the results were erratic.
When the results varied so much from test to test, even though the plant was stable, it was hard to believe the results.
Therefore, it was common practice to discard " bad" leak rate tests, that is, those above 1 spm in unidentified leakage, whose results we did not believe.
s I recall that my interpretation of applicable technical specification requirements was that, after a " good" leak rate test result was obtained, we had 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> to get another " good" result.
Looking back on that practice, now, I would say that we were complying with the letter, but not the spirit, of the technical specification.
Because we sometimes had a problem getting " good" leak rate test results at TMI-2, my recollection was that there was a lot of concern to get such results.
The concern (as I viewed it) would come from supervisors asking whether we got a " good" leak rate.
I assumed that the reason that they asked is because we had trouble getting acceptable results.
I am aware that Mr. Russell of the NRC's Office of Nuclear Reactor Regulation has accused Edward Frederick of lying because Mr. Frederick testified that he never performed leak rate tests on the day shift.
I remember that once I relieved Ed at the conclusion of the day shift (7-3) and asked him whether he "got a leak rate" and he said that "we don't do them on day shift" (or words to that effect).
I remember being annoyed because I thought that everyone tried to run a leak rate test every shift.
I think that incident, though, proves that Mr. Frederick was telling Mr. Russell the truth when he testified that he did not recall doing leak rate tests on the day shift.
I remember that there were several changes made in the _
computer program at different times to improve the leak rate calculations.
The personnel in charge of the computer told us that changes were being made and that we might see a difference in the results.
However, the operators were not informed of what was changed; we just were expected to keep performing the tests.
I remember thinking that the tests were almost meaningless because we got such inconsistent results.
I am told that certain events occurred in October 1978 that led to a Licensee Event Report about leak rate tests at TMI-2.
However, I recall that I continued the interpretation of the leak rate test procedure that we were required to get a
" good" leak rate test result at least once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
I also recall that, for a time, we rounded off unidentified leakage below 1.5 gpm so that the test was deemed to be within technical specification limits.
I remember that most, if not all, of the control room operators balked at rounding off test results, but that was the prevailing interpretation at the time.
I do not recall whether Mr. Haverkamp, the NRC Resident Inspector, approved the rounding-off, but Mr. Capra told me in my September 28, 1984 NRR/OI interview (at page 70) that Mr. Haverkamp had done so.
It is not surprising that I would not know whether Mr. Haverkamp approved the rounding-off of unidentified leakage: he would have dealt with more senior personnel on that type of issue.
I was aware, prior to the TMI-2 accident, that the l.__
addition of hydrogen during a leak rate test sometimes affected makeup tank level, but I never added hydrogen to affect the leak rate test results.
If I added hydrogen during a leak rate test, it was because it was needed, not to alter the test results.
Not only was the addition of hydrogen not prohibited during a test, but it was necessary to keep oxygen from entering the system and possibly causing corrosion.
I remember that when Joe Logan became the TMI-2 Unit Superintendent, the one thing that he always checked when he came into the Control Room was whether we had enough hydrogen in the reactor coolant system.
Therefore, our shift always tried to keep up the hydrogen level because of Mr. Logan's concern about hydrogen.
To the best of my recollection, I never added water to the reactor coolant system during a leak rate test without accounting for it in the computer.
As I understood it, we were permitted to add water during a leak rate test if we accounted for it.
I am aware that I have been accused by Mr. Russell of adding hydrogen (but not water) to affect leak rate test results.
I testified under oath before Mr. Russell that I did not do that, and I repeat that here.
I resent Mr. Russell concluding otherwise.
I do not believe that there is any evidence that I did so.
My shift logged hydrogen additions; if we had been trying to hide anything, we would not have done so.
My testimony on these subjects has been consistent -. - - _.
throughout, and I do not believe Mr. Russell has any evidence to the contrary.
I do remember having the vague feeling that Hal Hartman may have done something to affect leak rate test j
results because of the manner in which he would tell us that he got a " good" leak rate test.
I did not, and do not, have any proof of that.
I am aware that the NRC is accusing me of performing leak rate tests with a faulty level transmitter, which would make the test results invalid.
I did not try to manipulate leak rate test results by using a faulty level transmitter.
I do recall that we had continuing problems with one or both transmitters exhibiting oscillations and also not agreeing with each other.
As I recall, one or both transmitters were awaiting or undergoing maintenance the majority of the time prior to the accident.
Leak rate tests were not the only things that CRO's had to do; we had many surveillance tests to run.
Many times, our shift had to stop a leak rate test because an operator performed a plant evolution that would invalidate the test.
In fact, there came a time on "C" shift when we taped a 3x5 card to the panel showing makeup tank level to inform the other operators that a leak rate test was in progress so as to avoid altering the plant conditions during the test.
Joe Congdon was sometimes absent-minded and the card helped him and the rest of..
a the shift to avoid inadvertently doing things that would invalidate the test.
It is possible that leak rate tests on our shift may have been turned in when operators caused changes to the plant that should have invalidated the tests, but if that happened I believe that it was inadvertent.
During much of the time that TMI-2 was operational, hydrogen had to be added by an AO from outside the Control Room.
Therefore, it is difficult to fairly accuse my shift of adding hydrogen to affect a leak rate test during that period because the operators could not control the timing of hydrogen additions.
Based on our observations, hydrogen sometimes (but not always) had an effect on makeup tank level, but the effect was not consistent.
In fact, I told Mr. Russell and Mr.
Christopher that I never gave an Ao instructions to add hydrogen to affect a leak rate test result, and that is true.
I worked with Joe Congdon and I do not believe that he ever added hydrogen or water to falsify a leak rate test.
I know Joe well, and I know his family, and he would not have done that.
Chuck Adams was also a stickler for detail, and he would never have permitted us to " cut corners", or told us that a little subterfuge was okay.
Now that I have become involved in this proceeding, I would be grateful to the Presiding Board if it would exonerate me of any involvement in leak rate test falsification, because I did not engage in any such activities.
I also do not believe.
+y
- en e
w-r egyg
-- -= +
e
O that anyone on my shift did so.
I take my responsibilities at San Onofre very seriously, and the operators who are responsible to me have been trained by me to strictly adhere to all procedures, especially leak rate test procedures.
I believe that my experience with the TMI-2 accident and the TMI-2 leak rate investigations has made me a far better operator and supervisor, and I hope the Presiding Board will make that finding about me.
I sincerely wish to remain a licensed operator and supervisor in the nuclear industry.
I also sincerely appreciate the opportunity through this testimony to tell my side of the story, after years of simply answering investigators' questions. _ _ - _ -.
,