Possible HS2 Euston station de-scoping

[HSTEd]

If we think 9 platforms for 18tph is overly ambitious, maybe we should consult with the control room staff responsible for operations into Fenchurch Street.

 

[Bald Rick]

If we think 9 platforms for 18tph is overly ambitious, maybe we should consult with the control room staff responsible for operations into Fenchurch Street.

Somewhat different servicing arrangements at turnaround, but the principle is sound.

 

[The Planner]

If we think 9 platforms for 18tph is overly ambitious, maybe we should consult with the control room staff responsible for operations into Fenchurch Street.

Apples and apples though isn't it though, Fenchurch St isn't dealing with long distance trains that require longer turnarounds.

 

[HSTEd]

Apples and apples though isn't it though, Fenchurch St isn't dealing with long distance trains that require longer turnarounds.

Many of the HS2 trains are not particularly long distance though.

Manchester in travel time terms is as far from Euston as Shoeburyness is from Fenchurch Street.

Birmingham, Manchester and Leeds won't really be long distance destinations any more.
There will be long distance trains with multi hour journeys, but many of the trains will not be.

 

[100andthirty]

Building on HSTEd's point, it's an interesting notion that long distance trains take longer to turn around than short distance train. I often how quickly trains are turned round at Victoria. Whilst the distances are smaller, the journey times - at least for Birmingham are comparable to those for London to the south coast. Clearly a 400m long train might take longer to alight and load than 240m long trains, but even then having entry and exit points are several location and not just at one end will help to minimise this impact. What is needed is very careful critical path analysis to determine the minimum platform occupation time and then work had to try and reduce it further.

Others have mentioned Fenchurch St. I'd mention Brixton on the Victoria line. Clearly the Victoria line is not the same as HS2/Euston, but delivering 36 trains per hour though two platforms required very careful optimisation of all the participants in the process and is a testament to the original design where the run in and run out times were minimised by having a high-ish speed crossover on the approach and over-runs beyond the platforms.

Similar systems thinking should be applied to the redesign of Euston to make 10 platforms work

 

[Bald Rick]

Many of the HS2 trains are not particularly long distance though.

Manchester in travel time terms is as far from Euston as Shoeburyness is from Fenchurch Street.

Birmingham, Manchester and Leeds won't really be long distance destinations any more.
There will be long distance trains with multi hour journeys, but many of the trains will not be.

They’ll still need longer to load / unload, and clearing of litter, possibly restocking trolleys / galleys etc., which commuter operators don’t need to worry about.

Also, many of the HS2 trains will be 400m long, which will clearly take longer to get in/out of the platform (even at 60kph).

Building on HSTEd's point, it's an interesting notion that long distance trains take longer to turn around than short distance train. I often how quickly trains are turned round at Victoria. Whilst the distances are smaller, the journey times - at least for Birmingham are comparable to those for London to the south coast. Clearly a 400m long train might take longer to alight and load than 240m long trains, but even then having entry and exit points are several location and not just at one end will help to minimise this impact. What is needed is very careful critical path analysis to determine the minimum platform occupation time and then work had to try and reduce it further.

Others have mentioned Fenchurch St. I'd mention Brixton on the Victoria line. Clearly the Victoria line is not the same as HS2/Euston, but delivering 36 trains per hour though two platforms required very careful optimisation of all the participants in the process and is a testament to the original design where the run in and run out times were minimised by having a high-ish speed crossover on the approach and over-runs beyond the platforms.

Similar systems thinking should be applied to the redesign of Euston to make 10 platforms work

A lot of work has been done on that for the existing railway - I was doing it, at Euston, nearly 30 years ago!

In the ‘old days’ longer distance trains needed to be tanked, reservations removed / installed, cleaned, buffets restocked, parcels loaded / unloaded, etc. Much of that doesn’t apply now. Also, the termini turnaround on a long distance service would often correspond to the traincrew’s break, and they would take the train back out. This is less of an issue for the commuter termini where there are a lot more trains for drivers to take back out.

But most importantly, a long distance train has much more opportunity to be delayed compared to a shorter distance (but similar duration) service, and therefore more time is allowed at terminal to soak up delays.

