Necessity of fixed distant signals

[brad465]

While watching some drivers' eye view videos along routes with bi-directional working (such as the SEML from Tonbridge-Ashford), I wondered if the fixed distant signals for wrong-sided working were necessary, and why a multi-colour distant signal couldn't be used? I can understand a fixed distant signal where the "red" is the buffers at the end of a single line, or where the red signal is at a token withdrawal machine (like on the Isle of Grain line). But if there is a permissible through line beyond, what's stopping the provision of a green or double yellow option that could allow a train to not have to slow down so much?

Another similar example is where a stop signal doesn't go above yellow, but again the line continues beyond the following signal. This is present at Taunton on Platforms 2 and 3 in the London-bound direction: the signals cannot physically show higher than yellow, even if the route has already been set at the junction beyond with a yellow/green signal.

 

[Railsigns]

The permissible speed through the crossover that returns trains to the right-direction line after running wrong-direction along a bi-directional line will usually be sufficiently low as to require the last wrong-direction signal to be approach released from red. The distant signal on approach is therefore restricted to showing a yellow aspect.

If the bi-directional signalling continues beyond the crossover, the distant signal should be able to display a green aspect, because it's possible for a train to continue running in the wrong direction beyond the crossover.

 

[Notabene]

At Hunmanby, down trains approaching on the single line usually see, a distant showing either yellow or double yellow, then a two aspect red/yellow signal before the station and the junction to double track. At the station there is a two aspect r/g starter.

Not a fixed distant but an unusual arrangement.

 

[edwin_m]

If the bi-directional signalling continues beyond the crossover, the distant signal should be able to display a green aspect, because it's possible for a train to continue running in the wrong direction beyond the crossover.

Or would it still be a fixed distant on the grounds that this move would be rare - trains would normally return right line at the next crossover. If they didn't it would be because of some work under way on the other track around the crossover itself, which would probably put a speed restriction on passing trains anyway.

 

[james_the_xv]

The permissible speed through the crossover that returns trains to the right-direction line after running wrong-direction along a bi-directional line will usually be sufficiently low as to require the last wrong-direction signal to be approach released from red.

This is the reason 99% of the time. Another one is coming into P3 at Oxenholme from the Windemere Branch, the signal at the London end of P3 releases from red on all main routes, so the distant is fixed (I think it may be a board now)

 

[Railsigns]

Or would it still be a fixed distant on the grounds that this move would be rare - trains would normally return right line at the next crossover.

I'm not aware of any bi-directional lines where the wrong-direction distant signals are fixed for that reason. The ECML, the GEML, and the LT&S lines for example have long stretches of bi-directional signalling, and there are working distant signals on approach to intermediate crossovers. The cost of providing a green aspect on a distant signal during a resignalling is minimal.

 

[Annetts key]

Some of the “unusual” arrangements are however due to cost reasons.

It’s likely to become less likely now that computer based interlocking is normally used for new schemes.

But back when it was relay interlocking that had to be manually wired up, there was enough of a cost saving that it was sometimes done.

To provide a green aspect you don’t just need a signal head with a green lamp assembly. You also need the extra control circuitry, an extra relay or relays and the lineside cable used for the signalling needs more cable cores. Plus the extra expense of the time and resources to carry out the safety and functional testing when it is commissioned.

This has changed with computer based interlocking (signalling).

The example given for Taunton may be one. But as I don’t know the area that’s well, it could be other reasons.

But there are some signals elsewhere (mainly exits from sidings, loops or similar or on branch lines or goods lines or signals on reversible signalled lines) where the signal has less aspects than you may expect.

Within the S&T department, we have diagrams (aspect sequence charts) and tables (control tables) that detail the technical requirements for each aspect of each signal.

 

[Railsigns]

I've just looked into the situation at Taunton. It's another case of signals being restricted to a yellow proceed aspect because the next signals beyond are always approach released from red.

 

[brad465]

I've just looked into the situation at Taunton. It's another case of signals being restricted to a yellow proceed aspect because the next signals beyond are always approach released from red.

Which I don't understand because, while a position 1 junction indicator lights up to access the Up Main, the line speed is constant throughout (40mph) from departing the platforms until joining the Up Main (where it increases to 100mph).

 

[Belperpete]

Which I don't understand because, while a position 1 junction indicator lights up to access the Up Main, the line speed is constant throughout (40mph) from departing the platforms until joining the Up Main (where it increases to 100mph).

If there is a junction indicator for a route, then the signal must not clear for that route before the point at which the junction indicator is readable.

 

[brad465]

If there is a junction indicator for a route, then the signal must not clear for that route before the point at which the junction indicator is readable.

If you're referring to how an indicator must light up before the colour light changes from red when a route is set, that's not what I'm referring to. I'm talking about why a restrictive aspect on approach to the junction signal is necessary here, given the line speed is the same both approaching the signal and through the crossover.

 

[Railsigns]

If there is a junction indicator for a route, then the signal must not clear for that route before the point at which the junction indicator is readable.

That is only true if the route is approach released from red, and even then it's not always true.

 

[Belperpete]

If you're referring to how an indicator must light up before the colour light changes from red when a route is set, that's not what I'm referring to. I'm talking about why a restrictive aspect on approach to the junction signal is necessary here, given the line speed is the same both approaching the signal and through the crossover.

