Train EG37 was approximately 2,000 meters in length and weighed 13,032 tonnes. The Driver had made two brake applications prior to Connors range and had experienced no problems. Shortly before reaching the top of the range the Train Crew contacted Train Control and were told the line was clear to Yukan.

On reaching the top of the range the train slowed to 35 kph and reduced speed a further 15 kph when the Driver notched down to idle. EG37 slowed a further 3kph to 17 kph before gaining speed descending the range. The Driver began to apply dynamic brake at the top of the range but it had not fully applied before the train started to gain speed.

When the train speed had reached 32 kph the Driver made an initial brake application, (a reduction of brake pipe air to 450 kpa), as this had no effect further reductions were made. At 48 kph a full service brake application was made. These brake applications had no effect on reducing the speed of EG37. At 52 kph an emergency brake application was made but the train accelerated to 93 kph. At this point the train had entered a left hand curve where the front locomotives and 28 wagons separated from the remainder of the train. This portion came to a stop approximately 2 kilometres down the track.

The leading end of the remainder of the train had derailed on the left hand curve. A total of 73 coal wagons, 2 locomotives and one ELRC unit were destroyed. Estimated cost for rolling stock replacement was $20,000,000. There was also 700 meters of rail, 500 sleepers, 1000 meters of overhead wiring and a landslip detector destroyed. The estimated cost of recovery and the subsequent investigation was in the vicinity of $725,000.

There where found to be 2 contributing factors to this accident. The first was a communication interrupt between the Electric Locomotive Remote Control (ELRC) unit and the command locomotive at the lead of the train. These interrupts are common at the top of the range but usually only last for a short time period. This was not the case in this circumstance. The second contributing factor was an errant O-ring, which had lodged in a figure 8 shape across the seat of the brake pipe cut off valve in the ELRC unit, preventing brake pipe isolation from the 2 remote locomotives, which were still trying to pump the brakes off.