Should the HS2-HS1 tunnel under London be uncancelled?

[The Ham]

@John Jefkins

Still awaiting a response to the point about the 65 minutes claim:

I come back to the point that I've asked you about several times now.

How is St Pancras an hour better if a lot of the delay at Euston is too ensure that they make their booked service?

Either:

- travel from Euston had a delay of (say) 100 minutes with (say) 35 minutes of that being to ensure that the passenger gets there on time for their connection with the remaining 65 minutes required to make a connection which most would be able to do in under 30 minutes and quite a few in around 20 minutes and the delay at St Pancras is 35 minutes

OR

- the total delay at Euston is 65 minutes, of which (say) 35 minutes is to ensure that a passenger makes their connection, however there's also a (say) 35 minutes delay at St Pancras (but almost zero for changing platforms) to ensure that they make their connection

Bottom line of a passenger is wanting to get the 12:25 HS2 service from the central London station what time do they want to be at that station?

At Euston is that 11:20 or some other time earlier than that (say 10:45)?

At St Pancras is that 12:25 or some other time earlier than that (say 11:50)?

Edit:

Predicted passenger numbers for HS2 were not maximised, so even without making changes, there's a fair chance that passenger numbers could be higher than the predictions.

Reattached, so you don't miss it.

 

[Benjwri]

2 Eurostar trains/hr (only 4 movements)

Eurostar capacity is limited by space at St Pancras and work is already being undertaken to fix that. You cannot assume 2tph, effectively capping international trains to the UK permanently.

I am also still awaiting your answer to if you plan to fit ETCS to HS1, and the javelins, which while a long term plan is not funded.

TfL from saving £1 billion on not needing to upgrade Euston tube

Fake money. TfL doesn’t have a spare billion to spend on their own issues. The billion you quote would’ve been part of Crossrail 2, and if that does go ahead the station will still be built to the same cost. (and funded likely in large by the government)

 

[yoyothehobo]

1) As we're only talking about the first 1,000 metres in and out of St Pancras instead of Euston, your HS2 speed of 35 mph for the 1st 700 metres compares very well.
Your speeds along the NNL are incorrect as TfL trains aren't limited to 25mph along the same stretch. Its 35mph around the curves and up to 40mph elsewhere.
That means the 35mph out of Euston is EXACTLY THE SAME speed as 35mph around the Camden Market curve and 40mph along to the St Pancras throat.

You might want to tell Network Rail that their section appendix is wrong if you are adamant its greater than 25mph at any point between St Pancras and Primrose Hill.

 

[tomuk]

1) As we're only talking about the first 1,000 metres in and out of St Pancras instead of Euston, your HS2 speed of 35 mph for the 1st 700 metres compares very well.
Your speeds along the NNL are incorrect as TfL trains aren't limited to 25mph along the same stretch. Its 35mph around the curves and up to 40mph elsewhere.

The key thing you are ignoring is that people are not starting their journeys at either Euston or St Pancras. Its that better interchange speed that cuts far more off their total journey time (up to an hour) compared with more like 2 minutes extra journey time into or out of St Pancras compared with Euston.

And virtually none of them are starting their journeys from Stratford International, Ebbsfleet or Ashford.

 

[Bald Rick]

very quickly as your inability to answer questions asked or see things from a different point of view is really getting tiresome:

1) As we're only talking about the first 1,000 metres in and out of St Pancras instead of Euston, your HS2 speed of 35 mph for the 1st 700 metres compares very well.

You might be talking about the first 1000 metres, I‘m talking about the distance from the bufferstops to Primrose Hill.

Your speeds along the NNL are incorrect as TfL trains aren't limited to 25mph along the same stretch. Its 35mph around the curves and up to 40mph elsewhere.

Again, it is the NLL, not NNL. TfL trains are limited to the speeds I quoted, this comes directly from the Sectional Appendix (which another poster has helpfully attached, see just above this post). Don’t forget that the length of the train is important with respect to speed restrictions.

That means the 35mph out of Euston is EXACTLY THE SAME speed as 35mph around the Camden Market curve and 40mph along to the St Pancras throat.

It isn’t, see the Sectional Appendix. AND NO NEED TO SHOUT.

The key thing you are ignoring is that people are not starting their journeys at either Euston or St Pancras.

I think you are ignoring it to be honest. The majority of people that will use HS2 services are travelling from places where it makes no difference at all whether they go to St Pancras or Euston, and for many OOC will be easier. The concept that it equates to a 65 minute journey time reduction for a large percentage of passengers is fantasy. I‘ve asked three times, and now a fourth - please give some specific examples of journeys where there would be a 65 minute journey time saving through your proposal, along with the number of people who currently make that journey each year.

The rest of your post will just have to wait.

 

[John Jefkins]

@John Jefkins

Still awaiting a response to the point about the 65 minutes claim:



Reattached, so you don't miss it.

The 65 minutes is an “interchange penalty” used in the PDFH to predict potential passenger numbers which grow by 47% if you remove the double interchange of needing to travel between the Thameslink or Javelin or Eurostar or Kings X trains from say Cambridge to reach HS2 trains leaving Euston instead of St Pancras.

The key thing thus is the uplift of revenue from getting 15 million extra passengers ( ie about 50% of 70% =35% of say 42 million ).
15 million people paying over £60 for fares in the 2040s brings in an extra £1 billion a year. But we’d also get extra people travelling to reach GWR trains to Bristol or Heathrow trains etc from say Cambridge or Luton or Peterborough or North London via St Pancras who cannot so easily reach Euston and who would use these 12 trains per hour to an Old Oak Common connection instead of using the slower and much more overcrowded Elizabeth line.

very quickly as your inability to answer questions asked or see things from a different point of view is really getting tiresome:





You might be talking about the first 1000 metres, I‘m talking about the distance from the bufferstops to Primrose Hill.




Again, it is the NLL, not NNL. TfL trains are limited to the speeds I quoted, this comes directly from the Sectional Appendix (which another poster has helpfully attached, see just above this post). Don’t forget that the length of the train is important with respect to speed restrictions.



