ML20197J437

Forwards Human Factors Engineering Branch Review of 850325- 29 Appraisal of Emergency Response Facilities.Size of Technical Support Ctr Adequate for 30-35 People Assigned

ML20197J437
Person / Time
Site:	Columbia
Issue date:	04/05/1985
From:	Russell W Office of Nuclear Reactor Regulation
To:	Scarano R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
References
CON-WNP-0795, CON-WNP-795 NUDOCS 8504160015
Download: ML20197J437 (8)

Text

.

g gt f f WWW M'

,q.

1985 DIN"5:

APR 5 Central Files EFEB Files (2)

DHFS Files (2)

WRegan HFEB Members MEMORAf100f1 FOR: Ross A. Scarano, Director Division of Radiation Safety and Safeguards Region V FROM: William T. Russell. Acting Director Division of Human Factors Safety Office of Nuclear Reector Regulation

SUBJECT:

HUMAri FACTORS ENGIt'EERING PRANCH IflPUT F0P. TUE tHP-P EMERGENCY RESPDNSE FACILITIES APPRAISA'. REPORT Genrge Lapinsky of the Hunan Factors Enginearing Granch was involved in the liarch 25-29, 1905 appraise 1 of the Energencv Response Facilities et WNP-2. The results of his review arc enclosed for your ute in documenting the overall appraisel results. Any questions regarding this natter should be dirceted te Mr. Lapinsky (492-9694).

C;':'*C f:' wit 5 ,

a. , .%,a-du ~

Willian T. Russell, Acting Director

-! ' ' ,/ Divicion nf Hunan Factors Sa'etv Office of fluclear Peactor Regulatfon

Enclosure:

As stated cc: E. Willians M. Schuster, R. V R. Fish, R. V G. Temple, R. Y m

~

8504160015 850405 F ADOCK0509g9 b (<

v .

" ' "" M . . . . . . lM . . .... .. ..l m ....... . . . . . . . .liFEB, , , , , , , ,,,,, Di@ OD,,,l ,,pl D @ Q ... ,,,,,,,,,, ,,,,,,,,,,i,,,,,,,,,,,,, ,,,,,

i

"~4 .G.4pAnskyi mec...S!Mois.s.. .....WHRegan.... . . PHr.schhm. . . . !. . . . Wnuas e n . . . . . . . . . . . . . . . . . . . . . ....................

em lt . . . 4. ./. 4. p/.e. 5. . . . . l . .. ... 4.

. . 4./../.5.5./.

. /. 8. 5. 8 .5.

. .4.. ../.

} .4/

.. .. .. .t. ....... . ../.8. 5. . . . 3......

. . . ./ 8 5

,re- rony ve . c s,c kl.Cu 9 m QFCf((4 { Df CQP D QQCY P d8 W I M H" 2

HUMAN FACTORS ENGINEERING BRANCH INPUT FOR THE WNP-2 EMERGENCY RESPONSE FACILITIES '

APPRAISAL REPORT During the week of March 25-28, 1985 an appraisal of the WNP-2 Emergency Response facilities (ERF) was done. The Human Factors Engineering Branch (HFEB) participated in the appraisal and was responsible for the review of certain inspection items in IE Inspection Procedure 82212, Rev. 5. The results of the HFEB review are presented here by inspection cuidance number.

1.0 Technical Support Center (TSC) 1.1.1.1 Size: /

The TSC was designed to accommodate a minimum of 25 people. Useable floor space was estimated by the licensee to be about 3,100 square feet. The staff estimated the number of people assigned to be 30-35 and the area of assigned workspace to be about 2,500 square feet. The TSC walk-through did not provide an opportunity to do a dynamic evaluation, but in the subjective judgment of the staff (based on static, visual inspection), the size of the TSC is adequate for the 30-35 people assigned.

1.1.1.2 Layout:

The layout of the TSC was designed without benefit of a task analysis or information flow analysis. However, users were involved early in the design, and any deficiencies or recomnendations that arose during drills and exercises were formally reported and factored into the design. In addition, the licensee compared the WNP-2 design with the design of other licensee's ERFs to identify whether any important design aspect had been overlooked.

Because of time constraints, the TSC walk-through did not afford the staff an opportunity to evaluate the dynamics of the TSC layout. However, based on static visual inspection and the licensee's onocing deficiency-reporting-

' system, the staff judges the layout of the TSC to be adeouate to support its function.

