the Increasingly Frustrating Thameslink

[BzRail]

That's interesting. The signal failure was at 0620. I was at St. Albans station a little over an hour later. It was not in chaos. There were some trains running a little late, and a cancellation (no driver), but everyone got on trains and none were too heavily loaded. Perspective please!

No it wasn't. I was waiting for the 0531 from Haywards Heath which was delayed by the same 'signalling' problem. So it must have happened at least an hour earlier.
It is worrying that even you Rick may be being given misleading information.

It was almost amusing that any hint of an apology had been removed from the (new?) robo-announcements. The tone was 'train late, too bad'.

At least there was also a human announcer making an effort.

 

[SpacePhoenix]

I know the core would never get any relief tunnels bored because of cost but if cost was not an issue would the geology prevent tunnels for relief tracks being bored?

 

[Bald Rick]

No it wasn't. I was waiting for the 0531 from Haywards Heath which was delayed by the same 'signalling' problem. So it must have happened at least an hour earlier.
It is worrying that even you Rick may be being given misleading information.
.

That made me sit up, as I rather haughtily regard myself as a paragon of virtue. So I checked my facts. And you are right!

The failure was first reported at 0535. It didn't get reported on the system I use - which is normally pretty swift - until 0620.

Nevertheless it rather adds to the argument, in that I was at St Albans 2 hours after the failure, and the north was not in chaos.

 

[Bishopstone]

Another cancellation at Balcombe this afternoon, giving a two hour gap between southbound trains (15.21 - 17.21) unless they put a stop order on another service, although there is no such indication on the Thameslink Twitter feed or LDB.

 

[FlippyFF]

Is the building that's bounded by Farringdon platform 4, Cowcross Street, Farringdon Street and Clerkenwell Road on National Rail property? I presume like most modern buildings it will 'soon' need to be demolished and rebuilt, if this were to happen could a third track and platform for Thameslink be provided?
(Apologies if this should be in a different thread?)

Simon

 

[Bald Rick]

Is the building that's bounded by Farringdon platform 4, Cowcross Street, Farringdon Street and Clerkenwell Road on National Rail property? I presume like most modern buildings it will 'soon' need to be demolished and rebuilt, if this were to happen could a third track and platform for Thameslink be provided?
(Apologies if this should be in a different thread?)

Simon

No it isn't, and trust me, there would be no point.

 

[21C101]

That made me sit up, as I rather haughtily regard myself as a paragon of virtue. So I checked my facts. And you are right!

The failure was first reported at 0535. It didn't get reported on the system I use - which is normally pretty swift - until 0620.

Nevertheless it rather adds to the argument, in that I was at St Albans 2 hours after the failure, and the north was not in chaos.

To be fair thats because most trains at that time of day come out of one or other of the depots/sidings at Bedford, and quite a lot of even southbound fasts at that time of day are heading for the Wimbledon loop or the southeast division( effecitvely they are ECS trains doing a fast run to London to make best use of them). An hour later and I suspect it would have been a barrel of laughs at St Albans.

 

[DelayRepay]

To be fair thats because most trains at that time of day come out of one or other of the depots/sidings at Bedford, and quite a lot of even southbound fasts at that time of day are heading for the Wimbledon loop or the southeast division( effecitvely they are ECS trains doing a fast run to London to make best use of them). An hour later and I suspect it would have been a barrel of laughs at St Albans.

I was at St Albans at about 8:30 and to be fair, I didn't notice chaos either. I think there was one cancellation heading north and a couple of 5 minute delays (although I was going South so didn't pay a lot of attention to the northbound departures). But it wasn't chaos this time. The extent of the delays/cancellations was just a normal day on Thameslink.

 

[BzRail]

I was at St Albans at about 8:30 and to be fair, I didn't notice chaos either. I think there was one cancellation heading north and a couple of 5 minute delays (although I was going South so didn't pay a lot of attention to the northbound departures). But it wasn't chaos this time. The extent of the delays/cancellations was just a normal day on Thameslink.

