ML20215J291

Transcript of Jj Blessing 841214 Sworn Statement Re Breach of Physical Security at TMI on 790713

ML20215J291
Person / Time
Site:	Three Mile Island
Issue date:	12/14/1984
From:	Blessing J AFFILIATION NOT ASSIGNED
To:
Shared Package
ML20215J278	List: ML20215J273 ML20215J291
References
NUDOCS 8610240380
Download: ML20215J291 (3)

Text

.

ENCLOSURE 3 SWORN STATEMENT OF J. J. BLESSING TAKEN BY DI ON DECEMBER 14. 1984 k

8610240300 860401 PDR ADOCK 00000320 P PDR

.- . . - -- =- - - .- - -__ _ __ -

k NOTE: This cnclosure discusses information

,. ~ which is th3 subj;ct of an ongoing OI investigation. This enclosure may not be disseminated outside the NPC without coordination with NRR Flace: King of Prussf a, Pennsylvanta and the pennission of the E90 or the Director, OI. Internal access and distribution should be on a Date: December 14, 1984 "need to know basis." ..

STATEMENT 1

1 John Jacob Blessing, hereby make the following voluntary statement to i

Investigators Peter J. Connolly and Rfchard A. Matakas who have identiffed themselves to me as[Investfgators with the U.S. Nuclear Regu.latory Comis-g sfon. I make this statement freely with no threats or promises of reward 1 being to made to me. Investigator Connolly is preparing this statement for me

)  ?*~' at my request.

I am presently employed as an Engineering Specialist with United Energy l Service, Reading, Pennsylvania. I previously was employed with Metropolitan

Edison from October, 1976 to Pay, 1981. I was a Control Operator at Three Mile Island, Unit 2 (TMI-2) from November,1978 untf1 May,1981. I lef t Met Ed in order to seek better job opportunity.

l Mr. Connolly and Mr. Matakas have questioned me concerning an incfdent fnvolving a breech of physical security at THI-2 on July 13, 1979. According to the documentation maintained by Region 1 Office of Investigation, at 11:00 p.m. on July 13, 1979 a safety related nuclear river pump (NRP-1A) trfpped for no apparant reason. I recall that incident.

I was on duty that night in the

{ Control Room of Unit 2 along with my associated Raymond Bocher. I cannot i

recall specifically who was my Shift Supervisor or shift Foreman that evening.

Several day {after this inefder.t while I was on shif t with Raymond Booher, he

/

told me tiatahad a conversation with Jim Weagle at a local bar in the Middletcier., Pennsylvanta area. According to Bocher, Neagle had admitted to him that be had climbed into the nuclear pump house and trfpped the pump. To i

the best of my memory, Booher did not provide any additional detafis of

{ Neagle's conversation. I suggested to Bocher that he inform Bernte Smith, the l Shtft Supervisor that evening in the Control Room about Neagle's admtssion.

l

, Bocher initially did not want to discuss the incident with Smith, but shortly i

i s t

l l

l

.e ..

l thereaf ter, Smith walked into the Control Room and upon that Bocher informed hfe of Neagle's coment regarding the tripping of the pump. , I, was present when Bocher told Smith about Neagle's admission. Smith, in s'qy o'ptnfon, laughed the incident off and did not conduct any followup investigation of that particular matter to sy knowledge. .

Several months later, while I was in training, I heard that Bocher was in trouble over the pump trip incident and was being investigated by the FBI. I discussed this brfcfly with Bernte Smith and I recall Smith making a statement to se

• don't say anything, or you will be involved in the investfgation.' I Q

l was not interviewed by the FBI or NRC on this matter. As a result of the investigation which was conducted into the incident involving the pump trfp, j Bocher recefved a suspension of ten working days without pay from the Company.

(I The reason for his suspension was that he had failed to notify management regarding his knowledge that Neagle was involved in the trippfng of the pump.

However, as I indicated earlfer in this statement, I was present when Bocher had informed Bernte Smith of the incident and Weagle's admission only several days after the incident had occurred. I do not know why Booher did not inform r.enagement that he had nottffed Smith shortly after the incident had occurred.

70 my knowledge, Smith did not come to Bocher's defense stating that he had been informed of the incident. I cannot recall any additional information at.

this time. .

I have read the foregoing statsment consisting of two pages. I have made and initialed any necessary corrections and have signed by name in ink in the .

margin of each page. I swear that the foregoing statement is true and cor-rect.

Signed on December 14 1984 A,o f

.b, at>m[.

SIGNATURE: M chnJ/ cob 81essi Subscribed and sworn to before me this 14th Day of December,1984 at King of Prussia, Pennsylvania.

INVESTIGATORS: _

d/

P.J.Cffnnolly,Invekigator WITNESS: t / _ E Je/ .1 l

- _ _ _ _ _ - _ -..- -_7 , - , _ . . , . , . - _ - -__- __-, -.,, -y y .---, . _

,,-y

• ^ .,

[  %,

UNITEo STATES Enclosure 14 j

s I NUCLEAR REGULATORY COMMISSION wasamotow. o. c. sosss

.....f

January 4,1985 U.S. CC -

1 JZ 13 13. 9 37

,  ; ;;fr-- c .

i.'. l . *... . : .. ;

MEMORANDUM FOR: Harold R. Denton, Director i Division of Human Factors Safety THRU:

Hugh L. Thompson, Jr., Director Division of Human Factors Safety [

FROM:

William T. Russell, Deputy Director Division of Human Factors Safety

SUBJECT:

RESULTSOFJOINTOI/NRRINTERVIEWWITHMARKS.Ch

Reference:

~~ -

.1 .

~

'and W. J. Dircks (EDO) dated April 2, 1984

~

Staff TMI Requirements-Discussion of Pending_ Investigation-2.

Memorandum dated May 3,1984, from H. R. Denton (NRR) to 8. B. Hayes (OI-)

Subject:

NRR Review of 01 Investigation Materials Concerning Hartman Allegations of Falsification of Leak Rate Data at THI, Unit 2 3.

Memorandum from W. T. Russell (DHFS) to H. R. Den (NRR) dated December 14, 1984

Subject:

Results of Joint OI/NRR Investigation and Evaluation of Dennis I. Olson ~

' The purpose of this memorandum is to document the results of the join interview to with Mr. provided confirm information Mark S.earlier Coleman.by Mr. The interview Coleman that waswith Mr. C in conflict

with statements made by Mr. Dennis I. Olson during his November 15, 1984 interview with 01/NRR.

subject of Reactor Coolant System (RCSThe interviews of both individuals d

irregularities at TMI-2 during the per)od i Septemberleak rate surveillande te 30,1978 through the date of the accident Mr. Coleman were March 28, 1979. During that time frame Mr. Olson and the same shift. ifcensed Control Room Operators (CR0s) a,t TMI-2 assigned to

_ NOTE:

j This memorandum and enclosure discuss inforination which is subject of an-ongoing 01 investigation. This memorandum and enclosure may not be disseminated outside the NRC without coordination

01. with NRR and the permission of the EDO or the Director, basis.Internal access and distribution should be on a "need to know

+ w,s

~ a<

Y

l l

l Harold R. Denton January 4, 1985

Background

As a result of a Comission meeting on March 23, 1984, NRR was directed by Reference 1 to review 01 investigation materials concerning falsification of RCS leak rate surveillance tests at TMI-2 and refer back to OI those matters which required further investigation. The results of NRR's review was provided in Reference 2.

The review determined that follow-up investigation by 01 and further evaluation by NRR was needed in the case of seven licensed operators.

Mr. D. I. Olson, currently a Senior Reactor Operator (SRO) at Waterford 3, was one of the seven individuals identified as needing additional investigation and evaluation. Mr. Olson was interviewed under oath in New Orleans, Louisiana by 01/NRR on November 15, 1984. During the interview, many of the statements made by Mr. Olson did not appear credible in light of the technical evaluation of Mr. Olson's leak rate surveillance tests and statements!made by Mr. Coleman on April 10, 1980. The results of

~

the joint 01/NRR investigation and evaluation of Mr. Olson is documented in Reference 3. Because of Mr. Coleman's candid admission in April 1980 that he was involved in leak rate falsification at TMI-2, when first questioned about his involvement, it was recomended in Reference 2 that no follow-up investigation of Mr. Coleman was required.

Past involvement in TMI-2 Leak Rate Testing Irregularities 4

On December 14, 1984, a joint 01/NRR interview with Mr. Coleman was held in the Office of Investigations Field Office, Region I. The interview was conducted under oath and in the presence of Mr. Coleman's personal attorneys:

Mr. Bart Gephart and Ms. Jane Penny of the law firm Killian and Gephart. The interview was conducted by Mr. Keith Christopher, Director, Office of Investigations, Region I and Mr. Robert Capra, Technical Assistant Division of Systems Integration, NRR.

A complete sumary of Mr. Coleman's interview is provided as an enclosure to this memorandum. The enclosure also incluoes a Report of Interview of Mr. Coleman's April 10, 1980 interview with IE and 0IA and a sumary of leak rate surveillance tests involving both Mr. Olson and Mr. Coleman during the period under investigation. The following key points were discussed during the interview:

1.

Mr. Coleman had little faith in RCS leak rate test calculations which were performed to show compliance with the Technical Specifications.

2. Leak rate test results were erratic. Acceptable leak rate test results were more difficult to obtain as the date of the accident approached.

3. The operators on his shift, including Mr. Olson, were concerned about operating the plant with high identified leakage caused by valve leakage from the top of the pressurizer. Mr. Coleman stated at one point Mr. Olson recomended to his Shif t Supervisor that the plant should be shut down to repair the leakage.

4 It was common practice to run leak rate tests several times per shift until an acceptable result was obtained. The acceptable tests were retained and the unacceptable tests were thrown away.

Harold R. Denton January 4,1985 5.

Mr. Coleman believed that unacceptable test results were thrown away so they would not be seen by the NRC. These tests were thrown, away at the i direction of or with the knowledge and consent of Mr. Coleman's Shift l Foreman and Shift Supervisor.

6 According to Mr. Coleman, no changes were made in leak rate test practices following the October 18, 1978 incident that resulted in the generation of LER 78-62/1T.

7.

Contrary to the comitment contained in the narrative of LER 78-62/1T, Mr. Coleman was not instructed to enter the Action Statement of the Technical Specification when leak rate test results exceeded the Limiting Condition for Operation.

8.

Mr. Coleman believes it was comon knowledge among , operators, including the operators on his shift, that hydrogen additions to the MUT would influence MUT level indication and consequently leak rate test results.

9.

Mr. Coleman used hydrogen additions to the MUT to assist in obtaining leak rate test results that were within the acceptance criteria of the Technical Specifications.

10. While Mr. Coleman could not confirm whether Mr. Olson had ever made hydrogen additions to the MUT to alter leak rate test results, he mentioned adding hydrogen for this purpose to Mr. Olson on at least one occasion. Mr. Coleman stated that Mr. Olson acted like he did not want to hear about it. Consequently, Mr. Coleman did not recall discussing the subject with him again.

11.

Mr. Coleman stated that he was aware that adding water to the MUT during the last few minutes of a leak rate test could produce the same effect on leak rate test results as adding hydrogen to the MUT. Mr. Coleman stated that he use this method to manipulate leak rate test results during the last two months before the accident since hydrogen could no longer be added to the MUT remotely from the Control Room.

12. Mr. Coleman could not state whether Mr. Olson ever manipulated leak rate tests by the addition of water to the MUT during the last few minutes of the test. However, after reviewing the leak rate test data, he confinned method.

it would be hard to believe that Mr. Olson did not use this Conclusions During the period, Mr. Coleman served as a CR0 at TMI-2 on the same shift as Mr. Olson, Mr. Coleman was involved in activities associated with leak rate test falsification. While Mr. Coleman could neither confinn nor deny that Mr. Olson was also involved in these same actions, it is difficult to believe that Mr. Olson either did not participate in leak rate test manipulation or have direct knowledge that this activity was going on around him.

9 Harold R. Denton January 4, 1985 Mr. Olson stated that following.the October 18, 1978 incident, he was advised of the pmper interpretation of Technical Specifications and that he never threw away unacceptable leak rate tests following that incident. He stated he would staple the bad leak rate test results to the good test and turn them into his Shif t Foreman. There is no indication that the Action Statement of the Technical Specifications was entered by any of the operators on their shift, either before or after the October 18, 1978 incident. Had Mr. Olson properly applied the requirements of the Technical Specifications, the Action Statement would have been invoked whenever unidentified leakage exceeded 1 gpm. Mr. 01 son's claim that he did not throw away bad leak rate test results and turned them into his Shift Foreman confifets with Mr. Coleman's statements that his Shift Foreman and Shift Supervisor "did not want to see bad test results."

Mr. 01 son's statement that he was not aware that hydrogen additions to the make-up tank could alter make-up tank level and consequently influence leak

_ rate test results is also in conflict with Mr. Coleman's statement that this phenomenon was connon knowledge among CR0s. Mr. 01 son's statement that he was not aware hydrogen additions were_being made for that-purpose-is-in -

conflict with Mr. Coleman's statement that he mentioned using this technique to Mr. Olson on at least one occasion.

Mr. 01 son's statement that he never added water to the make-up tank during l the last few minutes of a leak rate test in order to influence the outcome of a leak rate test also does not appear credible in light of the technical data and Mr. Coleman's admission that Mr. Coleman used this technique to manipulate test results. Mr. Coleman and Mr. Olson were involved together in only two tests that exhibited this pattern. However, Mr. Olson was involved in five tests with Mr. Wright (the other CR0 on their shift) and one test where Mr. Olson alone was involved where water was added to the MUT during .

the last few minutes of a leak rate test. Ihese additions produced a higher i

MUT level increase than was recorded in the CRO's Log or compensated for in the leak rate test calculation.

~

In sunmary, based upon Mr. Coleman's testimony, coupled with the technical evaluation of the leak rate tests involving Mr. Olson I believe the conclusions reached in Reference 3 remain valid and that Mr. Olson was not truthful in answering questions regarding his role in leak rat'e test

. -manipulation.

o7 William dlDeputy Director T.1 Russell, Division of Human Factors Safety Office of Nuclear Reactor Regulation

Enclosure:

As stated cc: 8. Hayes K. Christopher, RI J. Liebeman

_u- . . . a. .-a.  :.. . . . _ ..

l .

.> - v agr rr .

a 3F ar DISTRIBUTION M

Docket Files 55-06404 Docket Files 55-06121 JAN 0 -11933 TMurley RI RMartin RIV RCapra ci LWiens

! WRussell E MORANDUM FOR: W1111aa T. Russell. Deputy Director MThompson Division of Human Factors safety. NRR t

3 '

FROM: Robert A. Capra. Technical Assistant

.; Division of Systems Integration. NRR

SUBJECT:

SWg4ARY OF JOINT O!/NRR INTERVIEW WITH MARK 5. COLEMAN The purpose of thfs memorandum is to document the results of the jotat DI/NRR interview with Mr. Mark 5. Coleman. The purpose of the interview with Mr. Coleman was to either confirm or refute statements made by Mr. Dennis I.

- Olson during his November 15,1984 interview with O!/NRR. The interview of both individuals dealt with the subject of Reactor Coolant System (RCS) leak rate survet11ance testing irregularities at TMI-2 during the period Septem6er 30,1978 through the date of the accident. March 28,1979. During that time frame Operators (CR0s} at Mr. THI-2 01assigned son and Mr. to the Coleman same shift. were licensed Control Room The intervfew with Mr. Coleman was held in the Office of Investigations Field Office. Region I. The interview was conducted under oath in the presence of Mr. Coleman's personal attorneys: Mr. Bart Sephart and Ms. Jane penny of the law fira Killian and Gephart. The interview was conducted by Mr. Keith Christopher. Of rector. Office of Investigations. Region I and me. Throughout

, the intervfew Mr. Coleman was very coo

questions in a straightforv:rd manner. perative Mr. Coleman and appearedconfirmed to that answer pre. all accident leak rate testing irregularities took place at TMI-2. Mr. Coleman' y admitted his involvement in activities that were in violation of approved M plant procedures and TMI-2 Technical specffication requirements. A complete i . saunary of his interview is provided below.

After being placed under oath. Mr. Coleman was asked to describe his employ-ment history prior to the accident at TMI-2 on March 28, 1979. Mr. Coleman stated that he served six years in the U. 5. Navy's nuclear power program

prior to joining Met-Ed in January 1974. He was hired as an Auxiliary Operator on TMI-1 and served in that capacity untti he went over to TMI-2 as a CR0 trainee in 1976. He received his Reactor Operator's license in October

, 1977. During the majority of the period under investfgation. Mr. Coleman stated he was assigned to Shift *D." Shift "D" consisted of the following licensed individuals:

' NOTE: This memorandum and enclosures discuss information which is the subject of an ongoing 0! investigation. This memorandum and enclosures may not be disseminated outside the NRC without coordination,with NRR and the permission of the ED0 or the Director, v4. an6. u.. 66... .nu ui m ivu6 ion am uiu oc un . ,cco so anew

................... A s.11.. .. . . . . ... . . ..... .............. . . . . . . . . . . . . . . . . . . . . . .... ............... . . . . . . . . . . . . . . . . . . . . . . . . . . . - ~ . - . .

