Is the Thameslink idea flawed

[Minstral25]

As a regular commuter I am begining to wonder if some of the aspirations for Thameslink are flawed.

I think it is a good idea to try and put regular trains through the London core so there is an opportunity for people to get through London quickly. In effect a fast tube line where the trains go much further.

However what are those platforms at City Thameslink going to be like when thousands of passengers are waiting for trains and there is a small delay. I know what it is like at London Bridge on a platform for a single train going south to Brighton, at each door around 20/25 people waiting for the train to pull in.

At City Thameslink they will be waiting for a Brighton service, a Sevenoaks service, a Guildford service, a Redhill/Gatwick Service, an Eastbourne service, a Dartford Service - each of which will pass within around 15 minutes. (I think 24 trains per hour is 6 per 15 minutes)

I predict pandemonium and unsafe numbers on Platforms - currently they are spread around 6 platforms - this will push hundred's of people to a single platform for all of those trains.

To make it worse, train doors are often open a few minutes early at platforms so early arrivals can sit on the train and wait and later arrivals can get straight on. Not so on Thameslink where the train pulls in and everyone has to get in at once.

Dwell times will be too long for 24 trains an hour due to overcrowded stations

And if there is a problem and a 15 minute delay - are the platforms going to be wide enough or are passengers going to have to wait outside to be let in the station?

Northbound is no better with Bedford/Perterborogh/Cambridge/Hertford etc being served.

 

[gordonthemoron]

It's a pity Thameslink stations don't have the capacity for platforms on either side of the train, so passengers alight on one side and board on the other. This works successfully at Hauptbahnhof, Karlsplatz & Marienplatz stations in Munich.

 

[Metroland]

To run 24 trains per hour you would need dwell times of around 30 seconds, this is fairly typical of London Underground lines. The Central, Northern and Victoria lines among others all regularly carry this sort of traffic often with standing room only with trains of excess of 1000 people on board, this making capacity around 30,000pph. This is aimed to be increased by at least 20% by the introduction of moving block reducing headways further.

Lines around London Bridge and parts of the GEML also carry this sort of traffic in the peak.

Passenger dwell times is the chief reason for capacity constraints, rail has the advantage of having very long vehicles where passengers can get in and out of very quickly, making its ultimate capacity higher than other forms of transport.

Other than reducing the headway, the the next step is to lengthen trains.

But Thameslink would be nowhere near the flow of many metro lines, there are parts of the Hong Kong and Moscow metros that carry over double that number per hour, with proposals for parts of the Hong Kong system to carry 100,000pph per direction.

 

[tbtc]

To run 24 trains per hour you would need dwell times of around 30 seconds, this is fairly typical of London Underground lines. The Central, Northern and Victoria lines among others all regularly carry this sort of traffic often with standing room only with trains of excess of 1000 people on board, this making capacity around 30,000pph

True. But on the London Underground, most people want to board the next train (yeah, there are branches on lines like the Northern, but most of the passengers will use the next service to arrive at a platform.

On Thameslink, a good number of passengers will be for stations north/ south of the "core" section, which is going to mean a lot of people for Brighton/ Cambridge etc waiting around getting in the way when the next service to Luton/ Eastbourne arrives in the platform.

Also, whilst most London Underground delays are "self contained" (e.g. a breakdown/ delay on the Northern Line shouldn't affect the Central Line), putting services to lots of different destinations though this one two tracked tunnel in central London means that a delay at Cambridge could mean knock on delays at Gatwick/ Sevenoaks etc as one train running into the core section late could knock a lot of other tightly timed trains out of their time slot.

 

[Metroland]

Many 'cross city' services exist in France, German, Holland, Switzerland, to use a few examples with a similar high intensity metropolitan service, ultimately running out to the outer suburban or beyond areas.

You are quite right to point out time keeping will be key to maintaining service reliability, as it's harder to do that where the service is not self contained. But looking at foreign examples, its not impossible.

 

[David Sinnett]

Interesting to see how many trains are delayed now at somewhere like ST.P before they go to 24tph. If they can't maintain the current flow to time what chance have they next year...whenever.

 

[SWT Driver]

I'd compare the idea to the same deluded & scatty idea that Railtrack had with WCML with moving block signalling.

It's fantasy, it'll never work because people don't act the way some crap computer simulation thinks they will, I mean it doesn't & won't take into consideration the scatty office woman dropping her shoe on the track, a idiot teenager putting a drinks can in the door, so that it doesn't close & so on.

