ML20237G517
| ML20237G517 | |
| Person / Time | |
|---|---|
| Site: | Three Mile Island  |
| Issue date: | 09/08/1986 |
| From: | Stier E GENERAL PUBLIC UTILITIES CORP., STIER, E.H. |
| To: | |
| References | |
| LRP-A-001A, LRP-A-1A, NUDOCS 8708140108 | |
| Download: ML20237G517 (276) | |
Text
{{#Wiki_filter:_,_
--"~7r7me--.r--
- ~ !
n w; w nr,n wm [l il HL g n j\ lI u TMI-2 h I REACTOR COOLANT INVENTORY BALANCE TESTING NilCLEAR REGULAT0kY COMMI$$10N Docket No, _ b Ofte tal Eth. h __: ! la the matter of _ d/ - ~ . _ . _ .-- R - - -
- JDENilFIED Staff #
Applic. ant _ V RICLIVID _. RfJECTED . Intmenor ___ Om S S14 . l 0,,,, n ,oa., 0 49d...; ew
' \
1 1 l I l l l 1
?ANIEXIGj;%,, Dl1
~ - " - - - - _ _ _ _ - _ _ . . . ._
l i i TMI-2 i REACTOR COOLANT ! INVENTORY BALANCE TESTING l l i , PREPARED FOR I GPU NUCLEAR CORP. I BY EDWIN H. STIER ! 1 INVESTIGATIVE STAFF: ) l FREDERICK P. DE VESA ROBERT T. WINTER 1 i i [ SEPTEMBER 5, 1985 { i l VOLUME II (B) CURRENT GPU SYSTEM EMPLOYEES : ASSESSMENTS OF INVOLVEMENT ! l IN LEAK RATE TESTING ' l
.1 1
t l l ~ i i INDIVIDUALS INCLUDED IN VOLUME II i 1 1 l ADAMS, Charles BEZILLA, Mark l COLEMAN, Mark CONAWAY, William f 1 CONGDON, Joseph R. l FAUST, Craig FELS, William i FREDERICK, Edward l GARRISON, Jack f GUTHRIE, Carl l HERBEIN, John i HOYT, Kenneth ILLJES, Theodore l KUNDER, George MARSHALL, Walter J. l MCGOVERN, Hugh MEHLER, Brian MILLER, Adam MILLER, Gary P. POTTS, William ROSS, Michael J. ' SMITH, Bernard TROFFER, George ZEWE, William 1
I I j i 4 4 JOHN HERBEIN began employment with Metropolitan Edison Company in September 1967. He is presently employed by.Penelec as Vice President, Station Operations. Between March 1978 and March 1979, he was the Vice President - Generation for Metropolitan Edison Company assigned to Headquarters in ! l Reading, Pa.1 He did not maintain an office at TMI. Hercein was ultimately responsible for generation at all facilities owned by Metropolitan Edison Company. And he was ! the highest level corporate officer directly responsible for the management of TMI.3 During 1978, he supervised j i cenerating station operations through Lawrence Lawyer, manaaer
)
of Operations for Metropolitan Edison Company, and other subordinates assigned to Headquarters in Reading, Pa. In 1979, as a result of a reorganization, Gary Miller, TMI Station Superintendent, began reporting directly to Herbein.4 Although there is no direct evidence, or even allegations, that Herbein was aware of, involved in, or willfully tolerated improper leak rate test practices at TMI-2, there is some l circumstantial evidence from which it could be inferred that he 1 was aware that the leak rate test was not precise in measuring leakage and was difficult to administer. It should be noted l 1 l l l l C -
JOHN HERBEIN that'Herbein is at th'e highest level of management for which there is any evidence that management knew ' of problems with leak rate testing at TMI-2. However, after closely . examining such evidence, we have concluded that it is insufficient to support a finding that Herbein was involved in, or willfully l tolerated improper leak rate test practices or the failure to take appropriate action at TMI-2 in response to indications of l excessive unidentified leakage. ; As Vice President - Generation at company headquarters in ! Reading, Herbein had virtually nothing to do with the conduct j or evaluation of leak rate tests.. He did not, of course, , I perform any tests personally, nor did he exercise any direct? responsibility for the oversight of leak rate testing at TMI-2.5 According to Herbein, he visited the Island ' once every five or six weeks and spent only brief periods in the control rooms, primarily to demonstrate his concern for the activities of his subordinates.6 Because of the breadth of his responsibil-ities, Herbein stated that he relied heavily on his managers to keep him informed of problems that required his. attention.7 c l JOHN HERBEIN i Herbein testified that he does not remember being informed ! I about serious problems associated with the performance of leak l rate tests in 1978 or 1979.8 He testified further that he did not remember that there was excessive leakage at that time, j nor did he recall any discussions with his subordinates regarding the possibility that TMI-2 might have to be shut down because of unsatisfactory leak rate test results.9 Herbein denied that there had been a decision to improperly keep TMI-2 i in operation in violation of Technical Specifications until TMI-l was placed back into service after refueling.10 He I i pointed out 'that Control Room Operators had the authority to l shut down the plant when necessary without higher approval. l , 1 ! One example of the evidence which suggests that Herbein may have had knowledge that repeated unsatisfactory leak rate test results were a problem. at TMI-2 is contained in the public statement made by the U.S. Attorney at the time Met-Ed entered l its quilty plea. In that statement, the U.S. Attorney briefly describes a telephone conversation, that allegedly occurred after the October 18, 1978, NRC inspection, which may have placed the TMI Station Superintendent, or Herbein, on notice of l l ? l JOHN HERBEIN i the difficulty experienced in performing leak rate tests at l TMI-2. The U.S. Attorney's statement says that: I i l Durina the conversation the operations l personnel alerted the Station Superintendent and/or Vice-President for Generation that i because of numerous " bad" leak rate tests I l obtained at Unit 2, the NRC's interpretation of the leak rate technical specifications would result in repeated shutdowns of the facility.ll The reference to the " Station Superintendent and/or I I Vice-President for Generation" indicates that the U.S. Attorney ) l was uncertain about which individual participated in the telephone call. In fact, the U.S. Attorney specifically states l that "The call was made to either TMI's Station Superintendent or Metropolitan Edison's Vice-President for Generation, or both."12 We specifically questioned Herbein about his alleged participation in this telephone conversation. He testified that he did not remember participating in such a discussion, and further stated that, if he had been involved in such a conversation, he thinks that he would have remembered it.13 None of the other witnesses that we interviewed could recall participating in, or being aware of such a conversation. Gary Miller, TMI Station Superintendent, specifically denied any i JOHN HERBEIN l { knowledge of this conversation.14 We have been unable to i l find any other evidence that confirms Herbein's participation i in this telephone call. l
)
The conversation would have alerted Herbein to the fact j that, operators were having difficulty obtaining satisfactory 4 i test results. However, the telephone conversation, as l j described by tihe U.S. Attorney would not have put Herbein on l notice of the practices of discarding unsatisfactory test j results, filing tests without proper evaluation, i exploiting j f unreliable data and manipulating tests. I I
- lJ 2
When questioned specifically about improper leak rate test practices, Herbein testified that he was not aware that operators were discarding and failing to document leak rate test results at TMI-2.1'I Moreover, he stated that he was not aware that they were manipulating test results and that he did not believe that management would have directed or ordered them to do so.16 We found no evidence to refute his testimony. None of the witnesses we interviewed has testified that Herbein was aware of leak rate test misconduct at TMI-2. Logan, Miller, Pctts, Troffer, and other high level managers who frequently
JOHN HER.BEIN communicated with Herbein have testified that they themselves I were not aware of such misconduct.1 Herbein did not attend TMI-2 Plant Operations Review Committee (PORC) or Plan of the-Day (POD) meetings, or engage in day to day activities that would have brought leak rate test misconduct to his attention. l l There were some documents routinely $ transmitted to { management in Readina, which, if closely scrutinized, would ' i have provided some indications that ! i operators were having difficulty performing leak rate tests satisfactorily. Each. day l ! a Daily Plant Status Report was circulated among management I personnel, includino the Unit and Station Superintendents, as well as their superiors at Met-Ed Headquarters in Reading, Pa. Significant issues raised in the report were discussed each mornina in a conference call i r, which they all generally participated. The report showed a figure for unidentified leakage. On numerous successive days, the same figure was reported.18 The Unit and Station Superintendents had attached to their copies of the Daily Plant Status Report copies of the shift turnover notes prepared by the Shift Supervisor on the llpm-7am shift.19 An example of the kind of entry that might be made in the shift turnover notes is the following notation made on October 18, "Still could not cet a leak rate - 1900 today is deadline. Doing hand calculations. i
W ,. ,, g
'/
v , _f
.) .. , ,
'(_0 JOHN TiERBEIN^ ! <
.Cannot find leak. Don't believe it is in the RB." It does not appear that copies of tho- shift turnover notes were y-attached to- the Daily Plant Status Peports transmitted to management in. Reading.
( ,. Although other witnesses differfd on whether Herbein would l generally parti ci pa te in the daily r.orning ' conference call, i
, Herbein believes that his participation was infrequent because his duties required him to go to other plant si tes .'and , as'a result, he was not always in the office. Additionally, although Herbein admitted receiving . the Daily Plant Status Report "on occasion," he did not believe that he, received it
( i "on a frequent basis" or "that it was something I focused- l I I on."23 Notwithstanding the discrepancies in the testimony ; I concerning the frequency of Herbein's. participation- in. the l morning conference calls and his receipt of the . Daily Plant
.{
Status Report, the participants we questioned testified that they did not compare entries in the Daily Plant Status Reports from day-to-day and did. not notice that identical leak rates-were reported for several days. Moreover, these witnesses did ; not recall discussing continuing. problems with leak rate testing in their conference calls. l
- 7- -[ corrected 1/17/86]
1 ,
',, l ,
x . -
N JOHN HERBEIN annot find lead. Do'n ' t believe it is in the RB." It does
\
not \ appear that copies of the shift turnover notes were attach d to the ' Daily Plant Status Reports transmitted to managemeNn . in Reading.21
\ u Although ther witnesses recalled that Herbein would j 1
generally parti ipate in the daily morning conference call, j s N- 1 Herbein believes khat his participation was less ' frequent because his duties r gpiqed him to go to other plant sites and, i 4 - a as a result, he was always in the office. Additionally, although Hep
\.not in admitted receiving the Daily Plant Status Report #on occas{on," he did not believe that he received it "on a fi equent basrs\ " or " that it was something I
\
focused on."23 Notwithstanding the discrepancies in the testimony concernino the frequency of Herbein's participation I in the mcrning conference calls and a receipt of the Daily Plant
- x l
Statusc Report, the participants we \guestioned
. s testified that , they did not compare entries in the ' Daily Plant Status Reports f rom day-to-day and did not notice tha't identical leak
}; s rates were reported for several' days. Moreover, these witnases did not recall discussing- continuinq p blems with leak ra'te testing in their e i;ference calls. 24 .
\ \
x \ o S i A t I
.w m- .A
f JOHN HERBEIN 1 Given the broad range of Herbein 's . responsibilities for i several generating plants, and aiven. the fact that he relied upon subordinates in the chain of command to bring significant matters to his attention, it is likely that Herbein did not focus upon the detailed entries in the Daily Plant Status Reports concerning leak rates. It must also be recognized that ; the groups outside of the chain of command specifically designed to identify such problems, e.g. PORC, QA, GRC & GORB, i l did not express concerns about leak rate test practices. l The stroncest indication that Herbein may have had knowledge of improper leak rate test practices flows from the fact that he signed LER 78-62 and submitted it to the NRC on November 1, 1978. However, the extent of Herbein's knowledge of such practices was severely limited by misleading statements included in the LER. Such statements did not truly describe the events which gave rise to the LER and did not accurately describe the action taken in response to indications of excessive unidentified leakage. As described in more detail in volume I of this report, LER 78-62 resulted from an NRC Inspection on October 18, 1978. At that time, an NRC Inspector discovered several test data sheets which indicated that leak rate test results had been obtained l JOHN HERBEIN showina unidentified leakage in excess of 1.0 gpm, and action required by the Technical Specifications had not been taken. As a result of the NRC Inspection, PORC determined that a reportable occurrence had taken place, and an LER was to be prepared for submission to the NRC.25 The final version of LER 78-62, however, conveyed misleading information to Herbein, and to the NRC, recarding the failure of Operations personnel to enter the action statement in response to unsatisfactory leak rate tests between October 16 and October 18, 1978. Among other things, the LER inaccurately reported that excessive unidentified leakage had been reduced to allowable limits on October 18, 1978, "by determining a portion of this leakage to be identified leakage ~ l from the Reactor Coolant System and to be well within the limits of T.S.3.4-6.2.c."26 However, the method that was actually employed to bring the leak rate test result within limits at that time was to round it off to a whole number, thereby making a test result up to 1.5 gpm acceptable.2 We also discovered that the language in the LER that refers to " identified leakace" was not included in the draft LER that was submitted by PORC Chairman, James Seelinger, to George Troffer, manaaer of Quality Assurance in Reading, Pa. It was
1 JOHN HERBEIN apparently added to the LER during the review process in l Reading.28 After reviewing the documents in question, Herbein l acknowledged that his signature appeared on the letter transmitting the LER to the NRC on November 1, 1978.29 i However, we cannot conclude from the evidence that Herbein had knowledge of the true facts underlying the LER, or how the j l problem was actually remedied, prior to signing and transmitting the LER to the NRC. When questioned, Herbein stated that he had no independent recollection of the LER and that he could not recall any discussions regarding its preparation or follow-up.30 He l also stated that he did not remember being aware that TMI-2 personnel had resorted .to rounding off leak rate test results in order to remain within acceptable limits.31 According to Herbein, LER's were drafted by Licensing after consultation with the appropriate individuals at TMI. The draft report would be reviewed by Potts, Troffer and the GRC at Reading, and finally submitted to Herbein for his signature. He would not approve an LER unless it was previously approved by Licensing.32 Typically, he would have the LER, a proposed
]
'l i
JOHNtHERBEIN i I transmittal letter' for his signature, and perhaps some other related documents in a file for his review.33-
~
Herbein indicated that if he had problems with the wording i of an LER, he would generally talk to.his Licensing group and ! that they . would address his concerns, and note the necessary changes. He stated that the Licensing group would- normally consult with TMI before making any changes to an LER.34 he communications necessary to agree upon a change in the LER could take place among any number of persons involved _ in the - review process.35 Herbein stated that he did not know that LER 78-62 contained inaccurate information and he did not know how the change in question- was made.36 He agreed that it was. probably made in Reading, but maintained that the change in the language would not have been done without' getting the information from someone at TMI.37 His testimony in 'this regard was consistent with the testimony of both Troffer and Potts. Unfortunately, because of the- lack of documentation, we cannot be certain who was responsible for the additional. misleading language in the final draft of the LER; however, we believe that such language was supplied by individuals at TMI. l E _..
i i i l JOHN HERBEIN Given Herbein's position and the nature of the review ! process, it is not likely that he was responsible for the i change in question. Herbein stated emphatically that it was not his policy to condone the submission of inaccurate .or misleading information to the NRC. We found nothing to i refute his statement. 1 J l l JOHN'HERBEIN i 1- Herbein 2/15/84, p. 5. I' 2 ( ~. Ibid.,-p. 6. 3 Ibid., p. 7. 4 Herbein'2/8/85, p. 47; 2/15/84, pp. 5-6. 5 Herbein 4/23/84, p. 6. 6 Herbein 2/15/84, p._9: 2/8/85, p. 9. 7 Herb'ein 4/23/84, pp. 6-7. 8 Herbein 4/23/84, p. 12; l 2/8/85, pp. 27-30, 40; 2/15/84, po. 12-14. 9 Herbein 2/8/85, pp. 40-44. 10 Herbein 2/15/84, pp. 26-27. 11 Statement of Facts submitted by the United States, (United States of America v. Metropolitan Fdison Company), February, 1984, (Tab 2). 12 Ibid., p. 13, (Tab 2). 13 Herbein 2/8/85, p. 40. 14 Miller, G. 2/15/85, pp. 75-76. 15 Herbein 4/23/84, p. 12. 16 Herbein 2/8/85, pp. 28-30; l 2/15/84, pp. 28-29. 17 Logan 3/27/85, pp. 61-64, 116, 125-130; 4 Miller, G. 2/7/84, p. 23; 2/15/85, pp. 28-32; Potts 1/6/84, p. 19; Troffer 3/20/85, p. 10. : t 18 Daily Plant Status Reports, (Tab 10). Vol. III, Table 4. [ corrected 1/17/86]- l l s
JOHN HERBEIN 19 -Shift Supervisor Meeting Minutes. October 5, 1977, (Tab 23). . Logan 3/27/85, pp. 28-29. 'f
)
20 Shift Supervisor Turnover Notes, October 18, 1978, (Tab 10). 21 Shift Supervisors' Meeting Minutes, October 5, 1977, I (Tab 23); f Zewe 2/27/85, pp. 86-87. 22 Herbein 2/8/85, pp. 8-10; Lawyer 11/10/83, pp. 18-19; l Logan 3/27/85, pp. 21-23; Miller, G. 2/7/84, p. 11; 2/15/85, pp. 9-10. l 23 Herbein 2/8/85, p. 9. J l 24 Miller, G. 2/15/85, pp. 31, 56. Logan 3/27/85, pp. 25-27. 25 Letter, J. Seelinger to B. Grier (NRC), Prompt Report i to Reportable Occurrence, October 19, 1978, (Tab 31); ' Bezilla 2/22/85, pp. 11-14. 1 26 TMI-2 Licensee Event Report 78-62/1T, November 1,.1978, (Tab 29). 27 Vol. III, Table 1- ! 4 Vol. IV (MPR . Repor t ) , Appendix K, Test Numbers 147, i 146, 145 and 144. ' 20 TMI-2 Licensee Event Report 78-62/1T (Draft),' November' l 1978, (Tab 28); j LER 78-62/1T, November 1, 1978,-(Tab 29). i 29 Herbein 4/23/84, p. 24. j 30 Ibid. l [ corrected 1/17/86]
l 4 i i l JOHN HERBEIN 31 Herbein'4/17/85, p. 36.
)
32 Ibid., pp. 5-8. I i 33 Ibid., p, 11, 34 Ibid., p. 14. i 1 35 Ibid., p. 15. 36 Ibid., pp. 39, 46-47. 1 37 Ibid., pp. 37-40. i 38 Troffer 3/20/85, p. 49; l Potts 4/18/85, p. 17. I 39 Herbein 4/17/85, pp. 29-30. I i i l 1 1 I [ corrected 1/17/86]
JOHN HERBEIN l l Herbein 2/15/84, p. 5. 2 . Ibid., p. 6. 3 bid., p. 7. 4 - Herbein.2/8/85, p. 47; 2/15/84, pp. 5-6. 5 Herbei 4/23/84, p. 6. 6 Herbein '15/84, p. 9. l l 7 Herbein 4/2 84, pp. 6-7. 8 Herbein 4/23/8 p. 12; 2/8/85, p. 27-30, 40; l 2/15/84,NRp. 12-14. 9 Herbein 2/8/85,'pp. 0-44. 10 Herbein 4/23/84, pp. 2 27. 11 Statement of Facts submitbed by the United States, j (United States of America ds Metropolitan Edison ~ Company), February, 1984, (Tag 2). 12 Ibid., p. 13, (Tab 2). 13 Herbein 2/8/85, p. 40. 14 Miller, G. 2/15/85, pp. 75-76. 15 Herbein 4/23/84, p. 12. 16 Herbein 4/23/84, p. 29. 17 Logan 3/27/85, pp. 61-64, 116, 125-130, ; Miller, G. 2/7/84, p. 23; j 2/15/85, pp. 28-32; j Potts 1/6/84, p. 14; 3/7/85, p. 19; . Troffer 3/20/85, p. 10. ' 18 Daily Plant Status Reports, (Tab 10). Vol. III, Table 4.
N \ JOHN HERBEIN
\
Shift Supervisor Meetina Minutes, October 5,'1977, (Tab 1{\N 23). l N Logan 3/27/85, pp. 28-29. 20 N Shift Supervisor Turnover Notes, October 18, 1978, (Tab 10). 21 N 1977, Shilg23); (Tab Supervisors' Meeting Minutes, October 5, N-Zewe 2/2.7/85, pp. 86-87.
- 22 Herbein 2/ 85, pp. 8-9; Lawyer 11/1 83, pp. 18-19; Logan 3/27/8 , pp. 21-23; Miller, G. 2/ 84, p. 11; 2/15 85, pp. 9-10.
23 Herbein 2/8/85, p. 9. 24 Miller, G. 2/15/85, p. 31, 56. Logan 3/27/85, pp. 26- . 25 Letter, J. Seelinger to B. Grier (NRC), Prompt Report to Reportable Occurrence,xOctober 19, 1978, (Tab 31); Bezilla 2/22/85, pp. 11-14. 26 TMI-2 Licensee Event Report 78- /lT, November 1, 1978, (Tab 29) . 27 vol. III, Table 1; Vol. IV (MPR Report) , Appendix K, Test umbers 147, 146, 145 and 144, 28 TMI-2 Licensee Event Report 78-62/lT (Draf , November 1978, (Tab 28); LER 78-62/lT, November 1, 1978, (Tab 29). 29 Herbein 4/23/84, p. 24. 30 Ibid. N N
l JOHN HERBEIN l 31 Herbein 4/17/85, p. 75. 32 Ibid., pp. 5-8. l_ 33 Ibid., p. 11. 34 bid., p. 14. 35 Ibhd.,p. 15. 36 Ibid. pp. 39, 46-47. 37 Ibid., . 37-40. ! i ' I t 38 Troffer 3 /85, p. 49 I' Potts 4/18 , p. 17. d 39 Herbein 4/17/h , pp. 29-30. i
\
- t. D
[. 1 1 KENNETH HOYT began employment with Metropolitan Edison Company on April 2, 1971. He is presently employed . by GPU Nuclear as an Engineering Assistant (Sr. III). Between March 1978 and March 1979, he worked.as a Shift Foreman assigned to TMI-2. He was generally responsible for supervising Control I Room Operators Harold Hartman, Raymond Booher and John Blessing. The Shift Supervisor that he normally reported to was Bernard Smith. Analysis of employee testimony, leak rate test records, and related plant documents, reveals that Hoyt failed to properly supervise the performance of leak rate tests and also failed to ; ensure that required action was k taken in response to l 1 indications of excessive reactor coolant . system (RCS) leakage-during January 1979. There is insufficient evidence from which i we can conclude that Hoyt participated in, or .knowinoly tol.erated, manipulation of leak rate test results. Supervision of Leak Rate Testing There is substantial evidence that Hoyt failed to properly supervise the performance of leak rate tests that were conducted.
'l a
.1 KENNETH HOYT on his shift. He - tolerated and participated in a practice of discardino and failino to document all unsatisfactory leak rate test results that were conducted on his shift. He also approved of the filing of numerous satisfactory test results which'were clearly invalid and which did not-accurately. measure i RCS leakage. These conclusions are ' supported by analysis of plant records and Hoyt's testimony.
TMI-2 records indicate that between March 1978 and March 1979, Hoyt reported for duty and worked as a Shift Foreman on approximately 96 shifts during which leak rate tests presumably could have been performed.1 Despite a company practice of attempting to perform a leak rate test at least once per shift, only 32 leak rate tests were filed during all of the shifts supervised by Hoyt.2 Hoyt sioned as the approver of nearly all of these tests.3 Aside from the fact that no tests were l filed showing l unsatisfactory leakage, there no l are obvious patterns ! associated 'with leak rate testing supervised by Hoyt.4 In an effort to explain the absence of filed tests on so many shifts, i Hoyt and others have claimed ' that they did not attempt to run I tests on-the day shift because of other work that was coing on in the plant.5- This explanation is not l supported by the
i
.,)
l KENNETH HOYT plant records pertaining to leak rate testing on Hoyt's shift. l Although there are n u m e r o'u s . 7 a m - 3 p m shifts in . which no test results.were filed, there is an almost equal number of shifts ! on which tests are not filed on both the 11pm-7am and 3pm-llpm shifts.6 There were a number of tests filed on the day shift and there were also a number of instances of consecutive assigned . shifts supervised by Hoyt when no leak rate test result was filed.7 These gaps in filed tests appear j throughout the entire period of TMI-2 operation.8-( l Unlike some of the other shifts, Hoyt's shift seems to have been somewhat successful in obtaining satisfactory results even during periods of higher than normal RCS leakage.9 As will be discussed later, there is evidence which suggests that some ; l of these successful test results obtained during the high leakage periods were manipulated by the Control Room Operators - working under Hoyt's supervision. l 1 l l l In any event, there is considerable testimony.to the effect l l that the absence of filed tests in general was brouaht.about by the' practice- of -simply discarding all unsatisfactory test results.10 Hoyt's testimony makes.it clear that he tolerated and even participated in this practice. Indeed, Hoyt testified that l I KENNETH HOYT I nearly all unsatisfactory test results. that were obtained on his shift were brought to his attention. Thereafter, either he. or his ' Shif t Supervisor would throw them ' in the trash can without~ making any further documentation of the test result.11' ' l ! Hoyt initially maintained that ' only inaccurate or invalid l test results ' were discarded in this fashion. He stated - that other plant instruments would be reviewed ' and if ' they. did : not agree with the. test results, the test would be deemed invalid and would be discarded.12 1 j In an attempt to explain the absence of any filed test results that were unsatisfactory, Hoyt testified that he did-not recall ever receiving an unsatisfactory leak rate test result that he believed to be ~ accurate and in agreement -with other plant instruments. 13 He admitted being aware, however,. l l that no log entries, no Exception and Deficiency reports, 'and no other forms of documentation were made with respect to the . so called invalid test results that were discarded. -His explanation for his tolerance of these procedural violations was that he and the' operators " sort of fell into a rut".14 , He denied that there was an express agreement among supervisors , not to comply with company procedures that' required documenta-- tion of.all surveillance tests.15 , 4- 3 i
l KENNETH HOYT Ultimately, upon further questioning after a 12 day recess , I from his interview, Hoyt altered his initial position with respect to discarding unsatisfactory test results. During his second interview session, he acknowledged that in 1978 and 1979, he had no way of really determining whether the unsatisfactory test results that were being discarded were invalid.16 He also conceded that it was the practice at that time to discard any test result showing leakaae in excess of 1.0 gpm, irrespective of actual plant leakage.17 Despite his l l admissions in this regard, he continued to maintain that he l believed that his shift was in compliance with the requirement to maintain leakage below the 1.0 gpm limit.18 Hoyt testified that although he did not directly do anything to prevent the practice of throwing away test results, he did express concern to Bernard Smith, his Shift S upe r tri so r , about the need to repeatedly run leak rate tests and about the need for a more accurate computer program. He did not recall Smith's response.20 When Hoyt was questioned about whether his tolerance of this practice was prompted in any way by upper level management pressure, he stated that he could not recall feeling under pressure.21 He did state, however, that a Shift Supervisor, i 5-
l i l KENNETH HOYT l I probably Smith, instructed him on one occasion to discard 1 ! unsatisfactory leak rate test results and not to leave them J lying around on the desks.22 He thought that this j instruction was prompted by an NRC inspection during which l several unsatisfactory leak rates were observed on a Shift ; 1 Supervisor's desk. He assumed that as a result of this ! inspection, top management had directed that unsatisfactory I test results should no longer be left lyino around before being , 1 discarded.23 i I l In addition to participating in the discarding of all j unsatisfactory leak rate tests, Hoyt also approved numerous satisfactory tests for filing that were obviously inaccurate or l otherwise performed in violation of test procedures.24 Indeed, our analysis revealed that more than half of the tests approved by Hoyt should have been determined to be invalid.25 l l At least four of the tests approved by Hoyt contained obviously inaccurate information on the test printout. itself.26 A few others were performed during non-steady state conditions and seven satisfactory tests performed in December 1978 were performed with faulty makeup tank level instrumentation.27
KENNETH HOYT. Hoyt's testimony makes it clear that he approved these invalid test results without any real concern for their accuracy, but simply because the results showed leakage under 1.0 qpm. According to Hoyt, in 1978, he believed that the leak rate test was not really capable of producinq a precise measurement of RCS leakage and that test'results were generally inaccurate.20 As a practical matter, he felt that the- test-was more useful as a tool for monitoring RCS leakage trends. Nevertheless, since the company and his supervisors required that the test be used to measure leakage, he turned in the paperwork to satisfy that requirement.29 When questioned about his test review process, Hoyt { initially testified that it was his practice - to make visual observations of various plant instruments and indications of RCS leakage during his tours of duty. When he was given leak
]
rate test results to approve, he would review the information on the printout and compare it to what he had already learned j from tourino the plant. If the test appeared to be valid, he - would sian and submit- it.. If he found the numbers to. be t inaccurate, he would refuse to approve and the test would instruct the Control Room Operator to run another test.30
- 'I i
KENNETH HOYT After further questioning and being confronted with some-of .l I the obviously invalid tests that he approved, Hoyt. altered'his
)
{ initial position somewhat. 1 i First of all, he clarified his statement regarding- the I nature of his test ' review process ' by pointing out' that when i I test results were submitted to him,-he.only alanced quickly at -l the numbers on the printout- to see if they looked realistic.31 He stated that he could not really' compare them' ( with his prior read'ings ' f rom :other plant instruments since he-would usually be reviewino the printout hours -later and plant" l conditions would not remain i that steady over time.32 Ultimately, Hoyt admitted that it was the practice at the time to merely sign and turn in all satisfactory test results reaardless of their inaccuracy.33 Test Manipulation l l The NRC has alleged that improper water -and hydrogen f additions were made curing several of the leak rate tests that were - approved by Hoyt.34 ' This is not surprisinq since - most ' of the tests run under Hoyt's- supervision were run by- Hartman and Booher. Hartman, of course, has admitted that- he manipulated tests.35 And he has also testified that Bocher a
KENNETH HOYT l manipulated tests as well.36 Hartman refused to be interviewed during .this investigation. Our review of his prior testimony, however, . did not uncover statements which clearly incriminate Hoyt. in test manipulation. On the contrary, Hartman's testimony, although somewhat ambiguous, suggests that Hoyt was not aware that operators were adding water and i hydrogen during leak rate tests for the purpose of manipulating j test results.37 i In his'NRC interview of March 26, 1980, Hartman first said that he did not know if his supervisors were aware of test l manipulation by Control Room Operators.38 He then added that in explaining to Hoyt and Smith why he could not get.a good leak rate, he would say things like,
...well, you know I don't want to fudge them !
and I'd keep telling them, I says I'm tired of wrestling with these, with these leak rates. I don't want to, we have got a problem here, why don't you do something about it.39 However, Hartman stated that he did not advise his supervisors when he added hydrogen to the makeup tank.40 Finally, during his last interview with the NRC on July 26, 1983, he again stated that he was not sure if his superiors
I l KENNETH HOYT were aware of test manipulations.41 Later in the same interview, however, he acknowledged that he was not aware of l any Shift Foreman or Supervisor who knew that hydrogen was being added to the makeup tank for the purpose of test l manipulation.42 l ! Our analysis of records pertaining to leak rate tests l approved by Hoyt also failed to produce conclusive evidence that he was aware of test manipulation by Control Room i Operators on his shift. l l l l The NRC has alleged that two of the numerous tests approved by Hoyt involved hydrogen additions during the course of the test. One of the tests was performed by Hartman on December 15, 1978, and the other was performed by Booher on December 26, 1978.43 We were able to confirm the Hartman addition, but we were not able to confirm the alleged addition in the Booher test. It should be noted that in an unsworn interview during our investigation, Booher denied that he ever added hydrogen durina a leak rate test in order to manipulate the test result.44 Hoyt also denied being aware that any operators under his ! supervision were adding hydrogen for the purpose of falsifying I ,
]
1 l KENNETH HOYT i: test ' results.- Accordina to him, he.did not even think it was possible to influence- test results by making hydrogen additions.45 In addition to using hydrogen, the NRC alleged that several of the tests. approved by Hoyt were influenced,by unrecorded ~or under-recorded water additions.46 Hartman admitted that he 47 However, Hoyt and Bocher manipulated tests in this manner. - testified that he was not aware that his operators were addina water to the makeup tank to falsify test results in this way.40 There is no conclusive evidence in the test records that repudiates Hoyt's testimony. Several of the alleaed water additions could not be confirmed because of- the previously discussed problems with the makeup tank level-instrumentation.49 The slight flattening ' of the makeup tank level recorder trace that appears in five tests during. early March 1979 could easily have been caused by a number of events i other than jogged water additions.50 For example, the strip chart for the test of March 15 appears more likely to have been influenced by a hydrocen addition than by a water addition.51:
]
i i KENNETH HOYT ! In any event, there is little evidence to suqqest that this f type of covert test manipulation would have been known to ! Hoyt. Among the eaveral tests that Hoyt approved that may have been manipulated, our analysis revealed only_ two in which there l was some evidence that Hoyt may have been aware of the improper t j conduct of the operator.52 5 Both of these tests were performed by Hartman, and involved water additions that were loqqed but were not accounted for in the test calculations. Both tests were performed during a time l when operators had not been able to obtain a satisfactory test l
)
result for several shifts.53 Hoyt loqqed each of these water ! i additions in the Shift Foreman Log but then approved the test results which clearly made no reference to the water additions l that had been made.54 Since Hoyt obviously had personal knowledge of the water additions in question, he of course should have realized that the leak rate tests that he had approved were not accurate. When confronted with these tests, Hoyt conceded that there were unaccounted-for water additions that he had in fact loqqed in the Shift Foreman Log. He denied, however, that he was aware of any effort by Hartman to manipulate the test re-sults.55 Hoyt explained his approval of the tests by saying
i KENNETH HOYT j that he did not know what he was thinking' at the time, but that there must have been an oversicht on ~his part.56 Given, the possibility of an oversicht and-the lack of direct evidence to L the contrary, we cannot conclude. that Hoyt knowinaly tolerated 'l
)
! leak rate test manipulation by Hartman in'these cases. 1 l l . Operator Response to Indications of Excessive Leakace t-The practice of discardina unsatisfactory results without documentation has made it difficult to identify every brief period when there were indications that unidentified leakaae l' l exceeded 1.0 apm on shifts supervised by Hoyt. However, durina l
-the periods we did identify when unidentified leakage appeared to exceed 1.0 opm, Hoyt was present and responsible for. super-l vising leak rate testing only in January 1979.57 .
