• Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!

Colour Light Signals - Light Order

Status
Not open for further replies.

Tomnick

Established Member
Joined
10 Jun 2005
Messages
5,826
Design of the Paddington scheme seems to have been done by the interns. Quite why the default lie of the points beyond a red SN109 for Down trains being stopped there was head-on straight into the Up Fast, instead of flank protection, as used for the previous 150 years, into the Down Relief, was never satisfactorily explained at the enquiry.
The distance to the point of conflict with the Up Main was much greater than the distance to the Down Relief would've been though, so a collision would be potentially more severe but also much less likely to occur in the first place. On that basis it seems to have been a reasonably sound decision. I guess that it would've made the interlocking arrangements much more complicated too, as the points would be too far apart to work as a co-acting pair without being an operational pain in the posterior.

Although it sounds pedantic it is a major distinction in signalling design. As a "semaphore" for example a starter signal it can show a green aspect if the advanced starter was red. A Day Colour Light Signal would have to be three aspect and show yellow.
The Rule Book (S7 1.6) still covers the possibility of a colour light signal that can't display a yellow aspect reading onto another stop signal - I can't think of a single current example, but presumably it was acceptance at some point in history.

If you have stopped or nearly stopped at either of the following types of signal at danger and that signal changes to a proceed aspect or indication, you must be prepared to stop at the next stop signal worked by the same signalbox.

• A colour light signal that cannot display a yellow aspect.
• A semaphore signal.

This does not apply to the signal controlling the entrance to an intermediate block section.
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

Surreytraveller

On Moderation
Joined
21 Oct 2009
Messages
2,810
Presumably, for road traffic signals, red at the top makes the signal more visible over a line of traffic.
And in addition, road vehicles are able to stop within a distance they can see to be clear, so if the aspect of a traffic signal is not obvious, then a road vehicle will be able to stop
 

eastwestdivide

Established Member
Joined
17 Aug 2009
Messages
2,532
Location
S Yorks, usually
Back on the ground-mounted signals, there was one at Worksop, red on top. Not sure if it's still there.
Seen here on a railtour in 1984 I think:
Worksop.jpg
 

Taunton

Established Member
Joined
1 Aug 2013
Messages
10,018
The distance to the point of conflict with the Up Main was much greater than the distance to the Down Relief would've been though, so a collision would be potentially more severe but also much less likely to occur in the first place.
But that's just the way it had been laid out. The whole layout, tracks and signalling, was new. They could have designed what they wished.
 

MichaelAMW

Member
Joined
18 Jun 2010
Messages
1,011
I believe that is theoretically a semaphore signal with its arm removed leaving just the spectacle lenses.

They were used in places where it was always gloomy and dark meaning sighting an arm is difficult to do and there will never be enough daylight to stop the oil lamp being seen through the lens.

Although it sounds pedantic it is a major distinction in signalling design. As a "semaphore" for example a starter signal it can show a green aspect if the advanced starter was red. A Day Colour Light Signal would have to be three aspect and show yellow.

The "day" is not generally used but "Day Colour Light Signals" was originally the correct term. This is because Semaphore signals are also colour light signals, but because the oil light is too weak to be seen during the day. The signal has a semaphore arm attached for use in daytime.
The Southern Region Central Division platforms 9 & 10 at Wimbledon had no-arm semaphores at the south end of the platforms. They were removed in the early 1980s when Wimbledon 'B' box closed and control taken over by the then new Victoria panel, which is at Clapham Jct.

I would be inclined to question the OP's initial question as colour-light signals were in use marginally earlier than road traffic lights, although not by a lot as far as a quick search can tell. It would have been better to ask why road didn't follow rail - although I appreciate that point has since been raised in this thread!
 

jopsuk

Veteran Member
Joined
13 May 2008
Messages
12,771
there are of course modern signal heads with:
3 aspect, all three colours from the same "lamp"
4 aspect, two "lamps" with a gap in between, where the top can show red or yellow, the bottom green or yellow.
 