 

[yorksrob]

I hope they keep the current Euston concourse. It's a design classic.

 

[Bletchleyite]

Building on HSTEd's point, it's an interesting notion that long distance trains take longer to turn around than short distance train.

It's not only about stocking the buffet, it's also about having a long enough layover to keep things punctual. Yes, you can clean on the move (VTWC did that to pretty good effect, you rarely saw a litter-strewn Pendolino) but that doesn't get you back on time if you're late.

 

[100andthirty]

They’ll still need longer to load / unload, and clearing of litter, possibly restocking trolleys / galleys etc., which commuter operators don’t need to worry about.

Also, many of the HS2 trains will be 400m long, which will clearly take longer to get in/out of the platform (even at 60kph).



A lot of work has been done on that for the existing railway - I was doing it, at Euston, nearly 30 years ago!

In the ‘old days’ longer distance trains needed to be tanked, reservations removed / installed, cleaned, buffets restocked, parcels loaded / unloaded, etc. Much of that doesn’t apply now. Also, the termini turnaround on a long distance service would often correspond to the traincrew’s break, and they would take the train back out. This is less of an issue for the commuter termini where there are a lot more trains for drivers to take back out.

But most importantly, a long distance train has much more opportunity to be delayed compared to a shorter distance (but similar duration) service, and therefore more time is allowed at terminal to soak up delays.

All really good points, and as far as I understand it the long Virgin/Avanti timings at Euston are both to cover delays and to try and keep the units on their assigned paths to fit in with getting trains to maintenance at the right time. When the staff need to, turnaround can be really quick.

The other thing is to put extra effort into managing out delays. But I wonder what might come out if an expert group was presented with the challenge of:

"What would I do to the station/train/servicing vehicle design and deployment of staff" to achieve a turnaround in (let's say) 15 minutes. You are allowed to think the unthinkable".

By 15 minutes I mean from wheel stop to wheel start.

 

[quantinghome]

It's not only about stocking the buffet, it's also about having a long enough layover to keep things punctual. Yes, you can clean on the move (VTWC did that to pretty good effect, you rarely saw a litter-strewn Pendolino) but that doesn't get you back on time if you're late.

Isn't the idea that the timetables will be planned for a 330kph top speed, but the trains will be able to do 360kph in service, thus maintaining punctuality? Great to have multiple layers of mitigation against delays built-in but it does feel it's moving into 'nice to have' territory.

 

[Bletchleyite]

Isn't the idea that the timetables will be planned for a 330kph top speed, but the trains will be able to do 360kph in service, thus maintaining punctuality? Great to have multiple layers of mitigation against delays built-in but it does feel it's moving into 'nice to have' territory.

I think the effect of considering adequate layover as a "nice to have" will be akin to considering Manchester Piccadilly platforms 15-16 a "nice to have", to be honest. Or of cutting testing on an IT project. It's easy to do, but it has long-lasting effects.

If you're building a multi-billion pound railway, saving in effect a tenner (OK, I'm exaggerating) by lopping one platform out doesn't sound sensible to me.

 

[CW2]

I think we can probably agree that it is possible to turn a long-distance train around in less than 20 minutes. But for that to take place you need:

• Additional staff for cleaning and servicing
• A very high degree of punctuality of the incoming services
• Controlled passenger boarding i.e. passengers kept hanging around on the concourse waiting for their train

Now scale that up from just one quick turnround to 18 trains per hour doing much the same. It doesn't take long for the extra staff costs and passenger inconvenience, plus delays / cancellations due to poor punctuality for the additional platform to pay for itself - probably many times over.

In timetabling terms this also means Euston HS2 becomes the permanent constraint on the entire network. Where you might have "hard points" in the timetable at other termini e.g. Birmingham or Manchester, now you end up resolving those by adding in pathing time - which is exactly what you don't want to do on a high speed railway.

 

[The Planner]

In timetabling terms this also means Euston HS2 becomes the permanent constraint on the entire network. Where you might have "hard points" in the timetable at other termini e.g. Birmingham or Manchester, now you end up resolving those by adding in pathing time - which is exactly what you don't want to do on a high speed railway.

