PURLEY 

4th MARCH 1989

Involving Littlehampton Driver 

Bob Morgan

adapted from extracts of the D.o.T. report

by A. Cookson

Deputy Chief Inspecting Officer of Railways

On Saturday 4 March 1989 at Purley in the Southern Region of British Railways. At 13.39, the 12.50 train from Horsham to Victoria was struck in the rear by the 12.17 train from Littlehampton to Victoria. The train from Horsham had just departed from Purley Station and had crossed from the Up Slow line onto the Up Fast line. The train from Littlehampton had been travelling at speed along the Up Fast line. The leading 6 vehicles of the 8 coach train from Littlehampton were derailed and deflected to the left down an embankment. The seventh coach of the train from Littlehampton was also derailed as were the rear two coaches of the win from Horsham.

Five passengers were killed and 88 persons, including 3 railway staff, required hospital treatment. Of the injured, 32 wen: detained in hospital, some with serious injuries but by the time I opened my Inquiry only 4 remained in hospital.

The accident caused considerable disruption to rail services. The Up and Down slow lines, which were blocked to traffic to facilitate the recovery work, were re-opened at 05.17 on 6 March. The Up and Down Fast lines were re-opened, subject to a 20 mile/h speed restriction and without the damaged crossover being replaced, at 15.43 on 6 March. The removal of the vehicles from the side and bottom of the embankment was a difficult task and the last one was taken away by road on 9 March. The re-instatement of the crossover was completed and normal operations were restored at 08.00 on 27 March.

The Site of the Accident

Purley Station is located some 13 ¼ miles south of London on the line from London (Victoria) to Brighton and the South Coast. To the north of Purley Station the line is carried some 20m above the surrounding residential area on a tree covered embankment. There are four tracks, from west to east they are the Up Fast, Down Fast, Up Slow and Down Slow lines respectively. The Up direction of travel is towards London. Purley Station has six platforms with the additional Up and Down Loop lines lying to the east of the other four lines. Immediately to the north of the station there are double ladder crossovers from the loop lines to the Slow lines and then to the Fast lines.Tthe accident occurred at the point where the crossover from the Down Fast line joined the Up Fast line.

Immediately to the south of the station the loop lines diverge into two double track branch lines to Caterham and Tattenham Corner. The line to Tattenham Corner pass beneath he Fast and Slow lines. The Fast and Slow lines continue to Stoats Nest Junction, approximately a mile south of Purley Station, before splitting to form the alternative Redhill and Quarry line routes.

The maximum permitted line speed on the Fast lines in the vicinity of Purley is 90 mile/h. The maximum permitted speed through the crossovers to the north of Purley Station is 25 mile/h. The railway is electrified on the 750 V dc conductor (third) rail system. In the area of the accident the traction current is supplied from a substation at Purley which is remotely supervised from a control room at Selhurst. From this control room it is possible to remotely switch  and monitor the electrical supply and distribution from several substations over a large area. The supervisory instructions arc carried by means of trackside cables which also transmit the state of the circuit breakers in the substations to the control room. Alternative electrical supplies am available to this area from adjacent control rooms including one at Brighton.

The Signalling Arrangements.

Train movements in the Purley area are signalled in accordance with the British Railways Board Track Circuit Block Regulations. All running signals controlling main line movements are of the four aspect colour-light type and are equipped with the Automatic Warning System (AWS). These signals are capable of displaying four different aspects, namely, a red stop aspect, a single yellow or double yellow caution aspect, or a green clear aspect. The correct sequence of aspects displayed by the signals. Where lines diverge the four aspects arc supplemented by junction indicators which take the form of a row of 5 white lights mounted above the main aspects and angled in the direction of the divergence from the main route. 

The signalling, which was installed in 1984, employs conventional relay interlocking. The interlocking for the Purley area is housed in a purpose built relay room adjacent to Purley Station. The signalling is, however, controlled from the Three Bridges Signalling Centre some 16miles south of Purley. Responsibility for the operation of Three Bridges Signal Centre is split between a number of signalmen each of whom control a specific section of railway. The overall supervision is undertaken by a Regulator and an Assistant Regulator. 

The whole of the area controlled by the Signalling Centre is shown on a continuous diagrammatic panel which displays the signal routes set, the track circuits occupied by trains and other displays relating to the lie of points, alarms, etc., so that each signalman is able to observe what is happening in the area for which he is responsible and also in the areas adjacent to his own. The identity of each train is shown on the display by alpha-numeric codes. These train descriptions are moved along the panel from signal to signal automatically by the passage of trains. The stepping forward of these descriptions is triggered by the train physically pausing a signal showing a proceed aspect. Should a train pass a signal at Danger its description will not step beyond the last authorised position but is retained there. 

The signalman sets the routes for trains by pressing first an entrance button and then an exit button For the mute required. When the route is physically set it is indicated by a line of white lights along the route on the panel. Signal indications for controlled signals ate shown on the panel by either a red light for a red signal aspect are by a green light for any proceed aspect; yellow aspects are not indicated separately from green aspects. Some running signals are either fully or semi-automatic in that the control of the signal is affected by the passage of a train. Without any action on the part of the signalman, the signal reverts to Danger as a train passes it and then as the train travels away from it passing other signals it changes to display progressively less restrictive aspects. 

1.8 Signal T168 on the Up Fast line which protects the crossovers immediately to the north of Purley station is a semiautomatic signal. In its automatic mode the signal operates for a succession of trains travelling along the Up Fast line. With it operating in the automatic mode the signalling interlocking is designed to prevent the setting of the crossover route from the Up Slow to the Up Fast line. The signalman is able to change the signal from the automatic mode to controlled operation at any time and the signal will then, after the passage of the next train, remain at Danger. The signalman would hen, subject to other controls allowing it, be able to set the crossover route. 

The signalman may also replace Signal Tl68 to Danger at any time. If he does so when a train is already approaching along the Up Fast line and has occupied any track circuit, the Comprehensive Approach Locking is designed to prevent the conflicting crossover route from being set until a 2 minute timing control has operated. This ensures that the route remains locked until sufficient time has relapsed for the approaching train either to have stopped at Signal T168 or to have passed it. If Signal T168 is replaced to danger before the train has occupied track circuit, the approach locking does not operate because the driver of the approaching train will not see a signal revert to a more restrictive aspect. Similar approach locking controls an provided for trains routed from the Up Redhill line to the Up Fast line through Stoats Nest Junction. 

The AWS provides both an audible and visual warning to the driver of the signal aspect. It is operated by magnets positioned between the rails approximately187m before the signal to which they apply. With the signal displaying a green aspect a bell sounds and the indicator displays an all-black disc. The driver is not required to the AWS for a green aspect. With the signal displaying a red, yellow or double yellow aspect a warning horn will sound. The driver has to acknowledge the AWS by depressing a button which silences the warning horn and causes the indicator disc to display black and yellow segments as a reminder to the driver. If the driver does not acknowledge the warning within 3 seconds the brake of the train will be automatically applied. The AWS system does not distinguish between red, single yellow m double yellow aspects. 

The Trains

The 12.50 train from Horsham to Victoria was a &car electric multiple-unit (EMU) No. 3441 of Class 423. This class of EMU was introduced in 1967. The formation of the train was as follows: 

Coach 76378 (leading). Driving Trailer Couch with 18 First Class seats in compartments and 38 Standard Class seats in an open saloon and 8 in a compartment. 

Coach 62261 Motor Coach (Non-Driving) with Guards Brake Compartment. 58 Standard Class Seats in two open saloons.

Coach 70894 Trailer Coach with 98 Standard Class seats in an open saloon. 

Coach 76377 Driving Trailer Coach with 18 First Class seats in compartments and 38 Standard Class seats in an open saloon and 8 in a compartment.

Each vehicle was 19.74m long. The driving wailers weighed 35 tonnes, the trailer 31.5 tonnes and the motor coach 49 tonnes giving a total weight of the train of 150.5 tonnes. The train was time tabled to depart from Purley Station at 13.34 and to cross from the Up Slow line to travel along the Up Fast line.

The 12.17 train from Littlehampton to Victoria was formed on departure from Worthing of two 4-car EMU of the Class 421/2 which were introduced in 1970, The units were No. 1280 leading and No 1295 trailing and the formation was as follows:

Unit 1280 

Coach 76730 (leading) Driving Trailer Coach with 18 Fust Class seats (in compartments) and 36 Standard Class seats (in open saloon).

Coach 62368 Motor Coach (Non-Driving) with Guards Brake Compartment. 56 Standard Class sears (in open saloon).

Coach 71048 trailer Coach with 72 Standard Class seats (in open saloon).

Coach 76801 Driving Trailer Coach with 24 Fust Class seats (in compartments) and 28 Standard Class seats (in open saloon).

