ML20207M091

Proposed Findings of Fact & Conclusions of Law of GP Miller.* Miller Did Not Put Pressure on Operators to Obtain Leak Rate Test Results Not Exceeding Tech Spec Limits.W/Certificate of Svc

ML20207M091
Person / Time
Site:	Three Mile Island
Issue date:	01/08/1987
From:	Hensley M HUNTON & WILLIAMS
To:
References
CON-#187-2151 86-519-02-SP, 86-519-2-SP, LRP, NUDOCS 8701130083
Download: ML20207M091 (50)

Text

_

p 2/S/

DOCKETED U9iRC UNITED STATES OF AMERICA '87 JAN -9 P3 :18 NUCLEAR REGULATORY COMMISSION r,F F .

DOCr, .pj REFORE THE PRESIDING BOARD In the Matter of )

) Docket No. LRP

)

INQUIRY INTO THREE MILE ISLAND ) ASLBP No. 86-519-02 SP UNIT 2 LEAK RATE DATA FALSIFICATION) 1 PROPOSED FINDINGS OF FACT AND CONCLUSIONS OF LAW OF GARY P. MILLER Michael W. Maupin Maria C. Hensley HUNTON & WILLIAMS F# Box 1535 Ric mond, Virginia 23212 (en :) 788-8200 Counsel for Gary P. Miller l

i 8701130083 870108'~

PDR ADOCK 05000320 0 PDR bS 6

3 TABLE OF CONTENTS I. INTERPRETATION AND IMPLEMENTATION OF TECHNICAL SPECI-FICATION REQUIREMENTS FOR REACTOR COOLANT SYSTEM UNIDENTIFIED LEAKAGE .............................. 2 II. PRESSURE ON OPERATORS TO OBTAIN LEAK RATE SURVEILLANCE TEST RESULTS NOT EXCEEDING TECHNICAL SPECIFICATION LIMITS ........................................... 3 III. KNOWLEDGE OF DIFFICULTIES OPERATORS WERE HAVING WHEN CONDUCTING LEAK RATE SURVEILLANCE TESTS ........... 5 A. Direct Evidence .............................. 6 (1) Control Room Operators .................. 6 (2) Shift Foremen .......................... 7 (3) Shift Supervisors ....................... 9 (4) Supervisor of Operations ............... 10 (5) Unit 2 Superintendent - Technical Support ............................ 11 (6) Unit 2 Superintendent .................. 16 l (7) Entities Outside the Formal Chain of Command ................... 18 (8) U.S. Attorney's Statement of Facts ........................... 22 B. Circumstantial Evidence ..................... 24 (1) Plant Daily Status Reports (PDSRsC ..... 24 (2) Shift Turnover Notes .................. 27 C. Overall Finding on Knowledge of Difficulties ............................. 29 1

l IV. KNOWLEDGE OF DISCARDING UNACCEPTABLE LEAK RATE SURVEILLANCE TESTS ............................... 30 V. KNOWLEDGE OF LEAK RATE FALSIFICATIONS ............ 31 VI. DERELICTION OR CULPABLE NEGLECT IN ALLOWING LEAK RATE FALSIFICATIONS AT TMI-2 .......................... 32 l A. Miller's Responsibilities .................. 36 l B. Discarding Leak Rate Test Results ........... 39

! C. Leak Rate Data Falsifications ............... 42 VII. CONCLUSION ...................................... 45 l

. . . . . - - . - - .. . .- _ _ -- . -_ l IDS - _ _

)

UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARD

)

In the Matter of )

) Docket No. LRP INQUIRY INTO THREE MILE )

ISLAND UNIT 2 LEAK RATE ) ASLBP No. 86-519-02 SP DATA FALSIFICATION )

)

)

PROPOSED FINDINGS OF FACT AND CONCLUSIONS OF LAW OF GARY P. MILLER

1. The Presiding Board has been charged in the Com-mission's Notice of Hearing with making findings of fact on certain specific issues surrounding reactor coolant system (RCS) leak rate data falsifications that occurred at Three Mile Island, Unit 2 (TMI-2) from February 2, 1978 until March 28, 1979. We have been instructed to make such find-ings in sufficient detail to determine the involvement of any individual who may now or in the future work at a li-censed nuclear facility. In Re Incuiry into Three Mile Island Unit 2 Leak Rate Data Falsification, CLI-85-18, 22 NRC 877, 880 (1985). We turn now to our findings with re-spect to C-ary P. Miller, a party to the present proceeding.

See Memorandum and Order (Rulings on Petitions to Inter-vene, Scheduling of Prehearing Conference and Representa-tion of Counsel) at 2-3 (Feb. 14, 1986). We shall deal with

-- __ _--,y_

_ y

the issues in the order in which the Commission has identi-fied them.

2. Gary P. Miller was TMI Station Superintendent for the period February 2, 1978 until sometime in early March 1979, when his job title was changed to Station Manager.

Miller, ff. Tr. 5039 at 2. He also served as TMI Unit 2 Superintendent during the period February 2, 1978 until late 1978, id. at 2, when Joseph Logan assumed an active role as Unit 2 Superintendent. Exhibit 1-A, Vol. VI(G),

11/18/83 Logan Statement to NRC at 1.

I. INTERPRETATION AND IMPLEMENTATION OF TECHNICAL SPECIFICATION REQUIREMENTS FOR REACTOR COOLANT SYSTEM UNIDENTIFIED LEAKAGE

3. Miller did not run leak rate surveillance tests or have managerial responsibility for the performance of leak i

rate tests (LRTs). Exhibit 1-A, Vol. V:(I), 2/7/84 NRC OI 1

Interview of G. Miller at 14. He did not recall ever observing an operator perform a leak rate calculation. Id.

at 13. He did, however, have some familiarity during 1978 l and early 1979 with the Unit 2 Technical Specifications.

l Miller, ff. Tr. 5039 at 15. He knew that Unit 2 was re-quired to use NRC " standard" Technical Specifications. Id.

at 16. As a result of his experience with Technical Speci-fication development and his other responsibilities for l

l l

l l

Unit 2, Miller had some familiarity with Technical Specifi-cation 3.4.6.2. He knew prior to March 28, 1979 that per-missible unidentified leakage was one gallon per minute.

Id. at 16. He also knew the leakage limits for identified and pressure boundary leakage. Id. at 16. He knew that the Unit 2 Technical Specification required that an accept-able unidentified leak rate.be obtained every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

Id. at 16. He recalled that the practice was instituted at Unit 1 of performing a leak rate test (LRT) once per shift and that this practice was carried over to Unit 2. Id.

Miller also understood that if the leak rate limiting con-ditions for operation were exceeded, it was necessary to reduce the leakage or to shut down the plant. Exhibit 1-A, Vol. VI(I), 2/15/85 Stier Interview of G. Miller at 35-36.

II. PRESSURE ON OPERATORS TO OBTAIN LEAK RATE SURVEILLANCE TEST RESULTS NOT EXCEEDING TECHNICAL SPECIFICATION LIMITS

4. Edwin H. Stier, a private attorney who conducted an extensive investigation into leak rate surveillance testing practices at Unit 2, see Exhibit 1-A, Vol. I, was asked during the hearing whether, based on the results of his investigation, he inferred that Miller had exerted pressure upon operators to obtain leak rate surveillance test results not exceeding Technical Specification limits.

d

O

._4_

l Tr. 1969-70 (Stier). Stier replied that he "did not find that the evidence indicated that Miller exerted that kind of pressure on anybody in the operations department." Tr. ,

1 1979 (Stier). There was no testimony to the contrary. Se-I nior operations personnel were unanimous in saying that ,

l

" management" did not exert pressure on operations personnel 1

to violate the Technical Specifications or Administrative Procedures or to engage in manipulation in order to "get good leak rates." See Exhibit 1-A, Vol. VI(F), 3/29/84 NRC OI Interview of Hitz at 56, 62, 90; id. Vol. VI(I), 3/29/85 Stier Interview of Mehler at 24; id. Vol. VI(K), 3/29/84 NRC OI Interview of Zewe at 22-23; id. Vol. VI(B), 4/9/84 NRC OI Interview of Bryan at 18, 24; id. Vol. VI(B),

3/30/84 NRC OI Interview of Chwastyk at 33; Floyd, ff. Tr.