It is clear q that he did not take steps to ensure. that proper action .was taken in response to excessive RCS leakaae at~that time. 1 Our analysis of that period reveals the unidentified leakaae gradually increased to above 1.0 opm on or- about January 3, 1979, and remained above 1. 0 g pm until January 15, 1979, when the plant was shut down.58 Hoyt and the Control
.j Room Operators under his supervision were not in the control i
{ _ - _ _ _ - - - - _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ -_ - - - __--__=_:_-____-__:
[ i> i. l KENNETH HOYT room in early January but were present in the control room for t I j five consecutive days on the 7am-3pm shift from January 10 to ' i January 14,.1979.59 i i Hoyt's shift did not file a leak rate test on January 10 or January 11, 1979.60 By this time, it was becoming increasingly ' difficult to obtain a satisfactory result, and there is considerable testimony that it was becoming necessary to discard numerous leak rate tests before obtaining a satisfactory result.61 There are also indications that i increased RCS leakage from pressurizer valves and other sources was of some concern among Operations personnel.62 On January 11, a Radiation Work Permit was issued for Hoyt, l a Health Physics Technician, and a Maintenance Foreman to enter l" the reactor building inside the "D-ring" to check for leaks at the pressurizer.63 Two other permits were issued for routine entry and inspection on the same day.64 Sump collection was on the , increase and a considerable amount of water was added to the makeup tank during this shift.65 1 Plant records also reveal that Hoyt and the others all received. radiation doses, but there is no record of the nature and amount of RCS leakage that they observed.66 Since there i 14 - l L
f I l )' l l I KENNETH HOYT i is no leak rate test on file, it is likely that any test that I i was performed on that day yielded unsatisfactory results and was discarded. There is no record of official entry into the i action statement because of RCS leakage. On the contrary, the Control Room Log shows that power was increased during the shift.67 On the next two shifts, indications of excessive RCS leakage continued. Plant records reveal that, at 21:53, 1 Coleman filed a leak rate test that was satisfactory only because a water addition was made during the test.68 On January 12, 1979, Hoyt and his crew again reported for work on the 7am-3pm shift. Plant records indicate continued excessive RCS leakaae and no leak rate test was filed. The next two shifts were also unable to obtain a satisfactory leak rate test result.70 On January 13, 1979, at approximately 10:40, Hoyt and-Hartman filed a leak rate test showing unidentified leakage at 0.2639 apm after almost 40 hours had passed without a 1 satisfactory result. This test was one of the previously discussed tests that were influenced by an unrecorded water addition during the test. 15 - i l
e ;= . l l l.
.KENNETH HOYT i
.1 Hoyt approved this test despite indications that RCS
\
\
l leakage was not within acceptable limits and the test result ! was not accurate.72 i At 12:45 on the 'same day, Hoyt and personnel from Maintenance and Health Physics again entered the reactor building to check for leaks at the pressurizer.73 i; ' t l Hoyt's .bg entry regarding this activity reads as follows: f Made inspection in Rx Bldo to check leakage from RC-V-133B, 134B, 135B. Total leakage from these valves is approximately the same ! as Thursdays [l/11] inspection. Survey of l cther areas (?) showed no other leakaoe. l Leahace is from body to bonnet areas of i these valves and appears to be safe yet to op'e r a te . 7 4 1 Since the leak rate test that was submitted contained no ! reference to this leakage, there is no official record quantifying the amount of leakage that was perceived at that time. An informal memo prepared by George Kunder, Unit ! Superintendent - Technical Support, pertaining to the same leakage, estimated the leakage to be approximately 2.0 opm and t. characterized it as identified.75 Apparently, it 'was this f
-l16 -
'~
h.
P1 4, s c. k l KENNETH H0YT l ! I-i characterization that provided the justification to continue
't operating,the plant.
e, On[ January 14, 1979, Hoyt again reported for work on the I i day shif t.76 The two previous shifts had not been able to ) i i( I obtain a satisfactory leak rate test and, continued to 1 experience indications of leakage.77 excessive RCS ' Plant records reveal that Hoyt'c' . shift also had not filed a - sa tist'actory leak rate test.78 They continued to operate the plant despite indic.ations that the pressurizer? leaks and other l l ! ' N 4 RCS leakaae ccotinued.
<% 3 1-1 1-
)
Based on Hoyt's testimony, it is clear that they had no basis for continuing to operate the plant. 4 i l I l It should be noted that Hoyt testified that he understood \ it to be part of his responsibility as a Shift Foreman to ensure that leakage remained within limiting conditions for operation.79 He further stated that if he believed that leakage exceeded 1.0 ppm, b? would notify management and recommend a shutdown of the plant.80 He conceded that he had never made such a recommendation, but explained that he believed that leakaae at TMI-2 never i 1 l 4
I KENNETH HOYT exceeded 1.0 gpm.81 After considerable questioning recarding the basis for his beliefs in light of his attitude about the accuracy of leak rate testing, he eventually conceded that he was not really able to be sure that leakaae remained under 1.0 qpm.02 However, he stated that he would make routine tours of the plant and would observe plant instruments to approximate
- <CS leakage. He maintained that he did not really believe that l they were operating very much in excess of 1.0 opm and did not think they were jeopardizing plant safety.83 '
When he was questioned specifically about the events of January 1979, he stated that he had very little recollection of the period, nor did he remember any discussions about leakace problems with specific individuals.84 He did remember entering the reactor buildina to very quickly observe a steam leak coming from the pressurizer and estimating the leakaae to be approximately 0.5 apm.85 He conceced that he could not accurately measure the leakaae, and that he only provided a very rouah estimate of the leakage because he was probably asked to do so by his Shift Supervisor.86 i He also conceded that, aiven his inability to quantify the leakace he observed and the information available from other plant instruments, there was a aood possibility that they were
KENNETH HOYT operating with unidentified leakage in excess of 1.0 gpm.07 l Nevertheless, he could not explain why he did not direct entry l into the action statement and proceed to shut down the plant.88 He said that he realized now that they probably should have shut down the plant, but at that time they must have believed that they were operatina safely and close to or under 1.0 gpm.09 i I 1 l KENNETH HOYT l ! A Vol. III, Table 2. 2 Ibid. 3 Ibid. 4 Ioid, l 5 Hoyt 2/14/85, p. 39. 6 Vol. III, Table 2. 7 Ibid. 3 Ibid. 9 Ibid.; Vol. IV (MPR Report), Figure IV-20. 10 Adams 3/19/85, p. 13; Coleman 2/5/85, pp. 32-33; Congdon 2/13/85, p. 29; Faust 2/19/85, pp. 110, 132-133, 136, 138-139; Hoyt 2/26/85, pp. 4, 8; ,i Illjen 2/7/85, p. 1; i McGovern 2/6/85, p. 109; ) Mehler 2/28/85, pp. 57-60, 63-64, 66; j Miller A., 3/21/85, pp. 20-23; I Zewe 2/27/85, pp. 10-11, 37. 11 Hoyt 2/14/85, pp. 43-44. 12 Ibid., pp. 40-41, 56. 13 Ibid., pp. 41-43. 14 Ibid., pp. 46, 48-51. l 15 Ibid., pp. 46-47. 16 Hoyt 2/26/85, p. 12. 17 Ibid., pp. 13-14. 18 Ibid., p. 16. 19 Ibid., pp. 18-19. l _ _ _ _ _ _ _ _ .0
j l l \ KENNETH HOYT I 20 Hoyt 2/26/85, p. 19. 21 ~ Ibid., pp. 20-22. 22 Ibid., pp. 22-2'3. 33 Ibid., pp. 22-26, 34 Vol. III, Table 2; Vol. IV (MPR Report), Appendix K, Test Numbers 221,. 220, 214, 212, 205, 188, 106, 102, 97, 94, 86, 83, 81, ' 64, 30. 25 Vol. III, Table 2. 26 Ibid., Test Numbers 205, 200, 189, 188. 27 Ibid., Test Numbers 106, 102, 97, 94, 86, 83, 81. 28 Hoyt 2/14/85, pp. 10-13. 29 Ibid., pp.-12-13. 30 Ibid., p. 18, 31 Ibid., 'p. 52. 32 Ibid., p. 52. 33 Hoyt 2/26/85, pp. 44, 46-47, 75. 34 Vol. IV (MPR Report), Appendix K, Test Numbers 102, 95, 81, 64, 61, 30, 22, 15, 13. 35 Hartman 4/27/80, p. 97; j 7/26/83, p. 2; I 3/26/80, p. 19. 36 Hartman 7/26/83, p. 1. 37' Ibid., pp. 1-2. 38 Hartman 3/26/80, p. 27, i 39 Ibid., p. 30. i L
)
[ corrected 1/17/86)
-- - - _ _ - - - -_ - ---_- -- - a
KENNETH HOYT 40 Hartman 4/27/80, pp. 57-58. l 41 Hartman 7/26/83, pp. 1-2. 42 Ibid., p. 2. 43 Vol. IV (MPR Report), Appendix K, Test Numbers 95, 81. 44 Boober 10/17/84, p. 8. 45 Hoyt 2/26/85, pp. 68-69. 46 Vol. IV (MPR Report), Appendix K, Test Numbers 102, 81, 64, 61, 22, 15, 13. 47 Hartman 7/26/83, p. 1. 48 Hoyt 2/26/85, pp. 139-140. 49 Vol. III, Table 2. 50 Vol. IV (MPR Report), Appendix K, Test Numbers 22, 15, 14, 13, 10. 51 Ibid., Test Number 10. 52 Ibid., Test Numbers 64, 30. 53 Vol. III, Table 1. 54 vol, IV (MpR Report), Appendix K, Test Numbers 64, 30. 55 floyt 3/27/80, p. 8. 56 Hoyt 2/26/85, pp. 107-115, 139-140. 57 Vol. III, Table 2. 58 Vol. IV (MPR Report), Figure IV-20. 59 Vol. III, Table 2. 60 Ibid. [ corrected 1/17/86}
i k l KENNETH HOYT 0 Ibid., p.-19. 21 Ibid., pp. 20-22, 22 Ibid., pp. 22-23. 23' bid., pp. 22-26. 24 Vo . III, Table 2; Vol. IV (MPR Repor t) , Appendix- K, Test Numbers 221, 220, 4, 212, 205, 188, 106, 102, 97, 94, 86, 83, 81, 64, 30, 25 Vol. III, able 2. 26 Ibid., Test Numbers 205, 200, 189, 188. 27 Ibid., Test N bers 106, 102, 97, 94, 86, 83, 81. 28 Hoyt.2/14/85, pp. 10-13, 29 Ibid., pp. 12-13. 30 Tbid., p. 18. U 31 Ibid., p. 52. 32 Ibid., p. 52. 33 Hoyt 2/26/85, pp. 44, 46-47, . 34 Vol. IV (MPR Report), Appendix Test Numbers 102, 95, ! 81, 64, 61, 30, 22, 15, 13. l 35 Hartman 4/27/80, p. 97; 4/29/80, p. 164;
- 7/26/83, p. 2; j 3/26/80, p. 19.
36 Hartman 7/26/83, p. 1. l_ 37 Ibid., pp. 1-2. _l 38 Hartman 3/26/80, p. 27. l l 39 Ibid., p. 30. a l l y l [ k
l t KENNETH HOYT 1 I i l 40 Hartman 9/17/80, pp. 57-58. 41 Hartman 7/26/83, pp. 1-2. 42 Ibid., p. 2. 43 Vol. IV (MPR Report), Appendix K, Test Numbers 95, 81. 44 B oher 10/17/84, p. 8. l j I 45 Hoy 2/26/85, pp. 68-69. 46 Vol. V (MPR Repor t) , Appendix K, Test Numbers 102, 81, ! 64, 61, 22, 15, 13. 47 Hartman 7 26/83, p. 1. ! 48 Hoyt 2/26/8 pp. 139-140. i 49 Vol. III, Tabl 2. 50 Vol. IV (MPR Re rt), Appendix K, Test Numbers 22, 15, 14, 13, 10. 51 Ibid., Test Number 1 . j 52 Ibid., Test Numbers 64 30, 53 Vol. III, Table 1. l 54 Vol. IV (MPR Report), Appen ix K, Test Numbers 64, 30. l 1 55 Hoyt 3/27/80, p. 8. i 56 Hoyt 2/26/85, pp. 107-115, 139- 0, 57 Vol. III, Table 2. l l 58 Vol. IV (MPR Report), Fiaure IV-20. 59 Vol. III, Table 2. 60 Ibid. 1 c k
i KENNETH HOYT. i 61 Congdon.2/13/85, pp. 33-34; Coleman 2/5/85, pp. 9, 39; Weaver 3/5/85,.pp.-'31-32; Adams 3/13/85, p. 9; Conaway 2/21/85,.p. 41; A. Miller 3/20/85, pp. 100-102; i 3/21/85, pp. 20, 40-41; Faust 2/20/85, pp. 37-38, 52, 85-86; 2/19/85, pp. 135-136; i Mehler 2/28/85, pp. 61-62, 73; Logan 3/27/85; p. 76.. l 62 Zewe 2/27/85, p. 63 ;- j Adams 3/19/85, pp. 21-28; Faust 2/20/85, pp. 36-38; i Mehler.3/29/85, pp. 4-10; .; Hoyt 2/26/85, pp. 104-105. H 63 'TMI Radiation Hork Permit' Number 14328, January 11, 1979 (Tab 9). 64 Ibid., Numbers 14329, 14330, January 11, 1973, (Tab 9). 65 Vol. III, Table 4. ; 66 TMI Radiation Work Permit Numbers 14328, 14329, 14330, January 11, 1979, (Tab 9). 67 TMI-2 Control Room Log, January 11, 1979 (Tab 48). 68 Vol. IV (MPR Report) , Appendix ' K, Test Number 65. 69 Vol. III, Table 2. 70 Ibid., Table 1. 71 Vol. IV (MPR Report), Appendix K, Test Number 64. 72 Hoyt 2/26/85, pp. 107-114. 73 Vol. III, Table 4. ; 74 Ibid; TMI-2 Shift Foreman Log,-January 13, 1979, (Tab 49). 75 Kunder Informal Notes from " Ops Notes" File,' January 13, 1979 (Tab 45). 76 voi, III,' Table 2. i
f KENNETH HOYT 77 Vol. III, Tables 1 and 4. l 78 Ibid., Table 2. 79 Hoyt 2/14/85, pp. 6, 11-12. i
~
80 Ibid., p. 16. 81 Ibid., pp. 16-17. 82 Hoyt 2/26/85, pp. 13-14. 83 Ibid., pp. 13-15. - 84 Ibid., pp. 84, 94-96. 85 Ibid., pp. 90-92. 86 Ibid., pp. 94-97. 87 Ibid., pp. 127-131. 88 ( Ibid., pp. 101-102. 89
)
Ibid., pp. 125(a), 130. l l l l I i
1 a I
)
l l l TIIEODORE ILLJES began employment with Metropolitan Edison j Company on February 9, 1971. Between March of 1978 and March r I ! of 1979 he worked as a Control Room Operator assigned to j
}
TMI-2. Normally, C. F. Mell and John Kidwell were the Control J I Room Operators assigned to his shift. William Conaway was the i Shift Foreman and Joseph Chwastyk was his Shift Supervisor. i Analysis of employee testimony, leak rate test records, and related plant documents indicates that Illjes did not properly utilize leak rate tests for the purpose of measuring reactor coolant system (RCS) leakage and did not take required action in response to the indications of excessive leakage. There is also strong circumstantial evidence that Illjes knowingly added hydrogen during leak rate tests in order to manipulate test results. These conclusions are based on the following information.
j i l THEODORE ILI JES .j 4 Utilization of Leak Rate Tests to Measure Leakage j i l 1 i There is substantial evidence that Illjes failed to use the J leak rate test to measure RCS leakage. Although Illjes claimed i that the leak rate test was a useful tool for approximating leakage, he also stated he did not consider the test accurate , I or reliable as a means of measuring leakage.1 It is clear from his testimony that he did not really attempt to utilize j the test for this purpose.2 Analysis of plant records pertaining to leak rate testing by Illjes provides support for this conclusion. Plant records l indicate that Illjes worked a total of 121 shifts from March 1978 to March 1979.3 Despite the policy of performing a leak rate test on every shift, records exist for only 43 leak rate tests on shifts worked by I11jes.4 l Illjes was the performer or control panel operator for. 28 of those tests. approximately forty percent of the filed tests were obviously ; curate, or were otherwise performed in violation of test procedures.5 I
-l 1
THEODORE ILLJES i Discarding-Test Results The evidence supports the conclusion that Illjes participated in the discarding of numerous unsatisfactory leak i
)
{ rate test results while assigned as a Control Room Operator at-TMI-2. 1 Illjes testified that it was standard practice on his shift' ' to perform a leak rate test at least~once a' shift. Many times l more than one test would be performed because.of the difficulty of obtaining satisfactory results.6-i k Despite his statement that leak rate test results in' excess og 1.0 gpm would sometimes. precipitate a review of plant instruments or a search' for leaks. Illjes . conceded that all test results above 1.0 gpm be would ultimately discarded q regardless of their validity.8 yg 1eaks were found and quantified, the amount of identified leakage would be reflected l in the results of the next test. That' test would be filed if it demonstrated leakage below 1.0 gpm.9 Illjes. testified that test results above 1.' O . gpm were sometimes given to his foreman to keep' him advised of plant conditions.10 However, he stated that 'either he or his i l
.. .. .. ...-J
i i i THEODORE ILLJES ( l foreman would discard all the test results above 1.0 gpm. l .- Illjes recalled that more tests were discarded than were filed.11 He believed, but could not be certain, that his l i Shift Supervisor was aware of the practice.12 ! 4 Pa tterns of test records for shifts worked by. Illjes are consistent with his testimony. The most consistent patterns of shifts worked by Illjes where no tests were filed are in January 1979 and between March 15 and 28, 1979. MPR's analysis shows that during these periods leakage was high and other ] I factors were present that would have made it difficult to I obtain satisfactory leak rate test result's, i l Filing Inaccurate Test Results l I 1 In addition to his involvement in discarding all unsatisfactory test results, it is clear from the evidence that 1 Illjes participated in the filing of numerous satisfactory test ! results without any real concern for their validity.14 Illjes confirmed that it was the practice on his shift to file all tests that yielded results under 1.0 gpm even if other plant instruments indicated that the results were clearly inaccurate.15 He stated that this practice was followed in order to keep advancing the 72-hour period during which a
j j i THEODORE ILLJES' satisfactory test result was needed.16 Test validity made no difference because he-pointed out that he probably believed all tests were inaccurate at that time.17 Our examination of the 28 test.results filed by Illjes, or filed when he was assigned to the control' room panel, revealed I that over half of them should not have been accepted as valid. i Several of those tests contain obviously. inaccurate information on the test' sheet, were performed during non-steady- state conditions, or were otherwise conducted contrary. to ' test procedures.18 Nearly one fourth of the filed tests recorded large negative values for unidentified leakage.19 Five filed tests were performed inaccurate
+
with an makeup tank level .( 4 transmitter supplying data to the computer.20
)
When confronted with these inaccurate test results, Illjes generally testified that they must ' have been accepted because it was the practice at that time to file any test demonstrating satisfactory results. He stated that he had no specific recollection of the faulty makeup tank level ' instrument and he I denied that he ever intentionally switched to the more erratic l level transmitter in order to enhance the possibility of I obtaining satisfactory test results.21 l I
J l
' THEODORE ILLJES' There is some evidence to suggest- Illjes. knowingly performed leak rate tests using the f aulty' . makeup tank level transmitter.
Between December 5, 1978 and January 10, 1979, ten tests ] filed on Illjes' shift were performed with the inaccurate transmitter.22 Five of these tests were i makeup tank level performed by 'Illjes or were performed when he was assigned to l the control room panel. l There is testimony that the makeup tank -level f instrumentation problem. was well known in the Operations 1 1 Department.24 This testimony is . corroborated by . plant-records that demonstrate an awareness- of the problem, particularly on the shift to which Illjes was assigned. .During 1 the relevant period, Illjes' shift submitted two work requests , i to repair the level ' transmitter and filed a Shift Supervisor j turnover note indicating that the transmitter was out-of-service.25 Despite this, Illjes maintained in his testimony that' he i did not recall ever . receiving instructions. to avoid . using the inaccurate transmitter while running leak rate tests.26' He acknowledged, however, that the transmitter should ~ have been declared out-of-service, and that leak rate tests should only' l l i
1 1 l THEODORE ILIJES l l l l i have been performed using the accurate transmitter.27 Our j l j review of the makeup tank level strip charts pertaining to l filed tests on his shifts did not reveal any case in which the accurate trasmitter was used.28 ] l
.l
]
Additionally, it appears that on December 5, 1978, the j inaccurate transmitter was switched to the computer l
'l specifically for the purpose of running a leak rate test.29 j l
Illjes was assigned to the panel on that date and he conceded i 1 that he would have been the person most likely to have switched ! the transmitter. I l l Manipulation of Test Results , The NRC originally alleged that there were three improper hydrogen additions and six improper water additions that were made during leat rate tests performed by Illjes or when he was s , assigned to the control room panel.31 We have concluded that two of the alleged water additions were actually hydrogen additions.32 We have also determined that a logged water addition on March 14, 1979, influenced the outcome of a leak rate test run by Illjes.33
I 1 THEODORE ILLJES . After examining the records pertaining to these tests and the :related testimony, we have concluded that there is substantial evidence that suggests that 111jes added hydrogen -l q to the makeup tank for the purpose of manipulating leak rate i test resulti There is insufficient evidence to conclude that i he added water to the. makeup tank for that purpose. Illjes denied that he ever added hydrogen to the makeup tank for the purpose of manipulating leak rate test results. Moreover, he testified that he was not even aware that hydrogen additions to the makeup tank could change indicated makeup tank level or influence the test results.34 His testimony is inconsistent with the testimony of . his former Shift Supervisor, Joseph Chwastyk. ') i On April 24, 1984, Joseph Chwastyk was interviewed regarding his knowledge of improper leak rate testing practices at 'IMI-1. In discussing the addition of hydrogen for purposes of test manipulation, he stated that he was not sure when he first learned of hydrogen's effect on leak rate tests, but he believed that it was some time in the latter half of 1978 while at TMI-2. He stated that he was advised. that a satisfactory test result would occur when hydrogen was added to the makeup tank during a leak rate test.36 When asked who :
1 l l l l 4 THEODORE ILLJES j i brought this to his attention Chwastyk stated, "I think I { remember, specifically, it was Ted Illjes."37 Chwastyk i stated that as a reault of this conversation, a test was conducted during which he personally observed that a hydrogen I I addition did have an effect on indicated makeup tank level and d 1 therefore on the leak rate test results.38 As a result, Chwastyk remembers cautioning all his operators not to add I i hydrogen while conducting leak rate tests.39 i l Plant documents indicate that this experiment was conducted on February 19, 1979.40 1 il 1 An examination of leak rate test data for that date reveals I I that two leak rate tests were performed and filed one half hour apart while Illjes and Chwastyk were working.41 Because of j perturbations in the makeup tank level strip chart traces for 1 these tests, the NRC has alleged that there were jogged water additions made during both tests. As a result of our examination of plant documents, we believe tW t .e alleged water additions were in fact hydrogen addition . Both tests of February 19 have unidentified leakage results below 1.0 gpm. An analysis of plant leak rate test data indicates that it was highly unlikely that two satisfactory tests would have been performed on the same shift.42 l4 1 THEODORE ILI.JES
,q I
The . Shif t Supervisor turnover notes for February 19, 1979, j l -written by Chwastyk, contain the following, "At 60" level in I
-l MUT & 5 -
6 psig H 2 Overpressure we get good LR."4 The i clear implication of this is that the experimental ' hydrogen - j addition was made during the course of the second leak rate test on February 19, 1979. A review of the makeup tank level J strip chart reveals an elevation in the trace close to the 60 -- inch level at a point corresponding to the running of the test.44 In light of the foregoing, we have concluded that 1 the offsets in the strip chart for that date were caused by hydrogen additions. In both tests conducted on February 19, Illjes was the CRO controlling the panel.45 He, therefore, would have been responsible for adding hydrogen to the makeup tank. Before confronting Illjes with the leak rate tests' of February 19, he was shown data regarding a test performed on I February 17.46 During that test, he was both the test l 1 performer and the operator assigned to the control room panel. MPR found that a hydrogen addition occurred shortly before the end of that test and that it had an effect on the test result.47 I THEODORE ILLJES After' observing the strip chart, Illjes ' conceded' there was a rise at: the point hydrogen appeared to have 'been: added.'48 However he continued to maintain that he had no knowledge of.' hydrogen's effect at that time.49 The next . successful tests performed by'that shift were the tests ' of February 19, which were also shown to .Illjes. He acknowledged .that there were offsets in makeup tank level. trace, and he also conceded that the filing of two tests below 1.0 gpm on one shift was an unusual occurrence. Af ter reading Chwastyk's turnover notes of February 19 that referred to the effect of hydrogen on a leak rate test, Illjes )
'l admitted that the note apparently referred to the second leak l rate test conducted by his shift on that day.51' However, he-insisted that Chwastyk simply may have been referring to the fact that a certain makeup tank level and pressure was needed for the performance of a satisfactory leak rate test and was not necessarily referring to adding hydrogen during the course of the test.52 He emphatically denied telling Chwastyk that hydrogen could affect the results of leak' rate tests.53 :
He conceded that if' an experiment, such as the one mentioned by Chwastyk, had been conducted, he would have been i the most likely person to have added the hydrogen during the l
THEODORE ILLJES. _i course of the test.54 Illjes also' denied being told..by Chwastyk or anyone else not to~ add hydrogen during the course-of leak rate tests.55 ) Given the number and the pattern of possible- hydrogen additions, and the testimony of . Chwastyk , it is probable that Illjes was aware of the effect of hydrogen additions on makeup tank level and employed hydrogen additions -to manipulate test I results. There is far less evidence to suggest . similar test i manipulation on his part by adding water during'the tests. ] l l 1 We were able to confirm only th'ree of the 12 NRC-alleged unlogged water ' additions on tests from Illjes' shift.56 When usked to review the data for the test during which he operated the panel, Illjes conceded that there appeared to be a water l addition during the course of the test.57 - He - said that he could not remember why he made the addition, but speculated that he must have been unaware that the leak rate test was-in progress and he must have added water to make an' adjustment in 8 rod position. The leak rate test performed by Illjes on March 14, 1979, included two recorded water additions indicating a 500 . gallon j addition to the makeup tank.59 However, the makeup tank 700 gallon addition. strip chart indicated approximately a _ 13 - i
l l l THEODORE ILIJES d l l l After a review of the logs, Illjes stated that it was possible that water was added in order to adjust control rod i i l position.60 He pointed out that it was his practice to make- f every effort to avoid adding water during the course of leak 1 rate tests. If an addition were necessary, he would be sure to l l
. 61 i account for the addition in the leak rate calculation. j i l l I It is not possible now to determine the necessity of the I water additions in question. In any event, given the small number of tests involved and the lack of any testimony linking Illjes to this type of manipulation, we cannot conclude that he added water for the purpose of falsifying test results.
-l Qperator Repsonse to Excessive Leakage It is clear that there were at least two periods when Illjes was working when there were indications that unidentified leakage exceeded 1.0 gpm and the required action was not taken.
Illjes stated that he understood that it was his responsibility to monitor plant leakage by the performance of leak rate tests. However, for reasons he could not clearly explain, Illjes stated that he believed he was not required to , shut down the plant as a result of leak rate tests showing l l l l THEODORE ILlJES i l l ) leakage in excess of 1.0 gpm, unless he was not able to obtain l I
- j. a satisfactory result within 72 hours fron the last satisfactory test.0 Ile stated that if an unsatisfactory l
leak rate test were obtained within the 72-hour period, the results would be used as a tool to analyze leaknge and steps !, I would be taken to remedy the leakage problem.64 However, if j 1 this could not be done, Illjes believed no further steps were I 4 i necessary to implement the action statement unless a ! satisfactory test result could not be obtained within 72 l J hours.65 l l l Illjes had no specific recollection of ' the reason why he had adopted such an interpretation of the Technical Specifications.66 He stated that it must have resulted from j the following:
. He did not believe it was his responsibility to interpret the Technical Specifications and, as a Conrol 1
Room Operator, he did what he understood his Shift 1 Foreman wanted him to do.67 ' l l l He did not recall anyone telling him to do anything different.68 I I l 1 1 1
l THEODORE ILIJES : l
'l
.- Leak rate test results were generally inconsistent and
{ not considered to be accurate.69 l l December 26, 1978 - January 15, 1979 0 During this period, only four shifts out of 13 on which Illjes worked produced filed leak rate test results. O MPR's i analysis of leakage reveals that other plant instrumentation l l' such as the sump level and makeup tank level trend demonstrated 1 unidentified leakage in excess of 1.0 gpm. Illjes had no specific recollection of this period of j l 2 time. However, he stated that an increase in sump l collection would have caused him to perform a leak rate test. Plant documentation indicating that unidentified leakage exceeded 1.0 gpm was reviewed with Illjes. Although he had no specific recollection of the events, Illjes conceded that test results might have been discarded, even if other plant instrumentation indicated excessive unidentified leakage.74 He also admitted that it was possible that leak rate test gesults would be' discarded in a situation where a leak could be identified but not quantified.75
1 1 J THEODORE ILIJES I i
)
I March 15 - 28, 1979 j MPR's analysis of this period indicates there was an increase in unidentified leakage in March 1979, which 4 l approached close to 1.0 gpm on March 16, and remained fairly l l constant thereafter.76 However, as - discussed in Volume I, . ! 1 operators would have perceived unidentified leakage to be much ! greater. l l An examination of leak rate test data for shifts on which Illjes was working during this petiod reveals 11 consecutive assigned shifts where leak rate tests were not filed. I Given Illjes testimony regarding the practice of performing leak rate tests on every shift and discarding all tests above 1.0 gpm, it is likely that numerous test results above 1.0 gpm i obtained during this period were discarded and appropriate ! I action was not taken in response to such results. 0 4 Tnere does not appear to have been any information available at that time that would have justified either the discarding of unsatisfactory results or the failure to take j required action ! in response to such results. This is ' particularly true cince during this period the operators were manually re-calculating the test results generated by the j j j
.a
q I i I THEODORE ILIJES l ( computer to correct for a calculation error in the computer program. Therefore, the Operators should have perceived the j recalculated results for unidentified leakage to be more l accurate and reliable. l 1 1 l l l l l I l'
\
THEODORE ILLJES 1 Illjes 2/7/85,.pp. 10-11, 29-31, 35, 41, 43; 2/11/85, p. 19. j 2 Illjes 2/7/85, pp. 17, 27, 30, 41; 2/11/85, p ., 23. 3 Vol. III, Table 2. 4 Ibid. l 5 Ibid. 6 Illjes 2/7/85, pp. 8-11. 7 Ibid., pp. 11-12. 8 111jes 2/7/85, pp. 12, 14, 16, 33, 39, 51-52, 78; l j 2/11/05, p. 13. 9 Illjes 2/7/85, pp. 13, 16-17. 10 Ibid., p. 14. 11 Ibid., pp. 13-16, 37, 45. 12 Ibid., p. 46 l 13 Vol. IV (MPR Report), Section IV. 14 Vol. III, Table 2. 15 111jes 2/7/85, pp. 17, 33-34; 2/11/85, pp. 17-19. 16 Illjes 2/7/85, pp. 33-34, 114. 17 Ibid., pp. 30-31, 41, 43. 18 Vol. III, Table 2. 19 Ildo-20 Vol. IV (MPR Report), Appendix K, Test Numbers 127, 111, 102, 68, 67; Figure V-5. [ corrected 1/17/86] j _ _ __ _ ____ _ ___ _ i
THEODORE ILLJES 1 Illjes 2/7/85, pp. 10-11, 29-31, 35, 41, 43; 2/11/85, p. 19. 2 Illjes 2/7/85, pp. 17, 27, 30, 41. ; 2/11/85, p. 13. 3 1. III, Table 2. 4 Idi , s 5 Ibid. 6 Illjes-2 /85, pp. 8-11. l 7 Ibid., pp. 1-12. 8 Illjes 2/7/85, pp. 12, 14, 16, 33, 39, 51-52, 78; ! 2/11/85, . 13. 9 Illjes 2/7/85, pp. 13, 16-17. 10 Ibid., p. 14. I 11 Ibid., pp. 13-16, 37, 4 . j 12 Ibid., p. 14. l 13 Vol. IV (MPR Report), Sectio 1 IV. l 14 Vol. III, Table 2.
\
15 Illjes 2/7/85, pp. 17, 33-34; 2/11/85, pp. 17-19. 16 Illjes 2/7/85, pp. 33-34, 114. i 17 Ibid., pp. 30-31, 41, 43. 18 Vol. III, Table 2. 19 Ibid. 20 Vol. IV (MPR Report), Appendix K, Test Numbe s 127, 111, 102, 68, 67; Figure V-5. i
THEODORE ILI.JES , a i
.21 Illjes 2/7/85, p. 87. j 22 Vol. IV (MPR Report), Appendix K, Test Numbers 111, 102, 100, 89, 87, 82, 71, 68,,67, 66.