Unit 1295

Coach 76816 Driving trailer Coach with 24 First Class seats (in compartments) and 28 Standard Class seats (in open saloon).

Coach 71063 Trailer Coach with 72 Standard Class seats (in open saloon).

Coach 62383 Motor Coach (Non-Driving) with Guards Brake compartment. 56 Standard Class seats (in open saloon).

Coach 76745 Driving Trailer Coach with 18 First Class seats (h compartments) and 36 Standard Class seats (in open saloon).

Each vehicle was 19.74m long. The driving trailers weighed 35.5 tonnes, the trailers 31.5 tonnes: and the motor coach 49 tonnes giving a total weight of the train of 303 tonnes. The train was timetabled to pass Stoats Nest Junction travelling along the Up Fast line at 13.38 ½ .

The damage caused to the vehicles of the trains in the collision and subsequent derailment was as follows:

12.50 Horsham to Victoria

Coach 76378 - No damage (leading vehicle)

Coach 62261 - No damage

Coach 70894 - Trailing end of vehicle, left hand side (in direction of travel) badly damage internally and externally over the length of the bogie with a 1.5 metre length of bodyside totally disintegrated from floor and roof. The electrical control and heating jumper receptacle boxes destroyed and wiring- damaged in this area Other minor damage sustained along left side of the coach.

Coach 76377 - Extensive damage to trailing end of the vehicle, left hand side (in direction of travel) having taken the full force of the impact. From the driving cab, with its handbrake column sheared, the bodyside was totally disintegrated for or 3 metre length, with severe damage sustained for a further 3 metre length including door pillars, windows and all wooden stepboards. The first two seating areas were severely damaged and various windows throughout the remainder of the coach were broken. The underframe equipment, including the truss-bars (strengthening members of the underframe) and brake reservoir tanks, was badly damaged. Both bogie frames were twisted and their suspension damaged with the wheels of the leading bogie being torn away in the impact. 

Class 421/2 Unit No. 1280

Coach 76730 - Coach at bottom of bank with both bogies detached. Severe body damage to right side of cab front, trailing end of vehicle tom away and flattened to floor level and centre roof section torn away. Extensive damage to under frame cross-members and longitudinal truss bars. Both bogies twisted and suspension damaged.

Coach 62368 - Coach down bank with both bogies detached. Severe body damage with luggage area bodyside panels smashed in and roof section split. Extensive damage to underframe cross members and longitudinal truss-bars. Both bogie frames twisted and suspension damaged.

Coach 71048 - Coach down bank on its left side with bath bogies detached. Severe body damage at leading end and to left hand side of vehicle. Extensive damage to underframe cross members and longitudinal truss-bars. Both bogie frames twisted.

Coach 76801 - Coach down bank on its right side with trailing bogie detached. Severe body damage to corners of vehicle. Moderate damage to underframe drawgear and one solebar bent. Bogie brakegear bent.

Class 421/2 Unit 1295

Coach 76816 - Coach down bank on its left side held by the buckeye couplers. Moderate body damage to left side of vehicle. Underframe equipment damaged and bogie headstocks and brakegear bent.

Coach 71063 - Coach part way down bank on its left side held by the buckeye couplers Superficial &age to body, underframe and bogies.

Coach 62383 - Derailed all wheels but still on track bed. Superficial damage to body, underframe and bogies.

Coach 75745 -  No damage and not derailed.

EVIDENCE

Working of the Trains

Driver E W Sellwood (Eastbourne) drove the 12.06 train from Eastbourne to Victoria on the day of the accident. It passed through Purley traveling along the Up Fast line at about 13.06. Mr Sellwood said that Signal T182 displayed two yellows, Signal T178 a single yellow and Signal T168 a red aspect. He said Signal 7168 "Came off as I was crawling along the platform" and he had seen a train going round the curve in the line ahead of him beyond Signal T162. As far as Mr Sellwood was concerned the signals were functioning as they should have done.

Driver A JP Lawless (Victoria) drove a Gatwick Express along the Up Fast line through Purley at about 13.33 on the day of the accident and he said all the signals were clear. He was a regular driver on the Gatwick Express and that journey was the fourth he had made that day. The only problems he had experienced with the signalling in the Purley area was having to stop the train for a red signal which was then cleared to green.

Signalman D J Owen had signed on duty at 13.10 and taken charge of that part of the Three Bridges Signalling Centre panel which controls the Purley area. He had a signalman at Three Bridges since September 1983. When he commenced work on the day of the accident the panel was working correctly and no abnormal train movements were taking place and there was nothing to cause him any concern.

Signal T168 was in 'Auto' as the Gatwick Express was approaching Stoats Nest Junction and he took it out of 'Auto'. As the Gatwick Express went past Signal T168 the signal indication on the panel went to red and stayed at red. The indicator lights for the platform and overlap track circuits cleared as the Gatwick Express continued its journey towards Croydon. It was normal for the train from Horsham to be routed onto the Up Fast line behind the Gatwick Express and this he did once the Tattenham Corner train had left Platform No. 5.

2.5 He operated the entry button at Signal T170 and the exit button at Signal T162. He saw the panel indications for 1639 and 1641 points go to the reverse position and the route lights illuminated showing that the route was set. He was not sure of the position of the train from Horsham because the track circuit indication extends from Signal T180 to T170 and he knew only that the train was between the two signals. He watched the indications as the train from Horsham departed and as soon as it occupied the first track circuit beyond Signal T170 the signal went back to red.

2.6 All the track circuit indications on the panel went to red indicating a failure of some kind. He was told there had been a major accident by a driver using the signal post telephone at Signal T153. After about 30 to 40 seconds the indications began to return to the panel and then he could see what had happened. When the indications returned the lH05 description for the train from Littlehampton was in the berth for Signal T168 but there were no track circuit indications illuminated for that train. The only track circuit showing occupied was the train from Horsham was standing which was just on the trailing end of 1639 points.

As soon as Mr Owen was told of the accident he immediately told Mr Timms, the Regulator, and asked him to can the emergency services while Mr Owen used the direct telephone link to the Selhurst Electrical Control. He told the electrical controller what had happened and asked for the traction current to be discharged in the whole area. The electrical controller told Mr Owen that there was a loss of indications at the Selhurst control and that he could not confirm the current was off.

Mr Owen said that he had always intended to allow the train from Horsham to follow the Gatwick Express and run ahead of the train from Littlehampton and that he had not changed his mind. That was the normal pattern of services which was repeated each half hour and it was also the normal method of operation to take Signal T168 out of 'Auto' as the Gatwick Express was approaching or going over Stoats Nest Junction. That day, as the Gatwick Express passed through Purley station, Mr Owen said his recollection  was that the train from Littlehampton was on the country side of Stoats Nest Junction and between Signals T178 and T182.

He said it was not unusual for the trains from Horsham to be a minute or two late or for the trains from Littlehampton, which run non-stop from Gatwick, to be slightly early. He said, however, there was no point in allowing the train from Littlehampton to run before the train from Horsham because the train from Littlehampton would be held at Croydon waiting for the correct departure time and the train from Horsham would be delayed that much more. It was, therefore, not unusual for trains to be checked or stopped at Signal T168. Mr Owen said he would consider changing the sequence of trains if the train from Horsham was 5minutes or more late.

Driver V A R Brown drove the 12.50 train from Horsham to Purley. Earlier in the day he had driven a train from Horsham to London and back. There was nothing unusual about the second journey as far as Purley. As he approached Purley station Signal T170 was displaying a red aspect which cleared to a single yellow aspect with junction indicator lights for the route to the Up Fast line. He could not remember whether the signal cleared as the train ran into the station, which he said it often did, or if it cleared immediately the train came to a stand.

He thought the stop at Purley was for the normal length of time. The signal was still showing the same aspects when he started the train. He looked back along the train as it departed and in doing so had looked toward platform No.1 but had not noticed the aspect Signal T168 was showing. He drove the train over the crossovers at about 25 mile/h. The train was about three coach lengths onto the Up Fast line when he felt a series of violent snatches from the rear of the train. He shut off power and applied the brake but the train was virtually at a stand anyway.

He realised something serious had happened to the rear of the train. He got out of the driving cab on the embankment side of the train. At first he thought only his train was involved and it was some time before he realised another train was involved. He saw the driver of a Light locomotive standing on the Down Fast line jump down and use the telephone at Signal T153 and so he knew the signalman had been advised what had happened. By that time passengers were beginning to get out of the front of his train.

Mr Brown applied a short-circuiting bar on the Down Fast lime. In the meantime Driver Luxford (Three Bridges), who had been travelling as a passenger on the train, had used a telephone adjacent to the Down siding to speak to the signalman and he shouted across that the Selhurst Electrical Control could nor confirm the current was off. Mr Brown obtained another short-circuiting bar and applied it towards the rear of his train on the Up Slow line. Passengers were now beginning to get out of his train on the offside and he was concerned that the current was not off. He then, with other members of staff, did his best to ensure that passengers made their way to the Down side and then to the station without stepping on any conductor rails.

