Executive summary
A train controller starting his morning shift on 23 September 2008 unknowingly planned to direct a freight train along a line that was occupied by another freight train, which was standing awaiting routing through an area where a signalling fault was under repair. He was not aware the second train was stationary on the line. A potential low-speed, head-on collision was avoided when the first train was subsequently routed along the adjacent line after the signal failure had been partially corrected.
The existence of the second train was not known to the train controller because the senior controller in charge of the previous shift had omitted to record the movement of the train on the train control diagram, and it was not showing on the mimic screen in the national train control centre owing to the signal failure. Neither the current train controller, nor the previous controller and a trainee controller he was mentoring had noticed that the second train, which was a scheduled service, was not displayed on the train control system.
The train controller who omitted to record the second train on the train control diagram was suffering from fatigue caused by an excessive planned and unplanned work roster that offered limited opportunity to sleep, in spite of his working hours closely conforming to the minimum requirements of the network service provider.
Investigations into previous train control incidents have led to recommendations about the potential use of existing onboard train technology to give train control live tracking of train locations, which could have helped avoid this incident by showing the existence of the second train in spite of the signalling failure.
KiwiRail management has introduced a new fatigue policy since the incident, and it has previously responded to fatigue related recommendations resulting from investigations into previous incidents. Since this incident the Transport Accident Investigation Commission has made 3 new recommendations to the rail regulator concerning train controller rostering, shift handover procedures and the retraining of train controllers after extended breaks from operating critical systems.
(Note this executive summary condenses content to highlight key points to readers and does so in simpler language and with less technical precision than the remainder of the report for the benefit of a non-expert reader. Expert readers should refer to and rely on the body of the full report.)
Related Recommendations
The trainee controller having to refresh his knowledge in double-line automatic signalling rules at the same time as undergoing desk certification complicated the desk certification process, requiring more handovers during a shift. There would be some safety benefits in KiwiRail’s training system for train controllers in specifying a maximum period for which new and returning train controllers can be away from using critical systems such as double-line automatic signalling before requiring refreshing in these systems.
Standards for maximum working hours and minimum rest periods on the train control roster could result in a train controller being fatigued at the start of a shift even in the absence of any other fatigue-inducing factors. The train control roster policy including, but not limited to, standards for maximum working hours and minimum rest periods should be reviewed to ensure it is designed to mitigate fatigue and promote wellness.
During this incident, 2 successive train control handovers were conducted in a fashion that did not detect the presence of all train movements, and on each occasion the trainee, then the incoming controller, assumed the responsibility for the desk before they had acquired full situational awareness of the status of all movements within their area. The area of train control would benefit from some form of external review of processes to identify if these practices were confined to this one event or symptomatic of a wider systemic issue.