4894 at 5; see, e.Q., Exhibit 1-A, Vol. VI(G), 2/26/85 Stier Interview of Hoyt at 20-21. We find, therefore, that Miller did not exert pressure on operations personnel to violate the Technical Specifications or Administrative Pro-cedures in order to obtain acceptable leak rate test re-sults.

1 I

.-. .~ -.

. III. KNOWLEDGE OF DIFFICULTIES OPERATORS WERE HAVING WHEN CONDUCTING LEAK RATE SURVEILLANCE TESTS

5. Miller testified that he was not aware during the time period in question that Unit 2 control room operators were having significt.nt difficulty achieving the acceptable leak rate surveillance test results required by Technical Specification 3.4.6.2.d. Miller, ff. Tr. 5039 at 17. He thought that acceptable leak rates could be consistently obtained. Id. at 18. Miller defined the term " acceptable" to mean valid; he believed the leak rate test war a valid test of actual plant leakage. Tr. 5075 (Miller). Miller testified that he was never advised prior to March 28, 1979 that acceptable leak rates could not be consistently ob-tained under the Unit 2 Technical Specifications. Miller, ff. Tr. 5039 at 18.

6. Miller's faith in the accuracy of the leak rate surveillance test resulted from his knowledge that surveil-lances were dry run tested before fuel loading. Tr. 5091 (Miller). If problems were identified during the " red-j lining," or dry run testing process, the engineer responsi-ble for that particular surveillance had the responsibility l of submitting changes to make the surveillance operation-ally acceptable. Tr. 5091 (Miller).

l

6-

7. We have reviewed all the available evidence bearing on Miller's knowledge, or lack of knowledge, of leak rate surveillance testing difficulties. We divide this evidence into-two separate categories: (A.) direct and (B.) circumstantial. The direct evidence consists of testimony from Miller's subordinates in the formal chain of command; information from oversight entities outside the chain of command that were at least potentially in a posi-tion to alert Miller to leak rate testing difficulties; and an allegation, made in the U.S. Attorney's Statement of j Facts in the government's criminal action against Metropol-itan Edison-Company,that Miller or John Herbein, or both, were alerted to operators' difficulties by a telephone call from the TMI-2 control room. The circumstantial evidence

' consists of certain documents that were routed to Miller and contained information that could have alerted Miller to operators' difficulties. We take each category of evidence in turn.

I A. Direct Evidence i (1) Control Room Operators

, 8. Miller testified that for both morale and safety i

reasons he attempted to visit the control room once a week, although he probably failed to make it that often. Miller,

! ff. Tr. 5039 at 14; Tr. 5066 (Miller). His contact with

• Mee v,.e6w--.

m.

control room operators was, however, limited. See, e.g.,

Exhibit 5-A, Att. 28, 3/28/85 NRC OI Interview of Germer at 13; id. Att. 13, 1/29/85 NRC OI Interview of Congdon at 9.

In previous investigations, operators testified that they were not aware of whether Miller knew of operators dis-carding leak rate tests, id. Att. 4, 7/26/83 NRC OI Inter-view of Hartman at 2; could not recall Miller commenting on the method of performing leak rates, id. Att. 3, 11/15/84 NRC OI Interview of Booher at 35; and did not discuss leak rate tests with Miller, id._ Att. 10, 9/28/84 NRC OI Inter-view of Cooper at 81. Given these operators' testimony, which was not contradicted before us, we find that Miller was not informed by them of ongoing difficulties with leak rate surveillance testing.

(2) Shift Foremen

9. Miller testified that during 1978 he attended I

[ monthly Shift Foremen meetings. Miller, ff. Tr. 5039 at

14. Miller recalled t hat there were no restrictions on the l

l topics that could be discussed at those meetings, but he believed they dealt principally with staffing and mainte-nance matters and occasionally with an operating matter.

! Id. at 15.

g

10. Shift' Foremen were questioned about their recol-lections of the topics of discussion at monthly Shift Foremen meetings. Those questioned could not recall dis-cussing leak rate testing or invalid leak rates. Conaway, ff. Tr. 3097 at 2; Exhibit 1-A, Vol. Vl(F), 2/14/85 Stier Interview of Hoyt at.29; Exhibit 5-A, Att. 30, 1/4/85 NRC OI Interview of Guthrie at.11; Exhibit 1-A, Vol. VI(A),

3/8/85 Stier Interview of Adams at 44-45, i 11. Minutes were kept of Shift Foremen meetings. Sag Exhibit 1-A, Vol. V(C), Tab 23. During his investigation, Stier reviewed those minutes to see if problems with leak rate testing were discussed, Exhibit 1-A, Vol. II(B), G 1 Miller at 5, 6, and he reported no such discussions. Our own independent review of the available minutes reveals that while the' general subject of surveillances arose at

one Shift Foremen meeting, Exhibit 1-A, Vol. V(C), Tab 24

. (1/4/79 Meeting Minutes), there is no indication that prob-

, lems with performing Surveillance Procedure 2301-3D1 in particular were discussed. We have also reviewed the I

record for evidence that any Shift Foreman discussed leak rate testing problems with Miller outside of the Shift Foremen meetings and have found none. We find, therefore, that Miller was not informed by Shift Foremen of ongoing difficulties with leak rate testing.

_g_

(3) Shift Supervisors

12. Miller also attended monthly Shift Supervisors' meetings. Miller, ff. Tr. 5039 at 14. Miller and other witnesses recalled that there were no restrictions on the subjects discussed at Shift Supervisors meetings. Id. at 15; see, e.a., Exhibit 1-A, Vol. VI(J), 2/8/85 Stier Inter-view of Smith at 37; id. Vol. VI(F), 3/29/87 NRC OI Inter-view of Hitz at 88. Shift Supervisors could not, however, recall discussing problems with leak rate testing at such meetings. See Exhibit 1-A, Vol. VI(J), 2/8/85 Stier Inter-1 view of Smith at 37; id. Vol. VI(B), 4/9/84 NRC OI Inter-view of Bryan at 23; id. Vol. VI(K), 2/27/85 Stier Inter-view of Zewe at 26; Mehler, ff. Tr. 3842 at 3; Hitz, ff.

Tr. 3664 at 4. One Shift Supervisor, Chwastyk, testified before us that he recalled that problems with computer-l assisted leak rate calculations were discussed at the

" shift supervisors' level." Chwastyk, ff. Tr. 3407 at 2-3.

Chwastyk had stated in an earlier interview that he assumed that problems with computer-assisted leak rates were dis-cussed at Shift Supervisors' meetings " simply because it was a natty problem." Exhibit 1-A, Vol. VI(B), 4/24/84 Stier Interview of Chwastyk at 22. He could not, however,

, "really specifically recall any discussions at a Shift Su-1

l. pervisor Meeting." Id. He believed that a computer person t

i

discussed.the topic of leak rates at a Shift Supervisors' meeting, and he said that in his opinion that discussion would be available in Shift Supervisors' meeting minutes.

Id.

13. As part of his investigation, Stier reviewed Shift Supervisor meeting minutes for information that might indicate that problems with the leak rate test procedure or related to the issue of discarding leak rate test results were discussed. Exhibit 1-A, Vol. II(B), G. Miller at 5.

