Could Double Deck trains be viable in the UK?

[Spartacus]

Jeez, this chestnut's so old a starving squirrel wouldn't touch it.

 

[DanTrain]

not really... a bus has {usually} one door... meaning 70 people per door... an 8 car train will have 2 doors per carriage... meaning 16 doors... 1000 passengers / 16 doors = 62.5 passengers per door.

If you want to challenge (un)loadong times, go to Paris Gare du Nord, in particular the RER platforms. It can take a good 5 mins to unload and reload a full double-decker train, which must severely limit capacity. Now compare that to the wide, spacious, quick loading 700s that run on Thameslink!

 

[Teflon Lettuce]

If you want to challenge (un)loadong times, go to Paris Gare du Nord, in particular the RER platforms. It can take a good 5 mins to unload and reload a full double-decker train, which must severely limit capacity. Now compare that to the wide, spacious, quick loading 700s that run on Thameslink!

you mean the wide spacious quick loading 700s that everyone seems to hate? of course the natural question is how wide are the doors on a 700 compared to those on a DD RER train?

 

[Bald Rick]

you mean the wide spacious quick loading 700s that everyone seems to hate? of course the natural question is how wide are the doors on a 700 compared to those on a DD RER train?

About the same. But the passengers are generally more polite on the former.

Incidentally, there are plenty of people who are quite fond of the 700s, myself included. Better than any other stock that us MML TL commuters have had in living memory. Of course, this is an opinion, and others are available.

 

[DanTrain]

you mean the wide spacious quick loading 700s that everyone seems to hate? of course the natural question is how wide are the doors on a 700 compared to those on a DD RER train?

People hate 700s because they have horrible seats and they aren’t 319s. I don’t much like them myself, but they do a job and they do it well. They also just more pleasant than the RER!

 

[Bald Rick]

People hate 700s because they have horrible seats and they aren’t 319s.

A very few people, many of whom are on this site. I would suggest that a very significant majority of people who use them couldn’t care less - to them it’s a train.

 

[Teflon Lettuce]

A very few people, many of whom are on this site. I would suggest that a very significant majority of people who use them couldn’t care less - to them it’s a train.

well tbh I haven't been on a 700, I have however been on a new GWR intercity... and the seats on that were a lot more upright and less padded than an old 125.. so perhaps it's a case of the railways having forgotten the lesson learnt by their bus cousins... that almost upright seatbacks and minimal padding lead to very uncomfortable journeys... and eventually passenger loss...

anyway, back to the thread in hand.... it's been quoted above that the RER DD's take about 5 mins to load/ unload... but what is the OVERALL capacity{ in passengers per hr} on those lines with entirely DD stock compared to using SDs? that, after all, is the crucial statistic to consider...

 

[AM9]

well tbh I haven't been on a 700, I have however been on a new GWR intercity... and the seats on that were a lot more upright and less padded than an old 125.. so perhaps it's a case of the railways having forgotten the lesson learnt by their bus cousins... that almost upright seatbacks and minimal padding lead to very uncomfortable journeys... and eventually passenger loss...

anyway, back to the thread in hand.... it's been quoted above that the RER DD's take about 5 mins to load/ unload... but what is the OVERALL capacity{ in passengers per hr} on those lines with entirely DD stock compared to using SDs? that, after all, is the crucial statistic to consider...

The newest double deck trains on the RER, the 5-car class Z22500 each have a capacity of 1337 passengers over a length of 112m. That is 11.9 passengers per metre length. The single-deck 12-car class 700 trains on Thameslink have a capacity of 1754 passengers over a length of 240m. That is 7.3 passengers per metre length.
In dwell times, a suggestion is made that a fully loaded RER can take up to 5 minutes to empty. Lets say that well trained commuters can do it in half of that time, 150 seconds. So a minimum headway of dwell plus 60 seconds (3.5 minutes) would enable 17 trains per hour giving a total passenger flow of 17 x 11.9 = 202 passengers per hour per metre length. ISTR the class 700 has been specified to enable 50% of a full train to alight and 50% to board within a single 45 second dwell which is effectively a full unload in the same time. Allowing the same headway of dwell plus 60 seconds (1.75 minutes) would enable 34 trains per hour giving a total passenger flow of 34 x 7.3 = 248 passengers per hour per metre length.
So even ingoring the effects of the restricted width of the UK loading gauge, a line served by well designed single deck trains will have a higher passenger capacity than one served by double deck trains. In addition, the single deck line will not need an oversized infrastructure, nor will there need to be special provision in every car for persons of limited mobility, (the class 700 has total accessibilty in its centre cars but a totally flat floor throughout its length allowing the elderly and those with slightly limited agility to use the whole train).

