Signalling for running wrong line Oxford to Radley

[Nicholas43]

GWR is currently running some shuttles Oxford<>Radley. They run 'wrong line' Hinksey to Radley. Why does the driver have to be talked past each signal, and have a pilotman? I thought there was advanced bidirectional signalling on this stretch.

 

[Freightmaster]

GWR is currently running some shuttles Oxford<>Radley. They run 'wrong line' Hinksey to Radley. Why does the driver have to be talked past each signal, and have a pilotman? I thought there was advanced bidirectional signalling on this stretch.

Bidirectional signalling on that particular stretch is only available on the up line,
but for operational reasons, these shuttles have been using the down line in
both directions, which is not bidirectional (the only two 'wrong road' signals
on the down line being in the Culham and Appleford area, i.e. south of the
line block)



MARK

 

[Railsigns]

Bi-directional signalling is usually designed on the premise that a train entering at one end will continue in the same direction at least as far as the next place where a crossover is located. It doesn't usually cater for trains changing direction somewhere in the middle, which would need more controlled signals and more signalled routes in the interlocking.

 

[Nicholas43]

Thanks, both. But it still puzzles me. On the Up line, there's a long berth from signal 2283, on the approach to Appleford, and signal 2331, on the approach to Kennington Junction. How is that different from the (apparent) berth between 2332 and 2310 on the Down line? If the signalling knows there's only one train in a given berth, why can't it cope with that train reversing? (edit): Or is it because these berths are now split by the broken bridge, and may safely have two trains in them?

 

[JN114]

Thanks, both. But it still puzzles me. On the Up line, there's a long berth from signal 2283, on the approach to Appleford, and signal 2331, on the approach to Kennington Junction. How is that different from the (apparent) berth between 2332 and 2310 on the Down line? If the signalling knows there's only one train in a given berth, why can't it cope with that train reversing?

In short/simplistic terms - the signalling (either right or wrong-direction) is triggered/enabled by a train entering the section in that direction. With the train reversing mid-section, the right-direction signalling isn’t enabled hence all the reds and pilotman required.

 

[Nicholas43]

In short/simplistic terms - the signalling (either right or wrong-direction) is triggered/enabled by a train entering the section in that direction. With the train reversing mid-section, the right-direction signalling isn’t enabled hence all the reds and pilotman required.

Sorry, don't understand this. As the train approaches Radley wrong direction, the down signal at the Oxford end of the platform clears to green.

 

[zwk500]

Thanks, both. But it still puzzles me. On the Up line, there's a long berth from signal 2283, on the approach to Appleford, and signal 2331, on the approach to Kennington Junction. How is that different from the (apparent) berth between 2332 and 2310 on the Down line? If the signalling knows there's only one train in a given berth, why can't it cope with that train reversing? Or is it because these berths are now split by the broken bridge, and may safely have two trains in them?

AIUI The berths are set up to trigger on certain Track Circuits or Axle Counter sections being triggered in a certain sequence. The train reversing would not be detected correctly and the UDA would fire (*** or *X** indication).

 

[Nicholas43]

AIUI The berths are set up to trigger on certain Track Circuits or Axle Counter sections being triggered in a certain sequence. The train reversing would not be detected correctly and the UDA would fire (*** or *X** indication).

UDA?

 

[zwk500]

UDA?

Not sure exactly of the acronym, but think its Unexpected Detection Alarm. Essentially the Berth registers that it is occupied, and so requires to display an indication in the berth, but the Train Describer system has not fed it a headcode from an adjacent berth and the signaller has not interposed a headcode manually, so it auto-fills with either **** or *X** depending on the system setup.

In the case of wrong-road working on the Down Oxford, there is no wrong-road signal for Up working between Hinksey and the Bridge, the next signal (OD2310) being between Culham and Appleford so the route cannot safely be set from OD2332 normally. Therefore the signal must be talked past at Danger. If they were using the Up Oxford I would hazard a guess that the reversal would need talking by as the next signal north of Radley (OD2331) would not be visible from the platform, and with the train having already progressed southward the Track Circuit sequence would be wrong and so the berth would not populate.