1.1.1.6 Display Interfaces: '

The TSC has several methods available for display of information which fall into two general categories: (1) computer-baseddisplays,and(2) non-computer-based displays. The non-computer-based displays include plant status boards, accountability logs, significant event logs, historical event records (hand-written Vu-graphs), and printed output from the facsimile copier. The staff found the non-computer-based displays to be easily readable and understandable, and adequate to support TSC functions.

The computer-based displays, on the other hand, had a number of shortcomings that could adversely affect the function and/or the timeliness of response of the TSC. The staff recommends that the licensee address the issues described below as areas needing improvement.

The first observation has to do with several issues that interactively culminate in a condition known as " lock-out." " Lock-out" is a condition in which the computer will not respond to commands from the user's terminal -

the user cannot even escape or interrupt the program to start over.

The issues involved in this problem (from the human factors perspective) are:

(1)computerresponsetimesarevariableandoftenveryslew;(2)nofeedback is provided to tell the user that a command has been received and/or the computer is working on the command; (3) the system " stacks" commands and works on them sequentially, and (4) the systen provides no meaningful error messages and has no " HELP" function. These issues interact in the following way:

Because the response time of the computer is variable, users have no consistent time constant by which to judge whether the computer is responding normally. Because response times are sonetimes very long (greater than 30 seconds), the user cannot quickly discriminate between a slow response and ne response. In addition, the users are generally provided with no feedback messages that could alleviate the problems of slow and variable respense,

i.e., visual or audible indications that a comand has been received and is being executed.

Given the above situation, the staff observed that users had difficulty determining whether a program was executing properly. A comon user response was to hit a function button repeatedly after 5-10 seconds, then to variously hit the " break" key, the " return" key, and/or a control character string. If none of those actions elicited a response, the user would assume that the terminal was " locked-out" and the user would either reset or turn off the tenninal and attempt to reload the program using control character comands.

If unsuccessful, the user would then call a computer professional for help.

o The staff noted that even if a program was executing correctly (albeit slowly) after 5-10 seconds, users tended to assume that the command was not being executed and would start entering new commands, etc. as described above. Since the computer " stacks" comands, the likelihood of sequencing errors is increased and may, in fact, account for some of the " lock-out" problems being experienced. The situation is further exacerbated by the lack of meaningful error messages and the lack of a " HELP" function.

In summary, the users of the computer terminals in the TSC may be

" locked-out" for seconds to minutes. The system provides little or no user feedback for the user to determine program status, and provides few meaningful error messages and no " HELP" function.

The second observation by the staff was that TSC and EOF terminals could affect the " Emergency Classification" status bar on the SPDS in the control room without the knowledge or consent of the control room operators.

The Graphic Display Systen (GDS) is a software subsystem that provides various displays of plant operational data, e.g., major plant variables' current values and trends,= isolation valve status, control rod position, etc.

One of the features of the GDS is a status bar representing " Emergency Class'ification." This status bar is on all cf the top-level or current 4

• e

4 conditions display pages. It is intended to alert the Control R'som, TSC, and EOF personnel to possible changes in Emergency Classification Levels, i.e.,

Unusual Event, Alert, Site Area Emergency, General Emergency. The bar is normally areen but changes to yellow, red, and magenta as the event classification escalates. The inputs determining the classification level are provided by a software routine that takes into account plant symptoms and radiological release data.

The staff found that the Emergency Classification bar could be put into an alerting state from the E0F by doing dose projections on another software subsystem that feeds a calculated value to the Emergency Classification

, routine in GDS.

Although the staff does not believe that this problem will compromise the function of the TSC or E0F, it is a disrupting influence and should be corrected by enhancing facility security at the EOF or terminal security at the TSC and E0F. In addition, confirmatory prompts should be added so that the user knows that he is about to change emergency levels, e.g., " Based on plant conditions, your current proiection of dose rate will change the Emergency Classification Level to (Unusual Event, etc.). Do you want to do this? If so, call the control room, x , and the TSC, x  ; inform them of the change in status, they type YES, and press return. If not, press

'F17' to return to the menu." Such prompting would minimize inadvertent changes to the GDS displays-.

The staff also observed that there is no direct display of maxinum integrated dose or dose rate at the 1.2 mile site boundary. Since this value may affect emeroency action levels, and since it is already automatically calculated and is an input to the GDS Emergency Classification status bar algorithms (see above), the staff feels that it should be displayed or made easily displayable. Currently, personnel in the TSC and Control Room must estimate values from grid map displays which are virtually unreadable (data is printed inyellow).

