ML20040E235
| ML20040E235 | |
| Person / Time | |
|---|---|
| Issue date: | 12/11/1981 |
| From: | Keyserling W Advisory Committee on Reactor Safeguards |
| To: | Major R Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1400, NUDOCS 8202040048 | |
| Download: ML20040E235 (12) | |
Text
- - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _
Q 7- /%' C
- w. MONROE KEYSERLING. Pn.D.
/g[
!~!
4 CG cT. PAUL CTOE"lT NA 5 BacoMLaws. MASSACHUSETTS 02146 e
(en?) s??.nos?
December 12, 1901 MEMORANDUM TO: R. Major, ACRS Staff Engineer FROM:
W. Keyserling, ACRS Consultant
SUBJECT:
Hinth Tater Reactor Safety Research Information Meeting --
Fresentations on Human Factors Research l
On October 27, 19S1, I attended the Minth Water Reactor Safety Research Information Meeting held at the National Bureau of Standards in Gaithers'ourg, Maryland.
As an ACRS Consultant, I was requested to review the papers presented during-the session on KRC sponsored Human Factors Research.
I have enclosed a brief sumnary of each presentation, along with my comments on the quality of the associated research pro-ject.
On November 2, 1931, I attended the ACRS Subcommittee on Human Factors meeting in Washington, D. C.
During this meeting, Mr. R. DiSalvo of the NRC presented a summary I
of ongoing and planned research activities in the Human Factors area.
Following this meeting, I prepared a report for Mr. D. Ward, Subcommittee Chairman.
The recommendations contained in that report were based on my observations during the two.ectings in Gaithersburg and Washington.
I apologize for the delay in forwarding this report to you.
8202040048 811211
~
FRESE:;TATIO?:S ON HUMAN FACTORS RESEARCH NINTH WATER REACTCR SAFETY RESEARCH INFORXATION REETING October 27, 1981 h tional Bureau of Standards Gaithersturg, '?.tryland i repared by:
M. 1.eyserling ACRS Consultant s
=
9TH WATER REACTOR SAFETY RESEARCH IllFORitAT10!1 !!EETIr4G PREll!!!!iARY AGEllDA FOR HurlAtl FACTORS RESEARCH PRESEllTAT10 tis Lecture Room A October 27, 1981 9:15 - 5:00 9:15 Welcome and Introductory Remarks - R. DiSalvo (f4RC) 9:30 A Validated Task Analysis for Reactor Operators - J.S. Eckel (General Physics) 10:15 Coffee Break 10:30 Time Response Data on Safety-Related Operator Actions - E. Kozinsky (General Physics) 11:15 Human Performance Modeling and its Validation - A. Swain (SNL) and H. Guttman (SNL) 11:45 Results of Graphic Display Evaluations during Transients at the Loss-of-Fluid Facility - 0. !! eyer (EG&G) and M. Bray (EG&G) 12:30 Discussion 1:00 Lunch u
-f 2:00 Results of Surveys on Control Room f.nnunciator Systems and Control Panel Design Changes - W. Banks (EG&G) 2:30 Simulator Technology and Practices - A Comparison among Several Industries -
P.Haas(ORiiL) 3:15 Coffee Break 3:30 System Sensitivity to Human Errors and Use of Actuarial Data in Determining Human Error Rates - R. Hall (Bill) 4:15 Preliminary Findings of the Human Factors Society's Study of fluclear Reactor Safety - H. Snyder (VPI)
..A Validated Task Analysis for Reactor Operators P. Haas, ORNL S. Eckel, General Physics Corp.
SUMMARY
The purpose of this study was to demonstrate how task. analysis procedures can be used to document the activities of key control room personnel in responding to abnormal / emergency conditions in a nuclear power plant.
Four abnormal / emergency events (nuclear instrument failure, small break LOCA, steam generator tube leak, and inadvertent safety injection at power) were studied on the Sequoyah, Westinghouse PWR simulator. The task analysis was supplemented with data collected by an on-line simulator monitor system which measured operator responses and errors in carrying out required tasks.
