Train Protection in the UK
More articles on this site about signalling are listed in the Signalling Index
In spite of the excellent safety record of railways as a means of transportation, there have been occasions when drivers have allowed their train to pass a point where they should have stopped. Many of these incidents have resulted in collisions, some involving loss of life and most involving damage to equipment or property. Most incidents are the result of a driver failing to ensure that his train stops at a stop signal. In the UK, this has become known as SPAD or Signal Passed At Danger.
Such incidents have occurred on railways ever since they began in the early 19th century and various systems have been introduced to try to prevent them. These have taken the form of both warning and train stop systems. In the UK, a warning system is used on most main lines. An alarm sounds in the driver's cab whenever a train approaches a caution or stop signal. If the driver fails to acknowledge the alarm, the train brakes are applied. The system is called AWS (Automatic Warning System)
AWS - Automatic Warning System
Fig 1: Schematic showing arrangement of AWS Ramp on the approach to a signal.
It was realised (even before WW1) that some sort of automatic and enforceable warning was needed. This (after a number of experiments and some complete systems had been tried) eventually took the form of a track mounted, non-contact inductor which became known as AWS (Automatic Warning System). The AWS "ramp" as the inductor is known, is placed about 185 metres (200 yards) on the approach side of the signal (diagram, left).
The AWS ramp is placed between the rails so that a detector on the train will pass over it and receive a signal. The ramp will thus warn the driver of the status of the signal. The French railways use a similar system called "the Crocodile", the Germans, the "Indusi".
Fig 2: Schematic showing position of AWS Ramp in the track on the approach to a signal.
The AWS ramp contains a pair of magnets, the first permanent, the second an electro-magnet linked to the signal to provide an indication of the aspect. The ramp is placed between the rails so that a detector on the train can receive the indication data. The more observant passenger on a station platform can often see the ramps between the rails. They are usually a dirty yellow.
In operation, the train first passes over the permanent magnet and the on-board receiver sets up a trigger for a brake application. Next it passes over the electro-magnet. If the signal is green the electro-magnet is energised, the brake trigger is disarmed, a chime or bell rings in the driver’s cab and a black indicator disc is displayed. The driver takes no action. If the signal is yellow or red, as shown above (Fig. 2), the electro-magnet is de-energised, so a siren sounds in the cab and the disc becomes black and yellow. The driver must "cancel" the warning, otherwise the automatic application of the train brakes is triggered. The photo (left) shows the AWS "ramp", as the equipment is called, mounted at the approach to a signal.
It can be seen from the above that the British AWS allows the driver to cancel a warning as he approaches a signal. This means that, if he cancels the warning and still fails to stop, his train could collide with the train in front. There have been some well documented examples of this in the recent past. The only way of preventing this situation is by adopting a system of enforcement.
A very simple system of enforcement is used by the London Underground. It is called the trainstop. It is a mechanical arm fitted to the track next to each signal. When the signal is red, the arm is raised and will physically hit a trip device fitted to each train should the train pass the signal. This causes the train to stop by cutting off the power to the motors and applying an emergency brake application. When the signal shows a proceed indication the trainstop arm is lowered and the train can pass unhindered. The system is a simple form of Automatic Train Protection (ATP). It was first used in the UK in 1904, the idea having been imported from the US. It has been used by a number of other systems around the world. The modern versions occur in various Automatic Train Protection (ATP) systems available today which are based on electronics. See Metro Signalling and ATP
In spite of the installation of AWS over most of the UK's main line railways, there has been a gradual increase in the number of signals passed at danger (SPADs) in recent years and some serious collisions as a result. In an attempt to reduce these, a number of suggestions were made to reduce the impact (pun intended) of SPADs. One of these is the Train Protection and Warning System or TPWS, which has now become standard across the UK.
Fig. 3: Schematic of TPWS setup on the approach to a stop signal. The Arming Loop switches on a timer and the Trigger Loop assesses the time elapsed to determine the speed of the train. If the time is too short, showing the speed is too high, the trigger will activate the train brakes.
The idea behind TPWS is that, if a train approaches a stop signal showing a danger aspect at too high a speed to enable it to stop at the signal, it will be forced to stop, regardless of any action (or inaction) by the driver. The equipment is arranged as shown left.
For each signal equipped with TPWS, two pairs of electronic loops are placed between the rails, one pair at the signal itself, the other pair some 200 to 450 metres on the approach side of the signal. Each pair consists of, first an arming loop and secondly, a trigger loop. The loops are activated if the signal is showing a stop aspect.
The pair of approach loops first met by the train at 400 to 200 metres before the signal, are set between 4 and 36 metres apart. When the train passes over the arming loop, an on-board timer is switched on to detect the elapsed time while the train passes the distance between the arming loop and the trigger loop. This time period provides a speed test. If the test indicates the train is travelling too fast, a full brake application will be initiated. In case the train passes the speed test successfully at the first pair of loops but then fails to stop at the signal, the second set of loops at the signal will cause a brake application. In this case, both loops are together (see photo - right) so that, if a train passes over them, the time elapsed will be so short that the brake application will be initiated at any speed.
