Re: Derailment at Concort West - Reason

[Bill <wguest@xtra.co.nz>]

Michael Walker wrote:

> At the risk of being severely flamed and sent lewd pictures in my e-mail,
> shouldn't the loco driver take at least some of the blame.

I never flame or post obscenity, Michael.

> After all, he DID misread the signal.

He read the signal, but mis-interpreted its meaning. He was, in his mind,
somewhat forward of his train. He was expecting a normal speed indication at the
signal in question, followed by a caution at the following signal, just exactly
as had happened over 100 times before. When he got the caution (reduce speed, be
prepared to stop at the next signal), he "fitted" it into his expectations.
Tchnically, your statement is quite correct. My point is that "conditioning" is
a quite normal human behavioural phenomenom, and should be guarded against.

> Or are signals in New Zealand prone to ambiguity?

No, the signal was not ambiguous.

> In addition, after recent debates on signaling, shouldn't there be enough
> warning from the first caution signal to slow down to an acceptable speed to
> be able to safely take the points?

There was. But by the time he saw the stop/low speed through points signal (home
signal), he had no show of slowing to a safe speed, let alone stop. Main line
speed 80km/hr, downhill grade, safe speed through turnout 15km/hr, visibility to
home signal approx 350 metres (as I remember).

> Or are the speed restrictions applicable by such signal indications
> incorrectly set by a big enough margin for train drivers to be able to take
> points at too high a speed whilst within the
> correct authorised speed limit for a caution signal?

No, this is not the case, but he only gets one chance! No cab indicator, no loco
assistant, and no radio message telling him of the change in the routine he has
followed 100 times before.

> If NZ signals are ambiguous and speeds under caution are too high over
> there, then the engine driver would be relatively innocent. Otherwise in a
> worst case scenario he was driving too fast and at least partially ignored a
> signal indication in which case he SHOULD take some of the blame.

He was allocated the primary responsibility for what happened. In my view, the
system made in inevitable that such an accident would eventually happen.
Further, there was a readily available technology (control to cab radio) that
could have alerted him to the intended change in plan. If he had been told once
by radio, I believe it would have broken the conditioned way of thinking he had
adopted.

> I guess I just get annoyed at the hand-holding that seems to be expected in
> this response - the train controller might as well drive the train for him if
> he needs to tell him the signal indications and point settings in advance. If
> a particular train driver cannot respond accurately to unexpected signals then
> either his training has been inadequate or he shouldn't be driving.

You should work for Tranz Rail. That was their attitude. So why did it happen
again? In my view, a more modern and satisfactory approach is to consider the
organisational and human factors involved in accidents, and to reshape systems
and adopt practices that minimise the likelihood of human error.

The most important work in this area in recent years has been undertaken by
Professor James Reason of University of Manchester. His book "Human Error" was
published in 1990, and he has recently published another called (I think, but I
have lent my copy out) "Managing the Risks of Organisational Accidents".
Excellent reading.

I notice another post to this message points out that it is not a good idea to
place a low speed turnout in a main line. I totally agree. This design principle
recognises the possibility of human error, and alters the system so that the
consequences of an error are reduced. Good thinking.

Bill