John Webb

Established Member
Joined
5 Jun 2010
Messages
3,047
Location
St Albans
there are of course modern signal heads with:
3 aspect, all three colours from the same "lamp"
4 aspect, two "lamps" with a gap in between, where the top can show red or yellow, the bottom green or yellow.
My experience with modern signals is that for 4-aspect signals the lower aspect shows red, yellow or green and the upper aspect yellow only.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
My experience with modern signals is that for 4-aspect signals the lower aspect shows red, yellow or green and the upper aspect yellow only.
The Ansaldo ones in the Cheadle Hulme area work differently - green at the top I think.
 

Choo choo 66

Member
Joined
26 May 2020
Messages
16
Location
Kernow
there are of course modern signal heads with:
3 aspect, all three colours from the same "lamp"
4 aspect, two "lamps" with a gap in between, where the top can show red or yellow, the bottom green or yellow.

I’ve never seen 4 aspect like that. I’ve only ever seen them where the bottom lens will show red, yellow or green and the top only show yellow where the signal is showing a double yellow
 

Taunton

Established Member
Joined
1 Aug 2013
Messages
10,018
I would be inclined to question the OP's initial question as colour-light signals were in use marginally earlier than road traffic lights, although not by a lot as far as a quick search can tell.
I think the first actual colour light scheme (as opposed to mechanical semaphore lenses over electric lamps as on the early London Underground etc) was the Westinghouse project on the Mersey Railway, installed with electrification in 1903, power points (compressed air), etc. Westinghouse did the lot, overseen by George Westinghouse from the USA personally.
 

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
But that's just the way it had been laid out. The whole layout, tracks and signalling, was new. They could have designed what they wished.

But going from 6 tracks to 4 tracks is inevitably going to result in some lines converging with others, regardless of the exact details of the layout.
 
Joined
31 Aug 2019
Messages
341
Location
IW
The similarities between road traffic signals and railway signaling is purely in your own mind.
They both use lights and in some implementations look similar to each other physically. Thats about it.
 

Taunton

Established Member
Joined
1 Aug 2013
Messages
10,018
But going from 6 tracks to 4 tracks is inevitably going to result in some lines converging with others, regardless of the exact details of the layout.
And that's exactly why, for generations of previous signal engineers, where you can (and you can here) the points would lie into the Down Relief (next track to the right) rather than head-on into the Up Main. Always called "Flank Protection". I see that Network Rail have even put up a page (here) intimating they might have heard of it (it's down as "jargon buster" so they obviously regard such an expression as a bit of a joke).

 

swt_passenger

Veteran Member
Joined
7 Apr 2010
Messages
31,263
My experience with modern signals is that for 4-aspect signals the lower aspect shows red, yellow or green and the upper aspect yellow only.
That’s definitely my understanding based on a few previous discussions in the forum, and observations of signals...
 

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
And that's exactly why, for generations of previous signal engineers, where you can (and you can here) the points would lie into the Down Relief (next track to the right) rather than head-on into the Up Main. Always called "Flank Protection". I see that Network Rail have even put up a page (here) intimating they might have heard of it (it's down as "jargon buster" so they obviously regard such an expression as a bit of a joke).


I get that. But, as stated, it (possibly) was judged in the design process that the longer run before merging onto the Up Main gave a driver a longer period of time to realise their error and stop in time (compared to a "side swipe' onto the Down Relief).

In addition, the driver would've ended up on bit of track that they should not be on - any train crossing to the Down Main does so much earlier before the track "runs out",which would've been another clue to the driver something was amiss.

A case of lower likelihood/greater severity playing off against higher likelihood/lower severity.

What also did not help was the relative inexperience of the Thames Trains driver, exacerbated by inadequate driver training, who tragically lost his life as a result.

In any case, when the route was reinstated a few years ago, the flank protection now directs trains to the Down Relief.
 
Last edited:

Taunton

Established Member
Joined
1 Aug 2013
Messages
10,018
Ladbroke Grove was not the only serious accident where flank protection had been thrown to the winds. Once single lead junctions became fashionable from about 1970 onwards there were a whole slew of head-on collisions at them (I can count half a dozen just sitting here), invariably fatal to the driver(s) and sometimes others, yet the design engineers just didn't understand.

