Accident at Clapham

At 8:10 a.m. on Monday, 12th December 1988, a crowded commuter train ran head-on into the rear of another train standing in the cutting just to the south of Clapham Junction station. After the impact the second train veered across the tracks and struck a third train, resulting in the deaths of 35 people with nearly 500 injured, 69 of them seriously.

The first train was the 7:18 a.m. service from Basingstoke that was being held at a signal on the up fast line. The second train was the 6:14 a.m. service from "Poole" (actually started from Branksome due to some previous vandalism on the track) which was proceeding under clear signals on the up fast line behind the Basingstoke train. What Driver John Rolls of the "Poole" train would have thought when his train rounded a sharp curve to reveal the Basingstoke train at a stand a short distance ahead can only be imagined. Despite applying full emergency braking there was no way the train could be more than slowed and the inevitable collision occurred. The force of the collision derailed the leading coach and forced it across the tracks to the right where it caught the 8:03 a.m. empty coaching stock train to Haslemere, on the down fast line, a glancing blow, partly de-railing the Haslemere train. This second collision almost certainly prevented the "Poole" train from moving further across the tracks.

The most serious injuries, and all the fatalities, were to the occupants of the first three of the "Poole" train's coaches. The front third of the leading coach was totally disintegrated whilst the rest of this coach, and the one behind it, had the left-hand side ripped off. The rear coach of the Basingstoke train was lifted bodily into the air and came down lying on it's left-hand side on the embankment above the wall of the cutting whilst it's rear bogie landed on top of the third coach of the "Poole" train, going through the roof above that coach's luggage compartment.

The subsequent enquiry endeavoured to establish the causes of the accident, and came to the following conclusions:

The signalling system had failed because during alterations to the signalling system a wire that should have been removed was not. It was still in the system and was making an electrical contact with its old circuit and was therefore able to feed current into the new circuit when the new circuit should have been dead. That current prevented the signal from turning to red.

The Weekly Operating Notice which the Southern Region of British Railways (BR) issued to train crews to keep them up-to-date had an entry in the issue for Saturday, 10th December 1988 which read:
"Signal WA25 has been abolished and a new 4-aspect automatic signal WF138 has been provided,. . . ."
It was that new signal WF138 which two days later in the morning rush hour of Monday, 12th December failed to prevent a second train from occupying the same track as an earlier one, and failed to stop the front of the second from running into the back of the first.

The basic premise behind the signalling was one of failsafe, i.e. if anything were to go wrong then the system should default to a safe condition, otherwise known in the rail industry as a right side failure. What happened at Clapham was not failsafe as when something did go wrong it allowed a signal to show a proceed aspect when it should have been showing a red aspect. A situation such as this comes under the term of wrong side failure.

Save for one circumstance, the track circuit (DL) running between Clapham Junction "A" Signalbox relay room and signal WF138 should have had current running permanently through it. That circumstance would have arisen when the wheels of a train had just passed signal WF138 and had therefore moved on to the stretch of rail through which runs the appropriate track circuit (DL). When those train wheels reached the section of track the current should have shorted out, the track repeater relay should have lost its energy, its contacts open, and the signal should have changed to red. The signal would have been similarly affected by the movement of a train onto the adjacent track circuit, DM. If a signal which is being operated by a relay is sufficiently close to the signal box, then there will only be the need for one track relay (TR) in the track circuit, controlling that signal. If, however, the signal is further away there may be a need for a track repeater relay (TRR) to carry out the necessary function
However, the relay could only perform its function as a switch cutting off the current provided that it was not by-passed, and therefore rendered inoperative by any false connection of wires or any false feed of electricity.

