ECML Disruption (shock!) (01/06/2026)

[Nicholas Lewis]

I still haven’t worked out how it works or why it still exists at all. Even when I started at SBB 30 years ago, we no longer had a method in use that was anywhere near as complicated and slow. Today, it is estimated that more than 95% of lines are equipped for bidirectional operation. Even the few lines without bidirectional operation can be used with signalling and without paperwork.

Yesterday, a points failure occurred on the busiest section of Switzerland’s east-west axis, resulting in a reduction from two to one usable tracks. After the usual “chaotic phase”, a plan was drawn up (or rather, a pre-prepared operational plan was retrieved from the electronic “drawers” and adapted). Thanks to this plan and a modern, flexible signalling system, 17 trains per hour were able to run over this single-track section. Theoretically, more would be possible if trains were run in groups. Naturally, staff and rolling stock deployment also had to be adjusted accordingly. Thanks to the regular timetable, the scenario was repeated every hour with only minor adjustments.

We could only wish for such luxury and even in the limited areas where we do have reversible signalling or SIMBIDS (simplified bi-directional signalling) installed even using that in the wrong direction is fraught with issues. Thus abandon the service has become the norm all too often but we are pretty good and getting things all square for the next day so organisationally UK is still has embedded capability and that is to be celebrated.

 

[hexagon789]

Was that delay the longest to a service that actually completed it's scheduled journey?

I'm sure we had a 480+ mins on an LNER London to Edinburgh last year on a completed journey.

There have been longer delays than that, even in relatively recent memory.

 

[Segedunum]

I'm sure we had a 480+ mins on an LNER London to Edinburgh last year on a completed journey.

There have been longer delays than that, even in relatively recent memory.

The one that sticks out in my memory was when SWR services got stuck in some extremely cold weather around the New Forest several years ago - I'm sure the delay to one service was something like 1270 minutes.

 

[harz99]

I'm sure we had a 480+ mins on an LNER London to Edinburgh last year on a completed journey.

There have been longer delays than that, even in relatively recent memory.

Yes of course, but I was talking about the disruption yesterday which this thread is about, not historical data...I'll amend the post slightly.

 

[Segedunum]

Yes of course, but I was talking about the disruption yesterday which this thread is about, not historical data...I'll amend the post slightly.

Likely the GC service: https://www.realtimetrains.co.uk/service/gb-nr:Y14826/2026-06-01/detailed#allox_id=0

That left York first and was the earliest scheduled - most delayed LNER service arrived London 191 late (1E25 from Aberdeen).

1S31 arrived Newcastle even later than scheduled but was binned there: https://www.realtimetrains.co.uk/service/gb-nr:G80273/2026-06-01/detailed

 

[Falcon1200]

At the end of the day, it doesn’t matter which train is available the next morning for a particular service.

It absolutely does! When trains are on a mileage-based maintenance schedule they have to presented to a suitable depot before running out of miles, and therefore have to be allocated to a diagram finishing at the depot, or at least to a diagram allowing a simple and not too disruptive change to another diagram. Before we start considering fuelling needs (not an issue in Switzerland of course).

 

[hexagon789]

The one that sticks out in my memory was when SWR services got stuck in some extremely cold weather around the New Forest several years ago - I'm sure the delay to one service was something like 1270 minutes.

I remember reading of that one, must've been awful conditions.

Yes of course, but I was talking about the disruption yesterday which this thread is about, not historical data...I'll amend the post slightly.

No worries. In which case, I think your example was the longest from last night.

 

[QueensCurve]

I still haven’t worked out how it works or why it still exists at all. Even when I started at SBB 30 years ago, we no longer had a method in use that was anywhere near as complicated and slow. Today, it is estimated that more than 95% of lines are equipped for bidirectional operation. Even the few lines without bidirectional operation can be used with signalling and without paperwork.

Yesterday, a points failure occurred on the busiest section of Switzerland’s east-west axis, resulting in a reduction from two to one usable tracks. After the usual “chaotic phase”, a plan was drawn up (or rather, a pre-prepared operational plan was retrieved from the electronic “drawers” and adapted). Thanks to this plan and a modern, flexible signalling system, 17 trains per hour were able to run over this single-track section. Theoretically, more would be possible if trains were run in groups. Naturally, staff and rolling stock deployment also had to be adjusted accordingly. Thanks to the regular timetable, the scenario was repeated every hour with only minor adjustments.

I have always been impressed by the bidirectional signalling in Switzerland and the almost universal provision of high speed turnouts to use it. I recall an August morning in 1996 where we lurched onto the other line at St Leonards in the Rhone Valley and I looked out through th droplight to see two men working to collect what I believe were body parts from the 4 foot.