It isn’t, see the Sectional Appendix. AND NO NEED TO SHOUT.




I think you are ignoring it to be honest. The majority of people that will use HS2 services are travelling from places where it makes no difference at all whether they go to St Pancras or Euston, and for many OOC will be easier. The concept that it equates to a 65 minute journey time reduction for a large percentage of passengers is fantasy. I‘ve asked three times, and now a fourth - please give some specific examples of journeys where there would be a 65 minute journey time saving through your proposal, along with the number of people who currently make that journey each year.

The rest of your post will just have to wait.

According to experts who do passenger forecast calculations these interchange penalties DO make the most difference on these LONGER journeys where people are carrying luggage or are doing it for leisure and taking children or buggies and where a car journey is an attractive alternative. If you make them walk from St Pancras to Euston you reduce passenger numbers.

Yes it’s the NLL not NNL but as both routes start off with the same 35mph it’s actually only about 700 metres extra at the same speed.

I too will create a diagram to show how it’s actually only about 90 seconds extra time to save about an hour of interchange penalty.

View attachment 177156


And virtually none of them are starting their journeys from Stratford International, Ebbsfleet or Ashford.

As this diagram shows the speeds are 35-40 mph for the straight route through towards Primrose Hill. I’ve assumed 35mph for the 200m long trains these would almost all be - due to the Camden Market curves. This equates to exactly the same 35 mph speed on your data for the exit from Euston.

We are thus only extending that 35 mph by about 700 extra metres to add just over a minute to journey times whilst saving an interchange penalty of about an hour. ( I’ll work exact data out and come back to you).

Case proven. You really cannot claim that using the NLL adds significant time. It doesn’t as these trains would be running slowly into or out of platforms and we’d be packing up to 20 trains per hour along this cross-London route.

Think of Thameslink across London. This isn’t high speed rail.
It’s high capacity rail.

 

[Class15]

@John Jefkins

Does your plan involve completely taking over the Primrose Hill line for HighSpeed trains and making them unavailable for freight use? If so, when the line through Gospel Oak is shut (e.g. for engineering works) where are you sending the freights? When the line through Gospel Oak is open, then the Primrose Hill line is not a huge loss for the freight sector, but if not you’re in trouble.

 

[yoyothehobo]

The 65 minutes is an “interchange penalty” used in the PDFH to predict potential passenger numbers which grow by 47% if you remove the double interchange of needing to travel between the Thameslink or Javelin or Eurostar or Kings X trains from say Cambridge to reach HS2 trains leaving Euston instead of St Pancras.

The key thing thus is the uplift of revenue from getting 15 million extra passengers ( ie about 50% of 70% =35% of say 42 million ).
15 million people paying over £60 for fares in the 2040s brings in an extra £1 billion a year. But we’d also get extra people travelling to reach GWR trains to Bristol or Heathrow trains etc from say Cambridge or Luton or Peterborough or North London via St Pancras who cannot so easily reach Euston and who would use these 12 trains per hour to an Old Oak Common connection instead of using the slower and much more overcrowded Elizabeth line.


According to experts who do passenger forecast calculations these interchange penalties DO make the most difference on these LONGER journeys where people are carrying luggage or are doing it for leisure and taking children or buggies and where a car journey is an attractive alternative. If you make them walk from St Pancras to Euston you reduce passenger numbers.

Yes it’s the NLL not NNL but as both routes start off with the same 35mph it’s actually only about 700 metres extra at the same speed.

I too will create a diagram to show how it’s actually only about 90 seconds extra time to save about an hour of interchange penalty.


As this diagram shows the speeds are 35-40 mph for the straight route through towards Primrose Hill. I’ve assumed 35mph for the 200m long trains these would almost all be - due to the Camden Market curves. This equates to exactly the same 35 mph speed on your data for the exit from Euston.

We are thus only extending that 35 mph by about 700 extra metres to add just over a minute to journey times whilst saving an interchange penalty of about an hour. ( I’ll work exact data out and come back to you).

Case proven. You really cannot claim that using the NLL adds significant time. It doesn’t as these trains would be running slowly into or out of platforms and we’d be packing up to 20 trains per hour along this cross-London route.

Think of Thameslink across London. This isn’t high speed rail.
It’s high capacity rail.

The diagram shows that the route towards Primrose Hill is 15-20mph not 35mph. You can repeat it many times, but it will never be true. The 35mph/40mph speeds are to the East of the St Pancras Junction.

 

[Benjwri]

Think of Thameslink across London. This isn’t high speed rail.
It’s high capacity rail.

Which is exactly the issue. You’re talking about interchange penalty but by running the train slow into actual London you’re creating a penalty for anyone not going to the St Pancras area.

Also worth noting such a penalty is partially present because people want to get there early for their advance tickets. That won’t go away with a shorter interchange.

 

[30907]

The 65 minutes is an “interchange penalty” used in the PDFH to predict potential passenger numbers which grow by 47% if you remove the double interchange of needing to travel between the Thameslink or Javelin or Eurostar or Kings X trains from say Cambridge to reach HS2 trains leaving Euston instead of St Pancras.

But it isn't a double interchange - only two trains are involved. It is certainly a longer distance, rather more than the equivalent at Waterloo or Manchester P but on a par with Glasgow, but only two trains are involved.

But we’d also get extra people travelling to reach GWR trains to Bristol or Heathrow trains etc from say Cambridge or Luton or Peterborough or North London via St Pancras who cannot so easily reach Euston and who would use these 12 trains per hour to an Old Oak Common connection instead of using the slower and much more overcrowded Elizabeth line.

So: passengers heading from TL to GWR change at STP instead of XFD (Farringdon). Allowing for the increased interchange time at STP vs ZFD (I'm talking actual physical distance, not NRT allowances), how much time is saved by this?
(obviously interchange TL-HS2 would be improved by some minutes)

. If you make them walk from St Pancras to Euston you reduce passenger numbers.

Agreed, but not by the amount you are claiming - on a typical journey of 2hr the difference is at most 10min, so below 10%.

