ML19308C172
| ML19308C172 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 07/30/1979 |
| From: | Chris Miller NRC - NRC THREE MILE ISLAND TASK FORCE |
| To: | Chipman G NRC - NRC THREE MILE ISLAND TASK FORCE |
| References | |
| TASK-TF, TASK-TMR NUDOCS 8001210507 | |
| Download: ML19308C172 (2) | |
Text
ps nouq
,(,
'k UNITED STATES 7}p g
NUCLEAR REGULATORY COMMISSION g.-
9 C WASHINGTON, D. C. 20555 j[
July 30, 1979 i
MER0RANDUM FOR: Gordon Chipman FROM:
C. O. Miller
SUBJECT:
FACTORS IN CONTROL ROOM DESIGN AND LAYOUT IISELF WHICH CONTRIBUTED TO THE TMI ACCIDENT (PRELIMINARY)
Re your memo of 24 July, listed below are factors about the TMI control room design and layout which contributed to the accident and which have been proven to my satisfaction already.
I have listed these observations within a narrow interpretation of " design"; thus, not including certain system characteristics, procedures determination, personnel selection and training factors which in the final analysis, must be considered in the total human factors analysis.
/
1.
The display of PORV position which did not portray the physical status but rather just the control signal to the solenoid.
2.
The display of the positions of block valves 12A/B which, in juxtaposition to the other similar indicating lights, precluded ready assessment of
- e ir status, t
3.
The use of pen oscillographs o m rwise called indicator-recorders to portray critical information such as pressurizer level, particularly when the recording medium (the strip chart) moves at only about 0.016 inches per minute.
(Instructors at the B&W simulator claimed 40 sec resolution.
I submit they may be optimistic.)
3 4.
The Baily meters have poorly human engineered dial presentations i
both in terms of scale indications and index discrimiracion.
Some of those associated with key parameters most likely contributed to lack of perception during dynamic events during the accident sequence; especially considering the absence of normal-caution-warning bands on the dials themselves or on the instrument housing.
5.
The basic concept of lights showing positions of valves etc.
i rather than acceptability of that position for a given phase of operation.
l l
l 800123ogp}
f
.o Gordon Chipman "
i
)
/
6.
The basic concept of the total instrumentation system which focuses upon.1tatic information vis-a-vis those dynamic para.neters which become critical in a hazardous transient.
L' 7.
The proliferation of non-critical information adjacent to those displays and controls which assume major importance during emergencies yet all presentations appear equally important visually.
}
Items 1 & 2 above are relatively explicit in application to the TMI l
sequence of events.
Items 3-7 (and probably many more) must be considered collectively as creating a condition where the operators had to do excessive integration of information to evaluate properly the nature of the emergency.
The basic theory I see at this time regarding this accident from a human factors point of view is a false hypothesis, or perhaps a series of false hypotheses, influenced by the design basis instrumentation and training approach used at TMI coupled with inadequate displays to inter-rupt such hypotheses or otherwise allow grasp of the total dynamic situation.
]
i M
b y
I C. O. Miller, Consultant NRC/TMI Special Inquiry Group i
cc:
E. K. Cornell R. C. DeYoung
,i
~.
,v..
.e.
-,,., +.,-,