 

[The Box Photter]

I'd compare the idea to the same deluded & scatty idea that Railtrack had with WCML with moving block signalling.

It's fantasy, it'll never work because people don't act the way some crap computer simulation thinks they will, I mean it doesn't & won't take into consideration the scatty office woman dropping her shoe on the track, a idiot teenager putting a drinks can in the door, so that it doesn't close & so on.

Spot on.

 

[jopsuk]

I appreciate it would have made the junction on the south side of the Thames even more complex, but surely there would have been potential for improved resiliance (and longer dwell times) if the TL lines at St P and Blackfriars had been arranged as four lines, with each station having a northbound and southbound wide island platform each. At St Pancras, the "outside" faces could have handled the GN services, with inner going MML, and at Blackfriars (say) inner to London Bridge, outer to E&C.

Then, increase dwell times at those stations, and instruct passengers at City and Farringdon to catch the first train and change at St Pancras or Blackfriars.

Would have required substantially more infrastructure work, I do accept, creating much more cost. However, it would surely have stood much more chance of being able to deliver the promised frequency.

As a thought- if TL is getting an eventual single common fleet, what about platform edge doors through the core? Surely the 8 car/12 car problem could be overcome by some sort of SDO system (for the platforms)?

 

[yorkie]

The concept is not flawed at all! However our implementation of it is seriously flawed due to the ridiculous nature of the stations, St Pancras is a prime example and there is no excuse. At places like St Pancras and Blackfriars there should be island platforms, with each platforms dedicated to services to/from a particular route.

Dwell times will be too long for 24 trains an hour due to overcrowded stations
.

I agree totally. Can DafT not see this?

Only an incompetent fool would suggest 24tph could be achieved.

Do the DafT still believe that 24tph can be achieved? It wouldn't surprise me if they do, sadly.

 

[Kingfisher200262]

I agree totally. Can DafT not see this?

Only an incompetent fool would suggest 24tph could be achieved.

Do the DafT still believe that 24tph can be achieved? It wouldn't surprise me if they do, sadly.

I'm not sure if it's in the public domain, but I believe that the spec has been revised back to 20 tph.

Question is, would a train builder that could have offered a 20 tph product, but decided not to offer because they couldn't do 24 tph feel sufficiently aggrieved to shout foul and cape the tender process???

 

[SWT Driver]

The concept is not flawed at all! However our implementation of it is seriously flawed due to the ridiculous nature of the stations, St Pancras is a prime example and there is no excuse. At places like St Pancras and Blackfriars there should be island platforms, with each platforms dedicated to services to/from a particular route.

I agree totally. Can DafT not see this?

Only an incompetent fool would suggest 24tph could be achieved.

Do the DafT still believe that 24tph can be achieved? It wouldn't surprise me if they do, sadly.

So where would you build such island platforms then, there's nowhere for them to be built, besides it's impractical & restrictive.

 

[yorksrob]

I suppose if you had one island platform and one single platform at each end (say Blackfriars and St Pancras) you could make the centre platform reversible and increase flexibility greatly without increasing the station footprint that much. E.g. the box at Blackfriars could have been built wider and Blackfriars could spill onto where the bay platforms used to be.

 

[Minstral25]

So where would you build such island platforms then, there's nowhere for them to be built, besides it's impractical & restrictive.

I would have built 2x island platforms in the St Pancreas Box to receive/dispatch trains to BedPan lines and GN lines - plus I suspect they may be enough space at Blackfriars to do a two island platforms as well (maybe E&C route and London Bridge route).

But Island platforms will not solve the problems of overcrowded platforms - just mean operational flexibility will be better. Farringdon and City Thameslink are going to be massive problems.

 

[Metroland]

I'm not sure if it's in the public domain, but I believe that the spec has been revised back to 20 tph.

Question is, would a train builder that could have offered a 20 tph product, but decided not to offer because they couldn't do 24 tph feel sufficiently aggrieved to shout foul and cape the tender process???

Wasn't it something to do with knocking the idea of ATO on its head?

Actually 30 trains an hour was achievable out of Liverpool Street at the turn of the last century with steam locomotives and semaphore signals. 37 trains per hour are run on many metro lines with high degrees of punctuality. Notwithstanding the point about a diffuse network, its not quite in the land of the fairies some of the armchair operators are describing.

