ML20155J653

Partially Withheld Results of Joint Ofc of Investigations/ NRR Evaluation of Er Frederick Re Leak Rate Testing Irregularities.Accuracy of Statements Questioned.Personal Involvement in Irregularities Discounted.W/O Encl 2

ML20155J653
Person / Time
Site:	Three Mile Island
Issue date:	05/30/1985
From:	Russell W Office of Nuclear Reactor Regulation
To:	Harold Denton Office of Nuclear Reactor Regulation
Shared Package
ML20151L202	List: ML20151L198 ML20155J582 ML20155J591 ML20155J599 ML20155J625 ML20155J636 ML20155J641 ML20155J649 ML20155J653 ML20155J677 ML20155J684 ML20155J692 ML20155J704 ML20155J710 ML20155J713 ML20155J726 ML20155J745 ML20155J753 ML20155J770 ML20155J778 ML20155J792 ML20155J808 ML20155J825 ML20155J835
References
NUDOCS 8605270146
Download: ML20155J653 (24)

Text

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

i

~

Enclosure 7 j +~ , ,,,\ UNITED STATES l

]3 g a NUCLEAR REGULATORY COMMISSION WWASHINGTON, D. C. 20586 f *g -

May 30,1985 l l  ;

~

l .. f MEMORANDUM FOR: Harold R. Denton, Director , t

, Office of Nuclear Reactor Regulation -

. j 1 i

1. FROM: William T. Russell, Acting Director '

i Division of Human Factors Safety, NRR j ,

SUBJECT:

RESULTS OF JOINT 01/NRR INVESTIGATION AND EVALUATION OF  !

l EDWARD R. FREDERICK ,

Reference:

1. Memorandum from S. J. Chilk (SECY) to B. B. Hayes (01) l

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

Subject:

i Staff Requirements-Discussion of Pending Investigation- ,

l MI

' 2. Memorandum from H. R. Denton (NRR) to'B. B. Hayes (01) ,,,

dated May 3, 1984, Subject NRR Review of 01 Investi-i gation Materials Concerning Hartman Allegations of j Falsification of Leak Rate Data at TMI, Unit 2 l

L 3. Memorandum from W. T. Russell (DHFS) to H. R. Denton 5 i (NRR) dated January 24, 1945,

Subject:

Follow-up

] Action on Additional TMI-2 Operators 6 1

) The purpose of this memorandum is to document the results of the joint OI/NRR  !

i investigation and evaluation of Mr. Edward R. Frederick, currently a Licensed  !

j Senior Reactor Operator (SRO) at Three Mile Island, Unit 2 assigned as an i

I Instructor Y in the Corporate Training Division, and to provide a reconnenda-  !

tion regarding whether his current SR0 license should be revoked.. modified, or suspended under 10 CFR 55.40 or other enforcement action taken under '

10 CFR Part 55.50, due to his involvement in preaccident leak rate testing  !

irregularities at TMI-2. ,

j Background '

As a result of a Consission meeting on Maren 23, 1984 NRR was directed by l j

Reference 1 to review OI investigative materials concerning falsification of reactor coolant system (RCS) leak rate tests at TMI-2 and refer back to 01 5

those matters which required further investigation. The results of NRR's i

- I

! NOTE: This memorandum and Enclosure 1 discusses information which is the

} subject of an ongoing 01 investigation. This memorandum and i Enciesure 2 discusses information that is maintained in the NRC's 1

Privacy Act System of Records (NRC-16). This memorandum and l

{ enclosures may not be disseminated outside the NRC without coordi-j nation with NRR and the permission of the EDO or the Director. 01.

Internal access and distribution should be on a "need to know" basis.

1 1.

i i

i Harold R. Denton May 30,1985,

~

i i

review was,provided in Reference 2. The review determined that follow-up .

investigation by OI and further evaluation by NRR was needed in the case of  ;

seven currently licensed operators. E. R. Frederick was one of these seven - i

.' individuals. At a follow-up Connission meeting on May 23, 1984 NRR pro- l posed joint 01/NRR investigations and evaluations of these individuals.  !

Subsequently, NRR issued letters to these individuals under 10 CFR 55.10(b) ,

requesting additional information regarding current performance. Based upon '

the joint investigation into the individuals' past involvement in improper .

1 activities at TMI-2 and an evaluation of the individuals' subsequent per- i j formance, NRR would recommend what action, if any, should be taken against  :

j the identified operators. As a result of some of the early interviews with i

. these individuals and further technical analysis, Reference 3 recommended

that three addition'al currently licensed operators also be interviewed. NRR ~

4 issued similar letters to these individuals under 10 CFR 55.10(b). Of the

! ten individuals identified for investigation, seven are currently licensed

at Three Mile Island, Unit 2, two are licensed at Waterford 3 and one is licensed at San Onofre 2 and 3. -

4 Past Involvement in TMI-2 Leak Rate Testing Irregularities On January 30, 1985, a joint Office of Investigations (01)/ Office of Nuclear Regulation (NRR) interview of Edward R. Frederick was held in the law office of Killian & Gephart, Harrisburg, Pennsylvania. The interview was conducted in the presence of Mr. Frederick's personal attorneys. The purpose of the interview was to determine Mr. Frederick's role, if any, in improper activities associated with RCS leak rate surveillance testing at TMI-2 prior to the accident on March 28, 1979'. At the time..Mr. Frederick was a licensed Reactor Operator (RO) and served as a Control Room Operator (CRO) on Shift "A." A detailed evaluation of Mr. Frederick's interview is included as Enclosure 1 to this memorandum.

Mr. Frederick denied that he personally attempted to influence the outcome of ,

leak rate test results by the addition of water and/or hydrogen to the l make-up tank (MUT) during the tests. His testimony regarding his lack of personal involvement in leak rate test manipulation or falsification is '

consistent with NRR's technical analysis. 4r. Frederick also denied that he had personal knowledge of or was aware of other operators intentionally manipulating leak rate surveillance tests. However, many of Mr. Frederick's responses to questions concerning preaccident reactor coolant system leak rate surveillance testing at TMI-2 were inconsistent with the testimony given by the majority of other CRO's interviewed to date. His testimony is also inconsistent in part with the testimony given earlier the same day by Mr. Craig-C. Faust, the other CR0 on shift with Mr. Frederick durin the six month period under investigation. The majority of conflicts deal w th the number of times the test was conducted; the difficulty in obtaining test results-that were within the allowable limits of the Technical Specifications; what was done with test results that were unacceptable; and interpretation O and compliance with the Technical Specifications and Limiting Condition for Operation associated with RCS leakage. During the interview, Mr. Frederick provided the following information:

l

1. '

j -

Harold R. Denton May 30,1985 i l -

w , >

j 1. Althqugh the THI-2 Technical Specifications required that the RCS leak ,

rate surveillance test be conducted once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, Mr. Frederick i

! recalled that the test was conducted daily, usually on the mid-shift. _

1 If a successful test result was obtained on that watch, Mr. Frederick l 1

stated that he wculd not run it again that day. Note: The majcrity of 5

! operaters interviewed have stated that the test was conducted once per  !