It would take a while for the delays to build up on the Northern end. My 5.31 from Haywards Heath was 30 mins late at London Bridge, but RealTimeTrains shows only 10 mins down at Bedford. It was not until nearer 7am South and 10am North that delays significantly increased, I assume due to congestion through the problem area.
This would tally with what Rick and others have said.

 

[DelayRepay]

I have a question which maybe one of the Thameslink drivers on this thread can answer: Why do the 377s always have problems opening their doors at St Pancras? It always seems to take for ever and non-regular passengers start to worry.

 

[Class377]

Would I be right in saying the 377s use GPS to detect where they are before releasing doors, and with St Pancras being underground they struggle to get a position? I seem to remember a driver announcement with a similar message back when they were introduced, and with the omnishambles that the electrostars have been for the route it wouldn't surprise me if the problem persisted.

 

[asylumxl]

I have a question which maybe one of the Thameslink drivers on this thread can answer: Why do the 377s always have problems opening their doors at St Pancras? It always seems to take for ever and non-regular passengers start to worry.

Don't need to be a driver to answer that. The design of the station means that the waves created by the OHLE can't escape and cause interference, making it hard for the train to pick up the signal from the beacon on the line.
--- old post above --- --- new post below ---

Would I be right in saying the 377s use GPS to detect where they are before releasing doors, and with St Pancras being underground they struggle to get a position? I seem to remember a driver announcement with a similar message back when they were introduced, and with the omnishambles that the electrostars have been for the route it wouldn't surprise me if the problem persisted.

Have you ever used GPS in a concrete bunker?

 

[AM9]

Don't need to be a driver to answer that. The design of the station means that the waves created by the OHLE can't escape and cause interference, making it hard for the train to pick up the signal from the beacon on the line.
--- old post above --- --- new post below ---



Have you ever used GPS in a concrete bunker?

That post offers two different answers to the question:

1) that the 'waves' can't escape causing interference, i.e. radiated emissions

and

2) using GPS in a concrete bunker as a comparison for the STP situation. Well a bunker would just be a screen to external signals (whether it had any power circuits in it or not) which would then result in a low available signal strength from the external source.

If NR have installed a local repeater in the station box, the outside signals are irrelevant, in fact they would exacerbate the situation with multipath interference. So it looks like a simple EMC issue with the line borne harmonics and chopper components created by the electronics in the Electrostars. One would hope that the 700 GPS set is fully immunised from the trains' onboard power equipment including components exported to the OHLE by other trains in the section.

 

[asylumxl]

It didn't offer two explanations. It's purely pointing out that GPS wouldn't work down there.

It is purely a matter of interference.

 

[AmeralGunson]

Wouldn't it be easier just to install a button marked, "release doors" in the driving cab? It seems to me that there's a little too much to go wrong with all this GPS technology.

 

[hassaanhc]

Wouldn't it be easier just to install a button marked, "release doors" in the driving cab? It seems to me that there's a little too much to go wrong with all this GPS technology.

If they can't detect where they are, then only the front 8 coaches are opened with the override, as the shortest platforms these call at only fit 8 coaches (although Mitcham Junction and Hackbridge where these don't go only fit 7 Electrostar coaches, and only just fit full 8 319 and 455 with front cab off the platform).

 

[Taunton]

How on earth can such fundamental problems not have been sorted out years after both trains and the new station were commissioned?

 

[tsr]

Electrostar Mitrac computer interfaces have emergency door release controls which can override the SDO system onboard. I don't know about Thameslink stock, but Southern 377s should allow the driver to select the doors to be released by touching them on the screen. This is usually only if the train's location is not listed on-screen or can't be selected for some reason. The door release buttons are then activated.

However, there will quite possibly be other steps taken before this functionality is used by a driver, and even if not, there is some delay.