 

[O L Leigh]

But Liv St has a 6 track throat.

I personally can't see the proposed frequency working out in practice, especially when the services arrive from and depart to such a wide range of places. I doubt there was enough space to make the St Pancras "box" wide enough to accomodate more platforms, so throughput will depend on sorting all the trains out before they enter and leave the core.

O L Leigh

 

[Metroland]

That would be on each line, Google LNER Jazz trains sometime. 16 four wheeled coaches providing 848 seats, 24 trains an hour would be normal climbing the Bethnal Green Bank on the down suburban line. Sykes lock and block semaphore signalling, Buckjumpers and later LNER N7s. Wonderful.

 

[O L Leigh]

I've tried the Google approach and can't find the wheat for the chaff. Perhaps if you could provide a link...?

O L Leigh

 

[Daniel]

With regards to this whole "24tph" m'lark - the Jubilee Line currently does operate a 24tph service - (this is before all this fancy TBTC and/or ATO comes in and increases that).

Dwell time targets are usually around 32-36 seconds.

This considering that the line is two tracked.

And the late running is awful.

From my time with the drivers over there I know how bad it can get - me and my driver very very rarely got off on time. But theres no noticeable impact to customers for overall late running on LU, because they don't advertise a timetable. If every train driver is running 1 hour late, the customers still have the same frequency of service at the same time.

But on NR, when the trains have advertised departures, late running will be noticed - and this is before all the complications of different destinations etc - which itself won't help as it will severely handicap you with trying to reform the service - (as on LU, each train will take the same amount of time to go to xxx and back).

 

[Metroland]

I've tried the Google approach and can't find the wheat for the chaff. Perhaps if you could provide a link...?

O L Leigh

There's not an awful lot there I must admit, you would need to consult books on the subject. But there was an excellent article in the November 2006 railway magazine.

Very briefly the Jazz service of the Great Eastern Railway had its roots in the 1864 Act of parliament, which authorised the replacement of the old Bishopsgate terminus with Liverpool Street. The Great Eastern railway compensated those that had their houses demolished a two pence return from their homes out on parts of Walthamstow and Lower Edmonton.

Commuting became extremely popular and cheap housing spread out along the Chingford and Enfield town lines. With the fares held down, soon intense overcrowding became commonplace.

In 1901 the Great Eastern looked to electrify the line, it was claimed electrification could offer a more intensive service. But because the fares were capped, the GER could make no return from the service and their locomotive superintendent, James Holden was given the job of designing a steam loco that could get to 30mph in 30 seconds with heavy trains. Anyway, the Decapods he designed were too heavy for the lines and eat coal. But the point was made and 1886 'Buckjumpers', an 0-6-0 tank were used on the new services.

In preparation for the new services extensive research was conducted, and it was reckoned with improved track and signalling steam could still equal electric (at least at that time).

The critical length was the mile between Liverpool street and Bethnal green, with signals difficult to spot in the smoke-filled tunnels below Bishopsgate, followed by a 1 in 70 climb. So to help drivers, banner repeaters were fitted. Some signals were relocated and Sykes lock and block was fitted with some track circuits.

Trains weighed to simulate crush loading conditions were timed by stop watch and to mark exact spots were additional banner repeaters or stop signals were needed, engineers threw out bags of whitener to mark the spot!

Anyway, it was possible to increase suburban line capacity by 50% to 24 trains per hour with 2-1/2 min headways on semaphore signals.

6 minute turnarounds were provided at Liverpool Street, so information had to be clear and modifications to passenger flow in the station were done. Cantrails above second class coaches were given a blue band, first class bright yellow and so on. The trains were said to be 'Jazzed up', hence the nickname.

And the cost? £80,000 in 1920 prices.

 

[O L Leigh]

Thanks for adding details.

*Further Googling*

Ah, I think you mean "Buckjumpers". These were GER S56's classified as J69's by the LNER.

O L Leigh

 

[Metroland]

You are quite correct, typo on my part!

 

[Richardr]

The other issue is that signage to customers will need to be accurate. At the moment, St Pancras has been accurate, in my experience, but some of the stations south of that are not always accurate as to which train is about to arrive - especially Blackfriars. If all the trains are similar and in the FCC logo, then the chances of getting on the wrong train are high.