- shift, plant conditions permitting and provided the plant computer was  !

l available. Of the 25 tests involving Mr. Frederick during the period .

i under investigation, 14 were done on the mid-shift, 9 on the swing-shift l l and 2 on the day-shift. Of the 11 tests performed on the day- or swing- ,

4 shift, 6 times an acceptable test had been conducted earlier in the day. i

!- In one case, 2' acceptable tests had already been conducted that same  ;

i- day. The difference between Mr. Frederick's recollection of test  !

l frequency and actual test frequency may be attributed to difficulty of  :

j recall with the passage of time since the difference between daily  :

4 testing and testing each shift if plant conditions are stable is not a  !

significant difference in the number sf tests performed. This is due to i j the high frequency of unstable plant conditions (e.g., start-ups,  ;

shutdowns, power changes, etc.) between October 1978 and March 1979.

l

! 2. Mr. Frederick stated that to-the best of his recollection, he never l threw away any leak rate surveillance test results. He said that all ,

lake rate tests he ran were turned over to the Shift Foreman for his  ;

! review. He did not know if they were subsequently thrown away or not.

Note: Tha mejority of operators interviewed have indicated that in u lid -

l or ucacceptable test results were routinely discarded, In a March 29 ,

1984 01 icterview with Mr. Frederick's Shift Supervisor, Mr. Zewe, t i stated that the discarding of invalid tests was accepted practice at the  !

J time. Mr. Faust also stated in his interview that invalid tests were i j routinely discarded. Thus, while it is possible that Mr. Frederick did l

! turn all leak rate tests over to his Shift Foreman for review, it does 1 i not seem credible that W. Frederick would not know that test results

• l showing greater than 1.0 gpa for unidentified leakage were routinely I j thrown away. ,

l 3. Mr. Frederick stated that he interpreted Technical Specification 3.4.6.2 ,

as requiring the operater to enter the Action Statenant .of the Technical  :

Specification whenever they received the results of a leak rate test l

, that exceeded the Limiting Condition for Operation, unless he could j

! explain why the test result was invalid. Note: The majority of j operators interviewed including Mr. Faust, T ieved that they were not .

required to enter the Action Statement, no matter how many bad leak rate ,

! tests.were received, as long as they had not exceeded the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> -

l

} required test interval without obtaining a successful test result. j i While Mr. Frederick's interpretation of the requirements of the l l TechrLical Specification is correct, there is indication that j

Mr. Frederick did not comply with this interpretation between October ~

l 1978 and March 1979. For example, he testified that he would not be l

! surprised if there were times when they ran a leak rate test on their. '

shift, obtained an unacceptable result and by the end of the shift they J

, 1

!. J t >

I l

i I i . i Harold R. Denton May 30,1985 l

! stilihadnotgottenanacceptableresult. Since the Action Statement  !

l requires that leakage be reduced within acceptable 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 />, had Mr. Frederick .

! followed his interpretation of the requirements of the Technical - -

Specification, as is stated, there would have been periods of time where

the plant should have been shut down.

4. Mr. Frederick also testified that he did not recall having difficulty  :

i . getting leak rate test results within the Technical Specification limits' and that if he had any problems he usually knew why. Note: The majority

of operators interviewed, including Mr. Faust,19dicatRThat leak rate l

test results were difficult to get and that as the date of the accident r

! approached.. it became more difficult to get an acceptable test. result.

j Mr. Frederick's statements regarding the difficulty of getting test j' results also appears to conflict with his 'ater testimony that it would i not surprise him if there were several times where they were not able to i get an acceptable test result their entire shift. -

j .  :

i

5. Mr. Frederick testified that he did not log the re:ults of leak rate '

tests in the CRO's Log for any tests (i.e., good v; bad) because items j of a routine nature were not required to be logged, Ncte

Administra- l tive Procedure 1012 requires that the test title, test number, and the

j. start and completion times of all surveillance tests rergired by the <

i Technical Sp'ecifications be logged in the CRO's Log. Shift "A" was the )

i only shift that did not routinely log satisfactory test results. None .

l of the six shifts logged unacceptable test results. .

t

6. Administrative Procedure 1010 required that problems encountered while (

performing surveillance testing should be documented on an " Exception  ;

. and Deficiency List." The list was to be attached to the surveillance '

test data sheet. Mr. Frederick stated that AP-1010 was not applied to leak rate tests, although he did not know why it had been excluded. ,

j Note: None of the six shifts applied AP-1010 requirements to leak rate i i surveillance testing.

i i 7. Although he vaguely recalled two other operators (Mr. Hartman and

Mr. Booher) complaining about the difficulty of surveillance testing in

! general, Frederick stated that he personally had no problems running the

! test. He attributed this to the fact he was very careful in setting up

! the required initial conditions for the test. He stated that he questioned all leak rate tests that were received. He stated that he i

< could usually predict what it should be. If it wasn't what he expected, I

he would question it. He examined each test result carefully and if he I i understood it, then it was alright. Note

These statements do not ,

j coincide with the results of the technical analysis which shows that two 1 l cf his tests were conducted during unstable plant conditions and six I i tests-were performed using the unstable level transmitter (LT-1) as 1 input to the computer for MUT level indication. In four of these tests, i -

there was a significant difference between the output of LT-1 and LT-2.

f Had Mr. Frederick reviewed the data carefully, as he testified, he

! should have seen these discrepancies.

I f

l

~ ~ . _ _ _ _-____ _ . - _ _ _ . _ _ . _ _ _ _ _ _ . . _ _ _ _ . _ _ _

1 1

l Harold R. Denton -

5- May 30,1985

8. Freder)ckstatedthatLER76-62/1Tneithe'ralteredhisinterpretationof the Technical Specification 3.4.6.2 nor did it change the way he performed 4 the test or handled the test results. Frederick stated that he did not -

have a preaccident recall of the LER.

9. With respect to hydrogen aoditions, Frederick stated that he was not aware, prior to the accident, that hydrogen additions to the MUT could j affect MUT level indication and consequently leak rate test results. He did not recall ever dis:ussing the effect with anyone or knowing that individuals were adding hydrogen in an attempt to influence the test results. Although Frederick was told that Mr. Faust stated he (Faust) was aware prior to the accident that hydrogen additions were alleged to alter the test *results, Mr, Frederick does not recall discussing the subject with Mr. Faust or any other cperator. Mr. Frederick's testimony that he did not add hydro 3en to manipulate leak rate test results is consistent with NRR's technical analysis of his leak rate tests.

10. With respect to water additions, Frederick stated that he never inten-tionally added water and n:t properly included the amount in the leak rate calculatier, in an attempt to influence the leak rate test results.

! He also stated that he was not aware of other operators manipulating tests in this manner. Note: Although, water additions, that were not Os completely acccunted for in the test calculation, occurred during three of Mr. Frederick's tests, there are credible reasons why that may have accidencly occurred. In additien, the number of tests performed by Mr. Frederick, during the high identified leakage period of February and March 1979, where water was added and properly compensated for, supports Mr. Frederick's testim:ny.

In sunrr.ary the technical analysis supports Mr. Frederick's testimony that he did not intentionally add water and/or hydrogen to the MUT during the leak rate tests for the purpose of manipulating the test results. However, other aspects of his testimony decling with difficulty in obtaining satisfactory test results, the disposition of unsatisfactory results, and his overall knowledge of leak rate testing irregularities raise questions regarding the accuracy of his responses durin, the interview.

Current Performance G

b

. _ , , ---.,.--,,---.-+----y, - - , , - - - + - - , . - - - - , - , -,r--.--- - . - - . - - - ,. , , - , , -

Harold R. Denton 6- May 30,1985

( -

~

d O .