'"'..g"k .................... - - . . . . . ~ ~

.................. .................... ..................... ..u................ ..................... ..............-...* " " * " " * * "

• ac ronu m no,eo acu oua OFFICIAL RECORD COPY *W^**"*" -

+~ ~- - *= -.. : ; squ" *=" "" ***

A  ;- r 7 M' t* * *"* - **: * *****t'\,***'

., e .e r, g g ,- . s - .

G. , . .

a v ' < G'

.4 s  ; -

!..r*'

F W F F r Mr. Russell 2-JAN 0 41385 SHIFT 'D"

). Shift Supervisor: Gregory Hitz i'

Shift Foreman: Adam Miller CRO: Dennis Olson j CRO:

7.1 Lynn Wright

• t CRO: Mark Coleman Mr. Coleman April 10,1980was shown a copy of the " Report of Interview" of Mr. Coleman's interview with IE and 0!A on this same subject. Mr. Coleman confirised that the infomation contained in the Report of Interview was'true and correct to the best of his knowledge (see Enclosure 1). '

Mr. Coleman was provided a copy of TMI-2 Surveillance procedure 1301-301

• Reactor Coolant S

( Specification Lintting Condi(TS)ystem Inventory" and a copy of TMI-2 T

.y tion for Operation." After reviewing the documents.

a Mr. Coleman stated that leak rete tests were routtnely done by using the plant computer. The test duration was usually cw hour. At the end of the test the operator would need to provide infomation to the computer to account for operator-induced changes to the RCS such as water additions to themake-uptank(MIT). Mr. Coleman confirmed that the Technical Specification 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. However required that a leak rate test be run at least once every leak rate (i.e., un It was operating policy to try and obtain a satisfactory identified leakage within the Technical Specification limit of 1 spm) once per shift if the plant was in steady state and the computer was available. Mr. Coleman stated that only good leak rate tests

'l were retained and logged in the CR0 Log. Leak rate tests showing -

unidentified leakage that did not meet the acceptance criterla of the te Technical Specifications (i.e., greater than I spa) were thrown away and were

- not logged in the CR0 Log. Me stated that bad leak rate tests were thrown

%~ '

away at the direction of or with the knowledge and consent of his Shift Foreman and Shift Supervisor.

' Mr. Coleman was asked how he interpreted the requframents of the Technical i

Specification with respect to when the Action Statement was required to be i

.j entered. Mr. Coleman said that since the leak rate test was required only once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, he and the other operators interpreted that to mean that once a good leak rate test result was obtafned, they nad the fell 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> to a

3 get another good leak rate result before they had to enter the Action Statement. Coleman stated that re were received during the 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />,gardless they would of not howentermany the 69d leak statement.

Action rate tests NOTE: Surveillance procedures 2301-301 states that if unidenti-fled leakage exceeds 1 ppm proceed with Action Statement 3.4.6.2.b.

4

.............n... ..........n....un. .......o.a.ouan o*~.ona-"".

""-" ~"" "*a"""""*"" * " " * " * " " " *

.................. .................... . . . . . . . . . . ~ . . . . . ~ . ~ . ~ . . ~ . - ~ . ..--...~-~.- ~ ~ ~ . - ~ ~ ~ - --

.................. .................... ..................... -...........-- .................... . . - - ~ ~ ~ ~ - - - - - - -

ac ro== sia no,so. =cu es.o

..... . . . . OFFICIAL RECORD COPY _. * " " * "E f.

. - . . . . . - . ..~.w..-- -

r,

- - - - , ,, .- ~ ~ - . . . , . ...,_1

--r - m , ., 7 -ec.r " ~ ~

~

=W.* ~ " ' ~ ., R.  :-

-- y

_ _ .m _ ~ .. _ . . . . _ . . . . - .-.m_

y j .

Mr. Russell ,

JAN 0 4 E5 4

Action Statement 3.4.6.2.b reads: 'With any Reactor Coolant System leakage greater than any one of the.above

. limits. excluding PRESSURE SOUNDARY LEAKAGE. reduce the

,., leakage rate to within Ifmits within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> or be in at least NOT STANDBY within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in COLD

{n. SHUTDOWN within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />."

Mr. Coleman stated that when a ba'd result was obtatoed. they would rerun the test until an acceptable result would be obtained. In addition, when actual

' leakage was suspected, operators would be tasked with looking for leaks in the plant. Mr. Coleman could not recall ever entering the Act, ton Statement of Technica1>5peelfication 3.4.6.2 A copy of LER 78-62/IT along with an attached routing sheet with the type-i written names and handwritten initials of control room personnel was provided

" to Mr. Coleman for his review. The routing sheet was used to indicate which individuals had seen the LER. Mr. Coleman confirmed that he had seen the LER

) before and that the initials on the routing sheet beside his typewritten .name

4 were his. Mr. Coleman stated that while he had seen the LER and initialed i j the routing sheet it was just one of many documents that he may have looked I d

at and initialed during one of his watches. At that time, the document did not stand out as being significant in his mind.

NOTE: During a routine inspection of TMI-2 operations on October 18, 1978, an NRC inspector discovereo several bad leak rate tests lying in the control room and that

.: THI-2 had been operating for two or three days with 1 '

unidentified leakage exceeding the Technical Speciff- .

cations. The incident resulted in the submittal of I Lice,nsee Event Report (LER) 78-62/1T.

7 i The LER states in part: *This event was caused by misinterpretation of the requirements of the technical

. specifications. The appropriate personnel will be

instructed on the requirements of applicable sections

.,, of the T.S. and the requirement to isreediately invoke a applicable action statements when the provisions of the 1 LCOs [Itatting condition for operatton] are not met.' '

r Mr. Coleman was asked if his understanding of the requirements of Technical

( 5pecification 3.4.6.2 was changed in any way following this incident.

Mr. Coleman stated that his understanding had not changed. He could not recall receiving any directions from his supervision to perform leak rate tests differently following the incident. He stated that leak rate tests d, that did not meet the requirements of the Technical Specifications continued 1" to be thrown away and, as best he could recall, he never entered the Action Statement of the Technical Specification. Mr. Coleman stated that it was his understanding that leak rate test results that did not meet the requirements 1' of the Technical Specifications were to be thrown away so the NRC would not

m. He stated Ents was not told to Alm d' rectly, Dut i t was implied,

> . ........S .........ee th . . . .e................. . . . . . . . . . . . . . . . . . . . . . ................~... ..................... .......~~~~~~~ ~~~~~ ~

"* **

• 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..................... . . . . . . . . . . . . . . . . . . . . . ..................... .......~.--~~ ~ ~ ~ . .

• • " I .................. .................... ..................... . . . . . . . . . . . . . . . . . . . . ..................... . . . . . ~ - - - - - ~ ~ ~ -

=ac ronu m no, o.=acu e n o *"**2"*-**

i OFFICIAL RECORD COPY --

~~'~

. . . . . . . . s i. .r . wes . . . .=mem. - e e ~ * - ~ w- % - =. u - - ' *-

, ' ~~'"" . 5" ': '~~ 'q.7. "" "K"Y~pr'r~f7? ~~ .~ w M C' ~ ~ ~

~ ' ~ ~ ~ ' ~ " ~ ~ ~

,3

..~a.aa......._._....._ .. .. ,. _ . , . .

~

• s y er y ar ,

Mr. Russell 4 JAN 0 41385 s

NOTE: Mr. Coleman's testimony that bd was not instructed in J the proper interpretation of'the Technical d Specifications and that he was still directed to throw q away bad leak rate test usults after the October 18

}j 1978 incident conf 1fets with testimony given to O! by ,

Mr. Olson.

4 M., Mr. Colaman stated that good leak rate tests results were harder to get as q the date of the accident approached. He stated pressure was exerted on operators to get good leak rates. The pressure increased the closer they came to the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> limit they believed was required by the Technical Specifications. Mr. Coleman stated that he could recall one instance where he was told he had no other duties during that watch except to keep running leak rate tests untti he received one that was within the acceptance criteria y_ of the Technfcal Speciffcations. In spite of the pressure. Mr. Coleman stated that he was never directed by his supervisors to intentionally 9 manipulate the test results. Mr. Coleman stated the pressure stemmed from F.i not wanting to have to shutdown the plant. With the plant off-line. they fi were not generating electricity and consequently the company was not making any money. He stated that he and the other operators owned stock in the

'1.

company and they felt it was their job to generate electricity.

4 Mr. Coleman was questioned about the high amount of identified leakage they were experiencing from the pressurizer prior to the accident. Mr. Coleman believed the high leak rates were caused by leakage from the p0RY and not one of the two code safety valves. He thought that a 1ct of the problems they were experfencing with obtainfn good leak rate test results for unidentified

.~ ; leakage were a result of the hf h value of identified leakage. In spite of.

the high leakage 1t was his bel of that the PORY block valve was not shut for N two reasons. First o to relieve p ussure Mr. Coleman stated they to keep the reactor fromwanted trippingthe on p0RV high pressure to be available in the event of a transfent. Second, he believed that there was a concern. if they shut the block valve they afght not be able to reopen it as a result of the high differential pressure that would be present across the valve seat. He said the high leakage was a concern to operators. While he stated that he never made a reconnendation to shutdown the plant to repair the leakage he F stated that Mr. Olson made comments to that effect during one of their d routira, on-shift meetings with their Shift Supervisor, Greg Hitz.

i Mr. Coleman said the general attitude at the plant at that time was to waf t i untti the plant came down for some other reason and they would make the

.; repairs at that time rather than' shutdown the plant specifically to repair the leaking valve.

Mr. Coleman also attributed problems with obtaining good leak rate test

, results to errors in the computer program for calculating leak rates. He a '

fsit, based upon discussions with Mr. William Falls, the computer programer.

that Mr. Fells was working on trying to resolve some of the problems.

Mr. Coleman was then questioned about the effect of hydrogen additions to the nus on sean rase sen,rusuiss. rw. coseman sta ;ea snat ne nrst searnec taat

. . . . . . . . . ~ . . . . . . . . . .....-~~..~.

" **"" Q . . . . . . . . . . . . . . . . . . . ..................... . . . . . . . . . . . . . . . . . . . . . ..................... .... ~ - ........... . . . . . . . . . . ~ ~ ~ ~ . ~ . . . -

Saft h ..................

l ..................u ..................... ...o............... ...................n ..."*aa""a*". " " " " * " * * " *

• cfoau m no,so..acu os.o

• um m **

- .. -.... OFFICIAL

....  % m RECORD ..m . COPY...

~m ---. 7- - r -- ; ,n.: .. ;v, ;. mm.gyy,r.p-. ~_- .~ n--~ .. .? - - - . - ---.

~'

Mr. Russell 5- JAN 0a m:,

hydrogen additions could influenca the outcome of a leak rate test from

~! Mr. Harold Martman. He said his shift normally reiteved Mr. Martman's shift d and that Mr. Hartman had told him about it during one of their'shtft turnovers. Mr. Coleman said for a period of time, hydrogen could be added directly to the PtJT from the control roca. Mr. Coleman stated that he 1 personally had observed the rise in MUT level indication when hydrogen was added. He stated that this level rise could be accomplished with a at,rmal.

e *' controlled addition of h I duration addition Cf.e.,ydrogen. It was not necessary to make only a short

!  : one or two seconds). Ne said while

• chemical.

additions

• during leak rate testing were to be avoided according to the .

procedure, he did not consider a hydrogen addition to the make-up tank to be a chemical addition. Me also stated, the rise in MJT level indication produced a transitory effect lasting only a few minutes. Therefore if an operator was gofng to use hydrogen to influence the leak rate test hydrogen i

_ would need to be added very near the end of the test.

M Mr. Coleman believed it was casuon knowledge by CRos. Shift Fommen and shift J

/ Supervisors that hydrogen addf tfons could alter MIT 1evel indication and.

l therefore. influence leak rate test results. However, he could not recall ever discussing this with anyone except Mr. Martman. Mr. Coleman was asked 1 if he ever discussed this subject with Mr. Olson. Mr. Coleman stated that i one time he recalled mentioning to Mr. Olson, during the performance of a leak rate test, that they could use hydrogen to increase the MUT level to help them pt a good leak rate test result. According to Mr. Coleman.

Mr. Olson covered his ears" and " acted Itke he didn't want to hear anything about it." Mr. Coleman said he did not ever recall discussing that matter with Mr. Olson after that.

NOTE: Mr. Olson testified that he never added hydrogen to the V.

MIT for the purpose of altering a leak rate test. He j -

stated that he was not awam of any operator adding hydrogen for that purpose. Mr. Olson stated that he did not have an understanding at that time that hydrogen additions could affect leak rate test results or that he had ever heard from anyone that hydrogen a

could alter the results.

Mr. Colanan could not recall ever seeing Mr. Olson add hydrogen to the MUT for the purpose of altering a leak rate test. He noted, however, that it was hard for him to believe that any of the operators did not know that this was

occurring.

Mr. Coleman was then questioned on the use of water additions to influence

' leak rate tests. Mr. Coleman was shown a listing of leak rate test results involving the operators on hfs shift. A copy of these summary sheets is provided as Enclosure 2 to this memorandum. It was pointed out that every test performed on their shift between December 27,1978 and the date of the accident (16 tests total) involved either a hydrogen addition (1 test). a water addition that was not included in the computer calculation (3 tests) or a water aco m on wnere sne amount inclucea in t1e computer ca lculation was

. u...n.......... ..................n. .no.u...o....ua o . . . . . .. . . m p o.

• u o.* m n a n o *. .m num."""""*" """"""*""

............u.... ..........n......... ......... mon.*... n.............*om m*m.ounoom ."**"""*"""

"> .................. .................... ..................... .................... . . . . . . . . . . . . . . . . . . . . . m...m m - m m

-- . m m m .

acronusiano e ==c=o 4o OFFICIAL RECORD COPY

. - - . - - m . - -..

• "'; *%7_*?

1--.rym ~p- _. - . . t- .-. g-} ,mp.7;; jy.m:g"- ?m- - -- *~ ~ - - * '

" ~

il ' ' ' ' . $"

i W G7 p g,

. Mr. Russell JAN 0a m less than the amount indicated on the M strip chart (12 tests). Out of the 4

12 tests involving under-recorded wa.ter additions.10 tests show that water was added during the last 10 minutes of the test. After reviewing the data.

Mr. Coleman was asked if his shift had discovered that by adding water to the

.; E during the last few minutes of the test (thus compressing the hydrogen gas blanket and tecreasing the pressure in the M) they could produce the

' same effect on M level indication as was possible by the addition of hydrogen. Mr. Coleman confirmed that he was aware that by adding water during the last few minutes of the test, he was able to get the M 1evel indicator to read a higher value than the actual amount of water added to the tank. Thus. Mr. Coleman conffreed that he was able to add for example,150 gallons of water to the tank and have the M 1evel indicator. show an apparent increase of 180 or 200 gallons. He would log the 150 gallons in the CRO's Log and include the 150 gallons as an " operator-caused change" in the computer. However, since the computer would read the final level in the M 1._. as 30 to 50 gallons higher than actual level, the leak rate calculated by the q computer would be .5 to .8 spe lower than the true leak rate. Mr. Coleman w

M did not recall if he had discovered this phenomenon himself or whether he had

'.j

- been told about it by another operator. Mr. Coleman was then asked if this ,

.1 method was used in Ifev of hydrogen additions because it was not possible to

~i add hydrogen from the control room during the last few months before the accident. Mr. Coleman stated that was the reason. Me said since you needed to add hydrogen very close to the end of the test it was not very practical to try and have the Auxt11ary Operator (AO) add hydrogen locally and be able

, to insure it would be added to the M at just the right time.

It was pointed out to Mr. Coleman, that 2 of these tests involved both

~

Mr. 01 son and himself. Two other tests involved Mr. Coleman by himself and.

Mr. Coleman and Mr. Wright. Mowever. 5 tests involved Mr. Olson and

Mr. Wrfght and 1 involved Mr. 01 son alone. Sased upon the fact all three i - operators on the shift were involved in these types of additions it appeared

/

that all three operators used this method of adding water to manipulate leak rate test results. While Mr. Coleman confirmed that he was involved in this activity, he could not confirm or deny whether Mr. Olson intentionally added water in this manner to manipulate leak rate test results. He stated.

however, that based upon the data he could see where it would be hard not to conclude that Mr. Olson was aware of and used this method to alter leak rate test results. Mr. Colesen was asked if water additions needed to be made close to the end of the test for legitimate operational reasons such as to move control rods. Mr. Coleman stated that more and more water needed to be added as the date of the accident approached due to the increased valve leakage from the top of the pressurtzer. However he did not believe water would have to be added close to the end of the test on a routine basis for aty legitimate reason. He stated that because of the increased leakage, boron would concentrate in the pressurtzer. In order to make-up the water lost due to leakage and to equalize the boron concentration, dominera11 red (DI) water would be added to the make-up tank and pressurtzer spray valve would be placed in manual.