 

[Teflon Lettuce]

The newest double deck trains on the RER, the 5-car class Z22500 each have a capacity of 1337 passengers over a length of 112m. That is 11.9 passengers per metre length. The single-deck 12-car class 700 trains on Thameslink have a capacity of 1754 passengers over a length of 240m. That is 7.3 passengers per metre length.
In dwell times, a suggestion is made that a fully loaded RER can take up to 5 minutes to empty. Lets say that well trained commuters can do it in half of that time, 150 seconds. So a minimum headway of dwell plus 60 seconds (3.5 minutes) would enable 17 trains per hour giving a total passenger flow of 17 x 11.9 = 202 passengers per hour per metre length. ISTR the class 700 has been specified to enable 50% of a full train to alight and 50% to board within a single 45 second dwell which is effectively a full unload in the same time. Allowing the same headway of dwell plus 60 seconds (1.75 minutes) would enable 34 trains per hour giving a total passenger flow of 34 x 7.3 = 248 passengers per hour per metre length.
So even ingoring the effects of the restricted width of the UK loading gauge, a line served by well designed single deck trains will have a higher passenger capacity than one served by double deck trains. In addition, the single deck line will not need an oversized infrastructure, nor will there need to be special provision in every car for persons of limited mobility, (the class 700 has total accessibilty in its centre cars but a totally flat floor throughout its length allowing the elderly and those with slightly limited agility to use the whole train).

which then begs the question.... why are DD trains seen as the solution to adding capacity in foreign climbs?

 

[AM9]

which then begs the question.... why are DD trains seen as the solution to adding capacity in foreign climbs?

I think that the generous structure gauge has over the years allowed them to pursue the double deck solution to more capacity per train without evaluating the true capacity, i.e. the capacity of a line. The demand for space on the RER was far below current levels when the decision was made to bore oversize tunnels. Here, the option for higher and (more importantly wider) rolling stock on the UK's first two 'RER like' railways just wasn't there, at least not at a justifiable cost. So the UK has worked at the affordable features of a high-density railway that do play a significant role in delivering the continuous high throughput with typical passenger and driver behaviour, i.e.:

very short dwell times/headways - adaptive signalling systems based on multiple balises and variable working headways
consistent running characteristics - brought with ATO increasing the opportunity to reduce headways
train designs able to take advantage of the minimal headways - better acceleration and braking
maximised passenger capacity - better standing areas
maximised passenger mobility - through gangways with enhanced passenger information systems
maximised ingress/egress speeds - wide fast automatically opening doors​

Many of those features have been incorporated in one or more of the latest designs of high density EMUs, (classes 700, 707, 345) and their related classes currently in manufacture.

 

[Bald Rick]

which then begs the question.... why are DD trains seen as the solution to adding capacity in foreign climbs?

@AM9s otherwise excellent analysis makes two assumptions:

a) dwell time for the RER is based on anecdote
b) it makes no allowance for certain RER stations having two platforms per direction of line, which allows longe4 dwells.

RER Line A manages 26tph through the core in high peak hour.

Nevertheless it is all about dwell time, and single deck trains can simply have more and bigger doors. There are plenty of new and recent high capacity heavy rail metro systems around the world that have stuck with single deck, eg the Munich S-Bahn

 

[AM9]

@AM9s otherwise excellent analysis makes two assumptions:

a) dwell time for the RER is based on anecdote
b) it makes no allowance for certain RER stations having two platforms per direction of line, which allows longe4 dwells.

