ML19290E625

Submits Info Re Essex Corp Audit of VEPCO Human Engineering Evaluation of Control Room.More Comprehensive Evaluation Must Be Performed & Assignment of Backfit Scheduling Often Does Not Agree W/Safety Interests

ML19290E625
Person / Time
Site:	North Anna
Issue date:	03/10/1980
From:	Mallory K ESSEX CORP.
To:	Beltracci L Office of Nuclear Reactor Regulation
References
NUDOCS 8003140305
Download: ML19290E625 (5)

Text

.. .

4

~

[0 -3 3 h ESSEX March 10,1980 To: Leo Beltracchi From: Ken Mallory [

Subject:

Findings of the Audit on North Anna _Upit 2 On March 3 and 4,1980, the Essex Corporation performed an audit of a VEPCO human engineering evaluation of the control room at VEPCO's North Anna - Unit 2 Nuclear Power Plant. Essex had two basic objectives in the audit:

1. To assess the completeness of the VEPCO review.

2. To examine the criteria used by VEPCO to judge the criticali:y of specific problems.

The Essex audit was brief and summary in nature. Although our evaluation of VEPCO's results is believed to be valid, the specific human engineering deficiencies observed in the plant must be considered tentative.

1.0

SUMMARY

CONCLUSIONS

a. While the VEPCO control room review uncovered some problem areas, it overlooked many others. The number and potential severity of problems uncovered in a very brief audit shows that a more comprehensive human engineering evaluation must be performed before decisions on the human engineering fitness of Unit 2 can be made.

b. VEPCO's assignment of human engineering backfits to "Short Term" and "Long Term" of ten does not agree with safety interests. We speculate that one primary criteria was the time needed to implement the backfit.

We disagree with some of the backfit scheduli .g proposed by VEPCO.

2.0 GENERAL CONCERNS Each of the concerns described below is given a Subjective Risk Assessment Weight based on the likelihood that a particular aspect of CR design will lead to an operator error in a safety-related activity. This likelihood is based on the opinion (s) of the Essex human engineer (s) reviewing the CR.

ESSEX CORPORATION

• 333 North Fairfax Street. Alexandria. Viq; inia 22314. (703) 548-1500 80031ko

Category 1 - High Risk of Operator Error in Safety-Related Activity Catego y 2 - Moderate Risk of Operator Error in Saf ety-Related Activity Category 3 - Risk of Operator Error in Safety-Related Activity

a. Safety injection Status - While there is a matrix of valve position indication lights that provice containment isolation status, there are no integrated er grouped displays for safety injection status (Category 1).

b. Lamo Test - There is no convenient means to test indicator lights, other than annunciators. During emergency operations, bulb f ailure could be misleading and time consuming (Category 2).

c. Trending - VEPCO should verif y the Unit 2 on-the-board trending capability agamst potential trending needs during pre- and post-accident operations.

Trending needs should recognize that operators may take some time to diagnose a failure and may trend many parameters during the diagnosis (Category 3).

d. Annunciators - VEPCO recognized the need f or prioritization and some reorganization of annunciators; however, when a failure is cleared the annunciator lamp is extinguished without any notification on-the-board to the operator (the alarm computer lists all " clears"). Since the list may not be read for some time, there should be some means to immediately inform the operator when system is brought back online.

• Prioritization - Categcry I e Clearing Notification - Category I e Reorganization - Category 3

e. Process Controllers - The Hagan " Manual-Auto" process controllers provide meter readouts of valve position derived from control signals. VEPCO should determine where direct readouts would be needed to maintain cognizance over valve status (operating and failure) (Category 2).

f. Procedures - VEPCO has not performed a human engineering evaluation of procedures. While a brief review found several strengths in Unit 2 documenta-tion (bi-level organization, detailed steps), some weaknesses were noted (excessive cross-referencing, long instructions - 52 words in one sentence, instructions containing up to four (4) steps, instructions nested in " Notes" and

" Cautions," incomplete instruction, poor English, insufficient inf ormation to make decisions, inefficient organization, etc.) VEPCO should evaluate the f ormat, content, typography, inf ormation sufficiency, and accessibility of every cont ol room procedure (Category 1).

g. Control Room Staffing - The CRO indicated that four (4) operators are involved in emergency operation: one at primary console; one at secondary console; one at back panel (with backup. turbines, etc.); and one reading procedures and performing some functions on the vertical panels. Based on observations .of three (3) procedure walk-throughs, Essex believes tha- more than one, and perhaps as many as three, operators are essential to eme gency actions at Unit 2. Operators should be trained and rotated as teams.