Only last week, and I know it was an emergency timetable, but the train I got north from Blackfriars had Kentish Town on the front whereas the announcements said Bedford.

Often trains due at Blackfriars or City Thameslink dissappear from the boards at the very last minute, with no information as to whether they are still due or cancelled..

Such confusion will not work once the addditional destinations north are added.

 

[mackenzie_blu]

Just put some actua;y figures in.

Time from wheels stopping (with the doors opening, driver closing them and making sure the doors are safely closed) to wheels moving again is 50sec. Yes 50 seconds in peak.

This doesn't actually sound a lot does it but I did some timing of this at City and Blackfriars on a regular basis and found most 319 and 377 do it in around this time so its possible.

What does worry me is have anyone seen what happens before the Ashford/Rochester service? Commuters line the platforms by the yellow line blocking all other commuters so image what its like when you start mixing all these different lines on platforms that aren't ideal for all this. I await to see the final design of these trains as I fear the ones we are going to get are going to make matters worse. The easy boarding/exiting of the Alstrom bid made me back this train above the Siemens/Bombardier bids.

 

[Kingfisher200262]

To run 24 trains per hour you would need dwell times of around 30 seconds, this is fairly typical of London Underground lines. The Central, Northern and Victoria lines among others all regularly carry this sort of traffic often with standing room only with trains of excess of 1000 people on board, this making capacity around 30,000pph. This is aimed to be increased by at least 20% by the introduction of moving block reducing headways further.

Lines around London Bridge and parts of the GEML also carry this sort of traffic in the peak.

Passenger dwell times is the chief reason for capacity constraints, rail has the advantage of having very long vehicles where passengers can get in and out of very quickly, making its ultimate capacity higher than other forms of transport.

Other than reducing the headway, the the next step is to lengthen trains.

But Thameslink would be nowhere near the flow of many metro lines, there are parts of the Hong Kong and Moscow metros that carry over double that number per hour, with proposals for parts of the Hong Kong system to carry 100,000pph per direction.

I would have thought that one of the main constraints on Thameslink running 24 tph is the flat junctions that other traffic cross. It's not really fair to compare Thameslink with a tube line, which is just that; an end to end railway, with nothing else interfering with it.

 

[jopsuk]

Aye- the Sub Surface lines have much lower frequncies, even when you tot up cumulative frequencies (such as Baker Street-Aldgate stations) due to the conflicting movements.

 

[Metroland]

Hence why 24 is more reasonable than the 37 possible on metro lines.

Don't forget Crossrail will be even more intense though.

 

[Drsatan]

37 trains per hour is only possible on metro services (e.g. the Circle line) because trains on those lines tend to be self-contained and won't be affected by delays from other services, which will have a knock-on effect on punctuality.

Likewise, with Thameslink, I can imagine a service to Bedford being delayed by a service to Cambridge, and services south of the Thames being delayed by other trains on the Brighton line. If something goes wrong (e.g. a signal failure) then due to the intensity and complexity of services in the "core" between Blackfriars and St Pancras then the timetable's going to go out the window.

Has nobody at DafT thought that most of the passengers boarding and alighting at St Pancras will be carrying luggage, since most passengers getting off there will be catching domestic or international services? 30 second dwell times won't be long for a family group (who may be foreign and hence not have a good grasp of the English language) to either get on or off. I expect that a few suitcases will become stuck in doors as a result which will delay the service even more. By all means increase dwell times at St Pancras to a minute or more, but then it wouldn't be possible to run 24 trains per hour.

 

[gordonthemoron]

who may be foreign and hence not have a good grasp of the English language

you'd be surprised

 

[Metroland]

The DfT and consultants spend many years modelling new transport schemes, passenger flow around stations, and modelling luggage (those wheeled suitcases are considered to be a real pest!), entrances and exits. Data from real life experience is fed it to a lot of computer models and there are many modelling software packages that look at passenger flow, signalling, train performance, train driving performance and so on.

Passenger modelling is a very important part of this, many rail stations are full to the brim of passengers and can cause a lot of delay in peak periods, Birmingham New Street being one such example. This is all taken into account when assessing new schemes.

Indeed a whole profession is dedicated to it and conferences are held.

http://landor.co.uk/conferences/movingit/programme.php

http://www.landor.co.uk/transportmodelling/index.php

http://www.amazon.co.uk/Modelling-Transport-Juan-Dios-Ortuzar/dp/0471861103