Not sure it will, any touch point with the conventional network are going to be the pivots. HS2 will have such a high PPM expectation that Handsacre, Crewe etc are going to be very much first on the graph and on the day any other train even mildly late is going to be kept out the way.

 

[HSTEd]

Ultimately I would expect Birmingham International to need more platform space.
After all, worst come worst if a train is too late to make it's slot on the core, you can axe it and put the people on the next train south.

 

[quantinghome]

I think the effect of considering adequate layover as a "nice to have" will be akin to considering Manchester Piccadilly platforms 15-16 a "nice to have", to be honest. Or of cutting testing on an IT project. It's easy to do, but it has long-lasting effects.

Platform 15-16 is a very obvious cause of delay though. Trains timetabled less than every 5 minutes on the most heavily congested route in the North of England and completely inadequately-sized platforms is hardly a useful comparison. The debate at Euston is whether the average time between arrivals per platform should be 33 or 37 minutes. I get that we should build for a reliable railway, but it does look quite generous.

Given there is no rolling stock depot at the London end there's no need for ECS moves. Given the short duration of journeys, restocking could be done at the 'country' end of the service. OOC provides a useful point for cleaning staff to come on board and do a sweep prior to arrival at Euston. What is there that actually needs doing at Euston which takes half an hour?

Can someone tell us whether a thorough root-cause analysis been undertaken of HS2 punctuality? As in, HS2 defines a certain acceptable probability of service delay - say 1% of trains delayed more than 5 minutes (it can never be zero), or some other form of metric. Then the infrastructure is designed to provide that level of punctuality. Sources of delay are identified, probabilities assigned, mitigations proposed and so on, and the most cost-effective solution is reached. Maybe I'm wrong, but it feels like layer upon layer of robustness is being added at great expense without detailed assessment of the merits.

 

[ABB125]

Ultimately I would expect Birmingham International to need more platform space.
After all, worst come worst if a train is too late to make it's slot on the core, you can axe it and put the people on the next train south.

Is Birmingham Interchange a 2-platform design?

Given there is no rolling stock depot at the London end there's no need for ECS moves. Given the short duration of journeys, restocking could be done at the 'country' end of the service. OOC provides a useful point for cleaning staff to come on board and do a sweep prior to arrival at Euston. What is there that actually needs doing at Euston which takes half an hour?

Some good ideas here. I especially like the idea of staff boarding at OOC to clear litter, since it's probably possible for one member of staff to get a fair way through a 200m unit in the couple of minutes to Euston. Also good because a lot of passengers will leave at OOC, so there will be more space for the cleaners to operate. (The cleaners also get to spend all day shuttling between EUS and OOC, getting their mileage up! )

 

[Bald Rick]

But I wonder what might come out if an expert group was presented with the challenge of:

"What would I do to the station/train/servicing vehicle design and deployment of staff" to achieve a turnaround in (let's say) 15 minutes. You are allowed to think the unthinkable".

By 15 minutes I mean from wheel stop to wheel start.

That’s exactly what I was doing nearly 30 years ago, and has been done several times since. The issue is, someone comes up with a solution, the trains are sent to the scrapyard (or not ordered) and then at the first sign of trouble the timetable collapses.

Now in the HS2 case, I would expect some turnarounds to be as short as 20 mins, ie those that are on the dedicated network. Long enough to get everyone off, quick service and and open for boarding 10-15 mins before departure (concourse space for intending passengers is another factor for consideration). But others will be half an hour+.

Can someone tell us whether a thorough root-cause analysis been undertaken of HS2 punctuality? As in, HS2 defines a certain acceptable probability of service delay - say 1% of trains delayed more than 5 minutes (it can never be zero), or some other form of metric. Then the infrastructure is designed to provide that level of punctuality.

I’d be astonished if this hasn’t been done. But the new infrastructure design is only on HS2 - there’s plenty that can go wrong on the existing lines that will affect HS2 services.

Is Birmingham Interchange a 2-platform design?

4 platforms, plus 2 through lines.

