Readers of Rail Engineer are well acquainted with the huge construction projects around the country, aiming to increase capacity by building additional platforms and/or providing grade separated junctions. Examples include Norton Bridge, Heathrow Airport junctions, London Bridge, Reading, and Peterborough.
These schemes are extremely costly. They involve massive and very visible civil engineering works that create a significantly enlarged footprint of the station or junction, accompanied by substantial, mostly behind-the-scenes, alterations to the signalling and telecommunications (S&T) infrastructure, not forgetting the extensive operational changes that Network Rail and the TOCs have to plan for and implement.
With project costs coming under intense scrutiny, the Digital Railway is looking at options to deliver significant increases in performance and capacity as an alternative to the need for extensive remodelling works, which to a certain extent, are dictated by the design requirements of colour light signalling, as this article describes.
Limitations of multiple aspect signalling
Colour light signalling began to be installed on main line railways from the 1920s, providing train drivers with vastly improved signal sighting compared with semaphore signals, and facilitating the introduction of centralised signalling control centres.
However, colour light signals, in addition to indicating what we now call ‘movement authority’ (MA), also provide information to drivers as to the location at which braking should commence in order not to exceed the limit of the current MA. This is achieved by a red signal being preceded by a yellow caution aspect in three-aspect areas, or a double yellow followed by a single yellow in four-aspect areas.
It follows that the first caution signal must be positioned at least at braking distance from the signal displaying red. Distance ‘a’ (three-aspect signalling) and ‘x+y’ (four-aspect signalling) must be a minimum of braking distance. In four-aspect areas, some flexibility in signal positioning is permitted within the stipulation that the distance between the single yellow aspect and the red aspect (‘y’) shall be no less than one-third of the actual signalling braking distance between the double yellow aspect and the red aspect (x+y).
To get an appreciation for the signal spacing distances involved, on a level gradient, the braking distance for a line speed of 125 mph is 2,054 metres.
The full requirements are detailed in Railway Group Standard GK/RT0075 and it is incumbent upon the signal design engineer to ensure that the position of lineside signals shall be compatible with the braking performance of rolling stock so that trains moving at the permissible speed can stop within the actual signalling braking distance.
Put simply, when creating a signalling scheme plan, firstly signals are positioned to protect junctions, and control movements starting from platforms and sidings. The preceding signals must be positioned to provide the necessary braking distance. It is these significant distances involved that may impact upon performance and capacity, as will be outlined in a moment.
European Train Control System (ETCS)
ETCS is not a complete signalling system in its own right but provides an interface between signalling trackside infrastructure and individual trains. The Driver Machine Interface (DMI) in the driving cab displays the distance for which the train is authorised to travel, and the maximum speed allowed. If the onboard computer predicts that these values are likely to be exceeded, the system intervenes to safeguard operation of the train.
At ETCS Level 2, the onboard equipment transmits and receives data from the signalling centre via the GSM-R radio network and the Radio Block Control. Balises in the four-foot communicate with the train to provide position references. All other conventional signalling equipment is provided, including train detection, point operating machines, interlocking, and signaller interface.
Lineside signals may or may not be provided, but retaining them allows trains to run on the route whether or not they are fitted with ETCS. Not providing lineside signals, as on the Cambrian line’s early deployment scheme, means that only trains with a healthy ETCS may operate on the line.
The important difference, compared with conventional multiple-aspect signalling, is that braking distances are continuously re-calculated by the ETCS onboard computer in accordance with the MA received from the interlocking, and geographic data such as speed limits. To achieve an accurate stopping position, the driver will look out of the cab window and observe the physical location of the end of MA. This is achieved by the provision of either conventional colour light signals (if fitted) and/or non-illuminated, reflectorised Block Markers.
The positioning of Block Markers, unlike colour light signals, is not constrained by braking distance and, in conjunction with extra train detection sections, additional Block Markers may be provided to allow trains to close-up, thereby increasing capacity.
MA is continuously updated on the DMI, allowing the driver to accelerate immediately when conditions ahead improve, rather than having to wait for the next signal to come into view.