 

[zwk500]

But it isn't a double interchange - only two trains are involved. It is certainly a longer distance, rather more than the equivalent at Waterloo or Manchester P but on a par with Glasgow, but only two trains are involved.

PDFH recommends that interchanges involving 2 stations (e.g. Euston to St Pancras) are treated as two separate interchanges, and the walking time is included in the journey time. I think that's personally a little pessimistic - London transfers should probably be 1.5 changes or similar as a tube/walking transfer is not the same as a separate change of train.

There's been a lot of talk of the interchange penalties in PDFH but relatively little discussion of the elasticities or actual method to forecast demand. Important to note that those interchange penalties are applied onto Generalised Journey Time schedules, which for long-distance journeys involving multiple changes of trains can be somewhat finickity to work out, and PDFH is also primarily for domestic journeys (all the reasearch that feeds it is based on UK domestic journeys) and I'd argue that the interchange penalty is largely absorbed into the security penalty when considering international rail from London. There's also the matter that you could make equally convincing arugments for St Pancras to be considered anywhere from 2 to 4 separate stations for PDFH purposes as you could for the entire King's Cross/St Pancras complex to be considered 1 interchange for PDFH purposes.

 

[Bald Rick]

But we’d also get extra people travelling to reach GWR trains to Bristol or Heathrow trains etc from say Cambridge or Luton or Peterborough or North London via St Pancras who cannot so easily reach Euston and who would use these 12 trains per hour to an Old Oak Common connection instead of using the slower and much more overcrowded Elizabeth line.

I think that is very unlikely. Anyone from Luton or Cambridge going to Heathrow will use Thameslink to Farringdon, and the Elizabeth Line to Heathrow. From arriving at St Pancras to arriving at Heathrow is typically 50 minutes including the change at Farringdon, although I have personally done it in 45. Going via St Pancras via your imaginary link and OOC is two changes, and couldn’t possibly be less than 45 minutes, but more likely 50-55. Why would people make it more difficult for themselves for little or no gain?

As this diagram shows the speeds are 35-40 mph for the straight route through towards Primrose Hill. I’ve assumed 35mph for the 200m long trains these would almost all be - due to the Camden Market curves. This equates to exactly the same 35 mph speed on your data for the exit from Euston.

Perhaps I can help explain how to interpret the sectional appendix, because you have not got this right.

The HS1 sectional appendix is in km/h. For trains heading from the international platforms towards the connection with the NLL it is 40km/h the whole way. (As stated previsouly, there is no connection from the domestic platfroms to the NLL link, but it is reasonable to assume that if there were, it would be 40km/h max, but very likely lower). 40km/h is approximately 25mph.

Moving on to the NR East Anglia Sectional Appendix, as pasted into the thread by others above, the connection from the HS1 Silo Curve leads on to the North London Incline, and the connection from that line onto the NLL is 15mph, as marked by the “15”; this is at Camden Central Junction, 4miles and 64 chains as shown in the table to the left of the diagram. At that point the NLL is indeed 35/40mph (the 40 mph applying to passenger services), however that speed applies only until 4miles 71 chains, ie 7 chains (141 metres) further on, where the asterisk (*) on the line concerned denotes a change of speed. From here through Camden Road station it is 20mph, as shown by the “20” on each line through Camden Road station. As any Javelin or HS2 train that would use this imaginary route is longer (or very early so) than 141 metres, one part of it would always be on the 20mph restriction or the 15mph restriction, and therefore unable to make use of the 40mph.

Heading towards Primrose Hill the line remains at 20mph, until the next asterisk at 5m18ch (ie 27 chains - 543 metres - further on from the start of the 20mph restriction). It is then 15 mph towards Primrose Hill, as denoted by the “15” on each line beyond the asterisks. A 400metre HS2 train would have only 143metres of track where it is exclusively in the 20mph stretch, and I doubt if it would be realsitic to expect drivers to accelerate to 20mph before almost immediately braking back for 15mph; they would just do 15mph all the way. A 200 metre train or Javelin might be able to accelerate - briefly - up to 20mph, but I rather suspect that the standards on drivability would kick in and it will be 15mph the whole way.

Therefore best case for a 400metre HS2 train would be to accelerate to 25mph from the bufferstops to the braking point for reducing to 15mph at the North London line junction (almost exactly a mile from the buffer stops) and then 15mph from there for the 3/4 mile or so to your proposed portal at Primrose Hill. Only when the back of the train - a further 400 metres at 15mph - is on your proposed new line can it then commence acceleration up from 15mph. Feel free to do the calculations using expected braking and acceleration rates, not forgetting to allow for the length of the train.

I can’t see how it can be less than 6 minutes to the rear of the train clearing Primrose Hill (still doing 15mph) and probably more, by which time an equivalent HS2 train from Euston is already about to depart OOC. For example a mile at 15mph from the front of train passing Camden Road Central Junction to back of train clearing the existing line at Primrose Hill takes 4 minutes alone. I’ll be happy to stand corrected if I am proven incorrect through your calculations.

Case proven. You really cannot claim that using the NLL adds significant time.

Let’s see what your calculations show, based on the actual linespeeds, and not those that you seem to think it is, before you start saying something is proven.

 

[Nym]

I do seem to despair at these blue sky thinking questions, because I fully believe the OP is an “expert” working on franchise bids, evidenced by the quality of these bids I’ve seen. Unfortunately my real life role needs me to deal with similar levels of fixation.

It honestly does feel like either someone trying to crowd source a solution from actual experts and is very much symptomatic of well, the industry we work in.

Fortunately I don’t need to deal with this for HS2 or Infrastucture, but these behaviours are rife in the rail industry, and are always quick to blame “consultants” when they in themselves are not an intelligent client and do not listen to reason from consultants saying that what they’re asking for is wrong. Hence, consultant delivers to spec, needs to be redone when goes to approvals, new consultant time.

 

[Backroom_boy]

It honestly does feel like either someone trying to crowd source a solution from actual experts and is very much symptomatic of well, the industry we work in.

