Railway signalling is a system used to control railway traffic safely, essentially to prevent trains from colliding. Being guided by fixed rails, trains are uniquely susceptible to collision; furthermore, trains cannot stop quickly, and frequently operate at speeds that do not enable them to stop within sighting distance of the driver. This necessitated the establishment of strict guidelines for time keeping and railway chronometers in 1891 by the general time inspector Webb C. Ball of Cleveland, Ohio, USA. In the UK, the Regulation of Railways Act 1889 introduced a series of requirements on matters such as the implementation of interlocked block signalling and other safety measures as a direct result of the Armagh rail disaster in that year.
Most forms of train control involve movement authority being passed from those responsible for each section of a rail network (e.g., a signalman or stationmaster) to the train crew. The set of rules and the physical equipment used to accomplish this determine what is known as the method of working (UK), method of operation (US) or safeworking (Aus.). Not all these methods require the use of physical signals and some systems are specific to single track railways.
The earliest rail cars were first hauled by horses or mules. A mounted flagman on a horse preceded some early trains. Hand and arm signals were used to direct the “train drivers”. Foggy and poor-visibility conditions gave rise to flags and lanterns. Wayside signalling dates back as far as 1832, and used elevated flags or balls that could be seen from afar.
The simplest form of operation, at least in terms of equipment, is to run the system according to a timetable. A fixed schedule is drawn up with which every train crew must be familiar. Trains may only run on each section of track at their scheduled time, during which they have 'possession' and no other train is permitted to use the same section.
When trains are running in opposite directions on a single-track railroad, meeting points ("meets") are scheduled, at which each train must wait for the other at a passing place. Neither train is permitted to move before the other has arrived. In the US the display of two green flags (green lights at night) is an indication that another train is following the first and the waiting train must wait for the next train to pass. In addition, the train carrying the flags gives eight blasts on the whistle as it approaches. The waiting train must return eight blasts before the flag carrying train may proceed.
The timetable system has several disadvantages. First, there is no positive confirmation that the track ahead is clear, only that it is scheduled to be clear. The system does not allow for engine failures and other such problems, but the timetable is set up so that there should be sufficient time between trains for the crew of a failed or delayed train to walk far enough to set up warning flags, flares, and detonators or torpedoes (UK and US terminology, respectively) to alert any other train crew.
A second problem is the system's inflexibility. Trains cannot be added, delayed, or rescheduled without advance notice.
A third problem is a corollary of the second: the system is inefficient. To provide flexibility, the timetable must give trains a broad allocation of time to allow for delays, so the line is in the possession of each train for longer than is otherwise necessary.
Nonetheless, this system permits operation on a vast scale, with no requirements for any kind of communication that travels faster than a train. Timetable operation was the normal mode of operation in North America in the early days of the railroad.
With the advent of the telegraph in 1851, a more sophisticated system became possible because this provided a means whereby messages could be transmitted ahead of the trains. The telegraph allows the dissemination of any timetable changes, known as train orders. These allow the cancellation, rescheduling and addition of train services.
Train crews generally receive their orders at the next station at which they stop, or they are sometimes handed up to a locomotive 'on the run' via a long staff. Train orders allowed dispatchers to set up meets at sidings, force a train to wait in a siding for a priority train to pass, and to maintain at least one block spacing between trains going the same direction.
Timetable and train order operation was commonly used on American railroads until the 1960s, including some quite large operations such as the Wabash Railroad and the Nickel Plate Road. Train order traffic control was used in Canada until the late 1980s on the Algoma Central Railway and some spurs of the Canadian Pacific Railway.
Timetable and train order was not used widely outside North America, and has been phased out in favor of radio dispatch on many light-traffic lines and electronic signals on high-traffic lines. More details of North American operating methods is given below.
Trains cannot collide with each other if they are not permitted to occupy the same section of track at the same time, so railway lines are divided into sections known as blocks. In normal circumstances, only one train is permitted in each block at a time. This principle forms the basis of most railway safety systems.