Conclusion During the period Mr. Frederick was licensed as a Control' Room Operator at TMI-2 prior to the accident, he states that he was not involved in nor had .

knowledge of leak rate testing practices that were contrary to plant proce-dures and approved Technical Specifications. Mr. Frederick also denies that he was involved in leak rate test manipulation or falsification or that he had knowledge that such actions were going on around him. While the technical analysis supports his testimony that he did not intentionally add water and/or hydrogen to the MUT during leak rate tests for the purpose of manipulating test results, other aspects of his testimony dealing with difficulty in obtaining satisfactory test results, the disposition of ,

unsatisfactory results, and his overall knowledge of leak rate testing irregularities raise questions regarding the accuracy of some of his '

responses during the interview.

bN e

~

Harold R. Denton May 30,1985 In sumary while I suspect the accuracy of some of Mr. Frederick's state-ments regarding leak rate testing irregularities, I am unable to > rove that he lied. I do not believe that he was personally involved !n leac rate _. I surveillance test manipulation or falsification . -

)

l MM T, h 1 ak '

William T. Russell, Acting Director i Division of Human Factors Safety Office of Nuclear Reactor Regulation

Enclosures:

l

~

As stated cc: B. Hayes '

K. Christopher  ;

f a

I.

l-

• i

[

.u p

i I

i 1

1_ .._ _ ...___ ___, _ _ _ _ . . . . _ _ _ _ _ _ _ _ _ . . . - _ _ _ _ , . _ . _ _ _ _ . . . _ _ . . . _ _ _ _ _ _ . . _ . . . _ _ _

NOTE: This enclosure discusses information which is the subject of an ongoing OI investigation. This enclosure may not be disseminated outside the NRC without coordination with NRR and the permission of )

the-EDO or the Director, OI. Internal access and distribution should i

! pe_on a "need to know" basis. i I - '

1 Enclosure !

s PAST INVOLVEMENT IN TMI-2 LEAK RATE TESTING IRREGULARITIES ]

I. Background f

On January 30, 1985, a joint Office of Investigations (01)/ Office of Nu. clear i Reactor Regulation (NRR) interview of Edward R. Frederick was held in the law office of Killian & Gephart. Harrisburg, Pennsylvania. The purpose of the interview was to determine Mr. Frederick's role, if any, in improper i activities associated with Reactor Coolant System (RCS) leak rate i

! surveillance testing at Three Mile Island, Unit 2 (TMI-2) prior to the 4

accident on March 28, 1979. Present during the interview representing the j NRC were: R. Keith Christopher, Director, Office of Investigations, Region I;' William T. Russell, Deputy Director, Division of Human Factors

], Safety; and Robert Capra, Senior Program Manager, Staff of the Executive j Director for Operations. Representing Mr. Frederick at the interview were .

his personal attorneys, Mr. Smith B. Gephart of the law firm Killian &  !

j Gephart, Harrisburg, Pennsylvania and Mr. Harry H. Voight of the law firm j LeBoeuf, Lamb, Leiby and MacRae, Washington, D.C.. A copy of the transcript '

i associated with this interview is provided as Attachment I to this enclosure.

Mr. Frederick is currently employed by General Public Utilities Nuclear Corporation (GPUN) as an Instructor V in the Corporate Training Division. i Mr. Frederick also holds a Senior Reactor Operator's (SRO) Ifcense for TMI-2. ,

j Mr. Frederick stated that prior to joining Metropolitan Edison Company

! (Met-Ed) in October 1973, he served as an Electrical Operator in the U. S.

! Navy's nuclear submarine program. He was first hired by Met-Ed as Auxiliary ,

i Operator (A0) and was promoted to Control Room Operator (CRO) at TMI-2 in  :

i 1975. He obtained his Reactor Operator's (RO) license in October 1977. He ,

5 remained a licensed CR0 at TMI-2 until after the accident in 1979. He was i l the'n transferred to the Training Department as an instructor. He was  !

t promoted to Supervisor of Non-Licensed Operator Training in 1981. He  :

received his SRO license in January 1982 and was subsequently promoted to Supervisor of Licensed Operator Training for TMI-2 and then for both TMI-1  !

and TMI-2. In July 1984, he was transferred to his present position in the j

! Corporate Training Division. i I

j The interview with Mr. Frederick concentrated on the period September 30,  :

! 1978 through March 28, 1979. During this time period, he served as a CR0 at

! TMI-2 assigned to Shift "A." At that time Shift "A" consisted of the i i following icensed individuals: ,

lO l I

1

i

~

SHIFT "A" Shift Supervisor

~

Bill Zewe .

Shift Fore: nan Fred Scheimann CR0 - Craig Faust CR0 Ed Frederick A sumary of the technical analysis of the leak rate tests involving

Mr. Frederick is provided in Section II of this enclosure. A sumary of the

interview with Mr. Frederick is contained in Section III of this enclosure.

The sumary of interview includes citations to the page numbers of the transcript from which the sumarized information was extracted. Overall conclusions regarding Mr. Frederick's involvement in improper activities associated with THI-2 leak rate surveillance testing are contained in Section IV of this enclosure, j II. Sumary of the Technical Analysis i

In late 1983 and early 1984, a techn'ical analysis of the leak rate surveillance tests performed during the last six months of operation of TMI-2 o was performed by the NRC. This analysis was done as technical support to the O Department of Justice (DOJ) in its criminal proceeding against Met-Ed. The results of this analysis were used in questioning Mr. Frederick during his interview. However, based upon information obtained during the 13 interviews of former THI-2 CRO's conducted to date, a reanalysis was performed during February and March 1985. The sumary of technical analysis provided in this 4 section relies on the updated 1985 analysis. The major differences between i

the two analyses, with respect to Mr. Frederick's tests will be explained at '

the end of this section. While the complete 1985 evaluation of leak rate l' tests at THI-2 is provided as Attachment 2 to this enclosure, a sumary of '

the conclusions regarding Mr. Frederick is provided here to help th'e reader  ;

understand the basis for the questions posed to him during his interview.

The test records show that of the 161 leak rate surveillance tests retained by the licensee during the period under investigation, Mr. Frederick was

involved in 25 of these tests. During the 25 tests, he was either the j individual who performed the surveillance test or.was the operator o'n watch i (i.e., the panel operator) who signed the CRO's Log for the period during

! which the test was conducted. The table below shows a breakdown of the j overall conclusions regarding Mr. Frederick's tests.

Breakdown of Leak Rate Tests Involving E. R. Frederick 1 ,

j Total number of tests on fi1e. . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . 25 ,

i Tests with no apparent prob 1am................................... 13 j Tests _ involving unstable plant conditions........................ 2 .

l Tests involving the use of an unstable make-up tank '

l l 1evel transmitter as input to the computer..................... 6

Water additi'ons that were not included in the calculation........ 2 i Water additions that were partially included in the calculation.. 1 ,

Tests indicating unidentified leakage in excess of the TS Limit.. 1, '

Tests involving hydrogen additions............................... 0  :

1  !

i l

l i

As can be'seen from the table above, 13 of the 25 tests involving l Mr. Frederick appear to have been conducted in accordance with the requirements and precautions of the surveillance procedure governing the ,

conduct of leak rate tests (SP 2301-3D1). The remaining 12 tests, while -

! yielding questionable or perhaps invalid results do not indicate a pattern of  ;

leak rate test manipulation. The basis for this conclusion is presented  ;

below. .