I think you are being generous

 

[The Ham]

The 65 minutes is an “interchange penalty” used in the PDFH to predict potential passenger numbers which grow by 47% if you remove the double interchange of needing to travel between the Thameslink or Javelin or Eurostar or Kings X trains from say Cambridge to reach HS2 trains leaving Euston instead of St Pancras.

The key thing thus is the uplift of revenue from getting 15 million extra passengers ( ie about 50% of 70% =35% of say 42 million ).
15 million people paying over £60 for fares in the 2040s brings in an extra £1 billion a year. But we’d also get extra people travelling to reach GWR trains to Bristol or Heathrow trains etc from say Cambridge or Luton or Peterborough or North London via St Pancras who cannot so easily reach Euston and who would use these 12 trains per hour to an Old Oak Common connection instead of using the slower and much more overcrowded Elizabeth line.

I understand that, what I'm asking is is the 65 minutes the total time penalty or is it the difference in time penalties between an in station (St Pancras) interchange and external interchange (between Euston and St Pancras)?

Is the Euston to St Pancras interchange 65 minutes?

Is the Euston St Pancras interchange (say) 90 minutes and the St Pancras interchange (say) 15 minutes giving a difference of 65 minutes?

The later clearly doesn't allow any extra time for making a connection to a long distance service.

Is the Euston St Pancras interchange (say) 120 minutes and the St Pancras interchange (say) 55 minutes giving a difference of 65 minutes?

That then allows for a interchange for a long distance train.

Which of the three is it?

As two don't give the answer you're after.

 

[Bald Rick]

Got a bit more time to go back to some other points.

2) There is absolutely no land take (apart from the 1st floor of one dentists surgery at Kentish Town Rd where a wider viaduct is needed to cross it).

There will be land take. All the reports done on this - at very high level - say there will be; this includes the Arup report you are relying on. Read pages 50 to 77 (Section 7). It mentions demolition of residential, commercial and industrial premises, even for the simplest option (which is less work than what you propose.) The segregated solution, which is similar (but still rather less) to what you propose, involves land take of parts of the various Camden Markets. Good luck with that! That’s just the permanent land take; during construction there will be much more, including the severance of roads throughout the area. This is not me saying this, it is the Arup report.

3) As said earlier the single freight train per hour using the route to Primrose Hill would continue to use one of the 8 slots in-between TfL trains to connect to the WCML at Willesden instead (just as they run in-between TfL trains now all the way from Stratford). There is room to expand freight 8 fold. TfL trains would shift over to the 2 restored tracks along the existing railway land and use restored platforms at Camden Rd.

As explained above - they can’t all be timetabled that way, because additional freight that joins the NLL from the Gospel Oak - Barking line fits into the spaces vacated by the trains pathed via Primrose Hill. There is not room to expand freight 8 fold. Not even 1/8th fold. I have seen recent work that demonstrates the capacity issues on the NLL - it is quite possible that Network Rail starts taking trains out of the existing timetable as it simply doesn’t work now. There is certainly not scope for diverting all the Primrose Hill trains via Gospel Oak on a permanent basis. There must be a freight route via Primrose Hill, as proposed in all the Arup options, so you need to amend your proposal to accommodate it.

4) As the Arup drawings show, the viaduct is already wide enough (at 8.4 metres) to take a pair of GC guage trains passing each other. It is only the central I frames of the bridge superstructures that get in the way, with the steelwork of 6 road-bridges needing to be replaced.

Without going down there, taking a line block and getting my tape measure out, I don’t know how to prove it to you. However gauging is about much, much more than static measurement of one point of a viaduct. GC gauge has a much bigger kinematic envelope (KE) than W10, with centre throw and end throw being particular issues on curves, especially tight curves as we have here. The tracks must be further apart throughout, and that causes gauge infringement with the parapet walls of the masonry viaducts. Then there is below the solebar, and all the kit we put there that is in the way of GC. Then of course the height - all that OLE is in the way and will have to go, along with the portals holding it up.

See below, from the Arup report, my bold and reddening.

Route Option R0 Option R4C
7.2.2 These options start at Old Oak Common Station. They use a single bore tunnel going eastwards to a tunnel portal in the Camden area from Chalk Farm to Primrose Hill. Then they join the NLL, upgrading its southernmost track to GC gauge and using this corridor to connect to the HS1 link near St Pancras.

7.2.3 Initially, there was an exception to this typical route, which was an Option RB variant. This variant considered a shared double track rail link, using both of the NLL tracks. The option was abandoned, owing to the difficulty of taking both NLL tracks and the extensive works required to upgrade these same tracks to GC gauge.

This SAVES about £6 billion, can be delivered EARLIER than a rebuilt Euston

1) Can you explain how much money the shorter tunnel saves money compared to the tunnel to Euston. Just for the tunnel section, nothing else. In your answer, take into account that the contracts have been let, the materials all procured, the TBMs are bought and being assembelled, the Euston portal well under construction, and all the geotechnical investigations, design, and prep works done and paid for. Also bear in mind that none of this investigation or prep work has been done for a tunnel emerging at Primrose Hill.

2) Can you set out some high level milestones that show how it can be delivered earlier than a rebuilt Euston. Start with the consents process.

3) Yet again, NO NEED TO SHOUT.

 

[Class15]

As explained above - they can’t all be timetabled that way, because additional freight that joins the NLL from the Gospel Oak - Barking line fits into the spaces vacated by the trains pathed via Primrose Hill. There is not room to expand freight 8 fold. Not even 1/8th fold. I have seen recent work that demonstrates the capacity issues on the NLL - it is quite possible that Network Rail starts taking trains out of the existing timetable as it simply doesn’t work now. There is certainly not scope for diverting all the Primrose Hill trains via Gospel Oak on a permanent basis. There must be a freight route via Primrose Hill, as proposed in all the Arup options, so you need to amend your proposal to accommodate it.