 

[21C101]

If they can't detect where they are, then only the front 8 coaches are opened with the override, as the shortest platforms these call at only fit 8 coaches (although Mitcham Junction and Hackbridge where these don't go only fit 7 Electrostar coaches, and only just fit full 8 319 and 455 with front cab off the platform).

Apparently, this has also been happening with SE electrostars in the Blackfriars bay platforms, which I doubt have any problems picking up GPS, resulting in the pax in the last 4 carriages having a long wait to walk through the train.

 

[jon0844]

If the problem persists at St Pancras, why not just put the S stop at the far end and make all trains stop there (complete with platform boards to show where the rear of a 4, 8 and 12 car train is) then have the driver manually open the doors every time?

Yes it will involve a fair walk for any 4 car services but there's no chance of any driver stopping short by mistake.

Doing this might just encourage the powers that be to seek a foolproof solution if and when people complain.

 

[Bald Rick]

Fortunately the 700s will be using the ETCS beacons to tell where it is, and won't be relying on the Hima-Sella kit which is what is susceptible to the interference.

Need to point out that it is a combination of the interference AND the absence of a GPS signal that causes 'computer says no'. If the train could get a GPS signal it would know where it is and enable the door release controls.

 

[Taunton]

Need to point out that it is a combination of the interference AND the absence of a GPS signal that causes 'computer says no'. If the train could get a GPS signal it would know where it is and enable the door release controls.

What genius specified a system for a route using tunnels through a city which used GPS? :roll:

If the problem persists at St Pancras, why not just put the S stop at the far end and make all trains stop there (complete with platform boards to show where the rear of a 4, 8 and 12 car train is) then have the driver manually open the doors every time?

Yes it will involve a fair walk for any 4 car services but there's no chance of any driver stopping short by mistake.

Now my experience is that this is exactly what is done at Farringdon southbound. I stepped down from the westbound Circle Line onto the north end of the southbound Thameslink platform, and the next train in was 4-car which rolled right past me to the very east end of the platform. Needless to say there is no indication (or none I saw) that a train was only 4-car and thus, although shown on the indicator there, would not be serving that part of the platform.

Maybe some realist has foreseen that if the number of cars was shown on the indicator it would be got wrong 50% of the time.

 

[OxtedL]

What genius specified a system for a route using tunnels through a city which used GPS? :roll:

It is designed so you use GPS when there is coverage, but otherwise can install beacons. The idea being you avoid spending lots on beacons.

 

[AM9]

Fortunately the 700s will be using the ETCS beacons to tell where it is, and won't be relying on the Hima-Sella kit which is what is susceptible to the interference.

Need to point out that it is a combination of the interference AND the absence of a GPS signal that causes 'computer says no'. If the train could get a GPS signal it would know where it is and enable the door release controls.

So out of the £6Billion or so budget for Thameslink, why can't they invest a few tens of £K to install something like this in the critical core locations:
http://www.gps-repeaters.com/?gclid=CjwKEAiA_NmkBRCe3ubC1aWAtEcSJACxkkbqxs-xcRbdU4aiHK0RAz2pQyJ3j3LkAkrObJ0ITV_A0RoCy5bw_wcB#
Even if they needed about four for STP, the cost would be peanuts compared with the potential for delays in the core stations at 24+ tph. The receiving aerials could be located vertically beside the MML tracks/platforms and if there was any offset, that would be programmed into the permissions for door opening locations on the trains. These things are cheap enough to install in offices, so a major commuter rail station could be regarded as equally if not much more important.

 

[SPADTrap]

So out of the £6Billion or so budget for Thameslink, why can't they invest a few tens of £K to install something like this in the critical core locations:
http://www.gps-repeaters.com/?gclid=CjwKEAiA_NmkBRCe3ubC1aWAtEcSJACxkkbqxs-xcRbdU4aiHK0RAz2pQyJ3j3LkAkrObJ0ITV_A0RoCy5bw_wcB#
Even if they needed about four for STP, the cost would be peanuts compared with the potential for delays in the core stations at 24+ tph. The receiving aerials could be located vertically beside the MML tracks/platforms and if there was any offset, that would be programmed into the permissions for door opening locations on the trains. These things are cheap enough to install in offices, so a major commuter rail station could be regarded as equally if not much more important.

