Links:

R.o.B

 

Presents

 

Kory McLeod

 

Singer/Song writer

 

Please have a listen:

Site 1

Site 2

Site 3

Last updated 12/04/2006

R.o.B Signalling and safety

Train Protection

Automatic Train Control (ATC)

As with many of the other railway companies the GWR main line has had several collisions between trains due to one passing a signal at danger as much as the other main lines but at the beginning of the 20th Century the Great Western Railway was the first to develop a wide spread system to warn drivers of signal aspects on the approach to the signal. 1900 saw a collision at Slough on the GWR main line when the 13.15 from Paddington to Falmouth ran in to the rear of a stationary local service in the station, the driver of the express having failed to stop at several stop signals on approach even though the weather was clear and visibility fair. The result was that some form of Audible warning apparatus should be used at signals to aid the driver's concentration and train safety, though it took several years for the final product to be installed on the main line. This involved an electrically insulated ramp made from steel at every distant signal in the four foot which forced a shoe on the locomotive. If the signal was at 'proceed' then the ramp, which was laid on a wooden board for electrical insulation, was electrified and when the shoe was pushed up the electrical circuit would cause a bell to ring. If the signal was at Caution the shoe would break an electrical circuit on the engine which would sound a steam whistle in the cab in the process until it was reset by the driver. By 1906 the GWR was developing the idea and was using it on its Henley branch where some improvements were made to give the train some form of automatic train control (ATC).

This was not the first use of some form of audible warning in train cabs however as the idea had been around since before the use of permanent semaphore signalling in 1840 with the NER developing the first really effective method in 1894 which used a metal arm which was sited in the '4 foot', between the running rails which was linked to the signal by a rod or a wire. If the signal was at 'caution' or 'Danger' the raised arm would hit a shoe on the engine which would apply a gentle brake application and sound an audible warning in the cab which the driver would then cancel. This was soon developed and fitted to around 1528 locomotives and 90 route miles, though it's primary use was as a fog warning rather than a full time warning element though it was certainly the fore-father of the GWR idea. It seems the NER had gained their idea from a similar method that had been in use some 10 years earlier in 1884 at Wimbledon on the London and South Western railways main line where a system called Kempe and Rowell's patent was used with some mild success though it was never fully taken up though it was supposed to be for all weather conditions and visibility levels. The reason why it was not such a success was probably more due to people's opinion that the driver had to be in full control of the train with all concentration on its progress and that automatic systems may lead to some complacency.

The problem with the early GWR system was that it only alerted the train driver to the fact that an approaching distant signal was at caution and did nothing to actually stop the train should the driver not acknowledge it. The companies engineers recognised this and set about developing a way that should the driver not acknowledge the alarm that the train would automatically stop with experiments taking place between 1908 and 1909. It still needed the ramp in the 4 foot at the distant signal but when the train passed over the ramp a Vacuum brake valve was also opened and could only be shut when the driver acknowledged the warning and if he did not acknowledge the train brakes would remain open the train drawing to a stop at a governed rate. This successful experiment was soon introduced to the GWR main line from Paddington to Reading between 1908 and 1910 with great success. By 1939 the GWR had fitted 3250 of its locomotives with the equipment and 2850 route miles were also equipped. By 1921 the NER had also improved their system and were using it on their main line between Shaftholme Junction in Doncaster and Berwick upon Tweed, though after an HMRI inspection in the August of that year a general announcement was made that the NER system should not be adopted anywhere else and the LNER when they took over in 1923 soon abandoned its use all together.

One reason for the lack of widespread development of the ATC has been put down not only to World War one but also to the Grouping of the railways in 1923 which saw a change in priorities of the Big Four railway companies to rationalisation of the wide varieties of locomotives, coaching stock and route miles. Safety though not totally ignored was put on a back burner for sometime to allow for system development, the exception once again appearing to be the GWR who did, as was noted above, continue with their ATC development. A multi-company committee was set up however between 1920 and 1922 to consider the way forward for ATC and was named the 'Automatic control committee'. There remit was to:

The conclusion of the committee was that there was a need for warning devices at distant signals but that some sort of stop device was more important at some stop signal locations. The committee also recommended that some form of standard procedure and mechanical equipment be established to dispense with fog signalmen at distant signals and that the best way forward was for Danger, Warning and Clear devices, though the use of cab signalling was totally ruled out. Needless to say that the idea of introducing such equipment to 24,000 distant signals, 38,000 stop signals, and 23,000 locomotives would cost in the region of 4,660,000 pounds and was something of a dampner on the idea especially when the country was in a huge depression.