This is not correct. While losing the Primrose Hill line would be a nuisance for network flexibility, particularly during engineering and disruptions, it is not needed on a regular basis. The freight can and does fit in via Gospel Oak when the Primrose Hill line is closed (it is a lot of the time). It sounds like a nice idea but there aren’t at all that many (if any) freights off the GOBLIN that take up slots used by Primrose Hill trains, because of the amount of the time that the Primrose Hill line is shut.

 

[Bald Rick]

This is not correct. While losing the Primrose Hill line would be a nuisance for network flexibility, particularly during engineering and disruptions, it is not needed on a regular basis. The freight can and does fit in via Gospel Oak when the Primrose Hill line is closed (it is a lot of the time). It sounds like a nice idea but there aren’t at all that many (if any) freights off the GOBLIN that take up slots used by Primrose Hill trains, because of the amount of the time that the Primrose Hill line is shut.

I’m sorry, but the timetable planning teams of Network Rail, Arriva Rail London, and the freight companies disagree with you. As do the Anglia operations team. And as they are responsible for it, I reckon their opinion carries some weight.

 

[30907]

PDFH recommends that interchanges involving 2 stations (e.g. Euston to St Pancras) are treated as two separate interchanges, and the walking time is included in the journey time.

My apologies, I stand corrected

I think that's personally a little pessimistic - London transfers should probably be 1.5 changes or similar as a tube/walking transfer is not the same as a separate change of train.

More than a little, but yes it may have a psychological effect.

 

[John Jefkins]

You said:

60km/h (37mph) from the bufferstops to 700 metres
65km/h (40mph) between 700m and 875m
70km/h (43mph) between 875m and 1065m
100km/h (62mph) between 1065m and the Euston tunnel portal at 1275m
145km/h (90mph) between 1275m and 2350m - approximately alongside Primrose Hill
160km/h (100mph) beyond 2350m, before stepping up to 180km/h then 230km/h in the Euston tunnels.

By comparison, coming out of St Pancras would be:

Exactly the same 37mph restriction for the first 700 metres (being a similar design of 3 roads out of St Pancras with the dualled viaduct plus a viaduct over Eurostar tracks)
40-45 mph alongside the TfL tracks for the next 1km (to the Camden Market S curve to Primrose Hill) and 35mph around those curves.
Then accelerating to 90 mph etc from just east of Primrose Hill.

The total distance from either St Pancras or Euston to the Primrose Hill junction would be the same 2km, so we're talking about 30mph slower for 1 mile = 2 minutes slower.
(It's actually about 1.4 km so less than a mile so I'm being generous with that 2 minutes).

ONLY TWO MINUTES SLOWER, but saving £6 billion and a 60 minute interchange penalty for 70% of the market - to earn approx £2 billion extra revenue annually in the 2040s.

This is not correct. While losing the Primrose Hill line would be a nuisance for network flexibility, particularly during engineering and disruptions, it is not needed on a regular basis. The freight can and does fit in via Gospel Oak when the Primrose Hill line is closed (it is a lot of the time). It sounds like a nice idea but there aren’t at all that many (if any) freights off the GOBLIN that take up slots used by Primrose Hill trains, because of the amount of the time that the Primrose Hill line is shut.

Thanks for your support. Goblin traffic uses just one extra of those 8 paths, so there are thus 6 other paths available for expansion.
My drawings also show how there is still room for night time usage of the Primrose Hill route (with room for a 3rd surface track past the Roundhouse).

 

[Class15]

Thanks for your support. Goblin traffic uses just one extra of those 8 paths, so there are thus 6 other paths available for expansion.
My drawings also show how there is still room for night time usage of the Primrose Hill route (with room for a 3rd surface track past the Roundhouse).

You’re most welcome.

Not quite - there are 4tph paths available for freight between Gospel Oak and Kensal Green Jn, and if you look at the timetable there are loads of spare gaps in the passenger services.

Take 4L89. It fits in between 2N16 and 2L16 at Camden Road, and as a result there is a 8 minute gap between those two services. That gap is there from all the way back at Willesden. As a result, 4L89 could be re-routed quite easily via Gospel Oak. It would also remove the dreaded crossover just west of Camden Road, which often delays westbound trains needing to cross the Primrose Hill line. In the other direction, it might be slightly harder over Kensal Green Jn but should still be doable.

 

[John Jefkins]

Got a bit more time to go back to some other points.



There will be land take. All the reports done on this - at very high level - say there will be; this includes the Arup report you are relying on. Read pages 50 to 77 (Section 7). It mentions demolition of residential, commercial and industrial premises, even for the simplest option (which is less work than what you propose.) The segregated solution, which is similar (but still rather less) to what you propose, involves land take of parts of the various Camden Markets. Good luck with that! That’s just the permanent land take; during construction there will be much more, including the severance of roads throughout the area. This is not me saying this, it is the Arup report.




As explained above - they can’t all be timetabled that way, because additional freight that joins the NLL from the Gospel Oak - Barking line fits into the spaces vacated by the trains pathed via Primrose Hill. There is not room to expand freight 8 fold. Not even 1/8th fold. I have seen recent work that demonstrates the capacity issues on the NLL - it is quite possible that Network Rail starts taking trains out of the existing timetable as it simply doesn’t work now. There is certainly not scope for diverting all the Primrose Hill trains via Gospel Oak on a permanent basis. There must be a freight route via Primrose Hill, as proposed in all the Arup options, so you need to amend your proposal to accommodate it.




Without going down there, taking a line block and getting my tape measure out, I don’t know how to prove it to you. However gauging is about much, much more than static measurement of one point of a viaduct. GC gauge has a much bigger kinematic envelope (KE) than W10, with centre throw and end throw being particular issues on curves, especially tight curves as we have here. The tracks must be further apart throughout, and that causes gauge infringement with the parapet walls of the masonry viaducts. Then there is below the solebar, and all the kit we put there that is in the way of GC. Then of course the height - all that OLE is in the way and will have to go, along with the portals holding it up.

See below, from the Arup report, my bold and reddening.







1) Can you explain how much money the shorter tunnel saves money compared to the tunnel to Euston. Just for the tunnel section, nothing else. In your answer, take into account that the contracts have been let, the materials all procured, the TBMs are bought and being assembelled, the Euston portal well under construction, and all the geotechnical investigations, design, and prep works done and paid for. Also bear in mind that none of this investigation or prep work has been done for a tunnel emerging at Primrose Hill.

