HS2 Old Oak Common

[LNW-GW Joint]

Interesting - do you know which bits are far further behind?

There are still no railway contracts let (slab track installation, OHLE, signalling), because of sky-high costs.
That's already nearly two years late on the original schedule.
Bromford tunnel is also over a year late on its original schedule.

 

[absolutelymilk]

There are still no railway contracts let (slab track installation, OHLE, signalling), because of sky-high costs.
That's already nearly two years late on the original schedule.
Bromford tunnel is also over a year late on its original schedule.

Oh dear! I was (probably naïvely) assuming that the railway contracts would be let soon and start early 2026, and take around 2 years but perhaps that was optimistic!

 

[Snow1964]

There’s a decent chance Old Oak might be completed this side of 2030, although with the construction industry who knows. But it’s there are other bits of the line far further behind.

The station box is completely dug, the slab floor has been under construction, and the most recent update suggested now building upwards in the hole, constructing platforms, then moving onto intermediate levels and the steelwork into 2026.

Separately this update (link) says the sprayed concrete lining tunnel linking Victoria box to Old Oak Common (the tapered tunnels with the junctions) will take until late 2025. By then the Northolt TBMs will have finished so that rather suggests fit out and track laying could commence from early 2026 (although doesn't mean it will start then)

Update: Old Oak Common tunnel construction works

The Old Oak Common tunnel is being constructed using the sprayed concrete lining (SCL) method. This involves excavating a pilot tunnel, enlarging it to its final size, and then installing the concrete lining and waterproofing. We last wrote to you in June 2026 to provide an update on the tunnel...

www.hs2.org.uk

Link also has handy map showing track layout (blue lines) at west end of Old Oak station

== Doublepost prevention - post automatically merged: 10 Nov 2024 ==

Just discovered there is published list of dates of work, including the installation of overhead wires and structures in Old Oak Common -Acton area

https://assets.hs2.org.uk/wp-content/uploads/2024/11/OOC-Central-section.pdf

 

[Bald Rick]

Separately this update (link) says the sprayed concrete lining tunnel linking Victoria box to Old Oak Common (the tapered tunnels with the junctions) will take until late 2025. By then the Northolt TBMs will have finished so that rather suggests fit out and track laying could commence from early 2026 (although doesn't mean it will start then)

Bear in mind there is a lot of work to get from finishing boring of the tunnels until it is ready to accept track. Not least that track doesnt go on the invert of the tuunel…

 

[Meerkat]

Bear in mind there is a lot of work to get from finishing boring of the tunnels until it is ready to accept track. Not least that track doesnt go on the invert of the tuunel…

Is the invert just filled with an obscene volume of concrete or are there channels left in it?

 

[Nottingham59]

Is the invert just filled with an obscene volume of concrete or are there channels left in it?

https://assets.hs2.org.uk/wp-content/uploads/2020/10/01150931/N30-Tunnel-Construction-Methodology.pdf

shows a typical cross section of an HS2 bored tunnel

 

[Backroom_boy]

So could possibly the next generation of tunnel boring machines lay the invert fill
along with the tunnel segments?

 

[stuu]

So could possibly the next generation of tunnel boring machines lay the invert fill
along with the tunnel segments?

Presumably there is a reason why it isn't just incorporated into the tunnel segments in the first place?

 

[SynthD]

There’s a lot of tbm activity behind the part that places the tunnel rings. There is a temporary rail line for various tasks, including re supplying the machine with rings and people.

 

[Nottingham59]

So could possibly the next generation of tunnel boring machines lay the invert fill
along with the tunnel segments?

I doubt if that would be feasible, as they need to be able to place the ring segements at different positions around the circumference to manage curvature (vertical and horizontal) in the tunnel and to correct any deviations to get back onto the exact alignment.

Maybe the slab track base is shaped with curved lower surface so that it can be installed into the bare tunnel?

 

[Bald Rick]

I doubt if that would be feasible, as they need to be able to place the ring segements at different positions around the circumference to manage curvature (vertical and horizontal) in the tunnel and to correct any deviations to get back onto the exact alignment.

Maybe the slab track base is shaped with curved lower surface so that it can be installed into the bare tunnel?

The base is usually cast in situ. Thats certainly how it was done for HS1 and Crossrail.
Engineering Tolerance for the embedded track is (obviously) much tighter than for the tunnel lining.

 

[Snow1964]

So could possibly the next generation of tunnel boring machines lay the invert fill
along with the tunnel segments?

No the TBM has grout mixing unit to inject it into voids, but no bulk concrete process, and they currently use bare segment bases as a roadway for TBM segment supply and muck removal.

Another reason is although the TBM follows a planned course, uneven hardness of the rock means the bored hole isn't always a perfect smooth circle, even if a segment ring was 1mm out of alignment (which is nothing on a 200mm thick segment) would you want a 1mm kink in track bed.

