ML20006E855
| ML20006E855 | |
| Person / Time | |
|---|---|
| Issue date: | 02/13/1990 |
| From: | Miller H NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | Roe J Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 9002260434 | |
| Download: ML20006E855 (5) | |
Text
'
a.,
w N.'
'y
',(
- FEB ! 31333 MEMORANDUM FOR:
Jack W. Roe, Director, Division of Licensee Performance and Quality Evaluation, Office of Nuclear Reactor Regulation FROM:
Hubert J. Miller, Director, Division of Reactor Safety, Region III
.j
SUBJECT:
INDIVIDUAL / CREW SIMULATOR CRITICAL TASKS REVIEW 5
This memorandum is in response to Ken Perkins' request of January 17,.1990, to' review the draft revision to ES-601 dealing with Individual / Crew Simulator Critical Tasks (ISCTs).
I fully agree with the idea that a more concrete and tangible definition of what constitutes an ISCT is necessary and commend Mr. Perkins efforts in moving in that direction.
Further, I understand the objectives of requiring that one ISCT failure will result in failure of the requalification exam. However, I believe this approach will add unnecessary complications to an already complex exam process and would be impracticable in-r many cases.
I am concerned that this approach departs from the original intent of the dynamic simulator portion of the requalification exam which was principally to evaluate performance of operators as a crew.
It is important to identify
-certain critical steps or elements that, if not performed properly by crew members, can be used to identify individual operator examination failures.
However, unless these are incidental to the crew evaluation objective, the exam development. process and the exam itself will.become exceedingly complex and significantly more resource intensive than they are now. There is no overriding need that we can see for requiring every exam to have.at least one critical element for each crew member and requiring that any critical element
' failure results in examination failure. For example, failure to properly classify a single event should not, in general, be grounds for exam failure.
Further, it appears that significant changes would be required to the rest of ES-601 to ensure that individuals are treated consistently when performing similar activities such as job performance measures-(JPMs). The additional effort to implement this requirement will detract from efforts to deal with the many other difficult aspects of the exam.
1
-The evaluation of individual operators should continue to occur principally through the examination of job performance measures and the written exam.
Also, the dynamic simulator exam can in fact yield useful insight on individual performance. However, I strongly believe that the current approach -- which stipulates that failure of one ISCT may, failure of two ISCTs will, result in exam failure -- is appropriate.
900213
%220R N
PNV 6 \\ /$ 41[
Jack W. Roe 2
FEB 1 a 13rs
)
- Each-written exam, each simulator scenario, and each JPM is unioue.- Each requires that the examiners observing and grading the individuals use judgement to arrive at a decision. To reduce all aspects of the process to a go/no-go
- situation, as proposed, is not appropriate.
The proposed use of a~ pass / fail approach to address specific knowledge deficiencies does not seem appropriate.
I recommend that a more innovative application of 55.59(b) resulting in a three tier approach:
pass, pass with remedial training, and fail, be investigated as an alternative. provides our specific comments on the draft revision. Enclosure 2 provides items which we believe should be evaluated to potentially enhance to process.
If you have any questions or would like to discuss our response further please contact me or Geoff_ Wright at FTS 388-5695.
T. O. Martin for Hubert J. Miller, Director Division of Reactor Safety cc:
W. Hodges, RI A. Gibson,.RII L. Callan, RIV.
R. Zimmerman, RV R. Gallo, RI T. Peebles, RII J. Jaudon, RIV D. Kirsch, RV K. Perkins, NRR t
R II RIII Rill Il fir
/cg i6in
%r 9
2
..s 4
ENCLOSURE 1 r
.D.1.C.(2)(a)3.: This position does not give credit to the crew for alternative actions.
It would appear more appropriate to fail the crew if they were unable to bring the unit to a " safe" stable condition.
- D 1.c.(2)(b)2.: This position when combined with the " definition" of ISCT's leads to a discrepancy between the various sections of the exam,- It appears, by the definition, that tasks which have no consequence to the scenario at hand, but could under other circumstances, can be classified as.an ISCT.