On double-tracked railway lines, which enabled trains to travel in one direction on each track, it was necessary to space trains far enough apart to ensure that they could not collide. In the very early days of railways, men (originally called 'policemen') were employed to stand at intervals ("blocks") along the line with a stopwatch and use hand signals to inform train drivers that a train had passed more or less than a certain number of minutes previously. This was called "time interval working". If a train had passed very recently, the following train was expected to slow down to allow more space to develop.
The watchmen had no way of knowing whether a train had cleared the line ahead, so if a preceding train stopped for any reason, the crew of a following train would have no way of knowing unless it was clearly visible. As a result, accidents were common in the early days of railways. With the invention of the electrical telegraph, it became possible for staff at a station or signal box to send a message (usually a specific number of rings on a bell) to confirm that a train had passed and that a specific block was clear. This was called the "absolute block system".
Fixed mechanical signals began to replace hand signals from the 1830s. These were originally worked locally, but it later became normal practice to operate all the signals on a particular block with levers grouped together in a signal box. When a train passed into a block, a signalman would protect that block by setting its signal to 'danger'. When an 'all clear' message was received, the signalman would move the signal into the 'clear' position.
The block system came into use gradually during the 1850s and 1860s and became mandatory in the United Kingdom after Parliament passed legislation in 1889 following a number of accidents, most notably the Armagh rail disaster. This required block signalling for all passenger railways, together with interlocking, both of which form the basis of modern signalling practice today. Similar legislation was passed by the United States around the same time.
Not all blocks are controlled using fixed signals. On some single track railways in the UK, particularly those with low usage, it is common to use token systems that rely on the train driver's physical possession of a unique token as authority to occupy the line, normally in addition to fixed signals.
Before allowing a train to enter a block, a signalman must be certain that it is not already occupied. When a train leaves a block, he must inform the signalman controlling entry to the block. Even if the signalman receives advice that the previous train has left a block, he is usually required to seek permission from the next signal box to admit the next train. When a train arrives at the end of a block section, before the signalman sends the message that the train has arrived, he must be able to see the end-of-train marker on the back of the last vehicle. This ensures that no part of the train has become detached and remains within the section. The end of train marker might be a white disc by day or a steady or flashing red lamp. If a train has entered the next block before the signalman sees that the disc or lamp is missing, he will ask the next signal box to stop the train and investigate.
Under a permissive block system, trains are permitted to pass signals indicating the line ahead is occupied, but only at such a speed that they can stop safely driving by sight. This allows improved efficiency in some situations and is mostly used in the USA, and in most countries is restricted to freight trains only, and may be restricted depending on the level of visibility.
Permissive block working may also be used in an emergency, either when a driver is unable to contact a signalman after being held at a danger signal for a specific time, although this is only permitted when the signal does not protect any conflicting moves, and also when the signaller is unable to contact the next signal box to make sure the previous train has passed, for example if the telegraph wires are down. In these cases, trains must proceed at very low speed (typically 20mph or less) so that they are able to stop short of any obstruction. In most cases this will not be allowed during times of poor visibility (e.g fog or falling snow).
An absolute block system is itself not entirely absolute. Multiple trains may enter a block with authorisation. This is necessary in order to split or join trains together, rescue failed trains and the like. In giving authorisation, the signalman also ensures the driver knows precisely what to expect ahead, and the driver must operate the train in a safe manner taking this information into account. Generally, the signal will remain at danger, and the driver will be given verbal authority, usually accompanied by a yellow flag, to pass a signal at danger, and the presence of the train in front will be explained. At locations where trains regularly enter occupied blocks, such as stations where coupling takes place, a subsidiary signal, sometimes known as a "calling on" signal, will be provided for these movements, otherwise they are accomplished through train orders.
Under automatic block signalling, signals indicate whether or not a train may enter a block based on automatic train detection indicating whether a block is clear. The signals may also be controlled by a signalman, so that they only provide a proceed indication if the signalman sets the signal accordingly and the block is clear.
One of the most common ways to determine whether a block is occupied is by use of a track circuit. The track at either end of each block is electrically isolated from the next block, and an electrical current is fed to both running rails at one end of the block. A relay at the other end is connected to both rails. When the block is unoccupied, the relay completes an electrical circuit, and is energized. However, when a train passes a signal and enters the block, it short-circuits the current in the block, and the relay is de-energized.