• "'c4 .................. ..................... . . . . ~ . . ~ . ~ ~ . ~..-~~~.~ ~..~.. ~ ~ ~ . . ~ ~ ~ - " " -

• " * *" O .................. ..................... . . . . . . . . . . . . ~ ~ . " . - . . . ~ . ~ ~ ~ ~...~~.~ ~. .~~.~~." " "" " " " ~ ~ ~ "

.................. . . . . . . . . . . . . . . . . " . . ..~*.*~..~ "." ~ " " ~ " ~ " ~ "*~"""~**-- " " * " " * " " " * " * * " * " * " " " " *

=ac ronu m no,so.wac= ono * " " * " "

. ...~.. a ... ,.... ~ em= w '~ ~ ~ ~~ -

OFFICIAL RECORD COPY

..- ' " "~~~'~~

, , e. , -

. =..v. . . - . . --- . ~ - . - -

i

.' Y' h 5 5 f f 4 Mr. Russell JAN 0 41885 NOTE: Mr. Olson stated that he could not explain why his shift needed to add water to the makeup tank 30 close ,

to the end of the test on a routine basis. He stated '

that the only possible explanation he could think of  !

was to change boron concentration in order to move control rods if they were right at the edge of the band they were supposed to maintain.

=

In addition to adding water to the WT during the last few minutes of the test, Mr. Coleman was asked about other methods by which water additions could be used to alter leak rate test results. Mr. Coleman stated it was possible to add water to the NJT and not include the addition in the computer calculation. He was'also asked if it was possible to add small amounts of water to the MUT over the course of the test such that this ' Jogged" water addition would be difficult to detect on the WT 1evel strip chart.

<- Mr. Coleman stated that both of these methods could be used; however, he had i no direct knowledge of arny operator making these types of additions to alter

' leak rate test results. Mr. Coleman also stated that while the operator who

-' ran the leak rete test was not normally the individual who had the panel.

kept the CR0 Log and made water additions, it was possible for either to make a water addition without the other operator being aware that the addition

.', was made.

Mr. Coleman retterated that he could not recall discussing methods by w5fch leak rate tests could be manipulated with anyone but Harold Hartman and Mr. 01 son on one occasion and dented that he had direct knowledge of anyone beside himself actually performing actions that would alter leak rate test results.

wassaaA s,laned bgg -

't Robert A. Capra. Technical Assistant Division of Systems Integration

Enclosures:

1. Report of Interview with Mr. Coleman

2. Sumary of Leak Rate Test Data cc: K. Christopher (RI) 4 DW/TNIH2/COLEMAN ENCLOSURES

~'c4 . p.H. U.. s d .. ..................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .................... . . . . . . . . . . . . . . . . . - . . - - - .

• • -O .RCAnrAm.1... ..................... ..................... ..................... ..................... ..................- - . - - . ~ .

1/4 /.?.s.,,,,, ,,,,,,,,,,,........, ...................,, . . . . . . . . . . . . . . . . . . , , , , , , , , , , , , , , , , , , , , , , , ...........___ _ - - ...

=ac rc== sie no,so.uncu om o OFFICIAL RECORD COPY ***"*"*-*

. . . . . . . . - . . - .. . . . ~ n . .a-u ,==<~ - - ---

. . . . - . . . ..~,.m. ,p,. . . .

,y , ,,.,. _. .

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

?

t - t

.} .

a.l Su (f4 v,

O.) 3

, . ENCLOSURE 1

, REPORT OF INTERVIEW WITH MARK S. COLEMAN BY IE & O!A ON APRIL 10, 1980 -

.t_

5 b.1

.l a

e i

e a

h a

k l

1 i

9

) p p . l _, -

• I - =^

( -f J K l h*l V i

• - = +=. . . .

. . . q -.,7 m- _

f

?, s. .,~ . ? *.

• AEPORT OF INTERVIEW Mark 5. Coleman Control Room Operator. Unit #2 Three Mile Island Nuclear

, 1 Station. Middletown, pennsylvania was interviewed onApril 10. 1980, by David H. Gamble. Office of Inspector and Auditor. U. 5. Nuclear Regulatory 3

1 Commission (NRC) and Donald C. Kir! patrick. Office of Inspection and i , Enforcanent. NRC.

Colanan wap accompanied by Eric L. S. Strahn, ettorney for the Metropolitan Edison Company.

,. At the conclusion of the interview Coleman nas asked whether he wo i

willing to sign a statement doc eenting what he had said. Colman initially

' .j replied that he was basically adverse to signing anything. After consulting with Colanan. Strahn requested that such a statement be prepared and they (Strahn and Coleman) would decide whether Coleman would sign it after rev it.

Gamble then drafted the attached four-page statement which Colanan carefulhy i

reviewed.

Strahn challenged the accuracy of several passages; however. Coleman affirmed each passage as being accurate. The only changes Colenan made were (1) grasmatical changes to page one which he initialled and (2) the rephrasing of a sentence on page three. After these changes were made. Coleman reread the entire statement several times and stated that it was all true. &

41ated that;,hgj,wan,te&N.Dmget_sut" At he didmetanosda.adcas Etatement which could be used to "hmmer."Aia.letsem

- ---~~--~ ~ ~ . . r--

. ; y n , y W. n . -- . C'* W '~"- ~~ ~~ ~ ~~~ ~

' k rk S. Coleman 2 The page three change concerned the response Coleman received from sup personnel after turning in a record of an unacceptable leak rate test.

During the interview, Colunan said he turned the test in to his foreman who turned it over to his shift supervisor (names not recalled). Colman did recall that the responsa he received (presimably from the shift supervisor) was i

to the effect of " Don't bring me this shit; I want good test results." Upon reviewing this ' passage. Strahn called this passage to Coleman's attention.

Coleman repeated that this was what ens said to him; but he asked that it not b identified as a ' quotation, since he might not have recalled the statenent w6rd-for-word.

.. Strahn then asked Coleman to identify who made the statement:

i

~

Coleman said it was the shift supervisor. As the discussion ensued Coleman then said that he reembered a group of supervisors being present when he was y

told this and he could not recall which member of the group ande the statanent.

Coleman said he honestly could not recall who was in the group, but it definit consisted of at least a foreman and a couple shift supervisors. Strahn questioned how more than one shift supervisor could have been present at any

. .l

.y one ties: Coleman replied that several shift supervisors are present on the day shift.

Strahn then. suggested that Colonen revise the ' Don't bring me any shit..." statement to clearly show its context. After several alternatives were suggested by Strahn (e.g. " negative test results are unacceptable"). Coleman agreed to substituting Kirkpatrick's suggested language:

• negative test results were not wanted to be seen" .

l .

Colanan said he understood the purpose of the leak rate test to be finding out how much identified and unidentified leakage existed. He' believed the basis for performing the test was the potential for radiation leaking to the atmosphare.

).'*

_ _ . . , e e -- ~ - - . m ~ ' W ~ ~ ~ '~~ ~ ~ ~ ~ ~~ '

Mark 5. Colenan 3 In response to 16 question as to whether there was a problem getting good t results, Colanan said there were a "lotta" leak rate tests with bad results.

Coleman attributed this problem to (1) changes in power. (2) a large known leak into the reactor coolant drain tank (RCDT), and (3) difficul, ties the 3

computer program had in taking into account differences in water density in j

the pressurizer and the RCDT. He said that he knew there were problems with s

the computer program because Bill Felds, the programer, was frequently work on it.

Golman mentioned a amber of ways that were commonly Ancest cauld laprove the test results:

Q).paping the RCDT; (2) putting antar.Ja..the ,

a make-up tank; (8) " flip-flopping" antar between the make-up taak and.the pressurizers (4) adding Iqydrogen; (ft).edding water without documentias.Aha

. addition in the computer; and, (6). adding water even with proper. h="atation in the computer (due to the proprea's failure to account for the1Ha p=edae in water densittaa). As indicated in the attached statement. Salas adding hydrogen during tests. His exact words on this point were,4&.awst

{ j gave added Igydrogen because, if there was a any to taprove your situation., ~

g I would do it".

Coleman also said that having to add Igydrogen to get a good leak rate test result did not indicate to him that there necessarily was a leak rate probles; Coleman believed that the discrepancies in the results caused by the hydrogen addition was a nomal function of the instruments. Coleman said that he knew that hydrogen addition would effect the leak rate'because he had observed the make-up tank level indication rise during hydrogen addition.

ll He said that this indicator tes the same one used by the computer.

(

Attachrnent: As stated l l

1 C'AW iM HKi Donald C. Kirkpatrick David H. Gamble l0G l

• • 29*- - " * * * * -

? ? ** ** *. ' *

• F YY 'U'

' ' ~

W. 5. IRlCLEAR REGULATORY CettIS$10N TRANSCRIPT OF STATDtENT DATE: .

TIME:

1 Mark 5. Colman do hereby aske the following free and voluntary statement to Mr.' David H. Gamble, who has identified himself to se as an Investigat for the U. 5. Nuclear Regulatory Ccamission. I understand this statment is

] being ande in connection with an official NRC Investigation and any, if l

necessary, be used in judicial or thinistrative action. I aske this statament with no threats having been ande against me or promises extended to me.

In sqy capacity as a control room operator at Unit #2 of Three Mile Island Nuclear Station I performed leak rate tests during the period from startup through March, 1979.

I do not believe that I have ever added unter to the system without doceenting the addition to the cceputer in an attempt to alter the results of a leak rate test to make it appear favorable, when it would not otharwise have been. If I have ever so added water without documenting it in the computer, it would have .

been inadvertent.

I believe I have added unter inadvertently during the time my partner was performing a leak rate test. In that situation, we discarded i

the test results since they were inaccurate.

a I know that there are a number of ways that test results could be altered. For ,

example, ona method - which was not prohibited by any procedures - was the addition of hydrogen during the test. I believe I have added hydrogen during a test

....-.........-.--.-,-.___,.-r-m-~~~-------

-v--- ~~

Mark 3. Coleman 3 ,

I have read the above statement, consisting of four pages and find it true and complete to the best of sy knowledge and belief.

E e

Subscribed and sworn to before me this 10th day of April 1980.

A e

G I

e 9

e S

+

aio M et *O -

B e4 ge 99 &y @ Me 4 6 SA S*P'M *D'N T' IE w y - ,, . - - - ,, . - - -- - --.. -

___mm-,, . , , , _ ,,-, y , ,_ ,, .-, - , , - - - - eyy .,

7 w

, . , , ., , U.S. NUCLEAR REGULATORY Com!5510N i .

SATE TIME I _ jf._,. k S.

I

. r mew do hereby make the fb11owing free and voluntary statssent to Mr. David H. Gambli. sho has fdentified himself to me as an Investigator for the U. S.

, Nochear ReguIatory Coenission.

I understand this statement

_- is betag made la connection with an officfal NaC Investigation

' and say, ff necessary, be used in judicial or administrative action.

I make this statement with no threats having been ande against me or promises extended to me.

.L y c. p , H, as . est c. . -

y ic- A va+ + z. f n.. /cle WJ R&v SML / Z wC,ej ja.L ak.

sth MJ b b r\ ok -

/

L. } pp.L 0 79.

1 2o ,J- A (: evt. b flatI W% ever odha mW +v 11 syshm-

-ifL. ,#

, c. y -,,d ;_O  % alJ? % % ~N w>;v- k . - a-%g& +s, aikr 1L re.s JIf., fa l<< L. c4k -

b /- ta r4 ih opptov~ hvwbin.

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

, H Nn oY '~'* dck p r-- ,,----- - - - - - - - -- -

---,---,m -

ww,-a-.,, -w-------m.,

.. = --

.v:2=

i. . .. . ,... . .:

3. . . . . . .. . . ..

s .

z.

! er% kee. k' 4< < 1F . D. km ea i 'so illed ~

wi%h -A euhny .;f. \

.? . . . L. .

ik . c y&T bk ad 44td. -

. , :'I . 1,a(:ei:1+. 2 w tA.  % *4 L  :

ha a je.d

• . &do M

' l.,k A rs4e 4{(k' f

.- & '.pe. & n > 3 (<?t<.. c.% ~.+c k . p

... 1- u n s. ,: .

n n --

I k.R.. .tM elis .shce- $ y.gwh.'eemk m l.

.T km.l LT. % 'd n-W .f wp L+ . 'ts k resulh coulA Le a ( M al T.c - er.

D.L . + s.

• . . ,n -s.)

u or>hW(4.A

. p.ceJ 4 .'-~ w n.

. 'd na.lfd+ ~~ A .A gaa _G .

~

% A st.. W T

_- A .A

~

a;

.ocica , . ]* '

.r.

i -ks.)-. 6

.k uLa, h al. M.,'

-a .a .481 u.-t[J & >2}.-< -

. M F M ib. r + ,- . s .

acecykh.le,adeA..o.-]. c.Al.

l c. e L c..w onek% s f

' n af1/f fld ik. . abb h.n k& I

.Gus a of

! Ye, w.4

$. cesbih.

t -

c. u .f k a.4(Adu. s. I w.n W A 4-' k-J'kV~fs

~4, L

?4- vs.s L cwk+ s.y .~~tS

' r~ I A.

J cemH ex Gy:

- 3 h cu.ssel Lf~ye, sAki%s - tL-- - - p- - - .k a p ta- u- in

...........u . L .. . .:L-S

~

3

"~. .

l c. h .& f .s p A . e a g b. ,.x w : $

a s -

.. .drL' sgevGn.r a5 %sfv:s sed, e n y h J ;. L +- LL e

.n u L a <t':' G A' .t des 1 4 pd~ ay,-

]) . . . '

• ->C.*16 '. % ? k4)e.t- Jah l emA

'% s:( .

eP.' 1L t-s t : f 3 A. A. d.

Y~ .

' ca yJrte. n.aK - L ds . .

' S ':' . s & v r.; u

' d u k u i. 4 afg .r -

-).? %nk e2 + f ~ $'1L. w(ks kK

'. LM

%k. M m . t h ' fL p . v.s. ~ ll '

'alda L . r<, v ft> f

.- .' Y w <.i eel e- L. Mi.

.. r.<e yk24.: d(b". v,Li s h .) -tL+- <<A-I6 e f - 64 A eW . ami . 1C. k c c. y A & k . ~ .s . w .4

1.

..- L. M.bikeA,- .  :.CL.; A ' wk ,. Z' d., Dr.

se~

,~ $. M.

kk %. (Y ate : ykt %. a r' n .u .t

f. on eag '

~

a eaq+M L kt,. >s~.A ;$

.-A p me.A i L * '

-/L ~.

- L

..) .%fare m .4L.%

% 4 I+ L m a i h s G/+ a v.A.s ar. 7L.

redifL - .

N:y',6W

- 2. <

. -%. efAu.N.f.M^<l-* L %s."' 7-

. d;=='- ?  %; 4L. eraA?n i F n .s e(w h + L+ fl- ss vr. s -

l wi<} 3=-el M e<.s (, . I'

_ ,__ __ _, ...-_----- -- - - -' ' ~ ~ ~ ' - ' ' ~ ' ~ ~ ~ ~ ' ' ' ~

• b.d '. 1 :. M s M 5...p citw'A Z' .. A a...

-q. .- _- .

[' .

kd" ht- T sk.h h %a .

s i .

• ci s y - H A ! ~ . g n d i 6,e _(3 _

b y .. + h . ma>l b N a.

. % . s e r , 3. . k

. % & '. g m . g . ;9-

a ae % ;+1es - n . + -

ed.e, -( naL  ;.;y

@4 L .* y. ~Gc. .< 4m.s . k w-h p et.svh A

7 ..

R,. g .,, n .

. h.: 9A L AtA.' LAQ.<n g  %  ?+L.,. u ,. .

. .s<.

i A A s ,.~ d u .1s & LufL.y W. -

.t f.h .T .JQ . a+.-

& f6 . p'.w>4cE ' .sa,. .

_ -t L .. e . .

E k+ WA  % .4a.

'cP- W .ya <s.

sk% f c..,.,.ys.-

a4 .Cl.;; .).f.,% ,4

- ~

kst- f , 7 . kef.ty .

• ~P ek 6W -a.4 <f'... t

,.g g 8 . - -

. . g . .

0 b f O A LG q gy _ ,,s g ..

. f fp:( . p ra. I .

\

j

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

ro McGovern.

at the site.

Earl Nemila and Mark 10, Coleman, 1980, on AprilI assis Office of Inspector and Auditor.Also participating in the interview was D As part of the interview operators a series of questions. The substance of theseestions qu. I askedand the each of the operator's in the two attached answers derived from ey memory and sparse notes, are briefly sheets.

which Mr. McGovervi and Mr. Hamila did.Each operator was asked to sign a wdt tion and editing of a statement, but decided not ti sign it.Str. Celesen parti to a question appears in the statement it is not included. Where the answer

. GmMd!&t Donale G. Kirkpatrick e

I 8 0

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

y.r,e- - y.-. 7 77 e e - ..-

. . . . . . . . . . ._ .-..-...s....-

\

ENCLOSURE 2 SUPMARY OF LEAK RATE TESTS AT THI-2 INVOLVING M. S. COLD %N AND D. I. OLSON DURING THE PERIOD -

09/30/78 THROUGH 03/28/79 -

4 i

1

- ww . ,~, _ 7.,my. gmx :.3;p_.~ :?

r-- ~~ - ' - ~ '

T~ 7 l

- . - - . , - . . . . . . . . . . . . . . ......x..