 

[The Planner]

Ultimately I would expect Birmingham International to need more platform space.
After all, worst come worst if a train is too late to make it's slot on the core, you can axe it and put the people on the next train south.

Is Birmingham Interchange a 2-platform design?

As far as I know, Interchange is 4 platforms, 2 islands. Tracks either side of the platforms and two through tracks, so six in total. Anything on the Curzon St spur doesn't have to interact until south of Interchange.

 

[Ianno87]

They’ll still need longer to load / unload, and clearing of litter, possibly restocking trolleys / galleys etc., which commuter operators don’t need to worry about.

All really good points, and as far as I understand it the long Virgin/Avanti timings at Euston are both to cover delays and to try and keep the units on their assigned paths to fit in with getting trains to maintenance at the right time. When the staff need to, turnaround can be really quick.

The other thing is to put extra effort into managing out delays. But I wonder what might come out if an expert group was presented with the challenge of:

"What would I do to the station/train/servicing vehicle design and deployment of staff" to achieve a turnaround in (let's say) 15 minutes. You are allowed to think the unthinkable".

By 15 minutes I mean from wheel stop to wheel start.

The other thing about Euston (current Avanti operation)is having long turnround times almost because you can; i.e. platform provision is relatively generous, so you may as well step down turnround times for resilience (plus helping diagramming etc.). Chances are they could do it a bit slicker, if they really had to, and do mitigations elsewhere to make this work.

Plus @Bald Rick's point about not taking this too far and paring to the bone too much.

 

[ABB125]

4 platforms, plus 2 through lines.

Thanks - I couldn't quite remember!

As far as I know, Interchange is 4 platforms, 2 islands. Tracks either side of the platforms and two through tracks, so six in total. Anything on the Curzon St spur doesn't have to interact until south of Interchange.

Makes sense - thanks

 

[Horizon22]

Many of the HS2 trains are not particularly long distance though.

Manchester in travel time terms is as far from Euston as Shoeburyness is from Fenchurch Street.

Birmingham, Manchester and Leeds won't really be long distance destinations any more.
There will be long distance trains with multi hour journeys, but many of the trains will not be.

Its more than manageable and many stations have slick operations to make this happen.

Now there are extra complications with long distance services which we assume are going to be of a certain standard, which might require a thorough clean, prep and a catering load for example. Saying that preparation of long-distance trains at terminals is now much easier than it was, especially with things like automatic reservations which you can see on the 80x fleets which was a substantial amount of time previously.

As @Bald Rick says 20 mins is manageable but not ideal (might be a peak service or only to Brum), but something like 30 mins would be perfectly adequate. As an aside, St Pancras on the HS side for Southeastern have turnarounds as low as 10-15 minutes. And the service is starting from scratch on dedicated lines so it mostly doesn't have to be woven into existing services and travel patterns.

 

[CW2]

Not sure it will, any touch point with the conventional network are going to be the pivots. HS2 will have such a high PPM expectation that Handsacre, Crewe etc are going to be very much first on the graph and on the day any other train even mildly late is going to be kept out the way.

I think possibly we are making the same argument in different terms.
Euston will have to be the point around which all HS2 trains will be timetabled. The timetable is determined by the platform availability / turnround times / station throat occupancy.
The HS2 stations at Birmingham, Manchester and Leeds will need to be secondary considerations as far as timetabling is concerned, especially if Euston is descoped. The flex of stopping patterns will need to be arranged around making Euston work first, and Birmingham Interchange / Birmingham Curzon Street / Manchester / Leeds after that.
I'm assuming that the start of HS2 services would trigger a major rewrite of the WCML and associated services, and in later years the MML and ECML too - as well as XC routes.
In all cases I would assume the HS2 services first on the graph, and everything else where best fits.

 

[Bald Rick]

Just as a matter of principle, it is good practice at terminal stations of long distance trains to not announce a departure until the inward service has arrived and the platform is clear of people. This is because of the large numbers of people with luggage / unfamiliar with arrangements etc and a need to keep everyone going the same way to minimise pedflow issues. Similarly, for long distance services a minimum time of 10 minutes is needed from being open to boarding before departure. These two factors alone drive a 15 min turnaround. Of course it can be (and is) done more quickly during disruption.