Southall was another one. The single-leads crossover set that the Down freight was taking had replaced a traditional double crossover, with parallel Down relief-main and Up main-relief. With that, the points on the Up main would have always been set over to the Up relief for flank protection while a Down crossing move was taking place, and the HST, at the speed it had reduced to by the collision point, would have been rather lurchingly diverted across to the relief, most likely without even derailing. I bet the design engineers never even thought of this.
 

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
Ladbroke Grove was not the only serious accident where flank protection had been thrown to the winds. Once single lead junctions became fashionable from about 1970 onwards there were a whole slew of head-on collisions at them (I can count half a dozen just sitting here), invariably fatal to the driver(s) and sometimes others, yet the design engineers just didn't understand.

Southall was another one. The single-leads crossover set that the Down freight was taking had replaced a traditional double crossover, with parallel Down relief-main and Up main-relief. With that, the points on the Up main would have always been set over to the Up relief for flank protection while a Down crossing move was taking place, and the HST, at the speed it had reduced to by the collision point, would have been rather lurchingly diverted across to the relief, most likely without even derailing. I bet the design engineers never even thought of this.

At Southall, the HST had only decelerated to around 80mph anyway, so would likely still have derailed or overturned in any case (I think the HST driver commenced emergency braking when the red signal came into view). Southall East crossovers are only 40mph IIRC.

I don't know how fast the old double ladder was, but if the replacement single lead was faster, trains crossing would spend less time conflicting with the main line, itself reducing collision risk.


The ultimate lesson in allow this is to not have trains pass red signals in the first place - which has significantly reduced with a greater focus on signal sighting, human factors, and national TPWS fitment.

I think the closest we've come in recent years to a serious collision as a result of a SPAD was the Wootton Basset incident.
 

merry

Member
Joined
19 Oct 2011
Messages
76
A couple of points:
- Quorn GCR Up platform down starter: In a conversation with the head of S&T, I was told it was a conscious design decision to fit a mechanical colour light (it is indeed a spectacle plate behind a single lens, electrically lit, operated by the usual lever and wire semaphore method - balance weight, crank, etc are under the platform). There was limited room for a semaphore alternative (well, perhaps e.g. a somersault arm would have been OK) due to passenger heads being at risk! The signal is only used when up trains that have terminated from Loughborough need to reverse, mostly Thomas shuttles and other special events. The trailing crossover north of the bridge had an FPL added to allow this as a passenger move. It was added some time after double track was reinstated. And yes, I'm a former S&T volunteer there! Sadly no time these days...maybe again in future. One can hope.

-single lead junctions / single ladders: In the 1980s or earlier, it was decided at a high level that to reduce maintenance costs, junction design should minimise (a) the number of times a switch is moved, and (b) the number of crossings, even if the number of switch ends increases a little. Apparently the maintenance cost of a switch end is lower than a crossing. A single "point" has one switch and one crossing, a diamond crossing contains 4 crossings (including 2 more costly obtuse crossings). A classic double junction has two switch ends and 6 crossings, 2 of which are obtuse. A single lead has 4 switches and 4 crossings. Apparently this is cheaper in maintenance and capital.
- Flank protection: To save wear and tear, and to reduce the risk of a failure (which is a risk every time a switch end moves), it was decided that unless there is a very good reason, a switch should lie in the last position set until a new route is called and set, not returning to 'normal' after a train passes. This improves efficiency in many locations and reduces the number of operations significantly. SPAD induced collisions weere deemed low risk given AWS, longer overlaps, and modern brake capabilities. The benefit was deemed to outweigh the risk, and given the pressures on BR costs at that time, who can blame BR's people? Since a few incidents culminating in Paddington & Southall, the practice of providing flank protesction has become more usual again at risk locations. 'Tis a safer practice and generally life has a higher value these days. Social change and all that...H&S has its place for sure.