He found that the minutes did not refer to leak rate test 3

problems. Id. There being no evidence that a Shift Super-visor discussed these problems with. Miller outside of the Shift Supervisors' meetings, we find that Miller did not learn of such matters from Shift Supervisors.

(4) Supervisor of Operations

14. James Floyd was the Supervisor of Operations for TMI-2 during 1978 and through the time of the accident.

Floyd, ff. Tr. 4894 at 1. For part of this time period, l Floyd reported directly to Miller in Miller's capacity as 1

Unit 2 Superintendent; from December 1978 onward, Floyd re-ported directly to Joseph Logan. Miller, ff. Tr. 5039 at 3-5.

v . . , - _,.-_-.---,_._.._.r,.._.,,,.,.,.,--,--.__y- - , , - . _ . _ _ - . . - - - _ , < - , - , - . - . , . - . _ , - - . . _ . . . _ . -

15. Floyd testified that at no time did he feel Mill-er was inaccessible to him; if Floyd had had a problem he could not solve, he would have taken it to Miller. Tr.

5006 (Floyd). Floyd testified, however, that prior to the Hartman allegations, he had faith in the validity of the leak rate test; he believed it was a "close enough (indica-tor of real unidentified leakage] to one gallon a minute that we could meet the one gallon a minute limit of the technical specifications." Tr. 4911 (Floyd); Floyd, ff.

Tr. 4894 at 5. Floyd testified that the problems operators were having with leak rate tests did not come to his atten-tion. Tr. 4983 (Floyd). Lacking feedback from his Shift Supervisors, Floyd never " created the time to look over (their] shoulder (s]." Tr. 4988 (Floyd). Floyd's testimony makes it highly unlikely that he would have discussed leak rate problems with Miller. We find that he did not.

(5) Unit 2 Superintendent-Technical Support

16. James Seelinger was Unit 2 Superintendent-Technical Support and Chairman of the Plant Operations Re-i view Committee (PORC) until early December 1978, when he became Unit 1 Superintendent. Seelinger, ff. Tr. 4623 at 1; Tr. 4628 (Seelinger). George Kunder replaced Seelinger as PORC Chairman and Unit 2 Superintendent-Technical Support in December 1978. Kunder, ff. Tr. 4800 at 1; Tr. 4783 (Seelinger).

17.. Seelinger testified that he talked with Miller about the events that occurred during an NRC inspector's October 18, 1978 visit to the TMI-2 control room. Tr. 4723 (Seelinger). The NRC inspector's visit, of course, precip-itated LER 78-62/lT, and we refer to this event as the "LER episode." While Miller had no current recollection of the LER episode, he did not doubt that he and Seelinger dis-cussed the violation and the LER. Miller, ff. Tr. 5039 at

17. Although he did not have current recollection of ev-erything he discussed with Miller, Seelinger testified that he believed he would have covered with Miller "our inter-pretation, the events that I was aware of as to how -- or surrounding that interpretation that I had played a part in, and that Mr. Haverkamp was aware of that interpretation and.had a problem with that interpretation." Tr. 4723 (Seelinger). Seelinger then testified that he and Miller "had a discussion about what was going on with Haverkamp.

We had a discussion as to how to handle, or if to handle l

l under the circumstances, whether this should be a re-portable event." Tr. 4724 (Seelinger). Seelinger also testified that he and Miller called someone in Reading about the Haverkamp findings. Tr. 4724 (Seelinger).

Seelinger believed he had received guidance from the person in Reading to "[h]andle the particular item as you would any other item that might be a reportable item." Tr. 4727 (Seelinger).

r

18. Seelinger's testimony on precisely what he said to Miller is far from clear. To approximate with any de-gree of certainty what Seelinger may have discussed with Miller, we must first examine what Haverkamp discussed with Seelinger. Haverkamp testified that while conducting a routine inspection of plant operations on October 18, 1978, he became aware that the results of recent leak rate tests indicated that unidentified leakage was greater than the TS limit of 1 gpm. Exhibit 20 at 2-3 (Haverkamp). It was at this time that Haverkamp discovered operators' incorrect interpretations of the 1 gpm limiti.g condition for op-eration (LCO), its relationship to the corresponding sur-veillance requirement and the actions required anytime the LCO was exceeded. Id. at 3-4. Haverkamp testified that he went to Seelinger's trailer where he informed Seelinger and Floyd 1! that the TS LCO applied to LRTs conducted more often than required by the 72-hour TS surveillance frequen-cy. Id. at 5. According to Haverkamp, Seelinger assured him that the plant would be operated in accordance with the

( TS action statements applicable to RCS leakage conditions.

Id. at 6. Haverkamp also testified that he was not aware i that operators were discarding LRT results, Tr. 2130 i

1/ Floyd did not recall discussing these matters with l

Haverkamp. Floyd, ff. Tr. 4894 at 7.

(Haverkamp), or:that operators were experiencing difficul-ties obtaining acceptable LRT results prior to October 16.

Tr. 2132 (Haverkamp). It was Haverkamp's impression on 4

October 18 that the problem of exceeding the LCO for unidentified leakage "was an isolated case." Tr. 2177 i

(Haverkamp).

19. Seelinger's recollection substantially agrees with Haverkamp's. 'Seelinger was the " principal person in-

! volved in technical specification interpretation," Tr. 4632 (Seelinger), and he and Haverkamp discussed " technical specification interpretation." Id. Seelinger had 7

testified in an earlier statement that he was not'sure the subject of keeping or discarding LRTs arose in discussions surrounding the LER. Exhibit 1-A, Vol. VI(J), 4/4/84 NRC OI Interview of Seelinger at 15. Seelinger testified be-e l

fore us that his early October concerns about operators' i

discarding LRT results and running multiple LRTs would have -

I "gone away" given the interpretation of TS 3.4.6.2 that he agreed to as a result of Haverkamp's visit. Tr. 4759-60 (Seelinger).

20. We are thus faced with determining what Miller might have known based on Seelinger's recollection, eight years after the LER episode, of what he discussed with Miller. A preponderance of the available evidence supports a finding that Seelinger discussed with Miller the fact i

i

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

4 that for the' period October 16-18, operators had obtained ,

LRT results showing unidentified leakage greater than 1 gpm and had failed to enter the action statement because of an incorrect interpretation of the applicability of the 72-hour surveillance procedure requirement. The weight of the evidence also supports a finding that Miller and Seelinger discussed whether failure to enter the action statement in response to indications of excessive unidentified leakage constituted an event that the licensee was required to report to NRC. The weight of the evidence does not support findings that (a) Seelinger informed Mill-er of operators' pre-October 16 difficulties in obtaining acceptable LRT results or (b) Seelinger informed Miller that he, Seelinger, knew that operators had been discarding LRT results for a short period in early October. See Tr.

I 4745 (Seelinger). These findings are buttressed by the Prompt Report of Reportable occurrence, which was prepared by Seelinger and is dated October 19. See Exhibit 1-A, Vol. V(C), Tab 31; Tr. 4780 (Seelinger). The Prompt Report j gives one the impression that the events of October 16-18 were isolated occurrences caused by a misinterpretation of 1

technical specifications. We find that that is the context in which Miller and Seelinger discussed the LER episode.

i .

o

- t

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

. 21. George Kunder assumed the position of Superintendent-Technical Support in December 1978. Kunder, ff. Tr. 4800 at 1. In his statement to the NRC, Kunder testified that he could not recall any discussions or meet-ings that dealt with the topic of problems with leak rates.

. Exhibit 1-A, Vol. VI(G), 9/28/83 NRC OI Interview of Kunder at 31. Although Kunder never specifically mentions Miller in this context, we find that given his limited role in leak rate testing, Kunder and Miller did not discuss prob-lems operators were having with LRTs.

(6) Unit 2 Superintendent j 22. Joseph Logan was hired by Metropolitan Edison in-January 1978 for the position of Unit 2 Superintendent.