At the exit briefing the licensee protested that this was not th'e case.

Further investigation confirmed that the staff was correct - there is no direct display of dose / dose rate at the site boundary. In addition, control room personnel stated that they probably would use the TRS-80 microcomputer to calculate the 1.2 mile dose rate because estimation from the grid map displays is so cumbersome.

The final observation regarding the TSC computer-based displays was that there was extreme glare on the cathode-ray-tube (CRT) screens that affected the readability of the displays. Of particular concern was the poor readability of the " slave" monitor in the center of the TSC which is intended for use by the Plant Emergency Director and his technical staff. There are several economical methods for reducing CRT glare, such as CRT hoods or egg-crate diffusers for overhead lighting. The licensee should investigate some of these alternatives.

1.2.6.1 Validation and Verification The computer programs supporting the TSC and EOF functions were subjected to an independent Verification and Validation (V&V) Program by NSS Incorporated.

The Program included fourteen simulated transients.and three major pre-ops tests. The front-end data acquisition system (TDAC) was validated by direct correlation with control room indications. Several volumes of V&V results were available, but because of time constraints the staff did not review l these documents.

1.2.6.3 Manual Systems For manually gathered data, the licensee has no extraordinary method for assuring the reliability and validity of the data. Crosschecks and confirmations are done only if the values are deemed to be questionable.

Formal procedures and checklists are generally not used for data gathering, and independent sources of irformation are generally not sought except when the primary source fails (as with meteorological data).

In summary, the staff confirmed that computer software was subjected to a V8V Program to assure its reliability; however, manual data gathering systems are rot normally subjected to a reliability check unless the data is deemed to be questionable.

1.3.1.4 Communication Interfaces During the walk-through the staff observed that the designated communications interfaces between the control room, TSC, OSC, and E0F are easily identifiable and conveniently located. Communication pro:edures were not exercised during the walk-through and, therefore, their adequacy could not be determined.

3.0 Emergency Off-site Facility (EOF) 3.1.1.1 Size The E0F was designed using the guidance of NUREG-0696 and NUREG-0654 as criteria for size. The useable area is estimated to be 20,000 square feet and the design assumption for number of personnel is 50. Generally, all work areas are multipurpose, that is, they serve as office and classroon space on a day-to-day basis and are "taken over" during emergencies. Because the primary use of these spaces is as offices or classrooms, they were amply

' sized and will accommodate greater than the 50 people assumed in the EOF design criteria. The staff concludes that these areas provide adequate workspace for EOF functions to be carried out.

,3.1.1.2 Layout During the initial design development, design guidelines were established for necessary adjacencies and sepa 3tions. These have generally been adhered to.

Although the EOF walk-through did not allow a dynamic evaluation of the EOF layout, the staff's visual inspection identified no obvious problem in the present arrangement. The staff suggests that the licensee consider moving

, the " Decision Center" to a more central location, such as the area now designated as the "Off-site Agency Coordination Center."

3.1.1.6 Display Interfaces '

The staff observed that the non-computer-based displays used in 'the E0F were consistent with those in the TSC and Operational Support Center (OSC), and were easy to read and understand.

Computer-based displays suffered from the same shortcomings as those in the 1SC (described in Section 1.1.1.6), except that glare did not seen to be as bad a problem in the E0F. In addition, output from the printer in the Meteorology and Unified Dose Assessment Center (MUDAC) was almost unreadabic because the ribbon needed changing. Staff suggests that routine maintenance and surveillance should be done on E0F equipment to avoid such problems

. during emergencies. Also, there was no users' manual or other documentation available in the Technical Data Center that described the operation of the Graphics Display System. Finally, hardcopy of operational data and trends is difficult and cumbersome to get - data values must be accessed through a start-up and testing program (STAR) and printed in textual form; copies of graphic displays from GDS cannot be produced at all in the E0F.

3.2.6.1 Validation and Verification and 3.2.6.3 Manual Systems The observations of the staff concerning the TSC (described in Section 1.2.6.1 and 1.2.6.3) apply to the E0F as well .

3_. 3 .1. 4 Communications Interfaces The walk-through of the EOF did not afford the staff an appropriate opportunity to evaluate the functionirg of the ccmmunications interfaces.