Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 1, 2018 19:13:32 GMT
We have to remember that TBTC is totally untested on a system as complex as the SSR. Aldgate, Baker Street, Paddington, Earls Court all pose major challenges. This is the thing with the new signalling, until its in place, no one really knows how well it is actually going to work. I think it will work well until there are trains that are delayed. What happens if two trains appear at the junction at the same time, one being late and one on time, which does the computer prioritise? Does it let the train on time go causing more delays to the delayed train, or does it delay the train that is on time causing two trains delayed, but by not as much as the delayed train being delayed? I would not want to say that quickly xD |
|
rincew1nd
Administrator
Junior Under-wizzard of quiz
Posts: 10,286 
|
Post by rincew1nd on Apr 1, 2018 19:50:52 GMT
There are still human overseers, I presume they can guide the decision making at junctions. They already make exactly this decision countless times a day.
|
|
North End
Beneath Newington Causeway
Posts: 1,769
|
Post by North End on Apr 1, 2018 20:43:56 GMT
There are still human overseers, I presume they can guide the decision making at junctions. They already make exactly this decision countless times a day. In reality it isn't always that easy to influence what the system decides to do. Even getting a train to use a particular platform at a terminus involves a fair few button clicks - hence why we often see nonsense like three empty platforms at Morden in the peak then an arriving train comes into the middle as per timetable, when best practice is of course to use the right-hand most available platform in that situation. Meanwhile, TBTC just doesn't push trains through junctions as efficiently as old signalling. To be fair, some of this may well be down to modern standards having to be applied at the design stage. But it's not all down to that. Look at the day when Golders Green platform 5 was out of commission all day, and by afternoon the service was running an hour late just due to everything running through the middle. Or during the recent snow day when the service was running late and we have blocking back from Belsize Park and Archway down to Camden Town simply because the system wasn't pushing trains through the junctions quickly. |
|
|
|
Post by superteacher on Apr 1, 2018 21:04:42 GMT
There are still human overseers, I presume they can guide the decision making at junctions. They already make exactly this decision countless times a day. In reality it isn't always that easy to influence what the system decides to do. Even getting a train to use a particular platform at a terminus involves a fair few button clicks - hence why we often see nonsense like three empty platforms at Morden in the peak then an arriving train comes into the middle as per timetable, when best practice is of course to use the right-hand most available platform in that situation. Meanwhile, TBTC just doesn't push trains through junctions as efficiently as old signalling. To be fair, some of this may well be down to modern standards having to be applied at the design stage. But it's not all down to that. Look at the day when Golders Green platform 5 was out of commission all day, and by afternoon the service was running an hour late just due to everything running through the middle. Or during the recent snow day when the service was running late and we have blocking back from Belsize Park and Archway down to Camden Town simply because the system wasn't pushing trains through the junctions quickly. Not to mention the time when the service was suspended south of Tooting and the system wouldn’t let trains into the SB platform until the previous train had cleared the siding, causing horrific delays. I’m assuming this and the Golders Green issue would have been subsequently fixed to prevent a repeat. |
|
|
|
Post by philthetube on Apr 1, 2018 22:39:05 GMT
In reality it isn't always that easy to influence what the system decides to do. Even getting a train to use a particular platform at a terminus involves a fair few button clicks - hence why we often see nonsense like three empty platforms at Morden in the peak then an arriving train comes into the middle as per timetable, when best practice is of course to use the right-hand most available platform in that situation. Meanwhile, TBTC just doesn't push trains through junctions as efficiently as old signalling. To be fair, some of this may well be down to modern standards having to be applied at the design stage. But it's not all down to that. Look at the day when Golders Green platform 5 was out of commission all day, and by afternoon the service was running an hour late just due to everything running through the middle. Or during the recent snow day when the service was running late and we have blocking back from Belsize Park and Archway down to Camden Town simply because the system wasn't pushing trains through the junctions quickly. Not to mention the time when the service was suspended south of Tooting and the system wouldn’t let trains into the SB platform until the previous train had cleared the siding, causing horrific delays. I’m assuming this and the Golders Green issue would have been subsequently fixed to prevent a repeat.That is a design problem and not a system issue, there is similar at Northwood, using conventional signalling, at other locations there are draw up signals to reduce speeds where conflicting moves and short overlaps could cause issues, eg Moorgate and I am sure at other sites as well. I am sure it would have been possible to design into the system that trains ran at reduced speed into Tooting when trains were leaving the siding, This is to do with mitigating the consequences of any spad.