Two tests, #10 (10/14/78) and #108 (02/07/79) were performed when the plant  ;

was not in a steady state condition. SP 2301-3D1 directs that the test be performed 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 operation. The procedure also cautions the operator to maintain the RCS and make-up system in a steady j state condition during the test by avoiding changes in valve line-ups, i coolers-in-service, pumps-in-service, etc. Power level changes should be minimized and the operator should avoid additions er removal of water from  !

the RCS and make-up system during the test. For the most accurate t i determination of the RCS leak rate, the initial and final conditions of .

reactor power, RCS temperr:ure, pressure and pressurizer level should be - l

identical. Test #10 was performed during the middle of reactor start-up f l following a reactor trip when the plant was not in a steady state condition. ,

l Test #108 was actually performed by Shift C, not Shift A.' The test was .

completed by Shift C eight minutes prior to watch relief and apparently not  :

O signed before shift turnover. Mr. Frederick signed the surveillance sheet as the individual performing the test. This test was also performed during unstable plant conditions caused either by changes in power level or by feed  ;

and bleed operations. These tests, while in violation of the limits and precautions of SP 2301-301 do not represent the type of operation that would

be used to manipulate leak rate test'results. .

j During the months of December 1978 and January 1979, one of the two level transmitters that provide make-up tank (MUT) level indication had a very i erratic and unreliable cutput. The output of one of the level transmitters would drive the MUT strip chart recorder in the control room while the other

level transmitter would provide MUT level indication to the plant computer.

The plant computer was used to run RCS leak rate surveillance tests at THI-2.  !

A selector switch in the control room allowed the operator to switch level transmitters feeding the strip chart recorder and the plant computer. When <

j the selector switch was selected to LT-1 the strip chart recorder would be l driven by LT-1 and LT-2 would provide input to the plant computer. When the

! selector switch was changed to LT-2 the strip chart recorder would be driven

~

by LT-2 and the computer would receive its MUT level indication from LT-1.

Because of the erratic nature of the output of LT-1 during this period, any i . leak rate test that was performed with the use of LT-1 providing input to the computer must be considered questionable. 0f the 54 tests conducted by all ,

1 shifts during this period only 16 were performed with the stable level i transmittee (LT-2) providing input to t1e computer during the test. During l this period, Mr. Frederick was involved in 7 tests, 6 of which used' unstable i LT-1 as input to the computer. These tests are: #64 (12/16/78), #73 1

, (12/23/78), #80 (12/30/78), #83 (01/02/79), #84 (01/04/79) and #85 l (01/05/79). Tests #64 and #85 do not appear to have been affected to a I I

i t

r

---,---------,n,-,,~n.----..-,n-n, _ ,.~...- ~ n n, ~...,-m-.-n - , - ~ ~ , . . - ~~.--v,,~m----w----

significant degree by the use of LT-1; however, the other four tests do show

, significant-differences between the output of LT-1 and LT-2. Because of the erratic nature of LT-1, the use of LT-1 as input to the computer would 'not -

provide assurance to an operator-that actual-unidentified reactop coolant system leak rate was less than the lechnical Specification limit of 1.0 gpm.

! Two tests, #12 (10/15/78).and #98 (02/02/79) involve water additions that were not included in the test calculations'." Test #12 involved the addition of 150 gallons of water near the beginning of the test and a large feed and bleed operation of approximately 600 gallons near the end of the test.

Neither operation is logged in the CR0 Leg or included in the test calculation. Mr. Frederick was the individual who had the panel during this

test, Mr. Germer, a CR0 trainee, was the individual who signed the surveillance sheet as performing the test. Test #98 was begun on Shift B and I completed on Shift A. The test contains a possible water addition of 110 gallons about 6 minutes before the end.of the test (Shift A). No log entry l was made in the CRO's Log and no water addition was included in the ,

computation. For this test it is notclear whether or not water was actually  !

added during the test, since the slope of the MUT level rise does not exhibit i the " normal" characteristics of a water addition. The evaluation relies on the offset in slope of the MUT level indication between the beginning and end 1

O of he test. Changes in pressurizer level and Tave do not account for the offset. Thus, this test is not conclusive from the standpoint of high confidence in the analysis. ,

One test, #149 (03/16/79) involves two water additions during the test. The actual change in MUT level indicates two additions were F.ade of 120 gallons each. The CRO's Log shows one addition of 200 gallons. The amount of water included in the computer calculation is 204 gallons. Mr, Frederick had the panel during this test and most likely made the wai.ar additions while Mr. Faust was the individual who performed the leak rate test. Since Mr. Frederick and Mr. Faust were involved in four other tests during the month of March 1979 where water was added and properly accounted for, this test appears to be an isolated incident,- It is believed that either the operators' interpretation of the' rise in MUT level indication (30 gal./in.)

was incorrect or the operator (Mr. Faust) intended to type "240" into the computer and transposed the numbers and actually typed in "204." Thus, this test is considered to be a simple human error.

4 Finally, one test, #12A (10/16/79) performed while Shift A was on watch, was ~

an unsigned, unapproved test indicating unidentified leakape of 2.56 gpm, which.is in excess of the 1.0 gpm Technical Specification imit for . ,

unidentified leakage. No water or hydrogen additions were made during this test that would influence the outcome of the test. This test was not filed with the other surveillance tests maintained by the licensee. Instead, this test was fTled with a set of tests that lead up to the submission of LER 78-62/1T on November 1, 1978. The significance of this particular LER is O

discussed during the interview with Mr. Frederick (See Transcript pages 58 -

through 64).

l 1

e

'1--g- e- p- - - -- ----y, e., - wy -+--m-ww vg---h w-,- --py-- m#mm,---,;w..-y,g-3 --- .-,.--a-w.=--e,1ew-enweWeur--we -DF-'wrem-ew---w--* drie w ey-- w o c--*- - -,e-er l

. As discussed in the beginning of this section, when Mr. Frederick was inter-viewed in. January 1985, he was questioned prior to NRC's updated analysis of the 161 leak rate tests. The earlier analysis included two tests where it appeared hydrogan was added during the test. The 1985 analysis concluded __-

that the spikes were not caused by hydrogen additions but by switching level transmitters between LT-2 and LT-1. The 1985 analysis considers leak rate tests performed during December 1978 and January 1979, which used unstable LT-1 as input to the computer as " questionable." Prior analysis included these types of tests in the " questionable" category only if it appeared that -

i water and/or hydrogen additions were made during the tests. Thus, Mr.

Frederick was not questioned during his interview regarding his knowledge of using the unstable level transmitter as input to the computer. Finally, the unsigned, unapproved leak rate test (#12A) associated with LER 78-62/1T was

] not included in the package of leak rate tests discussed with Mr. Frederick.

In sumary, the overall results of the NRC analysis for D0J used to question Mr. Frederick and the revised 1985 analysis support a finding that Mr. Frederick did not intentionally manipulate leak rate surveillance tests by the addition of water and/or hydrogen to the MUT during the period investi-gated. Although some of the tests contain elements that call the validity of the results into question, such as the tests performed using LT-1 as input to the computer during its unstable period, there is no pattern of manipulation, as is evident on some of the other shifts. This is particularly significant l during the months of-February and March 1979, when identified leakage was high, resulting in it becoming trore difficult to get acceptable test results for unidentified leakage. During this two month period, only 17 of the 60 tests retained, that were conducted by all.six shifts, appear to have been properly accomplished in accordance with SP 2301-3D1. Mr. Frederick was involved in eight of the 17 satisfactory tests.

III. Interview Summary

! During the period under investigation, September 30, 1978 through March 28, 1979, Mr. Frederick indicated that he was primarily assigned to "A" Shift.

The licensed members of his shift included: Shift Supervisor, Bill Zewe; Shift Foreman, Fred Scheimann; and Control Room Operators (CR0s) Ed Frederick and Craig Faust. See page 4.

