When it was announced that Virgin Trains was successful in being allocated maintenance depot capacity at Temple Mills in east London, there was a sort of reaction “we’ll that will mean that Virgin will be the competitor to Eurostar then.” But no sooner had the decision been made that I started to have this nagging feeling that this was not the last of this maintenance issue, and that not being allocated maintenance capacity at Temple Mills could even be a blessing in disguise.
Two pieces of information that have come to light since the Temple Mills decision further confirm this. First, Trenitalia is reported to be looking for a maintenance facility close to Paris, and is not abandoning plans for London. And second a source close to one of the other unsuccessful bidders for Temple Mills told me they are not giving up as they are sure there are bumps ahead before Virgin has everything they need in Temple Mills.
In the French national market both Proxima/Velvet and Kevin Speed/illisto are advancing with plans to not rely on existing workshops for their maintenance, and in the UK DfT has started to think openly about alternatives to Temple Mills.
So then – in typical #CrossBorderRail / #CrossChannelRail style – can we start to work out, publicly, where such maintenance facilities would be constructed?
What does any operator need to find? Ideally a brownfield site, something that is or was railway land, and has easy track access to the core high speed Channel Tunnel routes, so somewhere along HS1 to the Channel Tunnel, along LGV Nord in France, or potentially along LGV 1 in Belgium. Space to build a maintenance shed at least 400m in length would be ideal, although 200m would be enough at a push. And this site would ideally be as close as possible to one of the main terminus stations for Channel Tunnel routes – London St Pancras, Paris Nord or, possibly, Bruxelles Midi. For Trenitalia (and hence not for Gemini, Evolyn or Virgin, in the case of Temple Mills issues) a location that is also accessible from Paris Gare de Lyon from which their services depart to the rest of France would make additional sense. And more widely were it not possible to respect all of those criteria I suppose some sites in the middle of the routes – in particular in or near Lille – might need some additional consideration, although that might then necessitate services starting or terminating in Lille that no one has done to date.
So what have I found so far?
Everything is gathered together on this so far rather provisional map – zoomable version.
Explanations of each area, UK first:
Nothing is as good as Temple Mills, but there are a few options. Something at or close to the Barking container terminal (20km from St Pancras) might work, although track access off HS1 and the total space available are complicated. There are sidings in an old quarry at Northfleet, just beyond Ebbsfleet station on HS1. A depot here would likely accommodate max 200m trains, but it is brownfield land and is just 37km from St Pancras. At Ashford there is the maintenance facility for Hitachi’s class 395 trains, but access to the depot is third rail electrified only at the moment, and the site is very constrained. Ex railway land south east of Ashford station also merits a further look, although Ashford is 91km from London. Dollands Moor the under-used freight terminal at the UK end of the Channel Tunnel would likely have enough space, but as it is 107km from London its location is not ideal. While the Eurotunnel terminal in the UK is rather constrained, the Calais terminal has plenty of space and maintenance facilities for locomotives – but that is 166km from London.
Theoretically empty stock could be moved from Calais to Lille and services start there. But that is still 105km. Déliverance in Lille, a freight yard to the west of the city, has plenty of space and is 16km from Lille Europe station on existing infrastructure. Ronchin – a site used by SNCF track maintenance trains currently – is only 3km from Lille Europe, but a small connecting curve is missing, preventing easy access from there to Lille Europe.
And then to greater Paris. This is really complicated. The most obvious site to develop would be Drancy (12km from Paris Nord), but the track layout between La Plaine and Paris Nord is far from ideal, and getting paths could be complicated. Vaires-Torcy has plenty of space, but getting to Paris Nord from there is even more complicated – it is either a horrible 112km around the outer edge of Paris, or 34km but needing reversing at Pierrefitte. Drancy could be an alternative for the reversing manoeuvre. Villeneuve-Saint-Georges likewise has plenty of space, but is 104km from Paris Nord. These three sites could be of particular interest to Trenitalia as Paris Gare de Lyon can also be reached from these (41km from Drancy, 37km from Vaires-Torcy and 13km from Villeneuve-Saint-Georges).
Were a site needed only for Channel Tunnel services then there are a few more options. Persan-Beaumont (47km) and Creil (50km) would have adequate space, although routes to Paris Nord from either are on comparatively full tracks. Longueil-Sainte-Marie (67km) is furthest, but is on LGV Nord, but a tiny connecting curve onto the high speed line is missing, making this one complicated.