These isues also exist with conventional signalling. At Northwood it is impossible to let a train into the southbound platform when a train is signalled out of the siding, or if it is leaving the siding it is fully berthed in the N/B platform, it is also impossible to bring a train out of the siding when there is a train in the S/B platform.
It would have been possible , with the inclusion of draw up signals when the system was installed, but a decision was taken, presumably on cost grounds not to do this.
The situation will have been the same at Tooting.
In many cases the consequences of signalling installation decisions are not realised until it is required to be used in a non standard way.
|
|
North End
Beneath Newington Causeway
Posts: 1,769
|
Post by North End on Apr 1, 2018 23:40:18 GMT
Not to mention the time when the service was suspended south of Tooting and the system wouldn’t let trains into the SB platform until the previous train had cleared the siding, causing horrific delays. I’m assuming this and the Golders Green issue would have been subsequently fixed to prevent a repeat.That is a design problem and not a system issue, there is similar at Northwood, using conventional signalling, at other locations there are draw up signals to reduce speeds where conflicting moves and short overlaps could cause issues, eg Moorgate and I am sure at other sites as well. I am sure it would have been possible to design into the system that trains ran at reduced speed into Tooting when trains were leaving the siding, This is to do with mitigating the consequences of any spad.
These isues also exist with conventional signalling. At Northwood it is impossible to let a train into the southbound platform when a train is signalled out of the siding, or if it is leaving the siding it is fully berthed in the N/B platform, it is also impossible to bring a train out of the siding when there is a train in the S/B platform.
It would have been possible , with the inclusion of draw up signals when the system was installed, but a decision was taken, presumably on cost grounds not to do this.
The situation will have been the same at Tooting.
In many cases the consequences of signalling installation decisions are not realised until it is required to be used in a non standard way.
Here’s a couple of examples: 1) Stockwell southbound - which needless to say is one of the busiest platforms in the evening peak, enough to be a proverbial bottleneck. The glossy Thales publicity shows trains smoothly following each other. In reality, here you sit still in the tunnel for ages and ages by which time the preceding train is halfway to Clapham North, then all of a sudden it jumps up to 25 mph and you get limit of movement authority all the way into the platform. Great - the equivalent of one home signal just at the place you want a quick platform reoccupation. The same happens at Camden platform 2 - so no wonder when the job is up the wall we see chronic blocking back here. 2) Oval northbound. Every train comes in on a low brake rate. I’ve heard it said this maybe to do with the platform having an unusual change of gradient, but I’ve also seen documentation which says it’s to do with a LU requirement to have a particular number of train berths between Oval and Kennington. That’s fine when there’s trains in the section ahead. But the low brake rate is imposed even when the section ahead is empty. Even an old-fashioned draw-up signal would in most cases only impose speed control if actually needed, and this system is supposed to be “state of the art”. There are many, many such issues across the Northern. Another one is the totally pointless stop mid-way between Golders Green and Hampstead which some trains get. First Thales claimed it happened when there was a train in the platform ahead, but this was proved otherwise. Then it was a “blip with the points” whatever that means. Again this was disproved, and a software issue came to light. A mod was done and the situation was “improved” by changing the brake rate to stop people getting EBd in PM as the hand was dropping with no warnings. So now you don’t get EBd but even more time is wasted as the train now takes even longer to slow down. This has been the situation ever since... |
|
|
|
Post by banana99 on Apr 2, 2018 0:03:38 GMT
We have to remember that TBTC is totally untested on a system as complex as the SSR. Aldgate, Baker Street, Paddington, Earls Court all pose major challenges. Absolutely. It's a massive uptick on the relative straightforward tube lines that are very simple by comparison. However, isn't the goal to have trains running more reliably with those pesky humans? That way there will be less variance in the timetable? I think the root of all evil in this is the bloody customers (of course - as with any business we could do without them!). Dwell times will make this succeed or fail. The JL drivers don't do the long-running door-closing warning that happen on the SSL, the door alarm goes and the door shuts. If the SSL operators do the same and we get consistency across the network and if they get their dwell times down and if the passengers accept this by understanding that they will be waiting shorter intervals for the next train, then we have a good outcome on our hands. However, if the current practices continue then the timetable will go up the spout and there will be little or no improvement. Just my thoughts. |
|
|
|
Post by superteacher on Apr 2, 2018 7:35:30 GMT
Not to mention the time when the service was suspended south of Tooting and the system wouldn’t let trains into the SB platform until the previous train had cleared the siding, causing horrific delays. I’m assuming this and the Golders Green issue would have been subsequently fixed to prevent a repeat.That is a design problem and not a system issue, there is similar at Northwood, using conventional signalling, at other locations there are draw up signals to reduce speeds where conflicting moves and short overlaps could cause issues, eg Moorgate and I am sure at other sites as well. I am sure it would have been possible to design into the system that trains ran at reduced speed into Tooting when trains were leaving the siding, This is to do with mitigating the consequences of any spad.