Guard D J Stanford was in charge of the train from Horsham which arrived at Purley a minute late, As soon as the train stopped he stepped from the guards van, which was at the rear of the second coach of the train, onto the platform and he saw that Signal T170 was showing a single yellow aspect with the junction indicator illuminated. There were quite a few passengers waiting to board the train and that took about a minute. One of those waiting was Guard Barnes who entered the guards van. Mr Stanford checked the doors were closed, the signal aspect had not changed and gave the 'Ready to Start' bell signal to the driver.

The train started and proceeded over the crossover at what he estimated was between 15 and 20 Miles/h. The train was nearly over the crossover and onto the Up Fast line when "there was a mash and a judder, an almighty crashing noise and a terrible juddering". He was being tossed around and could not recall what happened next. When the train had stopped he saw Mr Barnes putting down track circuit operating clips on the Down Fast line and he knew the train was being protected. He walked back through the train to see what damage had been done and tried to calm passengers and get them to stay on the train.

He noticed three passengers who were bleeding in the rear coach and one lady lying unconscious. There was another person with her and so Mr Stanford tried to calm the other passengers. He was concerned to keep them on the train because he did not know if the current was off. He got out on the embankment side of the train and walked forward to find his driver and found that passengers were getting out. He tried to make sure that they did not go near the conductor rails. Having spoken to Driver Luxford and learnt of the problems of confirming the current was off, he gathered passengers together and, aided by another member of staff who was a passenger on the train, he walked them back to Purley station along the Down Fast line on which a short-circuiting bar was in position.

Guard B M Barnes caught the train from Horsham at Purley to travel to Victoria where he was due to sign on duty at 14.00. He noticed, while he was waiting on platform No. 3 at the top of the staircase from the subway, that a Bedford service and then a Gatwick Express went past along the Up Fast line. As the train from Horsham approached platform No. 3 he saw Signal T170 clear to a single yellow aspect and the lunar lights 30 seconds before the train stopped. He boarded the train with Mr Stanford.

As the train went over the crossover road onto the Up Fast line there was a terrible juddering and he thought the train had been derailed. He looked to the left hand side of the train and actually saw the other train going down the embankment. Leaving the train on its right hand side, while Mr Stanford went to the left, he placed track circuit operating clips on the Down Fast line.

He got back into the brake van and got the ladder out. In the compartment next to the brake van was a gentleman with two children and he was shouting he wanted to get out. Mr Barnes placed the ladder on the embankment side of the train and got him, his children and a lady out. He walked to the front of the train when there were people standing and he thought they were being taken away. He went round to the other side of the train and walked back on the Down Fast line side to the third coach where he put the ladder up and got people out making sun: they did not touch the conductor rails. A gentleman got out and said he was the last one in the third coach and Mr Barnes made his way to the last coach. He was told by a policeman that there was a seriously injured lady and, placing his overcoat on the ladder to turn it into a stretcher, he helped him place her on it. She was taken away by an ambulance crew.

Driver R G Morgan was the driver of the tram from Littlehampton. In order to enable him to assist my Inquiry he was given a limited immunity from prosecution by the Director of Public Prosecutions. At his request, on the advice of his legal representatives, he gave his evidence in camera.

Mr Morgan had been a driver for 22 years and during that time he had not been involved in any serious incidents nor had he passed a signal at Danger when not authorised to do so. He said he had bad no domestic worries, had been in good health, had not been taking medication and had not consumed any alcohol. He had just returned from an annual holiday.

On the day of the accident he booked on duty at Littlehampton at 07.18 and his first trip was to drive an empty Stock train from Littlehampton to Lover's Walk at Brighton. The booked departure time was 07.57 and he departed from Littlehampton at about that time and arrived at Brighton at 08.40 or 08.50. He then travelled back to Littlehampton as a passenger on train at about 09.50. With the guard he went to a cafe and ate breakfast. He also bought two sausage rolls. He then made his way back to Littlehampton and spent the time until his next job in the driver's mess room. He had not slept or dozed. There were others there and he had joined in the conversation and had spoken to the supervisor.

Shortly before the 12.17 departure time of the train from Littlehampton he left the mess room and made his way to the train which was already in the station. He entered the cab and prepared for departure carrying out a brake test in conjunction with the guard using the internal train telephone. He took off his jacket and hung it up which is what he normally did. The heater in the cab was on and the windows closed. The departure was normal and he opened the sliding window to look back along the train as it pulled away.

The first station stop was at Angmering where he stopped the train at the four car mark. The platform was also on the left-hand side am it was at Littlehampton and again he looked back through the open window on departing. He repeated this at the next station. At West Worthing the signal at the end or the platform was at red and there was a wait until it changed to a proceed aspect. The signal at the end of Worthing platform was at red because a second train was to be coupled to the rear of his train. While waiting he ate the sausage rolls he had bought. The other train was coupled and another brake test was made. He thought the wait at Worthing was between 3 and 4 minutes. The signal cleared to a green aspect for the departure of the train. Because the platform was on the right hand side he did not look out of the window.

He then ran on green signals to Lancing but received a yellow signal at Shoreham where he stopped the train at the eight car mark with the platform on the left hand side. A normal departure was made with signals displaying green aspects. Again he opened the window and looked back as the train departed. The next station stop was Hove where the train was routed into the loop line with the platform on the right hand side of the train. When the train departed from Hove with a double yellow aspect signal, the next signal was a single yellow, but the one after that changed to green before the train reached it. There were then green signals to just before Gatwick where a double yellow changed to green before the train reached it.

At Gatwick Airport the platform was on the left hand side and Mr Morgan supposed the station stop lasted about 2 minutes The train departed under green signals and again he opened the window and looked back. The next thing he could recollect was a green signal in the 'coved way' but he had no recollection of the next two Signals Tl82 and T178. He said that as the train travelling at 60 to 70 mile/h, approached the end of Purley station platform he noticed Signal T168 at red and he immediately made a full emergency brake application. He believed he did not cancel the AWS but that the power was shut off.

He realised that the train would not stop at the signal and he just hung on. He told me that "it ran through my mind it should not be that colour". He said everything happened so quickly, his train hit the other train a glancing blow, veered off to the left and went down the embankment. When the coach stopped he slid out from a gap in the cab and fell out onto the ground. He said he was mumbling as he tried to think how it happened. He still could not explain or understand it.

Guard A H Squires worked the empty stock train from Littlehampton to Brighton with Driver Morgan and ate breakfast with him. They had read newspapers and discussed the news generally and he believed Mr Morgan was his normal self. He had then later in the day been the guard on the 12.17 train from Littlehampton to Victoria. His evidence agreed with the evidence given by Mr Morgan on the working of the train from Littlehampton to Gatwick Airport.
He gave the signal for the train to start from Gatwick Airport station from the guards van in the second coach of the train. After the departure he went into the third coach and commenced checking tickets. While doing so he met Mr Knights, a revenue protection inspector, and they agreed to share the duties between them and to meet again in the guard's van. Mr Squires was the first back to the guards van. When Mr Knights returned he stood in the corridor at the open door to the guards compartment while Mr Squires stood inside the compartment with his back against an electrical cupboard.
They were chatting when them was a sudden emergency brake application and Mr Squires' initial thoughts were that a signal had been replaced to Danger in front of the train. He saw the emergency application register on the brake gauge. He had returned to the guard’s van when the train was in the Coulsdon area and had there been an earlier brake application he would have been aware of it. There was a series of bumps and he said that "the next thing we were flying through the air" and the electrical cubicle doors and fuses were falling like "autumn leaves".

Mr Squires said he could not remember everything that happened after that but remembered struggling to open a door and getting out of the coach and being concerned about protecting the train and being told by others that it had been done. Latter he found Mr Morgan lying on the ground covered with a blanket and mumbling. There was someone he did not know in the remains of the driving cab and Mr Squires asked what he was doing and received the reply that "the driver wants his key". He told the person not to touch it. 

Driver M Brown (Norwood) was at the controls of a Class 47 locomotive travelling light that was brought to a stand at Signal T153 on the Down Fast line about 550 metres north of Purley Station. He saw a train depart from Platform 3 and move across his path from the Up Slow to the Up Fast and he then became aware of the approach of another train travelling along the Up Fast at a speed he considered too great to be able to come to a stand before it reached the convergent point. The last coach of the slow train had not completed its manoeuvre over the trailing points on the Up Fast when it was struck in the rear by the fast train. Before the wreckage came to a stand, Driver Brown had climbed down from his cab and he then went to the signal post telephone. He telephoned Three Bridges Signalling Centre and advised the signalman that a major train crash had occurred to the north of Purley Station and that it was an emergency. At about this time another passenger train came to a stand at an adjacent signal on the Down Slow line. Driver Brown placed detonators from his own locomotive on the track in order to provide protection. He could see that short-circuiting bars had been put down across the conductor rails but could not obtain confirmation that the traction current had been switched off. After warning passengers on the train from Horsham of the possible danger from the live rail, he assisted other railwaymen, who had by then appeared on site, in conducting the passengers along the track to Purley Station.