Exhibit 1-A, Vol. VI(G), 3/27/85 Stier Interview of Joseph B. Logan at 3. From January 1978 until approximately November or December 1978, Logan was in a training program designed to result in his receiving his NRC SRO license.

Id. Logan visited the TMI-2 control room every day during the last three months of his training program to familiar-ize himself with plant personnel and plant operating proce-dures. Id. at 4. In December 1978, Logan actively assumed the position of Unit 2 Superintendent. Id. Vol. VI(G),

11/18/83 Logan Statement to NRC at 1.

l l

23. Logan testified that he was aware that Unit 2 op-erations personnel had problems obtaining consistent leak rate results, but he could not recall whether he became aware of that in January, February or March of 1979. Tr.

5123 (Logan). Logan also stated that he was aware that op-erators had difficulty obtaining acceptable leak rate test results. Exhibit 1-A, Vol. VI(G), 3/27/85 Stier Interview of Logan at 49. He attributed the difficulty to the LRT computer program. Id. at 72. Logan did not, however, re-call any extended period of time in which operators consis-tently obtained leak rate results greater than 1 gpm. Id.

at 65.

24. In his November 18, 1983 statement to the NRC, Logan testified that he did not recall discussing problems with the leak rate test program with Miller. Exhibit 1-A, Vol. VI(G), 11/18/83 Logan Statement to NRC at 5. Logan felt that in comparison with other plant problems, the problems with LRTs were not very significant. Id. We find, therefore, that Logan did not inform Miller of the operators' difficulty in obtaining acceptable LRT results.

r (7) Entities Outside the Formal Chain of Command

25. As we have found above,14 iller was not informed by those in the chain of command of operators' ongoing dif-ficulties with leak rate surveillance testing.2/ There were, however, a number of formal entities outside the chain of command that had independent oversight responsibilities for the safe operation of Unit 2. See Ex-hibit 1-A, Vol. V(B), Tab 14. These entities included the Plant Operations Review Committee, the General Office Re-view Board, Quality Assurance and the Generation Review Committee. Stier and his staff reviewed the records of each of these entities; they "found no evidence that leak rate test problems were raised with management by those groups." Exhibit 1-A, Vol. I at 138.E/

26. Quality Assurance had a particularly focused re-sponsibility for overseeing the performance of required surveillances. Administrative Procedure 1010 stated that the Supervisor of Quality Control was " responsible for providing surveillance of the Surveillance Test Program at a frequency, and in areas, selected by the OQA Department."

2/ We found in paragraph 20, above, that Seelinger did not inform Miller that operators were experiencing problems conducting LRTs prior to October 16.

1/ Stier concluded that PORC raised the issue only in the context of the LER. Tr. 1900 (Stier).

l l

i Exhibit L-A, Vol. V(C), Tab. 17 at 2.0, AP 1010 Section 2.4. The Manager of Quality Assurance stated that he believed Audit 78-31 was one of several audits regularly scheduled and conducted on a two-year cycle. Id. Vol.

VI(K), 3/20/85 Stier Interview of Troffer at 40-41. The

" purpose and scope (of Audit 78-31 was] to assess the ade-quacy of and compliance with the Technical Specification implementing procedures at TMI Units I and II." Id. Vol.

V(D), Tab 39, Memorandum from Loignon to Miller and Seelinger (Dec. 14, 1978). Technical Specification 4.4.6.2 was reviewed as part of Audit 78-31. Id. at p. 5 of 7 (Checklist 78-31 Tech Spec Appendix A).S! Audit'78-31 does not mention any problems operators were having conducting the reactor coolant system mass balance inventory test, one of the surveillances required by TS 4.4.6.2. We find, 1/ Item 6 on page 5 of 7 reads:

Sat. Unsat. NA Remarks 4.4.6.2 - How do S/' they don't these surveillances directly --

relate to gpm leakage? the leakage l rate has to be measured.

The Auditor explained that the T.S. 4.4.6.2 surveillances were mostly leakage indicators that had been documented as I sensitive enough to meet a mimimum threshold, but the leak

! rate test was the only required surveillance that would ac-l tually quantify the leak rate. Exhibit 1-A, Vol. VI(G),

j 4/22/85 Stier Interview of Loignon at 21; 23-24.

l i

t

therefore, that TMI entities outside the chain of command did not inform Miller of leak rate testing problems or improprieties.

27. NRC personnel were another source outside the chain of command that potentially could have uncovered problems with leak rate testing and brought those problems to Miller's attention. We examine NRC inspection activity in 1978 and 1979 for the sole purpose of determining what the NRC would have brought to Miller's attention, not for the purpose of casting blame or attempting to shift the re-sponsibility for leak rate testing improprieties at Unit 2.

28. There were a total of six NRC inspections con-ducted between February 1978 and March 1979 that discussed leak rate testing at TMI-2. Tr. 2174 (Haverkamp). Inspec-tion 78-10, conducted in February 1978, verified that the surveillance procedure had been prepared, approved and scheduled as required. Inspection 78-11, conducted during February and March 1978, included an inspector's evaluation of test results; he found them acceptable. Inspection 78-14, conducted in late March 1978, included verification by one or two inspectors that prior to initial criticality, technical specification 3.4.6.1 (RCS leakage system opera-ble) had been satisfied. Id. Inspection 78-30, conducted in early October 1978, involved a record review to verify that technical specification 3.4.6.2 (RCS leakage) had been

satisfied. Tr. 2174-75 (Haverkamp). Inspection 78-36, conducted in December 1978, included a review of records related to Surveillance Procedure 2301-3D1 to verify con-formance to requirements. Tr. 2175 (Haverkamp). Finally, Combined Inspection 50-289/79-01 and 50-320/79-01 discussed LER 78-62/lT and the appropriateness of licensees' correc-tive action. In it, the NRC Inspector verified that LER 78-62/lT met the reporting requirements of Technical Speci-i fications, "that appropriate corrective action (had] been taken, that the event was reviewed by the licensee as re-

, quired by Technical Specifications, and that continued op-eration of the facility was conducted in conformance with Technical Specification limits." Exhibit 20 at Exhibit E,

p. 11. Five of the inspection reports mentioned above

contained assurances that leak rate surveillance testing

, could be properly performed at TMI-2. Only one indicated a problem with leak rate testing and it concluded that the problem had been satisfactorily resolved. We find, there-fore, that the information available from the NRC to Miller in 1978 and 1979 would not have informed Miller of opera-tors' difficulties with leak rate testing.

(8) U.S. Attorney's Statement of Facts

29. The final piece of direct evidence from which it might be inferred that Miller was aware of difficulties op-erators were having conducting LRTs consists of a passage from the " Statement of Facts Submitted by the United States" in the criminal action against Metropolitan Edison Company. See Exhibit 1-A, Tab 2 at 13. The Statement of Facts (Statement) alleges that a Shift Supervisor stated that as a result of the October 18, 1978 NRC inspection a conference telephone call was made from the shift supervi-sor's office to "either TMI's Station Superintendent or Metropolitan Edison's Vice-President for Generation, or both." Id. At the time, of course, Gary Miller was the Station Superintendent. Acco'rding to the Statement:

During the conversation the operations personnel alerted the Station Superin-tendent and/or Vice-President for Gen-eration that because of the numerous

, " bad" leak rate tests obtained at Unit l 2, the NRC's interpretation _of the leak l rate technical specifications would re-sult in repeated shutdowns of the fa-cility.

30. The record indicates that Joseph Chwastyk was probably the Shift Supervisor to whom the Statement refers.

Chwastyk testified, however, that he did not know the Statement was referring to him until two months prior to his testimony before us. Chwastyk's testimony in this

, - , - - . _ - , , , ,-.,--._.w , , v - - - _ , - - - - - _ , - -. -- -w-----

hearing does not support the U.S. Attorney's allegation.