These isues also exist with conventional signalling. At Northwood it is impossible to let a train into the southbound platform when a train is signalled out of the siding, or if it is leaving the siding it is fully berthed in the N/B platform, it is also impossible to bring a train out of the siding when there is a train in the S/B platform.
It would have been possible , with the inclusion of draw up signals when the system was installed, but a decision was taken, presumably on cost grounds not to do this.
The situation will have been the same at Tooting.
In many cases the consequences of signalling installation decisions are not realised until it is required to be used in a non standard way.
And yet the Central line works perfectly well with allowing a train into the WB platform at Marble Arch once the previous train is fully berthed into the siding. It also allows a train into the WB platform at Woodford whilst a train is leaving the siding (no. 21 Road). Thales should be dragged along to the Central to see why a previous generation ATO system works better than theirs. It is totally ridiculous that a train cannot move into the SB platform at Tooting whilst the siding is occupied. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 2, 2018 8:35:01 GMT
Not to mention the time when the service was suspended south of Tooting and the system wouldn’t let trains into the SB platform until the previous train had cleared the siding, causing horrific delays. I’m assuming this and the Golders Green issue would have been subsequently fixed to prevent a repeat.That is a design problem and not a system issue, there is similar at Northwood, using conventional signalling, at other locations there are draw up signals to reduce speeds where conflicting moves and short overlaps could cause issues, eg Moorgate and I am sure at other sites as well. I am sure it would have been possible to design into the system that trains ran at reduced speed into Tooting when trains were leaving the siding, This is to do with mitigating the consequences of any spad.
These isues also exist with conventional signalling. At Northwood it is impossible to let a train into the southbound platform when a train is signalled out of the siding, or if it is leaving the siding it is fully berthed in the N/B platform, it is also impossible to bring a train out of the siding when there is a train in the S/B platform.
It would have been possible , with the inclusion of draw up signals when the system was installed, but a decision was taken, presumably on cost grounds not to do this.
The situation will have been the same at Tooting.
In many cases the consequences of signalling installation decisions are not realised until it is required to be used in a non standard way.
As far as I'm aware, this wasn't an issue at Tooting Broadway under conventional was it? The inner home was an auto, A701 C. Why should it not be possible to have the points at the end of the southbound platform set for Colliers Wood, with a train coming out of the siding onto the northbound? There should be no risk of collision at all in the event of an overrun |
|
|
|
Post by superteacher on Apr 2, 2018 8:42:46 GMT
This is to do with mitigating the consequences of any spad.
These isues also exist with conventional signalling. At Northwood it is impossible to let a train into the southbound platform when a train is signalled out of the siding, or if it is leaving the siding it is fully berthed in the N/B platform, it is also impossible to bring a train out of the siding when there is a train in the S/B platform.
It would have been possible , with the inclusion of draw up signals when the system was installed, but a decision was taken, presumably on cost grounds not to do this.
The situation will have been the same at Tooting.
In many cases the consequences of signalling installation decisions are not realised until it is required to be used in a non standard way.