Hope that explains the background, at least as I've had it reported.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
The only place I can think of with this is Birmingham New St where there also seems to be an approach control on the signals.
Birmingham New Street station, as rebuilt in the 1960s, was equipped originally with searchlight signal mechanisms in custom-designed enclosures. These were later all converted to conventional three-lens colour lights, many mounted horizontally due to space and sighting constraints under the canopies.

-single lead junctions / single ladders: In the 1980s or earlier, it was decided at a high level that to reduce maintenance costs, junction design should minimise (a) the number of times a switch is moved, and (b) the number of crossings, even if the number of switch ends increases a little. Apparently the maintenance cost of a switch end is lower than a crossing. A single "point" has one switch and one crossing, a diamond crossing contains 4 crossings (including 2 more costly obtuse crossings). A classic double junction has two switch ends and 6 crossings, 2 of which are obtuse. A single lead has 4 switches and 4 crossings. Apparently this is cheaper in maintenance and capital.
- Flank protection: To save wear and tear, and to reduce the risk of a failure (which is a risk every time a switch end moves), it was decided that unless there is a very good reason, a switch should lie in the last position set until a new route is called and set, not returning to 'normal' after a train passes. This improves efficiency in many locations and reduces the number of operations significantly. SPAD induced collisions weere deemed low risk given AWS, longer overlaps, and modern brake capabilities. The benefit was deemed to outweigh the risk, and given the pressures on BR costs at that time, who can blame BR's people? Since a few incidents culminating in Paddington & Southall, the practice of providing flank protesction has become more usual again at risk locations. 'Tis a safer practice and generally life has a higher value these days. Social change and all that...H&S has its place for sure.

Hope that explains the background, at least as I've had it reported.
Diamonds in double junctions and elsewhere definititely fell out of favour. Partly because when increasing the junction speeds at relaying, they often had to become switched diamonds, a type that is hated with a passion both by the PW and S&T engineering departments, as well as operators, as they are very difficult to keep in adjustment in changing temperatures and frequently suffer from detection failure, stopping trains. Hence long ladder-type junctions were preferred but again desired increases in junction speed meant a full double ladder junction often might not fit in the distance available. Single ladders and leads were often inevitable in modern layouts installed in the late 70s and 80s and there was a general desire that still exists to minimise the overall number of point ends where possible, as every one is a capital cost, a failure source, and a maintenance liability. Signal engineers of the time rarely had much say over the track layout configuration and usually had to work with what was provided. Flank point movement protection has always been applied where possible, but clearly in many single lead junction arrangements the flank points aren't there to move. It is not always a simple process to decide which points to move for every move and I will not try to defend the particular decisions made in the Ladbroke Grove junction design, but since then the science of junction overrun risk assessment has improved immeasurably and the outputs would drive such decisions and indeed the layout of the junction itself to a greater extent. The decision to install TPWS network-wide on a targetted risk assessed basis was also an outcome of the unfortunate series of incidents, and as others have mentioned has since eliminated junction collisions as a major cause of injuries and fatalities on the operating railway.
 
Last edited:

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
- Flank protection: To save wear and tear, and to reduce the risk of a failure (which is a risk every time a switch end moves), it was decided that unless there is a very good reason, a switch should lie in the last position set until a new route is called and set, not returning to 'normal' after a train passes. This improves efficiency in many locations and reduces the number of operations significantly. SPAD induced collisions weere deemed low risk given AWS, longer overlaps, and modern brake capabilities. The benefit was deemed to outweigh the risk, and given the pressures on BR costs at that time, who can blame BR's people? Since a few incidents culminating in Paddington & Southall, the practice of providing flank protesction has become more usual again at risk locations. 'Tis a safer practice and generally life has a higher value these days. Social change and all that...H&S has its place for sure.
I think you're slightly mixing up two issues here. Ever since route relay interlockings came in, points were left as they lay until specifically required to be changed (with the exception of trap points and a few others that had to be normal for safety reasons). However flank protection was perpetuated where the layout allowed, so at a double junction with a diamond the sets of points would not be left in positions that would allow conflicting moves over the diamond, so a route over the diamond would always call both sets of points accordingly.
Diamonds in double junctions and elsewhere definititely fell out of favour. Partly because when increasing the junction speeds at relaying, they often had to become switched diamonds, a type that is hated with a passion both by the PW and S&T engineering departments, as well as operators, as they are very difficult to keep in adjustment in changing temperatures and frequently suffer from detection failure, stopping trains. Hence long ladder-type junctions were preferred but again desired increases in junction speed meant a full double ladder junction often might not fit in the distance available. Single ladders and leads were often inevitable in modern layouts installed in the late 70s and 80s and there was a general desire that still exists to minimise the overall number of point ends where possible, as every one is a capital cost, a failure source, and a maintenance liability. Signal engineers of the time rarely had much say over the track layout configuration and usually had to work with what was provided. Flank point movement protection has always been applied where possible, but clearly in many single lead junction arrangements the flank points aren't there to move. It is not always a simple process to decide which points to move for every move and I will not try to defend the particular decisions made in the Ladbroke Grove junction design, but since then the science of junction overrun risk assessment has improved immeasurably and the outputs would drive such decisions and indeed the layout of the junction itself to a greater extent. The decision to install TPWS network-wide on a targetted risk assessed basis was also an outcome of the unfortunate series of incidents, and as others have mentioned has since eliminated junction collisions as a major cause of injuries and fatalities on the operating railway.
A single lead junction is a poor choice for capacity as well as safety reasons. Maximum capacity on a double lead is obtained by timetabling parallel moves in opposite directions to pass each other on the junction, as these moves do not conflict with each other and the junction is free for other moves for the maximum possible time. This isn't possible for the diverging route on a single lead. As the railway has got busier many single leads have been converted to double, usually as parallel ladders rather than diamonds, to achieve a capacity as well as a safety benefit. Wigston and Euxton are two that spring to mind.

Both parallel ladder double junctions and single leads have four switches and no diamonds, and I would guess that the former arrangement involves less movement of switches. The only operational advantage I can think of for the single lead is that it allows single line working on the diverging route. It seems to me there was an element of groupthink in the 70s and 80s regarding singling of double junctions, without thinking about whether there would be space available for a parallel ladder.
 

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
A single lead junction is a poor choice for capacity as well as safety reasons. Maximum capacity on a double lead is obtained by timetabling parallel moves in opposite directions to pass each other on the junction, as these moves do not conflict with each other and the junction is free for other moves for the maximum possible time. This isn't possible for the diverging route on a single lead. As the railway has got busier many single leads have been converted to double, usually as parallel ladders rather than diamonds, to achieve a capacity as well as a safety benefit. Wigston and Euxton are two that spring to mind.

Yes and no.

In some instances, single lead junctions meant that a faster junction could be provided without requiring any additional land.

An example is Ely North Jn, which is currently a 50mph single lead. Previously it was double junctions...but 20mph.

The extra 30mph makes quite a radical difference to the time a train physically occupies the junction (particularly a long freight train), but also the time taken to clear the section through to Ely Station, which in turn reduces the possible headway to a following train.
 

Bald Rick

Veteran Member
Joined
28 Sep 2010
Messages
29,070
Yes and no.

In some instances, single lead junctions meant that a faster junction could be provided without requiring any additional land.

An example is Ely North Jn, which is currently a 50mph single lead. Previously it was double junctions...but 20mph.

The extra 30mph makes quite a radical difference to the time a train physically occupies the junction (particularly a long freight train), but also the time taken to clear the section through to Ely Station, which in turn reduces the possible headway to a following train.

Indeed so. My stock answer to anyone who says ‘Ely North should be restored to its former layout’ is to ask - ok, which trains ar you going to take out of the timetable, as it has lower capacity?
 

Class 170101

Established Member
Joined
1 Mar 2014
Messages
7,908
Indeed so. My stock answer to anyone who says ‘Ely North should be restored to its former layout’ is to ask - ok, which trains ar you going to take out of the timetable, as it has lower capacity?

But why was it 20mph maximum speed?

The other thing to say about Ely North Jn, the layout as it stands to today, I doubt would ever be authorised today as a result of collisions elsewhere. I understood the junction was 'double' Blocked as a result (along with Ely West Curve being Uni-directional for many years) and that has / had an impact on capacity at this Junction. (is it still double blocked?)

