Track circuit

A track circuit is a simple electrical device used to detect the absence of a train on rail tracks, used to inform signallers and control relevant signals.

Principles and operation

The basic principle behind the track circuit lies in the connection of the two rails by the wheels and axle of locomotives and rolling stock to short out an electrical circuit. This circuit is monitored by electrical equipment to detect the absence of the trains. Since this is a safety appliance, fail-safe operation is crucial; therefore the circuit is designed to indicate the presence of a train when failures occur. On the other hand, false occupancy readings are disruptive to railway operations and are to be minimized.

Track circuits allow railway signalling systems to operate semi-automatically, by displaying signals for trains to slow down or stop in the presence of occupied track ahead of them. They help prevent dispatchers and operators from causing accidents, both by informing them of track occupancy and by preventing signals from displaying unsafe indications.

The basic circuit

A track circuit typically has power applied to each rail and a relay coil wired across them. When no train is present, the relay is energised by the current flowing from the power source through the rails. When a train is present, its axles short (shunt) the rails together; the current to the track relay coil drops, and it is de-energised. Circuits through the relay contacts therefore report whether or not the track is occupied.

Each circuit detects a defined section of track, such as a block. These sections are separated by insulated joints, usually in both rails. To prevent one circuit from falsely powering another in the event of insulation failure, the electrical polarity is usually reversed from section to section. Circuits are powered at low voltages (1.5 to 12 V DC) to protect against line power failures. The relays and the power supply are attached to opposite ends of the section to prevent broken rails from electrically isolating part of the track from the circuit. A series resistor limits the current when the track circuit is short-circuited.

Circuits under electrification

In some railway electrification schemes, one or both of the running rails are used to carry the return current. This prevents use of the basic DC track circuit because the substantial traction currents overwhelm the very small track circuit currents.

Where DC traction is used on the running line or on tracks in close proximity then DC track circuits cannot be used, similarly if 50 Hz AC electrification is used then 50 Hz AC track circuits cannot be used.

To accommodate this, AC track circuits use alternating current signals instead of direct current (DC) but typically, the AC frequency is in the range of audio frequencies, from 91 Hz up to 10 kHz. The relays are arranged to detect the selected frequency and to ignore DC and AC traction frequency signals. Again, failsafe principles dictate that the relay interprets the presence of the signal as unoccupied track, whereas a lack of a signal indicates the presence of a train. The AC signal can be coded and locomotives equipped with inductive pickups to create a cab signalling system.

There are two common approaches to provide a continuous path for traction current that spans multiple track circuit blocks. The simplest method installs insulated track circuit joints on only one of the two rails with the second being a path for the return current and a ground for the track circuit rail. This has the disadvantage of only being able to detect breaks in one rail so the more popular two rail system uses impedance bonds to permit traction current to pass between isolated track circuit blocks while blocking current at track circuit frequencies.

AC circuits are sometimes used in areas where conditions introduce stray currents, which interfere with DC track circuits.

In some countries, AC-immune DC track circuits are used on AC electrified lines. This is the predominant method of track circuiting on overhead electrified parts of the UK rail network. One method provides 5 V DC to the rails, one of the rails being the traction return and the other being the signal rail. When a relay is energised and attached to the track, normal voltage is 5 V DC. When there is a break in the circuit and there is no train, the voltage rises to 9 V DC which provides a very good means for fault finding. This system filters out the voltage induced in the rails from the overhead lines. These track circuits are limited in length to about 300m.

Jointless track circuits

Modern track is often continuously welded, the joints being welded during installation. This offers many benefits to all but the signalling system, which no longer has natural breaks in the rail to form the block sections. The only method to form discrete blocks in this scenario is to use different audio frequencies (AF) in each block section. To prevent the audio signal from one section passing into an adjacent section, pairs of simple tuned circuit are connected across the rails at the section boundary. The tuned circuit often incorporates the circuit to either apply the transmitted signal to the track or recover the received signal from the other end of the section.

Consider a railway with two block sections as in the diagram. Section 1 has frequency A injected at the left-hand end and received at the right-hand end. Section 2 continues from the right hand end of section 1 where frequency B is injected and then received at the right-hand end of section 2.

There is often a gap between where frequency A is received and frequency B is injected. This is referred to as a 'tuned zone' and is a section of track where the amplitude of frequency A reduces in the direction of section 2 and the amplitude of frequency B reduces in the direction of section 1. The tuned zone can be of the order of 20 m long.