As far as I'm aware, this wasn't an issue at Tooting Broadway under conventional was it? The inner home was an auto, A701 C. Why should it not be possible to have the points at the end of the southbound platform set for Colliers Wood, with a train coming out of the siding onto the northbound? There should be no risk of collision at all in the event of an overrun Yes, Tooting Broadway used to see regular terminating trains in the peak and it was never a big issue. |
|
North End
Beneath Newington Causeway
Posts: 1,769
|
Post by North End on Apr 2, 2018 9:26:24 GMT
As far as I'm aware, this wasn't an issue at Tooting Broadway under conventional was it? The inner home was an auto, A701 C. Why should it not be possible to have the points at the end of the southbound platform set for Colliers Wood, with a train coming out of the siding onto the northbound? There should be no risk of collision at all in the event of an overrun Yes, Tooting Broadway used to see regular terminating trains in the peak and it was never a big issue. The old layout at Tooting Broadway was very unusual in that the inner home was an auto, and the points could move with a train on approach. However, there’s some wires getting crossed with regards to a train not being able to arrive on the south with the siding occupied. This situation arose during a particular failure scenario, namely the SCS failure between adopting and Morden that happened a few months ago. It doesn’t arise during normal working - but does demonstrate how the system can bite in strange and highly disruptive ways when things go wrong. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 2, 2018 9:50:42 GMT
Yes, Tooting Broadway used to see regular terminating trains in the peak and it was never a big issue. The old layout at Tooting Broadway was very unusual in that the inner home was an auto, and the points could move with a train on approach. However, there’s some wires getting crossed with regards to a train not being able to arrive on the south with the siding occupied. This situation arose during a particular failure scenario, namely the SCS failure between adopting and Morden that happened a few months ago. It doesn’t arise during normal working - but does demonstrate how the system can bite in strange and highly disruptive ways when things go wrong. But just to clarify, under normal working, it is possible to bring a train into the southbound platform with the siding occupied or a train departing the siding for the northbound main? Or entering the siding from the northbound for that matter? |
|
North End
Beneath Newington Causeway
Posts: 1,769
|
Post by North End on Apr 2, 2018 12:26:29 GMT
The old layout at Tooting Broadway was very unusual in that the inner home was an auto, and the points could move with a train on approach. However, there’s some wires getting crossed with regards to a train not being able to arrive on the south with the siding occupied. This situation arose during a particular failure scenario, namely the SCS failure between adopting and Morden that happened a few months ago. It doesn’t arise during normal working - but does demonstrate how the system can bite in strange and highly disruptive ways when things go wrong. But just to clarify, under normal working, it is possible to bring a train into the southbound platform with the siding occupied or a train departing the siding for the northbound main? Or entering the siding from the northbound for that matter? Yes you can. As I say, the SCS failure at Morden was a particular scenario that caused particular issues. |
|
|
|
Post by philthetube on Apr 3, 2018 7:26:26 GMT
Some new steel posts have appeared north or Finchley rd northbound, between the Med and Jub lines, presumably to attach signalling equipment to, I was wondering what will happen further north where the Met is sandwiched between Chiltern and the Jub, there does not appear to be space between the Met and Jub, and I imagine that using the space by the NR tracks would not be practical because of possession issues.
|
|
|
|
Post by goldenarrow on Apr 3, 2018 14:32:36 GMT
The transmission pylons between Ladbroke Grove and Goldhawk Road are situated between the running lines although it looks like those lines are separated by an exceptionally larger distance than the 6ft way that separates each direction of the Metropolitan and Jubilee.
|
|
North End
Beneath Newington Causeway
Posts: 1,769
|
Post by North End on Apr 3, 2018 14:45:21 GMT
The transmission pylons between Ladbroke Grove and Goldhawk Road are situated between the running lines although it looks like those lines are separated by a 10ft way rather than the 6ft way that separates each direction of the Metropolitan and Jubilee. IIRC this is more to do with being a legacy of GWR broad gauge than anything else, and as such I wouldn't rely on it being classed as "10 foot" in the accepted sense. |
|
|
|
Post by goldenarrow on Apr 3, 2018 15:00:02 GMT
North End , Thanks very much, post corrected.
|
|
|
|
Post by shunterl44 on Apr 4, 2018 9:52:47 GMT
I posted this back in October but I now notice that cabling has commenced but there has been not obvious rectification work to the support system and so it is still all up and down. Shame.
|
|
Ben
fotopic... whats that?