2) Can you set out some high level milestones that show how it can be delivered earlier than a rebuilt Euston. Start with the consents process.

3) Yet again, NO NEED TO SHOUT.

Thanks Rick,
Since Arup's 2014 report, buildings alongside have been removed and there are now two wide passageways (Water Lane) that remove need of land take.
I have indeed used my Disto (more accurate than a tape measure) and the viaduct is 8.4 to 8.5 metres wide with 315mm brick non-structural parapets that could be removed.
Signaling is already cantilevered over the side - as could any walkway and/or narrower metal handrail, with easy scaffolding access from Water lane to remove parapets.

I see no demolition necessary apart from the roof and 1st floor of one dentist surgery at Kentish Town road (where the Victorians failed to 4 track the route and a wider viaduct is need to span over (and thus shelter) the rear of the school playground. That single demolition is shown as red dotted lines below. But the same dentists could re-open as a single floor building under the bridge and using the rear garden of the same property to add back the missing 1st floor space. Its already got a similar extension on the front.


2) As mentioned elsewhere by other people, freight trains have no problem running to Willesden. Avoiding the single Gospel Oak train, 6 other paths are available per hour.

3) Contracts are subject to change and those same TBMs would just stop north of Juniper Crescent (by Primrose Hill) instead of digging an extra 1km towards Euston.
The savings are in postponing Euston until a much smaller and more integrated station can be designed where HS2 and normal trains share the same (widened) surface level throat and share some platforms too in a much more flexible way. Most HS2 trains would run to St Pancras (all classic compatibles and when we get use of 2 of the 400m platforms - eg 5 and 10 some of the 400m trains too) with the Birmingham shuttle only starting to use Euston in the late 2030s or 2040s.

The boosted income comes from getting trains into central London earlier than Euston could be opened and in those interchange savings that boost HS2 (and cross London) markets.

What we need to do urgently is work out how to add an S shaped single tracked ramp within the O.O.C. box (very similar to the same ramp to Temple Mills at Stratford) to link Javelin trains to GWR platforms alongside and enable them to extend to replace the 4 paths of the Heathrow Express.

4) No consents (Limit of Deviation or Parliamentary) are needed as this is all within Railway Land. Its a railway upgrade where only Camden Council need to approve replacement work of the steelwork over the existing railway bridges to replace Victorian I beams with side-arched bridges that could be slightly wider. The iconic "Camden Road" side panel could possibly be added back (somehow?) or an open arch could replace it. Other minor bridges east of Camden Road could be simpler and cheaper concrete planks.

The key things though are that as TfL saves £1billion by not needing to upgrade Euston tube (and also saves money on yet more Elizabeth Line trains) they can afford about £50m to £60m to 4 track the route (re-decking bridges, adding track, restoring the old platforms and adding a lift and widening Kentish Town Road bridge).
HS1 Ltd get a huge boost to the value of their stations and HS1 itself, so they can afford to (a) add your junction links & signaling at St Pancras junction (b) widen the 1 track viaduct to 2 track and (c) add a reversable single track viaduct curving over the middle of the triangle to descend in the gap between tracks (or over the boat dock) to reach Javelin platforms. Engineers tell me having a concrete works on site in that triangle could actually reduce costs

They could also afford to add extra lifts by the Javelin platforms and fill in the draft atrium to add more seating space there. There is even height space there to add a balcony.
And of course it would be sensible to raise the floor level of the old platform alongside platform 10 to allow the flexibility of using it for rush hour Javelins.

So this doesn't just save HS2 costs. It provides much earlier income and boosted income too (by removing the interchange penalty).
It adds HS1 & Tfl to pay costs, whilst also enabling more lucrative use of most of the cleared land at Euston (as fewer platforms would ever be needed there).

You made a good point about 7 Javelins per hour at rush hour. Looking further at that
a) Two of them are to Ashford. By simply extending 3 of the HS2 trains in rush hour from terminating in a double slot at St Pancras to terminating at Ashford,
(i) we remove the need of 2 of those 7 Javelins and (ii) we create 3 extra slots of 7.5 minutes in rush hour to use for a 5th Javelin service to say Margate.
b) Another option would be to use 30 6 minute slots /hr instead of 24 7.5 minute slots as that adds room for your 3 extra Javelins (in those 6 extra slots).
c) My favoured option is to persuade HS1 Ltd to release platform 10 for say 90 minutes around rush hours to add 12 extra slots for each rush hour.
In reality, extra HS2 trains would certainly extend to Ashford in rush hour to return in less crowded hours afterwards and we could see a combination of these solutions.

Happy to underline or use bold instead of CAPITALS now I see the options. Apologies as I'm new to this site

That bridge is IMPOSSIBLE. Stop repeating it as an option.

There is no physical way in that distance to get a line up and over and down into St Pancras using the basic limitations of gravity.

I think you are being incredibly rude in your postings on here and you have not responded to any valid critiscism.

You can claim that only 4 javelins per hour use St Pancras normally, however at some point you will have to face that in reality at rush hour it is much more and this is likely when you will have your HS2 services arriving.

I am sure all the HS1 javelin customers will also be thrilled that they are now getting dumped in OOC rather than St Pancras.

That bridge isn't impossible as I have measured the gradients. Although it might need to span to the side dock I accept.

I'm new to this site and have been away (on train trips to York and Harrogate) with only access to a phone rather than this laptop. I'd not seen the questions as I had not seen the tiny page numbers on my phone. I'm only now getting time to respond. Please bear with me. I'm grateful for your comments and questions.

I really have answered the 7 Javelins per hour question - saying that (a) HS2 trains extending at rush hour to Ashford replace need of 2 of the 7, (b) 6 minute slots add room for 3 extra Javelins or (c) use of platform 10 at rush hours would add an extra 12 slots or 7.5 minutes to reverse Javelins towards OOC.