23 Ibid., Test-Numbers 127, 111, 102, 68, 67. ) i 24 Adams 3/8/85, pp. 47-53; H 3/13/85, pp. 140-141; - Congdon 3/13/85, p. 49; Faust 2/19/85, p. 16; j! 2/20/85, pp. 18-21; j Frederick 3/14/85, pp. 66-67, 71, 103-105; i Guthrie 2/12/85, pp. 16-17,.61-62; McGovern 2/6/85, pp. .34-37, 70-71,'86; Miller, A. 3/20/85, pp. 86-87. 1 25 TMI-2 Job. Tickets: C0013, December 5, 1978, C0345, December 25, 1978, (Tab 7); ) Shift Supervisor Turnover Notes, December 20, 1978,
)
(Tab 10). , i 26 111jes 2/7/85, p. 90. J 27 Ibid., pp. 80, 88. i 28 Vol. III, Table 2.
)
1 29 Vol. IV (MPR Report), Appendix K, Test Number 111. 30 111jes 2/7/85, p. 93. 31 Vol. IV (MPR Report), Appendix K, Test Numbers 130, . 102, 89, 67, 43, 39, 35, 34, 33 (See also Sections VI ! and VII). t 32 Ibid., Tast Numbers 34, 33. 33 Ibid., Test Number 11.- 34 Illjes 2/7/85, pp. 101, 109, 112, 121. 35 Chwastyk 4/24/84, p. 26. 36 Ibid., p. 27. i l I THEODORE ILIJES { 37 Chwastyk 4/24/84, p. 27. 38 Ibid., pp. 27, 35. 39 Ibid., p. 35. l 40 Shift Supervisor Turnover Notes, February- 19, 1979, (Tab 10); l J Vol. IV (MPR Report), Appendix K, Test Number 33. j l 41 Ibid., 'Iest Numbers 34, 33. 42 Vol. III, Table 1. 43 Shift Supervisor Turnover Notes, Feburary 19, 1979, (Tab 10). 44 Vol. IV (MPR Report), Appendix K, Test Number 33. 45 Vol. III, Table 2. 46 Vol. IV (MPR Report), Appendix K, i Test Number 35.
]
47 Ibid., Table VII-1. 48 Illjes 2/7/85, p. 111. 49 Ibid., p. 112. ! 50 Ibid., pp. 115, 118. l I 51 Ibid., p 120; Shift Supervisor Turnover Notes, February 19, 1979, j (Tab 10). ! 52 Illjes 2/7/85, p. 121. 53 Ibid., p. 120. 54 Illjes 2/11/85, p, 42. 55 111jes 2/7/85, pp. 102, 104. 56 Vol. III, Table 2; Vol. IV (MPR Report), Appendix K, Test Numbers 87, 44, 43.
THEODORE ILLJES 57 Illjes 2/7/85, p. 122. 58 DM. , p . 122. 59 voi, Iy .(MPR Report), Appendix K, Test Number 11. 60 Illjes 2/7/85, p. 126. 61 Ibid , pp. 72-73. 62 Ibid., pp. 27-28, 63 Ibid., pp. 29-30, 33, 57-58. 64 Ibid., p. 29-30 65 Ibid., pp. 29, 57-58. 66 Ibid., p. 35. 67 Ibid., pp. 34, 50-52. 68 Ibid., p. 50. 69 Ibid., pp. 29-30. 70 Vol. III, Table 2; Vol. IV (MPR Report), Appendix K, Test Numbers 71, 68, 67, 66. 71 Ibid., Figure IV-20. 72 Illjes 2/11/85, p. 11. 73 Ibid., pp. 10-28. 74 Ibid., p. 27. 75 Ibid., pp. 12-14. 76 Vol. IV (MPR Report), Figure IV-20. 77 Vol. III, Table 2. 78 111jes 2/7/85, pp. 8, 10, 12-13, 15-16. [ corrected 1/17/86) l
1 THEODORE ILI.JES
)
7 Illjes 2/7/8.1, p. 122. 58 Ibid., p. 122. I 59 Vol. IV (MPR Report), Appendix K, Test Number 11. 60 I ljes 2/7/85, p. 126. I I 61 Ibi ., pp. 72-73. I l 62 Ibid., p. 27-28. I 63 Ibid., p 29-30, 33, 57-58. l 64 Ibid., p. 2 -30 65 Ibid., pp. 29, 57-58. j 66 Ibid., p. 35. 67 Ibid., pp. 34, 50- . 68 Ibid., p. 50. 69 Ibid., pp. 29-30. l 70 Vol. III, Table 2; l Vol. IV (MPR Report), Append'x K, Test Numbers 71, 68, ) 67, 66. l 71 Ibid., Figure IV'20. 72 Illjes 2/11/85, p. 11. 73 Ibid., p. 10. 74 Ibid., p. 27. 75 Ibid., pp. 12-14. 76 Vol. IV (MPR Report), Figure IV-20. 77 Vol. III, Table 2. 78 Illjes 2/7/85, pp. 8, 10, 12-13, 15-16. t
I l l GEORGE KUNDER began employment with Metropolitan Edison Company in September 1968. He is currently the Manaaer of the TMI-2 Safety Review Group. Between March 1978 and December l 1978, he was Superintendent - Technical Support at TMI-1.1 1 In December 1978, he assumed the positions of Superintendent - Technical Support and Chairman of the Plant Operations Review Committee (PORC) at TMI-2. He held these positions until after l the accident in March 1979. At TMI-2, he reported directly to Joseph Logan, TMI-2 Superintendent.2 During the course of our investigation, we found no evidence to suqqest that Kunder was involved in, or condoned, improper leak rate testing practices at TMI-2. It does appear, however, that he participated in the decision to continue to operate TMI-2 in January 1979, despite indications of excessive unidentified leakage.3 As Superintendent - Technical support at TMI-1, Kunder had nothina to do with leak rate testina at TMI-2 durina a substantial part of the time under investigation. Once he was transferred to TMI-2 in December 1978, his involvement with leak rate testing was only indirect and minimal. He did not
l GEORGE KUNDER < l perform leak rate tests himself and he was not responsible for their oversight.4 ! l l Kunder testified that he was not aware of improper leak rate test practices, such as test manipulation by operators, at ! TMI-2.5 According to him, he did not even believe that it was possible to influence leak rate test results by adding hydrogen to the makeup tank.6 Kunder also stated that he did i not know of any pressure exerted on operators by management to improperly perform leak rate tests.
\
Although Kunder stated that he did not think that it was ! inappropriate to throw away leak rate test results under some circumstances, he denied knowing that tests were being discarded at TMI-2 in 1978-1979.8 He also said that he did not know of any decision to conceal unsatisfactory test results by failing to enter them into the Control Room Log.9 We found no evidence that Kunder was involved in, or condoned, the improper leak rate testino practices in i question. No testimony links him to leak rate test misconduct and we did not find any company records or documents that were sent to Kunder that would have put him on notice of such misconduct. i
)
1
GEORGE KUNDER The evidence does _ reveal, however, that Kunder was aware that there were serious problems with the accuracy of the leak rate test procedure at TMI-2. Logan testified that Kunder and other attendees at TMI-2 Plan of the Day (POD) meetings, on more than one occasion, discussed the difficulties in obtaining satisfactory and consistent leak rate test results. Logan further stated that to the best of his recollection, Kunder was lookina into those problems.10 Kunder acknowledged that he was aware of what he i l characterized as inconsistencies between leak rate test results and operator observations of actual plant' conditions.11 In his NRC interview, he admitted that there were complaints that the leak rate test results were higher than the leakaae perceived by Operations personnel, but he later took the position in our interview that he simply did not remember the I specifics of the problem.12 He remembered that his department was given the task of looking into the problem, but he did not get personally involved in its resolution.13 , He further testified that he has since come to learn that a I temporary change notice -(TCN) was prepared on March 16,-1979, ; to correct the ' problems with the procedure.14 Accordino to , Kunder, he does not remember much about the TCN and the l
1 I GEORGE KUNDER i discussions that led to its issuance and subsequent approval by i PORC. He also said that he did not remember being aware that the TCN failed to correct all of the calculation errors in the leak rate test procedure.15 ! Although it is clear that the TCN was indeed deficient, , I there is no evidence that Kunder was aware of that fact when it l was approved by PORC. Those individuals that we were able to i interview regarding the TCN all denied that it was intended to bring about more favorable, but biased test results.16 We found no documentary evidence to refute that testimony. Thomas Morck, the engineer who actually prepared the TCN, is no longer employed by Metropolitan Edison Company and refused to be l interviewed during our investigation. While we found nothing to suqqest that Kunder was aware of improper leak rate test performance, there is evidence thc' he failed to ensure that appropriate action was taken in response to excessive reactor coolant system (RCS) leakage in January 1979. As Superintendent - Technical Support and Chairman of PORC, Kunder had a responsibility to ensure that TMI-2 was operated in compliance with the Technical Specifications.17
-4 -
)
GEORGE KUNDER By the time he was transferred to TMI-2, LER 78-62, which acknowledged a. failure by TMI-2 operators to enter the action statement in response . to excessive RCS leakage, had already been i~ssued. PORC had a responsibility to. follow up with the corrective action that had- been promised in ' the LER. Despite
'this, nearly everyone who was questioned about this LER has stated that they did not recall receivina- instructions regarding its contents. The evidence shows that improper leak rate testing practices were not discontinued as a result of'. the O Nevertheless, company records reveal that on January LER.
l 10, 1979, Kunder, as PORC Chairman, reported to the General l Office Review Board that PORC had reviewed the LER and was satisfied with the corrective action taken.19 PORC records. reveal, however, that the PORC review was not completed until March 5, 1979.20 When confronted with these facts, Kunder essentially testified that he could not remember the LER. He also said that he could not tell, from his review of. the documents, exactly what had been done to take the corrective action required. by the LER, or what his involvement had been.22 Kunder explained that ordinarily PORC would not conduct an inquiry to determine if the required action had in fact occurred.23 GEORGE KUNDER l Despite .Kunder's inability to recall, the evidence ! l l indicates that he participated in a review of the follow up to the LER and that he was responsible for the closecut of the ! action item. These events occurred at the same time that he became aware of complaints about the accuracy of leak rate test' results.24 Thus, in early 1979, he was not only aware of the difficulty the operators were havino in obtaining satisfactory leak rates, but he also knew that, in the past, operators had not taken required action in response to unsatisfactory leak rate test results. Kunder testified that he did ' not draw any connections between these two pieces of information.25 There is evidence that Kunder was also aware of excessive j RCS leakage in January 1979. MPR's analysis shows that around January 3, unidentified leakage exceeded 1.0 gpm and remained above 1.0 qpm until January 15, when the plant was shut down. There were apparently several leaks that were contributing to this situation. Among them were leaks from pressurizer instrumentation valves.26 TMI-2 Plan of the Day (POD) i supplemental work lists suqqest that these' pressurizer leaks ! were identified by January 2, and that plans were beino made to attempt to repair the valves.27 Kunder stated that he remembered there was some leakage from the pressurizer root valves, but he couldn't remember a specific time frame.28 He
/
GEORGE KUNDER j acknowledged, however, that it is likely that he would have , been aware of these leaks if they were discussed at . POD meetinos.29 On January 11, Radiation Work Permits were issued for a ( Shift Foreman, a Health Physics Technician and a Maintenance Foreman to enter the reactor building inside the "D Ring" to. .a check for leaks at the pressurizer. Two other permits were j issued for routine entry and an inspection 'o n the -same O d :.y . Sump collection was on the increase 'and a considerable amount of water was added to the makeup tank durina this shift.31 Plant rbcords also reveal that employees inspected the leakage from the pressurizer valves, but there is no record of the amount of the leakage that they ! observed.32 Althouah no satisfactory leak rate test results were obtained, there is no record of an official entry into the action statement.33 Puring the next two days there were only two satisfactory .. leak rate tests obtained, both of which were influenced by water additions during the test.34 On January 13 at 12:45, Kenneth Hoyt and John McGarry entered the reactor buildina to check for leaks at the pressurizer. (- , l L l i l GEORGE KUNDER Hoyt's log entry regarding this activity reads as follows: Made inspection in Rx. bldg. to check leakage from RC-V-133B, 134B, 135B. Total leakage from these . valves- is approximately the same as Thursdays -[l/11] inspection. Survey . of other areas showed no other leakaae. Leakage is from body to. bonnet areas of these va and appears to be safe yet to operate.ges _ Hoyt . testified that he remembered entering the reactor building to very quickly observe a steam leak comina from the pressurizer and estimating the leakage to be approximately. 0.5 apm.36 He conceded that he could not accurately measure the l 4 leakaae and that he only provided a very rough estimate of the-leakage because that is all he was probably asked to do by his Shift Supervisor.37 Hoyt stated however, that he could not recall specific I discussions with anyone about the leakage. He' also conceded' .) that, given his inability to quantify the leakage he observed and the information available from other sources, there was a-good possibility they were operating with leakage in excess of Ic0 qpm.38- He could not explain why they did not enter the action statement and proceed to shut down the plant.
l GEORGE KUNDER John McGarry could not shed any further light on the issue. According to him, his observation of the leaking valves would have been solely for the purpose of assessing their I structural integrity. He stated that he would not have rendered any opinion regarding whether the amount of leakage required a plant shutdown. In any event, he did not recall any. specific discussions that he may have had at the time.39 l Neither Bernard Smith, Hoyt's Shift Supervisor, nor William i Zewe, who apparently was also working during this time period, l could remember participating in any discussions regarding the possible shut down of the plant.40 Kunder also' testified that he could not recall participating in any discussions or decisions regarding a plant shutdown in January 1979.41 1 An Operations note dated January 13, 1979, contains a diagram of the leaking pressurizer valves and states the following: RC-V135V decreased. RCV133B and 134B increased. Hoyt and McGarry say not justifiable to shut down yet based on valve leak. However, leak rate on MU tank acrees with what we're seeina in Rx Bldg. c
1 GEORGE KUNDER ! It continues in a.slightly different handwriting, [Approximately) 2.0 gpm leakage ' indicated on Makeup Tank level. Since no other leakage is existing the-RC-V-133B, 134B,- and 135B leakage is identified and
'below [ illegible).42 On the top right hand corner- of the note, Unit I i
Superintendent Joseph' Logan 's phone number is written next to his name. i Kunder testified ' that all of the handwriting on the note : his.43 was He maintained that he had no ' independent recollection of the events which caused him to write it and
.]
could not recall any discussions that he might have had with l anyone regarding a plant shutdown at that time.44 Accordinq to Kunder, he could have written the note for a variety of reasons and not necessarily because he was responsible for recommending the proper course of action to Logan.4 Logan has testified that he does not recall any discussions with Kunder about shutting down the plant because of RCS leakage in January 1979. He stated, however, that based upon his relationship with Kunder, it was likely that Kunder was assigned to review the leakage . problem and provide him with a report.46- Although he did not remember any discussion, Logan-i GEORGE KUNDER stated that he obviously must have been satisfied wit h Kunder's report.47 In addition to his inability to remember any discussion regarding the decision to continue operating the plant in i January 1979, Kunder was also unable to shed much light on the meaning of the note that he prepared at that time. He i conceded, however, that it would not have been correct to quantify leakage for leak rate test purposes by the method described in the note.40 In our view, the note and the surrounding circumstances indicate that Kunder was aware of a leakage problem that should have caused action to be taken as required by Technical Specifications. It is clear that the required action was not taken and neither Kunder nor any other witness has offered an adequate explanation. GEORGE KUNDER 1 Kunder 9/28/83, p.'4. 2 Ibid., p. 7. 3 Miller, A. 3/22/85, pp. 68-71; Mehler 3/29/85, pp. 13-16; g Logan 3/27/85, pp. 106-107. ' 4 Kunder 9/28/83, pp. 11-13; Letter, G. Kunder to N. Palladino, August 13, 1984, (Tab 38). 5 Kunder 9/28/83, pp. 38-40. 6 Ibid., p. 38. l 7 Ibid., p. 41. 8 Ibid., pp. 24-25. 9 Ibid., p. 28. 10 Logan 11/18/83, p. 3. 11 Kunder 3/6/85, p. 50. 12 Kunder 3/6/85, p.-51; 9/28/83, p. 14. 13 Kunder 3/6/85, p. 51. 14 Kunder 9/28/83, pp. 15-16. 15 Ibid., p. 45. 16 Frederick 12/10/82, pp. 3980, 3982, 3987; Guthrie 2/12/85, p. 69; Faust 9/25/81, pp. 874-889. l 17 Metropolitan Edison Co., Position Descriptions for Unit Superintendent Technical Support. (Tab 6). 18 Miller, A. 3/21/85, p. .71; Hoyt'2/26/85, pp. 59-60; Adams 3/13/85, p. 85; Smith 2/8/85, pp. 106-107 [ corrected 1/17/86)
GEORGE KUNDER 1 Kunder 9/28/83, p. 4. 2 Ibid., p. 7. 3 Miller, A. 3/22/85, pp. 68-71; Mehler 3/29/85, pp. 13-16; ogan 3/27/85, pp. 68-69. 4 Ku der 9/28/83, pp. 11-13; Lett , G. Kunder to N. Palladino, August 13, 1984, (Tab ). 4 5 Kunder 9 8/83, pp. 38-40. G Ibid., p. 4 . 7 Ibid., p. 41. -l 8 Ibid., pp. 24-25 9 Ibid., p. 28. 10 Logan 11/18/83, p. 3. 11 Kunder 3/6/85, p. 50. , 1 12 Kunder 3/6/85, p.. 51; 9/28/83, p. 14. 13 Kunder 3/6/85, p. 51. l l 14 Kunder 9/28/83, pp. 15-16. 1 15 Ibid., p. 45. l 16 I Frederick 12/10/82, pp. 3980, 3982, 39 7; l Guthrie 2/12/85, p. 69; Faust 9/25/81, pp. 888-889; l Hoyt 2/26/85, pp. 146-147. l 17 Metropolitan Edison Co., Position Descriptio for Unit Superintendent Technical Support. (Tab 6). 18 Miller, A. 3/21/85, p. 71; Hoyt 2/26/85, pp. 59-60; Adams 3/13/85, p. 85; < Smith 2/8/85, pp. 106-107. j l 1 l ___ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ __ _ _ _ ._ )
1 l l I GEORGE KUNDER 19 Kunder 3/6/85, pp. 10-11;
. General Office Review Board (GORB) Meetina Minutes, January 10, 1979, (Tab 25). I 20 Bezilla 2/22/85, pp. 33-34; Plant Operations Review Committee (PORC) Minutes, March 5-9, 1979, (Tab 26).
21 Kunder 3/6/85, pp. 24-26. I 22 Ibid., pp. 20-21. 23 Ibid., pp. 1~7-23. 24 Kunder 3/6/85, pp. 49-50, 25 Ibid., p. 51. 26 TMI-2 Job Tickets, Numbers CO237, CO238, C0240, December 16, 1978, (Tab 7); I McGovern 2/6/85, pp. 77-82.
]
27 Kunder 3/6/85, pp. 27-28; 1 Plan of the Day (POD) Meetina Minutes, Supplemental I Work List, January 9-14, 1979, (Tab 22). l 28 Kunder 3/6/85, pp. 27-28. 29 Ibid., pp. 28-29. l 30 TMI Radiation Work Permit Numbers 14328, 14329, 14330, January ll, 1979, (Tab 9). 31 TMI-2 Control Room Log, January 11, 1979, (Tab 18) ; TMI-2 Surveillance Procedure 2301-S1, Shift and Daily-Checks, January 11, 1979, (Tab 40). l 32 vol. III, Table 4. I 33 Ibid., Table 1 and Table 4. 34 Vol. IV (MPR Report), Appendix K, Test Numbers 65, 64.
GEORGE KUNDER.
- . 35 Vol. III, Table 4.
3'6 Hoyt 2/26/85, pp. 86, 91, 93, 94. 37 Ibid., pp. 95-97. 38 Ibid., p. 127. 39 McGarry 3/1/85, pp. 32-36. 40 . Smith 2/8/85, p. 126; Zewe 2/27/85, pp. 114-115. 41 Kunder 3/6/85, pp. 34, 39. 42 Kunder G.A., Informal Notes from " Ops Notes file, January 13, 1979, (Tab 45). 43 Kunder 3/6/85, p. 31. 44 Ibid., pp. 30-34. 45 Ibid., pp. 35, 39-40. 46 Logan 3/27/85, pp. 106-107. 47 Ibid., p. 107 48 Kunder 3/6/85, pp. 44-47.
.i
, , _ .)
[ corrected 1/17/86]
i il J. I GEORGE KUNDER l Vol. III, Table 4'. 36' Hoyt 2/26/85, pp. 86, 91, 93, 94. ;i 37 Ibid.,?pp. 95-97. . 38' id., p. 127. j 39 McG rry'3/1/85,.pp. 32-36. 40 Smith /8/85, p. 126; Zewe 2 7/85, pp. 114-115. 41 Kunder 3/ 85,.pp. 34, 39. 42 Kunder.G.A., nformal Notes from " Ops Notes file, January 13, 1 9, (Tab 45).- 43 Kunder 3/6/85, p. 31. 44 Ibid., p. 30. 45 Ibid., pp. 35, 39-40. 46 Logan 3/27/85, pp. 106-1 .
.. j 47 Ibid., p. 107.
48 Ibid., pp. 44-45, i I
-)
1 j
'l 1
i _ . _ . . _._ _ 1___ __ --_1------- - - - - - - - -
WALTER J. MARSHALL began employment as an Auxiliary Engineer with Metropolitan Edison Company in February 1977. He is presently employed at TMI-2 as an Operations Engineer. Between March 1978 and March 1979, Marshall was assigned as an Operations Engineer at TMI-2. Marshall functioned primarily as l the assistant to James Floyd, the TMI-2 Supervisor of Operations, and reported directly to him. Marshall held a Senior Reactor Operator's (SRO) license since 1978 and was familiar with the Technical Specification limits for reactor coolant system leakage.1 He stated he was also familiar with the leak rate test procedure.2 Marshall believed that it was the practice to perform a test each day, and did not know of any~ problems operators were experiencing in obtainina satisfactory leak rate test results.3 When questioned generally about his knowledge of leak rate test practices, Marshall testified that in 1978-1979 he was not aware of improper conduct such as discarding leak rate test results or manipulating tests by adding hydrogen or water to the makeup tank.4 E_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ - - - - - - _ - -
WALTER J. MARSHALL Marshall did not recall what action was required when leak rate test results showed unidentified leakage above 1.0 gpm. 5 He - pointed out that he did not regard it to be his i responsibility to ensure' compliance with Technical Specifications and leak rate < test procedures.6 Marshall was questioned. about the extent of 'his contact-with the control room while performino his duties. He _ .j 1 testified that he received verbal instructions daily from Floyd-about his activities.7 However, he regarded his primary , i responsibility to be supervising the operation of the
" polishers" (condensate polishers) located in the turbine-building basement. He estimated' he would spend 1
approximately 8 hours a day attendina to this duty. This was l particularly true between September 1978 and February 1979.9 Consequently, Marshall believed that he spent only about. 10 ) percent of his time in the control room.10 Marshall testified that when Floyd was absent, he would ~! take over- his routine duties. However, :Shif t Supervisors did ; not report to him concerning t' heir operation of the plant.11 ' Ma'rshall also stated he was not responsible for reviewing Shift Supervisor turnover notes.12 i I r _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ -- WALTER J. MARSHALL Marshall was assigned to review the Shift Foreman Logs once l a week, but only to insure that they complied with the administrative procedures on looging practices.13 Marshall would not normally review Control Room Logs. That duty was left primarily to Floyd.14 During his review of Shift Foreman Logs, Marshall did not pay a great deal of attention to leak rate test results.15 If he reviewed a Shift Foreman's Log and did not find any leak j rate test result, he would not have been concerned, and would i not have brought it to Floyd's attention despite his belief that a test was to be performed each day.16 Marshall pointed out that occasionally he would function as eaul rly ass gne Sh t Foreman. He did not recall any lea rate tests being performed on those shifts.17 Our review of leak rate test records did not reveal any tests approved by Marshall as a Shift Foreman.10 From time to time, Marshall would attend Plan of the Day (POD) meetings in Floyd's absence or to report on his own activities.19 Although part of the POD meeting was devoted j to a discussion of leak rate test results, he did not recall l 3- t i
WALTER J. MARSHALL any discussion of difficulty in obtaining satisfactory test results.20 1 1 I Marshall testified he had. recollection of' the ' circumstances' ! that led to the issuance of LER 78-62. He stated that he I overheard a conversation between Joseph Chwastyk. and perhaps Floyd, in which they discussed "whether they should be on a. four hour time clock or not" because -of a leak rate test I result.21 Marshall believes that -he knew the LER was connected with the subject matter of their conversation. Although he has no present recollection of being aware of the {
.LER's contents, he testified that he most likely had been aware i
of them since he signed the " sign-off sheet" and could tell from documents shown to him during the interview that he had. I been responsible for the followup action on the LER. However, Marshall had no recollection of being assigned the task of following up on the LER as a PORC action item. 2 j It appeared' to Marshall from the documents concerning the action item that the only thina he was required to do with the LER was to send it to the control room'and document that-the operators had read it.23 Since it took' five months for PORC' i to close the action item, Marshall was asked to explain the t delay. He testified that he had no recollection of the reason-
-4 -
l WALTER J. MARSHALL
.for the' delay, and does not recall being concerned about it.
He did not recall .anyone asking him ' why it' 'was 'taking so long.24' i Seelinger's letter- to the NRC of August 9, 1984, was l 1 reviewed with Marshall durino our interview. In this' letter,. I Seelinger asserted that he. had indeed advised- Operations personnel of the contents of the. LER. 5 Marshall, however, had no recollection of any meetings or discussions about the LER. 6
. I I
l
WALTER J. HARSHALL 1 Marshall 3/4/85, p. 6. l 2 Ibid., p. 7.
\
l 3 Ibid., p. 17. 4 Ibid., p. 27. 5 Ibid., p. 7. 6 Ibid., p. 27. 7 Ibid., pp. 4, 8. 8 Ibid., pp. 5, 19. J 9 Ibid., pp. 27. 10 Ibid., p. 5. $ 11 Ibid., pp. 19-20. l 12 Ibid., p. 19. l 13 Ibid., pp. 5-6. 14 Ibid., p. 10. , i 15 Ibid., pp. 23-24. 16 Ibid., p. 24. ' 17 Ibid., p. 14. 18 Vol. III, Table 1. 19 Marshall, 3/4/85, p. 15. 20 Ibid., p. 16. 21 Ibid., pp. 18-19. 22 Ibid., pp. 37-40. 23 Ibid., pp. 36-37. 24 Ibid., p. 43. t
) ; }
WALTER J. MARSHALL 35~ Letter, J. Seelinger to N. Palltidino (NRC), August 9, i 1984, (Tab 36). 26 Marshall 3/4/85, pp. 49-50. A I [ corrected 1/17/86] f l l
x WALTER J. MARSHALL 2 Letter, J. Seelinger to N. 1984, (Tab 36). Palladino (NRC), August 9, 26 Ibid., pp. 49-50.
\
s ( N
l t } HUGH MCGOVERN began employment with Metropolitan Edison Company on October 11, 1976. Between January of 1978 and November of 1978 he worked as a Control Room Operator (CRO) Trainee assigned to TMI-2. He became a licensed CRO in December of 1978 and remained assigned to TMf-2 in that capacity until the accident in March of 1979. For the most part, he worked on shift with CRO's Edward Frederick, Craig Faust, and Shift Foreman, Patrick Lydon while he was in training. Upon becomino licensed, he was assioned to work primarily with Earl Hemmila and Leonard Germer under the supervision of Shift Foreman, Carl Guthrie. Analysis of employee testimony, leak rate test records, and related plant documents suqqests that McGovern failed to properly utilize Reactor Coolant Inventory Balance tests (leak rate tests) for the purpose of measuring reactor coolant system leakage and also failed to take required action in response to the indications of excessive reactor coolant system leakage in March of 1979. 1 l
- _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ i
1 I HUGH MCGOVERN There is not sufficient evidence to support a conclusion that McGovern knowingly added water or hydrogen during leak i rate tests in order to manipulate test results. These conclusions are based on the following information. Utilization of Leak Rate Tests to Measure Leakage i There is substantial evidence that McGovern failed to properly employ the leak rate test for the purpose of measuring l reactor coolant system leakage. I McGovern essentially admitted that he did not really use , the test for this purpose. He testified that it was not a very accurate or reliable method of quantifying reactor coolant system leakage.1 According to him, leak rate tests performed with the computer often produced results that were unbelievable or inconsistent with readings from other plant instruments.2 Consequently, he viewed the leak rate test as a " paper" or administrative requirement that routinely had to be satisfied on a periodic basis.3 His recollection was that tests were run as often as two or three times per shift.4 The satisfactory results were turned over to the Shift Foreman for filino and the unsatisfactory rescits, showing unidentified 2-8 _--__________-__-_-_a
HUGH MCGOVERN leakage in excess of 1.0 gpm, were thrown away without further documentation.5 There was no real effort to verify the ; accuracy of the good or the bad results.6 According to McGovern, he probably threw away more tests than he filed. McGovern further testified that he believed all CRO's, Shift Foremen and Shift Supervisors at TMI-2 followed this same leak rate testing practice. He did not know how it originated and he did not recall that anyone was actually ever pressured or directed to treat leak rate tests in this manner. He just remembered that it had always been the practice to conduct leak j i rate testing in this way and that no one ever really questioned j it. For the most part, McGovern attributed this to a ceneral laxity and casual attitude about administrative requirements that existed at that time.8 He explained, however, that despite this attitude, he did not believe that the plant was being operated unsafely or with excessive leakage. McGovern testified that he continuously evaluated leakage by visually monitoring makeup tank level trends and other plant i parameters. ! He also said that inspections were routinely l 1 conducted and leakage was often identified, quantified, and controlled.10 3-
HUGH MCGOVERN he conceded that he was not able to accurately While system leakage with any degree of measure reactor coolant certainty, he felt that these unofficial methods allowed him to approximate or "ballpark" leakage and to recognize significant he always felt that increases.ll McGovern stated that leakage was under control and that he was therefore safely operating the plant. He said he believed that unidentified leakage generally was limited to under one gallon per minute, although he now realized that he was not able to conclusively demonstrate that fact during the period in question.12 Testing practices for shifts worked by McGovern seem to be consistent with his testimony. There were no filed tests showing leakage in excess of 1.0 gpm and there were many shifts in which no leak rate test result was filed.13 TMI-2 records indicate that between March of 1978 and March of 1979, McGovern reported for work as a Control Room Operator on approximately 85 shifts during which leak rate tests presumably could have been performed.14 Despite this fact, only 24 leak rate tests were filed for shifts on which McGovern was working.15 McGovern signed as the performer of 12 of these tests and was responsible for operating the control room l panel and maintaining the Control Room Loq during most of the
-4 -
HUGH MCGOVERN ( others.16 Although there was a policy of attempting to file a leak rate test at least once per shift, tests were actually 1 filed on less than one-third of the shifts that he was workina.I' Aside from the fact that no tests were filed showing leakage in excess of 1.0 ppm, there are no obvious patterns associated with leak rate testing done on shifts worked by McGovern. The records do not suggest a preference for filing tests during any particular time of day, and there are several instances of consecutive. assioned shifts when McGovern was working when no leak rate tests were filed. These gaps in filed tests appear from March 1978 right up through the accident.10 There is only one period where h'is shift filed fairly regular test results: February 3 through February 16, 1979.19 1 During that time, eight tests were filed out of ten shifts, but many of these tests were improperly performed as will be discussed later.20 i On the other hand, there is also one period between March 1 i i and March 16, 1978, when McGovern worked twelve consecutive assioned shifts without filing a test.21 Our analysis 2 5-
HUGH MCGOVERN reveals that this was a period in which increased reactor i coolant drain tank collection would have made it very difficult to obtain a satisfactory leak rate test because of the ! calculation error in the procedure. McGovern has testified that he does not recall that there were any specific periods when it was particularly more difficult to get good leak rate test results. According to him, it was difficult to get good leak rate test results from the outset. Filina Inaccurate Test Results In addition to discarding all unsatisfactory test results that he obtained, McGovern filed, or participated in the filinq ! of, numerous satisfactory leak rate test results that were inaccurate or performed in violation of test procedures.23 Almost half of the tests that were filed on shifts that he was working should have been determined to be invalid. A number of the tests performed while McGovern was on shift contained l obviously inaccurate information on the computer printout itself.24 Three of the leak rate tests submitted by McGovern's shift involved logged water additions durina the tests that caused unidentified leakage values to be reduced because of makeup tank level instrumentation problems. 5 J
I HUGH MCGOVERN Three of four satisfactory test results filed between December 16 and December 30, 1978, involved leak rate tests ) that were performed with faulty makeup tank level instrumentation.26 When confronted with these inaccurate test results, ! McGovern generally conceded that they should not have been filed. He explained, however, that since the results satisfied J l the administrative requirements, they were accepted.27 ] I For example, McGovern initially testified that he was aware that one of the makeup tank level instruments oscillated more widely than the other for a significant period of time during Unit 2 operations. He had also testified that it would have been his practice to shift to the more accurate instrument in order to perform a leak. rate test and in order to obtain a more l accurate result. l After being shown tests in which he participated that were performed with the faulty makeup tank level transmitter, McGovern at one point indicated that he did not recall the. specific tests but that they must have been inadvertently performed with the bad transmitter because he or his colleagues were not paying close attention.29 Later in his testimony, j
I l I HUGH MCGOVERN < he conceded that he probably ran leak rate tests knowina that the makeup tank level instrumentation was inaccurate. He I explained that the instrumentation was available for use, it f satisfied the administrative requirements, and running tests in this way was accepted practice.30 In any event, he denied that he intentionally switched makeup tank level instruments durina leak rate tests in order to manipulate test.results.31 In our analysis of the several which tests were run on McGovern's shift usino the inaccurate level transmitter, we found no pattern of switching instrum'ents before a leak rate test.32 Insofar as the loqqed water additions were concerned, McGovern testified that he did not recall, nor could he reconstruct, why it was necessary to add water during the tests in question.33 i He denied that he was aware at that time that i i properly accounted-for water additions would cause leak rate test results to reflect less unidentified leakaoe.34 i l Manipulation of Test Results Our investigation did not produce sufficient evidence to support a conclusion that McGovern was involved in the ! intentional manipulation of leak rate test results. l 8-
HUGH MCGOVERN-The NRC alleoed that hydrogen was added during two leak I rate tests that were performed December 30, 1978, and March 21, 1979, while McGovern was working. The test of December 30, -1978, was performed by McGovern
-with the faulty maksup tank level transmitter.36 .The obvious i disparity between the two makeup tank level readings and the-fact that no hydrogen was loqqed makes it impossible to_ confirm whether an addition was in fact made. Our analysis of the makeup tank -level strip chart trace, however, suqqests that faulty instrumentation .would have had a much greater effect on the test result than an unloqqed ' hydrogen addition.37 The test of March 21, 1979, was performed by Hemmila and Germer while McGovern was apparently a third ' operator _ working at that time.38 The Control Room Log does show that a hydrogen addition was made during the test, but our strip chart analysis reveals that it had little _or no effect on the leak-rate test results.39 In any event, it does not appear that' .