Guard S C Parr was in charge of the 13.40 Purley to Farringdon "Thameslink" service that was standing at Platform 6 awaiting its departure time when the accident occurred. Guard Parr, who was at the rear of his train, was told by his driver by means of the internal train telephone he could see that a serious train accident had occurred. Guard Parr immediately collected his emergency equipment and went towards the site of the accident. He saw that some track circuit clips had already been placed and he completed protecting those tracks that appeared not to have been dealt with. He also saw one traction short-circuiting bar in position and although he could not obtain an assurance that the isolation of the traction current had been confirmed he was advised that all the necessary protection had been carried out. He then went and assisted passengers to evacuate initially from an overturned carriage of the train from Littlehampton that was half way down the embankment and then the other train until the emergency services arrived and took over.

The most senior person present at Three Bridges Signal Centre was Regulator M T Timms who was being assisted by Assistant Regulator J.C. Underhill. The first indication they received that an accident had occurred was the indicated failure of the remote control signalling transmission system known as Time Division Multiplex (TDU) when an alarm bell rang and all the lights on the signal panel diagram shouted red. After about a half minute the signal lights reverted to normal and a few seconds later telephone messages were received by the Signalman Owen from Driver M Brown at a signal post telephone and by Assistant Regulator Underhill from staff at Purley Station. It was agreed that Mr Underhill would contact the British Rail telephone exchange in order to summon the emergency services. Using the appropriate emergency telephone number on the BR private line, he made contact with Miss J Hooker, a telephone operator at Waterloo. Because of initial confusion over the location of the incident, the caller was connected with the East Sussex Constabulary instead of the Metropolitan Police. It was over 5 minutes before Mr Underhill was connected with the Metropolitan Police. He then advised them that the incident was just to the north of Purley Station and could not identify the location nearer.

Meanwhile Mr Timms was telephoning Electrical Control Operator T L Foster, who was on duty in the Electrical Control Room at Selhurst to request an isolation of the traction supply in the area of the accident. Mr Foster was aware that there was something amiss when the Control Room lighting had dimmed and indications were received that two oil circuit breakers had automatically opened at the Croydon Sub-station that takes its supply from the National Grid. However, Mr Foster could obtain no indication of the state of the traction supply in the Purley area in spite of opening the circuit breakers controlling the immediate area. He therefore extended the area of isolation to the north by opening circuit breakers at the Selhurst Sub-station and requesting the Brighton Control Room operators to open the circuit breakers supplying the area from the south.

2.46 He explained that by 13.45, whilst he was confident that all the necessary circuit breakers had been opened, because he had no supervisory circuits available, he could not be absolutely certain that there was no train or other short circuit bridging the electrical section gaps and providing a stray supply to the area of the incident. He was therefore unable to give an undertaking to Regulator Timms that the conductor rails were not energised in the area of the incident and recommended that they be checked on site with test equipment. 

Signalling

Mr C Porter, the Regional Signal Engineer, explained how four-aspect colour-light signalling operated and in particular the comprehensive approach locking controls on Signal T168. He also explained the operation of the AWS and demonstrated the visual and audible indications given to a train driver. 

Mr Porter also defined various classifications or signalling equipment faults as follows:

Rightside Failures

All signalling equipment was designed to failsafe principles, which meant that with any failure of the equipment the system was designed so that, as far as practicable, the equipment fails to a safe condition and displays a more restrictive state to the driver or signalman than it would have done if it had not Failed.

These failures were referred to as Rightside Failures

Wrongside Failures A wrongside failure was a failure where something happens that should not happen and where the signalling system does not fail to a safe mode. There were some wrongside failure which were protected by other parts of the signalling system and are known as Protected Wrongside  Failures. For example, the failure of both filaments of a signal lamp was a wrongside failure but it would be 'protected' by the previous signal remaining at Danger when the controls on the previous signal railed to detect an electrical current flowing through the lamp of the other signal. An Unprotected Wrongside Failure was one which was potentially very serious and, for example, could be caused by a defective piece of equipment.

2.69 Mr N D Remfrey, a Technician Officer in the Signals and Telecommunications Engineers Department arrived at the Three Bridges Area Signalling Centre at about 13.55 to relieve another technician officer who informed him he had just had a telephone can from Brighton and had been told that the train describer had 'frozen'. They telephoned the Signalling Centre Regulator and were informed of the accident They made their way to the operating floor where they arrived at 13.58.

2.70 Mr Remfrey observed the indications on the signalman's panel. He said Signals T168, T170, T172, T174, T167, T153 and T154 were all showing a red indication as were shunt signals 1093, 1091 and 1096. He observed that the track circuit was showing occupied with the description 2CV7 (the train from Horsham) displayed within the track circuit.

Driving Technique

Mr G R Taylor, the Regional Chief Traction Inspector, was responsible for the maintenance of the standards of performance of footplate staff throughout the Southern Region. Assisting him were a team of 22 traction inspectors. Mr Taylor had been the Chief Inspector for 3 years, having been the Assistant Chief Inspector for 2 years, and a Traction Inspector for the previous 17 years. Prior to that he was a driver for 13 years.

He described to me the typical technique of drivers of trains similar to the train from Littlehampton on the journey from Gatwick Airport to Purley. Different driver's technique varied slightly and different techniques would be employed for different types of trains such as locomotive hauled trains or freight trains. From Earlswood to Quarry Tunnel, which is one mile and 353 yds long, there is a rising gradient. With power being applied normally on departing from Gatwick Airport station under clear signals the train would be travelling at 70 mile/h. At this point the train is a little under 4 miles from Purley and the gradient falls towards London. If the driver shuts off power at that point the train would be travelling at about 75 miles/h at Purley. Most drivers would not shut off power at that point.

After passing Signal T192, the next signal the driver would see would be the repeater signal for T190 provided because the minimum 7 second sighting time for that signal was not available, and then Signal T190 itself. The line then crosses over the Redhill line and it was on this section of line most drivers shut off power and the train would approach Purley travelling at 80 mile/h. The next signal is T188 and then comes the ‘covered way’, which used to be a tunnel, and if the controller had been kept open the train would be travelling at approximately 86 miles/h. After the ‘covered way’ there is the main Brighton Road and then the old sidings. Looking towards the centre span of a bridge which crosses both the Quarry and Redhill lines Signal T182 first becomes visible. On the approach to Signal T182 the next Signal T178 also becomes visible.

If on sighting Signal T182 it were displaying a double yellow aspect the driver would not immediately apply the brakes, he would reset the AWS but about 200 yds beyond the signal with the single yellow aspect of Signal T178 continuously in view he would start the brake application which he would continue, making allowances for the weather and rail conditions, so that the speed of the train would be reduced and it would roll into the station, and the driver would be able to stop the train at Signal T168 which was sited at the end of the platform.
Previous Incidents and Signalling Faults

Driver B Mathews (Victoria) was involved in an incident at Signal T168 in 1984. While driving a Gatwick Express he saw a series of double yellow signals and received the AWS warning which he cancelled before realising that Signal T168 was displaying a red aspect and that the previous Signal T178 must have been displaying a single yellow aspect. He was able to stop the train which was travelling at between 40 and 50 mile/h about a coach length past Signal T168. He immediately used the telephone to tell the signalman he had gone past the signal at Danger. In due course he made out a report to the Train Crew Manager and was subsequently disciplined.

Mr Mathews accepted the responsibility for passing the signal at Danger at the time and still did at the time of my Inquiry. He felt that with momentary inattention he did not register that Signal T178 was displaying a single yellow aspect instead of a double yellow aspect. He expressed reservations about the AWS system because it did not distinguish between single and double yellow aspects and that the repetitive resetting of the AWS when travelling under a series of double yellow aspects became an almost automatic reaction.

Driver D Creasey (Seaford) accepted the responsibility for passing T168 at Danger on 4 April 1986. He said he was running on a series of single yellow signals which turned to a double yellow as he approached them. He reset the AWS and kept the train running. Signal T178 was displaying a single yellow aspect which did not change to a double yellow. He reset the AWS and left the train to run before suddenly realising what he had done and although he made a full brake application the train passed the signal by 2 or 3 coach lengths. Mr Creasey was disciplined for passing the signal at Danger.