LEAK RATE TESTS AT TMI-2 INVOLVING M.S. COLEMAN (09/30/78 throvoh 03/28/79)

Test No. Date Start

• Ston* l Test CR0 Coments**

1 09/30 1734 1834 Coleman Wright No apparent problems.

1914 2014

' 3 10/01 1929 2029 Olson Coleman No apparient. problems.

0004f 0104f #10/02/78.

7 10/04 2034 2134 Olson Coleman No apparent problems.

0930f 1030f #10/05/78 33 11/02 1829 1929- Coleman Olson No apparent problems.

1839 1939 35 11/03 1 1758 Wright Coleman 1,658 No apparent problems.

708 1808 46 7 12/04 0628 0728 Coleman 111jes/

1038 0705 water addition 100 gil.

1138 Coleman Not logged in CR0 loc or inclu in calculation.

50 12/08 1617 1717 NcGovern Coleman 1622 water addition 60 gal.

2005 2105 Not logged in CR0 log or ine'.

in calculation.

54 12/10 1820 1920 Coleman Wright No apparent problems.

2220 2320 57 12/11 1659 1759 Coleman j

Coleman No apparent problems.

2059 2159 60 12/13 1950 2050 Coleman Wright No apparent problems.'

• it 2350 0050f #12/14/78 70 12/21 0259 .0359 Coleman Wright Unstable plant conditions for 0259 0359 LRT.

79 7 12/27 0810 0910 Coleman Phillippe LT switch for LRT. Possible 0825 0925 water additions: 30 gal at 082(

and 45 gal at 0845. Pzr Spray l back in auto'at 0835.

88 7 01/06 1029 1129 Coleman Olson LT switch for LRT. Possible 1044 1144 witer addition of 30 gal. at 1105-1125.

93 01/11 2153 2253 Coleman Olson 2230 water addition 240 gal.

2208 2308 CR0 log shows 200 gal added  ;

and included in calculation.

122 02/16 1203 1303 Olson Coleman 1245 water addition 60 gal.

1333 1433 1250 water addition 180 gal.

Not logged in CR0 log or includ in calculation.

Key on Page 2 l

. . - . . _ . . . . . . . . .-. p m ,7,

,7 7 ,. .-.-p.-.. .- .----- - - - - - - "

__s,._ a -,.r . n_ . : -

.n.

+

2_ r. - ; ,2.. ..;, :w LEAK RATE TESTS AT TMI-2 INVOLVING M.S. COLEMAN (CO (09/30/78through 03/28/78)

Test No. Date Start

• 5 top

• Test CR0 Connents**

129 02/25 2002 2102 Coleman Wight 2058 water addition 200 gal.

2117 2217 CR0 log shows 150 gal added and included in calculation.

i Water added 4 mins before end

> of test.

131 02/26 1839 1939 Coleman Olson 1922 water addition 180 gal.

2009 2109

- CR0 log shows 150 gal added.

• ' 154 gal included in the calcui tion. Water added 17 mins.

before end of test.

132 7 02/27 2150 2250 coleman/ Coleman 2255 2355 2240 water addition 150 gal.

Olson CR0 log shows 162 added and

' ~

included in calculation. A -

60 gal water addition may have

- don't earlier in the test. The 150 gal. addition was made 10 min. before the end of test.

138 03/04 0142 0242 Coleman

> Coleman 0240 water addition 270 gal.

1 0212 0312 CR0 log shows 238 added and included in calculation. Water added 2 mins. before end of te:

146 03/13 1105 1205 Olson Coleman 1200 water addition 330 gal.

1050 1150 CR0 log shows 305 gal added

• and included in calculation. ,

Water added 5 min. before end g of test.

d j .

• Start /stop - Top entry is clock time. The second entry is MIT strip chart time that corresponds with the leak rate test clock time.

• • Time shown in coments section is clock time.

7 Analysis findings questionable.

'1

]

i

-~r. c. 7--

---=s p u- *- * . . nc - pr n.~ pr ~ * - --

. ~ : --m -- ~

~ . -

-u. _ -. .w..~ - ~ .u.

{ .

i

SumARY OF LEAK RATE TESTS AT TMI-2 INVOLVING M.S. COLEMAN

, (09/30/78 throuch 03/28/79)

, 3. Leak rate tests with no aoparant hydrocen or water additions j Nun 6er Test No. Date 1 1 09/30/78 L.i 2 3 10/01/78 a 3 7 10/04/78 1 4 33 11/02/78 1 5 35 11/03/78

.b t 6 54 12/10/78

. .] . ,

. 7 57 12/11/78 6

8 60 12/13/78

2. Unstable olant conditions. Not able to evaluate Nun 6er Test No. Date 1 70 12/21/78

3. Hydrogen Additions f.,

NO"'

k 4. Unrecorded water additions /Not included in calculation j Number Test No. Date Amount Added

. 17 46 12/04/78 100 gal.

i 2 50 12/08/78 60 gal.

37 79 12/27/78 30 gal. & 45 gal.

47 88 01/06/79 30 gal. .

5. Recorded water additions (Amount on MUT strio chart different from logoed amount)

.}f Number Test No. "Datd Looned Amt. Calculation Amt. Stio Chart Time added befo 1

... end of test 1; -

1 g3 01/11/79 200 gal. 200 gal. 240 gal. 23 mins.

2 129 02/25/79 150 gal. 150 gal. 200 gal. 4 mins.

3 131 02/26/79 150 gal. 154 gal. 180 gal. 17 mins.

4 132 02/27/79 162 gal. 162 pa1. 60 gal. -- T '

150 gal. 10 mins.

-l.1 5 138 03/04/79 238 gal. 238 gal. 270 gal. 2 mins.

6 146 03/13/79 305 gal. 305 gal. 330 gal. 5 mins.

6. Overall Suninarv l.

LTRs on file involving M. Coleman: 20 '

LRTs involving water additions: 11 (55%)(100% for all tests after 12/21/78-9 tests)

Hydrogen additions 0 et Water additions Unrecorded "

5 Underrecorded 6* (4 tests involve water additions during the last 10 mins.

of the one hour test)

NOTE: 7 indicates analysis findings are questionable

~ ~ mr- .n-- y ,v.,;,p. :. y_ p:qygg--(, .. - - '

? ~ 3-

, a,_ em"-  % eL ane &.] " ^

• %'= .L*SDY ^'

• ' ~'~"*A****** *~^**

• A LEAK RATE TESTS AT TMI-2 INVOLVING 0. OLSON (09/30/78 - 03/28/79)

Test No. Date Sta rt* Scop

• Test CR0 Coments" 7 10/04 2034 2134 Olson Coleman No appamnt problems.

0930f 10308

1. 10/05/78 14 (?) 10/18 1221 1321 Olson Coooer

.y 1216 1316 1235 ' water addition 100 ga.

. 1255 water addition 60 ga

.! 1305 wrter addition 60 ga j

s Not lo9ged **. CR0 log or included in calculation 1

30 10/30 1131 1231 Olson

(, Frederick No apparent problems. l 1131 1231 '

33 11/02 1829 1929 Coleman Olson No apparent problems.

1829 1929

. _ 38 11/06 1931 2031 Olson Wright No apparent problems.

1945 2045

.s i

52 12/09 1149 1249 Hartman Olson

.j 1549 1649 No apparent problems.

e 66 12/17 0220 0320 Olson Olson No apparent problems.

. 0220 0320 88 (?) 01/06 1029 1129 Coleman Olson 1044 1144 1105-1125 water addition of 30 gal.

Not logged in CR0 log or included in calculation

93 01/11 2153 2253 Coleman

.7 Olson 2230 water addition 240 gal

' ^ 2208 2308 CR0 log shows 200 gal added . '

• - 200 gal included in calcu;-

i 1ation.

119 (?) 02/14 2046 2146 Kidwell Olson/ 2130 Hydrogen added to MUT.

2220 2320 Kidwell causing a 2" to 3" (60 gal j to90 gal)effect. No log entry. Kidwell relievec Olson at 2130.

122 02/16 1203 1303 01 son

! Coleman 1245 water addition 60 Sc 1333 1433 1250 water addition 180 gal Not logged in CR0 log or included in calculation.

131 02/26 1839 1939 Coleman Olson 1922 water addition 180 gal.

2009 2109 CR0logshows150galadded.

154'ga included in calcu-lation. Water added 17 mins.

before end of test.

Start /Stop - Top entry is clock time. The second entry is MUT strip chart time that corresponds with the leak rate test clock time.

Time shown in coments section are clock times. (?) Analysis findings cuestionable.

--- - .-- . . . -. - . - my-,ggr.- y9 4 . r :-

. *----.m- m- ' - - '

.m.. . m _- u......_---.._..-..- .a...-

~) .

LEAK RATE TEST AT 1MI-2 INVOLVING 0. OLSON (09/30 Test No. Cate Start

• Stop* Test CR0 Consnents**

132(7) 02/27 2150 2250 Coleman/ Coleman 2255 2355 Olson 2240 water addition 150 g G CR0 log shows 162 gal add

?. and. entered in calculatio Water added 10 min. befor

~i 133 EY.$aITNF Dn'tI_Id 1 02/28 1909 2009 Olson Wright

' 2159 2259 2005 water addition 150 9 CR0 log shows 100 gal add 1

and entered in calculatio.

Water added 4 mins before end of test. ,

137 03/03 0238 0338 Olson 01 son 0335 water addition 200 ga 0238 0338 CR0 log shows 150 gal adds and 152 gal included in ti

,- calculation. Water added 3 mins. before end of test 139 03/05 0320 M20 Olson Wright M 00 water addition 60 gal 0350 0450 0410 water addition 210 97 i

CR0 log shows 128 gal a m and included in calculatic Water added 10 mins before and of test.

140 03/06 0321 0421 Wright Olson 0420 water addition 210 ga 0351 0451 CR0 log shows 180 gal am

- and included in calculat;;

Water added 1 mirr before end of test.

7 -

141 03/08 0306 0406 Olson Wright' 0403 water addition 270 ga'

.; 0306 0406 CR0 log shows 228 gal adde and included in calculatfor Water added 3 mins. before end of test.

142 03/09 0323 0423 Wright

.;I Olson 0355 water addition 60 gal.

0323 0423 0415 water addition 180 gal j i CR0 log shows 180 gal acY;c and included in calculation Water addition 8 mins. befo end of test.

w <s.

146 03/13 1105 1205 01 son J Coleman 1200 water addition 330 gal

' 2050 .1150 CR0 log shows 305 gal added and. included in calculation Water added 5 mins before end of test.

Start /Stop - Top entry is clock time. The second entry is MUT strip chart time that corresponds with the leak rate test clock time.

• • Time shown in comments section are clock times.

(?) Analysis findings questionable.

. --,,-- - .- . ~ y c v,rg mm.w .4 x r r*"- A -'- "~ ~-

w - ..__-

y,

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

) .'*-

Supt 4ARY OF LEAK RATE TESTS AT TMI-2 INVOLVING D. OLSON (09/30/78to03/28/79)

Hydrogen Additions:

Number Test No.

j 1 (?) lig

-.I. Unrecorded Water Additions: .

Number Test No. AmoDnt Added

] 1(?) 14 220 gal

'i

, 2(?) 88 30 gal

,; 3 122 60 gal 180 gal -

_ Recorded Water Additions:

_. Ihm6er Test No. Legged Amount Included in Calculation Strip Chart Time added (f817) before end

-} g3 of test 1 200 gal 200 gal 240 gal 1 2 131 .

150 gal 154 gal 23 mins 180 gal 17 mir.:

l 3(?) 132 162 gal 162 gal 180 gal 10 mins 4 133 100 gal 100 gal 150 gal 4 min:

5 137 150 gal 152 gal 200 gal 3 mins 6 139 128 gal 128 gal 60 gal 19 mins 210 gal g mins

7 140 180 gal 180 gal 210 gal 1 mins 1.; 8 141 228 gal 228 gal 1

270 gal 3 mins

. g 142 180 gal 181 gal 60 gal 28 mins 180 gal 8 mins 10 146 305 gal 305 gal 330 gal 5 mins Overall Swenary LRTs on file involving D. Olson: 20

. LRTs involving water or Hg addition: 14(705)(100% for all test after 12/17/78)

H2 additions 1

Water additions

' Unrecorded 3 j Underrecorded 10*

• 8 tests involved water additions during the last 10 mins of

'sotal: 14 the one hour test.

l ,

(?) Analyses findings are questionable.

m-. y. - .:::~r, rt:r&- .py .. - s - ~- ~~~ ~ ~ v

_ " - ' '- * * " ' '~

. ,, Enclosure 15 UNITED STATES 8  % NUCLEAR REGULATORY COMMISSION *

, a a susenacton. o. c.rosss September 12, 1985 k*...* ,

MEMORANDUM FOR: William T. Russell. Acting Director Division of Human Factors Safety, NRR FROM: Robert A. Capra, Technical Assistant Division of Human Factors Safety, NRR

SUBJECT:

SUM 4ARY OF JOINT 01/NRR INTERVIEW WITH LEONARD P. GERMER The purpose of this memorandum is to document the results of the joint 01/NRR interview with Mr. Leonard P. Germer. Mr. Germer was interviewed in order to

'obtain additional-background information on the subject of reactor coolant system (RCS) leak rate surveillance testing at THI-2 during the period -

September 30, 1978 to March 28, 1979. During this period, Mr. Gemer was

' assigned to two different shifts at TMI-2. Three of the individuals with whom Mr. Gerwer stood watch are currently licensed operators and are the subject of a joint 01/NRR investi H. A. McGovern and C. L. Guthrie)gation and evaluation (Messrs. R. R. Booher.

The interview with Mr. Gemer was held in the law offices of Killian &

Gephart in Harrisburg, Pennsylvania. The interview was conducted under oath in the presence of Mr. Gemer's personal attorneys: Mr. Smith B. Gephart of the law fim Killian 8 Gephart and Mr. James W. Moeller of the law firm LeBoeuf Lamb, Leiby & MacRae, Washington, D.C. The interview was conducted by Mr. R. Keith Christopher, Director Office of Investigations, Region I and me. A copy of the transcript of the interview is attached as Enclosure 1. -

The page references cited in this sumary refer to the pages of Enclosure 1 from which the infomation was extracted. Enclosure 2 contains a copy of the

. NRR Evaluation of TMI-2 Reactor Coolant System Leak Rate Tests Performed Between September 30, 1978 and March 28,1979(UpdatedJuly 30,1985).

After being placed under oath, Mr. Germer was asked to describe his employment history. Prior to joining Metropolitan-Edison Company (Met-Ed) in 1977. Mr. Germer spent six years in the U. S. Navy and approximately eight years employed by three different companies unrelated to the nuclear industry. Mr. Gemer was originally hired by Met-Ed as an Auxiliary Operator (AO) at TNI-2. He served in that capacity until the Fall of 1978 at which '

timeheenteredtheControlRoomOperator(CRO)trainingprogram. Mr.

Gemer's duties as a CRO-in-training included standing watch in the control room on a regular basis with an assigned shift. Mr. Gemer remained in that position until the accident on March 28, 1979. He was then transferred to the TMI-2 accident recovery team. In August 1979. Mr. Gemer left Met-Ed NOTE: This memorandum and enclosures discuss information which is the subject of an ongoing investigation. This memorandum and enclosures may not be disseminated outside the NRC without coordination with NRR and the pemission of the EDO or the Director of 01. Internal access i and distribution should be on a "need to know basis.

Q' M -

P?

. William T. Russell September 12, 1985 to accept his present position as an A0 at Maine Yankee Atomic Power Plant.

Mr. Germer did not receive his. Reactor Operator's (RO) License at TMI-2 and is not licensed at Maine Yankee. See pages 3-6.

From October 1978 to December 31, 1978 Mr. Germer as assigned to Shift "E."  ;

When a sixth shift (Shift "F") was created on January 1,1979 Mr. Gemer was assigned to that shift. He remained with Shift "F" until the accident at TMI-2. During these periods, the other members of these shifts included:

' POSITION SHIFT "E" SHIFT "F" 10/78 - 12/78 01/79 - 03/79 Shift Supervisor: Bernie Smith Ken Bryan Shift Foreman: Ken Hoyt Carl Guthrie CR0 Harold Hartman Hugh McGovern CR0 Ray Booher Earl Hemila CRO-in-trai~ning John Blessing ,

~ ~

See pages 7 & 8.

Mr. Germer stated that as a CRO-in-training, he spent a large portion of his

time on shift studying material for his R0 license; however, when plant maneuvers or unusual surveillance tests were scheduled for his shift, he would physically manipulate the controls or perform the surveillance test under the supervision and direction of a licensed CRO. In performing his duties on shift, he did not work with any one particular CRO. Rather, he would normally work with the individual who was assigned as the panel operator on that particular day. See pages 9-11.

In describing his interactions with members of supervision and management.

Mr. Germer said that he would interface quite frequently with his Shift

• i Foreman on matters associated with shift routine and plant operations. He did not normally have much interaction with his Shift Supervisor or the TM1-2 Supervisor of Operations (Mr. James Floyd). He essentially had no dealings with higher level management such as the THI-2 Plant Manager (Mr. Joe Logan) or the Station Manager (Mr. Gary Miller). See pages 11-13.

RCS leak rate surveillance tests were normally run on the plant computer.