Euston will have to be the point around which all HS2 trains will be timetabled.

I’m not sure about that. Euston should be simple, even with one fewer platform - you can make the timetable work there by adjusting diagrams on turnarounds.

I suspect the HS2 timetable will revolve around Crewe or Piccadilly, at least for Ph1/2A.

Famously, the point from which the current WCML long distance timetable is built is Chester. (I’m not joking!)

 

[ABB125]

Famously, the point from which the current WCML long distance timetable is built is Chester. (I’m not joking!)

Is that due to lack of platforms at Chester?

 

[Ianno87]

Is that due to lack of platforms at Chester?

Letter in last month's Modern Railways from the relevant expert explained it; basically the number of West Coast Voyagers.

That basically sets the Euston > Chester > Euston round trip time at 5 hours (to not use too many sets) and the relative fixation of the Up and Down Chester paths. Those in turn fall on the opposite half hour to the Manchester-Crewe-Euston service (to give Crewe-Euston half-hourly), and thus the rest of the 20 minute pattern.

And there is a gap in Euston departures from xx33 to xx40 that corresponds to the Up Chester arrival each hour, to give it the most flexibility in which platform at Euston it can arrive in to make its return working at xx10.

 

[Nottingham59]

Clearly a 400m long train might take longer to alight

I agree with this. I really hope they've modelled passenger flow at Old Oak Common carefully. It would be a shame if the limiting factor on HS2 turned out to be OOC loading and unloading times.

 

[BayPaul]

Ulimately I would expect Birmingham International to need more platform space.
After all, worst come worst if a train is too late to make it's slot on the core, you can axe it and put the people on the next train south.

I could be wrong, but I don't think it will be possible to sensibly reverse a train at Birmingham Interchange, as the through lines are between the two island platforms. I believe that it will be possible to reverse a train at OOC, as there are 6 platforms available there, hopefully the two central ones will be reversible if necessary.

 

[Ianno87]

I agree with this. I really hope they've modelled passenger flow at Old Oak Common carefully. It would be a shame if the limiting factor on HS2 turned out to be OOC loading and unloading times.

Things like platform width, multiple platform entry/exits and how passengers are managed will help.

E.g. if you're in Coach 12, use entrance B and wait in the Purple Zone (or whatever).

 

[CW2]

I’m not sure about that. Euston should be simple, even with one fewer platform - you can make the timetable work there by adjusting diagrams on turnarounds.

I suspect the HS2 timetable will revolve around Crewe or Piccadilly, at least for Ph1/2A.

Famously, the point from which the current WCML long distance timetable is built is Chester. (I’m not joking!)

With 20 minute turnrounds for most services, and longer for the Scottish trains, on 18 tph you are pretty much full at Euston. Allowing for 400m long trains to clear the station throat, there is actually very little flexibility in the timetable. The only way you could create more flexibility would be by sub-standard turnround times, which is a performance risk. There is some flexibility to swap (say) a Manchester with a Leeds path or a Birmingham with a Newcastle, but part of the attraction of such a line is having an evenly spaced regular interval timetable (like the current WCML timetable). So having Euston to Manchester services at 00 / 20 / 40 minutes past every hour is a lot better than 00 / 15 / 43, for example.

I agree with this. I really hope they've modelled passenger flow at Old Oak Common carefully. It would be a shame if the limiting factor on HS2 turned out to be OOC loading and unloading times.

OOC dwells will be 2 minutes, and the trains are pick up only or set down only. All HS2 trains are (compulsory) fully reservable, so passengers should know which sector of the platform to be standing in for their specific coach when the train arrives.

 

[Ianno87]

I could be wrong, but I don't think it will be possible to sensibly reverse a train at Birmingham Interchange, as the through lines are between the two island platforms. I believe that it will be possible to reverse a train at OOC, as there are 6 platforms available there, hopefully the two central ones will be reversible if necessary.

Would probably be easier and less disruptive to divert to Curzon Street and send passengers forward on a waiting Euston service there without impacting a through dwell time.