RER Line A manages 26tph through the core in high peak hour.

Nevertheless it is all about dwell time, and single deck trains can simply have more and bigger doors. There are plenty of new and recent high capacity heavy rail metro systems around the world that have stuck with single deck, eg the Munich S-Bahn

Yes I agree my figures were based on a posted anecdote and I have witnessed headways closer than I've ever seen in the UK, except on LU (albeit some time ago of trains lining up at Auber with one in the platform, doors open, and the next one waiting with its cab peeping out of the tunnel ready to draw forward asap). The fact was that it was (and always will be) the passengers' behaviour that will determine the dwell time so two wide doorways serving two coach lengths of passengers cannot possibly be faster than two wide doorways serving a single flat level coach. That's ignoring the slower passengers climbing/descending the stairs.
Whilst typing the measures taken in the UK it did cross my mind how the next level of line capacity enhancement (whilst maintaining a two track formation) would be duplication of platforms. Much discussed here for the Thameslink core, - more as a means of handling passenger volumes on St Pancras (STP) platforms A & B than the total rate of passengers carried on the line, creating the space for additional tracks there and Farringdon (ZFD) would not only be very expensive, but also technically challenging because of the site constraints, causing severe disruption in construction. Only Blackfriars could relatively easily be modified that way, but I suspect that the ingress/egress there is not as high as STP is and ZFD will probably be when Crossrail opens.
Crossrail itself could of course be built with duplicate platforms in the core but given the very high costs associated with station boxes, it would have a major impact on the overall cost and without giving the existing platforms at the major interchanges similar treatment (if possible), there would be a risk of saturation from the combined enhancements of high density fast-unloading trains depositing a passenger load that could not be dissipated.
As you say, new systems are still being built with single deck trains and even in NYC, much of the busiest commuter lines are still single deck. It's only the longer outer-suburban terminating lines that seem to be relying on double deck for the highest capacity per train.

 

[Dr Hoo]

I do suspect that many of the drivers for double deck commuter trains elsewhere relate to cost of the trains and the operation rather than overall system capacity.

Besides needing fewer trains you also need fewer depot tracks and fewer crews. Particularly where tracks have to shared with freight and the infrastructure is not even owned by the commuter authority it is ‘desirable’ to operate as few trains as possible.

A single long dwell at a single city terminal is a relatively low price to pay. Where there are multiple station stops the pain rapidly ratchets up.

 

[30907]

I think you are near the mark with it being down to economics!

Paris and Brussels are the two cities that make most use of full length DD trains.
In Paris the cross city routes that were built for them have duplicate platforms in each direction, while in Brussels both Nord and Midi have ample platforms to allow 3 minute dwell times. The alternative would have been even more new trackage.
Although the routes cross the city, at peak times most passengers will be city centre commuters so not many boarding in the morning/alighting in the evening at main stations, which simplifies matters.

By contrast, DB and their competitors seem to use DD stock to allow shorter trains (compared with their predecessors - the IC2 is an obvious example, but it's true of lots of RE routes) - this means that platforms can be shortened when they are altered, as many need to be, to approved "accessible" heights. Of course they also save on train weight per head etc.
The longest DD I've seen in Germany is 7x26m (Metronom, Hamburg)

 

[AM9]

So on the face of it, it is very unlikely that the infrastructure on the classic network will get investment to allow double deck trains to run, especially where line capacity maximisation is the object. They just don't provide any cost-effective benefit to services that are already close to their capacity, especially on high density commuter routes.

 

[Teflon Lettuce]

out of interest... what impact would one way flow through carriages make to dwell times on DD's? what about platforms on both sides of the track? what about both together?

as I said before I find it strange that this country seems to do things the absolute opposite to the continent... we have DD buses and SD trains... and won't contemplate SD buses/ DD trains

 

[30907]

out of interest... what impact would one way flow through carriages make to dwell times on DD's? what about platforms on both sides of the track? what about both together?

as I said before I find it strange that this country seems to do things the absolute opposite to the continent... we have DD buses and SD trains... and won't contemplate SD buses/ DD trains

I think we have plenty of SD buses, even on routes without low bridges. And DD service buses aren't totally unheard of on the mainland- Berlin for starters, Paris historically.

