ML19208A912
| ML19208A912 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 07/24/1979 |
| From: | Chris Miller NRC - NRC THREE MILE ISLAND TASK FORCE |
| To: | Cornell C NRC - NRC THREE MILE ISLAND TASK FORCE |
| Shared Package | |
| ML19208A909 | List: |
| References | |
| TASK-TF, TASK-TMR NUDOCS 7909180250 | |
| Download: ML19208A912 (3) | |
Text
.
[pa "!c s
6, UNITED STATES
'\\
NUCLEAR REGULATORY COMMISSION
- .j 'N.,,01%,' ' N's J.
- j WASHINGTON, D. C. 20555 o.,
r 1
/
July 24, 1979
=.,*
MEMORA!!DL11 FOR:
E. Kevin Cornell, Staff Director, SIG FROM:
C. O. Miller, Consultant, SIG
SUBJECT:
MAN-MACHINE INTERFACE QUESTIONS RAISED BY TMI-2 AND POSSIBLE SHORT-TERM REMEDIAL ACTION The recently released " Lessons Learned Task Force Report...," NUREG-0573 st >es:
"A widely accepted lesson leamed from the TMI-2 acci&nt is that the man-machine interface in some reactor control rooms needs significant improvement."
Two short-term recommendations were i fered thereafter: direct indication of PORV and safety valve positions and better instrumentation for detection of inadequate core cooling.
(See pp. 7-8 of the report.)
Other short-term recommendations offered which were closely related to the man-machine question included improved analysis, emergency procedures
'and training to enhance operator performance during transients (pp.11-12) and an improved reactor operations comand function (p.12). However, no specific human factors inputs to these efforts were identified.
Prior documents, of course, have been critical of man-machine relationships in nuclear power control centers. These include:
WASH 1260 (1972)
WASH 1400 (1975)
Swain's Zion Report (1975)
EPRI Study (1976)
House Committee Hearin Aerospace Corp. Study'g Report (1976)
(1977)
Lewis Report (1978)
In addition to reviewing all of the above material.as pertaining to the subject at hand, the undersigned attended a very professional briefing of Commissioner Ahearne by Leo Beltracchi at a meeting chaired by Roger J.
Mattson on 20 July 1979.
The discussion was highly illuminating regarding the rationale leading to the aforementioned L2 recommendations and under-standing how the existent control rooms became configured as they are.
Efi.10f/
79 0918 ogge
'\\
E. Kevin Cornell July 24, 1979 Based upon this background, it is my opi~nion that NRC has still not attacked on a short-term basis some of the critical human factors problems illustrated by TMI-2; particularly the design deficiencies as they probably exist at some locations in terms of booby-trap display and/or control configurations.
This problem goes well beyond a particular display for a valve or even a particular requirement for instrumented data.
It is the general issue of operator errors poised to occur because of poorly engineered instrumentation and control panels.
Phrased differently, the short-term actions proposed by the L2 Task Force would preclude a repeat of TMI-2. But is it not logical to assume other confusing, if not simply inadequate, display and control configurations exist that could induce other equally critical operator errors? Certainly the references noted earlier suggest strongly an affirmative answer to that question.
Accordingly, I offer the following recorr.endations for consideration by the Commission for urgent action:
1.
Institution of a program to survey all operational and near-operational plants to look for and resolve on-site human engineering booby-traps... those control room deficiencies which are obvious, critical and amenable to simple solutions. This effort could probably be accomplished by specially selected teams '
using checklist type information already available from the EPRI and Zion reports.
2.
Institute an incident reporting and analysis systen emphasizing the reporting of human errors by control room and auxiliary operators with guarantees of anonymity if desired and reasonable protection against disciplinary action by either the government or the person's employer. Considerable precedent for this exists in NASA-administered programs both for their own operations and for the FAA.
3.
Implement a program to simplify emergency procedures including use of modcrn presentation methods which require response to symptoms first and integration thereof later to determine what transient' occurred, not vice versa which seemed to be the case with the cumbersome narrative-type TMI-2 orocedures.
4.
For training programs instituted by EPRI or others, incorporate a
~
segment which educates operators as to the nature of human error and how he or she can best avoid it.
E,0.0.18 O
C e
e 1 -
s E. Kevin Cornell July 24,1979 S.
Establish an appropriate organizational segment or otherwise staff NRC with qualified human factors personnel.
This should be an interdisciplinary group whose mission would be to provide technical support in the human engineering and the personnel subsystem fields which is a key contributor in the resolution of man-machine inter-face problems.
Recommendations (1) through (4) above might well be best implemented by a coordinated NRC,' EPRI, IEEE, AIF effort as distinguished from the regulatory process per se.
There have been several signals that the industry is anxious to resolve these problems; thus, they might welcome such an approach.
The urgency of these tasks can be appreciated by assessing liRC's posture in the eyes of the public when it becomes generally known how much background informatio.n was available to identify the control room de/iciencies, yet NRC's response had been to study the problem further.
The short-term corrective action proposed thus far is clearly meaningful, but it does not go far enough, soon enough, in the human engineering aspects of control pancis; nor is there current assurance that proper human factors technology will be applied tn either the short-term or long-term solutions.
I' stand ready to explain these recommendations further or my reasons for them.
0 5l
.#CL-C. O. Millei, Consultant NRC/TMI Special Inquiry Group cc:
M. Rogovin R. C. DeYoung G. Frampton, Jr.
e e
I e
8 SV1019 f.
l