Would it really be the same bit of kit found in any office though? I don't see why not but it just doesn't sound like the railway way :P

 

[21C101]

Would it really be the same bit of kit found in any office though? I don't see why not but it just doesn't sound like the railway way :P

Given that the railway official calamity proof version is suffering from EMC problems anything that is just rated for residential, light industrial and commerical use is likely to be fried in short order there.

 

[Taunton]

It is designed so you use GPS when there is coverage, but otherwise can install beacons. The idea being you avoid spending lots on beacons.

Ah. So what I should have said was :

"What genius specified a system for a route using tunnels through a city which used GPS, and a beacon system on a 25Kv railway susceptible to interference from 25Kv".

 

[Aictos]

If the problem persists at St Pancras, why not just put the S stop at the far end and make all trains stop there (complete with platform boards to show where the rear of a 4, 8 and 12 car train is) then have the driver manually open the doors every time?

Yes it will involve a fair walk for any 4 car services but there's no chance of any driver stopping short by mistake.

Doing this might just encourage the powers that be to seek a foolproof solution if and when people complain.

But you can't do that as it means inconveniencing the passengers who might have further to walk..... Not my own view of course but that of a ex FCC MD...

Take the Inner Suburban stations between Harringay and Watton At Stone via Hertford North, they can only take 6 car trains so a simple solution would be like you suggest to put a single S stop sign by the 6 car marker as the majority of trains are now 6 cars.

Now for operationally reasons, I think all trains should stop in the same position be it a Class 317, Class 365, Class 321 or Class 313 so regardless of the length of said train if it's booked to called at a station then it will fit thus reducing the risk of station overruns least it will provide a uniform stop instead of having trains stop at different parts of the platform.

 

[OxtedL]

I think most GPS repeaters work by simply re-broadcasting signals received on an outdoor antenna. [If you think about it, there's no other easy way of doing this without inventing your own indoor triangulation system, by which point track beacons are just as useful.]

This obviously means that the position a GPS receiver detects is that of the external antenna, wherever you are. This is useless as a safety feature unless the repeater has a very short range focussed on a specific area (and then the train is given the external coordinates), exactly the opposite of what an off-the-shelf repeater would be designed to do, and more what a track based beacon-type-thing is designed to do.

"What genius specified a system for a route using tunnels through a city which used GPS, and a beacon system on a 25Kv railway susceptible to interference from 25Kv".

A better question. I'm still not entirely convinced it is the right one to ask though - I imagine the system is a bolt on from the original system for Electrostars, used at places like (I think?) Victoria, possibly before it was required for Thameslink(?).

The problem sounds very specific to conditions at St Pancras as well, so perhaps it is something entirely unprecedented that even a good testing regime might have missed.

This is a lot of speculation though, and I'm happy to accept that maybe the kit is a load of crap.

 

[SPADTrap]

But you can't do that as it means inconveniencing the passengers who might have further to walk..... Not my own view of course but that of a ex FCC MD...

Take the Inner Suburban stations between Harringay and Watton At Stone via Hertford North, they can only take 6 car trains so a simple solution would be like you suggest to put a single S stop sign by the 6 car marker as the majority of trains are now 6 cars.

Now for operationally reasons, I think all trains should stop in the same position be it a Class 317, Class 365, Class 321 or Class 313 so regardless of the length of said train if it's booked to called at a station then it will fit thus reducing the risk of station overruns least it will provide a uniform stop instead of having trains stop at different parts of the platform.

It also means passengers would be stupid and run for trains that are stopping further down a platform, fall over then sue the company FAR more frequently than a stop short.