As a CRO, Frederick stated that his job was to manipulate the reactor controls and maintain records of the shift operations. He reported directly l to the Shift Foreman. They were in contact throughout the shift. The Shift Supervisor would supervise the "A" Shift on Unit as well, thus, the Shift Supervisor would split his time between the two plants. Consequently, Frederick would see Zewe less often. See page 5.

O-

-_- _ _ _ _ _ ___ o

i 1

l l

l

! Frederick stated that he did not have routine contact with the Supervisor of Operations at the time, J. Floyd. Mr. Floyd would come into the control room  ;

on day shifts; however, he did not recall discussing RCS leak rate __

surveillance testing with Floyd. His principal contact with the Plant Superintendent, J. Logan, was in a training capacity. Mr. Logan was trying to get a license on TMI-2, so Frederick had given him some system checkouts and showed him around the plant a few times. He did not recall discussing leak rate testing with Logan. Mr. Frederick stated that his contact with the Station Manager, G. Miller was even more limited. He said he seldom talked to Miller. See pages 6-7.

Mr. Frederick was asked to review a copy of TMI-2 Technical Specification

~

3.4.6.2 " Reactor Coolant System Operational Leakage." and a copy of TMI-2 Surveillance Procedure 2301-3D1, " Reactor Coolant System Inventory." After reviewing the documents, Frederick stated that the test was required to be run at least once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. That was the minimum number of times. He did recall that the test was conducted more often. He stated that he thought it was done daily and not once per shift. He believed if it was run successfully once per day that was good enough. See page 8.

Frederick stated that when he ran a leak rate test, he would give the results l O to the Shift Foreman whether they were acceptable or unacceptable. Then the Shift Foreman would discuss with the CR0s what action was required based upon the results of the test. See page 9.

With respect to follow-up actions Frederick said that in certain cases they would start looking for the source of the. leakage if the results were in excess of the Technical Specification limit. It was usually the Auxiliary Operators (A0s) that would do that. In addition, the A0s may have been able to readily identify something that they had changed in the plant that changed the leak rate. Frederick' stated that when a test result exceeded the limits specified in the Technical Specifications, he considered that they were required to enter the Action Statement of Technical Specification 3.4.6.2 provided the test was valid. See page 9.

NOTE: Part b of the Action Statement reads as follows: "With any Reactor Coolant System leakage greater than any one of the above limits, excluding PRESSURE BOUNDARY LEAKAGE, reduce the leakage rate 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 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 SHUTDOWN within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />."

NOTE: During an interview with Mr. Craig Faust, earlier that day, Mr. Faust stated that the Technical Specifications required the test to be run once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> but they ran them

" more frequently. If a bad leak rate test result (i.e.,

6 unidentified leakage. greater than the TS limits of 1.0 gpm) was received, they would throw the bad test away. As long as a good result was obtained once during the 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, it did not matter hcw many bad tests were run, they did not have to enter the Action Statement of the TS.

l  ;

i Mr. Frecesick reiterated that he turned in all bad leak rate tests to his i

Shift Foreman. He was not aware whether the Shift Foreman eventually threw '

l them aw'yT_ a He stated that they could have been. He did not personally recall ev.er throwing away test results. If it was bad, he would discuss it -' .

with the Shift Foreman and then would run another test. He stated that he did not recall Faust throwing tests away. He stated, "We usually did things pretty much the same way." See page 11.

. 1 i' Mr. Frederick also recalled that he did not have difficulty getting l acceptable leak rate test results. Frederick said the tests he had problems -

with, he was usually able to find a reason why it was invalid or why th.e

_ number was not what he expected it to be. He stated he could not recall periods where he would have to run several tests in a row before he could t finally get a good one. See page 11.

At this point in the interview, it was pointed out to Mr. Frederick that his testimony did not appear to match the testimony given earlier that day by the other CR0 on his shift. Mr. Faust. It was noted that there were only two CR0s on their shift and they were both in the control room at the same time every shift, yet from the differences in testimony between Mr. Faust and Mr. Frederick one would conclude that they must have been on different shifts. Three of these differences were discussed with Mr. Frederick. See page 12.

First, Faust's testimony was consistent with the testimony of most operators  ;

',' regarding the frequency of running leak rate tests. According to Mr. Faust leak rate tests were perfonned normally every shift, subject to plant l conditions and the availability of the plant computer. Frederick stated that ,

Mr. Frederick responded that leak rate '

tests tests were were done done on only theonce per day (.2300 to 0700).

mid-shift If it was done on that shift, he did not recall performing the test again that day. If it was not performed on the mid-shift it would be done during the day shift (0700-1500).

See pages 12-13.

NOTE: Since only good leak rate test results were retained by the licensee, the actual frequency that tests were run  ;

is not known. However, based upon the 161 tests that 1 were retained 71 (44%) were done on the mid-shift, 37 (23%) were done on the day shift, cnd 53 (33%) were performed on the swing-shift. (1500-2300). The record also shows that 29% of the time more than one good leak I rate test exists for the same day. Approximately half (50%) of the time only one good leak rate test exists for the same day and for 21% of the time, no leak rate

~

test exists for that day, excluding periods when the plant was in Mode 5 (i.e., cold shutdown).

Of the 25 tests involving Mr. Frederick, 14 were done O -

.on the mid-shift. 9 on the swing-shift and 2 on the day-shift. Of the 11 tests performed on either the swing or day shifts, 6 times an acceptable leak rate test i

l l

O '

had already been conducted earlier in the day. 'In one

~

case where Mr. Frederick signed the surveillance

- - sheet as performing the test, 2 acceptable leak rate

-tests had already been conducted that day.

Second, Mr. Faust test 4fied that bad test results were thrown away.

Mr. Faust's testimony again is consistent with the vast majority of other operators' testimony. Faust stated that if he could not determine why a test was invalid, then he would turn it in to the Shift Foreman. Mr. Frederick, on the other hand, stated that he never threw out a bad or invalid leak rate test. Frederick responded by stating: "Well, the details of the paper flow

. might just not be there. I don't know. I would leave them on his desk, the Foreman's-desk, to look at, or I would hand them to him personally and then we would discuss what to do with them. Maybe Craig (Faust) didn't do it that way. Maybe he made the decisions on his own." Frederick continued by saying i that if they were being thrown away, he either did not count it as signifi-cant and didn't remember it, or he was not aware they were being thrown away.

He did not recall either himself throwing any test results or he did not recall seeing other people throw them away. He said he is aware now that it was comen practice, but he does not believe he was aware of it then. See pages 13-15.

i~ Third, Faust stated that leak rate test results were hard to get, especially as the date of the accident approached. Frederick stated they were not hard to get. Frederick responded by saying that maybe he defined "hard to get" differently. "If the test didn't come out, I didn't consider that a difficulty or a problem to get upset about. I would just either run another one or just not run anymore. It was not a difficulty. It was a routine thing." Frederick was then asked if he could recall having to run one test after another and still not have a good leak rate test result by the end of his shift. Frederick stated that he did not recall, but it would not surprise him if that was the case several times. See pages 16-17.

l Frederick said that he did not concern himself much with. leak rates. He said l that Faust had gotten into the habit of putting a leak rate into the computer whenever they came into the control room to relieve the shift. He said that if that one came cut okay, that was all they would do for the night or during the shift. He said that he had gotten out of the habit of running the tests because of Faust's routine.- He said he would only run a test if it was his turn to take the readings and Faust had not already started a test. See -

pages 17-18.

NOTE: The routine described by Frederick appears to con-tradict his earlier statements that he only recalled the test being run once per day.on the mid-shift.

