ML20039D729

Response to Licensee Comments on HPI Question & Answers on NRC Oct 1981 Exams

ML20039D729
Person / Time
Site:	Three Mile Island
Issue date:	12/30/1981
From:	Aamodt N AAMODTS
To:
Shared Package
ML20039D726	List: ML20039D725 ML20039D729
References
NUDOCS 8201060065
Download: ML20039D729 (3)

Text

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -.

, A J. M -- ~ 12/33/81-f

' UNITED STATES OF. AMERICA NUCLEAR RECULATORY COMMISSION BEFORE THE ATOMIC SAFETY'AND LICENSING BOARD

)

In the Matter of )

METROPOLITAN EDISON COMPANY )- Docket 50-289 Restart-

)

(Three Mile Island Nuclear " -

_ Station, Unit No. l' )

AAMODT RESPONSE TO LICENSEE'S COMMENTS ON HPI QUESTION AND ANSWERS ON THE NRC OCTOBER EXAMINATIONS Li cen s e e ' s c o nm e'n t s under "A. Written Examinations" appear to support a position that the failure rate on the HPI question does not reflect a wide spread lack of under- ..

standing of the. basic throttling criteria listed in Section 6.B.3.ll.a of EP-1212-6B, i. e., the 500 s'ub co olin g

- - ~

_~..-- and- LPI ' flow criteria. For the sake of accuracy,_however,_r.

U: 5: .

Y'

' we would note that' three candidates-(C1, H2 ,~nd AA) failede

~

to reference one or more elements of these criteria as opposed to only one, AA, as suggested by Licensee.7_'

The' impact of EP-1202-4 on grading-is e v id e n't . Not fewer than fourteen candidates 3 made reference.to the 1600 psig.criterium of this procedure and were docked.4 Notifewer' -

1. C failed to list the 500 subcooling criteria.

2 H failed to list.the 1000 gpm per. leg criterie.

3 It should be noted that the HPI answers o f A(1. 7 grade _)r and - - -

F(1.8' grade) were.so_poorly reproduced th'at th ey a re ' no t1 in clud ed --

in this analysis.

4 Two by 0, three by 0.3, seven-by 0.5, one by 0. 8 and one by 1.0.1 hO

.- _ - _ _ _ __.. . . __ _..._ _ .__.~___. _ --

than five candidates who failed would have r2ceived a passing grade had their scores not been reduced. In addition, had reference to " going solid" been treated as suggested by Licensee, one more failing grade would have been a passing grade. In summary, if Licensce's assertions relative to the written examination had been available to and been acted upon by the SRC examiners, not nore than five candidates would have failed the HPI question. (This is not to imply that this is an acceptable level.)

It was evident earlier in this proceding that the NRC examiners and Licensee's training staff were functioning on different wavelengths with regard to questions'on the April examinations. Management capability in this area of communication capability when dealing with the Staff remains inadequate.

Licensee's assertions (Newton and others)5 to the contrary in the reopened hearing are challenged by this evidence.

47 7 There fore ,- the record should be reopened to examine this issue.

$~

d It should be'noted that' whereas Licensee appears-to so

, _ effectively' argue the adequacy of operator understanding-of .. -

HPI' throttling criteria, Licensee fails to take notice of the

~ ^

remarkable ~ improvement in candidates' performance on the B examination, where only three of sixteen failed the HPI question in contrast to eight .-o f seven t een on the A examination.

Applying the CHI Square test of si gn i fi can ce to these data yields a 0.09 level of confidence.

A level of confidence of 0.09 is not in itself over- -

whelming, but viewed in the context of two elements of the reopened hearing, the probability of such a second day 5

Transcript citations are missing since there has been a two week delay in receipt of transcripts of the reopened hearing by the Philadelphia Free Library.

e

-3 improvarent in grades being simply a matter of chance diminish substantially:

1. Licensee management was aware that this question was of particular interest to the Board. Did someone in management who was aware of the question on the A examination assume that this question would appear on the B examination and alert B examination candidates?

2. The unsupported, but unrefuted allegations related to managenent involvement in improprieties during the April examination.

The fact that overall the A and B groups showed no significant differences further enchances the likelihood that the marked difference in performance on the HPI question cannot be ascribed to chance. Since proctoring v.as by all reports more than adequate, management impropriety should not be precluded.

j] " The~~Aanodts note for the record that they have accepted Licensee's' arguments for the express ano singular purpose --

of point'ing out. possible Licensee management inadequaci es: c.

related to the issues of this proceding. No implic~ations- -

relative to the adequacy of the training and testing of the '

HPI question are to be inferred. Rather Licensee's recent change in the HPI training c*iteria might well be interpreted to relect on Licensee's inability.to establish firm training programs. .

Ruspectfully submitted, i

'/

' ' W A EV4'(shi{h

• kk- W ( f Norman:0. Aamodt-December 30, 1981