(I should point out I'm not a signalling engineer or signaller, so may have gotten this horribly wrong).

 

[louis97]

Bidirectional signalling on that particular stretch is only available on the up line,
but for operational reasons, these shuttles have been using the down line in
both directions, which is not bidirectional (the only two 'wrong road' signals
on the down line being in the Culham and Appleford area, i.e. south of the
line block)

Bidirectional signalling is available on both lines.

If you use the up line, you'd be on "normal" (ish) signalling to Radley then you'd need special arrangements for the return as the bi-di signalling won't be setup for a train returning mid-section.

If you use the down line, you can't set the route as the signal you set the route to is beyond the obstruction (the bridge), therefore you need to pass the signal at danger and the pilot. As you're passing the signal at danger you don't set the line up for bi-di working so the normal direction signals will clear in rear of the train. I suspect the train can then return towards Oxford as a ordinary train would, but not 100% sure on that.

 

[MarkyT]

From the Open Train Times map:

A route set in the up direction towards Radley on the down line from signal 2330 or 2332 at Kennington Jn, as shown by the blue arrow, would normally require train detection clear and opposing route locking disengaged all the way to signal 2310 south of Culham station. Because there's some track physically missing over the bridge (roughly around the red cross), and possibly some signalling cabling and other equipment disarranged in that vicinity, these conditions can't be proved adequately and there might actually be some rail vehicles sporadically entering the section to the south of the bridge bringing materials and plant in to the worksite, so these train detection sections, whether axle counters or tracks circuits, may be genuinely occupied. Hence there's absolutely no chance of being able to clear those signals at Kennington Jn.

== Doublepost prevention - post automatically merged: 2 May 2023 ==

I suspect the train can then return towards Oxford as a ordinary train would, but not 100% sure on that.

The down direction signalling on the down line should work normally for the return to Oxford.

 

[Freightmaster]

Bidirectional signalling is available on both lines.

In normal times yes, but as the line is currently severed at the Thames Viaduct,
it isn't currently 'available' as several people, including yourself, have explained

When I said "Bidirectional signalling on that particular stretch is only available on
the up line", the 'particular stretch' I was referring to was Kennington Junction to
Radley station rather than the entire section from through to Didcot North Junction.

Sorry for any confusion!



MARK

 

[Deepgreen]

Yes, it's a pity that the bi-directional running provision is not actually usable properly for this major issue.

 

[zwk500]

Yes, it's a pity that the bi-directional running provision is not actually usable properly for this major issue.

If the Bi-Di had been designed for the possibility of a single bridge failing it would have been decried as 'gold plating'...

 

[Nicholas43]

Thanks, everyone!

 

[Deepgreen]

If the Bi-Di had been designed for the possibility of a single bridge failing it would have been decried as 'gold plating'...

Oh, I agree - it's not a failing of the system, it's just unfortunate.

== Doublepost prevention - post automatically merged: 3 May 2023 ==

From the Open Train Times map:
View attachment 134325
A route set in the up direction towards Radley on the down line from signal 2330 or 2332 at Kennington Jn, as shown by the blue arrow, would normally require train detection clear and opposing route locking disengaged all the way to signal 2310 south of Culham station. Because there's some track physically missing over the bridge (roughly around the red cross), and possibly some signalling cabling and other equipment disarranged in that vicinity, these conditions can't be proved adequately and there might actually be some rail vehicles sporadically entering the section to the south of the bridge bringing materials and plant in to the worksite, so these train detection sections, whether axle counters or tracks circuits, may be genuinely occupied. Hence there's absolutely no chance of being able to clear those signals at Kennington Jn.

== Doublepost prevention - post automatically merged: 2 May 2023 ==


The down direction signalling on the down line should work normally for the return to Oxford.

As a slight aside, what does "Last Sent" mean regarding the Morris Cowley section on the OTT map?