I am despair about the fact that British Rail and their successors failed to provide bidirectional running even on the WCML. They didn't even install it on Beattock bank which would have paid for itself many times older in the >50 years since the intallation of the current signalling. I recall being delayed 40 minutes waiting for the pilotman when APT derailed at Milnthorpe in 1980 and about 1984 being delayed 2 hours south of Lancaster when the second train in front had failed in the platform. At least the platforms at Lancaster are now bidirectional. There are countless less memorable times where I have cursed the delay for want of a bidirectional facility.

 

[D1537]

It absolutely does! When trains are on a mileage-based maintenance schedule they have to presented to a suitable depot before running out of miles, and therefore have to be allocated to a diagram finishing at the depot, or at least to a diagram allowing a simple and not too disruptive change to another diagram. Before we start considering fuelling needs (not an issue in Switzerland of course).

Which is what happened; Heaton ended up with 3 extra sets, Craigentinny with 2 fewer, and Bounds Green with 1 fewer, hence the fiddling around this morning to start two southbounds and a northbound at Newcastle.

 

[Parham Wood]

I still haven’t worked out how it works or why it still exists at all. Even when I started at SBB 30 years ago, we no longer had a method in use that was anywhere near as complicated and slow. Today, it is estimated that more than 95% of lines are equipped for bidirectional operation. Even the few lines without bidirectional operation can be used with signalling and without paperwork.

Yesterday, a points failure occurred on the busiest section of Switzerland’s east-west axis, resulting in a reduction from two to one usable tracks. After the usual “chaotic phase”, a plan was drawn up (or rather, a pre-prepared operational plan was retrieved from the electronic “drawers” and adapted). Thanks to this plan and a modern, flexible signalling system, 17 trains per hour were able to run over this single-track section. Theoretically, more would be possible if trains were run in groups. Naturally, staff and rolling stock deployment also had to be adjusted accordingly. Thanks to the regular timetable, the scenario was repeated every hour with only minor adjustments.

If I recall correctly at one time on the Paddington to Bristol main line where bi-directional signalling was installed if any people were working on the track they did something to lock out bi-directional working on the basis that it would be unsafe for the staff on the track. Not sure if I have this correct but what do you do if staff are working on the line in Switzerland and bi-directional working is adopted?

 

[Horizon22]

This certainly involves a great deal of effort, but it is by no means an insurmountable problem. The key word is flexibility: flexibility in the deployment of rolling stock, in the deployment of staff, and in the infrastructure. However, all of this also requires a high degree of cooperation and standardisation between the TOCs and between the TOCs and the network operator (bidirectional signalling, rolling stock and staff freely deployable on many routes). If trains can be deployed flexibly and staff can also be deployed flexibly on these trains, things become much simpler. Fragmentation and bespoke solutions are the opposite. At the end of the day, it doesn’t matter which train is available the next morning for a particular service. The only thing that matters is that a suitable solution is found.

This also discounts any engineering works happening overnight. Many drivers are not available overnight either, especially if they are already delayed off duty.

 

[Segedunum]

This also discounts any engineering works happening overnight. Many drivers are not available overnight either, especially if they are already delayed off duty.

At least in this case there was some flexibility from an engineering work perspective. 1S84 (1933 Kings Cross - Edinburgh) only ran as that (and not as 1N14 to Newcastle) because there was a scheduled possession north of Newcastle which meant the usual last service (1S32 2003 Kings Cross - Edinburgh) couldn't get through in time. As it happens, it ran all the way through with a considerable delay.

Similarly, 1D36 2333 Kings Cross - Leeds was booked via Methley Junction due to a possession, but this was clearly eased in some form to allow it to run via Wakefield Westgate as a Mk4 set.

 

[QueensCurve]

If I recall correctly at one time on the Paddington to Bristol main line where bi-directional signalling was installed if any people were working on the track they did something to lock out bi-directional working on the basis that it would be unsafe for the staff on the track. Not sure if I have this correct but what do you do if staff are working on the line in Switzerland and bi-directional working is adopted?

Didcot to Bristols and Severn Tunnel Junction had bidirectional signalling installed about 1990 when I was iiving in Bath. I believe there was such a lockout but I still think the sense of it is wrong. One track should be used for the trains and the other locked out to make a safe zone for the track inspection work.

 

[Falcon1200]

I despair about the fact that British Rail and their successors failed to provide bidirectional running even on the WCML

I could not agree more! For some years I was Railtrack/Network Rail Controller for the Scottish section of the WCML and dealt with innumerable incidents where bi-di signalling would have hugely reduced the disruption and delay.

There are countless less memorable times where I have cursed the delay for want of a bidirectional facility.

Indeed; Even before I retired, in 2016, it could easily take up to two hours to introduce Single Line Working, thanks to the number of staff required and the remoteness of some locations. I doubt things are any better now.