During the six months Mr. Germer was on shift, he said that he performed numerous leak rate tests; however, he said they were all conducted under the auspices of one of the licensed CR0s. According to Germer, the CR0 " stood right over spy shoulder and looked when I punched them in, because a couple of times I hit the wrong key." During the time he was on shift he could not recall any specific meetings with his supervisors or manageme,nt to discuss 4

problems with obtaining satisfactory leak rate tests. See pages 13-15, After reviewing Surveillance Procedure 2301-3D1 " Reactor Coolant System Inventory," Mr. Germer was asked if he was given any guidance or particular instructions regarding precautions and limitations associated with running the tests. Mr. Germer recalled that tests were to be conducted only with the plant in a steady state condition (i.e., no transients or plant maneuvers taking place.) Mr. Gerner also noted that the first item under " Limits and Precautions" directed the operator to avoid the addition or removal of water from the reactor coolant and makeup systems during the test. See pages 15-17

^

l

. William T. Russell September 12. 1985 l

l Mr. Gemer was also asked to review TNI-2 Technical Specification 3/4.4.6.2 l 4

" Reactor Coolant System Operational Leakage." It was pointed out that

Technical Specification 3.4.6.2.b limited " unidentified leakage" to 1 gpm.

In order to demonstrate compliance with this limit. Surveillance Requirement i

4.4.6.2.d required the perfomance of a reactor coolant system water inventory balance (leak rate test) at least once per 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> during steady state operations. It was also pointed out that the action statement of this Technical Specification stated:

With any Reactor Coolant System leakage greater than any one of the above limits, excluding PRESSURE BOUNDARY LEAKAGE. reduce the leakage to within limits within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> or be in at least HOT STANDBY within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in COLD SHUTDOWN within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />.

See pages 17-19. .

Mr. Genner stated that they tried to run a leak rate test at least once per shift. Because the technical specifications required a satisfactory leak -

l rate test only once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, they ignored all unsatisfactory test results as long as they received a satisfactory test within the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> window. All unsatisfactory tests were thrown away and all good tests were retained and used as a basis for resetting the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> clock. Mr. Gerr;er did not recall it being too difficult to obtain test results less than 1 gpm. He said he was usually able to get a good test on his second or third try. When a bad result was obtained he would infom one of the CR0s or his Shift Foreman and perform the test again. See pages 19-23.

According to Mr. Genner, the practice of discarding leak rate test results that exceeded the limits of the technical specifications was coninon knowledge among the CR0s. Shift Foremen and Shift Supervisors. He said that he did not feel pressured to get a good leak rate test because he did not recall the .

tests being a big problem. He may have had to run two or three per shift

, but, overall he did not perceive it as a major problem. He did not recall it

, being more difficult to get good results as the date of the accident approached. See pages 23-25.

Mr. Germer was asked if he was ever told the reason leak rate tests were to be thrown away was so the NRC inspectors would not see them. He vaguely recalled being told to ensure bad test results were discarded because "an NRC guy found them laying out one time." He believed that his Shift Foreman was the individual who made this statement. He also thought that this direction was given to the entire shift. Mr. Gemer could not positive, but he believed Mr. Ken Hoyt may have been the Shift Foreman who made that statement. See pages 25-30.

A copy of LER 78-62/1T along with an attached routing sheet with the typewritten names and handwritten initials of control room personnel was provided to Mr. Germer for his review. The routing sheet was used to indicate which individuals had seen the LER. Mr. Gemer ccMirmed that the

' initials on the routing sheet were his; however, he did not have a preaccident recall of the incident. Sy pages 30-32. .

I NOTE: During a routine inspection of TMI-2 operations on October 18, 1978 an NRC inspector discovered several bad leak rate tests lying in the control room and that THI-2 had been operating for an extended period

~

. William T. Russell . September 12, 1985 of time with unidentified leakage exceeding the technical specification lir.it. The incident resulted in the submittal of LicenseeEventReport(LER)78-62/1T.

The LER states in part: "This event was caused by a misinterpretation of the requirements of the technical specifications....The- appropriate personnel will be instructed on the requirements of applicable sections of the T.S. and the requirement to imeciately invoke applicable action statements when the provisions of the LCOs [ limiting conditionforoperation]arenotmet."

While Mr. Germer could not recall the specific incident, he did not believe he was given any instruction or guidance on a revised interpretation of the requirements of the technical specifications following the October 1978 incident. Mr. Germer believed that the instructions he received about ensuring that bad leak rate tests were discarded, must have been associated with this incident.- See pages 30-37. .

Mr. Germer stated that there were several things that would cause him to consider a leak rate test invalid. These included such as actions as typing in the wrong numbers, hitting the wrong key, or forgetting to include in the calculationtheamountofwateraddedtothemakeuptank(MUT)duringatest.

When this situation occurred, he would throw the test away and start over. ,

At times, however, he would do everything correctly and still come out with results greater than 1 gpm. On those occasions, no action was taken because they were within the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> window. Mr. Germer could not provide a rational basis for why good test were accepted as valid and bad tests were considered invalid and thus discarded. While the results were variable, Mr. Germer believed at the time that the surveillance test results basically reflected true plant leakage. See pages 37-42.

Mr. Germer was asked if he was aware that operators may have been manipulating test results by adding water to the MUT tank during the test and either not including the water addition in the leak rate calculation or not including the full amount of water added in the calculation. Mr. stated that he personally never added water with the intent of influencing the outcome of a leak rate test and to the best of his recollection he was not aware of any other operator trying to manipulate test results in this manner.

He did not believe anyone ever told him how to get a " good" feak rate. Mr.

Germer stated that if he had wanted to influence the outcome of a test, he could figure out how to do it on his own. Mr. Germer said:

Nobody ever told me how to do it. I never even bothered. This is all hindsight. I never bothered to even look at how to get it to come out. I didn't really care. All I cared was getting the stupid thing done, hand it to the foreman, so I cculd go back and study, or learn something, because I had pressure on me to make sure I got marks to pass to get a license, okay? I was more concerned at the time to getting a license than I was worried about getting leak rates. It was part of the job to do the leak rate, and whatever else I had to do.

See pages 44-47.

l

, William T. Russell ,

September 12, 1985 Mr. Gemer stated that he was not aware at the time that adding hydrogen to the MUT could possibly influence MUT level indication and thus affect leak rate test results. Mr. Gemer'said that he did not believe the phenomenon

-, was true until the day before his interview, when it-was explained to him by his attorney. While the procedure prohibited chemical additions during a leak rate test, Gemer did not believe operators considered adding hydrogen to the MUT to be a chemical addition. He did not recall there being any prohibition against it, because technically it was not supposed to make any i difference. Mr. Germer could not recall observing any fluctuations, changes or erratic movement in MUT level indication when adding hydrogen to the MUT. <

)

Mr. Germer also stated that he had no personal knowledge of or reason to believe that operators were using hydrogen additions to assist them in getting good leak rate test results. See pages 50-56.

The next portion of the interview concentrated on MUT level transmitter problems. It was explained to Mr. Germer that during the months of December 1978 and January 1979, one of the two level transmitters (LT-1) providing MUT

~ level indication was providing a very erratic and unreliable output. Yet; during this period many operators, including Mr. Gemer, used LT-1 to provide MUT level indication to the computer during leak rate tests. Since there was a level transmitter selector switch in the control room,. operators could have easily used LT-2 (the stable transmitter) to provide reliable MUT level indication to the computer during the test. After reviewing some copies of MUT strip chart recordings which exhibited these characteristics, Mr.

Gemer said he could not specifically remember level transmitters being a problem at the time. He also did not remember if he was given any particular guidance or direction associated with switching level transmitters for leak rate tests. See pages 59-64.

It was explained to Mr. Germer that one of the allegations made by Mr.

Hartman was that operators would start a leak rate test using the MUT level.

l

' transmitter with the lower reading and at the end of the test, switch level transmitters such that the higher reading level transmitter would be read by

- the computer. Mr. Gemer stated that he did not recall anybody manipulating leak rate tests in that manner. See pages 64 & 65.

The last part of the interview centered on a review of the leak rate tests involving Mr. Germer. During the period under investigation, Mr. Gemer was involved in 14 leak rate tests, either as the individual performing the test

' and signing the surveillance test sheet or as the operator on the panel who signed the CRO's Log.

The following table shows a breakdown of Mr. Gemer's 14 tests by evaluation category:

Evaluation Category Number of Tests Tests with no apparent problems..................................... 5 Tests involving the use of an unstable MUT level transmitter t i to provide input to the plant computer............................ 5 i

l Hydrogen additions to the MUT during test........................... 2 Water addition during test that was not included in calculation..... 1 Test conducted during unstable plant conditions..................... 1 NOTE: While Mr. Gemer may have been involved in many more leak rate tests than the 14 indicated above, only 14 were retained and the remainder were discarded.

l

. William T. Russell September 12. 1985 After reviewing the tests Mr. Germer again stated that he did not personally attempt to manipulate the outcome of these tests. Mr. Gemer was asked if he had any knowledge or informatioh as a result of conversations observations, or anything else in his possession or within his memory that would indicate to him that other operators on his shift, specifically Messrs. Booher.

McGovern or Hemila, were manipulating leak rates during the time he was at TMI-2. Mr. Germer stated that as far as he knew, they were performing the tests properly. He did not recall anybody ever discussing with him ways to influence the outcome of a leak rate test. See pages 66-73.

Mr. Gemer stated his personal opinion that leak rate tests may not have been manipulated and that Mr. Hartman "was blowing off steam, because he was upset with somebody at the company." Mr. Gerwer was informed that four other individuals beside Mr. Hartman have admitted manipulating leak rate test results. Mr. Germer stated he was not aware of that, but it still did not alter his testimony. According to Germer:

I was never aware of people manipulating, and it is possibly due to the fact that I was just an RO trainee at the time, and really didn't get into that, and spent a lot more time studying and didn't-spend a lot of time on the board talking to the guys how they were doing it, because I considered the more important thing at the time to do was study.

See pages 74-77.

hh {-

Robert A. Capra, Technical Assistant Division of Human Factors Safety. NRR-

Enclosures:

. 1. Transcript of Interview

2. NRR Leak Rate Evaluation cc:

B. Hayes K. Christopher -

J. Lieberman J. Goldberg I

i

.

_):, .

- . , . . . - .._,---,.m . - , - . . , - - - - . , _____.,..,_-y, , , , _ _ _ _ . . _ _ _ , _ _ _

,__rm, - _ _ _ , , _ _ . , . _ , - . - - -, _ - _ _ _ - - _ _ . _ _ .#5,...,__._. ,

[ UNITED STA1 8 NUCLEAR REGULATOR nclosure 16 y

f wasHmomm. o. c.

( .....

/ September 26, 1985 MEMORANDUM FOR: William T. Russell Acting Director Division of Human Factors Safety, NRR ,

FROM: Robert A. Capra Technical Assistant  !

Division of Human Factors Safety, NRR

SUMMARY

OF JOINT OI/NRR INTERVIEW WITH EARL D. HEMMILA The purpose of this memorandum is to document the results of the joint 01/NRR interview with Mr. Earl D. Hemila. Mr. Hemila was interviewed in order to obtain additional background infomation on the subject of reactor coolant system (RCS) leak rate surveillance testing at TMI-2 during the period September 30, 1978 to March 28, 1979. During a portion of this period, Mr. Hemila stood watch with two of the individuals (Messrs. H. A. McGovern' and C. L. Guthrie) who are currently subjects of the joint OI/NRR investi-gation into leak rate testing irregularities at TMI-2 prior to the accident.

The interview with Mr. Hamila was held on March 28, 1985 in the law offices of Killian & Gephart in Harrisburg, Pennsylvania. The interview was conducted under oath in the presence of Mr. Hemila's personal attorneys:

Mr. Harry H. Voigt and Mr. James W. Moeller of the law fim LeBoeuf Lamb, Leiby & MacRae, Washington, D.C. The interview was conducted by Mr. R. Keith Christopher, Director, Office of Investigations, Region I and me. A copy of the transcript of the interview is attached as Enclosure 1. The page references cited in this sumary refer to the pages of Enclosure 1 from which the infomation was extracted.. Enclosure 2 contains a copy of the NRR Evaluation of TMI-2 Reactor Coolant System Leak Rate Tests Performed Between-September 30, 1978 and March 28,1979(UpdatedJuly 30,1985).

- After being placed under oath, Mr. Hemila was asked to describe his employment history. Prior to joining Metropolitan-Edison Company (Met-Ed) in the fall of 1976, Mr. Hemila spent six years in the U. S. Navy. Mr. Hemila was originally hired by Met-Ed as an Auxiliary Operator (AO) at TMI-2. He served in that capacity until entering the Control Room Operator (CRO) 1 training program for TMI-2. HereceivedhisReactorOperator's(RO) license in the Fall of 1978. In 1980, he obtained his Senior Reactor Operator's (SRO) license and was promoted to Shift Foreman. In 1981, he advanced to Shift Supervisor. Mr. Hemila left Met-Ed on November 30, 1982 and is currently employed as a Senior Start-Up Consultant with Nuclear Start-Up Services, Ann Arbor, Michigan. See pages 2-4.

I ND H.: This memorandum and Enclosure 1 discuss infomation which is the i subject of an ongoing investigation. This memorandum and Enclosure 1 l may not be disseminated outside the NRC without coordination with NRR l and the pemission of the EDO or the Director of 01. Internal access and distribution should be on a "need to know basis."

l

• - T h a /*

l

. . William T. Russell - 2- Siptember 26, 1985 For a short time after receiving his license, Mr. Hemila was assigned to "B" shift; however, when a sixth shift (Shift "F") was created on January 1, 1979, Mr. Hemila was assigned to that shift. He remained with Shift "F" until the accident. While assigned to Shift "F" the members of his shift included:

POSITION NAME .

Shift Supervisor: Ken Bryan Shift Foreman: Carl Guthrie CR0 Hugh McGovern CR0 Earl Hemila CRO-in-training Leonard Gemer See page 5. i In describing his shift organization, Mr. Hemila stated one CR0 would normally assume the panel, one CR0 would handle " switching and tagging" duties and the third CR0 would take the' log readings. The CR0s worked for the Shift Foreman, whose desk was in the TMI-2 control room. The Shift Foreman reported to the Shift Supervisor. Since the Shift Supervisor was responsible for both THI-1 and THI-2, he would nomally split his time between units while on shift. Mr. Hemila believed that the Shift Foreman and Shift Supervisor were cognizant of the actions of the CR0s during the shift. See pages 5-7.

Mr. Hemila stated that he did not normally have much interaction with the TMI-2 Supervisor of Operations (James Floyd). He believed that Mr. Floyd was knowledgeable about evolutions that were taking place in the plant; however, he did not deal with Mr. Floyd on routine operational matters. Hemila essentially had no dealings with other members of management such as the TMI-2 Superintendent of Technical Support (Jim Seelinger and later George -

Kunder),theTMI-2PlantManager(JoeLogan),ortheStationManager(Gary Miller). He did not recall any meetings with members of management to discuss problems associated with leak rate surveillance testing. See pages 7-10.

After reviewing copies of THI-2 Technical Specification 3/4.4.6.2, " Reactor Coolant System Operational Leakage," and TMI-2 Surveillance Procedure 2301-3D1, "RCS Inventory," Mr. Hemila stated that the technical specifications required a satisfactory leak rate test be performed once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Nevertheless, depending on the workload they tried to run at least one test every shift. Leak rate test results that exceeded the limits of the technical specifications were discarded. Hemila said operators believed it was okay to discard bad test results since the technical specifications required a satisfactory test only once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. See pages 11-14.

While Mr. Hemila believed that the leak rate test was a "somewhat valid" indicator of plant leakage, he did not believe it was possible for the test to accurately determine unidentified leakage within 1 gpm. Thus, he did not think there was anything wrong with throwing away results that exceeded the limit. He said he personally felt the plant was safe regardless of the results of the test. He did not recall ever entering the action statement of

,-, - - , - - - - - - -- - - - , - - - - - . _ , -.w-- ,c -

---

.-------e

. --,,e- , ~ - - - .- - - - , - - - - - - - - - , --w -

.- - William T. Russall 3- September 26, 1935 the technical specifications as the result of an unsatisfactory leak rate test. According to Hemila, the leak rate test was only used to satisfy the administrative requirements of the technical specifications, they used other sources to determine whether the plant was really leaking. M pages 15-18.

Mr. Hemila said leak rate tests were simple to run and were perfomed in a very routine manner through the use of the plant computer. The operator performing the test would try and establish stable plant conditioris prior to initiating the test. He would also normally infom the operator on the panel that the test was in progress. While test results sometimes exceeded the limits, he could not recall it being tco difficult to obtain satisfactory test i

results. See pages 19 and 20.

Despite the requirement to log the start and stop time of all surveillance I

tests required by the technical specifications, only satisfactory test results were logged. While Hemila could not recall much guidance being i

given in this area, he said that he was never directed by anyone not to log start times as an attempt to hide the fact they were having to run so many

_. tests. See pages 21 and 22.

On their shift, Mr. Hamila said the Shift Foreman (Carl Guthrie) usually saw or was made aware of all test results whether they were satisfactory or unsatisfactory. While unsatisfactory tests were discarded Hem 11a said 4

there was never any indication that tests were being discarded in order to I

hide them from the NRC. See,pages 23 and 24.