 

[JN114]

As a slight aside, what does "Last Sent" mean regarding the Morris Cowley section on the OTT map?

Last train description sent towards that bit of line

 

[MarkyT]

As a slight aside, what does "Last Sent" mean regarding the Morris Cowley section on the OTT map?

That's the ID of the last train to be sent to Cowley and remains after it disappears from the track-circuited area. A reminder for the signaller in the Train Describer system.

 

[Deepgreen]

That's the ID of the last train to be sent to Cowley and remains after it disappears from the track-circuited area. A reminder for the signaller in the Train Describer system.

Right, but OTT apparently doesn't show the actual train ID? Thanks.

 

[zwk500]

Right, but OTT apparently doesn't show the actual train ID? Thanks.

OTT will show the Train ID it's given by the data feed. For some, but not all, freight this is scrambled (known as 'obfuscated'), supposedly to protect commercial sensitivities but the industry has recognised that obfuscation doesn't really achieve anything now and so Train IDs are slowly getting unscrambled.

 

[oxfordrelayer]

Bidirectional signalling is available on both lines.

If you use the up line, you'd be on "normal" (ish) signalling to Radley then you'd need special arrangements for the return as the bi-di signalling won't be setup for a train returning mid-section.

If you use the down line, you can't set the route as the signal you set the route to is beyond the obstruction (the bridge), therefore you need to pass the signal at danger and the pilot. As you're passing the signal at danger you don't set the line up for bi-di working so the normal direction signals will clear in rear of the train. I suspect the train can then return towards Oxford as a ordinary train would, but not 100% sure on that.

== Doublepost prevention - post automatically merged: 4 May 2023 ==

blue circles on black lines are the axle counters

 

[MarkyT]

Something relevant here is that under track circuit block (TCB) regulations there must be a signal visible for restarting a train in the opposite direction at the point of reversal. That has been the case in new schemes for decades, although there may be a few very early colour light schemes still around which don't comply. Sometimes, on modern schemes, the method of working of the end spur of a single line branch is classified as one train without trainstaff (OT) rather than TCB. The original purpose of OT was to avoid physical tokens or multiple track circuit sections all the way to the buffer stops, but modern train detection coverage can be much simpler using axle counters, with sensors only at the other end of the single line from the terminus. Using OT in these cases is thus purely a cludge today to avoid needing a platform starting signal at the terminus, which can be very expensive if there's no need for any other signalling equipment, power or cabling around the terminus. Local instructions for traincrew reversing at these stub end branches on older non-compliant TCB or OT usually requires telephoning the signaller for permission before departure from the terminus however. This is often because of intermediate level crossings in the section of varying types that the signaller is responsible for, which may have public telephones or whose warning equipment might have failed.

 

[Deepgreen]

Something relevant here is that under track circuit block (TCB) regulations there must be a signal visible for restarting a train in the opposite direction at the point of reversal. That has been the case in new schemes for decades, although there may be a few very early colour light schemes still around which don't comply. Sometimes, on modern schemes, the method of working of the end spur of a single line branch is classified as one train without trainstaff (OT) rather than TCB. The original purpose of OT was to avoid physical tokens or multiple track circuit sections all the way to the buffer stops, but modern train detection coverage can be much simpler using axle counters, with sensors only at the other end of the single line from the terminus. Using OT in these cases is thus purely a cludge today to avoid needing a platform starting signal at the terminus, which can be very expensive if there's no need for any other signalling equipment, power or cabling around the terminus. Local instructions for traincrew reversing at these stub end branches on older non-compliant TCB or OT usually requires telephoning the signaller for permission before departure from the terminus however. This is often because of intermediate level crossings in the section of varying types that the signaller is responsible for, which may have public telephones or whose warning equipment might have failed.

Good detailed information, as ever - thanks.