Advantages of jointless track circuits:

• Eliminates insulated block joints, a component liable to mechanical failure (both of insulation and by introducing stress to adjoining rails) and maintenance.
• In electrified areas, jointless track circuits require fewer impedance bonds than any other double rail traction return track circuits.

Disadvantages of jointless track circuits:

• Restrictions on placing impedance bonds, hence any connection for electrification purposes, in or near tuned zones as this may upset the filter properties of the tuned zone.
• Electronic circuits are more vulnerable to lightning strikes.

Some popular AF track circuits used in the UK

Some of the earliest audio frequency track circuits still in use today were made by the Aster company in France. Frequencies of the Aster SF 15 type track circuit are 1700 Hz and 2300 Hz on one track and 2000 Hz and 2600 Hz on the other. SF stands for 'single frequency' and was the name given to the units made under licence by ML Engineering in Plymouth, UK. These frequencies are by definition unmodulated. The lack of modulation can lead to availability problems as well as making the signalling safety case difficult to produce. To address these problems, modulated track circuits were developed like the TI21 system.

TI21 type track circuits (now known as EBI Track 200) use eight nominal frequencies, from 1549 Hz to 2593 Hz for main line applications and eight frequencies from 5600 Hz to 8400 Hz for metro applications (designated TI21-M or EBI Track 300). Actual transmission is ± 17 Hz around the nominal frequency for main line and ±100 Hz for metro. The signal is FSK modulated at 4.8 Hz (20 Hz for metro) unless overridden by the MOD terminal on the front panel. TI stands for 'traction immune' and was the name used by ML Engineering in Plymouth. ML Engineering was taken over by various companies and is owned by Bombardier Transportation (2009). TI21 main line track circuits can be up to 1100 m in length. This can be extended to 2200 m with compensating capacitors.

To simplify traction pack design in locomotives, many track circuit manufacturers now transmit a unique code from the transmitter to the receiver. This offers improvements in availability, simplification in signalling system design and more robust safety cases. Such systems include the Siemens FTG S, Westinghouse (Invensys) FS3000, Bombardier EBI Track 400 and Alstom's Digicode and Jade.

CSEE UM71

CSEE (now Ansaldo STS) UM71 is another kind of jointless track circuit. It uses 1700 Hz and 2300 Hz on one track and 2000 Hz and 2600 Hz on the other.^[1] To reduce the chance of stray currents causing a wrong side failure the basic frequencies are modulated ±15 Hz or so. Different rates of modulation can be detected by equipment on the trains and used for ATC, so long as the transmitter end (Tx) is at the front of the train.

The TI21 and Westinghouse FS2500 jointless track circuits are similar to the UM71.

Data Pickup Unit

A jointless track circuit such as the CSEE can be divided with a Data Pickup Unit (DPU), which is cheaper than splitting it into two track circuits. A DPU avoids the need to change the frequency of a whole series of track circuits in a cascade. The DPU consists of a tuned coil which detects the presence or absence of current in the adjacent rail and picks up or drops a relay accordingly. One use of DPUs is for timing circuits. Each track circuit frequency has its own DPU tuned to that frequency. DPUs can be located almost anywhere; they overcome the limitation that Jointless tracks have a minimum length.

The UM71 DPU made by CSEE is triangular while the FS2500 DPU made by Westinghouse is rectangular.

DC Coded track circuits

In non-electrified areas, DC coded track circuits may be used. These modulate the current going from the powersource end to the relay end and control the signals and cab signals without the need for line wires. The modulated currents can be detected by equipment on the train to provide cab signalling. ^[2] They can be overlain by predictor systems to operate level crossings.^[3]

Brands of coded track circuit include:

• Microtrax by Union Switch & Signal.
• Invensys PSO 4000 ^[4]

Cut tracks

Where the length of a section exceeds the practical length of a track circuit, cut tracks can be provided. With a cut track, the relay of the last track cuts the powersource feed of the second last track circuit, and so on. Cut tracks are only suitable for unidirectional tracks.

Track circuits with ballast contamination will be shorter than those with good ballast, thus needing more cut tracks.

Failure modes and prevention

Wheels and brakes

Railway wheels are made from steel and provide a good short circuit from rail to rail (shunt resistance).

Longer trains with more wheels have better conductivity. Short trains or single engines can be a problem. Trains with a single Budd railmotor, which are also lightweight, and with discbrakes, had some problems when they stopped, and had to make a double stop to ensure good contact with the rails.

