An interim account of June’s passenger-train collision near Elstow has set out the most detailed sequence yet of the moments before impact, while leaving the central questions of cause and responsibility for the continuing safety investigation. The collision occurred on 19 June, about four kilometres south of Bedford station on the Midland Main Line.
Two services, one unexpected stop
At about 17:13, the 16:40 service from Corby to London St Pancras, designated 1H46, struck the rear of the 15:50 Nottingham to London St Pancras service, 1B67. Both trains were operated by East Midlands Railway. The leading train was stationary beneath the gantry carrying signal WH152 when it was hit.
The report says 1B67 had made an emergency brake application after receiving an Automatic Warning System, or AWS, warning at signal WH154. Footage showed the signal was green at the time, when the system should instead have provided a clear indication. The warning was not acknowledged within the 2.7-second response period, so the brake applied automatically. The train then stopped, and the signal behind it displayed red to protect the occupied line.
Investigators have identified 257 people on the two trains so far. The driver of 1H46 sustained fatal injuries. At least 160 people were injured and 102 were taken to hospital. Both trains derailed, although the vehicles remained broadly upright.
Recorded data from the following train
Forward-facing footage from 1H46 shows that WH154 was displaying red as the train approached. Its onboard recorder shows the AWS warning was acknowledged, but the train continued to accelerate and passed the signal at about 58mph. The same data indicates that full service braking began about 10 seconds before the collision, roughly 260 metres from the rear of the stationary train. An emergency brake application followed about two seconds later. The recorded impact speed was about 50mph.
Those observations describe what the available records show; they do not establish why the red signal was passed. The report notes that the recovered recording data has some limitations, including corruption during the collision, and that its timings carry a small degree of inaccuracy.
A separate warning-system issue
A further occurrence on 9 July has become part of the inquiry. Another Class 810 train approaching WH154 received an AWS warning while the signal was green. On that occasion, the warning was acknowledged within the required period and the train continued after the matter was reported.
Follow-up work found that an electromagnet in the AWS equipment at WH154 was 16mm below railhead level, outside the stated tolerance of plus or minus 12mm. The investigation says work is continuing to understand the interface between that equipment and the leading train’s systems. It has found no evidence that the electromagnet was not correctly energised. That later occurrence should not be treated as a concluded explanation for the collision.
What remains to be established
WH154 had no Train Protection and Warning System equipment, and no signal-passed-at-danger alarm was associated with it. The report says the signal was not legally required to have that equipment and had previously been assessed as low risk. Investigators will now examine the actions and factors affecting those involved, the braking and warning systems on 1H46, the stop of 1B67, signal visibility and performance, the emergency response and the way overrun risk was assessed at this location.
No final cause, finding of fault or safety recommendation has been made. The interim report is intended to preserve the evidence available so far; its conclusions may be revised as the full investigation continues.