Closing up on the train ahead
The example in Fig 3 shows a conventionally signalled converging junction. Two trains, 1P35 and 1K23, approach the junction at about the same time, and 1P35 is allowed to proceed, forcing 1K23 to come to a stand at signal 514, which is displaying a red aspect. It has to remain at a stand until the rear of 1P35 clears the overlap of signal 508, by which time 1P35 is running at line speed and over 1.5 miles from 1K23, which is starting from rest with an ever increasing distance from 1P35 ahead.
The overlap is a safety over-run, typically 180 metres, to provide a margin of protection against the driver slightly misjudging the braking, or the train’s brakes not operating to full efficiency.
Train 1K23 moves off with a single yellow aspect at 514, and the driver cannot accelerate up to line speed as he has to anticipate the possibility that the next signal is at red. For the next mile or so the train proceeds at a significantly lower speed than permitted since the driver doesn’t receive any movement authority update until the next signal, 508, comes into view, which will depend upon curvature and general visibility of the line ahead.
1P35 is making good progress and is now several signal sections ahead. When 508 comes into view it is displaying green, authorising the driver of 1K23 to accelerate to line speed, though by now the significant gap with 1P35 is a considerable waste of capacity.
A similar problem arises when a train catches up with the train in front, which may be doing a station stop, or the route has not been set at a junction ahead. A train running at line speed thus encounters a restrictive aspect requiring a brake application to be made. Passing a single yellow, the speed of the train will be reducing to a level substantially lower than line speed in anticipation of a red signal ahead, thus causing a loss of time and capacity. Once again, the driver cannot accelerate until the next signal becomes visible, although the latter may have cleared up in the meantime.
ETCS facilitates closing up
For the benefit of ETCS operated trains, additional block sections may be installed, as in Fig 4, permitting 1K23 to move forward from 514 to close-up on 1P35, maintaining a much closer but safe separation, thereby improving throughput.
This philosophy of providing additional block sections has been adopted throughout the Thameslink core route. A train approaching 510 or 514 will be detected by the signalling system as operating in ETCS-mode or not. A train not fitted or operating in ETCS mode will not be able to take advantage of the close-up facility, as the system will require the entire block sections clear, up to and including the overlap of 508, before 510 or 514 will display a proceed aspect.
A train approaching 510 or 514 with ETCS live will see a single yellow if the route is set up to the first marker board.
Green banner signal
Having the earliest possible view of the next signal considerably helps drivers to attain optimal speed. In certain situations, where the approach visibility of a stop signal is compromised and does not comply with signalling standards, a banner repeater signal may be provided on the approach to the signal to facilitate earlier advice to the driver of the state of the signal.
Traditionally, banner repeater signals simply replicate the signal ‘on’ or ‘off’ condition but, in the last few years, LED technology has enabled the enhancement of a green aspect to a banner repeater signal. This is intended to encourage drivers to ‘open up’ before the actual signal comes into view, particularly when the previous signal passed was showing an aspect other than green.
Diverging junctions
Signalling principles were reviewed in the 1970s, prior to the introduction of high-speed trains (HSTs). At this time, flashing yellow aspects were introduced to facilitate drivers receiving advance warning when a train is signalled through a fast diverging junction, enabling the driver to adjust the speed of the train in order to match, but not to exceed, the turnout speed.
However slow-speed turnouts continue to use the time-honoured approach release from red of the junction signal, with the inevitable slow crawl approach due to the sighting and driving technique issues already described.
ETCS obviates the need for flashing yellows and approach controls on junction signals, since the ETCS knows the route set and turnout speed, calculating the braking curve and advising the driver accordingly via the DMI display. For trains on which ETCS is operating, the ability of approaching signals to display flashing aspects, or be approach released from red or yellow, is not necessary, and will be disabled within the signalling system for the passage of each ETCS train. Only standard aspect-sequences will be displayed to these trains. Route or junction indicators will continue to operate.
Variable driving styles
TOC professional driving policies emphasise techniques for the avoidance of SPADs (Signal Passed at Danger) and examples taken from the ‘East Coast Professional Driving Policy’ booklet of 2013 includes the following requirements:
“Reduce train speed gradually and continuously when approaching a red signal, planning to stop well before the signal.”
“If a signal clears from a red to a cautionary aspect as the train approaches it, remind yourself of the cautionary aspect using risk-triggered commentary. Also, limit acceleration and do not exceed 30 mph.”
