SSR Resignalling

Did I miss something?

The last few posts suggest that a broken down train delay actually due to a non communicating train issue?

Even after entire S stock operations on the SSR is entirely signalled by CBTC, I recall seeing a post suggesting that Depot movements would still be handled locally. Presumably this means that every depot will eventually be installing some sort of "entry point" checking equipment where trains would be presented from the depot signalling system and held until the CBTC is ready to take over their routing etc.

Is there a case for installing(or commissioning) this kit now/early?

In the short term using this kit should if nothing else allow it to spot any non-communicating before they leave the depot - rather than wait until they try to enter an "active SMA" and cause unwelcome delays?

I don't know what the speed limit is, but on the Jubilee line a train is authorised by the controller to move to a specific location, who will have ensured that no other trains come into conflict with it. In the example I observed (quite a few years ago now) a failed train was authorised from the westbound platform at Canning Town to the centre platform at North Greenwich. This was after at least 15 minutes (possibly longer) of the driver (initially alone, and then with someone I presume was a technician) trying to fix the train (which iirc originally failed somewhere between West Ham and Canning Town). Initially the eastbound service operated normally, then with delays but eventually the entire line was suspended (although there was also an issue on the westbound somewhere between Baker Street and Wembley Park, I can't remember more precisely than that).

If the SSR operates similarly then the delay would be in large part not the movement to the siding/depot, but trying to fix then problem so you don't need to run at 10mph. At what point you stop attempting to fix it and move a train out the way is always going to be a judgement call, and one where the right answer is never knowable until after you've made the decision. Assuming that getting from the failure location to the nearest out of the way location at 10mph takes 10 minutes, then spending anything up to 9 minutes doing things to fix the problem is a guaranteed win overall. However, if you spend those nine minutes and the problem isn't fixed you've caused a 19-minute delay rather than a 10-minute one. But if you continue to try and fix it and succeed after a further three minutes then 13 minutes is better than 19. The calculation becomes even more complicated if moving the train will disrupt a different bit of railway - e.g. if there is a failed train at Bayswater, it's likely that at least some trains can be turned at HSK and the H&C service at Edgware Road will not suffer significant delay, but to move the train to a siding you're going to need to stop movements at one or the other.

As a layman, I’m finding that hard to comprehend that the failure of one train to communicate with CBTC means that the whole signalling area must be closed down. Is there really no train detection to allow safe isolation of the non communicating train while maintaining movements elsewhere?

The driver in restricted manual as mentioned before is limited to roughly 11mph and is driving to line of sight signalling. So the safest approach is to move all trains out of its path to avoid collisions

Trains behind a defective train in theory could run as the defective one would still be tracked via the axle counters

Nobody is doubting that what happens is because of safety, but the disconnect is at the extent of the area that needs to be shut down.
What is not obvious to a lay person why if you have a failed train on the eastbound at Paddington circle that needs to move to Edgware road, why that would require movement to be suspended all the way to Hammersmith. AIUI you haven't lost detection on any points, and you know where all the other trains are, so it is not intuitive why any train on a non-conflicting route needs to be stopped.

For example, if the non-communicating train was last proven to be in 2 at Paddington, and needs to move to platform 1 at Edgware Road, the lay person's assumption is that this will effectively set up a block section between those two points and that any train that doesn't require use of track in the shaded area in the diagram below will be able to move with the signal protection as normal. For example a train currently in either westbound platform at Paddington would not be affected.


When you say it doesn't work like that, we are not disagreeing that we are wrong about the status quo. What we are doing is failing to understand why the system does not/cannot work in a manner approximating what we expect because that would cause less disruption.
There may be a good reason why it is impossible for any system to allow movement of other trains that in the control area but nowhere near the failure, but if there is it is not obvious to a lay person, and so it needs explaining to us.

Theres always “ what ifs “ remember that

It still does happen. It did this morning on my District Line train outside Earl's Court.

I'd be interested to compare procedures here to the Jubilee, Northern and DLR that are also using various versions of Thales Seltrac.

All three of these lines/systems have their fair share of failed trains but they don't seem to block the line for 30 minutes.

So they have definitely designed out bi-di running from everything? I thought this was the source of much subsequent frustration on the Jubilee, as well as an extra cost to remove it from the design?

The next station is only a fraction of the story - its the next place of refuge where the train can be berthed so that other trains can pass it it which matters. ie, a siding, a bay platform, etc.

These are very comparatively rare and (for instance: Holborn on the Central line, Mansion House on the District line) even where they have existed they have often been removed.

Thankfully 10 mph running is exceptionally rare. As an aside, I've heard of it happening in the Central line too, although (especially on the Central line) it is possible for trains that can communicate with the signalling computers but not be driven by them to travel at normal speeds. This is because the line has many lineside signals, although when travelling in this mode there is a reduction in line capacity (the white signal aspect). The track recording and autumnal leaf-fall trains fall into this category. As far as I am aware, the other tube lines which have been automated do not have this facility (or if it is a design possibility it has not been enabled here in London). The subsurface will also be so afflicted, except on routes where operation with human - driven trains is being included in the overall scheme.

I was at Euston Square at about 8.30pm (Weds 2nd Oct 2019) and noted that trains were queuing to enter the station (both directions).

An S8 entering the CBTC zone was taking a very long time to get its 'authority to travel' and this was delaying trains behind it - the train had a blue signal light, but not white lights on the lamp clusters on the car sides.

Quite why there were queues on the eastbound was a mystery as trains were departing after what seemed to be normal length station stops.

I had my camcorder with me, so eventually everyone will be able to see this.

A post bump for metrider  , hopefully this should answer your question 

Indeed there are, mostly because of the ballast to slab track conversion works that have given a much improved track form and that in turn has facilitated significant speed uplifts on the straight single bore sections. Only remaining sections of bull head track on this section now are at Swiss Cottage and Lord's.

At the time of of that graph being drawn up, I believe there were some intervention vectors being tested out which might explain the erraticity to/from Finchley Road in particular. The track condition in that area is capable of supporting existing speed limits (30 SB & 25 NB I think) so I'm concluding that its more to do with a migration boundary being close by.

But in general, the quality of track in this area is fantastic compared to what it was a few years ago, testimony to the massive speed uplift in places.