Posts: 4,282
|
Post by Ben on Apr 4, 2018 14:35:48 GMT
Disappointing. No doubt they lack the time and money to encourage people to be neat. Whats the engineering joke - you can have it quick, cheap, and quality - but you have to pick two.
|
|
|
|
Post by goldenarrow on Apr 5, 2018 15:48:19 GMT
Given that there's still a way to go till it goes live, it's difficult to see the complete picture at this stage given the unique environment that the 4LM equipment will operate: in between an ATO line with some variations in equipment that will need to be worked around when the support structures are fully cabled up.
Hopefully as the conventional signalling equipment is removed, some of tangle of support structures and cabling will be untangled. At the moment the EB Met bridges past Kilburn are quite literally overflowing with cables.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 5, 2018 16:05:01 GMT
There is very little equipment to be installed trackside
|
|
|
|
Post by MoreToJack on Apr 5, 2018 22:39:53 GMT
I also wouldn't expect a particular amount of equipment to be removed either - at least not quickly. Legacy signalling is stil in place on both the Jubilee and Northern lines, as is a lot of other redundant cabling and structures.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 5, 2018 23:59:41 GMT
It’s written into the contract they have 10 years to remove the equipment but it has to stay for a minimum of 12 months
|
|
|
|
Post by toby on Apr 6, 2018 9:32:52 GMT
On the lines with legacy signalling (including SSR), how long would it take to switch over in the event of a catastrophic failure of the current signalling?
|
|
rincew1nd
Administrator
Junior Under-wizzard of quiz
Posts: 10,286
|
Post by rincew1nd on Apr 6, 2018 9:56:26 GMT
Switch over?
Do you mean going from CBTC/TBTC back to lights on sticks?
AIUI it's only possible to revert to legacy signalling for around a month after the new system goes live, after this time period it's new system or nothing.
|
|
|
|
Post by MoreToJack on Apr 6, 2018 10:53:14 GMT
On the lines with legacy signalling (including SSR), how long would it take to switch over in the event of a catastrophic failure of the current signalling? None of the currently auto lines have legacy signalling that is still commissioned, so it would be impossible to switch over. As does happen occasionally it is a full line suspension if the signalling system falls over. As the SSR migration areas are commissioned it is planned to retain staff in the Cabins/SCCs for a few weeks afterwards in the event of any issues. It is a very quick operation in the new SERs to change control from CBTC to legacy, although this link won't be retained forever. However, it would take far longer to then have to unbag all of the redundant lineside signals and perform a comprehensive test of the frame/panel to ensure it is all working still. And that's then also assuming that you have staff (both technical and operational) on hand and with the relevant licenses still in date. For any sort of failure it's easier to repair it and take the suspension than try to faff around switching everything back in. |
|
|
|
Post by londonstuff on Apr 6, 2018 18:02:04 GMT
IIRC, on the Jubilee when they were having all the nightmares with the installation and reliability they had to bite the bullet and just left the TBTC system on even though it was far from perfect, preferring to fix it while it was working rather than keeping it testing forever, which was far past the revised deadline anyway
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 6, 2018 22:54:20 GMT
The SSL upgrade will I’m sure give the designers and testers a headache when it comes to the complex junctions we have and I don’t expect it to be failure free for a while
|
|
|
|
Post by alpinejohn on Apr 7, 2018 7:08:13 GMT
In the past I have occasionally spotted works trains slotted in between normal services on SSL and deep lines. I know these are rare events so probably not welcomed by signallers but presumably when they do happen they really need to happen. As they usually trigger the “Special” message on platform displays, I assume paths for works trains are treated in much the same way as regular trains at least while conventional signalling remains available.
So come the switchover on SSR lines and conventional signals are decommisioned, how will these works trains talk to the new system? If they are completely banned from operation during service hours it would presumably complicate responding quickly to any urgent incident like delivering a replacement for a broken rail or set of points in any of the undergound SSR sections.
Have Thales/Bombardier been commissioned to kit out some of the DM/battery units with suitable comms equipment so they can still operate during normal service hours?
|
|
|
|
Post by superteacher on Apr 7, 2018 7:24:20 GMT
I believe that the engineering fleet have in-cab equipment which enables them to run under the new signalling.
|
|