McGovern had anything to do with this particular. test. Moreover, there is no testimony connecting McGovern to this type of test manipulation, and there is no corroborative evidence which suggests that he was involved in this kind of conduct. Under the circumstances, we cannot conclude that he 1 k; HUGH MCGOVERN-i 1 i manipulated a test result simply on the basis of a suspicious makeup tank level trace. i The NRC also alleged that improper and unlogged water additions were rade during seven of the tests that were performed while McGovern was on duty.40 Based upon our analysis of the strip charts, the alleged additions of December 24, 1978 and February 9, 1979, were actually made outside the test period, but were believed to have been made during the test because of chart time and logging errors.41 l Because of other irregularities in the tests of December 8, I i December 23 and December 30, 1978, MPR was unable to confirm the alleged water additions.42
]
The makeup tank level trace during the test of February 6, 1979, was so heavily ~ influenced by an apparent switch from automatic to manual mode control that we were unable to confirm the water addition alleged by the NRC during that test as well.43 We could confirm alleged water additions on March 19 and March 21, 1979, but we were unable to conclude that these additions were strongly indicative of leak rate test manipulation.
HUGH MCGOVERN The water additions of March 19 and March 21, 1979, - were logged 'and accounted for in the leak rate test but- were allegedly understated accordina to the NRC.45 We could not , find conclusive evidence that this was the case. On the ' contrary, it appears that the discrepancy between the amount of added water recorded by the operators during these tests a'nd that reflected in the strip charts was actually caused by the inaccurate makeup tank level instrumentation described in Volume I. Indeed, we discovered that, in addition to those tests listed by the NRC, there were three additional tests involvina accounted-for water additions on March 17, March 22,
.and March 27, 1979.46 Each of these tests was also significantly influenced by the combined effect of the inaccurate makeup tank level. instrumentation and the. water.
addition and would have yielded unsatisfactory results were it not for the addition.47 When McGovern was confronted with these tests, he denied that he had ever added water to manipulate test results and testified that he was not even aware that water addition made under these circumstances could produce more favorable test results.40 He also testified that he had never even heard anyone discuss this problem at that time.49 I l HUGH MCGOVERN Plant records support his testimony. Water additions were in fact Laing made guite frequently during the last half of March 1979 and they were approaching a frequency which, understandably, could have overlapped with the running of leak rate tests.50 i i On McGovern's shift, all five tests involving possible improper water additions were performed between March 17 and l March 27, 1979.51 Two of the tests included water additions which were made when the makeup tank level reached the low level point and were undoubtedly necessary to maintain proper makeup tank level. While it is not clear why the remaining three water additions were made, there is no evidence to suqqest that they were made in an effort to falsify test results. On the contrary, the lack of a pattern of improper water additions by McGovern, particularly during the first half of March when McGovern's shift filed no leak rate tests, 1 suggests that the water additions that were made were not intended to influence test results. Operator Response to Indications of Excessive Leakage The practice of discarding unsatisfactory test results without documentation has made it difficult to identify every
HUGH MCGOVERN brief period when there were indications that unidentified leakage exceeded 1.0 qpm at TMI-2. However, we were able to identify and closely examine three such periods. It seems that McGovern was working only during one of these periods, in March 1979. The events of this period are more fully discussed in Volume I of this report. For purposes of this section, this period is discussed only to the extent that it relates to the action taken by McGovern in response to the indications of excessive leakage. First of all, it should be pointed out that McGovern stated-that he was aware in 1978 that if unidentified leakage exceeded 1.0 qpm for several hours, he was required to reduce the leakage to within limits or to begin shutting down the plant.53 He admitted that the 1.0 gpm limit was not viewed as a requirement that had to be rigidly adhered to and that Operations personnel at TMI-2 did not strictly comply with the Technical Specifications reaarding reactor coolant system leakage.54 They viewed the 1.0 gpm limit more as a guideline than a requirement.55 As a result, he indicated that operators would generally monitor plant instruments, observe trends in various plant {
- HUGH MCGOVERN l
l conditions, and approximate reactor coolant system leakage. As i long as the leakage could be controlled, the plant would not be shut down.b According to McGovern, he never felt that the l plant was operating with leakage in excess of 1.0 gpm and he never felt any pressure from manaaement to operate the plant when it should have been shut down because of excessive reactor coolant system leakage.57 For the most part, plant records are consistent with McGovern's testimony. He was not assigned to the control room durino the period in October 1978 when unidentified leakage appeared to exceed 1.0 qpm.58 McGovern was assigned to the Control Room in late December 1978 for a few shifts, but stopped reporting for duty there in early January before reactor coolant system leakage became excessive.59 It was not until March 1979 that McGovern was working regularly in the Control Room during a time when there were indications that unidentified leakage exceeded 1.0 ppm. Our analysis indicates that unidentified leakaae increased during March 1979, until it approached close to 1.0 gpm on March 16, and remained fairly constant thereafter. However, for the reasons discussed in Volume I, operators would have perceived unidentified leakage to be much greater. I
HUGH MCGOVERN l I j Between March 1 and March 13, 1979, McGovern worked eleven consecutive assigned shifts during which no leak rate test was filed.60- Based upon his testimony and the testimony of l others, it is likely that numerous unsatisfactory leak rate l tests would have been discarded during this period. Despite j the fact that other plant insruments were showing leakage on the rise, there is no evidence that McGovern, or his colleagues, responded to these test results in accordance with the procedures. i McGovern stated that he did not specifically recall the events of this period and could therefore not explain what had occurred. However, McGovern acknowledged that it is probable that his shift would have been attemptina to obtain satisfactory leak rate tests.61 He further acknowledged that, based upon his experience, it was probable that if tests showing leakage in excess of 1.0 opm had been obtained they would have been discarded.62 l l 1 When asked whether there was any justification for simply I l discarding unsatisfactory tests in light of the indication of excessive leakage, he essentially maintained that computer g i tests were not that realistic and, in any event, that it was i 1 d
HUGH MCGOVERN standard practice to discard the unsatisfactory tests.63 However, the perceived unreliability of the computer results is insufficient justification for either discarding unsatisfactory test results or failing to take required action in response to such results. This is particularly true since during this period, the operators were manually re-calculating the test results generated by the computer to correct for a calculation error in the computer program. Therefore, the operators should have perceived the recalculated results for unidentified leakage to be more accurate and reliable. 16 -
7____--- HUGH MCGOVERN l i l 1 McGovern 2/6/85, p. 15. 2- Ibid., p. 16, 3 Ibid., pp. 14, 20, 27. l 4 Ibid., pp. 15, 17. l 5 Ibid., pp. 16, 21, 26.
.l
'6 _ Ibid., p. 26. l 7 Ibid., p. 23.
8 Ibid., pp. 20, 27, 29. 9 Ibid., pp. 30-32. 1 i 10 Ibid., p. 27. 11 Ibid., pp. 28-29. 12 Ibid., pp. 30-31. 13 Vol. III, Table 2. 14 Ibid. j 15 Ibid. 16 Ibid. :l j 17 Ibid. 18 Ibid. 19 Ibid. 20 Ibid.; ! Vol. IV (MPR Report), Appendix K, Test Numbers 59, 57, 55, 52, 51, 49, 42, 37. 21 Vol. III, Table 2. ; l 22 t$cGovern 2/6/85, p. 43. 17 -
HUGH MCGOVERN ' l ! 23 Vol. III, Table 2; ) I Vol. IV (MPR Report), Appendix K, Test' Numbers 179, '{ 178, 108, 94, 85, 84, 78, 52, 47, 37, 8, 6, 5, 2. l 24 Vol. III, Table 2; Vol. IV (MPR Report), Appendix K, Test Numbers 179, 178. 25 Vol. III, Table.2; Vol. IV (MPR Report), Appendix K, Test Numbers 85, 84, 78. 26 Vol. III, Table 2; Vol. IV (MPR Report), Appendix K, Test Numbers 94, 85, 78. 27 McGovern 2/6/85, pp. 81, 89. 28 Ibid., pp. 33-36. 29 Ibid., p. 86. 30 Ibid., p. 89. 31- Ibid., p. 90. 32 Vol. III, Table 2. 33 McGovern 2/6/85, pp. 95, 103. 34 Ibid., pp. 98-99. 35 Vol. IV (MPR Report), Appendix K, Test Numbers 78, 6. 36 Ibid., Test Number 78. 37 Vol. III, Table 2. 38 Ibid.; McGovern 2/6/85, p. 102. [ corrected 1/17/86) l i J
I HUGH MCGOVERN r l Vol. III, Table 2;
~
f ] i 1 Vol. IV (MPR Report), Appendix K, Test Numbers 179, { 178, 108, 94, 85, 84, 78, 52, 47, 37, 8, 6, 5, 2. 24 ol. III, Table 2; I Vo . IV ,MPR Repor t) , Appendix K, Test Numbers 179, 178. 25 Vol. I, Table 2; Vol. IV MPR Report), Appendix K, Test Numbers 85, 84, 78. 26 Vol. III, Ta e 2; ! Vol. IV (MPR Re ort), Appendix K, Test Numbers 94, 85, ' 78. 1 27 McGovern 2/6/85, pp. 81, 89. 28' Ibid., pp. 33-36. i 29 l Ibid., p. 86.
)
l 30 Ibid., p. 89. l 31 Ibid., p. 90. 1 32 Vol. III, Table 2. l 33 McGovern 2/6/85, pp. 95, 105-106. 34 Ibid., pp. 98-99. 35 Vol. IV (MPR Report), Appendix K, Test Numb rs 78, 6. 36 Ibid., Test Number 78. 37 Vol. III, Table 2. 38 Ibid.; McGovern 2/6/85, p. 102. I l 1 HUGH MCGOVERN 39 Vol. IV (MPR Report), Appendix K, Test Number 6. l l 40 Ibid., Test Numbers 108, 85, 84, 78, 52, 47, 7. ! 41 Ibid., Appendix H, Table H-1; Appendix K, Test Numbers 84, 47. 42 Ibid., Appendix H, Table H-1; Appendix K, Test Numbers 108, 85, 78. 43 Ibid., Appendix H, Table H-1; Appendix N, Test Number 52. 44 Vol. III, Table 2; Vol IV (MPR Report) , Appendix K, Test Numbers 7, 6. 1 45 Vol. III, Table 2; l i' Vol. IV, (MPR Report), Appendix K, Test Numbers 7, 6. 46 Vol. III, Table 2; f 1 Vol. IV -(MPR Report), Appendix K, Test Numbers 8, 5, 2. 47 Vol. III, Table 2. j f 48 McGovern 2/6/85,.pp. 98-100. 49 Ibid., p. 100. 50 Vol. III, Table 4. I 51 Vol. IV (MPR Report), Appendix K, Test Numbers 8, 7, 6, 5, 2. 52 Ibid., Tests Numbers 7, 2. l 53 McGovern 2/6/85, pp. 5-6. 54 Ibid., pp. 13, 27-29. 55 Ibid., p. 30. i
HUGH MCGOVERN 56 _ Ibid., p. 31. 57 Ibid., pp. 31-32. 58 Vol. III, Table 2. 59 Ibid. 60 Ibid. 61 McGovern 2/6/85, pp. 108-112. 62 Ibid., p. 110. 63 Ibid., pp. 111-112. 1 [ corrected 1/17/86] !
'I
HUGH MCGOVERN Ibid., p. 31. 57 Ibid., pp. 31-32. 58 vol. III, Table 2. 59 I id. 60 b . 61 McGov n 2/6/85, p. 108. 62 Ibid., p 110. 63 Ibid., pp. 11-112. l BRIAN MEHLER began his employment with Metropolitan Edison Company on May 9, 3967. He is presently employed by GPU t Nuclear Corporation as Radwaste Operations Manager for TMI-1. Between April 1978 and March 1979 he was employed as a Shift Supervisor at TMI-2. He was normally assigned to supervise a shift comprised of Shift Foreman, Charles Adams and Control Room Operators, Joseph Congdon and Martin Cooper. Analysis of . employee testimony and plant records reveals that Mehler failed to ensure compliance with the Technical Specifications and j procedures requiring documentation of reactor coolant system leakage. He also failed to ensure that required action was I taken in response to indications of excessive reactor coolant system leakage in October 1978, January and March 1979. There < l is insufficient evidence to conclude that he participated in, ! or knowingly tolerated, manipulation of leak rate test results. I i Supervision of Leak Rate Testing i It is clear from Mehler's testimony, as well as the testimony of others who worked under his supervision, that he ] l condoned the practice of discarding and failing to document I unsatisfactory leak rate test results obtained on his shift. l 1 l l l
BRIAN MEHLER Mehler was aware of the limiting condition for operation j for reactor coolant system leakage at TMI-2 and was also aware-of the requirement to proceed toward shutdown in the event that ! unidentified leakage exceeded 1.0 gpm for more than several hours.1 Mehler understood it was his responsibility as Shift Supervisor to assure compliance with this requirement. He l also acknowledged that the leak rate test was the means ! required by company procedure to be used to demonstrate compliance with the Technical Specifications. j Mehler denied being aware in 1978 or 1979 that the leak i rate test was inaccurate and unreliable as a means to measure reactor coolant system leakage for Technical Specification compliance. He stated it was the best tool available for the purpose. Further, he testified he did not recall any period j when either he or operators he supervised were concerned with i the reliability of leak rate test results. It should be noted that this testimony is not consistent with that of operators supervised by Mehler. Both Adams and l Congdon testified that th ey considered leak rate test results inaccurate and unreliable as a means of measuring reactor coolant system leakage.' Further, Adams stated that he believed he complained to his Shift Supervisor about the plant 1 l
l BRIAN MEHLER l l i operating without knowing what actual reactor coolant system 1 leakage truly was.6 Only after extensive questioning on the subject of test result variability did Mehler concede that he knew that test l 1 were
^
results variable to some extent and, therefore, not completely consistent with actual plant leakage. i He testified that as a Shift Supervisor it was rare for him to get personally involved in the performance of leak rate tests. The test results were not something he expected to be advised of on a day-to-day basis. The only time he expected to be advised was if there was a recurring problem or if he needed the test result to complete the Daily Plant Status Report. Mehler assumed there were no problems with the test unless something was specifically brought to his attention.O i Further, Mehler stated he would evaluate test results only , s if asked to do so, or in the absence of the Shift Foreman. As far as his evaluation was concerned, he only looked at the .
)
bottom line figures for reactor coolant system leakage.9 He l' I did not, as a practice, require his shift personnel to bring l I tests exceeding Technical Specification limits to his ! I O attention. J
~
1 BFIAN MEHLER l l l Although Mehler understood that one of his responsibilities was to ensure tha t the leak rate test was conducted properly in ]1 compliance with the procedure and Technical Specifications, the evidence indicates that he did not perform this responsibility effectively.11 It is clear that Mehler did not take any steps to prevent J operators on his shift from discarding test results without documentation wh en those results revealed unidentified leakage . k in excess of 1.0 gpm. l Mehler admitted that he was aware that operators were q discarding test results that were deemed to be inaccurate.12 i He stated he did not recall discarding any tests himself.I Mehler contended tha t he assumed that before discarding a test, his operators would evaluate the results and make a determination that the test was invalid.I He admitted that during that period, he did not pay much attention to the nature 5 However, it was his belief that of the evaluation process. othe r plant indications, such as makeup tank level and pressurizer level, would be checked by the operators before 6 determining a test to be invalid and discarding it. l i
o - _ - j 4 BRIAN MEHLER Further, Mehler stated that, if there were leakage indications, attempts would be made to identify the leakage for inclusion in the next leak rate test calculation. 'Once this occurred, the previous tests would be discarded.17 i During our interview with Joseph Congdon, he conceded that as it became more difficult to obtain satisfactory results, operators would no longer attempt to evaluate or verify leak rate test' results. Those showing unidentified leakage ' greater than 1.0 gpm were simply discarded.1 When confronted with Congdon's statements, Mehler nevertheless claimed that he believed that the process of evaluating leak rate test results was never abandoned, and that if tests were found to be valid, they would not be discarded.1 It should be noted that Charles Adams basically maintained the same position as Mehler until he was confronted with plant documentation. Only'at that point did he acknowledge that tests were discarded without l evaluation. Likewise, when Mehler was confronted with plant documentation for January 1979, he conceded his uncertainty about whether tests with unidentified leakage above 1.0 gpn were evaluated before they were discarded. He stated that it was possible that all test results showing leakage above 1.0 gpm were discarded without concern for their validity. 1 i [ corrected 1/17/86) E
y - , w.
-l
> 3 l
)
a l a
- , 1 x l
, s' !
l
\ s.
s BRIAN MEHLER Ti ', l
~~
- k. c.
,? , urther, x Mehler stated that, if there were leakage ) s ;- l l-
; indicat ons, attempts would be made to identify;the leakage for 1 3 - ,
Pinclusion in the nat leak rate test calculation. Once this l Joccurred,tt previous tests would be discarded.17
,s
/
,4 s o I
During our i Lerview with Joseph Congdon, he conceded that as it became more difficult to obtain satisfactory results, ' l operators would no 1,gger attempt to evaluate or - verify leak 1 A rate, test , results. Tho doding unidentified leakage grebter then 1.0 gpm were simply ' discarded.18 'When confronted with I l Congdon's , c ta t eme n t's , Mehle nevertheless claimed that he believed that the process of ev luating leab rate test results was n64er,4bcMoned, and thdt if _ sts were found to be valid, l I i they would not be discarded 19 t should be noted that l Charles Afais basically maintained the same position am Mehler
>(s until bei wm onf ronted with piant documer ation. Only at that s s point did h'e ~ acknowledged that tests (, ore discarded without-
-' I
[\ evaluation.20 ' 3 l - s
' *- ,k.
l Likewise, when Mehler was confronted 'thL plant sw ,
' Documentation for January 1979, he conceded ' his un ertainty about whtther tests with unidentified leakage above 1. 0. gpm were e va l'im t ed before they were d i sca rd ed . , He stated the it i
was possible that all test results showing leakage above 1. Sipm.wero? discarded without concern,for their validity.21
\
I s s 1 ,
BRIAN MEHLER Mehler testified that it was the practice on his shift to. j try to perform a leak rate test daily.22 He did not believe that many tests were discarded on his shift. Mehler assumed for . the most part that the results obtained were within the Technical Specification limits with the exception of the period just before the accident. According to Mehler, at that time, it was likely that more tests were discarded because of increased identified leakage flowing into the reactor coolant drain tank. l Mehler also knew thi _ Exceptions and Deficiencies were not
]
being filed with respect to leak rate tests. When asked if, as j l a Shift Supervisor, he were aware that the Exception and
]
Deficiency procedure applied to leak rate test surveillance he stated, "It never entered my mind if it d.id or it didn't". Mehler's testimony is not consistent with that of his Shift Foreman, Charles Adams, or with our analysis of plant records pertaining to leak rate testing at TMI-2. Adams testified that during the last four months of the operation of TMI-2, it became increasingly dif ficult to obtain satisfactory test results. Indeed, he stated that during this period up to four tests might be performed on his shift
BRIAN MEHLER before a satisfactory result was obtained. Additionally, our analysis of leak rate test records for the months of the January, February, and March 1979 indicate that most of ?. t a shifts supervised by Mehler failed. to file leak rate test results.27 Given the practice of performing one every shift, is likely that numerous tests were discarded throughout this period. This evidence is inconsistent with Mehler's statement that most tests were not discarded with the exception of the short period of high drain tank collection just before the accident. ; i In addition to questioning Mehler about his knowledge of test results over 1.0 gpm being discarded without evaluation, < 1 we questioned him concerning invalid results that were filed l because they showed unidentified leakage under 1.0 gpm. The clearest examples of such tests being filed occurred in December 1978 and January 1979 when a makeup tank level transmitter was malfunctioning. Mehler was questioned about tests performed on his shift, using that faulty transmitter. { I He stated that if he knew that an inaccurate makeup tank level transmitter was used during a leak rate test, he would ' consider the test invalid. Hrwever, Mehler stated repeatedly that he was not aware of the makeup tank level transmitter problem in 1978 and 1979.30 I
/
BRIAN MEHLER l l \ ! I l Both Adams and Congdon testified that the level transmitter ! l problem was common knowledge in the Operations Department at l I that time. Further, Shift Supervisors Hitz, Chwastyk, and Zewe all referred to the problem in their Shift Supervisor 1 turnover notes.32 The turnover notes of Hitz dated December j 20, 1978, indicate that the makeup tank level transmitter was l placed out-of-service because of the problem. i Mehler conceded that it was his practice to read the turnover notes from other shifts and that only important i i matters meant to be communicated to the oncoming shift were discussed in turnover notes. Although he did not recall l specifically, Mehler admitted after review of Zewe's turnover i notes of January 2 and January 4, 1979, that he was likely aware of the transmitter problem at that time. Mehler's shift showed an awareness of the malfunctioning level transmitter by avoiding its use for some leak rate tests. On four occasions, Mehler's shift used the accurate transmitter to supply data to the computer for leak rate tests. After reviewing the documentation relating to those tests, Mehler admitted that someone on his shift probably knew one of the transmitters was more accurate than the other. 1 i 1 l 1 BRIAN MEHLER l i l , We reviewed with Mehler the results of the leak rate tests performed with the faulty level transmitter and filed by his shift on December 30 and December 31, 1978. The results of the December 30 test show unidentified leakage to be -3.8 gpm, and the results of tile December 31 test indicate an unidentified leak rate of -2.4 gpm.37 The test of December 30 was started by the previous shift , { and completed by Mehler's shift. Upon completion of the test, ) l the inaccurate transmitter was then switched from the computer ) l 8 If a test result had not been filed by to the strip chart. Mehler's shift, 72 hours would have expired without a satisfactory test. Mehler admitted that the imminent expiration of the 72-hour period would have been communicated to his shi f t . He also agreed that the test with such a high negative leak rate, may have been filed because the 72 O hour time limit was close to expiring. Mehler had no recollection of this test. He stated that the decision to file the test would not have been discussed with him prior to its filing.41 When he was confronted with his testimony that problems would have been brought to his attention, Mehler contended that this situation was not a problem.42 _ 9_
i l i BRIAN MEHLER Mehler reviewed the leak rate test data for the next test filed by his shift on December 31, 1978.43 He recognized that the test was also performed with the inaccurate 1 transmitter and that it had been switched to the computer during his shift.44 After reviewing the data for the tests of December 30 and 31, 1978, Mehler acknowledged that it was possible for someone who was aware of the differences between the level transmitters to obtain a successful leak rate test result 50 percent of the time by using the faulty one. He acknowledged that this result would be obtained irrespec ti ve of actual conditions.45 He I 4 did not recall, however, any discussion about this possibility 4 among members of his shift or others.46 Manipulation of Test Results Although the evidence establishes that Congdon and Adams were involved in manipulating leak rate tests by adding hydrogen to the makeup tank, there is insufficient evidence to find conclusively that Mehler, as their Shift Supervisor, knowingly tolerated the practice. l
.l BRIAN MEHLER It is clear from Mehler's testimony that he was aware of the effect of hydrogen on makeup tank level indication as a i 1
4 j result of his experience at TMI-1. He stated that he believed it had the same effect at TMI-2.47 Despite his awareness of 5 this phenomenon, Mehler insisted that he did not communicate his knowledge to anyone nor did he have any recollection of any discussions about it. However, Mehler believed that snyone who added hydrogen to the makeup tank could see its effect.49 I He did not establish a practice on his shift of avoiding 0 hydrogen additions during leak rate testing. i { 1 Mehler testified that he was not aware that operators on his shift were . Sing hydrogen during the course of leak rate l I tests in an effort to manipulate the results. He stated that even though he was aware of the effect of hydrogen additions on makeup tank level readings, he was not aware that they could affect the results of leak rate tests. I 1 An analysis of test results obtained by Mehler's shift f between February 15 and March 1, 1979, indicates that hydrogen was added during three of the four tests filed in that ] period. When Mehler was confronted with the Shift i _ _ _ _____ _____-_ a
BRIAN MEHLER Supervisor turnover note written by Chwastyk, dated February 19, which refers to the effect of a hydrogen addition on a leak rate test, he could not recall having seen it.54 He also could not recall discussing the phenomenon with other Ehift Supervisors, as described in Chwastyk's testimony. Mehler could not recall the leak rate test performed by his shift on February 15, 1979.56 According to Condgon, he and other operators performed an experiment to determine the effect-on leak rate tests of adding hydrogen to the makeup tank. When confronted with the test data for February 15, Mehler did not concede that the rise on the makeup tank strip chart was attributable to hydrogen, despite the annotation on the chart indicating a hydrogen addition at that point. He stated that in order to reach a judgement on the test result and the possible effect of the indicated rise in makeup tank level, we would have to provide him with all the plant data for the time of the test.59 Mehler contended that the hydrogen experiment was not something he would have expected O his operators to bring to- his attention. However, Mehler admitted that even if he knew his subordinates had added hydrogen to affect a leak rate test result, he would have allowed that result to be filed, since he believed hydrogen would not affect the results .61
1 l BRIAN MEHLER There is no direct evidence that Mehler was aware of employees under his supervision adding hydrogen during leak rate tests in order to maniculate their results. Although Congdon was certain that Adams was aware, he only surmised that Mehler knew about the hydrogen additions.62 There is no evidence that Mehler was aware of, or , tolerated, any other method that might have been used for leak { rate test manipulation, such as " jogged" water additions.63 He also denied being aware that operators added water to the ] l makeup tank in an effort to take advantage of the makeup tank ) l instrumentation problem that caused an overstatement of the l 1 I level indication.64 ' Response to Indications of Excessive Leakage I Mehler was on duty during the three periods of high leakage 1 that we closely examined. He recognized that as a Shift Supervisor he had the responsibility to evaluate the safety implications of reactor coolant system leakage and would generally advise his superiors of his findings.65 He i acknowledged it was his responsibility to implement the action statement whenever actual leakage exceeded the Technical Specification limits. He believed that any time a " valid leak i 1 l I
BRIAN MEHLER i rate" was obtained, it would be necessary to reduce the leakage or take the necessary steps to shut the plant down.66 Mehler understood a valid leak rate to be one which was consistent with other plant parameters used to indicate potential reactor 1 coolant system leakage.67 Generally, Mehler was not able to recall very much about periods in which there was excessive reactor coclant system ; l leakage. He testified that he did not recall any situation ; l l where he considered the plant to be operating unsafely because
]
of excess reactor coolant system leakage or when it was necessary to shut the plant down.68 1 I October 16 - October 22, 1978 1 1 Subsequent to obtaining a successful leak rate test result at 19:28 on October 15, 1978, Operations personnel apparently < began to experience reactor coolant system leakage problems and difficulty in obtaining satisfactory leak rate test results.69 The last shift on October 16, 1978, performed a leak rate test that showed unidentified leakage at 2.56 O gpm. This test was not signed. Neither the Control Room Log nor the Shift Foreman Log contains any reference to action I being taken as a result of the unsatisfactory result. 14 --
I I BRIAN MEHLER 1 l On October 17, 1978, other Operations personnel were attempting ~ to identify the source of reactor coolant system leakage and had obtained anoth er unsatisfactory test result ; showing unidentified leakage to be 2.07 gpm. Again, there i was no reference in any of the logs to entry into ' the action statement as a result of the unsatisfactory test. I The 11pm-7am shift on October 18 performed a leak rate test 74 showing unidentified leakage to be 1.77 gpm. This test was j not signed. Kenneth Bryan, the Shift Supervisor for that ll shift, wrote in his turnover notes, "Still can not get a leak rate - 1900 today is deadline Doing hand calculations Can not find leak, don't believe it's in the RB." There is nothing in the plant records during that time to indicate that , k Operatione personnel were proceeding to shut down the plant. 1 Mehler's crew worked on October 18, during the 7am-3pm shift. Mehler stated that he had little independent recollection of this period. However, he acknowledged that it was likely that he read Bryan's shift turnover notes of October 18. He had no recollection of having seen the leak 1 I rate test results of October 16 and 17.79 ) i 1 l 4 i
-____________J
i BRIAN MEHLER Mehler was shown the four tests that were performed on his shift on October 18.80 He acknowledged that he probably saw them.81 After examining th e writing on the test sheets referring to the rounding off .of leakage results, Mehler recalled Floyd advising him that as a result of a conversation he had with NRC Inspector, Donald Haverkamp, he could round off test results. Mehler stated that the writing on the test conducted at 10:18, referring to rounding off to the nearest whole number, was his and that it was written in response to what Floyd told him. Further, it was Mehler's belief that he showed the results of the leak rate tests 'obtained on October 18 to Floyd.84 He had no specific recollection of performing the test signed by him at 10:18. However, he did say it was possible he had personally performed the test.85 Mehler reviewed the results of the test performed by Bryan's shift at 05:14 on October 18, and the note on the data sheet saying " rounds off high, but is corrected by leak rate 10-18-78 7:35:27 start time, ie into action action statement at 5:13:02 out of it 7:35:27".86 He stated that he did not recall seeing the 05:14 test or the handwritten note in October 1978.87 He could only recall a time when. he reported for duty on a morning shift and found the plant in the action l
BRIAN MEHLER statement preparing to shut down because of excessive reactor O coolant system leakage. However, he conceded that if this were the case on October 18, there should be a notation in the ! l Shi f t Foreman's Log. Mehler had no recollection of observing i such a log entry, and we found none during our review of plant re cc. rd s . O j 1 l Mehler then stated that perhaps they were not actually in jl i the - action statement, but only making preparations for taking the required action.90 When asked what such preparations , entailed, he expressed uncertainty and could not recall any a specific instructions or steps to be taken when preparing to enter the action statement. After reviewing documentation, Mehler conceded that th e round off appeared to be the means of avoiding the requirement to proceed toward a shutdown. Mehler had no independent recollection of LER 78-62, but acknowledged that he probably read it some time prior to the accident. He did not recall the Operations Memorandum written by Floyd on October 20, but stated tha t he probably read it also.94 He did not recall any discussion with Mark Bezilla about the contents of Cne LER.
BRIAN MEHLER Mehler did not recall any meetings or discussions with his superiors on the contents of the LER or the . Operations Memorandum.96 He stated that the only steps he would have taken as a Shift Supervisor to inform his subordinates of the contents of the Operations Memorandum would be to place it in the Operations Memorandum book for their review.9 Mehler acknowledged that his present understanding of the l Operations Memorandum is that the required action should .be taken any time that a leak rate test result indicates excessive unidentified leakage. However, he did not recall his understanding of the Memorandum in 1978. December 26, 1978 - January 15, 1Q79 MPR's analysis reveals that unidentified leakage exceeded 1.0 gpm on about January 3, 1979, and remained above that level until January 15, 1979. According to plant records, Mehler worked January 4 through January 10, and January 13 through January 15. Out of the 10 shifts he worked, leak rate test results were filed on only three occasions." Between January 7 and January 14, he worked six consecutive assigned shifts during which leak rate tests were not filed.100 BRIAN MEHLE2 Mehler was shown plant records pertaining to reactor j coolant system leakage for the period of January 7 through January 15. He was shown sump pump data, turnover notes, Radiation Work Permits (RWP's) and other evidence which ! indicated high leakage.101 Mehler conceded that in light of these indications of high reactor coolant system leakage, it was likely that his shift performed leak rate tests.1 I After reviewing his turnover notes for January 13, 1979, which state, ne e d leak rate," he acknowledged that his shift probably had been performing leak rate tests on that day despite the fact that no records were retained.103 After additional questioning, Mehler stated that leak rate test q results above 1.0 gpm that were obtained during this period, might have been discarded.104 When it was pointed out to i ( Mehler that the test results were most likely accurate, and should not have been disregarded, he could not explain why this had occurred.105 March 15 - March 28, 1979 l l l Unidentified ! leakage gradually increased during March i 1979. On March 16, 1979, it reached approximately 1.0 gpm and remained fairly constant until March 28, 1979.106
)
i l 1 l 1
l I, l BRIAN MEHLER I ! ) l Mehler's shift worked during most of that month. They were i present on March 1, 2,. 5, and 6, and from March 21 ' through j March 25.107 only two test results were filed on his shift between March 1 and 5. One test was filed between March 21 and 25.108 1 4 Given the testimony and the available plant documentation, I j it is highly probable that unsatisfactory test results were obtained by Mehler's shift during this period and were 1 discarded and not otherwise documented. It is also probable i that Mehler was aware of the increased leakage and the difficulty in obtaining satisfactory results. Indeed, Mehler recalled reporting the problem of excessive leakage from the safety relief valves (occurring just before the accident) to James Floyd. Further, Mehler specifically recalled that it was more difficult to obtain satisfactory test results just before the accident.I We have found no information that was available to Mehler at that time to justify the discarding of unsatisfactory results and the failure to take action as required by Technical
~ ._ _ - _ _ _ _ - -
i l BRIAN MEHLER -l i Specifications in response to results that exceeded limiting
.i
-conditions for operations. ;
i l
'ji I
i
)
l l l l l
l BRIAN MEHLER 1 Mehler 2/28/85, pp. 26-28, 32. 2 Mehler 2/28/85, pp. 25, 28, 64; 3/15/85, p. 160. 3 Mehler 2/28/85, pp. 25, 28, 45. 4 Mehler.2/28/85, pp. 52-54, 95, 101; 3/29/85, pp. 65-69. 5 Adams 3/8/85, pp. 14-15, 32; Congdon 2/13/85, pp. 37-38. 6 Adams 3/8/85, p. 73. 7 Mehler 3/29/85, pp. 65-69.