Cast iron brake shoes tend to clean the wheels of non-conductive debris (such as leaves and sand-based traction compounds), while disc brakes do not. As a result, some disc-braked vehicles have "scrubber pads" cleaning the wheels to aid in proper track circuit operation.

Relays

Track circuit relays, referred to by signal maintainers as "vital relays," are specially designed to reduce the chance of wrong-side failures. They may, for example, have carbon-silver contacts to reduce the likelihood of the wrong contacts welding shut after power surges and lightning strikes.

Circuit failures

The circuit is designed so that most failures will cause a "track occupied" or Track Occupancy Light (TOL) indication (known as a "Right Side" failure in the UK). For example:

• A broken rail or wire will break the circuit between the power supply and the relay, de-energizing the relay. See exception below.
• A failure in the power supply will de-energize the relay.
• A short across the rails or between adjacent track sections will de-energize the relay.

On the other hand, failure modes which prevent the circuit from detecting trains (known as a "Wrong Side" failure in the UK) are possible. Examples include:

• Mechanical failure of the relay, causing the relay to be stuck in the "track clear" position even when the track is occupied.
  • One perspex case warped in the heat, and touched the relay contacts, holding them up.
  • Another relay saw a metal washer slip off and jam the relay contacts up; the half-washers had to be replaced by full-circle washers.
• Conditions which partially or completely insulates the wheels from the rail, such as rust, sand, or dry leaves on the rails. This is also known as "poor shunting" ("failure to shunt" in North America and Australia). Sanding gear which operates on all wheels of a light engine can temporarily insulate it from the rails until the sanding ceases and the locomotive has moved further down the track.
• Conditions in the trackbed (roadbed) which create stray electrical signals, such as muddy ballast (which can generate a "battery effect") or parasitic electrical currents from nearby power transmission lines.
• Parasitic oscillations in the equipment that controls the track circuits.^[5]
• Equipment which is not heavy enough to make good electrical contact (shunt failure) or whose wheels must be electrically insulated.
• A rail break between the insulated rail joint and the track circuit feed wiring would not be detected.

Failure modes that result in an incorrect "track clear" signal (known usually in the US as a "false clear") may allow a train to enter an occupied block, creating the risk of a collision. Wheel scale and short trains may also be a problem. They may also cause the warning systems at a grade crossing to fail to activate. This is why in UK practice, a treadle is also used in the circuitry.

Different means are used to respond to these types of failures. For example, the relays are designed to a very high level of reliability. In areas with electrical problems, different types of track circuits may be used which are less susceptible to interference. Speeds may be restricted when and where fallen leaves are an issue. Traffic may be embargoed in order to let equipment pass which does not reliably shunt the rails.

Sabotage is possible. In the 1995 Palo Verde derailment, saboteurs electrically connected sections of rail which they had displaced to conceal the breaks in the track they had made. The track circuit therefore did not detect the breaks, and the engine driver was not given a "Stop" indication. Another form of sabotage, not intended to cause a train crash but merely to make trains stop and slow down unnecessarily in an effort to sabotage an economy or potential injuries, is to tie a wire between the 2 rails, causing a false obstruction signal.^[6][7]

Railhead contamination and rust

The track circuit relies upon an adequate electrical contact between the rail and the wheel; contamination can insulate the one from the other. A common problem is fallen leaves, though there have been cases where crushed insects have also caused detection failures.^[8]

A more persistent problem is rust. Usually the railhead is kept clean of rust by the regular passage of trains' wheels. Lines which are not used regularly can become so rusty as to prevent vehicles being detected; seldom-used points and crossovers and the extremities of terminal platform lines are also prone to rusting. Measures to overcome this include:

• Depression bars or treadles to detect vehicles;
• Stainless steel strips (often zig zag in shape) welded on the railheads;
• High voltage impulse track circuits;
• Track Circuit Assistor (TCA) - a train mounted system that breaks down the resistance of the rust layer;
• Axle counters over the affected section; and/or
• tunnel sticks whereby a track circuit cannot pickup unless a train is detected in the next track circuit.

Scale

Insulated blockjoints can be bridged by wheel scale in some circumstances causing one or two track circuits to fail. This problem may be reduced by having a pair of blockjoints in series about 4m apart. The short 4m section would not itself be track circuited.^[9]

Immunization

Electric locomotives must avoid generating noise in the frequencies that track circuits use. The SNCB Class 13 had such problems.

Transitory problems

A short, lightweight and fast train passing over an insulated blockjoint may disappear from the departing track circuit before it appears in the arriving track circuit, allowing false clear signals to be given. This problem can be overcome by introducing a time delay of say 1 – 2 seconds into the departing track circuit. Electronic track circuits such as the CSEE can easily incorporate such a time delay.