Chwastyk testified that he recalled walking into the shift supervisor's office when Seelinger was announcing to some-one over the telephone that he " wanted to declare a leak rate valid and write up an LER." Chwastyk, ff. Tr. 3407 at

4. During an interview with the U.S. Attorney's assis-tants, Chwastyk specifically told them that he did not know who Seelinger was talking to. Id. at 5. When the U.S. At-torney's assistants suggested Miller and Herbein as candi-dates, Chwastyk agreed that it could have been either one or both but said that he had no way of knowing. Id.

Chwastyk testified "it could have been anybody." Tr. 3512 (Chwastyk). Both Miller and Herbein denied any knowledge of a conversation such as the one described in the State-ment of Facts. Miller, ff. Tr. 5039 at 19; Exhibit 1-A, Vol. VI(I) 2/15/85 Stier Interview of Miller at 75-76; l Herbein, ff. Tr. 5268 at 10. Mr. Stier was unable to find i

any evidence confirming Miller's or Herbein's participation in this telephone call. Exhibit 1-A, Vol. II(B), Miller at 14; Id. Herbein at 5. We find, therefore, that Miller did not participate in such a telephone call.

B. Circumstantial Evidence (1) Plant Daily Status Reports (PDSRs)

31. Miller testified that he kept abreast of the daily status of the Units primarily through two mechanisms:

one was the Plant Daily Status Report (PDSR) and the other a morning conference call to Reading. Miller, ff. Tr. 5039 at 12. On days when Miller was at TMI, he reviewed the PDSR each morning and then held a telephone conference call with the Unit Superintendents and corporate personnel in Reading, who Miller believes also had copies of the PDSR.

Id. The PDSR listed the " primary leak rate," or unidentified leak rate, along with approximately 30 other important plant parameters. See Exhibit 1-A, Vol. V(B),

Tab 10. Miller would also review the comments on the right side of the one-Unit PDSR form and at the bottom of the two-Unit PDSR form that was adopted later, because he would l have expected any special problems to be addressed there.

Miller, ff. Tr. 5039 at 13.

32. Miller testified that in the years since the leak l

l rate difficulties came to light, he has had occasion to re-view carefully the PDSRs for the period October 1, 1978 through March 28, 1979. Miller, ff. Tr. 5039 at 18. He now knows that from time to time during that period the same unidentified leak rate was carried forward for three l

l t

or four consecutive days. Had Miller trended these leak-rate results, he might have been led to make additional in-quiries, and those inquiries might have led him to identify the difficulties that operators were having with the leak rate tests. Id. at 18-19. Miller testified, however, that he did not compare leak rate results on the PDSRs for sev-eral days at a time. He simply did not realize that such repetitive results were being recorded. Id. at 19.

33. We have reviewed the PDSRs for the relevant time period in an attempt to assess what Miller may have known during 1978 and early 1979. See Exhibit 1-A, Vol. V(B),

Tab. 10. At no time is the number reported for " primary leak rate" greater than 1 or less than .86 gpm, despite the probability that operators were obtaining leak rate test results of greater than 1 gpm.E/ We have also re-viewed the categories of comments contained in the PDSRs.

They are entitled " Tests and Special Operations in Prog-ress," " Required Surv. Tests which are overdue as of this date," " Remarks, Problem Areas, and occurrences" and "An-l ticipated Operations (Next 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />)." These categories of l comments do not contain references to problems in obtaining I

1/ The evidence establishes that LRT results of greater than 1 gpm were not filed. See Exhibit 1-A, Vol. I at 60.

Stier also indicates that there was an expected band of variation in LRT results. See Tr. 1417-18 (Stier).

1 1

1

I satisfactory unidentified LRT results. Egg id. Vol. V(B),

Tab 10. Indeed, there was a space on the PDSR for Shift Supervisors to " list out of spec. condition (s]." The unidentified leak rate was never listed as out of specifi-cation. Egg id.

34. For Miller to have known from the PDSRs that op-erators were experiencing difficulty running LRTs he would have had to realize that Shift Supervisors were reporting the same unidentified leak rate number for several days in a row. He did not. As noted below, Miller's days on-site were filled with a wide array of responsibilities associ-ated with running a nuclear power plant. Miller impressed the Board as a dedicated manager but one who had many competing demands on his time. We find it entirely under-standable that Miller would not have remembered from one day to the next the number for unidentified leakage re-ported on successive PDSRs, particularly in light of his lack of direct knowledge that operators were experiencing continuing difficulties obtaining satisfactory leak rate test results. We find, therefore, that Miller did not learn from the PDSRs that operators were experiencing dif-ficulties with leak rate testing.

(2) Shift Turnover Notes

35. Miller testified that in connection with leak rate investigations of recent years he has learned that there were probably attached to his copies of PDSRs certain

" turnover notes" written by the Shift Supervisors. Miller, ff. Tr. 5039 at 12. He could not recall seeing turnover notes. Id. But he acknowledged that he may have looked at the sheets if they were attached. Id. at 14.

36. The shift turnover notes were one method the out-going Shift Supervisor had to inform his replacement about the status of the Units. Miller, ff. Tr. 5039 at 12. In addition, the Shift Supervisors would normally discuss with each other during shift changes items in the notes. Exhib- ,

it 1-A, Vol. VI(H), 2/28/85 Interview of Mehler at 10.

37. We note that while Miller believes shift turn-l overs were probably attached to PDSRs, testimony from some Shift Supervisors indicates that they may not have been.

Exhibit 1-A, Vol. VI(J), 2/8/85 Stier Interview of Smith at 34; id. Vol. VI(K), 2/27/85 Stier Interview of Zewe at l

l 17-18 (uncertain of time period in which turnovers were at-l tached to PDSRs). In any event, several of the shift turn-overs for the period October 1978 through March 1979 contain notations referring to leak rates. Four of these i notations are simple declarations that are subject to 1

i l

l l

. benign interpretations; they indicate that the Shift Super-visor writing the note was aware that his shift had gotten a good leak rateE/ or was informing his replacement that the replacement needed to obtain a leak rate.1/ Only two notations made during the period October 1978 to March 1979 indicate that operations personnel had difficulty running LRTs. The first is contained in the 10/18/78 shift turn-over; item 11 states "Still could not get a leak rate --

1900 today is deadline [.] Doing hand calculations [.]."

Exhibit 1-A, Vol. V(B), Tab 10. This notation was made, apparently, during the time of Haverkamp's inspection. As we discussed above, we believe Miller would have known that for this short time period operators had obtained LRT re-sults of greater than 1 gpm unidentified leakage. The sec-ond notation is contained in item 16 of the 3/27/79 turn-over. It states "Didn't get good leak rate yet." There then a,7 ears, in what is apparently different handwriting, "need one by tomorrow. . . . Id. The very next notation s/ Item 6 on the 1/4/79 turnover, for example, states "we got good leak rate and Heat Balance 11-7"; Item 1 on the 2/19/79 turnover states "Got a good leak rate. Leakage out of RCS looks like is due to MU-R1 (outlet relief valve from MUT). (1) At 60" level in MUT & 5-6 psig H2 overpressure we get good LR. ... Exhibit 1-A, Vol. V(B), Tab 10.

7/ Item 38 on the 12/20/78 turnover states "need leak rate"; Item 25 on the 1/13/79 turnover states "need a leak rate". Exhibit 1-A, Vol. V(B), Tab 10.

is preceded by an asterisk and reads "Found instrument fit-ting leak on seal injection xmitte[r?)." Id.