A copy of LER 78-62/1T along with an attached routing sheet was provided to Mr. Hemila for his review.

! NOTE: During a routine inspection of TMI-2 operations on October 18, 1978 an NRC inspector discovered several bad leak rate tests lying in the control room and that TMI-2 had been operating for an extended period -

of time with unidentified leakage exceeding the technical specification limit. The incident resulted in the submittal of Licensee Event Report (LER) 78-62/1T on November 1,1978.

The LER states in part: "This event was caused by a misinterpretation of the requirements of the technical specifications....The appropriate personnel will be instructed on the requirements of applicable sections of the T.S. and the requirement to immediately invoke applicable action statements when the provisions of the LCOs [ limiting conditions for operation]arenotmet."

Despite his initials on the routing sheet, Mr. Hemila had no preaccident recall of the incident or the LER. He did not recall being directed at any time to discontinue throwing away bad leak rate tests, and he did not recall being instructed on the proper interpretation of the technical specifications. As best as he could recall, leak rate testing practices were the same at TMI-2 from the time he received his license until the date of the accident. See pages 24-30.

When unsatisfactory test results were received, the only follow-up action Mr. Hemila would take was to infom the Shift Foreman of the results. It I

_ ._ - -- -D

• William T. Russell - 4- September 26, 1985 was pointed out to Mr. Hemila that if a test was considered invalid, the procedure required the operator to note that in the remarks section of the test and to describe the action that invalidated the measurement.

Mr. Hemila said that action was not taken. He said the procedure was probably not followed in this case because the test was considered such a routine, everyday task that no one put much emphasis on it. jee,pages30-32.

Next Mr. Hemila was questioned about his knowledge of leak rate manipulation by operators through the use of unrecorded or under-recorded water additions.

Mr. Hemila stated that he did not know of anyone who intentionally added watertothemake-uptank(MUT)inanefforttoinfluencetheoutcomeofa leak rate test. Except for Hartman's allegations, he said that he had no

reason to believe that such actions were occurring at the plant. Hemila t

stated that as far as he knew, if water was added during a leak rate test and not properly included in the test calculation, it was due to operator error and not done to intentionally influence the outcome of the test See pages 34-39.

When adding water, Mr. Hemila stated that he would dial in the amount of water to be added on the batch controller. The amount of water logged in the CRO's Log and included in the leak rate test calculation would be taken off l the batch controller. Mr. Hemila said that he would only use the MUT strip i chart value to confirm that the amount being added was approximately what it should be. Mr. Hamila stated that he was not aware of any significant difference between the amount of water indicated on the totalizer compared

] with the amount indicated on the MUT strip chart. See pages 39-44.

Mr. Hemila was Informed that for a considerable period of time during the months of December 1978 and January 1979, one of the two MUT level trans-mitters (LT-1) provided a very erratic and unrelii.ble output. While Mr. Hemila could not recall any significant problem with the level trans-mitters, he stated it would most likely have been placed out-of-service and*

marked with an out-of-service sticker. He said whether it was actually declared out-of-service or not, prudent operating practice would dictate that you would use the stable level transmitter to obtain MUT level data for leak rate tests and not use the unstable transmitter. See pages 44-50.

Next Mr. Hemila was questioned about Hartman's allegation that operators i

used hydrogen additions to the MUT in order to influence the outcome of leak rate tests. Hemila stated that prior to Hartman's allegations being made public, he had no knowledge that hydrogen could affect MUT level and consequently leak rate test results. He said that operators were all sensitive to adding hydrogen because MUT pressure was one of the instruments theTMI-2PlantManager(JoeLogan)alwayslookedatwhenhewasinthe control room. If it was low, Logan would coment about it to the operators.

, As a result, Hemila stated that he would typically call up one of the A0s at the beginning of the shift and ask him to add hydrogen sometime during the shift. See pages 50-51.

Mr. Hemila did not believe there was any prohibition against adding hydrogen to the MUT during the course of a leak rate test, even though the procedure stated chemical additions should not be made during the test. Mr. Hemila 1

f reiterated that he had no preaccident knowledge that hydrogen could influence 1

w 4

meme-----vn - - , . , - , . - - - - xm-ann.e---- ~ m----~-----,-w--m-m--- v-e--m-em,-----w ,,-,-w .w.ww-,,wwem m a_, e-me-_ o----eemm.-me,,- ,

.

• William T. Russell 5- September 26, 1985 MUT level indication. He did not believe that others on his shift including Hugh McGovern or Carl Guthrie had that knowledge either. Mr. Hemila was asked how operators all around him could be manipulating tests by the addition of hydrogen to the MUT and he not know about it. Hemila disagreed that it was going on all around him but conceded that it could have been taking place on a limited scale, and he just didn't know about it. After the allegations were made public Hemila stated he was not contacted by anyone including McGovern or fauthrie who admitted they were aware that this was going on at the time. See pages 52-54.

During the next portion of the interview, it was pointed out to Mr. Hemila that our review of plant rpcords indicated he was involved in.15 leak rate surveillance tests that were retained by the licensee. Of the 15 tests, all but three contained actions that were contrary to the precautions and -

limitations of the approved test procedure. It was pointed out that th:

other 12 tests involved the following actions: -

Number of Tests Action 3

Water added near the end of the test. The amount of water included in the test calculation was significantly less than the amount indicated on the MUT strip chart.

1 Water addition that was not included in the test calculation. The test also used input from the unstable MUT level transmitter (LT-1).

3 Feed and bleed operation during the test that was not accounted for in the test calculation. One test also included the use of the unstable LT-1.

2 Hydrogen addition during the test. One test also ~

included the use of unstable LT-1.

2 No water or hydrogen additions; however, unstable LT-1 was used to provide input to the computer during the test.

1 Water or hydrogen addition (unable to differentiate).

See pages 62-66.

When asked if he could provide any additional information to help reconcile the differences between his testimony and the technical evaluation, Mr. Hemila stated:

I guess I would say that it bothers me when somebody casts son,e doubt upon what I have done, because I've had a lot of contact in the past five or six years with the NRC. I think you can go back and talk to any of them, and I have a very good professional reputation with those people.

Now, I can't sit here and give you a blow-by-blow account and say this is why this happened, this is why this happened. !

• can't do that. But I can tell ycu that there was never any water or hydrogen added to alter that leak rate on purpose, where I said: Hey, I know I've got to do this to alter the leak rate. I never did that. See pages 68 and 69.

h .

,. ,..- ,.--- ,.,-,.-,-~-.-.,-c.- -- --.-,

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

. W111'am i T. Russell S;ptember 26, 1985 It was pointed out to Mr. Hemmila that as the date of the accident grew nearer, the identified leakage was approaching the 10 gpm technical specification limit. Mr. Hemmila stated that he was never concerned about whether the plant should be shut down and the leakage fixed. He said it never crossed his mind. He said the attitude he had was: " Hey listen, there's people over me, and you've got the experts, et cetera. There's people looking at this problem. I mean, I just wasn't that concerned with it. I was worried about doing my day-to-day job." See pages 71 and 72.

Mr. Hemila was informed that according to the testimony of some operators, the leakage was affecting their day-to-day job by having to constantly recirculate the pressurizer to equalize the boron concentration, make frequent water additions and frequently pump down the reactor coolant drain tank (RCDT), instead of being able to concentrate on other important aspects of their job. , Mr. Hemila responded by stating that other than adding a little more water than usual, he did not consider it a big problem. See page 72.

9,L/o. Ce Robert A. Capra, Technical Assistant Division of Human Factors Safety NRR-

Enclosures:

1. Transcript of Interview

2. NRR Leak Rate Evaluation cc: B. Hayes K. Christopher

• J. Lieberman J. Goldberg 9

-* = e- + e

1. y UNilaD STATas Enclosure 17 l NUCLEAR REGULAT41RY COMMISSION

~

[.

E suasMesseTON,@, C.30506 August 27, 1985 -

MEMORANDUM FOR: Harold R. Denton, Director .

Office of Nuclear Reacecr Regulation ,

4 FROM: William T. Russell Acting Director Division of Human Factors Safety, NRR

SUBJECT:

RESULTS OF JOINT OI/NRR INTERVIEW WITH LYNN 0. WRIGHT

Reference:

1. MemorandumfromS.J.Chilk(SECY)toB.B. Hayes (01) and W. J. Dircks (ED0) dated April 2,1984,

Subject:

Staff Requirements-Discussion of Pending Investigation-

, DI

~

2. MemorandumfromH.R.Denton(NRR)toB.B. Hayes (01) -

dated May 3, 1984.

Subject:

NRR Review of 01 A,

Investigative Materials Concerning Hartman Allegations of Falsification of Leak Rate Data at TMI, Unit 2

3. MemorandumfromW.T. Russell (DHFS)toH.R.Denton (NRR)datedDecember 14, 1984

Subject:

Results of Joint 01/NRR Investigation and Evaluation of Dennis I. Olson

4. MemorandumfromW.T. Russell (DHFS)toH.R.Denton (NRR)datedJanuary4,1985,

Subject:

Results of Joint 01/NRR Interview with Mark S. Ccleman I

5.MemorandumfromW.T. Russell (DHFS)toH.R.Denton .

(NRR)datedJanuary 24, 1985

Subject:

Follow-up Action on Additional TMI-2 Operators The purpose of this memorandum is to document the results of the March 27, 1985 joint 01/NRR interview with Mr. Lynn 0. Wright. The interview with Mr.

Wright was conducted in order to obtain additional background information on l reactor coolant system (RCS) leak rate surveillance testing practices at TMI-2 during the period Septem6er 30,1978 through the date of the accident.

March 28, 1979. During that time frame, Mr. Wright was a licensed Control Room Operator (CRO) at TMI-2 assigned to Shift "D." Mr. Wright is currently self-employed and is no longer in the nuclear industry.

NOTE: This memorandum and enclosure discuss information which is the subject of an ongoing 01 investigation. This memorandum and enclosure may not be disseminated outside the NRC without coordination with NRR and the

, permission of the EDO or the Director 01. Internal access and distribution should be on a "need to know" basis.

l r _r e, -

7OQ $pP

---

m - - - e-----_ ,--e*- w e ,- w 4--._ _,%-,.n,

_ ._---,e. . -c 3-m

Harold R. Denton .2- August 27, 1985 l

Background )

As a result of a Comission meeting on March 23, 1984 NRR was directed by Reference 1 to review 01 investigative materials concerning falsification of RCS leak rate surveillance tests at TMI-2 and refer back to DI those matters i which required further investigation. The results of NRR's review was provided in Reference 2. The review detennined that follow-up investigation by 01 and further evaluation by NRR was needed in the case of seven licensed operators. Mr. Dennis I. Olson, currently a Senior Reactor Operator (SRO) at Waterford 3, was one of the seven individuals identified as needing additional investigation and evaluation. Mr. Olson was interviewed under oath in New Orleans, Louisiana by 01/NRR on November 15, 1984. During the interview, many of the statements made by Mr. Olson did not appear credible in light of the technical evaluation of Mr. Olson's leak rate surveillance tests and statements made in 1980 by his former shiftmate Mr. Mark S. Coleman -

Mr. Coleman was interviewed again on the subject by 01/NRR on December 14, 1984 in order to gain additional evidence that would either support of refute the statements made by Mr. Olson. The results of the investigation and evaluation of Mr. Olson and the results of the interview with Mr. Coleman -

were provided as References 3 and 4. respectively.

t Because of infonnation learned during the on-going investigation, Reference 5  ;

expanded the list of fonner operators to be investigated and evaluated from  !

seven to ten. One of the individuals added was the former Shift Foreman of

Shift "D " Mr. Adam W. Miller. Mr. Miller is currently the Manager, Plant i Operations THI-2.

Because of the conflicting testimony presented by Messrs. Olson and Coleman and because of the need to investigate and evaluate Mr. Miller, 01/NRR detennined that interviewing Mr. Wright would provide additional- background information on the leak rate surveillance testing practices of Shif t "D."

l Past Involvement in TMI-2 Leak Rate Testing Irregularities -

On March 27, 1985, a joint 01/NRR interview with Mr. Wright was held in the i . law offices of Killian and Gephart, Harrisburg, Pennsylvania. The interview j was conducted under oath and in the presence of Mr. Wright's personal attorneys. The interview was conducted by Mr. Keith Christopher Director, Office of Investigations, Region I and Mr. Robert Capra, currently on detail i

to DHFS/NRR.

A complete summary of Mr. Wright's interview is contained in the enclosed memorandum from R. Capra to me dated August 27, 1985

Subject:

Sumary of Joint O!/NRR Interview with Lynn O. Wright. Capra's memorandum also contains a copy of the transcript of Mr. Wright's interview and a copy of NRR's evaluation of TMI-2 leak rate tests perfonned during the period September 30, 1978 through March 28, 1979. The following key points were brought out by Mr. Wright during his interview:

1. Mr. Wright had little faith in the RCS leak rate test calculations which were performed to show compliance with the technical specifications. He believed the leak rate test results were erratic and provided results that were meaningless and did not reflect true plant leakage. '

2. While the TMI-2 Technical Specifications required a leak rate ,

surveillance test to be perfonned at least once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, it was common practice to run leak rate tests several times per shift until an

- - _ _ _ _ _ - , - _ , . - , _ , _ . - . . - . _ _ . _ _ _ . _ ,,___.-m_.. ,_,mm,. .___m . . ~ .

Harold R. Denton August 27. 1985 acceptable result was obtained. The acceptable tests were retained to show compliance with the technical specifications and the unacceptable tests were thrown away. Tests were thrown away with the knowledge and/or consent of his Shift Foreman and Shift Supervisor.

3. According to Mr. Wright, no changes were made in leak rate testing practices following the October 18, 1978 incident that resulted in the generation of LER 78-62/1T. Mr. Wright stated that at some point operators were told to retain all leak rate tests; however, that policy was discontinued after a short period of time and he did not know if that l guidance was associated with the October 1978 incident.

t

4. Contrary to the commitment contained in the narrative of LER 78-62/1T, Mr. Wright was not instructed to enter the action statement of the technical specification when leak rate test results exceeded the limiting condition for , operation.

j

. _, 5. Mr. Wright stated that prior to the accident, he was not aware that j hydrogen additions to the makeup tank (MUT) during the last few minutes

' of the test could influence MUT level indication and and consequently leak rate test results.

6. Mr. Wright stated that on occasion he did add hydrogen to the MUT near the end of a leak rate test; however, the purpose of these additions was to establish the same overpressure in the MUT at the end of the test as

was present at the start of the test. Mr. Wright said he believed this t

would minimize instrumentation errors. He stated that he did not add hydrogen in order to improperly manipulate leak rate test results.

7. Mr. Wright stated that prior to the accident, he was not aware that adding water to the RCS during the last few minutes oTa leak rate test-could produce the same effect on leak rate test results as adding l hydrogen to the MUT.

8. Mr. Wright stated that on occasion he did add water to the RCS during the last few minutes of a leak rate test; however, the purpose of these additions was to establish the same level in the MUT at the end of the test as was present at the start of the test. Mr. Wright said he believed this would minimize instrumentation errors. He stated that he did not make these water additions in order to improperly manipulate leak rate test results. Note: While Mr. Wright provided a reason for the water additions, he could not provide a reason why the amount of water included in the calculation was less than the amount indicated on the MUT strip chart recorder.

9. Mr. Wright stated operators felt pressure to get good leak rate test results in order to keep the plant operating and that the pressure increased the closer they came to the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> limit. Despite the pressure Mr. Wright stated that he was never directed by his supervisors to intentionally manipulate leak rate test results.

10. Mr. Wright stated that he had no knowledge of any operators, including either Olson or Coleman, falsifying leak rate tests by any means including: unrecorded or underrecorded water additions to the RCS, hydrogen additions to the MUT or taking ~ advantage of MUT level transmitter instrumentation errors. Except for the unreliability of the test. Mr. Wright believed that all leak rate tests were perfonned properly.

.- _- -. - _- . - - _ = - .

Harold R. Denton August 27, 1985 i

Conclusions All three CRO's on Shift "D" (01 son, Coleman and Wright) have been 3 interviewed by 01/NRR. While their testimony is consistent regarding certain i aspects of leak rate test practices on their shift. they are significantly different with regarding to their personal knowledge of or participation in leak rate test manipulation.

, During Mr. 01 son's testimony, he stated that following the October 18, 1978 incident, he was advised of the proper interpretation of the technical

, specifications and that he never threw away unacceptable leak rate tests

, following that incident. He stated he would staple the bad leak rate test i sheets to the good test and turn the package into his Shift Foreman. Mr.

i 01 son's testimony regarding technical specification interpretation and the discarding of test' results greater than 1 gpm conflicts with the testimony provided by Messrs. Coleman and Wright, both of whom agree that'no revised I interpretation of the technical specifications was given to the operators on l - -

their shift and that leak rate surveillance tests exceeding the limits of the technical specifications continued to be discarded up through the date of the accident.

. The testimony of Messrs. Olson and Wright regarding hydrogen additions conflict with that of Mr. Coleman. Both Olson and Wright stated that'they were not aware prior to the accident that hydrogen additions to the MUT could ,

affect MUT level indication and consequently leak rate test results. Mr.