An unintended flow of current, technically described as a "false feed" of current, removed from the signal the ability to "see" the train on the track that it was guarding. The Basingstoke train therefore became invisible to the signalling system. The signal, not knowing the train was there, did not turn to red to protect it and instead continued to show a proceed aspect and did nothing to stop entry onto that section of the track occupied by the Basingstoke train. It did, in fact, invite the "Poole" train onto the track by showing a proceed aspect

But how had this happened? The final work necessary to replace the old signal WA25 and bring into operation the new signal WF138 was carried out on Sunday, 27th November 1988. It involved two wiring jobs known as job No.104 and job No.201 which were both part of Stage 7B of the Waterloo Area Resignalling Scheme. The physical preparation for those two jobs had to be done during the working week before Sunday, 27th November 1988, and the actual completion of the work had to be done on that Sunday by connecting the wires. The conceptual preparation had been done much earlier in that the Design Office had prepared and issued the wiring diagrams for those job numbers

Under the previous wiring system an old wire ran from a relay to a fuse. That relay was called TRR DM because it was the track repeater relay for track circuit DM. Under the new system for the new signal WF138, the circuit was to go from TRR DM to the fuse by a different route, which was to include a further relay, TRR DL, the track repeater relay for the adjacent track circuit DL. New wiring had been prepared during the week to run from the relay TRR DM to TRR DL and then on from TRR DL to the fuse. The S&T engineer's task at the weekend had been to connect those new wires and disconnect the old wire. That disconnection should have been made at both ends of the old wire, both at the relay end at TRR DM and at the fuse end.
In fact no disconnection was made at the fuse end and, although at the other relay end the old wire was disconnected, it was not cut back as it should have been, nor was it secured out of the way of its previous contact. Although it was pushed away to one side, the wire was accordingly left long enough and close enough to its previous contact for it to be physically possible for it to return to its old position if the wrong circumstances arose.

Two weeks later, on Sunday 11th December 1988, some unrelated work took place to change over another relay, TR DN, which just happened to be immediately to the left of TRR DM. Whilst this was being done the position of the wires to TRR DM was disturbed, particularly that of the old wire between the fuse and the relay TRR DM, with disastrous consequences the following morning.

Now that its position had been disturbed by this Sunday's work, the old wire that had not been disconnected from the fuse, nor cut back or secured away from its old contact, was able to move back into its old position, and once there was able to make metal-to-metal contact with its old terminal and permit current to flow from the relay TRR DM to the fuse. As with the error of two weeks previously, this error went undetected.

Following the work the track was put back into commission between 4 and 5 o'clock on Monday morning and all was well for a while as trains kept running through the cutting without any delay. Not until Driver McClymont of the Basingstoke train stopped at a red signal and telephoned the signalbox, with the certain knowledge of a railwayman that the signal behind him had to be red. It was not. It had not "seen" Driver McClymont's train and continued to show a proceed aspect inviting the following train, the "Poole" train, to proceed towards the back of the Basingstoke train, which was concealed behind a left-hand curve.

Driver McClymont had had to make an emergency stop when signal WF138 changed from green to red as he was about 30 yards from it, resulting in him passing the signal at red and, once stationary, requiring him to telephone the Signalman and not to proceed until authorised to do so. The sudden change of aspect was because the preceding train had moved from the faulty track circuit, DL, where it was "invisible" to signal WF138, on to track circuit DM, which also controlled WF138 and was working properly. This was the first indication that anything was wrong.

There was a fourth tain involved, also on the up fast line and which had also been shown an erroneous proceed aspect by the faulty signal WF138. However, Driver Pike of this train saw the rear of the "Poole" train ahead of him and, with the aid of an emergency brake application, managed to come to a halt some 60 yards from the "Poole" train. Had he not done so the accident would have been even more horrific.

The Memorial that was erected for the victims of the Clapham Rail Disaster at Spencer Park.

Photograph by Gregory Beecroft

Clapham Memorial
Clapham Memorial The other side of the memorial showing the inscription.

Photograph by Gregory Beecroft

A close-up of the inscription.

Photograph by Gregory Beecroft

Clapham Memorial

This article is dedicated to the memory of all those who died in the accident and in particular that of Paul Perry-Lewis who was a business aquaintance of the SEmG webmaster.

The full report of the accident may be found at:

All photographs are copyright

return to Accidents index page

return to picture gallery page

This page was created 31 May 2009

SR Target

Valid CSS!    Valid HTML 4.01!