38. The most we can conclude from Miller's testimony and the turnover notes is that if Miller had access to these shift turnovers and if he had closely scrutinized the turnovers, he would have received some indication that for brief periods of time, operators were experiencing diffi-culty obtaining " good" LRT results. He would not have been aware of the continuing difficulty CROs were facing in run-ning LRTs. And he would not have been alerted to the fail-ure to file Exceptions and Deficiencies, the general prac-tice of discarding LRTs or the practice of filing tests without regard to their validity.

C. Overall Findina on Knowledae of Difficulties

39. Miller testified before us that he was not aware that operators were having difficulty achieving acceptable leak rate surveillance test results. We had an opportunity to assess his demeanor and credibility as a witness. We found Miller to be direct and candid in his responses. The weight of the evidence supports a finding that Miller did not know of difficulties with leak rate surveillance testing. We so find.

IV. KNOWLEDGE OF DISCARDING UNACCEPTABLE LEAK RATE SURVEILLANCE TESTS

40. Miller testified that he did not know that unacceptable leak rate surveillance test results were being discarded. Miller, ff. Tr. 5039 at 20. He said that he would not have condoned such conduct. Id.

41. We believe Miller would have known operators were discarding LRTs only if he actually saw someone throw away a LRT or was informed of the practice by a subordinate. We have reviewed the testimony of those who either were dis-carding LRTs or knew that operations personnel were dis-carding leak rate test results to see if anyone ever in-formed Miller of this practice. Aside from speculation that Miller would have known, see, e.c., Exhibit 1-A, Vol.

VI(K), 2/27/85 Stier Interview of Zewe at 50, it appears that no one actually informed him. See id.; Exhibit 1-A, Vol. VI(B), 4/9/84 NRC OI Interview of Bryan at 14; Exhibit 5-A, Att. 4, 7/26/83 NRC OI Interview of Hartman at 2; Ex-hibit 5-A, Att. 10, 9/28/84 NRC OI Interview of Cooper at 81.

42. We do not think it unusual that Miller would not have known operators were discarding tests. Miller was not the only visitor to the control room not to know of the op-erators' practice of discarding LRTs. See Tr. 5138 (Logan); Tr. 2130 (Haverkamp); Exhibit 1-A, Vol. VI(G),

9/28/83 NRC OI Interview of Kunder at 24; Marshall, ff. Tr.

4380 at 3; Exhibit 1-A, Vol. VI(C), 2/25/85 Stier Interview of Fels at 12; id. Vol. VI(G), 4/22/85 Stier Interview of Loignon at 27.

43. Discarding tests would, by its very nature, be hard to detect. See Tr. 1982-83 (Stier). Discarding was done primarily by control room operators, individuals with whom Miller had little dav-to-day contact. It was not for-malized in a written policy, but was rather an informal practice carried on at the operations level. Tr. 1416 (Stier). Miller would not have learned of this informal practice unless someone who knew of it informed him. We find no evidence that this was done. We, therefore, find that Miller did not know operations personnel were dis-carding unacceptable leak rate test results.

V. KNOWLEDGB OF LEAK RATE FALSIFICATIONS l

l

44. Miller testified that he did not know of data ma-nipulation or other improper actions taken to influence I Unit 2 leak rate test results. Miller, ff. Tr. 5039 at 20.

Miller's testimony was not contradicted by any witness in l

this proceeding. Stier was questioned on what the results of his investigation revealed with respect to Miller's knowledge of such improper leak rate testing actions as I

l

adding hydrogen or water to the make-up tank for the pur-pose of influencing leak ate test results. Tr. 1970-71 (Stier). Stier replied that "the evidence did not show that Miller had any such knowledge" of improper leak rate test practices. Tr. 1971 (Stier); see also Tr. 1985 (Stier). According to Stier, LRT manipulation developed from the CRO level of the operations department, and knowl-edge of test manipulation remained at a low level within the operations department. Tr. 1971 (Stier). We find, therefore, that Miller did not know of, and hence did not condone or direct, any action taken in an attempt to im-properly influence, or falsify, leak rate test results.

VI. DERELICTION OR CULPABLE NEGLECT IN ALLOWING LEAK RATE PALSIFICATIONS AT THI-2 l

45. The Commission directed this Board to address certain specific factual issues. 22 NRC at 880-81. It j then directed us to " render a recommended decision setting l forth the facts surrounding the . . . leak rate l

l falsifications at TMI-2." Id. at 881 (emphasis added); see l

also id. at 883. We have tended to equate leak rate "fal-l sifications" with intentional operator manipulations for l

l the purpose of influencing the outcomes of leak rate tests, i

" Falsification" would, therefore, include such actions as adding hydrogen gas or water to the make-up tank during the

4 performance of a leak rate test in an attempt to influence make-up tank level indication. See id.' at 880-81.

46. An understanding of the Commission's purpose in instituting this proceeding, however, leads us to conclude that the scope of our recommended decision is not limited to " falsifications" in the narrow sense. This proceeding is a direct result of Harold Hartman's allegations that prior to the TMI-2 accident, RCS leak rate surveillance tests "were at times purposely manipulated and records of unacceptable results were discarded . . . to cover up the fact that over an extended period of time the results of the tests exceeded technical specification limits for unidentified leakage." 22 NRC at 878. We think, there-

! fore, that the Commission intended to include the dis-carding of LRTs within the meaning of " falsifications." We i

do not, however, interpret the Commission's use of the term l

l " falsification" to include items B.2(a) and (b) of the Com-mission Order.E/

l l

1/ Item B.2(a) deals with TS 3.4.6.2 interpretation and j implementation, the LER episode and corrective action fol-l lowing the LER episode. Item B.2(b) deals with difficul-ties operators were experiencing in conducting LRTs and the pressure on operators, if any, to obtain LRT results that did not exceed TS limits. 22 NRC at 880.

l

47. Our inquiry is not limited to identifying-those persons who knowingly participated in leak rate falsifica-tion. We have been directed by the Comnission to render a recommended decision " identifying those individuals who participated in, or knew of and condoned, or by their dereliction or culoable neglect allowed the leak rate fal-sifications at TMI-2." 22 NRC at 881. (emphasis added).

The Commission twice used the words " dereliction or culpa-ble neglect," indicating that it chose those words careful-ly. See id. at 883. " Dereliction" means " intentional or conscious neglect." Webster's Third New International Dic-tionary at 607 (1976). Dereliction, as used in " dereliction of duty," has been defined as "a willful or fraudulent vio-lation or neglect of any official duty, and not the mere failure to do one particular thing." 26A C.J.S.

Dereliction at 499 (1956).

48. " Culpable neglect" has been variously defined.

See 25 C.J.S. Culpable at 29 (1966); 10A Words and Phrases 365-369 (1968). The term embodies elements of intent and action. In the civil context, it usually means more than the failure to use ordinary care, but less than gross care-lessness. 25 C.J.S. at 29. " Neglect" has been defined as "a failure to perform a duty." Home Owners' Loan Corporation v. Doolittle, 57 Ohio App. 329, 13 N.E.2d 920 (1937). In its modification of the word " neglect,"

" culpable" means heedlessness, and has the meaning "censur-able or blameworthy." Id. It has also been defined as "the neglect which exists where the loss or damage can fairly be ascribed to carelessness, improvidence or folly."

Ballentine's Law Dictionary at 295 (1969).

49. While the terms " dereliction" and " culpable ne-glect" are incapable of exact definition, we believe the Commission intended them to connote an element of scienter, or knowledge. Neglect may be either intentional or inad-vertent. By its use of the word " culpable," we believe the Commission intended that the neglect be ascribable to one's blameworthiness.E/ The nature of the present proceeding supports our interpretation of the Commission's Order.

This proceeding is a legislative-type hearing instituted for the purpose of developing the facts surrounding RCS leak rate falsifications. Our findings on individual re-sponsibility for involvement in leak rate falsifications

may be used in a quasi-punitive fashion to, for example,

, deny a license renewal. See 22 NRC at 884, n.3.

l L

l 9/ It is possible, of course, that we are misinter-preting the terms " dereliction" and " culpable ncglect."