The authors concluded that task analysis data could have important and useful applications in the areas of manning, training, workstation design, job performance aids, communications systems, and maintenance requirements for nuclear power control rooms.
COMMENTS Task analysis is an important tool which can be used in evaluating control rooms for human engineering deficiencies. This study demonstrates the usefulness of task. analysis ~ for a limited set of emergency conditions on one PWR simulator.
A major shortcoming with task analysis is the tremendous amount of time required to collect and reduce data.
For this reason, it is unlikely that the methodology would prove to be efficient in evaluating individual. control rooms for human design deficiencies.
Instead, this methodology is most appropriately suited for evaluating generic designs.
In summary, task analysis procedures will be very useful in improving the safety of nuclear power operations.
It is recommended that research in this area be continued, particularly for control room generic designs which are widely used.
i t
i i
Time Response Data on Safety Related-Operator Actions l
P. Haas, ORNL E. Kozinsky, General Physics Corp.
i
SUMMARY
4 In this study, both PWR and BWR simulators were used to establish a. data base of operator response times to abnormal conditions in a nuclear power plant.
Operator response times were measured using the computerized Performance Measurement System (PMS), developed by General Physics Corp. under contract to EPRI.
Using this system, the actions of both the simulator and operator were j
recorded, thus preserving a chronology of plant conditions and the operators' responses. By recording and analyzing the response times of operators durirg j
plant malfunctions, it was possible to construct a probability distribution of I
the time required for operators to perform corrective actions.
Data were also collected on Performance Shaping Factors (PSPs), such as procedures, control room design, operator training and experience, etc.
Cor-relational. analyses of the relationships among PSPs and delay times have not yet been performed,' but are scheduled at a later time in the study.
I COMMENTS f
This research has great potential for identifying shortcomings in control room designs, operator training, job performance aids, and emergency operating procedures.
i f
i i
i P
l
.. ~.., _.... - ~. _,. _, _, _.,.,
Sandia Human Factors Projects e Nuclear Power A.D. Swain, Sandia Human Factors Group
SUMMARY
During this presentation, Dr. Alan Swain reviewed approximately 20 projects under contract to the Sandia Labs in the area of nuclear power human factors research.
COMMENTS Dr. Swain covered a large number of topics in his 30 minute presentation.
Due to the fast pace and lack of detail, it was impossible to develop a critical opinion on any of the projects he discussed.
I s
Results of Graphic Display Evaluation During Transients at the Loss-of-Fluid
-Test (LOFT) Facility i
O.R. Meyer, E.G.&G. Idaho,-Inc.
SUMMARY
This presentation discussed ongoing research at E.G.& G. Idaho concerned with developing methods for evaluating human factors attributes of Safety Parameter Display Systems (SPDS).
A " static evaluation tachistoscopic experiment" was performed in which a subject's ability to detect abnormal operating conditions was measured using three different SPDS formats.
- 1) CRT simulated panel meters,
- 2) CRT " Deviation Bar Charts", and 3)
CRT " Star Diagrams".
Five different severe transients were presented to test subjects (LOFT reactor operators qualified to level IV) using the three idsplay formats.
Preliminary results indicated that the " Deviation Bar Chart" and " Star Diagram" formats were associated with improved detection performance when compared to the simulated panel meters.
Performance when using the " Star Diagram" was slightly better than performance when using the " Deviation Bar Chart".
COMMENTS The quality of this research appears to be quite good, and the results will i
be useful in developing methods for evaluating alternative SPDS designs.
Two questions should be raised, however, concerning the direction of this project:
1)
There appears to be an implicit assumption that multi-color CRT displays are superior to conventional panel meter displays.
Before a strong committment is made to using the CRT format for the SPDS system, additional studies should be performed using conventional panel meter displays. CRTs suffer'many shortcomings (e.g., limited display size, glare interference from ambient lighting) which were not evaluated in this study.