Driver D J Wright (Victoria) was the driver of a Gatwick Express which ran by Signal T168 by about a coach length on 16 November 1986. On the approach to a double yellow signal at Stoats Nest Junction he made a brake application but there was very little response. He sounded the warning horn continuously and believed he had attracted the attention of the driver of a train on the slow line. When the train had stopped he telephoned the signalman and told him what had happened. Although initially charged under the disciplinary procedure with passing a signal at Danger the charge was withdrawn after it was established that there was a fault with the brakes of the train.

Mr V C H Lambert (Bognor) who was now retired was a driver of some 40 yean experience when on 2nd January 1987 he was driving a train which was routed from the Redhill line to the Up Fast line at Stoats Nest Junction. He claimed Signal T178 was showing a double yellow aspect, a Gatwick Express passed travelling in the opposite direction on the Down Fast line as he approached Purley Station and then he saw Signal T168 at red. He made a full brake application but knew the train would not stop before the signal when to his amazement he saw a train crossing from the Up Slow line to the Up Fast line ahead of him. He estimated his train stopped two coach lengths clear of the other train.

After the train had stopped he asked the guard to telephone the signalman from Signal T168 because the guard was nearer to the telephone. The guard relayed an instruction from the signalman to continue to East Croydon. Mr Lambert spoke to the signalman by telephone from East Croydon. He said the signalman asked him "What have you to tell me driver?" and he replied "I haven't got to tell you anything other than what you already know -that I have over-shot a red light at Purley platform". He continued to Victoria Station where he was seen by a supervisor and later spoke to a Traction Inspector by telephone.

When charged under the disciplinary procedure with passing the signal at Danger his initial response was "I was not expecting 168 to be red but as I am the only person to see and know that – and after 38 years' driving - and well aware of the system I know it will be almost impossible for me to prove, as all signals applicable will have been tested and nothing found wrong". Mr Lambert accepted that the first occasion he had claimed that Signal T178 was showing a double yellow aspect followed by Signal T168 at red was at the formal disciplinary hearing. He had not mentioned this to the signalman, the supervisors or managers to whom he had spoken.

Mr Lambert said he was proud of his driving record of 40 years without making a mistake. In his evidence to me be was adamant that Signal T178 was showing a double not a single yellow aspect and he claimed that the signalling system allowed the conflicting movement to be made after Signal T168 had been put back to Danger after just a few seconds.

Mr A Galley, the Southern Region's Operations Manager, explained how allegations of signals being passed at Danger were dealt with, He said the incident would be regarded as a serious one. Normally the signalman, as well as the driver, would be the first to realise what had happened and a conversation would take place been them. Immediately following that the driver would be seen by an operating supervisor who would be in the vicinity to ascertain that the driver was fit and able to continue his driving duty and to make arrangements for the driver to be seen by a traction inspector for a much more detailed examination and record of the incident to be made. At the same time arrangements would be made by the signalman for Signal and Telecommunications staff to carry out full testing of the signalling equipment and as soon as possible the train would be taken out of service to have its brakes tested.

Following the incident if there were any dispute between the driver and signalman about the aspects of the signals he had seen the driver would be subjected to a medical examination which would concentrate on eyesight testing. Should there be an admission, or a conclusion, that the driver was in the wrong he would be dealt with through the disciplinary procedure. The same would apply if the error was by the signalman. The disciplinary procedure did not start with an assumption the driver was always wrong. A full check of the facts would be made first and if there were any form of dispute an inquiry would be held to establish the facts and conclusions drawn before any blame or discipline was started.
2.118 On occasions the driver will sincerely believe in his own view of the incident, despite evidence to the contrary. If the driver believes he has been wrongly blamed there was an appeal procedure at which the driver could be represented by an officer of his trade union or another advocate of his own choice. That appeal would be heard by an officer senior to the one who held the original disciplinary hearing. Mr Galley confirmed that Mr Lambert should have been seen in a 'face to face' interview and not interviewed over the telephone. Appropriate action had been taken against the Traction Inspector concerned.

Mr L H Page, the Area Signal Engineer (Maintenance) for the South Central Area, first learnt of the accident while at home shortly after 14.00. From home he made the necessary arrangements by telephone for dealing with the accident before going to site and taking charge of one of the testing teams.

In his normal duty of maintaining the signalling system he was assisted by two Signal Maintenance Engineers, one based at Clapham Junction and the other at Brighton. Located at East Croydon was a Supervisor and 18 staff to undertake the maintenance of signalling equipment in the Croydon and Purley areas. The maintenance work was based on a 6 weekly cycle. A technician would check the signal structure, clean the signal lenses, and oil the hinges and lock of the access door. Moving parts of point machines would be lubricated and track circuits checked for loose wires or connections. During the six-weekly cycle there was no specific test of the functioning of the equipment, though during the work on busy sections of line, the technicians would see the signals display the full range of aspects. Full functional testing of the signalling equipment was carried out to a laid down testing procedure quarterly.

Mr Page said that his staff would recognise my abnormal behaviour of the signal but that such failures were so rare that the chances of a technician seeing one was fairly remote. Normally reports are received from a signalman -use of what he has observed or had reported to him by a driver. When the problem has been identified and rectified the information is entered into a computer record system which stores the information in a simplified form. From these records Mr Page produced information on failures of Signals T186, T182, Tl78 and T168 at Purley. The records covered both actual faults and incidents when there was no fault in the signalling equipment.

The following incidents or failures had been recorded: 

Signal T186 

24 January 1985 - a 'rightside' failure occurred when the signal failed to clear when the signalman set the route. The failure rectified itself and no fault was found. 

21 December 1987 - there was a report that the signal was displaying a red aspect instead of a green aspect. No fault was found which would have caused this 'rightside' failure and it was possible that the signalman had replaced the signal to Danger. 

Signal TI82

18 June 1985 - a 'rightside' failure when the signal aspect went from green to red. This was not reported to the signal technician and was, therefore, not investigated. 

28 June 1985 - the green aspect lamp failcd. This was a 'wrongside protected' failure which would have caused the previous Signal T186 to display a red aspect. 

10 September 1985 - a 'rightside' failure when the signal went from green to red. No cause was found. 

30 March 1987 - a 'rightside' failure when the signal went from green to red. No cause was found but there had been problems with the detection on 1662 points which lay in the line ahead of the signal and a momentary loss of detection would have caused the signal to go to red. 

7 July 1987 - a 'rightside' failure when the signal aspect went from green to red. No cause was found. 

10 October 1987 -a 'righside' failure when the signal aspect went from green to red. No muse was found 

18 October 1987 -the signal was passed at danger. No fault was found with the signalling and the driver was held to blame. 

15 November 1988 -. a 'rightside' failure when the signal aspect went from green to red. No cause was found. 

26 March 1988 - it was alleged that the signal was showing a double instead of a single yellow aspect with the junction indicator lights illuminated for the route from the Up Fast line to Up Slow line through Stoats Nest Junction. All the signalling in the area was thoroughly tested and it was found that it was not possible when the junction route was set for Signal T182 to display more than a single yellow aspect. The driver involved reported that he only thought he saw the signal displaying a double yellow aspect but he could not actually be sure. 

Signal T178

24 October 1984 - a 'rightside' failure when the signal aspect went from green to red. This was not reported until June 1985 and, therefore, had not been investigated. 

18 April 1985 - a 'rightside' failure when the signal aspect went from green to red. No cause was found. 

4 October 1985 - a 'wrongside-protected' failure when the green aspect lamp failed causing the previous Signal Tl82 to display a red aspect. 

21 November 1985 - a 'wrongside - unprotected' failure when the signal was showing a green instead of a red aspect. The cause was found to be a basic error which resulted in the equipment for two track circuits of the same Frequency being housed in the same apparatus case, and it was possible to get cross-talk from one track circuit to the other via electromagnetic radiation. Track circuit frequencies were checked throughout the Southern Region; no other similar situations were found. The track circuits at Purley were modified. 

(This incident became known as the 'Fitzjohn' incident.) 

4 February 1986 - a 'rightside' failure when the signal aspect went from green to red. No cause was i found. 

Signal T168 

16 October 1984 - the signal was p a d at Danger. No fault wax found with the signalling system and the driver (Mr Mathews) was held to be responsible. 

5 May 1985 - a signalman realised that in certain conditions it was possible for the approach locking circuit on Signal T168 to be prematurely released; there was no incident which led to this discovery. When the design was checked it was found that one of the track circuits had been omitted from the approach locking controls. This was rectified and thoroughly tested to ensure that the approach locking was "absolutely perfect". 

22 May 1985 - a 'rightside' failure when the signal aspect went from green to red. The cause was established as the momentary loss of detection of 1639 points as a train passed on the adjacent line. 

10 June 1985 - a 'rightside' failure when the signal aspect went from green to red. The cause was again established as a loss of detection of 1639 points.

