London Waterloo station was recently the subject of a 24-day partial blockade as significant upgrades were made to the station. Platform extensions, trackwork modifications and signalling upgrades were all undertaken as London’s busiest station underwent its first major upgrade since the arrival of Eurostar in 1994.

Scheduled to last from Saturday 5 August to Tuesday 29 August – Monday was a bank holiday – this was a major piece of work. The five organisations that make up the Wessex Capacity Alliance are well integrated and the planning and preparations had been meticulous. Nothing was left to chance.

As a result, nothing should have gone wrong. Yet it did. The full reasons won’t be known until at least two enquiries are completed, but here are some early indications.

Network Rail CEO Mark Carne during a visit to the scene. Credit: Network Rail.

Network Rail CEO Mark Carne during a visit to the scene. Credit: Network Rail.

Slow-speed crash

The first sign of trouble came early in the morning of Tuesday 15 August, ten days after work commenced. The 05:40 to Guildford, 10-car train made up of a combination of Class 455 and 456 units, pulled out of Platform 11 on time.

Two minutes later, having reached a speed of 11mph, it veered to the left, struck a train of empty Network Rail wagons, and came to an immediate halt. Of the 23 passengers and two employees of South West Trains that were on the train, only three were treated at the scene by paramedics and none required hospital treatment.

An early investigation by the Rail Accident Investigation Branch (RAIB) revealed that the points were misaligned and had directed the passenger train away from its intended route. The misalignment was a consequence of a temporary modification to the points’ control system.

Use of the term ‘misalignment’ indicates that the points were not set for a particular route. Furthermore, for the train to take the wrong route, the gap between the wrong route’s stock rail and point blade must have been sufficient for the wheel flange. This implies that the points were around mid-position as the train left the platform.

The question then arises why should the points be in such an abnormal position. Two possibilities are that the trailing points were incorrectly set for the incoming empty stock move into platform 11 which then burst them, leaving them in mid position, or that, for some unusual set of circumstances, the points moved mid-position after the empty stock train arrived in platform 11.

Worryingly, this was a failure of the signal interlocking’s detection system as the initial RAIB release states that the train driver and signaller received indications that the points were correctly aligned, and thus locked.

The RAIB will conduct a full investigation to identify the circumstances leading to installation of the temporary control system modification, the safety measures provided while the temporary modification was in place, the checking and testing procedures applicable to the modification and any relevant underlying management factors. Although it will take several months to publish the formal report, the serious nature of this incident is such that the lessons from this incident for ongoing signalling project work need to be understood as a matter of urgency.

One key issue is the safety measures that were in place. Had the points been clipped and secured, the accident would not have happened. However, the Rule Book does not require this as the Person in Charge of the Possession only has to confirm with the signaller which routes are to be kept closed. It will be interesting to see how the RAIB views this requirement.

Those who planned the Waterloo works clearly considered that some form of physical protection was necessary. Hence wagons were deliberately placed to protect the workforce behind them from the live railway. It was a step well taken – without them the diverted train could have ploughed into people working on the station improvements. However, had the points been clipped there would not have been a derailment.

Credit: Tristan Appleby/Network Rail.

Credit: Tristan Appleby/Network Rail.

Testing overrun

Having recovered from the delays caused by the accident, news then broke, early in the morning of Tuesday 29 August, that the engineering works had overrun. All lines were open by 07:20, but the ensuing disruption lasted all day. Word was that this overrun was due to extended signalling testing.

There were more delays, albeit short lived, on the morning of Wednesday 30 August after a track circuit failure closed Platforms 1-3. Once again, there was some residual disruption even though the fault was cleared by 07:38.

Were these faults connected to the accident two weeks earlier? In a way that’s possible – having experienced such a serious fault, the signal checkers were no doubt particularly diligent and they had also been delayed starting their job. Being the last step in the process, it’s the testers that get the blame although the problem could have been the delay caused by the accident knocking-on through the project timeline.

Signal testing is performed to make sure that there are no remaining faults before a line is returned to service. If a fault is found, that’s a good thing as it can be corrected and the travelling public isn’t put at risk. But that correction takes time, and that’s a bad thing, although it’s far better to wait and only return a line to service when it is 100 per cent safe to do so.

And the track circuit failure the following day? Unfortunately, signalling faults do occur. On the 30th there was also a signalling fault between Farnborough Main and Basingstoke, which delayed Southampton-bound trains. The day before, a points failure at Kew Bridge and another at Portsmouth Harbour caused problems and, to cap it all, a broken-down train between Leatherhead and Effingham Junction closed all lines while a failed freight train between Eastleigh and Southampton Central did the same. So the Waterloo problem could have been mere happenstance.

It could also have been another example of new work disturbing old installations and revealing or causing faults. Only a full investigation will give all the answers, and Network Rail and RAIB are undertaking theirs as this is written. However, the barrier train did its job, no one was hurt, and a massive amount of work was achieved over the course of 24 days.


This article was written by Nigel Wordsworth.