It would be interesting to see therefore if the capacity is compromised that much if parallel working was maximised albeit with the lower top speed aginst today's layout. Secondly how it would be with parallel working from Ely to Kings Lynn and Peterborough only with the single lead remaining towards Norwich.

Ultimately of course the junction isn't the only thing preventing the increase in train services but the level crossings here as well.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
But why was it 20mph maximum speed?
Space available on the solum, not easy to expand due to surrounding marshy ground I believe.
The other thing to say about Ely North Jn, the layout as it stands to today, I doubt would ever be authorised today as a result of collisions elsewhere. I understood the junction was 'double' Blocked as a result (along with Ely West Curve being Uni-directional for many years) and that has / had an impact on capacity at this Junction. (is it still double blocked?)
Double blocking controls removed and disconnected bi-di routes reinstated since junction analysis revisited and TPWS fitted.
 

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
Space available on the solum, not easy to expand due to surrounding marshy ground I believe.

Double blocking controls removed and disconnected bi-di routes reinstated since junction analysis revisited and TPWS fitted.

The double blocking is very much still there (and related to the level crossings).

The other thing to remember is that the majority of moves on the junction are to/from Peterborough (the fully doubled route), even with doubling both King's Lynn and Norwich lines, you still end up with quite a few "conventional" conflicts left.
 

MotCO

Established Member
Joined
25 Aug 2014
Messages
4,085
Back on the ground-mounted signals, there was one at Worksop, red on top. Not sure if it's still there.
Seen here on a railtour in 1984 I think:
View attachment 90012

Is the reason for a top red on the ground signal also snow-related? If there was deep snow, and red was at the bottom, it would be obscured, and no light would show; the snow would have to be very deep to obscure the top red aspect.

I don't have any kind of visual impairment of which I'm aware, nor any kind of related mental impairment, but I find it hard to understand which button is which for door open/close, the "<|>" and ">|<" symbols just don't work for me well at all. As long as they're accompanied with words like "open" and "close" they're much easier to understand. Otherwise I have to think about them for a couple of seconds, or I press one of them at random, and then the other one if this doesn't work.
I guess if I used the door controls often enough I'd get used to them.
I suspect that many people encountering them for the first time find them hard to understand.
I also have problems with door buttons in lifts - I have to think twice which is which, which can be awkward if you are trying to re-open the doors quickly to let someone else enter without squashing them :(
 

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
I also have problems with door buttons in lifts - I have to think twice which is which, which can be awkward if you are trying to re-open the doors quickly to let someone else enter without squashing them :(

I've done that before - inadvertent (but well-intentioned) hammering of the "close" button as somebody else approaches.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
The double blocking is very much still there (and related to the level crossings).
Being AHBs, I would certainly expect aspect clearance delay for signals located within the crossing strike-ins, to maintain minimum road warning time under certain approach conditions. Wouldn't class that as double blocking though.
The other thing to remember is that the majority of moves on the junction are to/from Peterborough (the fully doubled route), even with doubling both King's Lynn and Norwich lines, you still end up with quite a few "conventional" conflicts left.
The double to Peterborough makes sense, as that's where the majority of the long, slow accelerating freights run that are the least desirable to stop and restart. The other routes via the single leads are mostly frequented by much shorter and faster-accelerating emus and dmus which clear the junction more quickly and whose timings can probably more readily tolerate an additional stop and restart or speed check on approach.
 

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
Being AHBs, I would certainly expect aspect clearance delay for signals located within the crossing strike-ins, to maintain minimum road warning time under certain approach conditions. Wouldn't class that as double blocking though.

No, it is double blocking - in all cases a freight stood at the signals protecting North Jn would hang back across the level crossings on approach, which the double blocking prevents.

There is also aspect clearance delay on the prior signals (that sit before the level crossings and the first red reached), which kicks in once the route across North Jn becomes free.

All observable on the OpenTrainTimes map: https://www.opentraintimes.com/maps/signalling/sbr
 
Status
Not open for further replies.

Top