There are separate machines that follow and lay the base slab, casting it in situ. This happens after other work is done (holes drilled for brackets, cables etc, and cross tunnels dug etc). If you did this before the floor slab would have problems of fine cement dust from drilling for these, which trains would blow all over the place (and presumably into own ventilation grills). Normally clean as much dust up as you can and any residual gets covered by the floor slab casting.

 

[Meerkat]

https://assets.hs2.org.uk/wp-content/uploads/2020/10/01150931/N30-Tunnel-Construction-Methodology.pdf

shows a typical cross section of an HS2 bored tunnel

View attachment 169076

That looks like a lot of concrete!

 

[Nottingham59]

That looks like a lot of concrete!

The thickness of concrete in the running tunnel itself doesn't look excessive to my untrained eye. Perhaps a couple of feet, which would include the depth of the slab track base?

The cross passages seem to have a lot, but they're only every 375m I think

 

[JCO1406]

Question, why does the GWR/Elizabeth line part of OOC have 8 platforms? 2 for the fast line and 3 for slow/lizzy line (1 being a bay platform). What would the other 3 be used for?

 

[Nottingham59]

Question, why does the GWR/Elizabeth line part of OOC have 8 platforms? 2 for the fast line and 3 for slow/lizzy line (1 being a bay platform). What would the other 3 be used for?

Everything passing will stop at OOC so each of the four tracks splits into two, giving eight platforms altogether

 

[Meerkat]

The thickness of concrete in the running tunnel itself doesn't look excessive to my untrained eye. Perhaps a couple of feet, which would include the depth of the slab track base?

The cross passages seem to have a lot, but they're only every 375m I think

I was just thinking in absolute terms, over miles of tunnel.

 

[swt_passenger]

Question, why does the GWR/Elizabeth line part of OOC have 8 platforms? 2 for the fast line and 3 for slow/lizzy line (1 being a bay platform). What would the other 3 be used for?

Can I point you to the track layout? None of the 4 relief side platforms are expected to operate as bays, terminating trains will run through to one of the 3 reversing sidings. As in the earlier reply by @Nottingham59 the main lines require 4 platforms for all services to call as planned.

HS2 Old Oak Common

Don't know, not too long, but wasn't that an emergency replacement not a planned job? It was very much a planned job: https://scotlandsrailway.com/projects/ravenscraig

www.railforums.co.uk

 

[corfield]

Can I point you to the track layout? None of the 4 relief side platforms are expected to operate as bays, terminating trains will run through to one of the 3 reversing sidings. As in the earlier reply by @Nottingham59 the main lines require 4 platforms for all services to call as planned.

HS2 Old Oak Common

Don't know, not too long, but wasn't that an emergency replacement not a planned job? It was very much a planned job: https://scotlandsrailway.com/projects/ravenscraig

www.railforums.co.uk

I played a train simulation game a few years back and 2 platforms per direction was essential for throughput on a busy line as it seemed to offset the time lost in dwell as trains behind continued forwards. I wasnt sure how decel/accel affected things but it was also noticeable that the faster the train, the more essential the 2nd platform was.

Adding more, eg. to 3 per track acheived nothing extra, I think because the numbers (in that game at least) meant you were now limited by what could fit on the track end to end, and 2 platforms had accounted fully for the “blockage” effect of stopping.

I’ve always wondered how this maps to reality, noting as you say OOC has 4 for reliefs and 4 for mains so that is 2 platforms per track in effect.

The tube on the other hand runs at the highest frequency and has (bar a few exceptions perhaps?) 1 platform per track.

I imagine there is theory on all of this

 

[SynthD]

That theory may also include how much track loop is needed before/after each platform, to not affect the train using the other platform or face. The diagram*, if it’s actually indicative of this detail, has some extra track per platform.

The diagram* has two To Be Confirmeds, one for a connection to the NNML towards Greenford, and one from the relief to the main, in the west. Are either of these confirmed yet?

*from swt_passengers post https://www.railforums.co.uk/threads/hs2-old-oak-common.240858/page-4#post-6187912

 

[DynamicSpirit]

I played a train simulation game a few years back and 2 platforms per direction was essential for throughput on a busy line as it seemed to offset the time lost in dwell as trains behind continued forwards. I wasnt sure how decel/accel affected things but it was also noticeable that the faster the train, the more essential the 2nd platform was.

Adding more, eg. to 3 per track acheived nothing extra, I think because the numbers (in that game at least) meant you were now limited by what could fit on the track end to end, and 2 platforms had accounted fully for the “blockage” effect of stopping.

I would expect that, very roughly speaking, two platforms per direction/track becomes useful if the dwell time is comparable to the interval between trains - because that's the point where a train could block the following train if there's only one platform. By the same logic, three platforms per direction is only going to become useful if the dwell time is something like twice the interval between trains - the point where a train could be in a platform long enough to block both the next TWO trains. That's not really plausible for GWR - and off the top of my head I can't think of anywhere on the UK network where that is likely - other than at terminus stations where trains dwell for a very long time. (Although ironically OOC is being built with three platforms per direction on the HS2 lines)

Three platforms per direction also has the disadvantage that it makes if impossible for one island platform to serve all trains heading in a given direction - which would be quite a disadvantage at Old Oak Common.