15 July 1985 -a 'wrongside protected' failure of a signal aspect lamp. The computer record did nor identify which aspect was involved. 

3 December 1985 - a 'rightside' failure when the signal aspect went from green to red. This was not reported to the signal technician at the time and was, therefore, no1 investigated. 

26 December 1985 - a 'wrongside-protected' failure of the red signal aspect lamp causing the previous Signal T178 to be held at red. 

4 April 1986 - the signal was passed at Danger. The signalling was thoroughly tested and found to be working correctly. The driver (Mr Creasey) was held to be responsible. 

16 November 1986 - the signal was passed at Danger. The signalling was thoroughly tested and foundto be working correctly. The records (incorrectly) indicated that the driver (Mr Wright) was to blame. 

2 December 1986 - a 'rightride' failure when the signal aspect went from green to red. This was notreported to the signal technician at the time and was, therefore, not investigated. 

2 January 1987 -the signal was passed at Danger. The signalling was thoroughly tested and found working correctly and the driver (Mr Lambert) was held to be responsible. 

21 September 1988 -a 'rightside' failure when the signal aspect went from green to red. No cause was found.  

During the course of the Inquiry reference was made to other signalling failures elsewhere on the Three Bridges area and Mr Page also gave evidence about these failures. These failures are not strictly relevant to the accident at Purley.

DISCUSSION

Queries as to whether it was a safe arrangement to cross trains from the Slow line to Fast at Purley while 'Fast' trains were travelling towards the junction were raised during the course of the Inquiry and also by many of those who wrote following the accident. On any other than the simplest and lightly used railways the transfer of trains from one line to another must take place. If it did not take place at Purley it would still have to take place elsewhere. The arrangement at Purley should create no greater risk than it does elsewhere. The risk is recognised and appropriate safety provisions are made. 

The stanard provision for such junctions are specified in the British Railway Board's signalling priciples and they have been included in the arrangements at Purley. The provision of a full overlap of 187m (200 yds) beyond the protecting Signal T168 and the approach locking controls; with the 2 minute delay are among the measures to avoid errors of judgement on the part of drivers or signalmen causing an accident. The overlap was adequate to prevent an accident on the previous four occasions when the signal was passed at Danger. I do not believe that increasing the overlap distance to the full braking distance is a practicable solution.

It is clear from the number of failures that the Three Bridges signalling system, did not operate to the very highest level of reliability that should have been expected from a new and modern installation. While the general situation was stil1 a matter for some concern, from the evidence given by Mr Page the situation was improving.

Of the failures of Signals T186, T182, T178 and T186, 16 were 'rightside' failures when the signal aspect reverted from green to red This is part of the fail safe design concept that should a failure occur the signal will revert from a proceed to a stop aspect. This can be caused by many things such as a momentary loss of detection on points or an interruption in the electrical circuit of an unoccupied track circuits because of the momentary nature of these events they frequently cannot be either reproduced or identified afterwards.

The reverting of a signal to Danger is in itself not dangerous. It may, however, result in a driver having to make an emergency brake application and could result in the signal being passed after it had reverted to Danger. The driver would not be blamed. Although regarded as potentially more serious the 4 'wrongside -protected' failures are in a similar category to the 'rightside' failures. The two 'wrongside unprotected failures were a very much more serious matter. Both stemmed from flaws in the design and installation procedures employed on the Three Bridges scheme. They should, in my opinion, both have been found during the design-checking of the system. They were, however, found and rectified prior to the accident on 4 Match 1989.

After the flaw in the approach locking on Signal T168 was found and remedied in May 1985 the system was tested and, in Mr Page's words, found to be "absolutely perfect". After the incident in January 1987 it was re-tested and was also tested again as part of the exhaustive testing following the accident. Mr Lambert remained convinced that Signal T178 was displaying a double yellow aspect and Signal T168 a red aspect despite the comprehensive technical evidence to the contrary. His claims as to how the approach locking of Signal T168 operated were based on a completely mistaken understanding of the system. The timing relays used in such controls are specially designed and manufactured 10 'fail safe'. Once the timer setting is accurately adjusted for 2 minutes in this case, it is sealed. If these timing relays fail, they do so by failing to run at all and &c locking would not be released and the conflicting route could not be set.

If Signal T178 was displaying a double yellow aspect or if Mr Morgan thought it was -he has made no suggestion it was - he would have had to make a brake application about 200 yds after passing Signal T178 and some considerable distance before the brakes actually were applied in order to be able to stop the Wain normally at Signal T162.

If Signal TL78 had been displaying a pen aspect he would not, of course, have made a brake application. It would have required a major 'wrongside' failure of the signalling system for Signal T178 to have been displaying a green aspect or a double yellow aspect with Signal 7168 at red. It is known from the evidence of Mr Sellwood, who had seen Signal T168 at Danger, that half-an-hour before the accident the correct sequence of aspects was being displayed. I find it inconceivable that a major fault in the signalling system could have suddenly occurred and then not be found afterwards.

Since the accident Mr Morgan has consistently said that Signal T168 was showing a red aspect and there is no evidence to show that this was not so. Following the previous incidents when Signal T168 was passed at Danger, the signalling was tested and no faults found. After the accident the signalling system was subjected to the most exhaustive testing and again no faults found. 1 consider therefore, that the sequence of aspects being displayed was correct with Signal T182 displaying a double yellow aspect and Signal T178 a single yellow.

The only alternative believe remains is that Mr Morgan failed 10 make any brake application on seeing the correct caution aspects being displayed by Signals T142 and T178. He must also have reset the AWS without heeding its warning. It was not unusual for trains to have to come to a stand at Signal T168 and Mr Morgan as an experienced driver over that route must have been well aware of that. It is unlikely, therefore, that he failed to reduce the speed of the train because he believed that Signal T168 'always came off allowing the train to continue.

Mr Morgan was an experienced and responsible driver but it is known that such drivers may suffer from an uncharacteristic lapse in concentration when caution signals are not recognised and the AWS is reset leading to a signal being passed at Danger. Fortunately, as Mr Warburton explained in his evidence, few of these incidents cause an accident but, nevertheless, the risk is there. This human behaviour problem has been recognised by all those involved in railway safety for some years and considerable research has been undertaken into it. The problem is complex and the work has not yet produced any positive measures which can be implemented to eliminate the problem.

The AWS used by British Railways was developed to attract the driver's attention to signals. When first introduced it was a significant and positive safety development but it has weaknesses. It does not distinguish between single and double yellow signals, it does not provide a lasting display of the signal aspect the warning referred to, and it can be reset unheeded by a driver whose concentration has lapsed. I believe that because of these inherent weaknesses the existing AWS equipment is not capable of usefully being developed further. While it is possible to devise other forms of AWS, which would address these problems, the timescale and resources involved are similar to those required for the development of an ATP system

The newer Automatic Train Protection (ATP) systems are mandatory rather than advisory. If the driver does not take the appropriate action to control the speed of the train in accordance with the signalling the ATP will take over the control of the train and bring it to a stand if necessary. Such systems cannot be reset and ignored by the driver. ATP is now employed by a number of Metro railways and some main line railways. I accept, however, that none of the existing systems matches exactly what is required for the British Railways network with its different types of trains and density of traffic. The problem that the application of any of the existing intermittent ATP systems would extend the headways is a significant one. The investment needed and the amount of work to be undertaken to install an ATP system is massive.

It is unclear why Signal T168 should have a higher than usual number of incidents of being passed at Danger. Although Signal Tl68 is more likely to be displaying a red aspect than, for instance, an automatic signal which only goes to red when a train is occupying the section of track it protects, I do not believe that this provides a satisfactory explanation. The minimum sighting time to be provided for any signal on a line with a maximum speed of 90 mile/h to comply with British Rai1waq.u own standards is 7 seconds. When w m p d with many of the signals in the Purley area the sighting of Signal T168 is poor being obstructed by Purley Station buildings. Its sighting distance is, however, marginally better than what is required to give the minimum 7 seconds.

Of perhaps greater importance is the sighting of the two previous Signals T178 and T182 which provide the advance warning that Signal T168 is at Danger. Both Signals TL78 and T182 have sighting distances considerably better than the minimum requirement; Signal T178 can be seen from Signal T182 a distance of over 1,100 yards. The braking performance of an EMU is such that there is no need to make the brake application until after passing Signal T182 and by then Signal T178 is in sight. Driver Morgan referred to looking ahead to see what aspect Signal T178 was displaying. I believe it is possible that the extended sighting of this signal may cause a driver to relax his concentration, not control the speed of the train properly and then find himself surprised by the short sighting distance of Signal T168.

