Turnaround Headcode

[cleatus_1]

Trains often turnaround with a sidings or a piece of tracks if there isn't a junction connecting both platforms. For example, Worcester Foregate Street, Longbridge, Paddington (XR), Adwick and more.

Makes me wonder why sometimes it uses one headcode, like Worcester Foregate Street - Henwick - Worcester Foregate Street.

Realtime Trains | 5T43 0926 Worcester Foregate Street to Worcester Foregate Street | 26/04/2025

Real-time train running information for 5T43 0926 departure from Worcester Foregate Street to Worcester Foregate Street on 26/04/2025. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

But sometimes it uses two headcodes, like Paddington - Westbourne Park Carriage Sidings - Paddington.

Realtime Trains | 3Y21 0600 London Paddington to Westbourne Park Cs | 26/04/2025

Real-time train running information for 3Y21 0600 departure from London Paddington to Westbourne Park Cs on 26/04/2025. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Realtime Trains | 3W16 0612 Westbourne Park Cs to London Paddington | 26/04/2025

Real-time train running information for 3W16 0612 departure from Westbourne Park Cs to London Paddington on 26/04/2025. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Anyone know when will which be used?

 

[Eloise]

In my experience it varies. You normally split it If the system requires it. Some ARS systems don't like the same headcode at the same timing point twice, some do. One headcode is also neater as it leaves the line on the graph for the entire time it is sat wherever it spinning round. If you plan it as two schedules then good practice is to join the lines up on the graph to show a train sat there e.g. 3A01 forms 3D02 but stands for 20 mins you would need to show one service having a 20 mins dwell to join them up. Not all do this so you run the risk of someone planning a schedule straight through a standing train. Some ops teams like to know what train ECS comes off and then forms hence two. No real hard and fast rule as you've found. Only thing I work to is where one headcode is used for the entire move this is normally the same as the service it forms e.g. 1A01 forms 1D02 the ECS would be 5D02.

Personally over the years I have moved to one shunt move unless the system dictates two.

 

[whoosh]

I'll just add that Westbourne Park has two reversing sidings, and it may be better to have the out and back ECS moves to be separate headcodes, as that would make it easier to step a train up/reform a train's diagram using another set from the other siding instead - for example, during disruption, or planned alterations due to engineering work.

So, with somewhere where there's more than one siding, it might be easier or provide more flexibility to go with the two headcodes option.

Also Westbourne Park, uniquely, has 'auto-reverse', so it may be a necessity as part of that system's operation, to have two separate headcodes - one for each part of the move being done.

 

[Horizon22]

I'll just add that Westbourne Park has two reversing sidings, and it may be better to have the out and back ECS moves to be separate headcodes, as that would make it easier to step a train up/reform a train's diagram using another set from the other siding instead - for example, during disruption, or planned alterations due to engineering work.

Westbourne Park has 3.

So, with somewhere where there's more than one siding, it might be easier or provide more flexibility to go with the two headcodes option.

Also Westbourne Park, uniquely, has 'auto-reverse', so it may be a necessity as part of that system's operation, to have two separate headcodes - one for each part of the move being done.

It's a different signalling system to really anywhere else (CBTC) and the signalling system will need separate headcodes.

 

[Wilts Wanderer]

The ARS at Paddington specifically does not accept headcodes with the same location twice, to my understanding. Otherwise it’s generally good practice to use a through headcode for an out-and-back shunt as it helps inform operational staff what the unit’s next passenger working is, which helps with regulation etc.

 

[Magdalia]

Thameslink/GN are not consistent: St Albans uses one headcode for moves in and out of the reversing siding, but Letchworth and Welwyn use two headcodes for moves in and out of the arrival/departure sidings and carriage sidings respectively.

 

[John Webb]

Thameslink/GN are not consistent: St Albans uses one headcode for moves in and out of the reversing siding, but Letchworth and Welwyn use two headcodes for moves in and out of the arrival/departure sidings and carriage sidings respectively.

Interesting observation. Watching the track berths on whatever system Railcam uses I thought the headcode was changed to the new departure one shortly before the train left the turnback siding?
(Curiously when first installed in the late 1970s the turnback siding had no track-circuit or readout on the then new eNtrance-eXit panel at West Hampstead PSB; signallers there had to use notes to remind themselves what was in it! It was only several years ago when the N-X panel replacement was being planned that it was given a T-C of its own.)

 

[Eloise]

Would need to check but you can effectively set swap in Welwyn and Letchworth sidings which may be done on occasions so the unit on may not be the unit off. At St Albans the unit in has to be the unit out so can be done on one headcode. Some ECS show shunt plus lengthy ECS move to a depot too all under one headcode.