~8 Mehler 2/28/85, pp. 11-16, 49-50. l 9 Ibid., pp. 48-50.
10 Ibid., pp. 56-57. 11 Ibid., pp. 25, 64. I 12 Ibid., pp. 57-58, 64. 13 Ibid., p. 63. 14 Ibid., pp. 58, 60. 15 Ibid., p. 60. 16 Ibid., pp. 60, 67-68. 17 Ibid., pp. 42-44, 68. 18 Congdon 2/13/85, p. 34. 19 Mehler 2/28/85, pp. 73-74. 20 Adams 3/8/85, p. 133; 3/13/85, pp. 3-5. j 21 Mehler 3/29/85, pp. 10-11. 22 Mehler 2/28/85, p. 54. j I! l [ corrected 1/17/86)
BRIAN MEHLER 23 .Mehler'2/28/85, pp. 62, 64. 24 Ibid., p. 74. 25 Adams'3/8/85, pp. 66-67. 26 Ibid., p. 81. 27 Vol. III, Table 2. 28 Mehler 2/28/85, p. 64. 29 Ibid., pp. 37-38. 30 Mehler.3/15/85, pp. 104-109, 131. 31 .Congdon 2/13/85, p. 49; Adams 3/13/85, pp. 140-141. 32 THI-2 Daily Plant Status Reports, December 20, 1978;. January 2, 4, 6, 9, 1979, (Tab 10). 33 Mehler 2/28/85, pp.-10, 13, 88; 3/15/85, p. 133. 34 Ibid., pp. 133-134. 35 Vol. IV (MPR Report), Appendix K, Test Numbers 109, 103, 96, 79. 36 Mehler 3/15/85, pp. 104-108. 37 vol, IV (gpR Report), Appendix K, Test Numbers 78, 76. 38 Ibid., Test Number 78. 39 Mehler 3/15/85, p. 120. 40 Ibid., p. .121. 41' Ibid., p. 122. 42 Ibid., p. 122. 43 Vol. IV (MPR Report), Appendix K, Test _ Number.76. 1 [ corrected 1/17/86] _ - !)
! BRIAN MEHLER l 1 ' Mehler 2/28/85, pp. 26-28, 32.
\
2 hier 2/28/85,.pp. 25, 28, 64;
- p. 160.
4 3 Mg\3/15/85, Mehl r 2/28/85, pp. 25, 28, 45. Mehler 2/28/85, pp. 52-54, 95, 101; 3'Y 29/ 85, pp. 65-69,. ; 5 Adams 3/8/B , pp. 14-15, 32; Congdon 2/1{3 85, pp. 37-38. 6 Adams 3/8/85, . 73. 1 7 Mehler 3/29/85, . 65-69. i 8 Ibid., pp. 11-16, 4 -50. ; 9 Ibid., pp. 48-50. , 10 Ibid., pp. 56-57.
\ '
11 Mehler 2/28/85, pp. 25, 64. l 12 Ibid., pp. 57-58, 64. l 13 Ibid., p. 63. 14 Ibid., pp. 58, 60. 15 Ibid., p. 60. 16 Ibid., pp. 60, 67-68. 17 Ibid., pp. 42-44, 68. 18 Congdon 2/13/85, p. 34. 19 Mehler 2/28/85, pp. 73-74, 20 Adams 3/8/85, p. 133; 3/13/85, pp. 3-5. 21 Mehler 3/29/85, pp. 10-11. 22 Mehler 2/28/85, p. 54. BRIAN MEHLER 23 Mehler 2/28/85, pp. 62, 64. N 24 ehler 2/15/85, p. 74. 25 Ad is 3/8/85, pp. 66-67. 26 Ibid. p. 81. l 27 Vol. II , Table 2. 28 Mehler 2/ /85, p. 64. 29 Ibid., pp. 3 -38. l 30 Mehler 3/15/85, pp. 104-109, 131. 31 Congdon 2/13/85, . 49; Adams 3/13/85, pp. 40-141. 32 TMI-2 Daily Plant Sta us Reports, December 20, 1978; January 2, 4, 6, 9, 19 9, (Tab 10). 33 Mehler 2/28/85, pp. 10, 1 , 88; 3/15/85, p. 133. 34 Ibid., pp. 133-134. 35 Vol. IV (MPR Report), Appendi K, Test Numbers 109, 103, 96, 79. 36 Mehler 3/15/85, pp. 104-108. 37 Vol. IV (MPR Report), Appendix K, Tes Numbers 78, 76. 38 Ibid., Test Number 78. 39 Mehler 3/15/85, p. 120. 40 Ibid., p. 121. 41 Ibid., p. 122. 42 Ibid., p. 122. 43 Vol. IV (MPR Report), Appendix K, Test Number 76. l BRIAN MEHLER i 44 Mehler 3/15/85, pp. 130-131. 45 Ibid., pp. 135-138, 46 Ibid., p. 141. 47 l Ibid., p. 2. l
'i 48 Ibid., pp. 2, 4-5.
49 Mehler 3/29/85, p. 25. 1 50 Mehler 3/15/85, p. 98; I 3/29/85, pp. 39-40. 51 Mehler 3/29/85, p. 60. 52 Mehler 3/15/85, pp. 3-4. 53 vol. IV (MpR Report), Appendix K, Test Numbers 38, 32, 28, 24. 54 TMI-2 Daily Plant Status Report, February 19, 1979, (Tab 10). 55 Mehler 3/15/85, pp. 4-5; 3/29/85, pp. 37-39; Chwastyk 3/30/84, p. 44. 56 Mehler 3/29/85, pp. 27-29. 57 Congdon 2/13/85, p. 52. 58 vol. IV (MpR Report), Appendix K, Test Number 38; Mehler 3/29/85, pp. 45-46. 59 Ibid., p. 50. 60 Ibid., pp. 27-31. 61 Ibid., p. 33. 62 Congdon 2/13/85, p. 62. 63 Mehler 3/15/85, pp. 56, 142. 64 Mehler 3/29/85, p. 63.
m l 1 BRIAN MEHLER 1 I 65 Mehler 2/28/85, pp. 7-8, 24-25. 66 Ibid., pp. 31-32. j 67 Ibid., p. 32. 68 Mehler 2/28/85, pp. 20, 94; 3/29/85, p. 20. 69 Vol. IV (MPR Report), Appendix K, Test Number 150. 70 Ibid., Test Number 149. 71 Vol. IV, Appendix K, Test Numbers 151, 150, 149. 72 Ibid., Test Number 148. 73 Ibid., Test Number 148. I 74 Ibid., Test Number 177. j 75 TMI-2 Daily Plant Status Report, October 18, 1978, ! (Tab 10). 1 76 Vol. III, Table 2. 77 Mehler 3/15/79, p. 6. 78 Ibid., p. 7. 79 Ibid., p. 16. 80 Vol. IV (MPR Report), Appendix K, Test Numbers 146, 145, 144, 143. 81 Mehler 3/15/79, pp. 29, 32. 82 Ibid., pp. 18-19. 83 Ibid., pp. 37-38. 84 Ibid., p. 29. 85 Ibid., pp. 38-40. 86 Vol. IV (MPR Report), Appendix K, Test Number 147. _________b
.i BRIAN'MEHLER I- .
.87 Mehler 3/15/85, pp. 16-17.
88 Ibid., p.,20. 89 Vol. III, Table 4; Mehler 3/15/85,;pp. 20-21, 43-44. 90 Ibid., p. 23. 91 Ibid., pp. 23-24. 92 Ibid., p. 78. 93 TMI-2 Licensee Event Report'78-62/1T,. November, 1978, (Tabs 28, 29); Mehler 3/15/85, pp. 72, 75. 94 J. Floyd, Operations ~ Memo 2-78-19,.0ctober 20, 1978, (Tab 34); Mehler 3/15/85, p. 85. 95 Ibid., pp. 76-77 96 Ibid., pp. 81, 83-84, 87. 97 Ibid., pp. 85-86. 98 Ibid., pp. 86-87. 99 Vol. IV (MPR Report), Appendix K, Test Numbers'72, 69, 63. 100 Vol. III, Table 2. 101 Vol. III, Table 4; Mehler 3/15/85, pp. 142-149; 3/29/85, pp. 3-7, 12-13, 16.
-102 Mehler 3/15/85, p. 148.
103 Shift Supervisor Turnover Notes,- January 13-, 1979, 1 (Tab'10); l
~
Mehler 3/29/85, pp. 7, 9. jl 104 Ibid., pp. 9-10. [ corrected-1/17/86)
-l
BRIAN MEHLER 105 Mehler 3/29/85, pp. 10-11. 106 Vol. IV (MPR Report), Figure IV-20. 107 Vol. III, Table 2. 108 Vol. IV (MPR Report), Appendix K, Test Numbers 24, 23, 4. 109 Mehler 2/28/85, p. 8. 110 Ibid., p. 64. j l [ corrected 1/17/86)
l l l BRIAN MEHLER 7 Mehler 3/15/85, pp. 46-47. 1 i 88 Ibid., p. 20.
]
I 89 Vol. III, Table 4;
)
M hier 3/15/85, pp. 20-21, 43-44.
]
l 90 Ib ., p. 23. 91 Ibid., pp. 23-24. 92 Ibid., p 78. 4 93 TMI-2 Lice ee Event Report 78-62/lT, Noventber , 1978, j (Tabs 28, 29 ; i l Mehler 3/15/85, pp. 72, 75. l 94 J. Floyd, operati ns Memo 2-78-19, October 20, 1978, (Tab 34);
.1 Mehler 3/15/85, p. 8 "
95 Ibid., pp. 76-77. ' 96 Ibid., pp. 81, 83-84, 87. 97 Ibid., pp. 85-86. i 98 Ibid., pp. 86-87. 99 Vol. IV (MPR Repo'rt), Appendix K, est Numbers 72, 69, 63. 100 Vol. III, Table 2. 101 Vol. III, Table 4; Mehler 3/15/85, pp. 142-149; 3/29/85, pp. 3-7, 12-13, 16. 102 Mehler 3/15/85, p. 148. 103 Shi f t Supervisor Turnover Notes , January 13, 1979 (Tab 10); Mehler 3/29/85, pp. 7, 9. 104 Ibid., pp. 9-10.
..__________N
i BRIAN MEHLER j 105 Mehler 3/29/85, pp. 10-11. I 106 Vol. IV (MPR Report), Figure IV-20. ! 107 Vo' . III, Table 2. ' 108 Vol. IV (MPR Report), Appendix K, Test Numbers 24, 23, 4. 109 Mehler 2 28/85, p. 8. l 110 Ibid., p. 1 . i
'I
)
]
1 i i l I i
\
N N 1
i l ADAM MILLER began employment with Metropolitan Edison on April' 4, 1973. He is presently employed as Supervisor of Operations, TMI-2. $etween March 1978 and August 1978 he was a Control Room Operator at TMI-2. From August 1978 until March 1979, he worked as a Shift Foreman assigned to TMI-2. On his I shift, he was responsible for supervising Mark Coleman, Leonard Wright, and Dennis Olson. Gregory Hitz was his Shift' ) i Supervisor during most of this time. I Analysis of employee tectimony, leak rate test records, and ! related plant documents reveals that Miller failed to properly supervise the performance of leak rate tests and also failed to l ensure that required action was taken in response to indications of reactor coolant system (RCS) leakage- during October 1978 and January 1979. There is insufficient evidence from which we can conclude that Miller participated in, or knowingly tolerated, manipulation of leak rate test results.
o -4
'(
, , ')
I ADAM MILLER Supervision of Leak Rate Tests s There is sobr: bantial evidence that Miller failed to properly supervise the performance of leak rate. tests that were conducted on his shift. He tolerated and participated in the practice of discarding and failing to document all test results that showed unidenti fied' RCS leakage in excess of 1.0 gpm. He also approved the filing of numerous satisfactory test results that were clearly invalid, that did not accurately measure RCS leakage, and that were performed in violation of company procedures. It is apparent from Miller's testimony that he was not seriously committed to ensuring that leak rate tests were properly conducted to measure RCS leakage. He testified that he was aware Technical Speci fications and procedures required the plant to be operated within limiting condition for operations for RCS leakage and that the leak rate test was the primary method for measuring that leakage.1 Miller admitted knowing that his responsibility as a Shi f t Foreman required him to monitor RCS leakage and to ensure that Control Room Operators working under his supervision complied , 2-
1 l l J ADAM MILLER with the Technical Specifications.2 Despite this awareness, i he acknowledged that he did not fulfill this responsibility.3 l i According to Miller, he developed a belief as a Control Room Operator that the leak rate test was inaccurate and could 3 not be relied upon to measure actual plant leakage.4 He i 1 explained that since the leak rate test could not ensure that the plant was actually within limiting conditions for operation, operators relied on their own subjective judgment to make this determination. He described thei r approach as more of a " subconscious" reliance on other indicators,of plant conditions. Consequently, the leak rate test was not considered important and it was conducted in a " frivolous manner." I Ultimately a " casual" and " cavalier" attitude developed towards demonstrating compliance with the Technical Specifications by means of the leak rate test.8 l Although Miller stated he did not believe there was any formal decision not to comply with the Technical Speci fications pertaining to RCS leakage, he conceded that, upon reflection, it appears that they were consciously disregarded.9 i i i ADAM MILLER Given Miller's general attitude about leak rate testing and , compliance with the Technical Specifications, it is not . surprising that he did little to ensure that leak rate testing was properly conducted on his shift. Miller acknowledged that it was the practice to discard all unsa ti s fa ctory test results and that he personally discarded some unsa tis fa ctory test results.10 According to Miller, any test result with uniden ti fied leakage above 1.0 gpm obtained within 72 hours from the last satisfactory test would be discarded. These intervening unsatisfactory test results 1 simply were not considered official test results for purposes i of compliance with the Technical Specifications.12 As far as he could recall, they were discarded without serious efforts to determine their accuracy.I If there were other indications of excessive leakage at the time an unsatis factory test result was obtained, there would be i a search for leaks, and attempts would be made to reduce the leakage. This process, however, would have little to do with the decision to discard the test result.14 Miller testified that two or three times as many tests were discarded than were filed.15 He also stated that it was
i 1 ADAM MILLER his practice to allow Control Room Operators under his supervision to discard tests without his prior review.16 l
)
In addition to the discarding of all unsatisfactory test results, Miller also admitted being aware that no Exception and Deficiency reports were filed and that no log entries were made. documenting the test results that were being thrown away. Although Miller stated that, at the time, he did not believe that such a practice was a violation of the Technical Specifications, he conceded that this logic was " perverted" and l resulted from a " selective" and " casual environment" regarding compliance with the Technical Specifications.18 Miller i admitted that, as a Shift Foreman, he perpetuated this practice and felt his Shift Supervisor was aware of it as well.19 l Miller's testimony concerning the discarding of unsatisfactory test results is corroborated by our review of plant documents relevant to his shift. TMI-2 records indicate that between March 1978 and March 1979, Miller reported for duty on 114 shifts during which leak rate tests presumably could have been performed.20 Despite Miller's stated practice of performing leak rate tests at least 5-
ADAM MILLER once per shift, only 42 leak rate tests were filed during the shifts worked by Miller. No test result showing leakage 2 above 1.0 gpm was ever filed. There were several instances of consecutive assigned shifts supervised by Miller where no $ : leak rate tests appear. The two periods when the fewest tests were filed were October 1978 end January 1979, when it would 3 have been rost difficult to obtain a satisfactory result. In addition to tolerating, and participating in, the discarding of all unsa tis fa ctory test results obtained on his shift, Miller also approved the filing of numerous satisfactory i leak rate test results that were inaccurate or were performed < in violation of test procedures.24 Almost two 'chirds of the tests that Miller approved should have been determined invalid.25 Six of the early tests filed by Miller, when he l was still a Control Room Operator, contained obviously inaccurate information or were performed in violation of company procedures. 6 Some sa ti s fa ctory leak rate test results approved by Miller in December 1978 involved tests ; performed with inaccurate makeup tank level instrumentation.27 l The last 12 tests that Miller approved between February 25 l l and March 28, 1979, included logged water additions during the tests that caused indicated unidenti fied leakage to be reduced 4
1 I ADAM MILLER because of makeup tank level instrumentation problems and a calculation error in leak rate test procedure. There is substantial evidence that these tests were manipulated. They will be discussed later in this report. l l During his initial testimony, Miller stated that he did not i believe that he would knowingly approve inaccurate test results and that he did not recall approving any tests that he knew to be inaccurate. However, a f ter being pressed on what standards he used to evaluate and approve test results, Miller I e conceded that there wau no standard other than wheth er th e j result was within acceptible leakage limits.30 Eventually, 1 l he admitted that he would approve leak rate test results which were sa t is f actory regardless of their accuracy or their agreement with what other plant instrumentation indicated l I leakage to be. Acc'ording to Miller, he approved all l l satisfactory results even when he was unsure whether leakage i i was within the limitations set forth in th e Technical l Specifications.32 After f ur th er questioning, Miller finally acknowledged that it was possible that the only real purpose for his review of test results was to enforce the practice of submitting only those leak rate test results showing unidentified leakage less
ADAM MILLER { i l than 1.0 gpm. 33 Miller stated that the practice of knowingly submitting inaccurate results did not cause him great concern at the-time.34 f Miller was also questioned about his attitude concerning l compliance with that portion of the leak rate test procedure which requires quantifying and entering operator-identified ; l leakage into the computer. He stated that he believed his l practice was to require the use of a Data Sheet 3 for this ! purpose.35 He further stated that once leakage was l identified, the same Da ta Sheet 3 would be carried over and 1 used for each subsequent leak rate test until the leakage was j quantified again or the leak was fired.36 Miller acknowledged, however, that there may have been occasions when identified leakage was so low that it would not be included in ! the test result, particularly if test results could be easily , obtained that were below 1.0 gpm.37 Miller was shown a leak rate test approved by him on 1 December 21, 1978.38 Two Data Sheet 3's were attached to the j leak rate test result. The Data Sheets reflected leakage that l i had been identified as early as December 1, 1978, but which had not been included in any test approved by Miller until the test I of December 21. That portion of the Data Sheet 3 entitled, 1 j
ADAM MILLER
" Evaluating Safety Implication," was signed by Hi tz, his Shift f
l Supervisor, on December 21.39 When asked why there was a l l delay of three weeks before the leakage identified on December 1 was included in the results of leak rate tests, Miller could not explain the delay. He did state, however, that the Shift Supervisor would fill out a Safety Evaluation only when he decided to make use of the Data Sheet 3 in a leak rate test.40 Miller was th en shown th e test results approved by him on January 6 and January 11, 1979.41 The test of January 6 included no operator-entered identified leakage, but, the test of January 11 did include operator-entered identified leakage that was documented by a copy of the same Data Sheet 3 used on the test of December 21, 1978. Although Miller admitted that he probably reviewed the Data Sheet 3 before approving the test result, he had no explanation why the same Data Sheet 3 from December 21 was used on January 11 and why it was not used in the previous test of January 6th. Miller was also asked to review plant documents pertaining to other RCS leakage which had been identi fi ed during this same period and which had not been included in the Da ta Shee t 3 of January 11. He had no explanation for the failure to account for this leakage.44 Regarding the use of the faulty makeup tank level transmitter for leak rate testing, Miller testified that he recalled be.ng 4 1 (
l 1 ADAM MILLER aware of the transmitter problem in 1978 and 1979 45 He ] stated that it would have been his practice when he first became aware of the problem to instruct Control Room Operators
]
under his supervision not to us e . th a faulty transmitter when ) performing leak rate tests.46 J> I l 1 An examination of the test results from December 8 to i December 21, 1978, does not completely substantiate Miller's i testimony in this regard. The test results reveal that between i December 8 and December 13, 1978, four tests were approved by Miller where the accurate transmitter was in fact switched to the computer for the performance of the tests.47 On the l o th er hand, after those tests, two consecutive tests were 1 approved by Miller which were performed with the inaccurate ' level transmitter. These tests occurred on December 17 and l December 21, 1978. Miller was shown a December 20 turnover note of his Shift Supervisor which stated, "MU-T [ makeup tank level transmitter] LTl OOS - I&C watching". He conceded that th e note meant that transmitter was not to be used and that this fact was probably brought to his attention during that time frame.50 When confronted with the makeup tank level strip chart for the test that he approved on December 21, 1978, which clearly ADAM MILLER demonstrated that the inaccurate transmitter was used for the test, Miller stated that he could not be certain that the inaccurate transmitter was used in the performance of the . test. With respect to the twelve tests in which there were accounted-for water additions, Miller testified that it was his practice to allow water additions during leak rate tests as long as they were accounted for in the test calculation. He testified that he did not think that he was aware of the language in the leak rate test procedure that stated that water additions should be avoided if at all possible. Fur th e r ,' he indicated that it was possible he did not even read the procedure.53 The facts surrounding these water additions will be discussed further in the next section of this report. Manipulation of Test Results It is clear from both the testimony of Coleman, and our analysis of leak rate test data, that numerous manipulated leak rate tests were approved by Miller. Never th eles s , our investigation did not produce sufficient evidence from which we could conclude with certainty that Miller tolerated, or participated in, the intentional manipulation of leak rate test results.
ADAM MILLER Coleman testified that he could not recall whether Miller was aware that leak rate tests were being manipulated through the use of accounted-for water additions.54 In his testimony, Miller denied being aware that operators under his supervision were relying on the makeup tank level instrumentation problem by making water additions during leak rate tests.55 Moreover, he stated that had he been aware of the instrumentation problem, he would have considered it serious and taken action to correct it. The only significant evidence which suggests that Miller was aware of test manipulation, stems from his approval of 12 consecutive tests in which water was added and accounted-for in the test results. Those tests were the vast majority of sa ti s f a ctory test results filed by all TMI-2 operators between February 25 and March ~ 13, 1979. This was a period when other shifts were experiencing increased difficulty in obtaining satisfactory results because of increased reactor coolant drain tank collection and a calculation error in the leak rate test procedure.59 To determine Miller's involvement in this type of test manipulation, we carefully evaluated the tests that he approved and questioned him at length about th em . i
ADAM MILLER j Miller stated that, as far as he knew, it was only
-t necessary to add water to maintain the required level in the j makeup tank or to adjust control rod position by changing boron concentration in the RCS. He further testified that the determination of the necessity for adding water during a leak rate test was left to the discretion of the Control Room -
q Operators. He stated that as far as' he could recall, he was not aware that the leak rate test procedure indicated that water additions should be avoided during the test.62 1 Miller testled that he required the use of Data Sheet 4's to document water additions.63 He did not recall having a problem getting Contro? Room Operators under his supervision to comply with this practice.64 Miller stated that his normal l practice would have been to review the contents of the Data Sheet 4, compare it with the data on the leak rate test result, and ensure that they were consistent. He believes that he would have also reviewed th e Data Sheet for the time that the addition was made, as well as the operation that necessitated the addition.65 Test data did not support his testimony in this regard. Eight of the 14 tests manipulated through the use of accounted-for water additions do not contain Data Sheet 4's. Those tests in which the Data Sheet 4's were not
l l ADAM MILLER . I filed were nevertheless approved by Miller. In the section on the Data Sheet identifying the operation that necessitated the l addition of water, there was usually the notation i
" makeup-to-makeup tank" or " increased makeup tank level". The ,
Data Shee t 4 for February 16, 1979, however, has no such entry. . l All of the water additions in the tests conducted between February 16 and March 28, 1979, were made within the last few minutes of the test period. This time is noted on the Data ;
}
Sheet 4's which were approved by Miller.68 i l Miller said he had no explanation why the Data Sheei 4's were not completed for each test in light of his stated I requirements. Moreover, he had no recollection of being concerned about it. Despite the repeated water additions by Control Room Operators under his supervision, Miller stated that he did not remember questioning anyone regarding the necessity o f. the additions. After reviewing the makeup tank level strip chart and logs, he concefed that it did not appear to be necessary to add water during the course of the tests that he 1 reviewed. Nevertheless, he probably would not have 2 required operators to advise him as to their necessity.
ADAM MILLER l 1 It was not his practice to review strip charts, or Data Sheet 4's, before approving leak rate tests. Further, he j conceded that th ere may have been times that he signed Data ' Sheet 4's as a matter of form and did not read them very ! carefu11y.74 l 1 Miller maintained that he was not aware of the makeup tank i 1cvel instrumentation problem in 1979. Therefore, even had he noticed that water additions had been made during tests, he would n ever th eles s have approved the results, as he did not believe that the results could be influenced by water additions that were accounted-for. I I Miller was advised that operators under his supervision had i explained the necessity for adding water during tests by stating that they might have been changing boron concentration l l to adjust control rod position. He conceded that he would have l approved such tests without the paperwork required for changes in boron concentration because there was a general practice of ; foregoing the paperwork on routine matters.76 l l 1 l 1 Miller was confronted with the Operations memo, issued by ! l James Floyd on January 25, 1979. In that memo Floyd states: l
i 17 ADAM MILLER I~ A recent NRC audit of Operating procedures . . . revealed a large number of discrepancies. There ! is no excuse for the majority of these items. .' They can be eliminated by the persons performing the procedures taking a few extra minutes to apply some attention to details. Each item in itself seems insignificant, but two pages full indicates a laxity on our part in properly executing the responsibilities that accompany the operating license. l 1 In an effort to ensure that procedurcs are l completed properly, please use the following guidelines . . . 6. When performing evolutions which require paperwork, boron dilutions, neutralizer tank dumps, etc. ensure that the accompanying paper work is properly completed.77 Miller testified that he probably received this memo.
~
Further, it was his practice to ensure that his Control Room Operators were aware of the contents of all Operations O of the memo j Memos. Al though he had no recollection in I question, Miller conceded that it appeared to refer to the l procedural requirement to complete certain paperwork whenever boron concentration was changed in the reactor coolant system.79 Miller also conceded that, despite the memo, he did not require operators on his shift to document, or seek his approval, when making changes in boron concentration. He said that Floyd's memo was probably disregarded. When pressed further for an explanation, Miller then said that
ADAM MILLER perhaps Floyd's memo was not disregarded, but probably was interpreted to apply only to large changes in boron concentrations and not routine changes. Miller was confronted with the leak rate test data for the manipulated tests that he approved in late February and March 3 1979. Miller acknowledged that although he had no clear recollection of being on the only shift to obtain successful leak rates in early March of 1979, he was probably aware of that fact at the time. Further, he conceded that it was likely that he had conversations about this with other shift personnel. Despite th e extraordinary success of his operators in obtaining sa tis f actory test results, he maintained that he never questioned them about the repeated water additions that were made at the end of the tests. 5 hhile there is no question that Miller's failure to evaluate the propriety of the water additions demonstrates a lack of concern for compliance with the leak rate test procedure, there is insufficient evidence from which we can conclude with certainty that he tolerated, or participated in, test manipulation. Although there are no recorded tests on Miller's I shift i where there are alleged hydrogen additions, there was a logged 1
l l 1 ADAM MILLER hydrogen addition during the test of November 5, 1978. That I test was performed by Miller's shift, but filed by the next j i shift. It was the only alleged hydrogen addition on Miller's { 1 shift. 'However, we were not able to confirm that the addition ) of hydrogen had any effect on the test result.86 i Despite the lack of documentary evidence that hydrogen was added on shifts supervised by Miller, Coleman admits that he had done so on a number of occasions.87 As with the water i j I additions, Coleman could not recall whether Miller was aware of his improper conduct. Miller testified he was not aware that Control Room Operators under his supervision were adding hydrogen to l manipulate leak rate tests.89 Further, he said that he was l not aware of hydrogen's effect on makeup tank level indication i and denied having discussions about it with other Operations l personnel.90 l A February 19, 1979, turnover note by Chwastyk referring to the addition of hydrogen during a leak rate test. was written just prior to Miller's shift reporting for duty.91 Despite the fact that Miller's Shift Supervisor would have received this turnover note, Miller stated that he could not recall any communication concerning its contents.92 ADAM MILLER Unlike the situation with the tests involving accounted-for water additions, there does not appear to be any obvious pattern of conduct by Control Room Operators under Miller's supervision which should have put him on notice that hydrogen additions were possibly being used to manipulate test results. The NRC also alleged that unlogged, or under-recorded, water additions were made during seven of the tests approved by Miller. However, based on our analysis of strip charts, we were not able to confirm any of the allegations. It appears that two tests were performed with faulty makeup tank level instrumentation that caused strip chart irregularities. These aberrations made conclusive identification of the alleged water additions impossible.94 One allegation was based on a mistaken determination of chart time." Another allegation pertained to a test conducted when the plant was not at steady state. This also made definitive strip chart analysis impossible.96 The remaining allegations involved strip charts that revealed normal traces, similar to other periods before and after the test. The data for these tests were reviewed by Miller during our interview. He denied that the strip chart traces indicated unlogged water edditions. Further, he stated he was not
ADAM MILLER aware in 1978-1979 that " jogging" water into the system was a 'l method of leak rate test manipulation.9' Operator Response to Indications of Excessive Leakage , I I d Among the three periods that we identified where there were ; indications that unidentified leakage exceeded 1.0 gpm, Miller was on duty during the October 1978 and January 1979 1 periods. Miller, at the time, recognized that he was ! responsible for monitoring and evaluating RCS leakage as a Shift Foreman.101 He also understood that it was his 1 1 responsibility to ensure that the limiting conditions for l1 j operation for RCS leakage were not exceeded.102 He further g 1 stated that if the limiting conditions for operation were .I l exceeded, it would have been necessary to enter the action ! statement. I Despite this basic knowledge of the action requirements, it does not appear that Miller made efforts to ensure that the required action was taken in response to indications of l excessive RCS leakage. This fact is clear from the analysis of plant documentation and Miller's own testimony. ADAM MILLER 1 October 16 - 22, 1978 l I 1 I on October 15, 1978, Operations personnel apparently began to experience difficulty in obtaining satisfactory leak rate test results. During the 3pm-11pm shift on October 16, an l unsigned test result was obtained showing unidentified leakage j 104 On the 11pm-7am shift on October 17, no to be 2.56 gpm. leak rate test result was filed. Th ere are entries in various i l plant records which indicate that this shift was searching for the source of uniden ti fi ed leakage.105 Miller's shift reported for duty on the 7am-3pm shift of October 17.106 i Coleman performed a leak rate test showing unidentified leakage to be 2.07 gpm. This leak rate test was not filed. Plant records reveal that there were ongoing attempts to identify RCS l leakage.108 j q l Miller stated that it was the practice on his shift for l Control Room Operators to discard tests with unidentified own, l leakage above 1.0 gpm on their without his 09 asked why particular was permission. When this test signed and apparently not discarded, he acknowledged that the reason may have been that there was concern about RCS
]
leakage. Although he had no specific recollection of that test performed by Coleman, Miller believes it likely that the test was presented to him for review.111 ' l
i i ADAM MILLER ! l l Despite th e perceived leakage on October 16 and 17, the time elapsed since the last satisfactory test, and the intervening unsatisfactory test results that had been obtained, j there is no indication that efforts were under way to shut down i the plant. Company records do not reveal any official entry into the action statement. On October 18, Miller's shift apparently worked relief I during the day shift. There is no indication that Miller j l was involved in th e series of leak rate tests run on this ] shift. He does not recall any of the events of October 18, and has no recollection of the round-o f,f policy that began that day.113 Neither could he recall the LER, although he believes he probably read it.114 Miller acknowledged that th e LER did not change his attitude toward the necessity to 1 enter the action statement when there were leak rate test ll5 results with unidentified leakage above 1.0 gpm. He believed it was possible that the LER was simply disregarded.110 December 26, 1978 - January 15, 1979 Plant records for this period indicate that Miller reported for duty on 12 shifts. There is a leak rate test result
t J
-l 1
ADAM MILLER l recorded for his shdft on only three occasions. Plant documentation indicates that, during this same period, actual' unidentified leakage exceeded 1.0 gpm.119 Plant data was reviewed with Miller, particularly those documents relating to makeup tank level, sump level and leakage problems. He reviewed the reactor building sump data for January 12 indicating that there were ten sump pump starts on that day.1 He also reviewed data for the month through January 13 indicating an ongoing unquantified leak in the pressurizer area. Miller acknowledged he was probably aware of this information. No leak rate tests were filed on his shift for January 12 or 13, leading us to auspect that unsatisfactory results were obtained and discarded.1 Miller was asked if there was any justification for discarding leak rate test results indicating uniden ti fi ed leakage above 1.0 gpm given the plant conditions at that time. He conceded that there would be no justification for doing so, considering the seriousness of the leakage problem.1 [ corrected 1/17/86]
i I l ADAM MILLER reco ded for his shift on only three occasions.118 Plant i i documen . tion indicates that, during this same period, actual ur.identifi>d leakage exceeded 1.0 gpm.119 I i Plant data. was reviewed with Miller, particularly those . I documents relating to makeup tank level, sump level and leakage ! I problems. He rev.i wed the reactor building sump data for ; i 1 January 12 indicating hat there were ten sump pump starts on j J that day.120 He also eviewed data for the month through I January 13 indicating an ngoing unquantified leak in the pressurizer area. Miller ackn wiedged he was probably aware of I this information.121 No leak rate tests were filed on his shift for January 12 or 13, leading us to suspect that unsa isfactory results were obtained and discarded.122 Miller was as d if there was any i jus ti fi ca tion for discarding leak rate test esults indicating uniden ti fied leakage above 1.0 gpm given the ant conditions at that time. He conceded that there would be no ' justification for doing so, considering the seriousness of th leakage problem.125 Miller believed that the only reason test' would have been discarded was that results were above 1.0 gpm, ad it was the practice at the time to discard thos e tests with t entry into the action statement. l i ADAM MILLER ( When confronted with the January 11 test result indicating unidentified leakage to be -0.16 gpm, when other plant conditions indicated leakage to be above 1.0 gpm, Miller stated that any test result exceeding 1.0 gpm would have been -{ discarded.124 He could offer no explanations why a test result demonstrating negative unidentified leakage was filed by , i this shift, other then that the practice was not to pay a great deal of attention to test results because they were believed to be unreliable.125 When asked to explain why he failed to properly respond to the excess leakage in January as required by the Technical Speci fi ca ions , Miller stated, "I can't".126
l 1 s l ADAM MILLER 1 A. Miller 3/20/85, pp. 8,9, 89. 2 Ibid., pp. 9-10. l 3 Ibid., p. 89, j 4 I A. Miller 3/20/85, pp. 32, 51, 73, 88-89; 3/21/85, p. 41. 5 A Miller 3/20/85, pp. 35-36. 6 Ibid., pp. 36-37. 7 A. Miller 3/20/85, p. 33. 8 Ibid., pp. 63, 64-65, 68. 9 A. Miller 3/20/85, p. 66; 3/21/85, p. 74. 1 1 10 A. Miller 3/20/85, pp. 54, 58-59, 70. I 11 Zbid., p. 65, 69. 12 Ibid., pp. 69-70. I l 13 A. Miller 3/21/85, p. 4. 14 Ibid., p. 5. l 15 A. Miller 3/20/85', p. 95. 16 Ibid., p. 99. 17 Ibid., pp. 74-84. 18 Ibid., pp. 59, 61-62, 64-65, 74-75, 91. 19 Ibid., pp. 58-59, 60, 70, 94. 20 Vol. III, Table 2. 21 Ibid. 22 Ibid. 23 Ibid., Table 2.