The Southern Region are examining other signals with a higher than usual history of being passed at Danger to see if a similar arrangement and sighting of signals exist. If my deduction is correct there appears to be no orthodox solution to the problem. 1 believe it would probably be worthwhile, as a short term expedient, to increase the sighting distance of Signal T168 by the provision of a banner repeater signal even though the conditions which would normally make one necessary are not present. The provision of the banner repeater may help reinforce the single yellow caution aspect of Signal T178.

I was concerned by the mistaken, but clearly sincerely held, belief of Mr Lambert that he did not need to tell anyone about the abnormal signal aspect he believed he saw because others would also know the aspect being displayed. This is, of course, not so and could only be achieved by installing elaborate monitoring equipment which would increase not only the cost of signalling systems but would also probably adversely affect the reliability of the system. It is of concern that some reports from drivers do not reach the signal engineering staff in time for them to carry out an investigation. Of equal concern is the absence of explanation back to the member of staff who made the initial report.

The initial collision took place at a closing speed of 30 to 40 mile/h with the two trains at a slight angle, The damage mused was consistent with the front right corner of the leading coach of the train from Littlehampton striking the rear left hand corner of the trailing [fourth) coach of the train from Horsham before sliding alongside the fourth coach and striking the rear right hand corner of the third coach. The damage to the rear end of the leading coach of the train from Littlehampton was extensive and the body was almost entirely destroyed. Those killed were, I understand. all travelling in this part of the train. The initial damage appears to have been caused by the heavy leading bogie of the second coach striking the body of the leading coach as the train 'jack-knifed'. The body structure having been seriously weakened by the impact was further damaged by the large trees growing on the embankment.

The remaining vehicles survived the fall down the embankment remarkably well; the body structures were not seriously broken, most windows remained intact, and the main internal fittings stayed fixed in place. There was however, a number of passengers seriously injured by being thrown about within the coaches as they plunged down the embankment. The evidence of some of the passengers, and of Mr Knights', graphically illustrated what occurred. I do not believe the newer designs of rolling stock would have survived significantly better in this form of accident.

Members of the Railway Inspectorate have encouraged the concept of 'black-box' recorders being installed on trains. Such devices, while not an accident prevention measure, would be of peat assistance in the investigation of accidents. British Railways are now actively pursuing the installation of such equipment. The specification outlined by Mr Warburton was an impressive one and while obviously possible for new rolling stock a simpler version may be necessary if it is to be introduced quickly or its installation extended to existing rolling stock.

Having arrived on site the emergency services co-ordinated their activities effectively and carried out their difficult tasks with their usual skill. The Fire Brigade and Police quickly took charge of the accident site and liaised with the railway staff in a way which I believe was generally satisfactory. I believe, however, the liaison arrangements could be improved still further in two aspects. Firstly, in addition to the emergency services incident control at the foot of the embankment. British Railways opened their own incident control at Purley Station. There was no direct communication links established between these two incident controls and liaison took place between individuals at the scene rather than between the controls. I believe the liaison arrangements would have been strengthened with better liaison between the emergency service and railway incident controls.

Secondly, having quite properly taken charge of the site the emergency Services then regulated access or other persons to the sire. Clearly a number of responsible railway engineers did not gain access to the site and the rolling stock as quickly as was desirable. The post-accident condition of the trains and the position of the driver's controls provide important evidence towards the investigation of any train accident. Had the braking performance of the train been in question much valuable evidence would have been lost by the time the examination was made. Clearly the police officers have a responsibility to protect property and evidence at the scene but it is equally important that railway officers (and members of the Railway Inspectorate) have access to the site as well. I welcome the concept of police officers accompanying the railway officer during the inspection and testing; it must be beneficial to both parties. The respective roles were not fully understood at Purley and action has already been taken to improve understanding and co-ordination.

Immediately following the accident there was doubt as to whether current had been cut off from the conductor rails. This was due in a large part to the difficulties caused by the loss of the supervisory circuits between the electrical control and the remote sub-stations. The circuits were lost when the line side cables were destroyed by the derailed train. I believe that, while this risk cannot be totally eliminated greater protection or duplication of these control and supervisory circuits would reduce the risk and the consequences.

The railway staff at the site of the accident quite properly, because there was doubt, applied short-circuiting bars and treated the conductor rails as if they were energised. Mr Foster, the Electrical Controller, did well in conjunction with the controllers of adjacent electrical areas m protect quickly the site and provide what assurances he could. Suggestions were put forward at the Inquiry that all trains should he equipped with indicator devices which when placed on the conductor rail would show whether or not it was energised. 1 believe such devices could be of assistance but would have to include some system for 'self-testing' to eliminate the possibility of a dangerous 'wrongside' Failure.

I believe that in general the train crews and other railway staff involved in the emergency arrangements and the evacuation of passengers behaved in a commendable fashion. I was less happy with the part played by the Station Manager who did not ensure on his arrival that the necessary liaison with the emergency services had been established. I was also concerned that a Signal Technician began a series of physical disconnections of the signalling equipment without fully checking and recording the state of the signals. While his actions were apparently motivated by safety considerations and in accordance with the Rule Book and Departmental Instructions, they may have made the investigation more difficult. British Railways have this matter under review and additional advice has been given.

CONCLUSION

There is corroborated evidence that T170 was showing a proceed aspect (a single yellow) and junction indicator lights for the movement of the train from Horsham from the Up Slow line to the Up Fast line. There is also clear evidence that the switch blades of the crossover tracks from the Up Slow line to the Down Fast line and from the Down Fast line to the Up Fast line were lying properly secured in the reverse direction required for such a movement.

It is  clear that Signal T168 on the Up Fast Line was showing a red stop aspect protecting the route set for the train from Horsham. There is no eye-witness evidence as to the aspect being displayed by the two signals before Signal T168 on the approach to it along the Up Fast line. I conclude, however, from the exhaustive technical evidence that the two signals were displaying the correct caution aspects, that is, a double yellow aspect at Signal T182 and a single yellow aspect at Signal T178.

The brakes of the train from Littlehampton were in proper working order as was the AWS equipment. I must conclude, therefore, that the driver of the train From Littlehampton failed to heed the caution aspects or Signals T182 and T178. He must also have failed to heed the alarm from the AWS which he must have twice reset. He made a brake application when the red aspect at Signal T168 came into view. Despite the full emergency brake application it was by then impossible for the collision to be avoided.

I  find that the collision was caused by the failure of the driver of the train from Littlehampton to control the speed of the train, in accordance with the preceding signals enable him to bring the train to a stand at Signal T168.

REMARKS AND RECOMMENDATIONS

This accident would have been prevented by an Automatic Train Protection (ATP) system. Justification of the capital expenditure involved in the installation of such a system cannot be provided on the basis of normal investment criteria. It is clear, however, that momentary lapses of concentration by train drivers do occur and that the Automatic Warning System (AWS), which is at present in use, can be reset without the driver heeding its warning. In order to ensure the continuation of the generally high safety standard of railway travel 1 recommend the installation of an ATP system on an 'high speed' and on all intensively used lines as quickly as possible.

The British Railways Board have already taken the decision to proceed with the development and installation of an ATP system. Having examined the various systems in use by other railway administrations they have included that none of the systems presently in use can be used with British Railways existing signalling and traffic patterns without some development. While acknowledging that this is correct I nevertheless recommend that the provision of ATP should not be delayed by an extended development period and that an existing system, which is proven and validated, should be used with a minimum of development.

As an interim measure in respect of the signalling arrangements at Purley I recommend that a 'banner' repeater signal which can be seen on the approach to Purley Station should he provided for Signal T168. (This signal was provided on 3 September 1989).

The British Railways are already considering the provision or 'blck box' incident recorders. I recommend all new builds of locomotives and multiple-units should be equipped with such recorders and that existing ones should be retrospectively fitted if it is practicable to do so. It may be appropriate to use a simpler recorder for existing rolling stock.

It is important that any signalling irregularities, whether actual or perceived, should be reported and investigated promptly, and any necessary remedial action when immediately.  It is equally important that those making the reports are kept as fully informed as possible. I consider the present arrangements for reporting, actioning, recording and reporting hack on action taken are not. as comprehensive as they should be. I recommend that a better regulated system be introduced as quickly as possible.

Footnote

Driver Morgan received a limited immunity from prosecution to enable him to give evidence to my inquiry. However, he was subsequently charged with manslaughter and endangering life and it was not possible to publish my report while proceedings against him were outstanding.
On 3 September 1990 Driver Morgan appeared at the Central Criminal Court and pleaded guilty to the charges of manslaughter and endangering life. He was sentenced to 18 months imprisonment with 12 months suspended.

UNOFFICIAL CIRCULAR REGARDING SOME FACTS BEHIND THE PURLEY CRASH

17th MARCH 1989

We understand that Bob Morgan has stated that he cannot remember what happened to cause the crash at Purley, when one considers the force of the impact and the fifty foot drop down the embankment, it is not at all surprising. British Rail however stated publicly on the day of the accident, that the crash was caused by human error. Since that date Bob Morgan has been hidden from all people, press and colleagues alike.