What TPWS Does
TPWS has certain features which allow it to provide an additional level of safety over the existing AWS system but it has certain limitations and does not provide the absolute safety of a full Automatic Train Protection (ATP) system. What TPWS does is reduce the speed at which a train approaches a stop signal if the driver fails to get the speed of the train under control to allow him to stop at the signal. If the approach speed is too fast, TPWS will apply a full brake but the train may still overrun the signal. Fortunately, since the train is already braking and there is usually a "cushion" of 200 yards (183 metres) between the signal and the block it is protecting, there will be a much reduced risk of damage (human and propertywise) if the train hits anything. With a possible total distance of 2000 feet (about 600 m) between the brake initiation and the block entrance, trains "hitting" the first loops at up to 120 km/h (75mph) could be stopped safely.
TPWS is also provided at many (about 3000)Permanent Speed Restrictions (PSRs) to ensure that a train does not pass through a restricted section of line (say one with a sharp curve) at too high a speed. However, there have been a number of issues related to the use of TPWS in these cases. Drivers have complainted that, although they were approaching the PSR at a speed which would allow the train to run at the correct speed within the restriction, they still got stopped by the TPWS "speed trap". This has led to some vigorus discussions between Network Rail, the train operating companies and the HSE.
An add-on to TPWS, called TPWS+ is provided at certain signals where train speeds are above 100 mph or 160km/h.
What TPWS Does Not Do
The safety effects of TPWS are limited by the fact that it is provided only for stop signals and that it cannot have any effect at caution signals. This means that there is a range of speeds at the higher level which will be excluded from full protection. In spite of this, it is suggested in published data that 60% of accidents due to SPADs will be prevented by the installation of TPWS at critical locations. This is achieved, it is said, at 10% of the installation costs of a full ATP system.
TPWS does not replace the existing AWS system. AWS is retained, so the driver will still get the warnings advising him of adverse signals. The TPWS equipment is designed to interface with the existing on-board wiring of trains so that it can be fitted quickly.
ATP (Automatic Train Protection) or TPWS
An increasing number of railways around the world are provided with ATP. ATP provides a either a continuous or regular update of speed monitoring for each train and causes the brakes to apply if the driver fails to bring the speed within the required profile. There are various versions of ATP, some of which are described in our Metro Signalling and ATP page. ATP is popular on metros because of the very dense train services provided and because many run for long distances in tunnels. New, or newly upgraded high speed railways also have ATP.
The main reason why existing railways have been slow to introduce ATP is because of the costs and because it is difficult to allow for the variable braking capabilities of different types of trains, in particular, freight trains. The varying size and braking abilities of freight trains means that data input for the on-board ATP computer has to be manual. Railway administrations have been reluctant to invest large sums of money in a safety system which, because of the possibility of manual input error, does not offer a total "vital" safety coverage. For the UK, the high price of full ATP has caused it to be rejected as the system-wide standard signalling safety system, so TPWS has been adopted as the nearest suitable and more cost-effective alternative.
For European main line railways, the required form of train protection on those routes intended for interoperable services (the TEN routes) will be based on the architecture of ERTMS, the European Rail Traffic Management System. The signalling part of ERTMS is called ETCS - European Train Control System. The system has been developed across Europe and installed on selected routes in a number of countries. In the UK, a trial version has been installed on the Cambrian route in Wales and the first section from Pwllheli to Harlech was commissioned on the 28th October 2010. A useful description of the trial installation has been published by the Rail Engineer magazine here. The ERTMS website has a description of the basic technical structure and operation.
It is worth saying here that passing a signal at danger is something that every driver fears and tries to avoid but knows that, in a moment of distraction or in an attempt to make up time, it will happen to him one day. The normal pattern of such incidents is that the signal (and, in case you ask, yes, it has happened to me) is passed because of an error of judgement in braking, not due to ignoring a signal aspect. Usually, such overruns are not at high speed and the overlap beyond the signal absorbs it so that the train does not enter the occupied block. Sometimes and more seriously, the overrun occurs when the driver misses a caution and cancels an AWS warning. Several collisions have occurred as a result of this and it is something which TPWS will almost eliminate.
One area where TPWS has turned out to be more trouble than it prevents is at terminal platforms. A too restrictive 10 mph speed limit on the approach to the buffers in a terminal platform has meant an increase in the time for a train to clear the routes into the terminus. At many termini in the UK, this has seriously affected capacity at peak times and has the effect of reducing the number of trains arriving and departing. This will lead to a reduction in service or will reduce recovery capability.
Since modern rolling stock is built to a high
standard of crashworthiness, a 20 mph buffer stop collision is unlikely
to cause a
serious deformation of a vehicle. Any speed restriction below
this level for
arriving trains causes a severe operating restriction on the terminus
apparent safety benefit. There should be an immediate increase in
terminal entry speed to 20 mph. Some terminals now have a 15 mph limit.
More articles on signalling:
The Development and Principles of UK Signalling - Metro Signalling and ATP - ATP Transmission and Moving Block Systems - ATO - Route Signalling - ATC - UK Warning & Protection Systems - UK Signalling - What the driver sees - Single Line Signalling - US Signalling