But even if it were the Commission's intent to place blame on an individual based on what a reasonable person in that individual's position should have known based on all the surrounding circumstances, our findings with respect to Miller would remain unchanged.

l

50. We found above that Miller did not know-that (a) leak rate surveillance test results were being discarded or (b) some operations personnel were manipulating data or taking other improper actions during leak rate surveillance testing in an attempt to improperly influence test results.

We must determine whether Miller's lack of knowledge of ei-ther (a) or (b) was due to his own dereliction or culpable neglect. To determine whether Miller is properly charge-able with dereliction or culpable neglect, we first need to develop an understanding of Miller's activities at TMI dur-ing the period February 2, 1978 to March 28, 1979.

A. Miller's Responsibilities

51. As TMI Station Superintendent or Station Manager during the period February 1978 through March 1979, Miller was responsible for the approximately 500 people employed at the TMI site, 100 of whom were supervisory personnel.

Miller, ff. Tr. 5039 at 5. Persons in three principal op-erating positions reported to him. They were the Unit 1 Superintendent, the Unit 2 Superintendent and the Site Maintenance Superintendent. Id. at 2-3. Joseph Logan was the Unit 2 Superintendent from December 1978 through

! March 29, 1979. Id. at 3.

52. Miller was also TMI Unit- 2 Superintendent for the period February to December 1978. Miller, ff. Tr. 5039 at 3. Four principal operating positions reported to Mill-er in his capacity as Unit 2 Superintendent. They were the Superintendent Technical Support, Supervisor of Operations, Supervisor of Maintenance and Supervisor of Radiation Pro-tection and Chemistry. Id. at 4. James Seelinger was Unit 2 Superintendent Technical Support until late 1978, when he was replaced by George Kunder. Id. at 5. James Floyd held the position of Supervisor of Operations throughout the February 1978 through March 1979 time frame.

Id. at 4-5.

53. Miller testified that during the period in ques-tion he usually worked six days a week, sometimes seven,

and on the average about 80 hours9.259259e-4 days <br />0.0222 hours <br />1.322751e-4 weeks <br />3.044e-5 months <br /> a week, carrying out his l

many and varied job assignments. Miller, ff. Tr. 5039 at l

l 5-6; see also Seelinger, ff. Tr. 4623 at 3. Miller spent from 20 percent to 40 percent of his time away from the Is-land, principally in meetings at Reading on a host of sub-

! jects. Miller, ff. Tr. 5039 at 6. Most of Miller's time from 9 a.m. to 5 p.m., Monday through Friday, was spent in meetings. Id. He dealt with large volumes of work-related l correspondence prior to and after normal working hours.

Miller estimated that he spent about 30 percent of his time dealing with personnel matters. Id. at 7. During the last

few months of 1978, Miller was heavily-involved in planning for the mid-February 1979 outage of Unit 1 and the antici-pated commercial operation of Unit 2. Id. at 9-10. Mill-er's personal staff consisted of a secretary and a " tech analyst," an individual that assisted Miller with adminis-trative and, to a limited degree, technical matters. Tr.

5043 (Miller).

54. Miller testified that despite his many and varied job responsibilities, he did pay attention to operating matters. Miller, ff. Tr. 5039 at 11. He necessarily depended on his subordinates to advise him of problems they believed needed his attention. Id. Miller relied princi-

' pally on Seelinger through 1978 and on Joe Logan thereafter to keep him informed of Unit 2 operations. Miller, ff. Tr.

5039 at 12.

55. Miller testified that Seelinger was, as a practi-cal matter, the principal Unit 2 manager during 1978 and until Logan took over. Id. at 3-4. Miller explained that j he, Miller, was the Unit 2 Superintendent and he "[did] not I

! run from that responsibility." Tr. 5049 (Miller). Miller testified that he relied on Seelinger because Seelinger had

(

an NRC license for Unit 2 and was in the Unit everyday.

j Tr. 5048, 5049 (Miller); see Tr. 4734 (Seelinger). Miller l

testified that if Seelinger knew of a problem that ought to l

l be dealt with by the Unit 2 Superintendent, Miller expected

Seelinger to notify him. Miller had enough confidence in Seelinger's technical ability and integrity to know that

.Seelinger would not leave something in the Unit unresolved.

Tr. 5048-49 (Miller).1E/

(B) Discardino Leak Rate Tests Results

56. We found above that Miller did not know op-erations personnel were discarding LRT results. We also found that Miller was not apprised of thi.s practice by his subordinates. To understand how Miller would not know op-erators were discarding LRTs it is necessary to evaluate leak rate testing in the larger context of plant'op-erations.

12/ Miller and Seelinger did not entirely agree on the ap-propriate characterization of Seelinger's role at Unit 2.

It appears to have been, however, a dispute over form and I not substance. Seelinger testified that his position was a

( " grade level" higher than that of the Supervisor of Op-erations. Tr. 4625-26 (Seelinger). By virtue of having

. the inclination to tackle the jobs that needed doing, Seelinger's " plate of jobs" was more than full. Seelinger, ff. Tr. 4623 at 6. Seelinger did not, however, want to leave the impression that Operations Department personnel reported to him. Seelinger, ff. Tr. 4623 at 3-4. Given Miller's testimony, we think it clear that Operations De-partment personnel did not report to Seelinger. It is also clear, however, that as a practical matter Seelinger was viewed as the senior manager of Unit 2 a good deal of the time. See Tr. 5219-20 (Logan); Tr. 2173 (Haverkamp); Ex-hibit 1-A, Vol. VI(H), 4/18/84 Stier Interview of Mehler at 8; see Exhibit 1-A, Vol. I at 133.

5 7 ._ Performing leak rate surveillance testing was only one of many surveillances operators were required to perform. See, e.c., Frederick, ff. Tr. 2447 at 2. Without belittling the importance of leak rate surveillance testing, we believe it can be accurately characterized as a routine undertaking performed by CROs as part of their daily activities. In the context of control room op-erations, the running of a LRT would not be a particularly noteworthy or memorable event. See, e.o., Tr. 2160 (Haverkamp).

58. TMI-2 had Administrative Procedures (AP) in place for dealing with surveillance test results. In the absence of notice or knowledge to the contrary, we believe a person in Miller's position would have been justified in believing that operations department personnel were following those procedures.

59. AP 6.10.ld. required "[rlecords of surveillance activities . . . required by . . . Technical Specifica-tions" to "be retained for at least five years." Exhibit 1-A, Vol. V(B), Tab 14 at 6-18. Leak rate tests that could not be performed successfully or failed to meet acceptable criteria were to be documented by the use of Exception and Deficiency sheets. Id. Vol. V(C), Tab 17, AP 1010, Section 2.3(c). Exceptions and Deficiencies were to be reviewed by Shift Supervisors. Id. AP 1010, Section 3.2.4. In the

y

- case of a Deficiency, the Shif t Supervisor was to determine whether a reportable event had occurred and, if so, bring the matter to the Unit Superintendent's or the Supervisor of Operations' attention. Id. AP 1010, Section 3.2.4.2.

In addition, the GMS Coordinator, an individual who sched-uled and coordinated required surveillance test programs, was responsible for informing the PORC Chairman and the Supervisor-Quality Control (QC) whenever surveillance tests d not meet acceptance criteria. Id. AP 1010, Section 2.3(c). The GMS Coordinator was to do this by submitting to PORC and QC copies of the test's E/D sheet. Id. Miller expected the GMS Surveillance Coordinator to schedule records of surveillance completion and track deficiencies overall for both Units. Exhibit 1-A, Vol. VI(I), 2/15/85 Stier Interview of Miller at 42.