When the writer of this report phoned the May Day Hospital at Thornton Heath, to enquire about his health. The hospital denied his existence. Someone week later, his colleagues had to obtain special permission to visit him.

His colleagues realise the need for his seclusion and the need to kep the press and the news media at a distance, but also note that the Southern Management lost no time in placing their own version to the press.

We are unable to say what happened at Purley on this day, suffice it to say it is on record with British Rail. ‘That this accident at this spot was predicted, and unless the problem is recognised must happen again.'

On the 21st November, 1985, the 06:20 Gatwick to Victoria Rail-Air Express, approached Purley with a green aspect at Stoat’s Nest, only to find another train in his path. The Signals changed from red to and then to two yellows. Driver Jeff Fitzjohn states, that although his signals showed a double yellow aspect at T178, he could see a train crossing from the local to the fast line at Purley. Signal T168, was at red (two yellows at T178, red at T168). Driver Fitzjohn was lucky enough to see the changes of aspect at Stoat’s Nest and realise there was something wrong, otherwise there would have been another 90 m.p.h. pile up.

British Rail stated that it was caused by irregular use of the Track Circuit Equipment, and that the fault had been rectified now?

On January 1st, 1987, Driver Vic Lambert, on the 06:17 bognor to Victoria, approached Purley with a double yellow aspect at T178, Purley signal T168 was showing a red light with a train crossing from the local to the fast inn Laberts path. Lambert was unable to stop and ran by T168, narrowly avoiding impact with the crossing train.

Communication with the signalman at Three Bridges showed that he did not know that Lambert had passed T168 and was on the point of collision. The Signalman did know however there was a further train right on Lambert’s tail. his instructions were that Lambert should move off to East Croydon as quickly as possible. That message was relayed in more colourful language than this, but the message was clear and it meant that a repeat of this accident at Clapham Junction could have taken place (in this case it would have been the forerunner to it).

Had Lambert not reported himself, no one would have ever known that this run by had happened or that a near disaster had taken place.

On subsequent disciplinary action and at the appeal Lambert’s evidence was not believed by British Rail, and no action to find out what really did happen.

We do not say there is a fault in the automatic system either at Clapham Junction or Three Bridges neither do we say that Signalman or Drivers are making mistakes. Both are working to what they believe are correct methods. We do feel however that there is something wrong with instructions and time tables that require signalmen to bring a fast express service to a stand to allow slower trains to pass ahead of it on the same line. Reference to the number of signals changing aspects will show that in most cases these occur at cross over points where one section of the line is on automatic control, the other on manual control.

Firstly, we question the wisdom of selecting a point where the line speed is at maximum 90 m.p.h. to use a a cross over point. Purley is just this position, where trains are crossed over, not only at random, but on a scheduled basis. The point being, that if a train is running late on the slow line, the trains on the fast line have to be brought to a stand to wait for the slow train to pass ahead of it because the schedule dictates that it must. Older types of signalling would have allowed signalmen to use their common sense, modern signalling methods do not permit this luxury.

Secondly, we bring to your notice, the design of this system where the “detectors” that update the signal computer and panel, bear little relation to the signals they are supposed to indicate on the signalman’s panel, signal T178 for instance. The distance of the detector is 615 feet past the signal (a twelve car train length), and until the train passes this detector, the signal panel still shows the train to back at Stoat’s Nest signals. This situation shows that the driver can have passed T178 (in which he cannot see it), and travel a further five seconds at 80 m.p.h., and the signal can be changed, by the simple fact that the signalman has switched from auto control to mania control, a fact that this is done by setting his road from slow to fast at Purley. If this is done in the period of lost movement (615 feet), the driver would never see it.

I have used the signal T178 as the demonstration signal, but this is true on all the signals involved in this accident. The signal on the fast line are largely “switched” to automatic as are the signals on the slow line and cannot be faulted and on straight line running, present no problem. The problem arises when the lines are switched from auto control to manual and the time delay in updating signal equipment. The driver is always ahead of  the information supplied to the signalman, in the case of T178, by as much as a thousand and ninety one yards plus 515 feet.

In the case of driver Lambert, this can be proved by the fact that just 7 seconds before his incident at T168, he passed a Gatwick Express on the down line, for this to happen, the cross over at Purley could not have been set, yet just seven seconds later a train crossed from the slow line to the fast and Lambert was accused of running by T168 Purley.

Driver Jeff Fitzjohn reported this, driver Lambert reported this and now driver Bob Morgan has a major accident at this point. Again I wish to go on record for the second time and state that this is bound to happen again. 

THE REMEDY IS SIMPLE AND OBVIOUS EVEN IF THIS REPORT IS NOT BELIEVED.  All trains entering the section from Stoat’s Nest, Redhill or Quarry line, should, through the automatic locking system, lock the cross over points at Purley, so that movements cannot take place all the time there is a train approaching on the fast lines.

-------------------------------------------

* In c1989, Eastbourne driver Cyril Ring was involved in accident whereby signal T168 was showing a red aspect when he was running through Stoat’s Nest (signal T178) under green aspects.

LOCOMOTIVE JOURNAL

JANUARY 2008 

17 years on – justice for ASLEF man convicted of manslaughter

THE Court of Appeal last month quashed a train driver’s 17 year old conviction for manslaughter.

Bob Morgan, a train driver and ASLEF member, was convicted of two counts of manslaughter on 3 September 1990. He was sentenced to 18 months in jail, of which 12 were suspended, after he had admitted passing signal T168 while it was red – an event which contributed to the Purley train crash of 4 March 1989. 

However, the original conviction had not taken into proper account the fact that this signal was dangerous. It had been passed at danger on four different drivers – between 16 October 1984 and the Purley tragedy. This occurred again on 6 June 1991. Since the fault with the signalling has been remedied, no trains have passed T168 at danger. 

At the trial neither the prosecution nor the defence were aware that signal T168 was a ‘multi-SPAD’ (regularly passed at danger) signal – which is why Bob was advised to plead guilty. Today a signal with this history would be immediately assessed for risk and action would be taken to remedy the danger. This did not happen in 1989.

The accident happened at 13.39 when Bob Morgan’s train, travelling from Littlehampton to London, crashed into the back of a train travelling from Horsham to London. The Littlehampton train careered off the track, down the embankment into gardens below. 6 people died and 80 were injured.

ASLEF general secretary Keith Norman, said he was proud that the union had stood by Bob Morgan throughout his 17-year ordeal, ‘I am delighted that Bob has been vindicated. It was important not only for his peace of mind and reputation, but for all other UK train drivers. Today’s verdict should ensure that no train driver in the future will have to face such a terrible ordeal.

‘We have established that where safety issues – like and unsafe signals – contribute to an accident, the driver will never again be the single person held to account – which is what happened in Bob’s case.

‘Over a period of several years, drivers found it difficult to see the signal in question – a fact proven by the number of different drivers who passed it signalled at danger.

‘The tragedy for Bob and his family was that no one in management admitted their inefficiency or neglect. Instead, they slunk away, happy to leave Bob to carry the can for them all. There was no consideration of corporate responsibility – simply the desire by managers to escape blame by pointing a finger at an individual. Bob paid the price for the inefficiency and carelessness of layer upon layer of management. 

‘We can never undo the harm Bob has endured - but we can admire the dignity with which he bore the burden of the last 17 years.

LOCOMOTIVE JOURNAL

APRIL 2009

 LOCOMOTIVE JOURNAL

MAY 2009

I set off from Head Office with a heavy heart on 7 April for an appointment I wished I didn’t have to make. Along with the President Alan Donnelly I made my way to Worthing Crematorium where the funeral of Bob Morgan took place.

Bob’s story has been told often in the pages of the Journal, from the tragic accident at Purley and his conviction and imprisonment for manslaughter to his name finally being cleared 14 years later at the High Courts of Justice.

Bob was the last man who should have had to face such unjust traumas. He loathed the publicity his case attracted and his relief was obvious when he told me, just a year ago, that ‘it was finally over’. And now, with so little time passed since he was freed of the burden, he has died. He was just 66 when he drowned in a boating accident. He was last seen one late afternoon sailing on the River Medina off the Isle of Wight .

It is a deeply sad story, and perhaps the only people who can fully appreciate its tragedy are those who have driven trains themselves. People who know the dangers and the constant challenges of our profession. ASLEF members.

As I looked around the packed crematorium I saw driver after driver who had turned up to pay their last respects, united in a ommon fraternal bond. I was moved by the comradeship that we share. ASLEF is more than a union.

KEITH NORMAN

GENERAL SECRETARY