This method does not explicitly need to check that the entire train has left the block. If part of the train is left in the block, that part will continue to be detected by the track circuit.
This type of circuit is used to detect trains, both for the purpose of setting the signal indication and for providing various interlocking functions — for example, not permitting points to be moved when a train is standing over them. Electrical circuits are also used to prove that switch points are in the appropriate position before a signal over them may be cleared. Modern UK trains, and staff working in track circuit block areas, carry operating clips so that, in the event of a derailment fouling an adjacent track, the track circuit can be short-circuited. This triggers danger signals on that track and can be used to prevent a collision before the crew is able to contact a signalman.
An alternative method of determining the occupied status of a block is using devices located at its beginning and end that count the number of axles entering and leaving. If the same number leave the block as enter it, the block is assumed to be clear. Although axle counters can provide similar functionality to track circuits, they also exhibit a few other characteristics. In a moist environment an axle counted section can be far longer than a track circuited one. The low ballast resistance of very long track circuits reduces their sensitivity. Track circuits can automatically detect some types of track defect such as a broken rail. In the event of power restoration after a power failure, an axle counted section is left in an undetermined state until a train has passed through the affected section. A track circuited section will detect the presence of a train in section immediately.
Most blocks are "fixed", i.e. they include the section of track between two fixed points. On timetable, train order, and token-based systems, blocks usually start and end at selected stations. On signalling-based systems, blocks start and end at signals.
The lengths of blocks are designed to allow trains to operate as frequently as necessary. A lightly-used line might have blocks many kilometres long, but a busy commuter line might have blocks a few hundred metres long.
A train is not permitted to enter a block until a signal indicates that the train may proceed, a dispatcher or signalman instructs the driver accordingly, or the driver takes possession of the appropriate token. In most cases, a train cannot enter the block until not only the block itself is clear of trains, but there is also an empty section beyond the end of the block for at least the distance required to stop the train. In signalling-based systems with closely-space signals, this overlap could be as far as the signal following the one at the end of the section, effectively enforcing a space between trains of two blocks.
When calculating the size of the blocks, and therefore the spacing between the signals, the following have to be taken into account:
Historically, some lines operated so that certain large or high speed trains were signalled under different rules and only given the right of way if two blocks in front of the train were clear.
One disadvantage of having fixed blocks is that the faster trains are allowed to run, the longer the stopping distance, and therefore the longer the blocks need to be, thus decreasing the line's capacity.
Under a moving block system, computers calculate a 'safe zone' around each moving train that no other train is allowed to enter. The system depends on knowledge of the precise location and speed and direction of each train, which is determined by a combination of several sensors: active and passive markers along the track and trainborne tachometers and speedometers (GPS systems cannot be used because they do not work in tunnels.) With a moving block, lineside signals are unnecessary, and instructions are passed directly to the trains. This has the advantage of increasing track capacity by allowing trains to run closer together while maintaining the required safety margins.
Moving block is in use on London's Docklands Light Railway, New York's Canarsie "L" Line, and planned for future use on the Jubilee Line. It was supposed to be the enabling technology on the modernisation of Britain's West Coast Main Line which would allow trains to run at a higher maximum speed (140mph), but the technology was deemed not mature enough, considering the large number of junctions on the line, and the plan was dropped. It forms part of the European Rail Traffic Management System's level-3 specification for future installation in the European Train Control System, which will (at level 3) feature moving blocks that allow trains to follow each other at exact braking distances.
On most railways, physical signals are erected at the lineside to indicate to drivers whether the line ahead is occupied and to ensure that sufficient space exists between trains to allow them to stop.
Older forms of signal displayed their different aspects by their physical position. The earliest types comprised a board that was either turned face-on and fully visible to the driver, or rotated so as to be practically invisible. While this type of signal is still in use in some countries (e.g. France and Germany), by far the most common form of mechanical signal worldwide is the semaphore signal. This comprises a pivoted arm or blade that can be inclined at different angles. A horizontal arm is the most restrictive indication (for 'danger' or 'caution', depending on the type of signal).
To enable trains to run at night, one or more lights are usually provided at each signal. Typically this comprises a permanently-lit oil lamp with movable coloured spectacles in front that alter the colour of the light. The driver therefore had to learn one set of indications for day time viewing and another for night time viewing.