VEPCO should evaluate the problems inherent in emergency operations with less than four operators, and take steps to assure that the optimum number are on duty at all times. Some of the Emergency Procedures are quite dif ficult f or one operator (Catego.y 1).

h. Switch Position Labelinn Inversions - In a sample rurvey of switch position labels, Essex f e. d some labels that were contrary to CR conventions. VEPCO should review all labels and correct those that differ from stereotype of convention (Category 2),

i. Labeling - VEPCO noted problems in labeling nomenclature, size of font, and contrast. Essex agrees with VEPCO's observation (Category 1 - some of the least readable labels were on ESF panels - labels were black on purple).

j. Display Location - VEPCO noted that some critical displays should be moved f rom the verticals to the console. Essex agrees with this observation (Category 3).

k. Color Coding - VEPCO should assess the likelihood that meanings for colors f or labeis, demarcation lines, indicator lights of all va-ieties and annunciators will become confused during high-stress operations (Category 2).

1. Emergency Eculoment - VEPCO should evaluate the number and accessibility of breathing devices (Category 3).

m. Meter Confusion - Some strings of vertical meters are six meters long, and at least one string has identical scales f or all meters. This could lead to reading confusion. VEPCO should assess means (e.g., bessell coloring, demarcation lines) to separate and group meters (Category 2).

n. Switch Guarding - VEPCO noted the need for guards on some switches. Essex concurs, particularly f or two "J-handle" switches near the edge of the secondary panel (Category 1).

o. Meter Ranges - VEPCO noted the need to mark nominal operating ranges on vertical meters. Essex agrees (Category 1).

p. Access to CR - Large numbers of VEPCO maintenance personnel (approxi-mately 10) were in the CR during both days of the Audit. Access to back panels, communications, and passageways was often times blocked. VEPCO should implement some means to significantly reduce the traffic in the CR (Category 3).

q. Recorder Doors - At least one and perhaps more of the strip chart recorder doors had def ective latches and were attached to the board with tape. Either the recorder latches should be backfitted or maintained adequately.

r. Communications - VEPCO identified a need to revise CR communications f acilities. Essex agrees (Category 2).

s. Operator /CR Integration - VEPCO has not performed any systematic walk-throughs of procecures in the CR. Such walk-throughs are essential to qualifying procedures as well as to identification of problems in panel design, manning, etc. VEPCO should perf orm thorough walk-through/ talk-throughs of all procedures (Category 1).

t.

Panel Arrangement - Unit 2 control panels are loosely arranged by system; however, in many important cases systems are fragmented and wide spread among other system. While some spreading is no doubt caused by separation criteria, much is not. VEPCO should undertake a study on means to improve layout without degrading training measurably. Sequential, functional, systems mimics are possibilities (Category 2).

u. Bottom of Panels - Like TMI - Unit 2,! North Anna - Unit 2 has portions of the vertical panels that are obstructed from view at the console. Unless VEPCO implements restrictions, some critical displays will be mounted in these obstructed portions of the vertical panels. This could cause serious safety problems (Category 3).

v. T sat Readouts - T sat digital readouts are placed on top of the cabinets containing in-core monitoring. Operators prefer these readouts to saturation monitors on vertical panels, but to read the digital T sat meter requires a step ladder. VEPCO should instal! a repeater readout, or move the current T sat meter, or find some other way to provide digital saturation information without requiring the operator to find and climb a ladder (Category 2).

w. Manual Process Controller - The small Hagan " manual" process controller does not provide a direct indication of valve status. VEPCO should investigate the impact of this feature on the information available to the operator when the valve is operating and failed (Category 3).

x. Timing - Three ESF operations require the operator to take action at specific times since: CDA. A small timer, actuated perhaps by the Containment Hi-Hi pressure, would enable the operator to take the needed action without guesswork (Category 3).

y. VEPCO Recommendations - In addition to those mentioned above, VEPCO mnde several recommencations from its study.

1. Alarm prioritization by sound (Essex Category 3).

2. Demarcation lines on controls and displays (Essex Category 1).

3. Improve meter / chart scaling (Essex Category 2).

4. Communications between personnel wearing emergency d ess (Essex Category 1).

z. Containment Isolation - Some containment isolation status indicators are below the line-of-sight of an operator working at the console (Category 2).

3.0 HUM AN ENGINEERING STRENGTHS Listed below are some of the exemplary features of the North Anna control room.

a. Demarcation line color on the console relate to lines on the vertical panels enclosing related displays.

b. Some EPs are clear, concise, and complete.

c. First-out annunciators help in diagnosis.

d. Status displays (annunciato.s) help in diagnosis,

e. Organization of annunciators and ver.ti'tal panel displays follow the organi-zation of systems on the console, f.

Switch guarding on back panels prohibits inadvertent operation.

It should be noted that each of the preceding items, except those identified by VEPCO, resulted f rom a very brief CR review. Therefore, the severity of problems underlying these concerns may be questionable.

.