_ ,In addition, Mr. Frederick and Mr. Faust were involved in the performance of 19 tests together. Mr. Frederick l signed the surveillance test sheet as the individual O

performing the test 11 of the 19 times. This fact -

appears to conflict with Mr. Frederick's testimony that he had gotten out of the habit of running leak rate i

tests because of Faust's routine.

s I

Frederick-was next questioned about the requirement to log all leak rate tests in the CRO's Log. He was shown a copy of Administrative Procedure 1012 " Shift Relief and Log Entries." Paragraph 3.3.17 states that the operator shall " Record title and number of the test perfomed, and the start .

and completion times or time of suspension of the test. The performance of ~

all periodic tests and inspections required by the Technical Specifications shall be recorded." Frederick stated that he did not log any of the data associated with leak rate tests. He cons.idered this test routine and did not ,

log evolutions of a routine nature. Frederick believed that another procedure AP-1001 stated that items of routine nature did not need to be logged. A copy of AP-1001 was not available to verify his claim. See pages 19-20.

NOTE: None of the six shifts logged all of the data required by AP-1012 in the CR0's Log; however, all other shifts, except Shift A routinely logged the completion time and the results of " good" leak rate tests. Leak rate tests t; hat did not meet the acceptance criteria were not logged.

Shift A did not log any leak rate test results, good or bad.

Mr. Frederick was also shown a copy of Administrative Procedure 1010, Technical Specification Surveillance Program." It was pointed out to O Mr. Frederick that Section 6.5 of the procedure required that problems encountered while performing surveillance testing shall be recorded on an

" Exception and Deficiency List" and that this list, when required, was to be attached to the surveillance test data sheets. Mr. Frederick stated that this' procedure was not applied to leak rate tests. He does not see how it could have been excluded, however, except by practice. He said.that if someone would have brought it to his attention at the time, he would have filled out the required form. See pages 21-23.

Next, Mr. Frederick described how he perfomed a leak rate test. He stated that the test was done completely by computer. You had to establish the

initial conditions for the test such as steady state, no power changes, etc.

Once you established the initial conditions you just typed the right code into the computer and it would collect the data at the beginning of the test and one hour later at the end of the test. He said that it was their practice to inform the other CR0 when the test was in progress so that they

~

could maintain steady state conditions in the plant so that the test would be as valid and as accurate as they could make it. Stee,pages 24-25.

Mr. Frederick was asked if he ever came up with an invalid test because of an operator-induced change. He stated that that did occur. Usually when it happened it was because of a water inventory change that was not accounted for in the calculation. He said if he ran the test, he would make a notation as to why he thought ~1t was invalid, or signed it, and gave it to the Shift .,

Foreman atd explained what happened. Again, he stated that he did not know if these tests were thrown out. He suspected that if they were invalid and he could prove it to the Shift Foreman then they were thrown away. Frederick believes that these invalid tests were discussed on several occasions with the Shift Supervisor, Bill Zewe. See pages 25-27.

- - _ __m._ m_. n -m v_-. , _ . ,, y. . . , , _ ,,,_a ,.,,.,m, , , , , . , _ ,

. , _ . . , _ _ , , , ,, , , , , , . y _, , ,. ,,_ ,, ___ _ _ _

l When askef if he had ever had a discussion with his Shift Foreman or his Shift Supervisor regarding how these invalid tests should be handled, he stated that-he could not recall his Shift Foreman telling him that the way he was handling them wasn't the way he wanted him to do it. Thus Frederick - '

believes that his Shift Foreman must have approved of the way he was doing it. See page 27.

Frederick was asked why Faust recalle'd throwing away tests and he did not.

Frederick stated, "It just may be that when Fred (Scheimann) talked with

- Craig,(Faust) about invalid leak rates, it was Craig that turned around and threw them in the wastebasket rather than Fred." See page 27.

When Frederick got a bad leak rate test result, he stated that he could

. usually tell why. He said that he doesn't know why he did not have as much trouble with the test, as apparently others did. He recalled that he spent a lot of time trying to establish initial conditions for the test before he would start it. He believes that that may be one reason why he had better results than others. He stated that he had vague memories about one or two He said operators that formerwho CRO's complained H. Hartman about andsurveillance testing')in R. Booher (Shift F weregeneral.

the individuals .

that he recalled. He remembered them because Shift F relieved their shift O quite often. Frederick stated that he did not recall either of them ever mentioning that they were having problems to the point where they would have to do things to influence the outcome of the test. See pages 30-31.

When Frederick was asked if he believed the information that he received from

! the computer with respect to the accuracy of the leak rate test, Frederick i

stated that he questioned all the results that were received. He felt that i

the test tolerance was to limited for the accuracy of the instruments involved. He claimed that when he ran a leak rate test, he could usually predict what it should be. If it wasn't what he expected, he would qdestion it. He said that he examined each one carefully and if he understood it, then it was alright. See page 32.

NOTE: As stated in the Sumary of Technical Analysis, Mr. Frederick was not questioned regarding his know- ~

ledge of the unstable HUT level transmitter (LT-1).

However, based upon Mr. Frederick's claim that he examined each leak rate test carefully and questioned any results that were not what he expected, it is hard to understand how 8 of his 25 tests involved either unstable plant conditions (2) or the use of unstable LT-1 as input to the computer (6).

Mr. Frederick confirmed that his understanding of Technical Specification 3.4.6.2 wat that if he ran a leak rate test and the results came out greater i

than one gpm for unidentified le'akage, he would enter the Action Statement of O the Technical Specification, provided he believed the test was valid. He stated that he would run another leak rate test immediately to confirm the first test. See pages 35-36. Frederick was asked if he had ever run a test

, . . - - - . . .- - . , - . - , - - . , , . ~ . - - - - - .., . , - - . - - . - .-----.,,._-,,,,_--,.,..__,.-,,,,--n- -,.-.,,.n,,,. -,,a-,. nw -- --,---w.,- , . -w-----

___ .____-_. _ =_ ._- . __ _ . - _. _ _ _ _ _ _

1 l ,

and the results were greater than the Technical Specification limit and at the end of four hours he still had not gotten a leak rate test result that was within the acceptance criteria. Mr. Frederick stated that he could not recall such a situation. It was pointed out to Mr. Frederick, that Mr. Fausr had testified earlier that at times their shift had gone through an entire shift without getting a leak rate test result within the limits of the Technical Specifications. See,page 37. ,

1 In discussing the surveillance frequency and the requirements of the Action  !

Statement, Mr. Frederick stated that the Technical Specifications required l that the test be run at least every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. However, he believes the test ,

was run once a day? He stated that "everybody seems to be saying that they j did them every shift, therefore, they had to get a correct one every shift. l Perhaps I avoided all of that because I only did them when the surveillance

. either called for it or someone made a special request." See pages 37-41. 1 Mr. Frederick was then shown a document which showed graphically, when leak rate tests were run. The document showed that on several occasions Shift A did run leak rate tests on the same day that another shift had gotten an acceptable leak rate test earlier in the day. Mr. Frederick agreed that it appeared tests were run more than once a day. See page 43.-

O Frederick was asked if he had ever gotten into a discussion with Mr. Faust regarding the interpretation of Technical Specification 3.4.6.2. Frederick did not recall any specific discussions but that was one of the topics that he enjoyed talking about to everybody. He stated that it's a comon problem in the industry to try and get a single interpretation from every Technical Specification for all the people that are using it. At TMI-2 the Technical Specifications were as open to interpretation as anywhere else. See pages 48-49. He also could not recall any specific conversations with h7s Shift Foreman regarding his interpretation of the Technical Specification. He believes that because of differences in interpretation, he developed his routine of performing the test, discussing it with his Shift Foreman and allowing the Foreman to make a decision. Frederick stated that he could make a suggestion and interpret the data for the Shift Foreman, but unless Frederick vehemently disagreed with him or his decision on the action necessary, it was up to the Foreman. See pages 49-50.