Coleman stated that it was connon knowledge among CRO's that hydrogen i

additions could alter MUT level and that he had mentioned the phenomenon to Mr. Olson on at least one occasion. Mr. Coleman testified that he personally added hydrogen for the purpose of altering leak rate test results and that while he could not recall ever seeing Mr. Olson add hydrogen for that i purpose, he found it hard to believe that any of the operators did not know ,

, that this was occurring. Mr. Wright testified that he had intentionally added hydrogen during leak rate tests; however, he stated his purpose was to l try and establish the same hydrogen overpressure in the MUT at the beginning and end of the test in order to minimize inaccuracies caused by instrumentation errors. The NRR evaluation of leak rate tests which were conducted during the period under investigation identified only one test, perfomed by Shift "D" personnel, during which hydrogen was added. Based upon the admissions of Coleman and Wright that they did add hydrogen during leak rate tests, several additional tests should have been identified; however, because of the unpredictability of MUT level response to a hydrogen addition and the fact that the majority of leak rate tests were not retained, '

the evaluation can not provide additional evidence to resolve these conflicts.

The testimony of Messrs. Olson and Wright regarding underrecorded water additions to the RCS during the last few minutes of a leak rate tests were also in conflict with that of Mr. Coleman. Mr. Coleman stated that he was aware that adding water to the RCS during the last few minutes of a leak rate test could produce the same effect on leak rate test results as adding hydrogen to the MUT. Mr. Coleman stated that he used this method to manipulate leak rate test results during the last two months before the accident since hydrogen could no longer be added to the MUT remotely from the control room. Both Olson and Wright stated that they were not aware of this phenomenon.

e

---

.m--._ --.m.m,-..--_--_,- - _ . _m.w,_ ,.m-,.w-m-mw---

Harold R. Denton August 27, 1985 The NRR evaluation of the TMI-2 leak rate tests shows that during 12 of the last 13 tests performed by Shift "D" personnel, water additions were made very close to the end of the one hour test period. In each case the amount of water logged in the CRO's Log and manually input into the computer for the calculation was less than the amount indicated on the MUT strip chart recorder. Assuming the response of MUT level transmitter feeding the strip chart recorder and the level transmitter providing input to the plant -

computer were the same, the unidentified leak rate calculated by the computer was less than it would have been had the MUT strip chart value of the addition been used. If an operator were aware of this phenomenon, as in the

case of Mr. Coleman, he could take advantage of the phenomenon to increase his chances of obtaining a satisfactory result. The surveillance procedure governing leak rate testing directed the operators to avoid water additions unless absolutely required. Thus, in order to detemine whether an operator was taking advantage of this phenomenon, it is necessary to determine why the water additior.s were made. -

As stated above, Coleman testified these water additions were made with the .

intent of altering leak rate test results. Wright stated that the water j additions were not required for operational reasons, instead he added water near the end of the test in an attempt to establish MUT level at the same ,.

value for the beginning and end of the test in order to minimize instrumentation errors. However, since Wright stated he used the MUT strip

chart to detennine the amount of water added, he could not explain why the i

" total number of gallons added" that was inserted into the computer was

! smaller than the amount indicated on the MUT strip chart. Mr. Olson i

explained that the water additions near the end of the tests were necessary for one of two reasons, either MUT level was approaching the lower end of the operating band or to reduce boron concentration in order to move control rods from their outer limit. Neither of 01 son's explanations appears credible.

First, a low level in the MUT would not have occurred prior to the end of these tests. In most cases water was added during the last five minutes of ~

the test. Second, none of the other five shifts, which were operating the 4

plant in the same manner and with the same control rod limits as Shift "D."

. required water additions during the last few minutes of their tests.

In sunnary, based upon Mr.Coleman's testimony and the NRR evaluation of tests involving Coleman, Wright and Olson that shows every test perfonned between January 11, 1979 and the date of the accident contains either. an underrecorded water addition during the last few minutes of the test (12 tests) or an unaccounted for feed and bleed operation (1 test), the weight of the evidence supports a conclusion that both Messrs. Olson and Wright were not forthright and candid in their testimony regarding their knowledge of or participation in leak rate falsification at TMI-2. My conclusions and reconnendations contained in Reference 3 regarding Mr. Dennis I. Olson are unchanged.

A William T. Russell, Acting Director Division of Human Factors Safety NRR

Enclosure:

As stated cc: B. Hayes K. Christopher J. Lieberman J. Goldberg

w DISTRIBUTION August 27, 1985 Docket File - 06131 Docket File - 06404 Docket File - 06121 TMurley, R1 l 8 Martin, RIV BSnyder MEMORANDUM FOR: William T. Russell, Acting Director DEisenhut HThompson Division of Human Factors Safety, NRR LWiens FROM: Robert A. Capra, Technical Assistant Division of Human Factors Safety, NRR

SUMMARY

OF JOINT 01/NRR INTERVIEW WITH LYNN 0. WRIGHT The purpose of this memorandum is to document the results of 'the joint OI/NRR interview with Mr. Lynn O. Wright. The purpose of the interview with Mr.

~' Wright was to obtain infomation on the subject of reactor coolant system' (RCS) leak rate surveillance testing at TMI-2 during the period September 30, 1978 to March 28, 1979.,

The interview with Mr. Wright was held in the law offices of Killian &

Gephart in Harrisburg, Pennsylvania. The interview was conducted under oath in the presence of Mr. Wright's persona'l attorneys: Mr. Smith B. Gephart of the law firm Killian & Gephart and Mr. Harry J. Voigt of the law firm LeBoeuf. Lamb, Leiby & MacRae, Washington, D.C. The interview was conducted by Mr. R. Keith Christopher, Director, Office of Investigations, Region I and me. A copy of the transcript of the interview is attached as Enclosure 1 The page references cited in this summary refer to the pages of Enclosure 1 from which the information was extracted. Enclosure 2 contains a copy of the NRR Evaluation of TMI-2 Reactor Coolant System Leak Rate Tests Perfomed -

Between September 30, 1978 and March 28, 1979 (Updated July 30,1985).

After being placed under oath, Mr. Wright was asked to describe his employment history. Mr. Wright spent three years attending Juniata College and about six months working at the Harrisburg Steel Company prior to going Metropolitan Edison Company (Met-Ed) in 1972. Mr. Wright worked at Met-Ed's Crawford Generating Station in Middletown, Pa. for two years before entering the Auxiliary Operator (AO) training program for TMI-1 in 1974. He served as an AD at THI-1 until entering the Control Room Operator (CRO) training program for TMI-2. He received his Reactor Operator's (RO) license in October 1977. He then served as a CR0 at THI-2 until May 1983 when he became an Operations Monitor in the -Quality Assurance Department. Mr. Wright left Met-Ed in January 1984 and is currently self-employed. See pages 3-5.

w js~1w

,  ?

eN N Pl%

NOTE: This memorandum and enclosures discuss information which is the subject of an ongoing 01 investigation. This memorandum and enclosures may not be disseminated outside the NRC without coordination with NRR and the permission of the EDO or the n4 .,..n n< nt in +.- 1 =r- ee n a di e

• ri hn+i n., omia ha c-

-o .................. .'.. "a?.ed

. t?..k6T...b55i5......

.... . ?. ... . ..................... .....................

EATEk anc rccu sia no,soe wacu oa4o AFFI A f. P 8' CDP m CODY .

• u.s. o po a ss a-4ei

_ _ _ _ _ _ _ _ _ _ _ _ _ - _ - - - - - - - - ---- - ---- - - -- ' - - - ~ '- - ~ ~ '

. ,Jg,].f amD%. , , y2 g ugustA 27, 1 4 ,4 .

During"D."

Shift the period under investigation, Mr. Wright was primarily assigned to Shift "0" consisted of the following licensed individuals:

SHIFT "D" Shift Supervisor: Gregory Hitz Shift Foreman: Adam Miller -

CRO: Lynn Wright CRO: Dennis Olson CRO: Mark Coleman See pages 5-7.

NOTE:

Mr. Hitz is no longer licensed and works for a private contractor.

Mr. Miller is a licensed Senior Reactor Operator (SRO) at TMI-2 and is currently on June 5,the danager, Plant Operations THI-2 and was interviewed by OI/NRR 1984. Mr. Olson is a licensed SRO serving as a Control Room Supervisor at Waterford 3 and was interviewed by OI/NRR on November 15, 1984.

Mr. Coleman is a licensed R0 at TMI-2 currently assigned to the Training Department at TMI and was interviewed by OI/NRR on December 14, 1984.

In describing the way his shift was organized Mr. Wright stated that one of the CRO's would be assigne'd as the primary plant ov erator, another as the secondary plant operator and the third was assigned as the surveillance operator. The primary plant o reactor plant operations (e.g.perator was in individual

, water additions to the RCS who andcontrolled hydrogen additions to the make-up tank) and maintained the CRO's Log. The secondary plant operator controlled evolutions associated with the steam plant and the cutbuildings. The surveillance operator was principally responsible for coordinating the accomplishment of assigned surveillance tests for the shift (e.g., RCS leak rate surveillance ' ests). t The three CR0s would normally rotate positions on their next watch. The CR0s worked directly for the Shift Foreman and received the majority of operational guidance from the Foreman.

The Shift Supervisor was a dual licensed individual responsible for the operation of TMI-1 and THI-2 and split his time between units. The Shift Foreman for THI-1 and TMI-2 reported to the Shift Supervisor. See pages 7-11. '

Mr. ' Wright's interface with the Supervisor of Operations,-Mr. James Floyd, regarding plant operations was minimal. If there were things that Mr. Floyd needed accomplished, he would normally speak with the Shift Supervisor or the Shift Foreman. 1.ikewise, Mr. Wright's interface with other members of management such as the Superintendent of Technical Support, the Plant Manager and the Station Manager was even more distant. See pages 12-18.

Mr. Wright was informed that the NRC had interviewed both Mr. Olson and Mr.

Coleman regarding leak rate testing irregularities at THI-2 and that Mr.

Coleman's testimony was significantly different from that of Mr. Olson. It was pointed out that the purpose of the interview with Mr. Wright was to try and resolve some of the differences by obtaining additional background information from Mr. Wright. See pages 20-21.

Cooies of TMT-2 Surveillance Procedure 2301-301. " Reactor Coolant System

,,,,,, y Inventory," and THI-2 Technical Specification 3/4.4.6.2, " Reactor Coolant __.

Syster Operationa1 trakage; were provided to +rMright for

. " . . " . . " . . " . . .. .. .. .. . . . . . . . . " ............. his review"

......"...."..'. ""="

,CATE) lce=u m no,soinewon.o ..

OFFICIAL RECORD COPY * " " "' * **

giamT. Russe -

August 27 Q 85, .

After reviewing the documents, Mr. Wright was asked to describe how his shift conducted leak rate surveillance tests. Mr. Wright stated that the test was usually run at least once per day and sometimes o'nce per shift, plant conditions permitting. Because of oscillations in the secondary side of the plant, however, it was difficult to establish true steady state conditions.

The surveillance operator would nomally start the test by inputting the required commands into the computer. At the end of one hour, the operator would provide the computer with information such as the amount of water added or removed from the RCS during the test and identified leakage. Then, the computer would printout the initial and final values of key plant parameters

.as well as the final value of total leakage, identified leakage and unidentified leakage. See pages 23-27.

Mr. Wright believed that the ongoing feedwater oscillations led to unstable primary plant parameters and thus leak rate surveillance test results that were unreliable. The results ranged from negative results to positive results that exceeded the limiting condition for operation in the technical specifications. Because of the variability in test results. Mr.' Wright -

stated that operators had little faith in the leak rate test procedure.

He stated that, test results which showed unidentified leakage in excess of the technical specification limit of 1 gpm were discarded and those below 1 gpm were retained ir. order to show compliance with the technical specification. He stated that as an operator watching the plant operate and respond, the leak rates test results obtained, whether " good" or " bad" bore

• little relationship to what he believed to be the true status of plant 1eakage. See pages 27-33.

While Mr. Wright believed that there was a general feeling of dissatisfaction with the leak rate test procedure, he stated operators were concerned more about other significant operational problems in the plant than they were about leak rate tests. Although Mr. Wright admitted that the leak rate surveillance

• test was the only quantitative method that could be used to show compliance with the technical specification, he stated operators used other plant parameters such as make-up tank (MUT) level indication, pressurizer level indication, reactor building sump levels and the number of sump pump starts per shift to detemine if there was a leakege problem in the plant. See pages 32-39.

According to Wright, leak rate test results that exceeded the technical specification limit were thrown away. He did not know when that practice started but it was connon practir.e. Although Mr. Wright did not recall ever being directed to throw test results away, they were discarded with the knowledge and/or consent of his Shift Supervisor and his Shift Foreman.

Mr. Wright did recall that for a short period of time he was told to keep all leak rate surveillance test results; however, this practice was soon discontinued. He stated that he was never given any indication or reason to believe that bad test results were to be discarded so the NRC would not see them. Mr. Wright considered the bad test results worthless and was not aware of any technical specification or administrative procedure that required them to be keep. He therefore did not consider discarding such tests improper at the time. See pages 36-41.

g't :t:ted t6ct W ; r d *::t c:d t: e N;cd " the CSC' LG --

oneico * * " " " '

He was not aware at the time of ar y requirement to 109 all te!sts. Nr.Wri$ht

,,,4 """ dTd"h6h"fi41Tde" v Disso"msny 74%s6H"al'1"Edi ts"WeFi"EdE74gTd',w3s"t'6"k fs1"ths"f scY" """""""""

that"they had"to run test 1 before they-could set"en" acceptable """

sh See ,pa,g,e,s,,,j,1

, ,$,,f,{,.,,,,,,,,,,,,,

c ac:= sia no,som cu om.o

, OFFICIAL RECORD COPY *"'"2"*-""

g ' M giam T. Russh %ep , q- 4 August 27, 1985 Mr. Wright was asked how he interpreted the requirements of the technical specifications with respect to when the action statement was required to be entered. Mr. Wright said that since the leak rate test was required to be run only every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, he and other operators interpreted that to mean that once a good leak rate test result was obtained, they had the full 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> to get another good leak rate result before they had to enter the action statement. Regardless of how many bad leak rate tests were received during the 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, they would not enter the action statement. See pages 43-45.

Note: Surveillance Procedure 2301-3D1 states that if unidentified leakage exceeds 1 gpm, proceed with action statement 3.4.6.2.b.

Action statement 3.4'.6.2.b reads: "With any Reactor Coolant System leakage greater than any one of the above limits, excluding PRESSURE BOUNDARY LEAXAGE, reduce the leakage rate to within limits within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> or be in at least HOT STANDBY within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in COLD SHUTDOWN within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />."

A copy of LER 78-62/1T along with an attached routing sheet with the typewritten names and handwritten initials of control room personnel was provided to Mr. Wright for his review. The routing sheet was used to indicate which individuals had seen the LER. Mr. Wright confirmed that the initials on the routing sheet appeared to be his; however, he did not have a preaccident recall of the incident described in the LER. See pages 45-50.

Note: During a routine inspection of THI-2 operations on October 18, 1978, an NRC inspector discovered,several bad leak rate tests lying in the control room and that THI-2 had been operating for an extended period of time with unidentified leakage exceeding the technical specification limit. The incident resulted in the submittal of Licensee Event Report

(LER)78-62/1T. -

l The LER states in part: "This event was caused by a misinterpretation of the requirements of the technical specifications. The appropriate personnel will be instructed on the requirements of applicable sections of the T.S. and the requirement to immediately invoke applicable action statementswhentheprovisionsoftheLCOs[limitingconditionfor operation] are not met."

Even though Mr. Wright did not recall the specific incident, he did not believe he was given any instruction or guidance on a revised interpretation of the requirements of the technical specifications following the October 1978 incident. He believed it was possible that the short period of time they were told to keep all leak rate tests, regardless of the results, may have been tha result of this incident. See pages 50-54.

Mr. Wright was asked if he ever complained to his supervisors about the difficulty he was experiencing in getting good leak rates. He could not recall any specific instance; however, he was sure there were discussions involving his Shift Foreman and Shift Supervisor regarding leak rate problems. Mr. Wright was informed that most other operators have testified that as the date of the accident approached, acceptable leak rate test

+

2t& U-TtuY% W'- r%t'*c-m ---

serica p

"""" -d tt "::n recolle' p tion. %-tIt de:ared ap to himthat they had h1 ways had trouble running

,_ ,3 ........mrmngm"..rnrtsnnirnuwmnhridsmrEnh@ Ter ""..". "."." ' "." "

........... ............. ...... ..................... ..................... ..................... . . . . . . . . . . . ~ . - - ... - .

em) .................. .................... ..................... ..................... ..................... . . . . . . . . . . . . . . . . . . . . .................

ccessuaiano,so,o cuo24o

,_ OFFICIAL RECORD COPY *"**""-"*8

s. - William T. Russell  % ,.

-se A g August 27,g increasing. Mr. Wright was aware that their computer specialist, Bill Fels, was cognizant of the problem and was working on modifications to the program; however, he was not aware of who in the management chain was specifically advised of the problem. He could only speculate that the Supervisor of Operations should have been aware of the situation. See pages 55-58.

