Philip Haigh
Echoes of Clapham at Waterloo.
“I hope RAIB does not spend its usual year before publishing Waterloo’s report. That something went wrong is clear, but the potential consequences of errors being made in temporary signalling changes are too great to wait.”
MONDAY December 12 1988 dawned cold and clear in South West London. It would end with commuters reading a banner headline in the Evening Standard- RUSH HOUR DISASTER - and a cover picture of crumpled coaches.
Two trains collided at 0810 that morning, and a third struck the wreckage seconds later. Thirty-five people died and nearly 500 were injured.
The accident shook British Rail. It occurred after Driver McClymont of the 0718 Basingstoke-Waterloo stopped at signal WF47 close to Clapham Junction, to report that the previous signal WF138 had reverted from green to red as he was 30 yards from it.
With the 0718 standing at WF47, the signal behind (WF138) should have been showing red. It was not, and this allowed the 0614 PooleWaterloo to approach, which it did under Driver Rolls’ control. Despite an emergency brake application, the 0614 struck the rear of the 0718 at around 35mph. Just as this happened an empty stock train, the 0803 Waterloo-Haslemere, passed on the adjacent line. Knocked sideways by the collision, the front coach of the 0614 from Poole hit the second coach of the empty stock train.
A fourth train was about to pass WF138, which was showing a single yellow despite the two trains beyond it. Driver Pike was 250 yards from WF138 when he spotted the danger and braked to stop just 60 yards from the back of the Poole train. When Pike’s guard walked back to check WF138’s aspect it was still showing a single yellow.
BR quickly admitted to faulty signalling. WF138 had been brought into service only the weekend before, following alterations in the relay room of Clapham Junction A signal box as part of BR’s Waterloo Area Resignalling Scheme (WARS).
Such was the severity of the accident that the Transport Secretary appointed Anthony Hidden QC to conduct a Judicial Inquiry (the first into a collision since 1876). His report found weak management and failings in preparation, supervision, inspection, testing and checking. He found inadequate training and problems from staff shortages and excessive overtime.
The immediate cause of the faulty signalling was a wire in the relay room that had been disconnected from a relay two weeks earlier. It should have been disconnected at both ends but was only disconnected at one. Although it had been pushed away from its old terminal, it had been connected since the 1930s, and when it was further disturbed by work on December 11 it moved back to its old position. When touching the terminal, it allowed current to flow that kept signal WF138 at yellow when it should have been red. No one checked that these changes had been correctly done.
WARS had started by resignalling Waterloo itself in stage 1 in 1984. This work involved shortening Platforms 1-4 to allow BR to use standard (rather than specialist) components in points.
It is these platforms that Network Rail has just spent most of August making longer. NR’s challenge was made harder by a derailment and collision in the middle of the closure, when August 15’s 0540 Waterloo-Guildford struck a stationary train of empty ballast wagons placed to protect NR’s engineering worksite.
An initial bulletin from the Rail Accident Investigation Branch (RAIB) on August 30 reported that the points the train was crossing were misaligned, even though the signalling system showed driver and signaller that they were correctly positioned.
This was a consequence of temporary modifications needed to allow tracks to be disconnected as part of the platform lengthening work. RAIB’s investigators spent
considerable time combing over what changes had been made in the relay room controlling this section of line.
RAIB’s brief words sound a disturbing echo from that awful December morning, just four miles down the line at Clapham Junction. I hope RAIB does not spend its usual year before publishing Waterloo’s report. That something went wrong is clear, but the potential consequences of errors being made in temporary signalling changes are too great to wait.
The jolt to BR’s safety culture and its signalling and telecommunications (S&T) department from the Clapham accident was nearly 30 years ago. That culture has changed beyond recognition since then - initially under BR, which had suffered two more fatal collisions (Purley and Bellgrove) in the months that followed Clapham.
Southall (1997), Ladbroke Grove (1998) and Hatfield (2000) then reminded the privatised railway in the most horrific manner that there was more to do. Britain’s railway is now the safest in Europe with many, many more passengers travelling than back in 1988.
NR completed its Waterloo project on August 29, the scheduled day but a few hours late such that trains were delayed. In the days that followed, signalling proved troublesome with further line closures needed to ensure it worked properly. This took the shine from NR’s considerable efforts in extending platforms for ten-car trains.
Disruption is inevitable if parts of the railway are to be rebuilt to allow longer trains, more trains, or both to run. NR came frustratingly close to completing Waterloo without unplanned disruption. Given August 15’s accident, NR did well to keep the overrun so short, but how it must wish there had been no delay and no adverse headlines.