60. TMI-2 had a valid procedure for dealing with leak rate surveillance test results. The evidence in this hear-ing establishes, however, that operations personnel did not follow this procedure. See, e.o., Exhibit 1-A, Vol. I at

60. In his report, Stier concluded that "[bly their very nature, the practices of discarding tests and failing to follow procedures for documenting test results tended to conceal such conduct." Id. at 135; see also Tr. 1982-83 (Stier). Discarding tests and failing to file Exceptions and Deficiencies " subverted the paper flow." Tr. 1983 l

(Stier). Stier testified that there was "a heavy reliance on the flow of paper from one level of management to the next, as a means of raising issues." Id. Thus, despite the fact that operators may have been discarding leak rate surveillance tests, a person in Miller's position could not reasonably have been expected to know that AP 1010 and.AP 6.10.ld were not being followed unless he had observed practices contrary to those administrative procedures or unless he had been informed. We found above that Miller  !

I did not know LRTs were being discarded. We find, there-fore, that Miller is not guilty of dereliction or culpable neglect in failing to know that operations personnel were  :

I discarding leak rate surveillance test results.

C. Leak Rate Data Falsifications

61. Hal Hartman, a TMI-2 control' room operator prior to the accident, alleged that RCS leak rate surveillance tests were at times purposely manipulated by various means.

22 NRC at 878. Hartman previously testified that he had no reason to believe that anyone above the CRO level knew of LRT data manipulation. Exhibit 5-A, Att. 4, 7/26/83 NRC OI Interview of Hartman at 2; Exhibit 1-A, Vol. VI(F), 7/26/83 Hartman Statement to NRC at 1-3. Hartman testified before us that he never approached "his supervision" with his con-cerns about the manipulation of LRTs at Unit 2. Tr. 2292 (Hartman).

62. It appears that manipulation for the purpose of improperly influencing LRT results was limited to only a few CROs. William Russell, Director, Division of Human Factors Technology, Office of Nuclear Reactor Regulation, ff. Tr. 651, and co-author of Exhibit 5-A, testified that he "did not conclude that there was willful manipulation on a widespread basis." Tr. 2359 (Russell). While there was testimony to the effect that some operations department personnel were aware of the effect of hydrogen on make up tank (MUT) level transmitters, many CROs and all Shift Foremen and Shift Supervisors testified that they did not know hydrogen was being added to the MUT for the purpose of influencing leak rate test results; neither were they aware of anyone using any other method of manipulating LRT re-sults.

63. Stier testified that based on the results of his t

investigation, he concluded that Shift Foremen as a class l generally were not aware of LRT manipulation. Tr. 1966 l '

(Stier). As for Shift Supervisors, Stier concluded that

"[t]he evidence indicating that Shift Supervisors condoned test manipulation is extremely weak." Exhibit 1-A, Vol. I at 129.

64. Chwastyk, a shift Supervisor, testified that he conducted an " experiment" to determine if hydrogen had an I effect on LRT result; he concluded that it seemed to have a i

. favorable effect. Chwastyk, ff. Tr. 3407 at 5. Chwastyk-could not, however, recall any specific discussions with Shift Supervisors other than the one relieving the shift on which he conducted the " experiment." Tr. 3477-78, 3480 (Chwastyk). Chwastyk testified that he could not recall discussing with Floyd, Logan, Seelinger or Miller the ef-fect he observed hydrogen to have on LRTs. Tr. 3483-86 (Chwastyk). In any event, Chwastyk testified that he was not aware that some operators were adding either hydrogen or water to the make up tank during the performance of leak rate tests in order to improperly influence test results.

Chwastyk, ff Tr. 3407 at 6.

65. Stier testified that his investigation did not show that Mr. Miller knew of improper leak rate test manip-ulations. Tr. 1971 (Stier). Stier said that based on the l conduct of employees and the patterns of LRT results, he I

believed that LRT manipulation developed within the op-erations department and that knowledge of test manipulation tended to remain at a fairly low level within the op-erations department. Tr. 1971 (Stier). Stier's testimony was uncontradicted. The documentary and testimonial evi-dence developed in this hearing supports such a finding.

I We find, therefore, that Miller did not have reason to sus-pect improper LRT manipulations. He is not, therefore, properly chargeable with dereliction or culpable neglect in l

l l

not knowing that some operations personnel were engaging in improper leak rate test manipulations.

VII. CONCLUSION

66. Based on our review of the relevant, material and reliable oral and documentary evidence in this proceeding, we conclude that Gary P. Miller did not put pressure on op-erators to obtain LRT results not exceeding TS limits, did not know of operators' difficulties with leak rate surveil-lance testing, did not know operators were discarding LRTs, and did not participate in, or know of and condone, or by dereliction or culpable neglect allow leak rate data falsi-fications at Three Mile Island Unit 2.

Respectfully submitted, GARY P. MILLER I

l W e 0.

M'ariaC.Hensley,Counfl Of Counsel Michael W. Maupin Maria C. Hensley HUNTON & WILLIAMS P.O. Box 1535 Richmond, Virginia 23212 Dated: January 8, 1987 l

i 4

DOL K.E T E C es p:C January 8,.1987m -9 P 3 38 0FF:0 - -

M UNITED STATES OF AMERICA DOCrli t% 1. " '. .M r-NUCLEAR REGULATORY COMMISSION 'M ' l BEFORE THE PRESIDING BOARD l

)

In the Matter of )

)

INQUIRY INTO THREE MILE ) Docket No. LRP ISLAND UNIT 2 LEAK RATE ) ASL3P No. 86-519-02 SP DATA FALSIFICATION ) l

)

CERTIFICATE OF SERVICE I certify that copies of Proposed Findings of Fact and Conclusions of Law of Gary P. Miller were served upon the following persons by deposit in the U.S. Mail, first class, postage prepaid, or in the cases marked by an asterisk below by Federal Express, and in the case marked by two as-terisks by Express Mail, addressed to them at the following audresses:

Secretary U.S. Nuclear Regulatory Commission Washington, DC 20555 ATTENTION: Chief, Docketing and Service Section

s

,

• Administrative Judge James L. Kelley Chairman Atomic Safety and Licensing Board Panel U.S. Nuclear Regulatory Commission 4350 East West Highway East West Towers Building - 4th Floor Bethesda, Maryland 20814

• Administrative Judge Glenn O. Bright Atomic Safety and Licensing Board Panel U.S. Nuclear Regulatory Commission 4350 East West Highway East West Towers Building - 4th Floor Bethesda, Maryland 20814

• Administrative Judge James H. Carpenter Atomic Safety and Licensing Board Panel U.S. Nuclear Regulatory Commission 4350 East West Highway East West Towers Euilding - 4th Floor Bethesda, Maryland 20814

• Mary Wagner, Esq.

Office of the Executive Legal Director U.S. Nuclear Regulatory Commission 7735 Old Georgetown. Road Bethesda, Maryland 20814

• Ernest L. Blake, Jr., Esq.

Shaw, Pittman, Fotts & Trowbridge 1800 M Street, N.W.

Washington, DC 20036

• James B. Burns, Esq.

Isham, Lincoln & Beale Three First National Plaza Suite 5200 Chicago, Illinois 60602

• Ms. Marjorie M. Aamodt Mr. Norman O. Aamodt 200 North Church Street Parkesburg, Pennsylvania 19365

w F

i ** Ms. Marjorie M. Aamodt Mr. Norman O. Aamodt P.O. Box 652 Lake Placid, New York 12946 77[dAut.

Maria C. Hensley, Counsel for Gary P. Miller Dated: January 8, 1987 f

I 1

- , - - -, __ _ - _ . , - - , . . , . . _ . . . , - - . , - - , . _ _ _ ,.c-, , , - , _ , _ _ . . - . _ . . _