Nigel Harris

Waterloo derailment raises urgent and disturbing questions

“It is said that when the driver phoned the signallers to report the derailment and collision, they grasp what had h because they we looking at a pane showing everyth be in order.”

“My blood runs cold to learn that a passenger train crashed on open points that supposedly fail-safe signalling failed to detect.”

It’s been a roller-coaster summer for Network Rail project delivery. August’s three-week Waterloo blockade sort-of started OK and was ticking over, despite some disruption - then the early morning commuter train derailment on August 15 brought more than wheels off the track and threatened a major overrun, with concurrent political meltdown.

Thankfully the project was handed back on the booked day, albeit a couple of hours late, and we’ve seen intermittent disruption since. But overall it wasn’t a disaster, and I have great respect for the ‘Orange Army’ who achieved a complete rebuild in a very short time.

It was then very satisfying to see the massive Thameslink possession (August 26September 2) handed back ahead of schedule, followed by some 99% PPM running and positive feedback from drivers about the 10km of new track, including the impressive Bermondsey dive-under. This was communicated well by NR’s increasingly effective social media team, who posted a fascinating driver’s eye view.

It was good to see these developments and all the Network Rail teams involved deserve credit - the long-running London Bridge/ Thameslink upgrade has been massive, but sadly it will still be recalled in Google searches for how it went wrong rather than the benefits it is delivering. Because it was handed back early with no fuss or delay, the Thameslink possession naturally barely featured in national news. The NR social media team does great work, but there’s still a mountain to climb to rebuild rail’s public reputation.

The Rail Accident Investigation Branch’s (RAIB’s) interim report into the Waterloo derailment and collision, which very nearly caused the blockade to overrun (staff were suspended and a crucial relay room quarantined for investigation), raises disturbing questions. To set context, Philip Haigh looks at this on pages 56-57, and he explains how a signalling technician’s wiring error led directly to the Clapham disaster of December 12 1988 that wrecked three trains and killed 35 people.

In the Hidden Inquiry into Clapham, the mood was that BR had been pressurising its technicians to do too much, too quickly, and that the stray wire that moved and made contact where it shouldn’t and caused the crash was a result of this high pressure. The outcome was that BR’s signal engineering function retreated into its shell and the Institution of Railway Signal Engineers (IRSE) took tight control of the signal testing process that emerged - and IRSE has maintained that rigid control ever since. This signal testing system passed in turn from BR through Railtrack to Network Rail, with most testers nowadays employed not by NR but by contractors.

I do not for a moment question that safety remains paramount, but increasingly over the past few years I have heard terms such as ‘cabal’, ‘stranglehold’, ‘law unto themselves’ - even ‘arrogant’ - used in reference to this very important job and its scarce practitioners.

To be honest, I hadn’t paid much heed and came to regard such ripples as the inevitable and harmless consequence of what happens when enthusiastic engineers proudly want to commission a job which operators are equally keen to see handed over, while those who have the responsibility of signing it off say: “Hold your horses, let’s make sure it’s safe first.”

But it seems to me that the Waterloo incident is a canary in the mine. Something just isn’t right. We need urgent answers - and action. I am alarmed by the early facts emerging through RAIB and the disturbing implications and questions they raise, specifically for signal testers and the IRSE.

RAIB reports that a set of points in Platform 11 were “misaligned”, and that as South West Trains’ 0540 Waterloo- Guildford departed it derailed on the half-open points, slewed to the left and collided with the barrier wagons on the adjacent line, protecting the worksite.

Such facts as we already know surrounding this incident horrify me - because both the empty stock movement into the platform and the passenger departure which followed were properly managed. The EMU ran into the platform under clear signals, running through the points which were wrongly set in the trailing direction. The train was then properly dispatched outbound, under clear signals.

It is being said that the Wimbledon signallers were looking at a panel which showed nothing wrong, but which had allowed an empty coaching stock movement in and then a loaded passenger movement out of Platform 11 over a set of points which were not merely unlocked, but half-open. The driver deserves our thanks. He noticed the open blades and braked - he was too late to stop in time, but his vigilance meant that he hit the barrier wagons alongside at a reduced speed of just 7mph.

But here is the most frightening thing. It is said that when the driver phoned signallers to report the derailment and collision, they struggled to grasp what had happened because their panel showed everything to be in order. Points locked, signals cleared - nothing wrong. My blood runs cold to learn that a passenger train crashed on open points that supposedly fail-safe signalling failed to detect.

As I see it, the questions which Chief Inspector Simon French’s RAIB investigators must now answer are profound. The IRSE and its signal testers must promptly explain:

How did a set of points on an operational passenger line OUTSIDE the possession area move, but come to be left lying in the wrong direction?

Why did the supposedly fail-safe detection system fail to detect that these facing points were closed and locked in the wrong direction?

How was it possible to set the route and clear signals for both ECS and passenger movements into AND out of Platform 11, over a set of points that were not merely unlocked, but open?

Why did the Wimbledon panel display indicate that all was well, for both movements, even after the 0540 to Guildford had derailed and crashed?

Why was this set of points, immediately beyond the barrier train separating worksite from operational platforms, not clipped and scotched to prevent movement?

That this happened at all is deeply worrying. That it happened as a result of actions during mid-blockade signal testing by specialists whose fundamental purpose is to ensure that precisely such an occurrence is impossible, strikes at the heart of what keeps us safe. Clapham’s failings over a rogue signal aspect leading to 35 deaths led to a justifiably intrusive and disruptive signal testing system. At Waterloo, that system failed and led to a driver and signaller seemingly being told all was well by the signalling when a facing point lay open in front of an approaching passenger train. We need to know why from RAIB - and urgently, too. Was it a mistake or incompetence? And who watches the watcher?