 

[AM9]

out of interest... what impact would one way flow through carriages make to dwell times on DD's? what about platforms on both sides of the track? what about both together?

as I said before I find it strange that this country seems to do things the absolute opposite to the continent... we have DD buses and SD trains... and won't contemplate SD buses/ DD trains

Buses don't run with minimum headways

a) because they don't run every 10 seconds, even on the busiest routes
and
b) because they can overtake when one is delayed by an overcrowded one.​

We also have smaller and more crowded roads so DD buses have been used to carry more passengers. Many of our railway bridges have enough clearance for DD buses. Less so on most of mainland Europe.

Conversely, we have a much smaller loading gauge on our railways than most of mainland Europe so other than as a proof of concept (which O V Bullied proved wasn't viable) our infrastructure isn't suitable for DD trains. Neither of those things are 'strange' to me as road and rail development are not directly connected in any way.

 

[DanTrain]

out of interest... what impact would one way flow through carriages make to dwell times on DD's? what about platforms on both sides of the track? what about both together?

If by this you mean the top deck empties at end A and loads at end B whilst the bottom deck empties at end B and loads at end A, then probably a bit. The problem remains though that the staircases are cramped and severely limit the amount of passengers who can leave the top amd bottom decks per minute. Also, the higher proportion of seated passengers, whilst more confortable, take longer to get up and gather their belongings and then leave than standing passengers do.

as I said before I find it strange that this country seems to do things the absolute opposite to the continent... we have DD buses and SD trains... and won't contemplate SD buses/ DD trains

We do have thousands of SD buses in this country, and DD trains on regional/IC routes are limited by infrastructure - I personally think they’re a good idea on some outer suburban and InterCity routes, but we simply don’t have the infrastructure (and since HS2 has to be classic compatible, this is unlikely to change!)

 

[swt_passenger]

So on the face of it, it is very unlikely that the infrastructure on the classic network will get investment to allow double deck trains to run, especially where line capacity maximisation is the object. They just don't provide any cost-effective benefit to services that are already close to their capacity, especially on high density commuter routes.

Exactly what the DfT/NR report says, that I posted a few days ago.

 

[Peter Mugridge]

If you want to challenge (un)loadong times, go to Paris Gare du Nord, in particular the RER platforms. It can take a good 5 mins to unload and reload a full double-decker train, which must severely limit capacity.

RER D at Gare du Nord has an issue in that it shares the tracks between there and Chatelet with RER B, which takes priority as it serves both the Paris airports. RER D has a lot of dwell time built in on the core section between St Denis Stade de France and Gare de Lyon to allow for this. I should point out that they have already started planning work on an additional pair of tunnels between Gare du Nord and Chatelet in order to allow each line to have dedicated track - once this is open the core section of RER D will see a service frequency boost and also reduced journey times*.

Elsewhere in Paris, including other parts of RER D, the Z20500 units load and unload with pretty much the same dwell time as everything else.


*By the time it's open, though, the RER D fleet will have been replaced with Z57000s.

 

[apk55]

What about on the high speed lines, which are built to a much larger loading gauge. HS1 could be used by double deck stock to Paris etc without problems. Loading time would not be a problem. HS2 should be built with with the possibilities of running double deck stock on the busiest sections such as

 

[swt_passenger]

What about on the high speed lines, which are built to a much larger loading gauge. HS1 could be used by double deck stock to Paris etc without problems. Loading time would not be a problem. HS2 should be built with with the possibilities of running double deck stock on the busiest sections such as

HS2 itself is being built to allow for double deck stock. Fact. Why do people keep suggesting it as if it’s not?

 

[Chester1]

HS2 itself is being built to allow for double deck stock. Fact. Why do people keep suggesting it as if it’s not?