Although he had never read the details of the Hartman allegations, he was generally aware that allegations were made that operators had used water and hydrogen additions improperly to assist in getting good leek rate test results. Mr. Wright stated it really didn't seem to him like there was anything that would influence the outcome of a test, "it appeared to have a mind of its own." He considered the computer program for leak rate calculations to be a " random number generator." Mr. Wright believe that in addition to the computer program not properly accounting for changes in plant parameters, problems: he believes that two other conditions also led to leak rate test (1) the difficulty in establishing steady state conditions, and (2) instrument inaccuracies and response time. See pages 58-68.

It was pointed out to Mr. Wright that the surveillance procedure cautioned operators not to perfonn certain evolutions during the performance of. leak rate tests including the addition of water to the RCS. Mr. Wright res by stating that the procedure advised against such additions; however,ponded if such additions were required, the amount of water added'had to be manually entered into the computer by the operator before the final calculation was performed. At times water additions during leak rate' tests could not be avoided. During February / March 1979 THI-2 experienced increased leakage from the top of the pressurizer through either the PORY and/or the safety valves.

During this period, frequent water additions were required to replace the water lost through the top of the pressurizer. Since the water additions

' became more frequent as the date of the accident approached, it was difficult to run test. Seea leak pages rate test without having to add water at some point during the 71-73.

In discussing how water additions were made to the RCS, Mr. Wright stated he would normally determine the amount of water actually added to the RCS by looking at the change in level on the MUT strip chart recorder. While the amount of water added was also indicated on the totalizer for the batch controller, he stated that he did.not use this value. He recalled that there were problems with the batch controller. At times the automatic features of the batch controller did not provide the full amount of water desired and at

, other times it would add more than the desired amount. See pages 68-74.

Note: One of the major points of conflict between the testimony of Messrs.

Olson and Coleman, when they were interviewed by the NRC on this subject, was their explanation of why 12 out of the 13 leak rate tests, performed by Shift "D" during the months of February and March 1979, showed underrecorded water additions very close to the end of the test.

An underrecorded water addition is one in which the amount of water logged in the CRO's 1.og and included in the leak rate calculation is less than the amount of water added as indicated on the MUT strip chart. Mr. Coleman testified that he was aware of the difference between the totalizer and the MUT strip chart and he took advantage of that differovo hv intentine llv addine water near tha end of the tMt knowing that the computer would calculate a icwer leak rate than would _ , ,

conca) ...................

" "'I th'e"ca ra .""Mr Otson;"on"tti rother* hand;* stated that*h r"""""""

................... .Ot W se**M' car) .................. .....................

ec e:cu sie no,so,==cu on.o OFFICIAL RFCORD COPY *"8""""-

J ill.f = T "=seE .-=emer -S. q

. August 27; M_ l was not aware of the instrument inaccuracy and that the water had to

.have been added for some legitimate operational reason.

According to Mr. Wright, water was added intentionally near the end of the I test; however, he disagreed with Mr. Coleman's reason for the addition.

Wright stated he personally added water near the end of the test in order to

! bring the level in the MUT back up to the approximate level it was at the 1

beginning of the test. He said since MUT level was read by the computer only at the beg hning and the end of the test, it was his theory that by making  :

the init,ial and final values of MUT level approxisaately the same, he could '

eliminate possible calibration errors in the MUT level instrumentation as one of the possible sources of inaccuracy in the calculation. It was pointed out to Mr. Wright that while his explanation provided a reason why the water was added, it did not explain why the amount added according to the CRO's Log and provided as manual input to the computer was less than the amount indicated l

on the MUT strip chart. If he were using the totalizer value, the difference was understandable; however, as indicated by his earlier testimony, he was i _

aware of the differences between the instruments and stated he used the MUT

strip chart recorder to determine the amount of water added. ~ Seepages

! 75-83. ,

Four of the 12 underrecorded water additions perfomed by Shift "D" involved l Mr. Wright. In two cases he was the operator who ran the test and in two cases he was the primary plant operator (i.e., in charge of water additions) during the period the leak rate tests were conducted. All four tests were performed with Mr. Olson and not Mr. Coleman. These four tests were reviewed in detail with Mr. Wright. In all four cases the water was added within four

minutes of the end of the test and the final value of MUT level was close to i' the' initial value. The difference between the amount of water added -

according to the CR0's Log and provided as manual input to the computer was between 22 and 5g gallons less than was indicted on the MUT strip chart. -See i

pages83-104.

• Next hydrogen additions were discussed with Mr. Wright. Mr. Wright stated that he added hydrogen to the MUT approximately four times while leak rate tests were being conducted. He stated the first time he added hydrogen  !

during a test, the leak rate test result was satisfactory. After that, he could not see that it had any influence on the test result. Unlike other operators who admitted adding hydrogen in order to affect MUT level indication, Mr. Wright said he was not aware that a hydrogen addition could influence level indication. He stated that when he added hydrogen during a test it was for essentially the same reason he added water (i.e., to have the final value of the parameter equal to its initial value). Thus, if hydrogen pressure at the beginning of the test was 18 psig, he would add hydrogen to the MUT to increase the pressure to 18 psig near the end of the

, test. Mr. Wright stated he aoded hydrogen many times during periods when leak rate tests were not in progress and he could not recall seeing any change in MUT level as a result. See,pages 104-108.

Note: Since MUT overpressure is not printed out on the leak rate surveillance test sheet, it is not possible to verify Mr. Wright's testimony on this matter.

cesier) Mr. W ipht stated that at no time was he aware that either Fr. Olson or N *.

,,,4 ' """"' C61 Esh Tddfd*Wh'tWP"ahd'ur"hydF606M"during"tM cpye"Of"4"1 erk" rate" test in"~~""~

- ~~~ order to-intentiona l4y fa1sMy*st-resultsv~. see pages. 108 L .10 9.. - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . . ,'

one) .................. ..................... ..................... ..................... ..................... .................... ................

cm mno,....e one OmCIAL RECORD COPY *"""O******

I

. ' ' ~ NTliam T. Russei rA %7 - August 2 h 985 Mr. Wright was also questioned about pressure being placed on operators to get a good leak rate test result. He stated that there was pressure. Each' of the operators felt they were competent operators and if you were getting close to the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> surveillance deadline, there was pressure to do the best you could to get a good leak rate. Mr. Wright was asked if he was ever told by his supervisor to "do whatever you can to get a good one." Wright responded by stating that he could not specifically recall such-an instance but he could postulate something along that line by having an operator continue to run tests until a result within the technical specification was l obtained. Wright considered the results meaningless; however a good test result allowed the plant to continue to operate. Mr. Wright stated the -

! operators wanted to keep the plant running. Whether it was right or wrong, l

in hindsight, he said the operators felt comfortable with the plant from a leakage standpoint. From everything he had observed in his duties and watching the panel, he was not concerned that he had a major problem with the reactor plant as far as leakage. He thought his problem was with the leak

rate calculation. See pages 109-112.

At this point in the interview, Mr. Wright was asked if he would agree or' disagree with the following statements regarding leak rate practices at THI-2 that were made by Mr. Coleman during his December 14, 1984 interview.

1. Only good leak rate tests were retained and logged in the CR0 log. Leak rate tests showing unidentified leakage that did not meet the acceptance criteria of the technical specifications were thrown away and not logged in the CRO's Log. Bad leak rate tests were thrown away at the direction of, or with the knowledge and consent of the Shift Foreman and Shift Supervisor. Mr. Wright agreed with this statement. See pages 122 & 123.

2. When a bad leak result was obtained, the test would be rerun until an acceptable test was obtained. Mr. Wright agreed with this statement. ,

See page 123.

3. With respect to the October 1978 LER, Mr. Coleman could not recall receiving any direction from his supervision to perform leak rate tests differently following the incident. Leak rate tests that did not meet the acceptance criteria continued to be thrown away, and as best he could recall, he never entered the action statement of the technical specifications. Mr. Coleman stated that it was his understanding that leak rate tests that did not meet the requirements of the technical specifications were to be thrown away so the NRC would not see them.

Mr. Wright agreed with the first sentence of this statement; however he disagreed with the rest. Wright.did recall a short period of time when '

all leak rate tests were retained but was not sure whether it was associated with this incident. He did not perceive the return to the old policy of throwing away leak rate tests as a problem since he would have assumed that the return to the old policy was as the result of the NRC reconsidering the need to keep them. See pages 123-127.

4. Mr. 'Coleman stated that good leak rate results were harder to get as the date of the accident approached. Mr. Wright disagreed with that statement. It was Wright's recollection that they were always difficult e ~t M W d - ~"11 my n.,<na hatag o n ..aiag1y m . difficult emer) .t...i.m. .e.

..................t.h. a. .n. . .a. .n.y. . .o..t.h. .e. r. . .p e. .r i. .o.See

d. page 127.

m ................ ..................... ,,,,,,,,,,,,,,,,,,... ...............,,

4C1CMtf ................... ..................... ..................... .....................

4 .................. ..................... ..................... ..................... .....................

cc ron m no,.oi .c *uo OFFICIAL RECORD COPY *""""-***

liam T. Ruf N S August N el985 5.

Mr. .

Coleman stated pressure was exer'ed rates.

t on operators to get good leak The pressure increased the closer they came to the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> limit.

Mr.. Wright agreed with that statement. See page 128.

6.

Mr. Coleman stated that he could recall one instance where he was told h had no other duties during the watch except to keep running leak rate tests until he received one that was within the acceptance criteria of the technical specifications. Mr. Wright could not recall such an incident. See page 128.

7.

Despite the pressure, Mr. Goleman stated that he was never directed by his supervisors to intentionally manipulate leak rate tests. Mr. Wright agreed with that statement. See pages 128 & 129.

8.

Mr. Coleman stated that the pressure stemed from not wanting to have to shut down the plant. With the plant off-line, they were not generatin any electricity and consequently the Company was not making any money.g Heistated that he and the other operators owned stock in the Company and they felt it was their job to generate electricity. Mr. Wright agreed with the spirit but not the tone of the statement saying that he never gave any consideration to the few shares of stock he owned *in the company but they were operators and it was their job to operate the plant and generate electricity. _See page 129.

9. Mr. Coleman stated that he first learned that hydrogen additions could influence the outcome of a leak rate test from Mr. Hartman. Mr. Wright stated that he did not recall any conversations with Hartman about hydrogen. He also did not recall any discussions with Coleman about hydrogen; however. Wright stated since they worked on the same shift it was possible that they had some. Mr. Wright stated that he had added t

l hydrogen several times during . leak rate tests but it was to bring t

hydrogen pressure back to its initial pressure it was not for the ,

\

purpose of altering MUT level. See pages 130 & 131.

10. Mr. Coleman stated that he personally observed a rise in MUT level indication when hydrogen was added. He stated that the level rise could be accomplished with a nonnal controlled addition of hydrogen. It was not necessary to make only a short duration addition of one or two seconds. Mr. Wright stated that he did not have that knowledge at that time.

. See pages 131 & 132. .

11. Mr. Coleman also ' stated the rise in MUT level indication produced a transitory effect lasting only a few minutes. Therefore, if the operator i

' was going to use hydrogen to influence a leak rate test, hydrogen would need to be added near the end of the test. Mr. Coleman said that it was comon knowledge among CR0s, Shift Foremen and Shift Supervisors that hydrogen additions could alter MUT level indication and therefore, influence leak rate test results. Mr. Wright disagreed with Mr. Coleman,

' stating that he did not recall having that knowledge at that time when they were doing leak rate tests. See page 132.

l

12. Mr. Coleman stated that he did not recall seeing Mr. Olson add hydrogen fn the MIT fne the nnennes nf altmeinn lomb entoe yo c o o r8 hmavor 4+

carict) Wa 5 hard for him to believe tbat any of the

...............g g. . g.g ....

operators dit not kn0W thi8

, g...yg. ..gg.ppgggjog.t5 ting"th3 t"it' *15" d bV TG t" t Rt'"" *""" * '

l ...............on.. ..................... ........o u ' " " "

.""~'nn cun> .................. .....................

..................... ..................... . . . . . . . . . . . . . . . . . . . . ..~..~..........

cer:cusiano,eomacuo2

, _ OFFICIAL RECORD COPY *" ""*

l A liam T. Russ3142 e %9 - AugustY,198 he added hydrogen, but the same thoughts were not going through his head when he was adding hydrogen evidently that were going through Mr.

Coleman's. See pages 132 & 133.

13. Mr. Coleman confimed that he was aware that by adding water during the last few minutes of the test he was able to get the MUT level indicator to read a higher value than the actual amount of water added to the tank.

Mr. Coleman was then asked if this method was used in lieu of hydrogen additions because it was not possible to add hydrogen from the control room during the last few months before the accident. Mr. Coleman stated that was the reason. Mr. Wright again stated that he had no such knowledge. He admitted that he added water near the end of some leak rate tests, but it was to bring the final MUT level back up to the original value at the start of the test. It was not done to manip'ulate test results as stated by Mr. Coleman. See pages 133-137.

Mr. Wright was also questioned on his knowledge of problems with the MUT .

level transmitters (LTs). In general, Mr. Wright stated that they did have problems with the two level transmitters (LT-1 and LT-2). He recalled that quite frequently one transmitter or the other was taken out of service and worked on by' contractor personnel. He did noti recall one transmitter giving them more problems than the other. See page 138..

It was pointed out to Mr. Wright that almost the entire time the plant was operating during the months of December 1978 and January 1979, LT-1 was providing a very erratic and unstable output. It was also pointed out that when LT-2 (the stable transmitter) was providing input to the MUT strip chart recorder,LT-1(theunstabletransmitter)wasprovidinginputtotheplant computer which was used to perform the leak rate surveillance test. Mr.

Wright was shown examples of leak rate tests which were perfomed using bo,th LT-1 and LT-2. Mr. Wright agreed that the tests which used LT-1 as input to the computer could not be considered valid tests. _See pages 138-141. ,

Mr. Wright did not recall how his shift ran leak rate tests during this period. He assumed that they would always try to have the stable level transmitter providing input to the computer for the period of the test. Mr.

Wright was asked if he was aware of any operators switching level transmitters during the test in order to take advantage of the highest level reading near the end of the test. Mr. Wright stated he was not aware of that situation occurring. See pages 141-144.

Mr. Wright was next asked if he agreed with some of the statements made by Mr. Olson when he was interviewed on this. subject by the NRC on November 15, 1984.

1. Mr. Olson stated that after the October 1978 LER was issued, he recalled being briefed on the proper interpretation of the requirement of the technical specifications and that following that time, he never threw away an unsatisfactory leak rate test. Mr. Wright stated that Mr.

Olson's statement was not consistent with his recollection. See pages 145-146.

9 ur nie-r .....a .s., wo use cr--1o. 19 ---r, ..s1, u4.s ,s,1..v

r. .

en, cn s drveillance teht program as it was wr'itten and implemen't ed and that he

.................d' dN't"HFvW"dH7'db6dsrMt"Eb'dl t"thE"sutVFit MMC'd"tWi~t"111 t!1t!"*ME"Wi l; ht """"""'

.................sta ted-tha.t.Jie. would.have to disagree.with N N.01sen.'.s-statement,.- See - -..........

cm) ................p. a .ge 146. .................... ..................... .....................

=c c:= m no, o.. mew ouo narieeas si n - ~~ ^

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

.2..

.Hi g am T. Russe 1 Q w 9 10 -

r August 2H99985 -

3. Mr. Olson stated that he would only make a hydrogen addition upon the specific recomendation of the chemistry personnel. He said that he never added hydrogen independently. Again Mr. Wright did not agree with this statement. The operators had a normal operating band for hydrogen pressure in the MUT. If pressure was low in the band they would add hydrogen. In addition, if hydrogen concentration was low in the RCS. the chemistry personnel may have recomended adding hydrogen; however, the the CR0 could add hydrogen independently. See,pages 151 & 152.

In concluding the interview, Mr. Wright was asked if he had any reason to believe that any operators, including Mr. Olson and Mr. Coleman, falsified any leak rate tests by any means including: underrecorded water additions, failure to record water additions, or hydrogen additions. Mr. Wright stated .

that he had no knowledge of such activities among any of the operators at that time. He also sa< d that while he personally added hydrogen and water during leak rate tests these additions were done only in an attempt to establish the same final conditions as were present at the beginning of the test. They were not done in an attempt to wrongfully manipulate test results. Except for the unreliability of the test Mr. Wright believed thht

all leak rate tests were performed properly. See,pages 147 & 148.

?. . .

1 i ha+ t .*W

! Robert A. Capra, Technical Assistant Division of Human Factors Safety, NRR

Enclosures:

1. Transcript of Interview -

2. NRR Leak Rate Evaluation l

l . cc: K. Christopher (RI) -

l e

="=> .R.T.s.ey.._ ..................... ..................... ..................... ................. _ . .. _ ... _ _ _ _ . _ _ _ .......

• * **> . 8.G4RT.4.t...... ..................... ............... _ _ ..................... _ - .... ~ .. - - . - ~ ~ . - ~~~~-

• * " > s37/85...... .....................

l c c cr.s w m no,so oew o u o OFFICIAL RECORD COPY *"*****" .

) .

I ., - __ _ _ . _ _ _ _ . _ _ _ _ _ _ . - - - - - -- -- -