I ADAM MILLER 24 ;Vol. IV (MPR Report), Appendix K, Test Numbers 213,- I 210, 204, 200, 192, 125, 123, 92, 88. , l 25 Vol. III, Table 2. ! 26 Vol. IV (MPR Report), Appendix K, Test Numbers 218, 213, 210, 204, 200, 192. ; 27 Ibid., Test Numbers 92, 88. ' 28 Ibid., Test Numbers 29, 27, 26,'25, 21, 20, 19, 18, 17, 16, 14, 12, 1. 29 A. Miller 3/20/85, pp. 43-44. 30 Ibid., pp. 49-50. , 31 Ibid., pp. 53, 56. 32 Ibid., p. 57. 1 33 A. Miller 3/21/85, p. 42. l l 34 A. Miller 3/20/85, pp. 93-94. j 35 A. Miller 3/21/85, pp. 5-6. 36 A. Miller 3/22/85, pp. 10-11, 12, 32-33. , 37 A. Miller 3/21/85, p. 11. 38 Vol. IV (MPR Report), Appendix K, Test Number 88.- 39 Ibid, k 40 A. Miller 3/22/85, pp. 4-8. 41 Vol. IV (MPR Report), Appendix K, Test Numbers 70, 65. l 42 A. Miller 3/22/85, pp. 54-56, 63. I 43 Vol. III, Table 4. ' 44 A. Miller 3/22/85, p. 56. i 45 A. Miller 3/20/85, p. 87; 3/21/85, p. 113; , 3/22/85, p. 17. 1
ADAM MILLER 46 A. Miller 3/21/85, p. 113. 47 Vol. IV (MPR Report), Appendix K, Test Numbers 108, 104, 101, 98. 48 Ibid., Test Numbers 92, 88. 49 Shift Supervisor Turnover Notes, December 20, 1978, (Tab 10). 50 A. Miller 3/22/85, pp. 16-17. 51 Ibid., pp. B-9. 52 A. Miller 3/21/85, pp. 30-32, 35-38. 53 Ibid., p. 39. 54 Coleman 2/5/85, pp. 63, 86. 55 A. Miller 3/21/85, pp. 45-46, 50-51, 54, 55; 3/22/85, pp. 87-88, 108. 56 A. Miller 3/21/85, p. 53. 57 Vol. IV (MPR Report), Appendix K, Test Numbers 29, 27, 26, 25, 21, 20, 19, 18, 17, 16, 14, 12. 58 Vol. III, Table 1. 59 vol, Iy (Mpp peport), Section III, and Figure IV-20. 60 A. Miller 3/21/85, pp. 33-35. 61 Ibid., pp. 28-29. 62 Ibid., pp. 29-30. 63 A. Miller 3/22/85, p. 107. 64 Ibid., p. 107. 65 ! Ibid., pp. 49-50, 54, 98, 124-125, 66 Vol. IV (MPR Report), Appendix K, Test Numbers 29, 27, l 21, 20, 19, 18, 16, 14. 67 Ibid., Appendix K, Test Number 36. l
ADAM MILLER 68 Vol. IV (MPR), Appendix K, Test Numbers 36, 26, 25, 17, 12, 1. 69 A. Miller 3/22/85, pp. 129-130. 70 Ibid., 3/22/85, pp. 51, 122. 71 Ibid., pp. 110-111, 120-123, 127-128. 72 Ibid., p. 52. 73 Ibid., pp. 49-52, 108, 127. 74 Ibid., p. 97. 75 A. Miller 3/21/85, p. 46; 3/22/85, pp. 51, 108. 76 Ibid., pp. 53, 93-95. 77 J. Floyd, Operations Memo, January 25, 1979, (Tab 55). 78 A. Miller, 3/22/85, pp. 84-85. 79 Ibid., pp. 85-86. 80 Ibid., pp. 112-113. 81 ' Ibid., pp. 113-115. 82 Ibid., pp. 115-120. 83 Vol. IV (MPR Report), Appendix K, Test Numbers 36, 29, 27, 26, 25, 21, 20, 19, 18, 17, 16, 14, 12, 1. 84 A. Miller 3/22/85, pp. 122-123, 134-135. 85 Ibid., 3/22/85, p. 134. 86 Vol. IV (MPR Report), Appendix K, Test Number 122, 87 Coleman 2/5/85, pp. 51-52. 88 Ibid., pp. 63, 86. 89 A. Miller 3/21/85, p. 102. 90 A. Miller 3/22/85, pp. 104-105. l l
ADAM MILLER 91 Shift Supervisor Turnover Notes, February 19, 1979, (Tab 10). 92 A. Miller 3/22/85, pp. 105-106. 93 Vol. IV (MPR Report), Appendix K, Test Numbers 161, 108, 88, 80, 70, 19, 14. 94 Ibid., Test Numbers 88, 19. 95 Ibid., Test Number 161. 96 Ibid., Test Number 108. 97 Ibid., Test Numbers 80, 70, 14. 98 A. Miller 3/21/85, p. 103; 3/22/85, pp. 9-10. 99 A. Miller 3/21/85, p. 109. 100 Vol. III, Table 2. 101 A. Miller 3/20/85, pp. 9-10. 102 Ibid., pp. 8-9, 42, 89. 103 Ibid., p. 42. 104 Vol. IV (MPR Report), Appendix K, Test Number 149. 105 Vol. III, Tables 1 and Table 4. 106 Ibid., Table 2. 107 Vol. IV (MPR Report), Appendix K, Test Number 148. 108 Vol. III, Table 4. 109 A. Miller 3/21/85, p. 95. ' 110 Ibid., p. 92. ! 111 A. Miller 3/21/85, p. 92. 112 lbid., pp. 96-97. 113 Ibid., p. 97. k z e g $3 g e-m mm mummui is in m
1 i ADAM MILLER 114 Ibid., pp. 62-63. 115 Ibid., p. 63. 116 Ibid., p. 71. l 117 Vol. III, Table 2. ! 118 Ibid.; Test Numbers 80, 70, Vol. IV (MPR Report), Appendix K, 65. 119 Vol. III, Table 4. l Vol. IV (MPR Report), Figure IV-20 i 120 TMI-2 Shift & Daily Surveillance Data Sheets, January 12, 1979, (Tab 40). 121 A. Miller 3/22/85, p. 69, Vol. III, Table 2.
]i 122 j 123 A. Miller 3/21/85, p. 71. !
Vol. IV (MPR Report), Appendix K, Test Number 65; 124 W A. Miller, 3/22/85, pp. 61-62. 125 A. Miller 3/22/85., p. 56-57. 1 126 Ibid., pp. 83-84. 1 I
GARY P. MILLER began employment with Metropolitan Edison Company (Met-Ed) in 1972. He is presently employed by Met-Ed as Director of Operations. Between March 1978 and March 1979, Miller functioned as the Station Superintendent for TMI. During 1978, as Station Superintendent, he reported directly to Lawrence Lawyer, then Manager of Operations for Met-Ed. In early 1979, his title was changed to Station Manager. Thereafter, Miller reported directly to John Herbein, then Vice President - Generation for Met-Ed. We did not ncover sufficient evidence to conclude that Miller was involved in, or tolerated, improper leak rate testing practices. As Station Superintendent, Miller had no direct involvement in leak rate testing. Miller testified he was not aware leak rate test results were being discarded at TMI-2.1 It was his belief that all surveillance records were retained and collected by the Surveillance Coordinator.2 Further, Miller testified that he was not aware operators were having i difficulty obtaining satisfactory leak rate test results at any time.3 1 l 1
i l GARY P. MILLER Miller stated he had no knowledge in 1978-1979 that some operators were manipulating the results of leak rate tests by adding hydrogen and water during the course of the test.4 Miller testified that if he had been aware of this activity he would have instituted disciplinary action against those individuals.5 We reviewed all available testimony from Shift Foremen and ! Shift Supervisors, as well as testimony from Joseph Logan, James Floyd and Jack Herbein in an effort to determine if .{ l anyone attributed to Miller direct knowledge of leak rate test ! improprieties. In no case did we find any testimony that Miller had direct knowledge of the discarding of leak rate test results, filing inaccurate test results, or that tests were manipulated by Operations personnel. i l Although Miller testified that he was not aware of improprieties in the performance of leak rate tests, he did acknowledge familiarity with the Technical Specifications on limiting conditions for operation ( LCO ' s) for reactor coolant system leakage and with the procedural bases for leak rate testing at TMI-2. Miller understood it was necessary to reduce the leakaae or to shut the plant down if the LCO's were exceeded.0 2-
GARY P. MILLER Further, Miller stated that he was familiar with the Exception'and Deficiency procedure and that it applied to leak rate test .results.7 He believed that operators- were following this procedure.0 Miller .could' neither recal'1 ever having. serious doubts about the leak rate test's reliability, nor having anyone-else express such doubts . to him.9 He expressed his own belief .in the basic reliability of the test results, although he recoanized that they were subject to variability from a variety of causes that were difficult to control.10 In an effort to understand his lack of knowledge'about the leak rate test improprieties at TMI-2, we questioned Miller about his daily activities- and responsibilities. As Station. Superintendent, Miller .had the responsibility of supervising the activities at both TMI-l and 2. However, in addition to his duties as Station Superintendent, Miller acknowledged that he also functioned as the Unit Superintendent for TMI-2 until December 1978.11 Miller testified that while he functioned in this capacity, he relied greatly on James Seelinger to perform the day-to-day duties as Unit Superintendent of TMI-2.12 l I i GARY P. MILLER Miller testified that up until Logan assumed responsibility i for TMI-2 in late December 1978, he typically worked 12 hours a day. During this time, he handled personnel problems, security matters, health physics problems, and any significant problems that arose at either Unit. Likewise, during this period, he was also deeply involved in planning for the TMI-l refueling and the commercial startup of TMI-2.13 As Station i Superintendent, Miller said he spent between 20 to 40 percent l ( of his time in Readino, Pa., at meetinos with Company management.14 i Miller testified that his day usually started with a morning conference call between TMI personnel and Management in Readina. Both Unit Superintendents participated in the j call.15 The purpose of the call was to allow the Unit i l Superintendents to advise Miller and others of various problems that they were concerned with at each Unit. During these conversations, those participating had a copy of the Daily I l Plant Status Report to which they could refer.16 i According to Miller, it was his practice to try to visit the control room once per week, usually on the 3pm-llpm I shift.1 Miller testified that although he visited the control room he did not recall observino the performance of 4-
GARY P. MILLER -l leak rate tests.18 Miller testified that he knew of the l company practice of performing a leak rate test on each ! l shift.19 However, it was not his practice to review those l l results.20 It was not his practice to attend Plan of the Day (POD) meetings or Plant Operations Review Committee (PORC) meetings.21 However, Miller testified that he did attend and conduct Shift Supervisor and Shift Foreman meetings.22 i l I We reviewed Shift Supervisor Meeting Minutes for any information which might indicate that there were problems with the leak rate test procedure or related to the issue of discarding leak rate test results. Those Minutes indicated that personnel matters were the main topic of discussion and did not refer to leak- rate test problems. Indeed, Miller I testified that it was his recollection that those meetings focused on personnel matters more than anything else.23 Since Miller's daily activities did not indicate direct knowledge of improper leak rate test pram ices, and there was no testimony that he had such knowledge, we examined the plant documentation and other sources of information then available to Miller which might have evidenced improper leak rate test
._ --.___----__--______a
GARY P. MILLER practices to him. In addition, we questioned Miller at length about the contents of the discussions which took place during the morning conference call. For the most part, the plant documentation consists of the Daily Plant Status Report and the Shift Supervisor turnover notes which werr_ usually attached. Since Miller testified that he attended the Shift Pupervisor and Shift Foreman meetinos, the minutes of those meetings were also examined. Miller was also questioned about his knowledge and involvement with LER 78-62. Miller stated that he relied on the mornino conference call as the primary means of keeping abreast of plant conditions and problems.24 The initiative for issues to be discussed during this call came primarily from the Unit Superintendents.25 Given the schedules of the individuals involved, usually only ' significant matters were discussed.26 The parties taking part in the conversation usually had'available for their review copies of the Daily Plant Status Report. It was not the practice to discuss all the items in this report.28 Miller explained that surveillance procedures were not normally focused on as a topic of discussion.29 He did not 1 l
l GARY P. MILLER recall any discussion about the inability to obtain satisfactory leak rate test results.30 However, he acknowledged that it was likely that problems related to reactor coolant system leakage, and the need to shut the plant down, would be discussed. Nevertheless, he had no specific recollection of such a discussion.31 t Miller recalled receiving the Daily Plant Status Report, but he had no recollection of receiving the Shift Supervisor
)
turnover notes attached to the report.32 l The minutes for the meeting held on October 5, 1977, were 'l reviewed with him. Item 3 under "New Items" in that report states: Mornino Report '[ Daily Plant Status Report} . It was decided that a xeroxed copy of the Shift Supervisor turnover will be hand carried to G.P. Miller and G.A. Kunder to eliminate so many phone calls in the morning. Action: Shift Supervisor.33 Miller had no recollection of the document or the meeting; but, after reviewina its contents, conceded that he probably did receive copies of the Shift Supervisor turnover notes.34 However, he stated he sas not certain that he read them and reiterated his belief that he relied primarily on the morning phone conversation to keep abreast of plant conditions.35
1 l GARY P. MILLER l < l 1 ? Further, Miller testified that, even if he had received them, i he would have relied on the individual Unit Superintendents to review Shift Supervisor turnover notes for items of significance.36 Miller admitted receiving the Daily Plant Status Report, and acknowledged that his name appeared to be signed on the document. However, he testified that his name was usually sianed on the report by the Shift Supervisor submitting it.37 According to him, the placement of his name on the report was a mere formality and did not mean that he reviewed its contents before it was submitted to management in O Reading. Miller said he was aware that the left hand portion of the l Daily Plant Status Report referrina to " primary leakage" was the unidentified leakage figure for the plant.39 On several occasions, successive Daily Plant Status Reports stated an identical figure for " primary leakage." The figure from an j earlier test was carried for a number of days.40 Despite his knowledge of the practice of performing a leak rate test on each shift, he had no recollection of observina the same 1 primary leakaae fiqure reported on successive days.41 On other occasions, there were significant differences in this
.- l GARY P. MILLER l fioure on successive reports. Miller . did not recall observing q primary leakage figures which changed drastically from 2
day-to-day. When questioned about this, he testified ~ that he expected a certain degree of variability in results from.one test to the next.43 Miller testified that in reviewing the Daily Plant Status Report he was more concerned with information contained on the right hand side, since that is where problem situations were likely to be indicated.44 He relied more on the Surveillance Coordinator to track leak rate test deficiencies and maintain records of surveillance completion.45 It should be noted that the only tracking performed by the Surveillance (GMS) Coordinator was to determine whether time limits for filing a-test were exceeded.46 It is 11Aely thu?. Miller received the Daily Plant Status Report and the Shift Supervisor turnover notes which were attached. He therefore would have had accese to reports and turnover notes containino information indicating leak rate test problems and irregularities. The Daily Plant Status Reports for December 31, 1978, January 3, and January 13, 1979, were reviewed with Miller.47
I I i GARY P. MILLER I Despite the references to serious leakaQO problems, Miller did not recall any problems during that period.48 q The Shift Supervisor turnover notes attached to the Daily , 1 Plant Status Report of January 13 state, "need leak rate". l k Miller testified he did not recall seeing turnover notes with i that terminology.49 He further stated that he expected that i any problem in obtaining satisfactory leak rate test results j l ove- an extended period of time would have been brought to his l attention. However, he had no recollection of this l occurring.50 Plant documentation such as Radiation Work Permits, sump data, and makeup tank level indication were reviewed with I Miller for the time period of January 9 through 13.51 This documentation clearly i~ndicated that unidentified leakage was above 1.0 gpm. Also, Miller was advised that over 60 percent of the shifts working during the month of January 1979 failed to file leak rate test results. Miller acknowledged that this was also the type of problem situation he expected to be brought to his attention:. However, he had no recollection of being advised of the problem. GARY P. MILLER l When it was pointed out to him that test results above 1.0 ' gpm were most likely discarded during this period despite the clear indications that actual unidentified leakage was above 1.0 gpm, Miller stated that had he been aware of this he would 4 have considered it improper conduct at that time.53 We also reviewed with Miller the Daily Plant Status Report i
.I and turnover notes for March 27, 1979. The turnover notes j l
stated, "didn't get good leak rate yet. Need one by tomorrow
....54 Primary leakage reported on the Daily Plant Status j 4
Report indicates unidentified leakage to be 0.4 qpm.55 j
)
J q l Miller acknowledged that as Station Superintendent he would j have assumed, upon reading the report, that the 0.4 gpm primary ) leakage ficure was obtained on the morning shift of March 27. Miller could not explain why the turnover note for March 27 reflected 1 the inability to obtain a satisfactory leak rate j while the report reflects unidentified leakage of 0.4 gpm.56 Miller conceded that, given his practice at the time, he most likely did not read the turnover note, but relied on the 1 morning phone call instead for information on plant status.57 ll In the absence of any evidence that problems with leak rate tests were discussed during the morning conference calls, and l l
.4 I
l l
.i
GARY P. MILLER qiven the fact that M'.11er did not focus upon leak rate matters contained in docus.ents submitted to him regularly, it is unlikely that Miller would have become aware of improper leak rate test practices in the ordinary course of performing his routine duties. However, since the report of an unusual event may have provided information to Miller about improper leak rate practices, we questioned him about LER 78-62. As discussed in detail in Volume I, LER 78-62 arose out of an NRC Inspection which discovered that leak rate test results showing unidentified leakage in excess of 1.0 gpm had been obtained and not filed, and that required action was not taken in response to such results. The NRC Inspector's discovery was then brought to the attention of PORC which determined that a reportable event had occurred and directed that an LER be prepared and submitted to the NRC. The LER was prepared under the direction and supervision of James Seelinger. The LER inaccurately described 'the events which gave rise to the LER and misstated the means by which unidentified leakage had been reduced to within allowable ! limits. Although Miller stated that he had no recollection of the LER, he conceded that he was most likely aware of it and l l
)
le 12 -
i l 1 l l I i GARY P. MILLER 1 i involved in discussions pertaining to it.58 However, we have I been unable to discover any evidence which indicates that ? Miller was advised by James Seelinaer, or anyone else, of the j facts underlying the LER, or the means by which the unidentified leakage was reduced. As far as we can determine, the LER conveyed to Miller only the information that was apparent on its face - that operators had obtained leak rate test results in excess of 1.0 gpm, that the unidentified i leakage had been reduced by determining a portion of this l leakage to be identified leakage from the reactor coolant i system, and that they had not taken the required action. Accordingly, we have found no evidence to suggest that Miller became aware of improper leak rate testing practices as a result of LER 78-62. The only other evidence we found from which it could be I
\
inferred that Miller was aware of, or consciously tolerated, improper leak rate test practices, consists of the U.S. Attorney's Statement placed on the record at the time Met-Ed entered its pleas in the criminal matter. In that statement, the U.S. Attorney briefly describes a telephone conversation that allegedly occurred after the October 18, 1978, NRC inspection, which may have placed Miller, or Herbein, on notice
1
]
GARY P. MILLER that there were problems obtaining satisfactory leak rate test l results at TMI-2. The.U.S. Attorney's statement says that: . i During the conversation' the operations f. personnel alerted the Station Superintendent and/or Vice-President for Generation that because of the numerous " bad" leak rate tests obtained at Unit 2, the NRC's interpretation of leak rate technical specifications . would result in . repeated shutdowns of the facility. The Company never advised the NRC that its ., interpretation of the technical -l specifications would repeatedly. shut down TMI Unit 2. Nor would the NRC otherwise expect such consequences since it was not informed that the leak rate test did not' function properly.59 Apparently, the U.S. ' Attorney was uncertain about who participated in that conversation since it is stated, "The call was made to either TMI's Station Superintendent or Metropolit.an Edison's Vice-President for Generation, or botic." Miller had no recollection of the incident referred to in the U.S. 1 Attorney's statement, nor did he recall conversations with any j
.i Operations personnel similar to that referred to in the. U.S.
Attorney's statement. None of the other witnesses we interviewed could recall participating in, or being aware : of, such a . conversation. We have not been able ' to discover any j other evidence that indicates Miller participated in such a telephone call. . i
'4 l
1
GARY P. MILLER Assuming, however, that a conversation did in fact occur as described by the U.S. Attorney, it would have alerted Miller to the fact that operators were having difficulty obtaining satisfactory leak rate test results. This is essentially the same information he could have learned from a careful reading of Daily Plant Status Reports and turnover notes. However, it would not have put Miller on notice of the improper leak rate i test practices that were occurring. Accordingly, we cannot conclude that Miller participated in, or consciously tolerated, improper leak rate test practices. 1 1 I l l l l
'(
GARY P. MILLER i 1 Miller, G. 2/7/84, p. 23; 2/15/85, p. 28.
~2 Ibid., p. 29.
3 Ibid., p.-32. 4 Miller, G. 2/7/84, pp. 38, 40; 2/15/85, pp. 42-43, 76. 5 Miller, G. 2/7/84, p. 41; 2/15/85, p. 43. 6 Ibid., pp. 19-20, 35. 7 Ibid., pp. 33-36. G Miller, G. 2/7/84,-pp. 22-23, 26-27; 2/15/85, pp. 33-36, 42. f-9 Ibid., p. 59. 10 Ibid., p. 57 11 Ibid., p. 3. 12 Ibid., pp. 4, 78. 13 Ibid., p. 6. 14 Ibid., p. 7. 15 Ibid., p. 8. 16 Ibid., p. 9. 17 Ibid., p. 28, 18 Miller, G. 2/7/84, p. 13. 19 Miller, G. 2/15/85,.p. 21. 20 Miller, G. 2/7/84, p. 25. 21 Miller, G. 2/7/84, p. 21; 2/15/85, pp. 25, 37. 22 Ibid., p. 14. [ corrected-1/17/86)
'T 1
GARY P. MILLER 23 Miller, G. 2/15/85, p. 26. 24 Ibid., pp. 8-9. 25 Ibid., p. 24. 26 Miller, G. 2/7/84, p. 11; 2/15/85, p. 23. 27 Miller, G. 2/7/84, p. 12. 28 Miller, G. 2/15/85, pp. 23-24. 29 Miller, G. 2/7/84, p. 12. 30 Miller, G. 2/15/85, p. 69. 31 Ibid., pp. 11, 25. 32 Ibid., pp. 11-12. 33 Shift Supervisors' Meeting Minutes, October 5, 1977, (Tab 23). 34 Miller, G. 2/15/85, pp. 17-18, 45-46. l 35 Ibid., pp. 12, 18. l 36 Ibid., p. 13. 37 Ibid., p. 14 38 Ibid., p. 14. 39 Ibid., p. 22. 40 TMI-2 Daily Plant Status Reports, (Tab 10); Vol. III, Table 4. 41 Miller, G., 2/15/85, p. 31. 42 Ibid., p. 56. 43 Ibid., p. 57. 44 Ibid., pp. 23, 32. l
~ ~
[ corrected 1/17/86)
.1
4 GARY P. MILLER 45 Miller, G. 2/15/85, p. 42. 46 Ibid., pp. 29-30; Beeman 9/29/83, p. 8. 47 TMI-2 Daily Plant orts and Shift Supervisor j
' Turnover Notes, 12/31/78, Status 1 Rep /3/79, (Tab 10).
48 Ibid.,.pp.-53, 67. 49 Ibid., p.-67 50 Ibid., pp. 69-70. 51 Vol. III, Table 4. 52 Miller, G. 2/15/85, pp. 50, 69-70. l 53 Ibid., p. 66. 54 Daily Plant Status Report and Shift Supervisor Turnover Notes, March 17, 1979, (Tab 10). 55 Ibid., (Tab 10). 56 Miller, G. 2/15/85, p. 72. 57 Ibid., p. 74. 58 Ibid., pp. 37, 74. 59 Statement of' Facts Submitted by the United States, (United States of America v. Metropolitan Edison Company), February 1984, (Tab 2). 60 Miller, G. 2/15/85, pp. 75-76. { l
'i
[ corrected 1/17/86] a
i GARY P. MILLER
-1 Miller,.G. 2/7/84, p. 23; 4 2/15/85, p. 28.
]
2 Ibid., p. 29. 3 fhid., p. 32. ! 4 Mil er, G. 2/7/84, pp. 38, 40; 2/15/85, pp. 42-43, 76. 1 5 Miller, G. 2/7/84, p. 41; d 2/15/85, p. 43. j i 6 Ibid., pp. 9-20,.35. 7 Ibid., pp. 33 4. 8 Ibid., p. 34. H l 9 Ibid., p. 59. 10 Ibid., p. 57, 11 Ibid., p. 3. I 1 12 Ibid., pp. 4, 78. 13 Ibid., p. 6. 'I 14 Ibid., p. 7. 15 Ibid., p. 8. 16 Ibid., p. 9. 17 Ibid., p. 28. 18 Miller, G. 2/7/84, p.-13. 19 Miller, G. 2/15/85, p. 21. 20 Miller, G. 2/7/84, p. 25. 21 Miller, G. 2/7/84,.p. 21; ; 2/15/85,1pp. 25, 37. l 22 Ibid., p. 14. y 1 16 - _l .n
GARY P. MILLER 23- Miller, G. 2/15/85, p. .26. 24 Ibid.,.pp. 8-9. 25; Tbid., p. 24, 26 Mi ler, G. 2/7/84, p. 11; 2/15/85, p. 23.
'27 Mille , G. 2/7/84, p. 12.
28 Miller, . 2/15/85, pp. 23-24. 29 Miller, G. 2/7/84, p. 12. , 30 Miller, G. 2 5/85, p. 6 9. . 31 Ibid., pp. 11, 5. 32 -Ibid., pp. 11-12. 33 Shift Supervisors' eeting Minutes', October 5, 1977, (Tab 23). 34 Miller, G. 2/15/85, p. 7. 35 Ibid., pp. 12, 17. 36 Ibid., p. 13. 37 Ibid., p. 14. 38 Ibid., p. 14. 39 Ibid., p. 22. 40 TMI-2 Daily Plant Status Reports, (Ta 10); Vol. III, Table 4. 41 Miller, G.,'2/15/85, p. 31. , 42 Ibid., p. 56. ;
.43 Ibid., p. 57.
44 Ibid., p. 32. 1 1
;\
- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ i
'l j
1 GARY P. MILLER. l l 45 Miller, G. 2/15/85, p. 42. L-46 Ibid., opp. 29-30. I 47 TMI-2 Daily Plant Status Reports,and' Shift Supervisor i s . T0 nover Notes, 12/31/78, 1/3/79, (Tab 10)'. 48 Ibi 4, pp. 53, 67. 49 Ibid., . 67. 50 Ibid., pp. 69-70. 51 .vol. III, T le 4. 52 Ibid., p. 50, 53 Ibid., p. 66. 54 Daily Plant Status eport and Shift. Supervisor Turnover Notes, March 17, 197 (Tab 10). 55 I, bid., (Tab 10). 56 Miller, G. 2/15/85, p. 72. 57 Ibid., p. 74. 58 I Ibid., pp. 37, 74. 4
.l 59 Statement of Fac'ts Submitted by t United States, ,
(United States of America V. Metrop litan Edison Company), February 1984, (Tab 2).
]
60 Miller, G. 2/15/85, pp. 75-76. 1 i 1 J 1 I
i i i i i i 1 WILLIAM 'POTTS began employment with Metropolitan Edison l Company in 1970. He is presently employed by GPU Nuclear Corp.- as ' Manager of Project Engineering at TMI-2. From March 1978
.until approximately December 1978, he was Supervisor of i l
Licensing for Metropolitan Edison Company assigned to 1 Generation Headquarters in Reading,- Pa. Wreafter, he became Superintendent - Technical Support at TMI-l and . remained in .l that capacity until after the accident.1 Durino the course l of our investigation, we found no evidence to suggest that j l Potts was involved in or condoned improper leak rate testing l practices. As Supervisor of Licensino in Readina, Potts had virtually ; nothing to do with leak' rate testing. He did not perform any tests personally nor did he exercise any direct responsibility for the oversicht of leak rate testing- at TMI-2. Potts testified that he was not aware of any improper conduct i relating to leak rate testina at TMI-2 during 1978.2 He-i specifically denied being aware of leak rate test manipulation, and stated that he did not know anyone else who was aware' of ' I test manipulation at that time.3 Potts further-stated that he was unaware of any management practice or policy that l i ___-a__1- - - - - - - _ - _
I' l WILLIAM POTTS encouraaed the manipulation of leak rate test results. According to Potts, he was not even aware that there were any problems with the leak rates at TMI-2, and he knew of no corporate decision to keep TMI-2 operating until TMI-l was returned to service after refueling.4 Georoe Troffer, Potts' immediate supervisor, corroborates Potta in this reaard.5 We found nothing in the company records to refute his statements and there is no testimony that links Potts to knowledge ot improper leak rate testing practices at TMI-2. The only connection that we discovered between Potts and leak rate testing during our investigation stemmed from his review of LER 78-62.6 As Supervisor of Licensing, one of his primary responsibilities was to review Licensee Event Reports (LER's) prior to their Gubmission to the NRC. As a result, he reviewed and approved LER 78-62 which conveyed misleading information to the NRC regarding the activities of October 16-18, 1978.8 This LER dealt with the failure of Operations personnel to enter the action statement in response to excessive reactor coolant system leakage on October 16, 1978. Our investigation revealed that the LER did not accurately describe the action that was taken in response to the 2- ) i
WILLIAM POTTS indications of leakage at that time and that some of the misleading language in the LER was probably added after the ; i report had been sent to Troffer for his review and submission to Herbein. In his testimony, Potts conceded that he probably l reviewed LER 78-62 but stated that he had no recollection of it at this time.9 He explained, however, that under the system in place in i 1978, an engineer assigned to his department would have been primarily responsible for reviewing the event in question and drafting the report for Potts' review and submission to George- l Troffer, Manager of Quality Assurance. Troffer would then send it directly to Herbein for Herbein's signature and submission ! to the NRC.10 Potts stated that .his review of LER's was primarily for the purpose of ensuring proper wordino, clarity, timeliness and adequacy of reported action.11 He stated that he would accept the report and facts as being true and did not investi-cate the underlying accuracy of the reports. According to Potts, if he found a report to be incomplete, he would discuss it with the individuals at TMI who were responsible for its preparation.12 If the report was acceptable, he would pass it on to George Troffer and the Generation Review Committee for 3-
WILLIAM POTTS further review and submission- to Herbein. Potts stated that, if changes were made to the report during the review process, they would be docunented in the file.14
.I l
I When questioned specifically about LER 78-62, Potts maintained that he did not remember anything about it. He ] l stated that he could not recall any discussions regarding this. . J LER or any discussions involving reactor coolant system leakage j in excess of 1.0 ppm, which was deemed satisfactory after rounding off computer printed leak rate test results.15 Potts further testified that he did not know how the LER was changed in Reading and why there was no documentation of that fact in the file.16 l He speculated that any one of several people in Reading could have read the original draft of the LER and found it to , be incomplete because it did not explain how unidentified leakage had been reduced, to within satisfactory limita. According to Potts, someone probably would have called, or traveled to, the Island and asked how the reduction in leakage took place before changing the report.17 Potts further said that if the change was agreed upon over the telephone, it might not have been documented, althouah it should have been.18 In 4 - r
1 i
?
WILLIAM P0TTS I l any event, he stated that he was certain that no one in Reading would have changed the LER unilaterally. According to him, it was against all custom and practice for someone from Licensing I to chance LER's without the concurrence of the individuals at TMI.19 His testimony on this issue was corroborated by George Troffer and John Herbein. I J i When confronted with the facts regarding the events of 1 October 16-18 which our investigation revealed, Potts conceded that the LER was misleading in its discussion of the response l by TMI-2 operators to leak rate test results in excess of Technical Specification limits. He also conceded that the LER failed to indicate that unsatisfactory test results were rounded off in an attempt to satisfy the action requirements of the Technical Specifications.21 He said that he did not believe that he, or any of his subordinates, would have approved the final version of the LER had they understood the facts.22 We found no evidence that refutes his testim >ny in this regard, 3 5-
s j_ 'i g. . ; .; q s ,/ 4
'\ ' ;
; '4 ,. ?;t ,'
- 4 e
; I j
1
< ,WlLLIAM POTTS 1 ~i r s
,t 1 potts1/'6/04,p. 5.