Whilst it is normal to associate the presentation of a green light with a safe condition, this was not historically the case. In the very early days of railway signalling, the first coloured lights (associated with the turned signals above) presented a white light for 'clear' and a red light for 'danger'. Green was originally used to indicate 'caution' but fell out of use when the time interval system was discontinued. A green light subsequently replaced white for 'clear', to address concerns that a broken red lens could be taken by a driver as a false 'clear' indication. It was not until scientists at Corning Glassworks perfected a shade of yellow without any tinges of green or red that yellow became the accepted colour for 'caution'.
Mechanical signals are usually remotely operated by wire from a lever in a signal box, but electrical or hydraulic operation is normally used for signals that are located too distant for manual operation.
On most modern railways, colour light signals have largely replaced mechanical ones. Colour light signals have the advantage of displaying the same aspects by night as by day, and require less maintenance than mechanical signals.
Although signals vary widely between countries, and even between railways within a given country, a typical system of aspects would be:
On some railways, colour light signals display the same set of aspects as shown by the lights on mechanical signals during darkness.
Signalling of British origin generally conforms to the principle of route signalling. Most railway systems around the world, however, use what is known as speed signalling.
Under route signalling, a driver is informed which route the train will take beyond each signal (unless only one route is possible). This is achieved by a route indicator attached to the signal. The driver uses his route knowledge, reinforced by speed restriction signs fixed at the lineside, to drive the train at the correct speed for the route to be taken. This method has the disadvantage that the driver may be unfamiliar with a route onto which he has been diverted due to some emergency condition. Several accidents have been caused by this alone. For this reason, in the UK drivers are only allowed to drive on routes that they have been trained on and must regularly 'practice' on lesser used diversionary routes to keep their route knowledge up to date.
Under speed signalling, the driver is not informed which route the train will take, but the signal aspect informs him at what speed he may proceed. Speed signalling requires a far greater range of signal aspects than route signalling, but less dependence is placed on drivers' route knowledge.
When the train is routed towards a diverging route that must be taken at a speed significantly less than the mainline speed, the driver must be given adequate prior warning. With speed signalling, the signals approaching the divergence will display aspects appropriate to control the train's speed, so no 'approach release' is required.
Under 'route signalling', the aspects necessary to control speed do not exist, so a system known as approach release is employed. This involves holding the junction signal at a restrictive aspect (typically 'stop') in order that the signals on the approach show the correct sequence of caution aspects. The driver will brake in accordance with the caution aspect, without necessarily being aware that the diverging route has in fact been set. As the train approaches the junction signal, its aspect may clear to whatever aspect the current track occupancy ahead will permit.
Where the turnout speed is the same, or nearly the same, as the mainline speed, approach release is unnecessary.
The consequence of a driver/engineer failing to respond to a signal's indication can be disastrous. As a result, various auxiliary safety systems have been devised. Any such system will necessitate the installation of trainborne equipment to some degree. Some systems only intervene in the event of a signal being passed at danger. Others include audible and/or visual indications inside the driver's cab to supplement the lineside signals. Automatic brake application occurs if the driver should fail to acknowledge a warning. Some systems act intermittently (at each signal), but the most sophisticated systems provide continuous supervision.
In-cab safety systems are of great benefit during fog, when poor visibility would otherwise require that restrictive measures be put in place.
The most recent train control systems use electronic systems to indicate the state of the track ahead on cab displays at all times and can halt a train automatically if a signal is passed at danger.
In the early days of the railways, signalmen were responsible for ensuring any points (US: switches) were set correctly before allowing a train to proceed. Mistakes were made which led to accidents, sometimes with fatalities. The concept of the interlocking of points, signals, and other appliances was introduced to improve safety. This prevents a signalman from operating appliances in an unsafe sequence, such as setting a signal to 'clear' while one or more sets of points in the route ahead of the signal are improperly set.
Early interlocking systems used mechanical devices both to operate the signalling appliances and to ensure their safe operation. From about the 1930s, electrical relay interlockings were used. Since the late 1980s, new interlocking systems have tended to be of the electronic variety.