With the conclusions not being settled another committee was formed in 1927 to take up where the previous one had left off as the idea was not widely supported though seen to be sound. This time round the committees remit was 'To review the recommendations made by the ATC committee of 1922 and to consider the and report what alterations, if any, should be made in that committee's conclusions and recommendations having regard to the developments that have taken place since that date.' Despite the committee stating that the GWR ATC was a good idea and could be employed elsewhere the other four companies decided on re-signalling and track circuiting rather than ATC. Even the HMRI seems to have been dubious about its efficiency and did not seem to be overly supportive of having ATC set up network wide, despite 2 major incidents in 1928 where drivers had passed signals at stop with the other three main line companies still hanging back not that they were against the idea of minimising train accidents and train crew or signalling errors.

The 1930's saw introduction on many lines of track circuiting and multi aspect electric signalling and a heightening of train speeds on main lines though there was still a lack of Automatic train control though the GWR continued to fit ATC to its distant signals until Nationalisation in 1948 and between 1927 and 1948 this paid off with accident passenger deaths results showing a record level of 2! During this period the LMS was also considering the introduction of ATC stating that 'The desirability of equipping all signal boxes on the most important routes with this appliance [ATC] is under review, although it has been decided to provide it in all new signalling schemes where conditions justify it.' To be fair to the HMRI, by this time they had seen the benefits of ATC and regularly recommended companies to install it on the main fast lines and especially at locations that had accident potential, but also at speed restrictions though the latter would take over 40 years to achieve.

Here it is worth a mention of another form of Audible warning for train drivers, was the use of detonators at key locations near signal boxes and signals. In later years detonators which are essentially small tightly packed explosive that are laid on or strapped to a rail are normally used to warn of people working on the track of a hazard ahead, the guard of a broken down train walking back to lay detonators should it be necessary. During the first part of the 20th century and before they were also used in fog to advise drivers of their approach to stop or caution signals with the signal man being able to lay the detonator on the rail head by means of levers in the signal box. One problem with this and was a common occurrence was that drivers or the engine crew may not hear them go off as the engines wheels ran over the top of them causing them to bang. To alleviate this problem the LMS doubled the detonators to make the bang louder.

In 1947 the GWR continued with its development of the use of ATC with experiments indicating whether the new colour signals were showing green, single yellow (caution) or double yellow (advance caution, used on high speed lines), the hooter in the cab that replaced the steam whistle many years before giving a double tone to indicate the double caution ahead. This was done by use of a 'magnet valve' which was polarised by electricity passing in to the ramp magnet causing a tone before the normal caution tone. Though by the time the experiments had got under way nationalisation stepped in and the project was abandoned by British Railways as being to expensive to install despite its effectiveness being proven at 95mph on the run between Reading and Maidenhead the train coming to a safe stand at Maidenhead station with an automatic brake application having taken place under test conditions.

By 1947 the HMRI was up in arms about trains passing signals at stop with several severe accidents in that year: The Southern railway seeing 4 separate collisions with 36 people killed:

Up to 1947 the Southern Railway had only had 4 collisions between 1923 and 1946 resulting in the deaths of 15 people. The LNER in 1947 saw 3 collisions with 52 dead, the GWR having none this put down to their use of ATC! Fog has always been a major problem on the railway especially in the days when oil lamps were used on signals as they were more often than not unable to penetrate the fog. The HMRI had regularly called for detonator placers for use in fog as standard throughout the Railways of Britain and it was soon believed a new Regulation of Railways act would be needed to enforce the use of ATC and detonators. Though with the modernisation of the railways in the 1950's and 1960's and introduction of Colour signalling the ATC became installed in most cases as the semaphore signals were replaced.

Though ATC is not fully fool proof [I'm not calling train drivers fools!], as drivers can push the acknowledgement button automatically without registering its tone it is still a highly effective method of alerting drivers to danger ahead and if the tone is, for some reason, not acknowledged then the train automatically stops. It has continuously proven that with its expanded use accidents caused by the use of ATC/AWS and the introduction of multi aspect signalling that driver errors have greatly reduced even though they do still sadly occur. One such incidence being the 'Ding Ding and away' syndrome where a train guard would give the right away and the driver would forget to re-check the signal aspect and route code before pulling away from the station. This has been partially resolved by the installation at stations of signal repeaters on platforms showing station staff and train crew if the signal for that line is 'Off' or 'On', 'clear' or 'danger'. The last installation of an ATC device was between Abbotswood Junction and Westerleigh Junction on the former LMS Birmingham to Bristol line which had been increasingly used by Western locomotives, with the last of the ATC ramps amazingly not being de-commissioned until 1976!