Originally Posted by
FloaterNorthWest
Helicomparator,
To quote the final AAIB report.
“It was not possible to determine precise timings but it was calculated that, before the helicopter reached Bothwell, the pilot was presented with a low fuel 1 warning caption, with the associated aural attention-getter. This aural attention-getter was acknowledged by the pilot. The warning caption then extinguished, before re-appearing after an undetermined interval. This, too, was acknowledged by the pilot. The caption extinguished again. The low fuel 2 warning caption then illuminated, with the associated aural attention-getter, and was also acknowledged. The time, in addition to the low fuel 1 caption then re-appeared a third low fuel 2 caption. This was acknowledged, before extinguishing again, leaving the low fuel 2 warning. The low fuel 1 and low fuel 1 warning caption then re-illuminated once more and was, again, acknowledged. After this, the low fuel 2 warnings captions remained illuminated for the rest of the flight.”
it would appear the warning system was working. The unexplained area is the continuing to fly and take tasks after they had illuminated.
Were they used to flying with Low Fuel captions so desensitised?
FNW
Yes I know, however what is not considered in the report is the integrity of these events stored in the memory of the central warning system. I get the impression that the system records a chain of events without time stamps, so a certain degree of guestimation is involved in matching the stored events to the preferred scenario.
The system thinks a caption was illuminated but clearly that system isn’t actually measuring light output from a bulb, it is something much further up a chain that thinks a caption OUGHT to be illuminated without actually knowing that it is.
The report isn’t explicit about how these events are detected and stored and so I remain a bit dubious as to whether they are concrete facts or just the most probable (in isolation) explanation. One thing I learnt from my days in HFDM is that it is very easy to imagine a match between flawed data and an expected scenario, when in fact no such match exists.
Given the choice between a pilot and 2 crew ignoring persistent warnings both trivial (prime pumps) and serious (low fuel x 2), over a prolonged period, vs a fault in the warning system that meant that, whilst the “brains” of the system wanted captions to be illuminated, in fact they weren’t, Occam’s razor might suggest that the latter is the simpler, and hence the correct, explanation.