 

[Nottingham59]

I would expect that, very roughly speaking, two platforms per direction becomes useful if the dwell time is comparable to the interval between trains - because that's the point where a train could block the following train if there's only one platform. By the same logic, three platforms per direction is only going to become useful if the dwell time is something like twice the interval between trains - the point where a train could be in a platform long enough to block both the next TWO trains. That's not really plausible for GWR - and off the top of my head I can't think of anywhere on the UK network where that is likely - other than at terminus stations where trains dwell for a very long time.

Modelling suggested that between one third and one half of HS2 passengers would detrain at Old Oak Common, heading for Crossrail, Heathrow and the Overground.

I'd imagine that a similar proportion might do the same from long-distance GWML services, which suggests relatively long dwell times.

 

[Bald Rick]

I would expect that, very roughly speaking, two platforms per direction/track becomes useful if the dwell time is comparable to the interval between trains

Strictly speaking, it is when the potential interval between trains is less than the dwell + platform reoccupation time + a perfromance margin.

In HS2s case, the length of trains and linepseed means the platform reoccupation time could be will be (at a guess) in the region of three minutes; add in 2 minute dwells and you are already in need of two each way before a performance margin.

 

[Snow1964]

Modelling suggested that between one third and one half of HS2 passengers would detrain at Old Oak Common, heading for Crossrail, Heathrow and the Overground.

I'd imagine that a similar proportion might do the same from long-distance GWML services, which suggests relatively long dwell times.

Does that mean the intention is to deliberately run trains half to third empty east of Old Oak Common. run 3 when 2 would be sufficient for the load.

Might be operational convenience, but not exactly fair when there are other GWR routes with crowding.

 

[DynamicSpirit]

Modelling suggested that between one third and one half of HS2 passengers would detrain at Old Oak Common, heading for Crossrail, Heathrow and the Overground.

I'd imagine that a similar proportion might do the same from long-distance GWML services, which suggests relatively long dwell times.

The difference would be that, when OOC was being designed, HS2 was, if I recall correctly, being planned to have about 18tph. I believe GWR on the other hand has about 6 long distance tph - so on average, a train every 10 minutes.

Even allowing for future growth and for the intervals between trains not being completely regular, you'd have to have extraordinarily long dwell times before 2 platforms per direction becomes inadequate for that level of service

 

[Snow1964]

I played a train simulation game a few years back and 2 platforms per direction was essential for throughput on a busy line as it seemed to offset the time lost in dwell as trains behind continued forwards. I wasnt sure how decel/accel affected things but it was also noticeable that the faster the train, the more essential the 2nd platform was.

I am sure it has also been said, that it works best when serving both sides of an island, and the turnout speeds are high enough to allow normal (not speed restricted) braking and accelerating without any slowdown approach control, or slow crawls when leaving the platform

 

[SynthD]

Does that mean the intention is to deliberately run trains half to third empty east of Old Oak Common. run 3 when 2 would be sufficient for the load.

Might be operational convenience, but not exactly fair when there are other GWR routes with crowding.

People from HS2 would also go west, eg to Heathrow. It might be nice if the peak trains leave Paddington with standing passengers, and leave OOC with everyone in a seat.

 

[MasterPhillips]

The difference would be that, when OOC was being designed, HS2 was, if I recall correctly, being planned to have about 18tph. I believe GWR on the other hand has about 6 long distance tph - so on average, a train every 10 minutes.

6TPH? more like 11 and that’s off peak. Not to mention 4TPH Heathrow express. So that’s 15 on the Mains alone at OOC.

 

[DynamicSpirit]

6TPH? more like 11 and that’s off peak. Not to mention 4TPH Heathrow express. So that’s 15 on the Mains alone at OOC.

But only 6tph of those are long distance ones (2*Bristol, 2* South Wales, 1* Cheltenham, 1*Devon/Cornwall). The others are more like local or regional commuter trains (To be fair I guess some of those commuter trains use the same tracks as the long distance trains and will use the same platforms at OOC so perhaps should be counted, but they are also formed of commuter stock that's designed for ease of entry/exit so are likely to have shorter dwell times).

 

[MasterPhillips]

But only 6tph of those are long distance ones (2*Bristol, 2* South Wales, 1* Cheltenham, 1*Devon/Cornwall). The others are more like local or regional commuter trains (To be fair I guess some of those commuter trains use the same tracks as the long distance trains and will use the same platforms at OOC so perhaps should be counted, but they are also formed of commuter stock that's designed for ease of entry/exit so are likely to have shorter dwell times).

Also 2TPH to Oxford/Worcester also formed of IET’s. But yes good point about the dwell times, I hadn’t thought of that. Ultimately 2 island platforms is the right amount and will definitely be needed at OOC.