No, I am referring to the issue that a signal's main aspect is visible from a much greater distance than a junction indicator is. Leading to the possibility that the driver could see say the green aspect, and think that he is taking the high speed route, and only later see that the junction indicator is also alight and that he is in fact taking the diverging route. Which is why routes with junction indicators are usually approach released ( or some other advance warning provided, such as flashing yellows, splitting distants, etc).

This approach release is generally not needed where both main and diverging routes are the same speed, or within 10 mph. Also not needed where the signal only has one route (i.e. there is no route where the aspect lights without the junction indicator). However, there was an exception to that: where the driver might think there could be another route. The standard example given for that was the case of a signal reading off the end of a bidirectional line, where a driver might think that there was a "straight ahead" route to continue running bang-road. In such cases, although the signal only has one route to return to the "right road", it is still provided with a junction indicator, and was often provided with the associated approach control too.

Some bidirectional signalling was done cheaply (e.g. SIMBIDs simplified bidirectional signalling). In such cases, providing the approach control meant that only a fixed yellow distant was needed, without an AWS electro. I seem to recall that simplified AWS was a feature of SIMBIDs schemes, such as the right direction AWS magnets not being suppressed for bang road moves.

 

[High Dyke]

As previously mentioned, it will come down to cost - it always does on the railway!

On the East Coast Main Line (ECML), near Peterborough, a fixed distant board was provided for special working (wrong direction over the Down Slow line) last year. This was for the duration of engineering work in the area. A colour light stop signal was also fitted, in conjunction with this. Both signal and sign are covered out of use now. The expense of providing full signalling for this would not be justified.

 

[Class15]

If there is a junction indicator for a route, then the signal must not clear for that route before the point at which the junction indicator is readable.

That’s not correct - at Camden Road the junction indicator for Gospel Oak comes on and signal clears as soon as the approaching train has left Caledonian Road & Barnsbury.

 

[Belperpete]

That’s not correct - at Camden Road the junction indicator for Gospel Oak comes on and signal clears as soon as the approaching train has left Caledonian Road & Barnsbury.

My original comment was made with respect to signals reading off bidirectional lines (the subject of this thread), sorry if I didn't make that clear.

I don't know the particular case you mention, but it is acceptable for both to clear immediately if the signal only becomes visible at the point at which the junction is readable, for example. There are other cases where a signal can clear before its junction indicator is readable, as I described in my subsequent post.

 

[brad465]

No, I am referring to the issue that a signal's main aspect is visible from a much greater distance than a junction indicator is. Leading to the possibility that the driver could see say the green aspect, and think that he is taking the high speed route, and only later see that the junction indicator is also alight and that he is in fact taking the diverging route. Which is why routes with junction indicators are usually approach released ( or some other advance warning provided, such as flashing yellows, splitting distants, etc).

This approach release is generally not needed where both main and diverging routes are the same speed, or within 10 mph. Also not needed where the signal only has one route (i.e. there is no route where the aspect lights without the junction indicator). However, there was an exception to that: where the driver might think there could be another route. The standard example given for that was the case of a signal reading off the end of a bidirectional line, where a driver might think that there was a "straight ahead" route to continue running bang-road. In such cases, although the signal only has one route to return to the "right road", it is still provided with a junction indicator, and was often provided with the associated approach control too.

Some bidirectional signalling was done cheaply (e.g. SIMBIDs simplified bidirectional signalling). In such cases, providing the approach control meant that only a fixed yellow distant was needed, without an AWS electro. I seem to recall that simplified AWS was a feature of SIMBIDs schemes, such as the right direction AWS magnets not being suppressed for bang road moves.

I see, is this something that could be improved in a future resignalling? All the signals in this case study are still filament bulb; LED would be brighter and a new signal setup could allow a flashing yellow option if needed.

 

[Annetts key]

The design principles still apply with LED signals. Don't forget that "brightness" is relative to the viewing angle. Conventional filament lamps have two glass lenses per aspect to focus the light from the filament. That applies to both the main aspect and each of the the junction indicator lights. LED signals also are focused. Hence from too wide an angle at distance, they too are hard to see in daylight.

With computer based signalling, flashing aspects essentially comes for free (as in no extra hardware is normally needed) so that's the currently preferred method (as I understand it).

 

[Belperpete]

With computer based signalling, flashing aspects essentially comes for free (as in no extra hardware is normally needed) so that's the currently preferred method (as I understand it).

In hardware costs it may come for free, but there are additional design, data prep and testing costs.

== Doublepost prevention - post automatically merged: 26 Feb 2025 ==

I see, is this something that could be improved in a future resignalling? All the signals in this case study are still filament bulb; LED would be brighter and a new signal setup could allow a flashing yellow option if needed.

It could be improved, but it would likely need more than just the signal converting to LED. The signal in rear would likely need additional aspects, and AWS equipment, for example.

 

[Annetts key]

In hardware costs it may come for free, but there are additional design, data prep and testing costs.

Of course, but then nothing in life is totally free. The point is that the costs are rather different compared to providing flashing aspects using relay based interlocking (quite a bit of additional circuitry is needed), flashing aspect control units and everything else needed for a conventional installation.