2 Ibid., p . 14 , u .
/
3 Ibid., p. 19. 4 Ibid., pp. 18-19. O' Troffnr 3/20/85, p,; 10. - p 6 TM1-2 l'icensee Event Report 78-6?/1[," ( Tabd,,r, 2 8., 29). 7 Potts 3/7/S% p.'3. N 8 TMid2LicenheeEvebtReportNo. 78 762-1T,. November 1, 1978, (Tab 29). ,
,e 9 Potts 3/7/85, p. 9.
' ') - T 10 Ibid.,#' pp . F7 6. 4
,1 , _;
9 1 J. 1 Ibid., p. 3. - 12 Ibid., pp. 3-4. ;, 13 Qi d,. , pp. 5 - t! . , ( 14 Potts 4/18/85, pp 13-12.' 15 Ibid., p. 15. " 16 Ijgf., p. 18. [* ,
/ '
17 ' Ibid., pp. 16-17 18 < Ibid., p. 18, 19 Ibid., p. 17, # 20 Troffer 3/20/85, pp. 34-35; Herbein 4/27/85, pp. 39-40. 21 potts 4/18/85, p. 20. , 22 Ibid., p. 21. 1 [ corrected 1/17/86] l l k l
k l WILLIAM POTTS l 1 Potts 1/6/84, p. 5. f i 2 Ibid., p. 14. j 3 bid., p. 19, 4 Ib\d., pp. 18-19. ; 5 Troff-r 3/20/85, p. 10. I 6 TMI-2 L'censee Event Report 78-62/lT,-(Tabs 28, 29). I 7 Potts 3/7/ 5, p. 3. 8 Document Rev ew Sheet for LER 78-62 1T, (Tab 30). 9 Potts 3/7/85, p 9. 10 Ibid., pp. 5-6. 11 Ibid., p. 3. 12 Ibid., pp. 3-4. 13 Ibid., pp. 5-6. 14 Potts 4/18/85, pp 11-12. 15 Ibid., p. 15. 16 Ibid., p. 18. 17 Ibid., pp. 16-17. 18 Ibid., p. 18. 19 Ibid., p. 17. 20 Troffer 3/20/85, pp. 34-35; Herbein 4/17/85, pp. 39-40. 21 Potts 4/18/85, p. 20. 22 Ibid., p. 21. MICHAEL J. ROSS began employment with Metropolitan Edison Company as a Station Shift Supervisor in 1970. He is presently employed by GPU Nuclear Corporation as Manager of Plant Operations for TMI-1. He began training for that position in January 1978 and has held it continuously since that time. During our investigation we did not find any evidence to suqqest that Ross was. either involved in or condoned improper leak rate testing practices at TMI-2. As Supervisor of Operations for TMI-1, Ross had virtually no involvement with the operation of TMI-2 during the period under investic-tion. For the most part, his exposure to TMI-2 was limited to four-hour periods that he spent there in training for the purpose of maintaining his TMI-2 Senior Operator's Licenea.1 The records pertaining to these training periods show that Ross spent less than 50 hours at TMI-2 between March 1978 and March 1979.2 During most of the time that he was at TMI-2, the plant was not operating at power or he was not present in the control room. Moreover, he does not appear to have been assigned to TMI-2 during any of the high leakage periods that
MICHAEL'J..ROSS we identified, or during any of the shifts in which manipulated leak rate tests were filed. Given his very limited exposure to TMI-2 during these training periods, it cannot be infer.'d from his mere presence that Ross would have become aware of the improper leak rate testing practices. We did not find direct evidence - f rom any other source which suggested that Ross condoned leak rate test misconduct at TMI-2. Ross has denied that he was aware of any such misconduct. Specifically, he t es ti fied that he did not know that any operators had added hydrogen or water to the makeup tank at TMI-2 in order to manipulate test results. He added that he had no reason to believe that other high level managers were aware of test manipulation. Ross also stated that he did not discuss problems pertaining to the performance of THI-2 leak rate tests with James Floyd, Shift Supervisors or Shift Foremen. He pointed out that he did not normally attend TMI-2 PORC meetings and did not attend TMI-2 POD meetings. His testimony in this regard is supported by the testimony of Floyd.b Ross acknowledged that he did attend staff meetings with TMI-2 managers from time to time, but said that they did not discuss leak rates or routine day-to-day operations at 1 2- [ corrected 1/17/86] 4 i
.I
( MICHAEL J. ROSS these meetings. Instead, they focused primarily on matters relating to the planning and scheduling of Unit operations. Ross stated that, to his knowledge, management had made no decision to continue to operate TMI-2 with excessive leakage until TMI-1 completed refueling and was placed back into operation.10 He also said that he did not know of any Shift Supervisor recommendation to shut down TMI-2 that was overruled because TMI-1 was out of service at the same time.11 We found no evidence to refute his testimony. Several TMI employees were questioned specifically regarding Ross's know-I ledge of improper leak rate testing practice, but no one has testified directly that Ross was aware of any wrongdoing. We reviewed all available TMI-2 records related to leak rate test misconduct, but found none that would have been reviewed by Ross and put him on notice of the misconduct. Mark Coleman was the only employee whose testimony even remotel:r implicated Ross in improper conduct. Coleman test-i fi ed that when he first began performing leak rate tests at TMI-2, he turned in a test result showing unidentified leakage in excess of 1.0 gpm and his Shift Supervisor told him that, "We don't want to see this shit around here." Coleman [ corrected 1/17/86] l 0
\
l
s MICHAEL J. ROSS l we dentified, or during any of the shifts in which manipulated l leak ate tests were filed.3 j l Given his very limited exposure to TMI-2 during these training pe ods, it cannot be inferred from his mere presence that Ross wou d have become aware of the improper leak rate-testing practice . We did not find direct evidence from any ; I other source which succested that Ross condoned leak rate test misconduct at TMI-2. < Ross has denied that e was aware of any such misconduct. { Specifically, he testified that he did not know that any l operators had added hydrogen r water to the makeup tank at ; TMI-2 in order to manipulats te results.4 He added that he had no reason to believe that oth r high level managers were l l aware of test manipulation.5 Ross 1so stated that he. did ( l not discuss problems pertaining to th performance of TMI-2 l leak rate tests with James Floyd, Shift upervisors or Shift Foremen.6 He pointed out that he did not attend TMI-2 PORC or POD meetings.7 His testimony in this rec d is supported by the testimony of Floyd.8 Ross acknowledged that he did 3 attend staff meetings with TMI-2 managers f rom ' ti _ to time, j l but said that they did not discuss leak rates or routine j day-to-day operations at these meetings. Instead, they cused y 1
( MICHAEL J. ROSS i prim rily on matters relating to the planning and scheduling of Unit o rations.9 Ross St tad that, to his knowledge, management had made no decision to ontinue to operate TMI-2 with excessive leakage refueling an6 was placed back into until TMI-l c pleted operation.10 He 1so said that he did not know of any Shift Supervisor recommen tion to shut down TMI-2 that was overruled because TMI-l was out service at the same time.11 refute his testimony. Several TMI We found no evidence t employees were questioned sp cifically regarding Ross's know-ladge of imprqper leak rate t eting practice, but no one has aware of any wrongdoing.12 testified directly that Ross wa We reviewed all available TMI-2 re ords related to leak rate test misconduct, but found none that would have been reviewed nduct. by Ross and put him on notice of the mis e testimony even Mark Coleman was the only employee wh remotely implicated Ross in improper conduct. Coleman test-ified that when he first began performing leak ate tests at TMI-2, he turned in a test result showing unidenti 'ed leakage im that, in excess of 1.0 gpm and his Shift Supervisor told shit around here."13 oleman "We don't want to see this 3-
MICHAEL J. ROSS j f i testified that Ross and Gregory Hitz were his first Shift )1 Supervisors, but that he couldn't remember specifically whether one of them was involved in this incident.14 i Despite Coleman's lack of recollection, it is clear that Ross was not the Supervisor who participated in the conversation. The first leak rate tests at TMI-2 were not performed until March 1978.15 By that time, Ross was in training to assume his position at TMI-1. TMI records reveal that he had been replaced by Hitz as Coleman's shift Supervisor . by January 1978.16 { I 4 i l I l 1 l j
' MICHAEL"J. ROSS 1 Ross-1/25/84, p. 8.
2 TMI On-The-Job Training Reviews / Experience' Records for M.J. Ross,.1978 and.1979, (Tab 54). I 3 ' Ibid., (Tab 54); Vol. III, Table'1; Vol. IV (MPR Report), Figure IV-20. 4 Ross'1/25/84, pp. 74, 78, 80-81. 5 Ibid., pp. 81, 93. l -~ 6 Ibid., pp. 92-93. 7 Ibid., pp. 26, 28. O Floyd 3/30/84, pp. 13-14. ! 9 Ross 1/25/84, pp. 9, 93. 10 Ibid., pp. 85-86. 11 Ibid., p. 86. 12 Smith 3/26/84, pp. 24, 42; Mehler 3/27/84,.pp. 32, <4 3 , 77; Hitz 3/29/84, pp. 38-39; ) 4/24/84, p. 57. 13 Coleman 4/25/84, pp. 5-6. 14 Coleman 2/5/85, pp. 33-34, 37, 44-45; 4/25/84, pp. 6-8. 15 Vol. III, Table 1. 16 TMI Units 1 and 2 Shift Assignment January 6, 1978, (Tab 4); Coleman 2/5/85, p. 46; ; Hitz 3/29/84, pp. 30-31. i [ corrected 1/17/86]- j
MICHAEL J. ROSS 1 Ross 1/25/84, p. 8. 2 TMI On-The-Job Training Reviews / Experience Records for M.J. Ross, 1978 and 1979, (Tab 54). b 3 I id., (Tab 54'; Vol III, Table 1; Vol. (MPR Report), Fiqure IV-20. 4 Ross 1/2 84, pp. 74, 78, 80-81. 5 Ibid., p. 1. 6 Ibid., pp. 9 -93. 7 Ibid., pp. 26, 8. 8 Floyd 3/30/84, pp. 13-14. 9 Ross 1/25/84, pp. 9, 3. 10 Ibid., pp. 85-86, 11 Ibid., p. 86. 12 Smith 3/26/84, pp. 24, 42; Mehler 3/27/84, pp. 32, 43, 7 - Hitz 3/29/84, pp. 38-39; 4/24/84, p. 57. 13 Coleman 4/25/84, pp. 5-6. 14 Coleman 2/5/85, pp. 33-34, 37, 44-45; 4/25/84, pp. 6-8. 15 Vol. III, Table 1. 16' TMI Units 1 and 2 Shift Assignment, January 6, 1978, (Tab 4); Coleman 2/5/85, p. 46; Hitz 3/29/84, pp. 30-31.
BERNARD SMITH began employment with Metropolitan Edison Company on October 27, 1958. He is presently employed as the Radwaste Operations Manager at TMI-2. Between March 1978 and March 1979, he was employed as a Shift Supervisor at TMI. At TMI-2, he was normally assigned to supervise a shift comprised of Shift Foreman, Kenneth Hoyt, and Control Room Operators, Raymond Bocher and Harold Hartman. Analysis of employee testimony and plant records reveals that Smith failed to ensure compliance with the Technical Specifications and procedurcs requiring the documentation of reactor coolant system (RCS) leakage. He also failed to ensure that required action was taken in response to indications of excessive reactor coolant system leakage. There is insufficient evidence from which we can conclude with certainty that he participated in, or knowingly tolerated, manipulation of leak rate test results. l 3
. _ _ _ _ _ _ _ _ _ _ _ _ _ 1
1 I I l 1 BERNARD SMITH l J Supervision of Leak Rate Testing It is clear from his testimony, as well as the testimony of others, that Smith condoned the practice of discarding and I unsatisfactory leak rate test results j failing to document l i obtained on his shift. l { i Smith initially testified that it was his understanding Technical Specifications required only that a that the satisfactory leak rate test result be obtained at least once every 72 hours. According to him, it was not necessary to enter the action statement, or proceed to shut down the plant, in response to test results showing unidentified leakage in e:: cess of 1.0 gpm, if such results were obtained within 72 ) hours from the last satisfactory test.1 ; i After continued questioning, he clarified his initial position and ultimately stated that he understood that it was l l necessary to enter the action statement any time that a valid or believable leak rate test result showing unidentified leakage in excess of 1.0 gpm was obtained.2
BERNARD SMITH I r I 1 he did not know if the f Smith also testified that
)
j containment atmosphere monitoring system was an effective tool l and for quantifying reactor coolant system leakage in 1979; that the containment sump monitoring system was helpful only in 1 1 leakage.3
~
of gross He determining approximate amounts acknowledged that the leak rate test was the primary method I measure unidentified reactor coolant that was employed to system leakage at TMI-2.4 q He also acknowledged that as a Shift Supervisor it was partly his responsibility to ensure that leak rate testing was and in accordance with the Technical performed properly Specifications and the Company's procedures. Specifically, he understood that he had the responsibility to evaluate the safety implications of RCS leakage and the authority to shut the plant down if leakage was unsafe or exceeded Technical i l Specification limits.6 Smith spent approximately half of his time in the TMI-2 control room and was kept generally advised of major plant activities and evolutions. He stated that he did not play much of a role in the supervision of leak rate testing. For the most part, leak rate testing was treated as a routine task that had to he completed by the operators with little oversight
BERNARD SMITH f by him.8 Direct supervision of the tests was conducted by l the Shift Foreman, who had primary responsibility to ensure ! that satisfactory results were turned in within required time limits.9 Smith testified that as a Shift Supervisor, he ! generally would only be advised about leak rate testing when , his shift was unable to obtain a satisfactory result. His testimony is confirmed by the testimony of his subordinates who , remember mostly that Smith would repeatedly instruct them to l "get a good leak rate." In any event, despite Smith's limited involvenient in leak rate testing, it is clear that he was well aware of the problems that operators were experiencing in obtaining accurate j and satisfactory results. l Smith testified that he believed that there were serious i problems with the computer program for leak rate testing at . l TMI-2.12 According to him, it was much harder to get good , leak rates at TMI-2 than it was at TMI-1.1 He felt the results that were obtained at TMI-2 were unreliable and often , 1 unbelievable. Although he did not recall specific problems or complaints from the operators, Smith testified that he was sure the operators believed that the inaccuracy of the leak rate test was a serivus problem since they were, in effect, BERNARD SMITH l
" flying blind".1 In fact, Hartman and Hoyt both testified that they complained to Smith about the inability to obtain accurate and satisfactory leak rates. Guthrie also i'
recalled complaining to his Shift Supervisor about this problem but could not remember whether his Shift Supervisor was Smith l' or Bryan at that time.17 Smith admitted that he personally did nothing to correct these problems with the leak rate test that came to his attention. He said he believed that the problem was with the computer program and that it was being worked on by the Company.1 In addition to knowing that the leak rate test results were not accurate, Smith also was aware of, and tolerated, numerous procedural violations in the performance of the f tests. He admitted that he knew that operators were discarding and failing to document unsatisfactory leak rate test results at TMI-2. Smith testified that as long as he could recall, it 1 was common practice to discard invalid leak rate tests at both l TMI-l and TMI-2.19 He believed that the practice originated l at TMI-l at the Shift Supervisor level.20 l l l
BERNARD SMITH He admitted that leak rate tests at TMI-2 were run l repeatedly and discarded until a satisfactory result was l 1 obtained. He estimated that there were probably two or three j tests thrown out for every one that was kept. His estimate was consistent with plant records which reveal that his shift filed approximately 30 leak rate tests in the roughly 80 shifts-that they reported for duty when the plant was operating. Smith acknowledged that he probably threw away some tests himself, but could not recall how many.23 , 1 Smith also testified that he was aware that no log entries ] I or Exception and Deficiency reports were made to document the < tests that were discarded.24 He stated that as far as he was concerned it was not necessary to document discarded tests. Moreover, he stated that it was his belief that the requirement to record Exceptions and Deficiencies did not apply to leak rate tests because they were performed on a per shift 6 basis. , When explaining the reason for the discarding of so many tests, Smith initially testified that test results were discarded at TMI-2 only if they were deemed to be invalid after a review of other plant instruments and plant conditions.27 After considerable questioning, he ultimately conceded there l l l BERNARD SMITH j f was really no way to verify th+ accuracy of leak rate test results and that all tests showing leakage in excess of 1.0 gpm l were deemed to be invalid.28 He admitted that, as a matter ) l of practice, unsatisfactory test results would be discarded and satisfactory tests would be filed, irrespective of actual plant leakage. i According to Hoyt's testimony, Smith not only tolerated this practice but may have enforced it as well. Hoyt stated that after an NRC inspector found a leak rate test lying on a Shift Foreman's desk, he was instructed by his Shift Supervisor to discard " bad" leak rate tests and not leave them lying around. Hoyt stated that he was almost certain that his Shift Supervisor at the time was Smith.30 Operator Response to Excessive Leakage The practice of discarding and failing to document unsatisfactory test results made it difficult to identify every brief period when there were indications that unidentified leakage exceeded 1.0 gpm at TMI-2. However, we did identify, and examine closely, a number of periods where plant data ) indicated that unidentified leakage would have appeared to exceed 1.0 gpm. Smith was working during some of these periods j l
! BERNARD SMITH i
and failed to ensure that action required by the Technical Specifications was taken. The events of these periods are fully discussed in Volume I of this report. For purposes of i this section, these periods are discussed only to the extent that they relate to the action taken by Smith in response to the indications of excessive leakage. It must be recognized at the outset that Smith did not personally perform or review leak rate tests on a regular b basis. Aside from needing the latest test results for insertion in the Daily Plant Status Report, Smith might not have necessarily been advised of the current leak rate unless there was a problem. However, there is considerable testimony and other evidence from which we can infer that Smith would have generally been made aware of problems with leak rate test results, as well as indic.ations of excessive leakage. He essentially admitted this in his testimony. He stated that as a matter of practice, he was informed if operators on his shift could not obtain a satisfactory leak rate test result.31 He also said that reactor coolant system leaks were brought to his ettention in order that he might evaluate them and determine the safety implications of the leakage. g
i BERNARD SMITH According to Smith, he would personally go out and inspect the leakage in an attempt to verify that it was contained in the plant and did not pose a danger to the public.33 On ) occasions, Smith said he would dispatch Auxillary ? other Operators to collect the leakage and to quantify it so that it He acknowledged could be accounted for in leak rate tests. v_ j that as a Shift Supervisor he would be the initial person to evaluate leakage and to decide whether the plant should be shut
. 35 down or continue operating.
Smith testified that he did not recall ever shutting down entering the action statement, because of the plant, or 6 attitude regarding the excessive leakage. Given his Specifications for RCS leakage, this is not Technical Smith's testimony regarding the action surprising. requirements for RCS leakage makes it clear that he did not properly enforce the relevant Technical Specifications in 1978 and 1979. As was pointed out previously, Smith initially testified that he believed that the Technical Specifications required only that a satisfactory leak rate test result be obtained at least once every 72 hours. He indicated that he thought it was not necessary to enter the action statement if intervening test results produced unsatisfactory results. _9_
BERNARD SMITH Later, he clarified his position and acknowledged that he understood that action to reduce the leakage or to shut down the plant was necessary in response to any valid test result that showed unidentified leakage above 1.0 gpm. He also admitted that he had received numerous unsatisfactory leak rate test results throughout 1978-1979 and that he did not direct 7 entry into the action statement."9 He explained, however, that he never really believed that there was leakage "above 1.0 gpm" at TMI-2.40 When pressed further in this regard, he conceded that given the problems with leak rate testing, there was no way to really be certain that leakage was at acceptable limits. When asked how he could justify continuing to operate the plant without being able to measure reactor coolant system leakage, he stated that he felt it was safe to do so. He also stated that he did not agree that the 1.0 gpm limit was the equivalent of safe operation of the plant.43 He indicated that despite their inability to precisely measure RCS leakage, operators could locate and evaluate leakage by making tours of the plant. Smith testified that as long as the leakage was under control and contained inside the plant, he
- l
I l BERNARD SMITH ] I 1 i i felt that it was safe to operate. He stated that he never reached a point at TMI-2 where he believed that leakage was at ] I an unsafe limit.45 l l Smith also explained that he did not recall that anyone any concerns about operating the plant with ever voiced unsatisfactory leak rates.46 Moreover, he stated that he was never instructed by his superiors that it was necessary to in response to all test results enter the action statement 47 showing leak'.ge in excess of 1.0 gpm. He denied that there was any conscious decision to keep TMI-2 operating until TMI-l ] restarted after the refueling cycle.40 l l Smith's testimony is not completely consistent with other i evidence we reviewed during the course of our investigation. 1 Smith worked regularly during the period that included the October NRC inspection, the circulation of Floyd's Operations memo dealing with the RCS leakage action requirements, and the distribution of LER 78-62. Smith stated that he remembers receiving the Operations memo and he acknowledged that he signed the document review form verifying that he had read the LER. He further stated, however, that . he remembered that the Operations memo prov!1ed information regarding the computer BERNARD SMITH program modifications but did not recall the instructions regarding the necessity to enter the action statement in response to unsatisfactory leak rate test results.50 Additionally, he stated that he could not explain how he interpreted the LER in 1978 since he was certain that it never caused any change in the way they "did business".51 He could not recall any discussions or meetings about the LER and could not explain how it failed to bring about any change in the leak rate testing practices at TMI-2.$ It should be noted that NRC Inspector Haverkamp has stated that Smith may have been one of the Supervisors present on the day that he interrupted a discussion regarding unsatisfactory leak rates. However, the schedule for Smith's shift suggests that he was not present at that time but was working on the 11pm-7am shift.54 Others have indicated that they definitely complained to Smith about leak rate testing practices. Hoyt and Hartman both testified that they expressed frustration to him o v r.r being requested to repeatedly run leak rate tests until a satisfactory result could be obtained. Guthrie testified i that on more than one occasion he expressed concern to his Shift Supervisor that the plant was being operated when it was not possible to demonstrate a satisfactory leak rate. He 1 W i
q l l i BERNARD SMITH k l could not remember specifically whether he had those j discussions with Smith or Bryan.$ Our examination of plant documents provides further evidence that Smith failed to ensure that proper action was i taken in response to RCS leakage on the following occasions. January 10-15, 1979 I Smith's shift did not file a leak rate test or. January 10 l or 11, 1979.58 By this time, it was becoming increasingly difficult to obtain a satisfactory result, and there is 1 considerable testimo[ny that it was becoming necessary .to ; discard numerous leak rate tests before obtaining a j satisfactory result.59 There are also indications that l l increased RCS leakage from pressurizer valves and other sources l I
)
was of some concern among Operations personnel.60 I i I I On January 11, a Radiation Work Permit was issued for Hoyt, a Heath Physics Technician, and a Maintenance Foreman, to I l enter the reactor building inside the "D-ring" to check for leaks at the pressurizer. Two other permits were issued for routine entry and inspection on the same day. Sump collection was on the increase and a considerable amount of water was added to the makeup tank during this shift.62 1 I
BERNARD SMITH Plant records also reveal that Hoyt and the others all received radiation doses; however, there is no record of the nature and amount of reactor coolant system leakage that they observed.63 Since there is no leak rate test on file, it is likely that any test that was performed on that day yielded , i unsatisfactory results and was discarded. There is no record i of official entry into the action statement because of RCS leakage. On the contrary, the Control Room Log shows that j 1 power was increased during the shift.64 { 1 On the next two shifts, indications of excessive RCS leakage continued. Plant records reveal that on January 11, at 21:53, Coleman filed a leak rate test that turned out to be l satisfactory only because of a water addition that was made during the test. l On January 12, Smith and his crew again reported for work on the 7am-3pm shift.66 Plant records indicate that excessive RCS leakage continued and no leak rate test was filed.67 The next two shifts were also unable to obtain a satisfactory leak rate test result. On January 13, at approximately 10:40, Hoyt and Hartman filed a leak rate test showing unidentified leakage at 0.2639 r
1 l BERNARD SMITH gpm after almost 40 hours had passed without obtaining a satisfactory result.69 This test was one of the previously discussed tests that were influenced by an unrecorded water addition during the test. Hoyt approved this test despite indications that it was not accurate and that RCS leakage exceeded the limiting condition for operation at the time. l l At 12:45 on the same day, Hoyt and personnel from f 1 Maintenance and Health Physics, again entered the reactor f building to check for leaks at the pressurizer. The Radiation i 1 Work Permit was approved by Smith. I Hoyt's log entry regarding this activity reads as follows: l 1 Made inspection in Rx Bldg. to check leakage j from RC-V-133B, 134B, 135B. Total leakage i from these valves is approximately the same I as Thursdays [ January 11] inspection. Survey of other areas showed no other leakage. Leakage is from body to bonnet areas of these valves and appears to be safe yet to operate.72 Since the leak rate test that was submitted contained no I reference to this leakage, there is no official record quantifying the amount of leakage that was perceived at that time; nor is there any evidence that Smith performed an evaluation of the leakage pursuant to Company procedures.
I l BERNARD SMITH l A memorandum prepared by George Kunder, Superintendent - 1 1 Technice.1 Support, pertaining to the same leakage, estimated the leakage to be approximately 2.0 gpm and characterized it as
" identified". Apparently, it was this characterization that provided the justification to continue operating the l
plant; however, no one on the shift who was working at the time j remembers these events very clearly.74 I
)
l l On January 14, Hoyt again reported for work on the day shift. The two previous shifts had not been able to obtain a satisfactory leak rate test and continued to experience i indications of excessive RCS leakage. Plant records reveal ) that Smith's shift also had not filed a satisfactory leak rate , test. They continued to operate the plant, despite indications that the pressurizer leak and other RCS leakage i
!4 continued. It is clear, from Hoyt's testimony that they had no f
basis to do so.77 j i i When questioned specifically about the events of January l 1 1979, Hoyt stated that he had very little recollection of the ! i O period. He did not remember any discussion with specific l 1 individuals. He did remember entering the reactor building to l 1 very quickly observe a steam leak coming from the pressurizer I I 79 He and estimating the leakage to be approximately 0.5 gpm. j l 1 l 1 a
l BERNARD SMITH conceded that he could not accurately measure the leakage and that. he only provided a very rough estimate of. the leakage because that is probably all he was asked to do so by:his Shift f Supervisor. 1 He also conceded that,.given his inability-to quantify.the leakage in question, and the information available from other plant instruments, there was a good possibility that the plant was operating with unidentified leakage in excess of - 1, 0 - I I gpm. Nevertheless, he could not explain why he did not ] direct entry into the action. statement and proceed to shut down the plant. He said that.he realized now that they probably should have shut down the plant but at that time they must have ) believed that they were operating safely and close t o ,- or. ' under, 1.0 gpm.0 In his testimony, Smith indicated that he had no 1 recollection of the events of January 1979. He did not. remember discussions with anyone regarding excessive RCS ; leakage, nor did he recall any evaluation of leakage for the purpose of deciding. whether to shut down the plant.84 Smith's general testimony,. of ' course, is consistent with his' failure to recommend a plant shutdown. 1 l l _____.__._._________________._.________.____.____m___ _ _ _ _ _ _ . _ _
I l BERNARD SMITH Smith was specifically questioned regarding what he probably would have done in response to the excessive leakage i at that time. He indicated that he would have allowed the continued operation of the plant after verifying that the ! I excessive leakage was contained and making an educated guess gegarding the amount. Although he could not point to anything in the Technical Specification or the company procedures which authorized estimating leakage that could not be measured, he stated that it was the practice at the time.86 Test Manipulation The NRC has alleged that improper water and hydrogen additions were made during several of the leak rate tests that j were perforr:ed by Smith's shift.87 This is not surprising l l since most of the tests run by his shift were run by Hartman and Booher. Hartman, of course, has admitted that he manipulated tests and he has also testified during other investigation.e that Bocher manipulated tests as well.bb Hartman refused to be interviewed during this investigation. Our review of his prior testimony, however, did not uncover statements which clearly implicate Smith in test manipulation. On the contrary, Hartman's testimony, alti.sgh somewhat ambiguous, suggests that Smith was not aware that operators were making water and hydrogen additions during leak rate tests for the purpose of manipulating test results.89 e- 1
+.
i r.; - /
)
BERNARD SMITH i I ,* <
'/ Ird his NRC interview of March 26, 1980, Hartman first said l thrit he did not know if his supervisors were aware of test
% B
"~
man'ih$,lation bf Control Room Operators. He then added that
, /
J fa>eicplaining to Hoyt anc1 Smith why he could not get a good r- ;
/3. leak rate, he'would say things like, "well, you know I don't
-l s want to fudge them and I'd keep telling them, I says I'm tired of wrestling hith tests, with those leak rates. I don't want to, we have got a problem here why don't you do something about
? it." 1 He also stated that Smith would often be on his back , to get a good leak rate and that, knowing Smith, he took it to ,, mean, "gdt one by hook or crook".92 Unfortunately, he never 4 A# explained what he meant by that statement. In his deposition that was taken during a suit between GPU and Babcock & Wilcox, Hartman stated that he did not advise his supervisors when ?- added hydrogen to the makeup tank.93 Finally, during his last interview with the NRC on July l l 26, 1983, Hartman again stated that he was not sure if his I 1 superiors were aware of test manipulation.b4 Later in the same interview, however, he acknowledged that he was not aware of any Foreman or Supervisor who knew that hydrogen was being added to the makeup tank for the purpose of test manipulation. BERNARD SMITH In his testimony, Smith denied that he was aware of test manipulation of any sort by operators working under his supervision.9 Moreover, he stated that he did not believe ) that tests coul6 in fact be successfully manipulated by hydrcgen additions.9 He was corroborated in this respect by Hoyt.9 No one else at TMI-2 has implicated Smith in test manipulation. Our analysis of records pertaining to leak rate tests run under Smith's supervision also failed to produce conclusive evidence that he was aware of test manipulation. We were able to confirm only a few of the improper water and hydrogen additions alleged by the NRC.99 Given the lack of any obvious pattern of systematic test manipulation and the self-concealing nature of the conduct in question, we cannot conclude that Smith was aware of wrongdoing on the part of the operators. Among the numerous tests that were performed on his shift, Smith was only directly involved in the approval of two of them.1
.l I
On October 16, 1978, during the llpm-7am shift, Smith approved a test result obtained by the previous shift and apparently left unsigned for several hours. The NRC has
BERNARD SMITH l alleged that an unrecorded water addition took place during the test; our analysis revealed what appeared to be " feed and bleed" operatiens going on at that time. The test printout ) also contained an error in the value for operator-entered identified. leakage and included a negative value for unidentified leakage. While Smith's approval of this test surely demonstrates his lack of concern for the accuracy of. those tests tt it were- filed, it is unlikely that he would have been involved in, or aware of, test manipulation that may have occurred several hours before he reported for work. Smith testified that he signed the test in question and conceded that it was inaccurate on its face.1 1 He stated, however, that he had no recollection of the particular test, but that he must have signed it because he generally accepted negative numbers at the time.102 Smith also approved a leak rate test performed by Bocher on March 15, 1979, which appears to have been influenced by a , hydrogen addition.103 Hartman has testified that Booher added hydrogen to manipulate tests, but he also testified that Shift Supervisors were not informed of these additions.1 There is no other evidence which suggests that Smith was aware of the addition if, in fact, it'took place.
I i BERNARD SMITH l 1 Smith 2/8/85, pp. 8-9. 2 Ibid., pp. 13-15. 3 Ibid., pp. 18-19. 4 Ibid., p. 19. 5 Ibid., pp. 6-7, 15-16. ) 1 6 Smith 3/26/84, pp. 25-26. l 1 7- Smith 2/8/85, p. 5.
]
8 Ibid., p. 16. 9 Ibid., pp. 22-23. j 1 10 Ibid., p. 16. 11 Hartman 10/19/79, pp. 49-50; Hartman 3/216/80, pp. 28-9; l Coleman 2/5/85, pp. 42-43, 47; Congdon 2/13/85, p. 38.
)
12 Smith 3/26/84, p. 24. l 13 Smith 4/25/84, p. 11. 14 Smith 3/26/84, pp. 52-53; 2/8/85, pp. 19-20 15 Smith 2/8/85, p. 76. 16 Hartman 3/26/80, p. 30; Hoyt 2/26/85, pp. 18-19. l 17 Guthrie 2/12/85, pp. 30, 33. 18 Smith 2/8/85, pp. 75-76; Smith 3/26/84, pp. 23-24. 19 Smith 2/8/85, p. 66. 20 Smith 3/26/84, pp. 13-14. 21 Smith 2/8/85,, p. 70.
BERNARD SMITH 22 Vol. III, Table 2. 23 Smith 3/26/84, p. 20. 24 Ibid., p. 13. 25 Smith 2/8/85, pp. 48-49. 26 Smith 3/26/84, p. 13; 4/25/84, p. 5. 27 Smith 3/26/84, pp. 18-19. 28 Smith 2/8/85, pp. 55-56, 75. 29 Ibid., pp. 65, 72. 30 Hoyt 2/26/85, pp. 22-26. 31 Smith 2/8/85, p. 16. 32 Ibid., p. 26. 33 Ibid., p. 27. 34 Ibid., pp. 28, 58-59. 35 Ibid., pp. 56, 128. 36 Ibid., pp. 95, 98, 12C. 37 Ibid., pp. 9-12. 38 Ibid., p. 14. 39 Ibid., p. 98. 40 Ibid., p. 77. 41 Ibid., p. 75. 42 Ibid., p. 76. 43 Ibid., p. 77. l 44 Ibid., pp. 77-78. 45 Ibid., p. 76. [ corrected 1/17/86] ; I _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ . /
l { i BERNARD SMITH 22 Vol. III, Table 2. j 23 mith 3/26/84, p. 20. ] 24 Ib d., p. 13. 25 Smit p. 49. 26 Smith 26/84, p. 13; 4 5/84, p. 5. 27 Smith 3/2 84, pp. 18-19. 28 Smith 2/8/85 pp. 55-56, 75. 29 Ibid., pp. 65, 72. 30 Hoyt 2/26/85, pp. 22-26, 31 Smith 2/8/85, p. 1 32 Ibid., 26. j
- p. 1 33 Ibid., p. 27.