Because HS2 is not popular and therefore the more positive aspects are often ignored. I think there is a reasonable chance of double deckers running between London and Birmingham/Manchester/Leeds in 2033 if the demand during the first couple of years of opperation is high. The decision on rolling stock for 2033 doesn't need to be made for another decade.

 

[JonasB]

which then begs the question.... why are DD trains seen as the solution to adding capacity in foreign climbs?

If you can't run more trains, doubledeckers can be cheaper than extending the platforms.

 

[DenmarkRail]

HS2 itself is being built to allow for double deck stock. Fact. Why do people keep suggesting it as if it’s not?

Because HS2 trains will be built for classic compatible lines, so it'll all be very good up until Lichfield, but once it enters the WCML, it'll be bye bye bridges. Thats why they won't be DD

 

[swt_passenger]

Because HS2 trains will be built for classic compatible lines, so it'll all be very good up until Lichfield, but once it enters the WCML, it'll be bye bye bridges. Thats why they won't be DD

I didn’t comment about the intended HS2 trains. In response to apk55’s point above, I said that HS2, the infrastructure, was being built to allow for double deck trains. Still a fact.

 

[AM9]

If you can't run more trains, doubledeckers can be cheaper than extending the platforms.

It's not about running more trains, the task is to carry more passengers. As this discussion has largely demonstrated, double deck trains may hold more passengers per train but normal passenger behaviour negotiating narrow stairs, queueing to get on or off and turning up on the platform in larger numbers means that the higher capacity trains cannot run as frequently as well engineered single deck trains which have a slightly lower capacity per train, (i.e. a double deck train will only give about 40% more capacity than a single deck).
As for your assertion that running double deck trains is cheaper than extending platforms, that is only true if the line structure gauge will allow passage of those trains. In the UK, it would mean rebuilding most overbridges, and many underbridges, moving every platform away from the running rails and as for tunnels! Despite the need for some stations to be modified or even rebuilt, longer platforms is nowhere near as expensive unless you only consider running double deck trains across the Fens and the Somerset levels.

 

[Teflon Lettuce]

It's not about running more trains, the task is to carry more passengers. As this discussion has largely demonstrated, double deck trains may hold more passengers per train but normal passenger behaviour negotiating narrow stairs, queueing to get on or off and turning up on the platform in larger numbers means that the higher capacity trains cannot run as frequently as well engineered single deck trains which have a slightly lower capacity per train, (i.e. a double deck train will only give about 40% more capacity than a single deck).
As for your assertion that running double deck trains is cheaper than extending platforms, that is only true if the line structure gauge will allow passage of those trains. In the UK, it would mean rebuilding most overbridges, and many underbridges, moving every platform away from the running rails and as for tunnels! Despite the need for some stations to be modified or even rebuilt, longer platforms is nowhere near as expensive unless you only consider running double deck trains across the Fens and the Somerset levels.

The thing is, if you design a double deck train with WIDE stairways and WIDE doors... add in a one-way flow through the carriage... and even add platform on BOTH sides of each track... then the DD train will load/ unload as fast, if not faster, than an SD train... as has been amply proved with buses the speed of boarding/ alighting has nothing to do with whether a vehicle is sd or dd but more to do with how many doors, position of doors and width of doors... in fact, as has been proved with buses an all seated DD will have dwell times shorter than a standee sd with the same number of passengers... due to the passengers wishing to board/ alight NOT having to fight their way past standees...

you say that DD trains can only improve capacity by 40%, but if a line is at absolute saturation as regards tph physically able to use the line then surely putting DD's on and INCREASING capacity by 40% should be welcomed...

out of interest... IF a line needs more capacity what is more expensive... converting the line to DD with the existing number of tracks? or doubling the number of tracks? Considering land prices in the places where this is likely to be needed then I'm guessing DD is the way forward...
but then again that would just be too "continental" for us Brits!

 

[Bletchleyite]

I think we have plenty of SD buses, even on routes without low bridges. And DD service buses aren't totally unheard of on the mainland- Berlin for starters, Paris historically.

The main reason is that standard bridge heights in the UK are higher than most European countries. This is curiously the exact opposite of the railway system. It doesn't really have anything to do with passenger flow etc.