“Treat approaching a red signal in the platform the same way as approaching any other stop signal.”
Such well meaning procedures are designed to reduce the SPAD risk but, ultimately, are cautious to the extent that individual driving styles may impact upon performance and have a knock-on effect upon the service.
Although the ETCS DMI provides the driver with a visual display of maximum permitted speed and target speed, controlling the actual speed is still in the hands of the driver. Consistent driving with actual braking that closely aligns with the calculated curve will be achieved by the provision of Automatic Train Operation. The Thameslink core is the first application on Network Rail of ATO and preparation, including driver training, is under way in readiness for the Class 700 fleet to commence ATO operation later this year.
The problem of through stations
Fig 5 shows a typical layout where points are in the overlap of the starting signals. The first train arrives in Platform 15 with the forward route set for its departure onwards from 389 to 405. Unfortunately, this locks the overlap points 627 normal, and the route cannot be set for a following train to approach 387. This subsequent train has to wait at 349 (which may be several hundred metres outside the station) until the first train has departed and cleared the overlap of 389.
As an alternative, Fig 6 shows a solution whereby the platform starting signals have their own, separate plain-line overlaps. Thus with a train in Platform 15 and route set from 389, the signaller can set the route for the following train from 349 to 387 so that this train can arrive whilst the first train is completing station duties. The next train can then be signalled into Platform 15 with the route set from 387 to 405, and so on, saving valuable minutes by obviating standing out at 349.
However, this option requires two platforms and a significantly longer station footprint to allow for the provision of the plain line overlaps, which will be costly and not always physically possible if the station is built within the constraints of a viaduct. This solution does provide immediate capacity improvement for all trains and is how the extensively remodelled layouts at Reading and London Bridge have been configured.
The ETCS solution in Fig 7 avoids the costly provision of another platform by the provision of additional block sections, allowing a following train to close up on the approach to the platform as the previous train departs. This solution would increase capacity without the need for expensive civil and track engineering works, but may not be the right solution if a second platform is needed for the layover of terminating trains.
Every ETCS fitted train will benefit from this option, but the overall improvement in capacity will incrementally increase as more trains are fitted with ETCS.
A comparison between the traditional signalling and ETCS solution can be observed in the revised layout at London Bridge High Level, where the Canon Street lines have three platforms and the Charing Cross lines have four platforms, allowing trains to alternate between platforms to increase the throughput of trains, whereas the Thameslink route adopts the ETCS solution and utilises only two platforms (4 and 5) and relies on the closing-up flexibility of ETCS, provided by additional train detection sections.
ETCS Level 3
Whilst the ETCS Level 2 solution, with additional hard-wired block sections, may facilitate the desired increase in line capacity, it contravenes the desirability for the ETCS project to reduce the amount of trackside signalling infrastructure. Level 2 additional block sections require the costly purchase and installation of trackside train-detection equipment, with ongoing maintenance resource implications. The affordability is questionable, and the many additional potential equipment failure points could compromise reliability of the ERTMS Level 2 operation.
ETCS Level 3 differs from Level 2 by obviating the provision of fixed blocks using trackside equipment. Safe separation of trains is achieved by fixed virtual blocks based on train position information reported by the train, or full moving block with dynamic handling of reported train position information. Level 3 requires each train to have a Train Integrity Monitor (TIM) in order to report that the train is complete (with no carriages or wagons left behind).
Clearly, Level 3 is the way forward, delivering better capacity, reducing costs and increasing reliability. Various Level 3 options are currently in development, including the possibility of overlaying onto legacy signalling systems with train detection, thereby providing a fall-back in the event of the failure of the onboard ETCS or loss of train integrity. Provision could also be made to operate trains not fitted with ETCS if lineside signals are retained.
Level 3, including ‘Hybrid Level 3’, which utilises existing train detection, was described more fully in issue 151 (May 2017). There are various Level 3 options at different stages of development. Hybrid Level 3 is the most advanced and is seen as the low-risk solution, given the anticipated simple and smooth migration path from existing trackside signalling systems and trains.
A demonstration of Hybid Level 3 is expected to be up and running in the coming months.
Read more: Discussion on “Making a Success of the Digital Railway”