All Javelin passengers could get on and off at St Pancras. But many passengers would stay on board to reach OOC (or ideally Heathrow if a Stratford international style ramp was added within the bare box so far completed at OOC - S curving over the platforms. Again its a 500m ramp at the gradient able to be used by those trains. I have checked that.

Almost all trains would use St Pancras so Nobody is being "dumped" at OOC.
4 Javelins and 8 HS2 trains per hour would stop at each of Stratford, St Pancras and Old Oak Common. Some empty to depot trains from OOC might bypass St Pancras.
The 12 trains/hr reversing in around 4 to 5 minutes at St Pancras would create a new ACROSSRAIL service.
ICE trains reverse in the same timeframe at Frankfurt HbH.

Option 1


Option 2

Option 3

 

[dosxuk]

What we need to do urgently is work out how to add an S shaped single tracked ramp within the O.O.C. box (very similar to the same ramp to Temple Mills at Stratford) to link Javelin trains to GWR platforms alongside and enable them to extend to replace the 4 paths of the Heathrow Express.

It's already too late for this. That sort of structure needs to be designed, engineered and integrated with the construction. To add it now would involve basically tearing down what's there and starting again from scratch.

Furthermore, a ramp coming up out of the low-level station wouldn't make it to ground level until it was well past the intended site for OOC - meaning trains using it - in the same way as Stratford - would not be able to call there.

 

[yoyothehobo]

That bridge isn't impossible as I have measured the gradients. Although it might need to span to the side dock I accept.

Care to tell us what the gradients are over what distance? What height have you assumed they need to clear the line beneath?

 

[John Jefkins]

Eurostar capacity is limited by space at St Pancras and work is already being undertaken to fix that. You cannot assume 2tph, effectively capping international trains to the UK permanently.

I am also still awaiting your answer to if you plan to fit ETCS to HS1, and the javelins, which while a long term plan is not funded.

Fake money. TfL doesn’t have a spare billion to spend on their own issues. The billion you quote would’ve been part of Crossrail 2, and if that does go ahead the station will still be built to the same cost. (and funded likely in large by the government)

HS1 already has the equivalent signalling. When we're saving billions a few tweaks to ensure this line can cope with 20 trains/hr (like Thameslink) is peanuts.

I have carefully avoided use of the 6 "Eurostar" platforms (except platform 10 for rush hours) to allow Eurostar to expand.
But to really expand, some international services will need to reach OOC to win the west of London and more of the Heathrow markets.

The billion isn't part of Crossrail. A station upgrade similar to Victoria's £900m upgrade is planned for HS2.

It's already too late for this. That sort of structure needs to be designed, engineered and integrated with the construction. To add it now would involve basically tearing down what's there and starting again from scratch.

Furthermore, a ramp coming up out of the low-level station wouldn't make it to ground level until it was well past the intended site for OOC - meaning trains using it - in the same way as Stratford - would not be able to call there.

As they've only built the box base so far, it doesn't need tearing down.
I worked on the Channel Tunnel project. When Maggie stood up in Bruges we had to redesign the whole terminal at Folkestone. That's why we use CAD these days

Care to tell us what the gradients are over what distance? What height have you assumed they need to clear the line beneath?

I agree it is tight, and it would need surveying. Having travelled around Japan though, these structures get thrown up there in days
Seriously though it does look possible because the lines drop under where the viaduct would cross. I'd need a site plan in Autocad to fully check it out though.

The curve across the triangle is 120m @ 3% where the line it's crossing is also dropping @ 3%? - so there could be a 7.2 metre height difference by that point.
More if the start point is higher (which is why a proper survey would be needed). That's the crucial bit.
If more room is needed then it would just need to start rising from a different point to increase the height difference where it crosses the tracks dropping down.
The rest of the route could be bridged over to the dock if there isn't enough of a gap between the tracks for the ramp back down. A solution could be worked out.

Getting into the Eurostar platforms is a lot simpler than reaching the Javelin platforms.
Most traffic would run across the throat as there really are not many Eurostar movements each hour - so this bridge is only needed if all the Javelins went to OOC.
There are multiple options for using St Pancras and HS1 platforms. HS1 don't want to discuss their "poor use" of the 6 400m platforms until they work out a deal with Eurostar competitors - who'd only frankly add a few trains each per day.
I still contend that to expand Eurostar, some of those trains should run to exploit the west of London market from OOC instead of all terminating @ St Pancras.
That frees up 2 of the 6 "Eurostar" platforms to reverse HS2 trains in - and extend HS2 trains to terminate at Stratford or Ebbsfleet (& Ashford at rush hour).
Ashford has Balfour Beatty sidings beyond the 4 platforms there as well as the Javelin depot for HS2 train stabling.

Question - what other depot space is their accessible from HS1 that HS2 trains or "Eurostar" competitors could use - as Temple Meads space is at a premium?
Another point could be that it could be cheaper to store many of the HS2 trains overnight to the east of London....

 

[Benjwri]

HS1 already has the equivalent signalling. When we're saving billions a few tweaks to ensure this line can cope with 20 trains/hr (like Thameslink) is peanuts.

HS1 doesn’t have ETCS, Javelins dont have ETCS. You’re going to have to fit it with it to run the HS2 trains down it, or pay to make the HS2 stock compatible.

You’ll also have to fit OOC with colour signalling or the Javelins with ETCS.

But to really expand, some international services will need to reach OOC to win the west of London and more of the Heathrow markets

Don’t get me wrong I would love this, but it isn’t going to happen. OOC has start construction. You would have to significantly relax security for it to happen, without devoting an entire island platform to international services, which obviously isn’t feasible, along with a very expensive redrawing at this late stage.

 

[yoyothehobo]

The curve across the triangle is 120m @ 3% where the line it's crossing is also dropping @ 3%? - so there could be a 7.2 metre height difference by that point.
More if the start point is higher (which is why a proper survey would be needed). That's the crucial bit.
If more room is needed then it would just need to start rising from a different point to increase the height difference where it crosses the tracks dropping down.
The rest of the route could be bridged over to the dock if there isn't enough of a gap between the tracks for the ramp back down. A solution could be worked out.