 

[Nicholas43]

Has a human had to instruct whatever controls signal 2332 something to the effect: "You can clear as usual if the route is set to Morris Cowley, but refuse to clear if the route is set to Didcot, even if all the automated detection is telling you there's nothing in the berth from you to 2310 ? You don't need to know why, but just for your information, the reason is that if you clear, the berth will get occupied by a train which mustn't go past Radley, and a human will have to do Weird Things before the right direction signals from Radley back to you can clear, to let the train come back to Oxford" ?
Or can 2332 work that out for itself, if a human just tells it "There's a total blockade at a point between you and 2310" ?

 

[MarkyT]

Has a human had to instruct whatever controls signal 2332 something to the effect: "You can clear as usual if the route is set to Morris Cowley, but refuse to clear if the route is set to Didcot, even if all the automated detection is telling you there's nothing in the berth from you to 2310 ? You don't need to know why, but just for your information, the reason is that if you clear, the berth will get occupied by a train which mustn't go past Radley, and a human will have to do Weird Things before the right direction signals from Radley back to you can clear, to let the train come back to Oxford" ?
Or can 2332 work that out for itself, if a human just tells it "There's a total blockade at a point between you and 2310" ?

I'd say the particular up routes [#1] on the down main from 2330 and 2332 towards Didcot will have been deliberately disabled in the interlocking. The up routes to Cowley and possibly to the up main may remain enabled. In SSI these restrictions could be imposed using the 'technician's terminal' functionality and I expect the more modern processor-based interlockings for the Oxford area housed at Thames Valley Signalling Centre will have exactly the same facility. Only authorised maintenance personnel will have access to and permissions for the tech terminal so signallers cannot override those commands once applied. The equivalent controls in older relay interlockings were achieved by 'route disconnection links (RDL) [#2]' in the equipment rooms, again only accessible to technical staff, not signallers.

#1 - A route in signalling is a unique path between an entrance signal and an exit at a destination signal or buffer stop.
#2 - A route disconnection link is a metallic conductor between two adjacent terminals that can be moved securely to interrupt the circuit for a specific relay associated with the particular route concerned.

 

[louis97]

Has a human had to instruct whatever controls signal 2332 something to the effect: "You can clear as usual if the route is set to Morris Cowley, but refuse to clear if the route is set to Didcot, even if all the automated detection is telling you there's nothing in the berth from you to 2310 ? You don't need to know why, but just for your information, the reason is that if you clear, the berth will get occupied by a train which mustn't go past Radley, and a human will have to do Weird Things before the right direction signals from Radley back to you can clear, to let the train come back to Oxford" ?
Or can 2332 work that out for itself, if a human just tells it "There's a total blockade at a point between you and 2310" ?

I don't think either quite explain what is going on. The signalling system will be setup for a possession of the line between Radley and Culham, therefore it isn't possible for the signaller to setup the route (Each signal has a number of routes from it, sometimes more than one to the same signal or buffer stop) towards 2310. There is no requirement for the signaller to tell the signal anything, it won't clear unless a route is set and the route towards 2310 won't be possible to set because of what I previously described. The route towards Morris Cowley is irrelevant here as it is another route entirely from 2332 and is unaffected by what is going on between Radley and Culham.

 

[MarkyT]

I don't think either quite explain what is going on. The signalling system will be setup for a possession of the line between Radley and Culham, therefore it isn't possible for the signaller to setup the route (Each signal has a number of routes from it, sometimes more than one to the same signal or buffer stop) towards 2310. There is no requirement for the signaller to tell the signal anything, it won't clear unless a route is set and the route towards 2310 won't be possible to set because of what I previously described. The route towards Morris Cowley is irrelevant here as it is another route entirely from 2332 and is unaffected by what is going on between Radley and Culham.

No doubt the signaller will have various reminders/collars set up on the workstation screens in accordance with rules and instructions to prevent any inadvertent attempt to set a route into the possession area. If I was a signaller there though, I'd still like to be reassured the particular routes have been disconnected by the techs. for the duration of the works though, partly because even though there is some physical track actually missing on both roads, this may not actually result in the actual train detection sections showing occupied, because axle counters are used through the area, which do not detect the integrity of the rails.