Transmission of status

Track circuit occupancy status, along with status of other signal and switch related devices, may be integrated with a local control panel as well as a remote rail control centre. If the track circuit contains a relay, it can be connected to a device for sending status information via a communications link. The status can then be displayed and stored for archival for purposes of incident investigation and operations-related analysis. Many signalling systems also have local event recorders for recording track circuit status.

Siding turnout

It is sometimes convenient to wire the detectors of a set of points through the track circuit over those points. This can be done in one of two ways:

• a contact of the points detector can shunt the track circuit when the points are reverse, putting the signals to red, however this is not failsafe.
• the track circuit can be split with extra blockjoints and the detectors in the points complete the track circuit when the points are normal and the signal is entitled to receive a green light. This is partially fail-safe.
• a second relay can be installed on the turnout, with its contacts wired in series with the main relay. This is fail-safe but expensive.

Track circuit clips

A simple piece of safety equipment that can be carried by trains is a track-circuit clip. This is simply a length of wire connecting two metal spring clips that will clip onto a rail. In case of accident or obstruction a clip applied to a track will indicate that that track is occupied, therefore putting signals to danger. As an example of use, if a train is derailed on a double track, and is foul of the second track, application of a clip to the second track will immediately return signals protecting the second track to danger. This procedure is a much more effective safety measure than attempting to contact a signalling centre by telephone because its effect is immediate and automatic.

History

The first use of track circuiting was by William Robert Sykes on a short stretch of track of the London Chatham and Dover Railway at Brixton in 1864.^[10] The failsafe track circuit was invented in 1872 by William Robinson, an American electrical and mechanical engineer. His introduction of a trustworthy method of block occupancy detection was key to the development of the automatic signalling systems now in nearly universal use.^[11]:3ff

The first railway signals were manually operated by signal tenders or station agents. When to change the signal aspect was often left to the judgement of the operator. Human error or inattentiveness occasionally resulted in improper signalling and train collisions.

The introduction of the telegraph during the mid-nineteenth century showed that information could be electrically conveyed over considerable distance, spurring the investigation into methods of electrically controlling railway signals. Although several systems were developed prior to Robinson's, none could automatically respond to train movements.

Robinson first demonstrated a fully automatic railway signalling system in model form in 1870. A full-sized version was subsequently installed on the Philadelphia and Erie Railroad at Ludlow, Pennsylvania (aka Kinzua, PA), where it proved to be practical. His design consisted of electrically operated discs located atop small trackside signal huts, and was based on an open track circuit. When no train was within the block no power was applied to the signal, indicating a clear track.^[11]:4

An inherent weakness of this arrangement was that it could fail in an unsafe state. For example, a broken wire in the track circuit would falsely indicate that no train was in the block, even if one was. Recognizing this, Robinson devised the closed loop track circuit described above, and in 1872, installed it in place of the previous circuit. The result was a fully automatic, failsafe signalling system that was the prototype for subsequent development.^[11]:6–9

Although a pioneer in the use of signals controlling trains, the UK was slow to adopt Robinson's design. At the time, many carriages on UK railways had wooden axles and/or wheels with wooden hubs, making them incompatible with track circuits.

Accidents

Caused by lack of track circuits

Numerous accidents would have been prevented by the provision of track circuits, including:

• Norton Fitzwarren rail crash (1890)
• Redfern (1894) - incorrectly set signal
• Hawes Junction rail crash (1910)
• Quintinshill rail crash (1915)
• Geurie crossing loop collision (1963)

Caused by track circuit failure

Much rarer are accidents caused when the track circuits themselves fail. For example:

• Cowan rail disaster, which occurred when sand on the rails insulated the wheels from the rails, causing a failure to shunt that allowed a trailing block signal to improperly display a clear aspect, resulting in a rear end collision.
• June 22, 2009 Washington Metro train collision, which occurred when a parasitic oscillation developed in a track circuit control module.^[5]

Broken rails

Because Track circuits operate by passing a current through one or both of the tracks, they can sometimes detect if a rail has broken completely. However, if the break is only partial or is at a turnout (set of points) detection may not be possible.

• Hither Green rail crash (1967) - derailment due to top of rail having broken off
• Weyauwega derailment (1996) - undetected broken rail in turnout.

See also

• Application of railway signals
• Track circuit interrupter
• Wrong-side failure

References