It appears to us that this type of ISCT is nothing more than a JPM or possibly even a question on the written exam (E.P. classification). Given this scenario, the individual who failed the ISCT would fail the ' exam while the individual who failed the sane task as a JPM or written question would not.
We do not believe that requiring each individual to perform an ISCT should be a requirement. The dynamic simulator was intended to examine crew performance and the JPM's were to evaluate the individual.
The concept that each individual must perform an ISCT defeats the concept of crew training / performance. Examinees may become reluctant to take action on situations they notice.but are not
.directly responsible for,-for fear that such action could result in additional-scenarios being run and therefore additional exposure to failure. Additional scenarios, if required, would still not guarantee that the specific individual would perform the ISCT.
D.1.c.(2)(b)1.: See comment on D.1.c.(2)(b)(2) above.
It seems that we may be confusing the need for additional training with the need to fail the individual on the exam.
Elements 2, 3, and 4.are excellent. Without these the successful completion of the task is not possible. Element 1 is extremely vague. One can say the virtually everything the operator does in the control room has plant and public safety significance. We need to further refine this element by possibly specifying that it must apply to the scenario being run.
We don't believe it is reasonable for anyone to determine how many different ways a task can be performed incorrectly therefore making a determination of t
the consequences impractical. This would require a failure analysis type PRA for each task to be reviewed. Further, virtually any action can complicate an event eventhough it may only be a minor problem. This idea needs to be further clarified with both good and poor examples being provided.
What is an essential safety action? This needs to be defined with examples.
We believe that the " incorrect reactivity control" should be amended to state if required by the scenario. Actions should have immediate consequences to the situation at hand and not just the possibility thereof.
l
g.
(
ENCLOSURE 1 (Cont'd) 2 Disagree with the work "could" in " effectively manipulate ESE controls that could lead to (a.) above." Replace with "did' which, as above, will tie misactions, or the lack of action. to real consequences-rather than hypothetical ones.
Disagree with the item that only challenges a plant safety system without any.
. negative consequences. Virtually any activity can challenge safety. What do we do'with the situation where the individual unnecessarily scrams the facility? This challenges both RPS and ESF systems.
In-general. we would recommend that, for all issues discussed, both good and poor examples be provided.
1 i
^l "i
1 l
L 1
.o
p.
p f-
.i ENCLOSURES 2 ITEMS FOR CONSIDERATION' S
_1.
Whether we keep the proposed plan to fail an individual for failure to properly perform an ISCT or maintain the current system, one of the following items should be performed, a.
' Tighten up on what constitutes an ISCT such that only those items i
which actually result in acute adverse consequences to the public or plant are considered. The examiners indicate that this approach may.
i not be workable given the present time limit'on simulator scenarios and fidelity of simulators.
b.
Form a working group, similar to that formed to develop the K&A catalogues, to nefine what tasks are to be considered by definition to be ISCT's and are caused to fail an individual.
2.
If "b" above is used and the one ISCT failure criteria is implemented then ES-601 reeds to be revised to reflect that tasks defined as critical
-Tasks, if improperly aerformed, will result in failure of the individual i
regardless of where t1ey are in the examination (ie. simulator, JPM's,~or written).
3.
Consider a three-tier approach in evaluating the results of an examination:
pass, pass with the remediation, fail. This approach would modify the-approach taken in "2" above.
a.
First option would be to use 1.b, above, leaving the pass criteria
- as'it presently exists.
If the individual failed two critical tasks y
.he/she would fail. However, if the~ individual failed only one then they would pass the exam but 55.59(b) would be invoked to require additional training for the task failed.
b.
A variation on this thene, using 1.b above, would be to require each individual to accomplish a specified number of " critical tasks".
These could be accomplished during the dynamic simulator and/or JPM
. walkdown. A pass rate percentage would be set. An individual would require retraining and examination on any failed tasks even if they passed that portion of the exam.
k i
2
-