- 2) The evaluations reported in this paper are best classified as basic research on the design of CRT displays. CRTs are now being used for many (non-nuclear) applications, and a data base of knowledge is i
i i
I
rapidly being developed.
In tems of cost-effectiveness, NRC sponsored literature searches may be preferable to NRC sponsored research.
Results of Control Room Annunciator Systems Evaluations William W.
Banks, E.G.& G.,
Idaho, Inc.
SUMMARY
This study was performed to (a) identify problems with annunciator display systems in reactor control rooms, (b) provide specific and generic solutions to these problems, and (c) provide direction for future improvements and research in the area of man / machine interfaces for annunciator display systems.
Investigators visited several reactor control rooms to examine annunciator systems. These systems were evaluated using existing military standards and by conducting structured interviews with operators.
Fifteen major problem areas were discovered as a result of the inspections.
Examples of these problems were dramatically illustrated through the use of 35mm color slides.
The author concluded his presentation by reccmmending that top-down systems analyses should be used ia the development of new annunciator systems.
At a minimum, the design process should include a safety missions statement, functional analyses, and task analyses to determine operator information re-qu:.rements.
COMMENTS The presentation provided graphic examples of human factors deficiencies in control room annunciator display systems. The Control Room Design Review procedures now being developed by the NRC shoull prove very useful in identifying similar problems in other nuclear power facilities.
Training Simulator' Requirements -- A Comparison Among Several Industries P.M..Haas, ORNL
SUMMARY
This study examined the use of training simulators in the nuclear power industry and in several other applications (e.g., airlines, railroads, military).
As a result of these comparisons, several shortcomings were discovered with the training programs currently used in the nuclear power industry:
1.
There has been a tendency within the nuclear power industry.to develop simulator training programs in the absence of clearly-defined training objectives.
2.
There has been little coordination between the use of simulators and other aspects of the overall training program (e.g., personnel selection and the " textbook" training curriculum).
3.
Cost / effectiveness analyses have not been used in the development of simulator systems. As a result, the overall efficiencies of existing simulator training programs should be subject to review.
The author concluded that the nuclear industry and the NRC should move toward developing a systems approach to training. Specific recommendations included the followinc:
1.
A cooperative effort by the NRC and the nuclear. industry leading to a' planning document for developing the systems approach.
2.
A clear delineation to the industry of NRC's planned regulatory involvement in operator selection and-training, and simulator certification.
3.
Continuation of effort to develop a job / task analytic data base for control room personnel.
4.
Continued improvement of ANSI /ANS 3.5.
COMMENTS Dr. Haas has developed sound ideas for the training of control room j
operators. His recommendations in this area should be seriously considered by the nuclear power industry and the NRC.
i l
s Human Factors Research Programs at Brookhaven National Laboratories R.E. Hall, Group Leader Risk Assessment and Engineering Analysis Group
SUMMARY
In this presentation, Dr. Hall reviewed several projects in the human factors area which are currently being performed at the Brookhaven National Laboratories.
COMMENTS Due to the fast pace of this presentation, it was diffiellt to critically evaluate the various projects discussed.
l Reactor _
Preliminary Findings of the Human Factors Society Study of Nuc ear _
Safety Research_
i Harry L. Snyder, Virginia Polytechnic Institute and State Univers ty
SUMMARY
d being l
This presentation discussed the current status of an ongoing stu yThree specific tasks w l
l performed by the NRC by the Human Factors Society.be performe Survey NRC offices and report to determine those aspects of nuclear power plant safety with human factors implications.
1.
lt s
Selectively check with the nuclear industry regarding the comp e enes l
2.
and accuracy of the study group's findings.
ht For each regulatory activity with human factors impacts ft 3.
1981. The final The first two tasks have been completed as of mid-August, l
to develop a task is currently underway, and the Human Factors Society p ans long-range plan for the NRC by mid-February, 1982.
COMMENTS None.
l l
(.,
f
.