I In describing how he would review leak rate test results, Frederick said that he would look at many things. He stated that if something occurred which i changed the steady state condition, he would imediately suspect the results of the test whether it came out less than one or greater than one. Frederick i did not consider it difficult to maintain steady state conditions on the back shifts, with the exception of feedwater oscillations that he stated were caused by the Integrated Control System (ICS) tolerances. See pages 50-51. j Occasionally, Frederick stated, he would get negative numbers. He said this

. could happen if'the unidentified leakage was close to zero, just due to instrument tolerances. He stated that if something happened to change identified leakage and you were not aware of it, that could also cause a negative leak rate. See page 51.

1 Frederick'itated that if he found out that water was added during a leak rate test, he w_ald normally not include the amount of water added in the calculation. Instead, he would just start a new test. He stated that even with this type of situation, he still turned in the invalid test to his -

Foreman. He did not recall throwing any of these tests away. See pages 52-54.

It was pointed out to Mr. Frederick that ihost tests conducted after 1

mid-February 1979, contained water additions during the test, including tests performed by Shift A. Frederick stated that was true because they changed their mode of operation because of the suspected leakage through one of the code safety valves on the top of the pressurizer. Frederick stated that they had to add a lot of water to make-up for lost inventory. See pages 52-53.

Frederick was then shown a copy of LER 78-62/1T. Frederick indicated that he had seen it several times before; however, the first time he was shown the document after the accident, he could not recall seeing it prior to the accident. Because he did not recall seeing it prior to the accident, although he had initialed the review sheet, he stated that the LER had no  ;

impact on the way he performed leak rate tests. See pages 58-59. J NOTE: LER 78-62/1T dated November 1, 1978, dealt with an incident O involving leak rate tests and the " misinterpretation" of the requirements of the Technical Specification 3.4.6.2 (i.e., the requirement to enter the Action Statement when a leak rate test exceeds the acceptance' criteria, regardless if the test was performed more frequently that the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> requirement levied by the Technical Specification).

Frederick stated that since he was inter)reting the Technical Specification correctly at the time of the incident, t1e LER did not alter his performance of leak rate tests. He stated, at that time, he believed that if you performed a leak rate test and the results were greater than the one gpm

, limit, you were bound by the results, unless the test was invalid for some reason. He stated that, "if you believe the result was greater than one, that was cause for some alarm. If you really believed it, you had to act upon it." See pages 61-63. -

In discussing hydrogen additions, Frederick stated that prior to the accident, he was nst aware that hydrogen could affect MUT level and consequently leak rate test results. According to Frederick, he does not think that he and Faust discussed the hydrogen effect. He stated that after the accident, when he heard about the Hartman allegations, he spent some time investigating whether or not he thought the effect was real. Frederick concluded that it l was not. He said that he examined the prints and diagrams of the MUT and i examined the effect of hydrogen additions on the MUT strip charts. It did not appear to Frederick that hydrogen additions would produce any type of consistent result. Frederick was informed that Faust testified earlier in O

the day that Faust was aware that hydrogen aeditions were alleged to be capable of influencing leak rate test results and that he-had a discussion l

l

with other operators about it prior to the accident. Frederick reiterated

that he-and Faust did not discuss it amongst themselves. He continued that he was not-aware that hydrogen could be used to alter leak rate test results.

He did not use hydrogen for the purpose of affecting leak rate tests and he -

did not recall ever entering into any discussions with other operators or '

supervisors as to the potential impact hydrogen could possibly have on leak rate test results. He said that it may have been possible that he talked with someone about it; however, he does not recall. See pages 68-75.

In discussing water additions, Frederick stated that water was routinely ~

added to change boron concentration or to make-up for lost inventory. '

Demineralized (DI) water could be added or else borated water could be added from the Boric Acid Mix Tank (BAMT), or the Reactor Coolant Bleed Tanks (RCBTs). Water would be pumped into the make-up tank and then transferred into the RCS via the make-up pumps. Frederick stated that when adding water. l he would use the totalizer on the batch controller to know how much water he '

added. Although, if he were adding water from the Boric Acid Mix Tank he could also use the stroke counter on the BAMT pumps, since they were positive

. displacement pumps. See pages 74-77.

When asked if he ever used the MUT strip chart to determine how much water was added, he said that he did not. He said that since water was being O transferred out of the MUT at the same time water was being added, you would not get an accurate answer unless you stopped the make-up pumps. See pages 77-78.

In adding water, Frederick stated that it was possible to control or throttle the rate of flow using MU-V-9 in the line going to the MUT. Frederick thought that it would be unusual to throttle MU-V-9 when adding water from the RCBT or when adding DI water. He said that you may want to add water in smaller batches, but he couldn't think of any reason to go to an intermediate position on MU-V-9. See page 80.

Frederick stated that the batch controller contained a current batch counter, a totalizer and an integrator. If he wanted to add 600 gallons in 200 gallon batches, he would set the maximum batch to 600 and the current batch to 200.

The water would start being added until the counter reached 200, at which time the water addition would automatically stop. The counter would then show that 200 gallons of the 600 gallons had been added. Frederick believed I that you would have to manually open and close MU-V-9 after each batch. See page 80.

Frederick believed that all water additions were logged in the CR0 Log. He stated that he was familiar with Hartman's. allegations that water was added and not recorded or underrecorded during leak rate tests. However, Frederick stated thaj he did not recall anyone telling him that they did it. He stated he had no knowledge at that time that that practice was being done. He O stated that if he personally failed to record a water addition it was a mistake. He did not do so with the intent of trying to change the leak rate

result.. Ffederick also stated that he never underrecorded a water addition for the same purpose. See pages 81-83.

~ '

The staff's overall copclusions regarding the individual leak rate tests involving Frederick were reviewed with him. It was pointed out that there were relatively few instances where water was either added and not included in the calculation or the amount included.in the calculation was not as large as the apparent amount added from looking at the MUT strip chart. The technical analysis supports Mr. Frederick's testimony that he does not appear to have been involved in the addition of either hydrogen or water to the MUT in an attempt to influence leak rate tests. See pages 84-98.

In explaining why he may have had better success than others, he stated that it seemed that if you set up the test right and took your time about getting the plant oscillations that were going on out of the way, it seemed to work out. In addition, Frederick believes that he had a good grasp of integrated plant response. That was one of the reasons he said he was picked as an instructor. See page 103-106.

IV. Conclusions l

Mr. Frederick denied that he personally attempted to influence the outcome of O 1eak rate test results by the addition of water and/or hydrogen to the MUT during the tests. His testimony regarding his lack of personal involvement in leak rate test manipulation or falsification is consistent with NRR's l

l i

technical analysis. Mr. Frederick also denied that he had personal knowledge i of cr was aware of other operators as intentionally manipulating leak rate I surveillance tests. However, many of Mr. Frederick's responses to questions concerning preaccident reactor coolant system leak rate surveillance testing at THI-2 were inconsistent with the testimony given by the majority of other

. CRO's interviewed to date. .His testimony is also inconsistent in part with the testimony given earlier the same day by Mr. Faust, the other CR0 on shift with Mr. Frederick during the six month period under investigation. The majority of conflicts deal with: the number of times the test was conducted; the difficulty in obtaining test results that were within the allowable limits of the Technical Specifications; what was done with test results that were unacceptable; and interpretation and compliance with the Technical Specifications and Limiting Condition for Operation associated with RCS leakage. ,

During the interview, Mr. Frederick provided the following information:

1. Although the TMI-2 Technical Specifications required that the RCS leak l rate surveillance test be conducted once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, Mr. Frederick recalled that the test was conducted daily, usually on the mid-shift.