You would be lucky to get over the first set of lines even if they are dropping as where you want to put the junction in you have restricted space due to the lines to the East which are climbing into the HS1 route. You are restricted by the lines that diveunder as well so you have a very short distance with which to lead into your viaduct that is at approximately 32mAOD. This then has to climb over a line which is at 35mAOD and then plus any gauge clearance plus height of the structure. I dont know what this is but 4m for train plus 1m? for OLE clearance +2m for structure and track so that is 7m. A total of 10m height you have got to gain in about 150m, so a gradient of 1 in 15, with a very similar descent on the north side towards the NLL that is at 35m AOD. This is also at a very tight gradient.

You cannot do much with the immediate throat of St Pancras either as to the east of the line is a lot of buildings and beneath the surface are the Thameslink lines which restrict your piling options for bridges.

I really dont think it is physically possible to put this link in without extending it all the way over the overground lines to the North and connecting off the Camden town lines there, but then that messes with a plan having segregated routes because you will have to cross again towards Primrose Hill and require massive amounts of landtake.

Ergo, i think the bridge is impossible.

 

[John Jefkins]

I think that is very unlikely. Anyone from Luton or Cambridge going to Heathrow will use Thameslink to Farringdon, and the Elizabeth Line to Heathrow. From arriving at St Pancras to arriving at Heathrow is typically 50 minutes including the change at Farringdon, although I have personally done it in 45. Going via St Pancras via your imaginary link and OOC is two changes, and couldn’t possibly be less than 45 minutes, but more likely 50-55. Why would people make it more difficult for themselves for little or no gain?



Perhaps I can help explain how to interpret the sectional appendix, because you have not got this right.

The HS1 sectional appendix is in km/h. For trains heading from the international platforms towards the connection with the NLL it is 40km/h the whole way. (As stated previsouly, there is no connection from the domestic platfroms to the NLL link, but it is reasonable to assume that if there were, it would be 40km/h max, but very likely lower). 40km/h is approximately 25mph.

Moving on to the NR East Anglia Sectional Appendix, as pasted into the thread by others above, the connection from the HS1 Silo Curve leads on to the North London Incline, and the connection from that line onto the NLL is 15mph, as marked by the “15”; this is at Camden Central Junction, 4miles and 64 chains as shown in the table to the left of the diagram. At that point the NLL is indeed 35/40mph (the 40 mph applying to passenger services), however that speed applies only until 4miles 71 chains, ie 7 chains (141 metres) further on, where the asterisk (*) on the line concerned denotes a change of speed. From here through Camden Road station it is 20mph, as shown by the “20” on each line through Camden Road station. As any Javelin or HS2 train that would use this imaginary route is longer (or very early so) than 141 metres, one part of it would always be on the 20mph restriction or the 15mph restriction, and therefore unable to make use of the 40mph.

Heading towards Primrose Hill the line remains at 20mph, until the next asterisk at 5m18ch (ie 27 chains - 543 metres - further on from the start of the 20mph restriction). It is then 15 mph towards Primrose Hill, as denoted by the “15” on each line beyond the asterisks. A 400metre HS2 train would have only 143metres of track where it is exclusively in the 20mph stretch, and I doubt if it would be realsitic to expect drivers to accelerate to 20mph before almost immediately braking back for 15mph; they would just do 15mph all the way. A 200 metre train or Javelin might be able to accelerate - briefly - up to 20mph, but I rather suspect that the standards on drivability would kick in and it will be 15mph the whole way.

Therefore best case for a 400metre HS2 train would be to accelerate to 25mph from the bufferstops to the braking point for reducing to 15mph at the North London line junction (almost exactly a mile from the buffer stops) and then 15mph from there for the 3/4 mile or so to your proposed portal at Primrose Hill. Only when the back of the train - a further 400 metres at 15mph - is on your proposed new line can it then commence acceleration up from 15mph. Feel free to do the calculations using expected braking and acceleration rates, not forgetting to allow for the length of the train.

I can’t see how it can be less than 6 minutes to the rear of the train clearing Primrose Hill (still doing 15mph) and probably more, by which time an equivalent HS2 train from Euston is already about to depart OOC. For example a mile at 15mph from the front of train passing Camden Road Central Junction to back of train clearing the existing line at Primrose Hill takes 4 minutes alone. I’ll be happy to stand corrected if I am proven incorrect through your calculations.




Let’s see what your calculations show, based on the actual linespeeds, and not those that you seem to think it is, before you start saying something is proven.

But we aren't talking about today's track or signalling. We're talking about maximum speeds around those radii - which is a lot nearer to 40mph than 15mph.
1km of the route from St Pancras to Camden Market is straight along newly laid GC tracks with no junctions or stations, so 40-45 mph is possible .
Again, this isn't now. Its re-signaled and relaid GC Gauge track over wider road bridges with potential to increase the radius a touch at each end - to raise it 10kph to 38mph.

I've drawn the radius out accurately on Autocad and compared that to maximum speeds for railway radii. Its a 350 metre radius - or a 43mph to 50mph max line-speed.
I assumed only 35mph but 45 mph may be a more likely max speed for most of this distance. Here is the evidence on Autocad. The purple construction circle is 350m radius.

 

[yoyothehobo]

Can you give some details of projects use have worked on as a track engineer and designer? I am not saying you havent but the people who you are disputing this with have given plenty of evidence of their experience on the forum. A simple wikipedia reference doesnt quite do it for me and i am not sure how comfortable i would be going through 350m radius reverse curves at 45mph in HS2 rolling stock

 

[dosxuk]

As they've only built the box base so far, it doesn't need tearing down.

You might want to check that - they actually completed the structure in February. The box was constructed from the top down, with the base being the last part to compete.

Platform / lift / escalator / HVAC positioning is literally cast in stone at this point. Deciding to cut a whole load out to build a big, heavy ramp through the middle will need it all redesigning and rebuilding. The structural loads on the structure will be completely different with the middle cut out to fit in your ramp.