34 Ibid., pp. 28, 58-59. 35 Ibid., pp. 56, 128. 36 Ibid., pp. 95, 98, 126 37 Ibid., pp. 9-12. l 38 Ibid., p. 14, 39 Ibid., p. 98. f 40 Ibid., p. 77. 41 Ibid., p. 75. 42 Ibid., p. 76. 43 Smith 2/18/85, p. 77. 44 Ibid., pp. 77-78. 45 Ibid., p. 76. I BERNARD SMITH j 46 Smith 4/25/84, p. 12-13. 47 Smith 2/8/85, pp. 96-97. 48 Smith 4/25/84, p. 14; 3/26/84, p. 26. 49 Smith 2/8/85, pp. 99-100, 103-104.
'50 Ibid., pp. 99-100. j l
51 Ibid., pp. 106-107. j l 52 Ibid., pp. 107-108. l I 53 Haverkamp 3/3/85, p. 4. l l 54 TMI Units 1 and 2 Operations Department Shift Assignment, and Shift Schedule, approved August 15, 1 1978, (Tab 4). ] 55 Hoyt 2/26/85, pp. 18-19; Hartman 3/26/80, p. 30. 56 Guthrie 2/12/85, pp. 30, 33. 57 Ibid., p. 33. 58 Vol. III, Table 2. 59 Guthrie 2/12/85, pp. 19-20, 65-67; I congdon 2/13/85, pp. 34-35; 2/21/85, p. 44; conaway/19/85, Faust 2 pp. 134-136; Faust 2/20/85, pp. 37-38; Miller A., 3/22/85, p. 71; Mehler 2/28/85, pp. 81-82, 87; Hoyt 2/26/85, pp. 101-102; Logan 3/27/85, pp. 107-108, 114, 118-119. 60 Vol. III, Table 4. , 61 TMI Radiation Work Permits: January 11, 1979, (Tab 9). 62 TMI- 2 Control Room Log, January 11, 1979, (Tab 48); Vol. IV (MPR Report), Appendix K, Test Number 65.
I BERNARD SMITH l 63 Mulleavy 3/5/85, pp. 15-16; ; Vol. III, Table 4. ' 64 Control Roc n Log, January 11, 1979, (Tab 48). 65 voi, Iy ( gt"6 Report), Appendix K, Test Number 65. 66 Vol. III, Table 2. 67 Daily Plant Status Peport and Shift Supervisor Turnover Notes, January 12, 1979, (Tab 10); TMI Radiation Work Permits, 14331; 14332; 14333; 14334; 14335, January 12, 1979: (Tab 9); Volume III, Table 4. 68 Vol. III, Table 2. 69 Vol. IV (MPR Report), Appendix K, Test Number 64. 70 Hoyt 2/26/85, pp. 107-114. 71 TMI Radiaton Work Permit, Number 14337, January 13, 1979, (Tab 9). 72 TMI-2 Shift Foreman Log, January 13, 1979, (Tab 49). 73 G. A. Kunder, Informal Notes from File " Ops Notes", January 13, 1979, (Tab 45). 74 Hoyt 2/26/85, pp. 118-119; i Guthrie 2/12/85, pp.59-60; I Vol. III, Table 2. 75 Vol. III Tables 1 and 4. 76 Ibid., Table 2. 77 Hoyt 2/26/85, pp. 122-128. 78 Ibid., p. 84. BERNARD SMITH 79 Ibid., pp. 85, 86, 89, 92. CO Ibid., pp. 93-97, 123-124. 81 Ibid., p. 127. G2' Ibid., pp. 102, 129-131. 83 Ibid., p. 130. 84 Smith 2/8/85, pp. 115, 126. 85 Ibid., pp.-127-128. 86- Ibid., p. 129. 87 Vol. IV (MPR Report), Appendix K, Test Numbers 64, 61, 31, 30, 22, 15, 14, 13, 10. 88 Hartman 4/27/80, p. 97; 4/29/80, p. 164; 7/26/83, pp. 1, 2. 89 Hartman 7/26/83, pp. 1-2; 5/22/79, pp. 15-16. 3/26/80, pp. 26-27 90 Hartman 3/26/80, p. 27. 91 Ibid., p. 30. 92 Ibid., p. 28, 93 Hartman 8/18/82, p. 276. 94 Hartman 7/26/83, p. 1. 95 Hartman 7/26/83, p. 2. 96 Smith 3/26/84, pp. 16-17, 22; 2/8/85, pp. 130-131. 97 Smith 3/26/84, pp. 30-31; 98' Hoyt 2/14/85, pp. 32-33; 2/26/85, pp. 67-69. 99 Vol. IV (MPR Report), Appendix K, Test Numbers 64, 30. l ,
BERNARD SMITH 100 Ibid., Test Number 150, 10. l 101 smith 2/8/85, p. 83. 102 Ibid., p. 83. 103 Vol. IV (MPR Report), Appendix K, Test Number 10. 104 Hartman 7/16/82, p. 109; < 7/26/83, p. 2. { l I l 1 [ corrected 1/17/86] j I l
BERNARD SMITH 10 Ibid., Test Number 150, 10. 101 Smith 2/8/85, p. 83. 102 id., p. 83. 103 Vo IV (MPR Report), Appendix K, Test Number 10 106 Hartm 7/16/82, p. 109. 1 4 l l l j
i GEORGE TROFFER began emp]ayment with Metropolitan Edison Company on April 29, 1981. He is presently employed by GPU Nuclear Corp. as a Manager - Site Facilities. Between March 1978 and March 1979, he hold the position of Manager of Quality Assurance at Generation Headquarters in Readina, Pa.1 In j that capacity, Troffer was responsible for the supervision of l I more than one hundred employees working in several departments )
.l at Three Mile Island and other Metropolitan Edison facilities, j i
He reported directly to John Herbein, Vice President for ) Generation at Metropolitan Edison. During the course of our investigation, we found no evidence to suggest that Troffer was involved in, or condoned, improper leak rate testing practices. 1 I As Manager of Quality Assurance in Reading, Troffer had virtually nothing to do with leak rate testing. He did not, of course, perform any tests personally, nor did he exercise any direct responsibility for the oversight of leak rate testing at TMI-2. Troffer testified that he does not remember being aware of any problems associated with the performance of leak rate tests in 1978 or 1979. He testified further that he did not
GEORGE TROFFER remember that there was excessive leakage at that time; nor did he recall any discussions with Herbein or his colleagues in Reading regarding the possibility that TMI-2 might be shut down because of excessive leakage.2 We found nothing in the company records to refute his statements and there is no testimony that links Troffer to knowledge of improper leak rate testing practices at TMI-2. Troffer's only apparent exposure to leak rate testing at TMI-2 came about as a result of hit review of LER 78-62 and his review of an audit of TMI-2 surveillance procedures that were conducted under his supervision. Neither of these two documents by themselves would have made Troffer aware of improper leak rate testing practices at TMI-2. Between November 1978 and January 1979, an audit (number 78-31) was conducted by the Quality Assurance Audit aroup under the direction of Gerald Loignon, a quality assurance engineer, for the purpose of assessing the adequacy of compliance with Technical Specification implementing procedures at TMI Units I and II. The final audit report, which was approved by Troffer, did not cite any deficiencies in the Unit 2 leak rate procedure, nor did it suggest that the audit team had discovered major problems with TMI leak rate testing practices. l l l 4 l
i I GEORGE TROFFER Loignon testified that apparently he did not uncover any 3 deficiencies with respect to leak rate testing at TMI 2. He stated that, if he had, he surely would have included this in his report.4 He attributed his failure to uncover these problems to the fact that his report was completed very shortly after the TMI-2 commercial startup, and therefore did not delve ! into leak rate testing in very great detail.5 He also pointed out that the practice of discarding and failing to l 1 document test results would have been difficult for someone conducting an audit to discover, since the evidence of the wrongdoing was being thrown away.6 In any event, it was clear that Loignon did not teport any leak rate test misconduct to his superiors and that his report would not have put Troffer on notice that there was any such wrongdoing. Troffer acknowledged that he approved the report ' but stated that he had no recollection of it at thi- time. He conceded that the report was somewhat superficial, but defended Loignon as a good auditor who was covering a lot of ground and simply did not spot the problems associated with leak rate testing.0 In addition to approving the audit report, Troffer also reviewed and approved LER 78-62.9 This report conveyed
Q GEORGE TROFFER misleading information to the NRC regarding the activities of October 16-18, 1978.10 This LER dealt with the failure of Operations' personnel to enter the action statement in response *' to excessive reactor coolant system leakage on October 16, 1978. Our investigation revealed that the LER did not describe accurately the action that was taken in response to the s indications of leakage at that time and that some of the misleading language in the LER was probably added after the report had been sent to Troffer for his review and submission to Herbein. Troffer testified that he had no specific recollection of LER 78-62.11 However, he testified that, as a matter of practice, three copies of each LER would be submitted to his # office from TMI and would be automatically distributed thereafter to him and to William Potts, Manager of Licensing.12 Troffer said that normally he would not review the reports at this stage since he expected the Licensing Department to begin the review process automatically without direction from him.13 i According to Troffer, once an LER was received in Reading, Potts would be responsible for ensuring that it was critten clearly and accurately and that it adequately described the
.a. s
GEORGE TROFFER d corrective action that was to be taken.14 If Potts felt that the LER was deficient in some respect, he would work with an assigned licensing engineer and, } if necessary, an individual ] from TMI to correct the wording.15 Troffer further testified that he did not remember any formal cystem for documenting changes to an LER that might be made in this way although he ) conceded that, in hindsight, there should have been some documentation.16 In any event, Troffer said that once Potts was satisfied with the LER, he would submit it to him as Manager of Quality Assurance, as well as to members of the Generation Review Committee (GRC) in advance of one of their regularly scheduled meetings. Troffer stated that he was also Chairman of the GRC.I The LER would then be discussed at a GRC meeting for about 10 minutes or less and the committee would make a recommendation to the Vice President. Troffer conceded that the GRC review was often meaningless, since individual committee members would have already reviewed the LER by virtue of their own positions.10 For example, as Manager of Quality Assurance, Troffer would have been responsible for reviewing LER's and then forwarding them to Herbein for his signature.
GEORGE TROFFER Thus, normally he would have already approved an LER that was pending before the GRC.19 Company records revealed that the LER was mailed to Troffer, reviewed and amended in Reading, signed by Herbein and submitted to the NRC in one day. When confronted with these facts, Troffer said he did not know how this had happened but speculated that perhaps the licensing engineer might have carried a copy to Reading in advance of the mailed copy.20 Troffer denied emphatically that anyone at Reading would have put an innaccurate date on the LER merely to satisfy the time requirements for reporting. According to him, they were absolutely honest with respect to the date and time of correspondence that Herbein sent to the NRC.21 We also asked Trof fer whether the change that was made to the LER in Readina might have been made without an investigation simply to make the LER sound more thorough. Troffer testified that he did not believe this was the case. According to him, the general practice was to be very honest about these matters.22 He was corroborated in this regard by Potts, who also testified that it was not likely that a chance in the LER would have been made without input from TMI.23 Unfortunately, because of the lack of documentation, we cannot I
l GEORGE TROFFER 1 I be cartain who was responsible for the additional language in j the final draft of the LER. There is no evidence from which we a can infer it was Troffer. l l l I i l l
l j GEORGE TROFFER I 1 Troffer 3/20/85, p. 2. 2 Ibid., p. 10. i 3 Met Ed OA Audit Report 78-31, January 24, 1979, (Tab 39). 4 Loignon 4/22/85, pp. 27-29. ! 5 Ibid., pp. 35-36. ! l 6 Ibid., pp. 29. 7 Troffer 3/20/85, pp. 40-41. j 8 Ibid., p. 45. 9 TMI-2 Licensee Event Report 78-62/1T (Draft), November, 1978, (Tab 28); TMI-2 Licensee Event Report 78-62/1T, November 1, 1978, (Tab 29). 10 Ibid. h i 11 Troffer 3/20/85, p. 11. l l 12 Ibid., pp. 22-23. l 13 Ibid., pp. 23-24. k j 14 Ibid., p. 27, 15 Ibid., p. 27. 16 Ibid., pp. 27-28. , i 17 Ibid., pp. 3-4 i j 18 Ibid., pp. 29, 39. l! 19 Ibid., pp. 38-39. j 20 Ibid., p. 35. 21 Ibid., p. 34. } 22 Ibid., p. 49. 23 Potts 4/18/85, p. 17. (corrected 1/17/86) i l l
GEORGE TROFFER l 1 Troffer 3/20/85, p. 2. i 2 Ibid., p. 10. 4 3 t Ed OA Audit Report 78-31, January 24, 1979, (Tab 39).- ) l 4 Lo non 4/22/85, pp. 27-29.
]
5 Ibid, pp. 35-36. )l 1 6 Ibid., p. 29. ]
\
7 Troffer 3 20/85, pp. 43-44. i 8 Ibid., p. 4 . 9 TMI-2 License Event Report 78-62/lT (Draft), November, 1978, (Tab 28), j l TMI-2 Licensee E nt Report 78-62/lT, November 1, 1978, (Tab 29). 10 Document Review Shee for LER 78-62/lT, (Tab 30). 11 Troffer 3/20/85, p. 11 l l 12 Ibid., pp. 22-23. 13 Ibid., pp. 23-24. 14 Ibid., p. 27. 15 Ibid., p. 27. 16 Ibid., pp. 27-28. 17 Ibid., pp. 3-4 18 Ibid., pp. 29, 39. 19 Ibid., pp. 38-39, 20 Ibid., p. 35. 21 Ibid., p. 34. 22 Ibid., p. 49. 23 Potts 4/18/85, p. 17. 1 i WILLIAM ZEWE began employment with Metropolitan Edison l l Company (Met-Ed) on Februar, 14, 1972. He is presently i employed as the Superintendent of Titus Generating Station. Between March 1978 and March 1979, he was employed as a Shift Supervisor at TMI. At TMI-2, he was normally assigned to supervise comprised l a shift of Shift Foreman, Frederick - 1 Scheimann, and Control Room Operators, Craig Faust and Edward l Frederick. 1 Analysis of employee testimony and plant records reveals l that Zewe failed to ensure compliance with the Technical Specifications and procedures requiring the documentation of RCS leakage. He also failed to ensure that required action was taken in response to indications of excessive reactor coolant system (RCS) leakage in October 197e, January and March 1979. There is no evidence that he participated in, or knowingly tolerated, manipulation of leak rate test results.
WILLIAM ZEWE I i Supervision of Leak Rate Testina ! 1 l It is clear from his testimony, as well as the testimony of others, that Zewe condoned the practice of discardino and failing to document unsatisfactory leak rate test results obtained on his shift. 1 Zewe acknowledged that he was aware of the limitino conditions for operation for RCS leakage at TMI-2 and that there was a requirement to proceed towards a plant shutdown in the event that unidentified leakage exceeded 1.0 opm for more than several hours.1 As a Shift Supervisor, Zewe under-stood that it was partially his responsibility to ensure compliance with these requirements. He stated that he believed that neither the containment sump nor the containment atmosphere monitorino systems could be used effectively for the purpose of monitoring unidentified RCS leakage and ensuring that such leaksce did not exceed 1.0 apm.3 He testified that the computerized leak rate test program was the method actually utilized to determine if unidentified RCS leakace was exceedina acceptance criteria.4 i l
]
l WILLIAM ZEWE l Zewe said that he was aware that leak rate tests at times provided variable results or results that were ridiculous.5 He also recalled that operators generally complained about ' the accuracy of the test results. They could not bh relied upon without verification from other plant instruments. Despite these statements, when he was asked whether the leak rate test allowed him to determine if RCS leakage was exceeding limits, he would not di$ectly answer the question. He initially stated that the leak rate test allowed them to operate the plant safely. When pressed further, he stated that they were using what they had available and that he felt'it was safe. When asked the question again, he finally stated, "I l repeat, that is what we used and that is what we had ; available. That is what we used it for. I didn't stop to try I to analyze it, like we are trying to do now. I just did it." Zewe testified that he now realizes that the leak rate test was not accurate; however, he could not explain why the inadequacies of the leak rate test program were not apparent to him in 1978.10 Nor did he recall ever personally taking any steps to' improve upon the accuracy of the test procedure.11 [ corrected 1/17/86] t
I 4 I WILLIAM ZEWE Ze e said that he was aware at that time that leak rate tests o en provided variable results or results that were i 5 that operators generally ridiculous He also recalled complained out the accuracy of the test results.6 They could not be r lied upon without verification from other plant instruments.7 l Despite these st tements, when he was asked whether the j
}
1eak rate test allowe him to determine if RCS leakage was j exceeding limits, he wou not directly answer the question. I He initially stated that te leak rate test allowed them to l operate the plant safely. Whe pressed further, he stated that they were usino what they had av ilable and that he felt it was I safe.8 When asked the question a ain, he finally stated, "I repeat, that is what we used a that is what we had available. That is what we used it fo . I didn't stoo to try to analyze it, like we are tryina to do n . I just did it."9 Zewe testified that he now realizes that e leak rate test was not accurate; however, he could not e. lain why the inadequacies of the leak rate test program were no apparent to him in 1978.10 Nor did he recall ever cerconally aking any steps to improve upon the accuracy of the test procedure i
)
I i l WILLIAM ZEWE f It is also clear that Zewe did not take any steps to l t prevent operators on his shift from discarding test results [ without documentation when the results revealed unidentified j leakage in excess of 1.0 ppm.12 { Zewe admitted that he was aware the operators were i 4 discardina test results that were deemed to be inaccurate; ' however, he had no idea of the number of tests discarded because he never thouaht about it.13 He stated that he gave operators on his shift the discretion to discard test results j 1 without consulting with him or the Shift Foreman.14 As a result, he would not have the slichtest idea how many tests they had discarded.15 When asked whether he knew that it was the practice at that time to discard all unsatisfactory results irrespective of their accuracy, Zewe stated that he was not sure. He conceded, however, that he could not recall any occasion when a test result showina excessive leakage was kept.16 Faust testified that he threw away as many as half of the test results he obtained, most of which showed unidentified leakage in excess of 1.0 ppm.17 Frederick indicated that he did not recall ever throwing a test result away since he turned them over to his Shift Foreman.10 WILLIAM ZEWE Our own analysis reveals that their shift filed test K4 results on only approximately one-third of the shifts that they s worked. No test result above 1.0 opm was ever filed.19 Zewe e also acknowledged that he knew there were occasions when his operators worked entire shifts without obtaining a aood leak rate test result. He understood that this probably meant they
.,.n had repeatedly run tests and discarded the results because they exceeded 1.0 gpm.20 .
L Zewe also knew that Exceptions and Deficiencies were not . I beina filed with respect to the discarded tests and that the tests were not beino documented in any way.21 He stated that he understood the administrative procedure regarding Exceptions and Deficiencies and also that it was his responsibility as a Shift Supervisor to evaluate and resolve Exceptions and Deficiencies. Nevertheless, he stated that he did not remember whether he ever instructed anyone to - prepare an Exception and Deficiency with respect to a leak rate test. He could not explain why that had not been done in licht 4 of so many invalid or unsatisfactory test results.23 He did not remember ever being directed by his superiors to not fils Exceptions and Deficiencies in connection with leak rate tests.24
~
\
5-B s
,.B .-
.g -
. . . s
T 4 9 WILLIAM ZEWE 34), f Although Zewe admitted that he was familiar with the Technical Specifications and procedures at the time, he claimed that it did not occur to him that this practice of discarding - leak rate tests violated Technical Specifications or Company . procedures.25 Manipulation of Test Results + i
' s During the course of our investigation, we found no ,
evidence that Zewe was involved in, or tolerated, manipulation ( , of leak rate tests. * ,
.x Zewe denied that he was aware that anyone had engaged in test manipulation.26 Moreover, there is no testimony from anyone else that in any way connects Zewe to test manipulation.
Our review of plant records revealed only two tests of a suspicious nature that were performed while Zewe was on ' duty.27 These were tests performed by Faust which contained accounted-for water additions that caused indicated unidentified leakage to be lower because of the makeup tank ,3 level instrumentation problems.28 We were unable to conclude .
.. i s
l . q
WILLIAM ZEWE with any certainty that these tests were manipulated; however, even assuming they were, there is no evidence that Zewe was asure of the manipulation. Response to Indications of Excessive Leakay The practice of discardino and failing to document unsatisfactory test results has made it difficult to identify every brief period when there were indications that unidenti-fied leakage exceeded 1.0 ppm at TMI-2. However, .we did identify and examine closely three such periods. It seems clear that Zewe was working during these periods and failed to take required action in response to indications that unidenti-fled leakage exceeded 1.0 gpm. The events of these periods are discussed in Volume I of this report. These periods are dis-cussed here only to the extent that they relate to tho' action taken by Zewe in response to the indications of excessive leakaae It must be recoanized that, as a Shift Supervisor, Zewe' did not personally perform or review leak rate tests in an effort to measure . RC$ leakage. Aside from reading the latest ' test results for insertion in the Daily Plant Status Report, Zewe indicated that,, as a Shift Supervisor, he might not necessarily l t l \ g... t
WILLIAM ZEWE have been advised of the leek rate unless there was a problem. He stated that he regularly conducted plant tours O and inspections for PCS leaks. He would also evaluate the safety implic.' ions of RCS leakage that was brought to his attention and would generally advise his superiors and other j supervisors of his findings in a variety of informal ways. 31 Zewe testified that he was generally aware of problems that were b 'ng expressed in the control room. He recognized l that he was ultimately responsible for ensuring that proper action was taken in response to excessive RCS leakage. I I For the most part, Zewe was not able to recall very much j about periods in which there was high unidentified RCS leakage. Instead, he was only able to provide some general statements about his knowledge of operator response to RCS leakage and his opinions of their practices. For example, Zewe acknowledged that there were problems associated with leak rate testing and that it did not allow for a completely accurate assessment of whether leakage was exceeding 1.0 gpm. He also stated that he believed that the 1.0 gpm limit was unreasonable and not necessarily related to the safe operation of the plant.35 He was not really concerned about the inability to precisely measure RCS leakage,
~b~
icorrected 1/17/86}
l ;l WILLIAM ZEWE ! l J hav been advised of the leak rate unless there was a proble,.29 He stated that he regularly conducted plant tours 1 O and inspgetions for RCS leaks. He would also evsluate the j safety im ications of RCS leakage that was brought to his attention an would generally advise his superiors and other supervisors of is findings in a variety of informal ways.31 i Zewe testified thh. he was generally aware of problems that were beinq expresse in the control room. And recognizeJ that he was ultimate 1 responsible for ensuring that proper action was taken in respon e to excessive RCS leakage.33 i For the most part, Zewe ws not able to recall very much 4 about periods in which there was hiah unidentified RCS } 1 leakage. Instead, he was only ab e to provide some general ! I statements about his knowledge of erator response to RCS leakage and his opinions'of their practic s. I For example, Zewe acknowledged that th e were problems ! I associated with leak rate testing and that it di not allow for l a completely accurate assessment of whether eakaae was enceedina 1.0 opm.34 He also stated that he bel eved that the 1.0 ppm limit was unreasonable and not necessarily related to the safe operation of the plant.35 He was not alty concerned about the inability to precisely measure RCS leaka e, ; I l
WILLIAM ZEWE since he believed that by reviewing the test results, toaether with other plant instruments, it was possible to datermine whether leakaoe was unsafe.36 He stated that he did not see anything wrong with operating the plant without knowing the amount of actual unidentified leakage. He said that he thought it was an accepted practice and it just wasn't something that was focused upon 4c that time.37 Zewe did not remember ever directing his subordinates to proceet towards shutdown because of excessive unidentified leakaae.38 Accorcina to Zewe, if he had caused entry into the action statement it would have been documented in the Control Room and Shift Foreman Logs.39 However, we found no such documentation for- the three periods we analyzed during which unidentified leakage appeared to exceed 1.0 apm. October 16 - 22, 1978 Subsequent to obtainino a satisfactory leak rate test at 19:28 on October 15, 1978, Operations personnel apparently began to experience RCS leakaae problems and difficulty in obtainina satisfactory leak rate tests.40 No tests were i a i
WILLIAM ZEWE filed during the first two shifts on October 16, 1978.41 Zewe, Faust, Frederick, and Scheimann reported for duty at 15:00 on October 16, 1978.42 At 20:35, a leak rate test was performed which showed unidentified leakage at 2.56 apm.43 This test was apparently discovered later by NRC Inspector, Donald Haverkamp. The test was not signed. Neither the Control Room Log nor the Shift Foreman Log contains any reference to the test or entry into the action statement as a response to its unsatisfactory result.44 On October 17, 1978, other Operations personnel were attemptina to identify the source of RCS leakage and had obtained another unsatisfactory leak r:a te test result (2.07 gr a) before Zewe's shift reported for duty.45 At 15:00 on October 17, 1978, Zewe, Faust, Frederick, and Scheimann acain reported for duty.46 Despite the previous unsatisfactory test results that had been obtained, they again failed to enter the action statement or proceed to shut down the plant. Zewe indicated that he had very little independent recollection of this period.4 Nevertheless, after reviewing various plant records pertainina to these events, Zewe l l i
WILLIAM ZEWE acknowledged that he apparently was working on his normal shift on October 16, 17, and 18, 1978.48 He also acknowledged that . plant records indicate there may have been some concern about excessive RCS leakage and that he may have been involved in searching for leakage at the time.49 He could not explain why he did not direct entry into the action statement.50 However, he acknowledged that such a decision would have been his responsibility.51 Since the test of October 16, 1978, was performed on his shift, Ze';e was asked whether he was ever questioned by anyone who was preparing the LER about his failure to enter the action statement. However, he did not recall such questioning. He further stated that if he had been questioned or criticized reaardina his misinterpretation of the Technical Specifica-tions, he would tend to recall that.53 Neither Frederick nor Faust had any explanation reaardina why they failed to enter the action statement at that time.54 December 26, 1978 - January 15, 1979 Zewe stated that he could recall very little about this period.55
WILLIAM ZEWE . Unidentified leakage reached 1.0 qpm on about vanuary 3, 1979, and remained at or above that level until January 15, t 1979, when the plant was shut down. Accordina to plant records, Zewe, Faust, Frederick, and Scheimann were on the llpm - 7am shift on January 2, 3, 4 and 5.56 Zewe's turnover notes contain entries that express an awareness of RCS leakage and problems with one of the makeup tank level transmitters.57 f Neveri.heless, on January 2, 4, and 5, Faust and Frederick turned in satisfactory test results that were obtained with the f faulty makeup tank level transmitter.58 Despite his comments in the turnover notes, Zewe stated that he did not recall the problem with the makeup tank level instrumentation.59 ] I After January 8, Zewe and his subordinates worked several I consecutive assigned shifts during which no leak rate test ! l result was filed.60 There are indications that they were aware RCS leakage was on the increase, particularly by 1 1 January 11. On that day, Hoyt entered the D-ring and attempted to estimate RCS leakage from the pressurizer valves while Zewe's shift was working.61 Given the evidence regarding the I practice of discardina unsatisfactory test results, it is likely that unsatisfactory results obtained during this period were discarded without documentation.
r I WILLIAM ZEWE It is clear that Zewe did not take action to shut down the plant despite the information that was available to him. When questioned about the key events of this period, he testified that he could not remember.62 l Faust testified that he did recall going through more than one shift without beina able to get a good leak rate test, but he could not remember the specifics.63 He further stated ., that had he obtained unsatisfactory test results that he could l1 l not invalidate, he would have turned them over to his Shift I Foreman for appropriate action. He stated that this probably occurred, but that he could not remember 'any specific ; incident.64 Like Zewe, Frederick testified that he could not remember the specific events of this period.65 March 15 - 28, 1979 Our analysis indicates that unidentified leakaae increased during March 1979, until it came close to 1.0 gpm on March 15, and remained fairly constant thereafter. However, as discussed in Volume I, Operations personnel would have perceived
.f unidentified leakage to be much areater.
I 13 -
WILLIAM ZEWE Zewe's shift turned in a negative test result on March 16, and then worked five consecutive assigned shifts durina which no test result was filed.66 Thereafter, satisfactory test results were filed on Zewe's shift on March 24, 25, 27 and 28.67 Each of these yielded satisfactory results because of water additions during the test that caused unidentified leakace results to be reduced for reasons previously explained.68 Zewe testified that he was not aware of this phenomenon.69 Given the testimony and the available plant documentation, it is highly probable that unsatisfactory test results were obtained by Zewe's shift during this period which were discarded and not otherwise documented. Despite his present inability to remember, it is also probable that Zewe was aware increased
~
of the leakage and the difficulty in obtainina satisfactory results. There does not appear to have been any information available at that time that would have justified the discarding of unsatisfactory results and the failing to take required action in response to those results. This is particularly true since, durina this peilod, the operators were manually I
- 14 -
1
WILLIAM ZEWE recalculating the t e s,t results generated by the computer to correct a calculation error in the computer program, thereby producinq more accurate and reliable resalts, l I 1 I
x I l- i L W.ILLIAM ZEWE
- 1. 'Zowe'2/27/85, pp. 11-12.
2 Ibid., p. 5. l 3- Ibid., pp. 7-8. ' 4 Ibid.,.pp. 5-9. t 5 -Ibid., pp. 29-30.
-6 Ibid., pp.-27,.33-34.-
7 Ib'id., p. 10, 8' Ibid., pp. 13-14.
'9 Ibid., p. 32.
10 Ibid., pp. 33, 35. 11 Ibid., p. 34. 1 12 Ibid., pp. 50-52. I i 13 Ibid., p. 11.
.4 14- Ibid., pp. 44-45.
15 Ibid., p. 45. If Ibid., p. 41.
]
17 Faust 2/19/85, pp. 95, 101-102. j 1 18 Frederick 3/12/85, pp. 76-77. I
'19 .-
Vol. III, Table 2. -j 20 Zewe 2/27/85, pp. 46-47. ! 21 Ibid., pp. 37-38, 22 Ibid., pp. 35-37, 23 Ibid., pp. 35-36. ) 1 24 Zewe 2/27/85, p. 37 J
~ ~
i [ corrected 1/17/86] 1 , H H
WILI,IAM ZEWE 1 Zewe 2/27/85, pp. 11-12. 2 \ Ibid., p. 5. 3 I id., pp. 7-8. 4 Ibi , p. 5. 5 Ibid., . 29-30.. 6 Ibid., pp 27, 33-34. 7 Ibid., p. 1. 8 Ibid., pp. 13- 4. 9 Ibid., p. 32. 10 Ibid., pp. 33, 35. 11 Ibid., p. 34. 12 Ibid., pp. 50-52. 13 Ibid., p. 11. 14 Ibid., pp. 44-45. 15 Ibid., p. 45. 16 Ibid., p. 41. 17 Faust 2/19/85, pp. 95, 101-102. 18 Frederick 3/12/85, pp. 76-77. Ng 19 Vol. III, Table 2. 20 Zewe 2/27/85, pp. 46-47. 21 Ibid., pp. 37-38. 22 Ibid., pp. 35-37. 23 Ibid., pp. 35-36. 24 Zewe 2/27/85, p. 37. t I l
11LsIAM ZEWE 25 Zewe 2/27/85, p. 52. 26 Zewe 3/29/84, pp. 18-19; 2/27/85, pp. 100, 108, 125-126. 27 Vol. IV (MPR Report), Appendix K, Test Numbers 9, 1. 28 Vol. III, Table 2. 29 Zeve 2/27/85, pp. 23-24. 30 Ibid., pp. 14-15. 31 Ibid., p. 17. 32 Ibid., p. 27. 33 Ibid., pp. 5, 83. 34 Ibid., p. 48. 35 Ibid., pp. 34, 49, 132. 36 Ibid., p. 48. 37 Ibid., p. 52. 38 Ibid., p. 42. 39 Ibid., p. 70. 40 Vol. IV (MPR Report), Appendix K, Test Number 150 (see also Section IV). 41 Vol. III, Table 2. 42 Ibid., Tables 1 and 2. 43 Vol. IV (MPR Report), Appendix K, Test Number 149. 44 Vol. III, Table 4. 45 Vol. IV (MPR Report), Appendix K, Test Number 148; Vol. III, Table 4. 46 Vol. III, Tables 1 and 2. 1 l 1
WILLIAM ZEWE 47 Zewe 2/27/85, pp. 65, 68, '0, 73. 48 Ibid., p. 65. 49 Ibid., p. 69. } 50 Ibid., pp. 70-71. 51 Ibid., p. 83. 52 Ibid., pp. 80-83. 53 Ibid., pp. 83-87. 54 Frederick 3/12/85, pp. 115-130; Faust 2/19/85, pp. 160-165. 55 Zewe 2/27/85, p. 85. 56 Vol. III, Table 2. 57 TMI-2 Daily Plant Status Reports and Shift Supervisor Turnover Notes, January 2, 3, 4, 5, 1979 (Tab 10). 58 Vol. IV (MPR Report) , Appendix K, Test Nurobers 75, 74, 73. 59 Zewe 2/27/85, p. 85. 60 Vol. III, Table 2. 61 Ibid., Table 4; Zewe 2/27/85, pp. 111-116. 62 Ibid., p. 115, 63 Faust 2/10/85, pp. 37-40. ' 64 Ibid., pp. 40-41. 65 Frederick 3/14/85, p. 109. 66 Vol. III, Table 2, Vol. IV (MPR Report), Appendix K, Test Number 9. 67 Vol. IV (MPR Report), Appendix K, 'est Numbers 4, 3, 2, 1. l I
WILLIAM ZEWE 68 Vol. III, Table 2. 69 Zewe 2/27/85, p. 126. i l 1 i i I i 1 1
.9 I
i e i _ -_ _ _ _ _ _ _ _ _ _ _ _ - _ - - - _ _ . _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ .}}