If a successful test result was obtained on that watch, Mr. Frederick stated that he would not run it again that day. Note: The majority of

operators interviewed have stated that the test was conducted once per shift, plant conditions permitting and provided the plant computer was i

l l

available. Of the 25 tests involving Mr. Frederick during the period under investigation, 14 were done on the mid-shift, 9 on the swing-shift and 2 on the day-shift. Of the 11 tests performed on the day- or - ,

swing-shift, 6 times an acceptable test had been conducted earlier in the day. In one case 2 acceptable tests had already been conducted that same day. The difference between Mr. Frederick's recollection of '

test frequency and actual test frequency may be attributed to difficulty of recall with the' passage of time since the difference between daily 1 testing and testing each shift if plant conditions are stable is not a significant difference in the number of tests performed. This is T e to the high frequency of unstable plant conditions (e.g., start-ups, shutdowns, power changes, etc.) between October 1978 and March 1979.

2. Mr. Frederick stated that to the best of his recollection, he never threw away any leak rate surveillance test results. He said that all lake rate tests he ran were turned over to the Shift Foreman for his review. He did not know if they were subsequently thrown away or not.

Note:. The majority of operators interviewed have indicated that invalid or unacceptable test results were routinely discarded. In a March 24, 1984 OI interview with Mr. Frederick's Shift Supervisor, Mr. Zewe, stated that the discarding of invalid tests was accepted practice at the time. Mr. Faust also stated in his interview that invalid tests were routinely discarded. Thus, while it is possible that Mr. Frederick did turn all leak rate tests over to his Shift Foreman for review, it does not seem credible that Mr. Frederick would not know that test results showing greater than 1.0 gpm for unidentified leakage were routinely thrown away.

Mr. Frederick stated that he interpreted Technical Specification 3.4.6.2

~

3.

as requiring the operator to enter the Action Statement of the Technical Specification whenever they received the results of a leak rate test that exceeded the Limiting Condition for Operation, unless he could explain why the test result was invalid. Note: The majority of operators interviewed including Mr. Faust,"TeTieved that they were not required to enter the Action Statement, no matter how many bad leak rate tests were received, as long as they had not exceeded the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> required test interval without obtaining a successful test result.

While Mr. Frederick's interpretation of the requirements of the

. Technical Specification is correct, there is indication that Mr. Frederick did not comply with this' interpretation between October 1978 and March 1979. For example, he testified that he would not be surprised if there were times when they ran a leak rate test on their shift, obtained an unacceptable result and by the end of the shift they still had not gotten an acceptable result. Since the Action Statement

. . requires that 'eakage be reduced within accestable limits within 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />s-or be in at least hot standby within tie next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />. Had Mr. Frederick followed his interpretation of the requirements of the O Technical , Specification, as he stated, there would have been periods of time where the plant should have been shut down.

i

4. Mr. Frederick also testified that he did not recall having difficulty

, getting leak rate test results within the Technical Specification limits i

and that if he had any problems he usually knew why. Note: The '

majority of operators interviewed, including Mr. Faust, Indicated that leak rate test results were difficult to get and that as the date of the accident approached, it became more difficult to get an acceptable test i result. Mr. Frederick's statements regarding the difficulty of getting i test results also appears to conflict with his later testimony that it would not surprise him if there were several times where they were not

. able to get an acceptable test result their entire shift.

5. Mr. Frederick testified that he did not log the results of leak rate I tests in the CRO's Log for any tests (i.e., good or bad) because items l of a routine nature were not required to be logged. Note:  !'

Administrative Procedure 1012 requires that the test YTETe, test number, and the start and completion times of all surveillance tests required by the Technical Specifications be logged in the CRO's Log. Shift "A" was the only shift that did not routinely log satisfactory test results.

None of the six shifts logged unacceptable test results.

6. Administrative Procedure 1010 required that problems encountered while performing surveillance testing should be documented on an " Exception and Deficiency List." The list was to be attached to the surveillance test data sheet. Mr. Frederick stated that AP-1010 was not applied to leak rate tests, although he did not know why it had been excluded.

Note

Fone of the six shifts applied AP-1010 requirements to leak rate surveillance testing.

7. Although he vaguely recalled two other operators (Mr. Hartman and Mr. Booher) complaining about the difficulty of surveillance testing in general, Frederick stated that he personally had no problems running the test. He attributed this to the fact he was very careful in setting up the required initial conditions for the test. He stated that he ~

questioned all leak rate tests that were received. He stated that he t

could usually predict what it should be. If it wasn't what he expected, he would question it. He examined each test result carefully and if he understood it, then it was alright. Note: These statements do not coincide with the results of the technical analysis which shows that two tests were conducted during unstable plant conditions and six tests were

! performed using the unstable level transmitter (LT-1) as input to'the computer for MUT level indication. In four of these tests of the cases, there was a significant difference between the output of LT-1 and LT-2.

Had Mr. Frederick revieweif the data carefully, as he testified, he should have seen these discrepancies.

~

8. FredeN ck stated that LER 78-62/1T neither altered his interpretation of the Technical Specification 3.4.6.2 nor did it change the way he l performed the test or handled the test results. Frederick stated that he did not have a preaccident recall of the LER.

_ _ _ , _ _ , _ _ _ _ _ , _ . , _ . , _ _ . , _ _ _ , _ . , _ , _ -,, _,.,_..._,_.,_.,_,__.....,,._..,__,._m -

l

9. With respect to hydrogen additions, Frederick stated that he was not aware prior to the accident that hydrogen additions to the MUT could

' affect MUT' level indication and consequently leak rate test results. He ~

I did not recall ever discussing the effect with anyone or knowing that individuals were adding hydrogen in an attempt to influence the test

results. Although Frederick was told that Mr. Faust stated he (Faust) was aware prior to the accident that. hydrogen additions were alleged tas alter the test results. Mr. Frederick does not recall discussing the subject with Mr. Faust or any other operator. Mr. Frederick's testimony that he did not add hydrogen to manipulate leak rate test results is consistent with NRR's technical analysis of his leak rate tests.

]

i

10. With respect to water additions, Frederick stated that he never intentionally added water and not properly included the amount in the leak rate calculation in an attempt to influence the leak rate test results. He also stated that he was not aware of other operators manipulating tests in this manner. Note: Although, water additions, that were n'ot completely accounted foT Tn the test calculation, occurred during three of Mr. Frederick's tests, there are credible reasons why that may have accidently occurred. In addition..the number of tests performed by Mr. Frederick, during the high identified leakage period of February and March 1979, where water was added and properly compensated O. for, supports Mr. Frederick's testimony.

In summary, the technical analysis supports Mr. Frederick's testimony that he did not intentiona11y add water and/or hydrogen to the MUT during the leak i

rate tests for the purpose of manipulating the test results. However, other aspects of his testlinuny dealing with difficulty lii obtaining satisfactory test results, the disposition of unsatisfactory results, and his overall knowledge of leak rate testing irregularities raise questions regarding the ,

accuracy of his responses during the interview. t 1

Emns O

- __