So what am I missing? What have I overlooked? What other sites should be considered? I will updated this post and the map accordingly.
And as I am travelling to the UK for an event in Ashford on 8th January 2026 that will be an opportunity to have a closer look at some of the UK sites.
[Additional Explainer – 18.12.2025, 19:15]
Seems that some things I thought were obvious here are not obvious to all. All of this would apply to whatever company needs maintenance space. It would not apply to Eurostar – they have adequate maintenance capacity already. It would apply to Trenitalia if they decide to carry on with the idea to run to London, and do this somehow in conjunction with their French routes. It might apply to Virgin Trains if something goes wrong with efforts to negotiate with Eurostar about access to Temple Mills. It would apply to Gemini and Evolyn if they can manage to find finance for rolling stock and develop their ideas further. It probably will not apply to SBB as they would likely do maintenance in Switzerland, but we will only know more about SBB’s plans in 2026.
[Ongoing updates]
Suggestions so far: Goussainville from Suzanne Veerman, north of Temple Mills from Andy Allan, Singlewell from AlisonW, Valenton from PGLux. I will assess all of these! Keep them coming!
Images used in this post
Train Photos
Dollands Moor
July 18, 2021
CC BY-SA 2.0
Barking is probably reserved for future development of Barking Riverside. And, there are 2 massive pylons immeadiately south of the existing tracks
How much an operator based in Calais would pay just to cross the Eurotunnel empty and reach London to start their services…
You’d only pay the per-train charge and not the per-passenger charge. And Getlink is very keen on more traffic through the tunnel. But 160km empty to London? Or 100km empty to Lille? Can’t see it happening, honestly.
Barking Part2: also, you can’t access it from STP as the line hasn’t come out of the tunnel yet!
co-ords 51.526977,0.137458
Yes, and there is an ongoing discussion about what the site might be used for, for rail freight as well. It looks… complex! (I might head and have a look around there on 9th or 12th January)
For Ashford, you wrote “At Ashford there is the maintenance facility for Hitachi’s class 395 trains, but access to the depot is third rail electrified only at the moment, and the site is very constrained. Ex railway land south east of Ashford station also merits a further look, although Ashford is 91km from London.”.
It’s a bit more complex than that as there are sites both south east of the station and south west. Both would require crossing the 3rd rail tracks.
1) South west is the location of the former Chart Leacon depot – demolished in 2014(?). A little of this will be used for sidings but I can’t find any mention of the rest of it. I couldn’t find anything more recent than 2021 about the Chart Leacon development – https://land-referencing.co.uk/media/2021/05/chart-leacon-pressing-ahead/. The lack of any mention makes me wonder if it’s been cancelled.
2) South east is the former Ashford works, last used by Balfour Beatty for crane repairs. It’s being redeveloped for non-rail use. Here https://www.ashford.gov.uk/news/latest-news/making-the-newtown-works-site-work-for-the-borough-s-interests/ is a description of the former works development.
Thanks! Was not aware of that one to the south west. That merits a further look, although the site looks rather short on the map. And the Balfour Beatty one I will eliminate then.
Off the cuff, naive open question: Could there be any value in making a depot site (or a pair of depot sites on each side of the channel) into a passenger station with a big parking lot, and running some trains ex depot in revenue service. Or adding a new station nearby, solely for trains originating/terminating at depot, or as a deliberately stopping service for budget travellers. Maybe as a short hop depot to depot service, with car hire available at each end, or a longer journey to a city centre destination.
The addressable market would be travellers from the hinterlands that would otherwise drive to an airport rather than travelling into a city centre to catch the cross channel train.
And could that approach commercially unlock greenfield sites?
I imagine that’s all a no, but I haven’t worked out the reasons why not to conclusively eliminate from consideration
That is basically Ebbsfleet – Northfleet!
Ebbsfleet is a massive carpark in the middle of nowhere, with solid road connections (also onto the M25). And then the depot location I list in the blog post is a couple of km from the station.
The only possible headache with these sorts of things is that starting or ending trains at interim stations, especially Ebbsfleet (or Ashford, theoretically) is that the Channel Tunnel parts of the stations are not staffed currently because Eurostar does not want to stop at these stations. But I think that problem is surmountable.