Great Western Coffee Shop

Various of the posts on changing passenger journey profiles, fare rises, overcrowding and more ... have lead me to do a bit of thinking about where we are, what has happened in the last 20 years, and where we might be headed 20 years from now.

Some lateral thinking ...

According to the ong Term Passenger Rolling Stock Strategy for the Rail Industry (http://www.raildeliverygroup.com/files/Publications/2016-03_long_term_passenger_rolling_stock_strategy_4th_ed.pdf) from the Rail Delivery Group dated March 2016, which appears to be the fire behind some of the smoke we're seeing saying how shocking it is that trains are so old, there were 12,968 passenger carrying vehicles at that date (page 12 - it's a long report!)

Compound growth rate in the 20 years from 1995 to 2015 was 3.8% per annum, meaning that passenger numbers in the final year of that period were 210% of what they were in the first year.  This growth is shown graphically on page 7 of the report and broadly corresponds to the growth in the number of passenger journeys over the same period which I showed at http://www.firstgreatwestern.info/coffeeshop/index.php?topic=17797.msg206896#msg206896 .

The increase in passenger numbers has not been matched by an increase in national fleet size:
"The 110% increase in passenger-miles in the 20 years between 1995/6 and 2014/15 was achieved with an increase of just 15% in the size of the total national passenger fleet."

How has this enormous increase been achieved?   The report says
"This major increase in fleet utilisation efficiency since 1995 was principally achieved by much improved marketing and utilisation of spare off-peak capacity. In addition, changes have been made to the type, configuration and functionality of much of the national fleet. These have increased the average capacity per vehicle."
but the passenger view might well be that part of the gain has been by loading trains with passenger numbers over and above what they can really hold in comfort, and indeed the report's attitude celebrates the economic gain of getting lots more people onto only a few more trains.

I also note
"The average fare paid per passenger mile has risen by only 5% in real terms over this 20-year period."
which is pretty astonishing when you consider that loading per carriage has gone up by 82%.  It means that in real terms, income per carriage has gone up by no less than 92% in real terms.  Gosh - I would love to be able to raise the income of our hotel by 92% per room in real terms, but then (perhaps) this rise is the industry digging itself out of a hole which was highlighted by the Serpell report answering the question "which part of the network is profitable", but which looked at answers based on cutting rather than promoting growth within existing resources.

Looking ahead ...

The genie is out of the bottle! It can't keep on growing as it has done.  A further 110% passenger grown won't fit on a further 15% extra fleet - unless you get more and more people into each train by replacing large areas of seating by standing room, or increasing the aggressive price differentials between trains that are now full ("peak") and trains that still have capacity ("off peak").  Modern technology would allow for some of this, with dynamic pricing. New trains will be able to sense their loads and advise people automatically that there's "room up front".  And longer operating hours (the 24 x 7 railway) will allow even better stock utilisation; it will also overcome the industry's headache of having to 'taxi' staff out to trains parked overnight some distance from depot.  Electric trains will run somewhat faster, and have considerably shorter service times and will be much much more reliable (see page 24) - again that's a gain.  Replacement of final slam doors, and redesign will help reduce station dwell times.

The report looks at a rise over the next 20 years from 12,968 vehicles to (median case) 19,476 - that's almost exactly a 50% increase.  I conjecture that a significant proportion of than increase is necessary to overcome overcrowding that already exists - demand has been running ahead of supply.  And I note that even if growth drops from 3.8% per annum to a half of that(1.9%), 43% extra capacity over the next 20 years would be required to stand still! (Yes, in saying that I have ignored the efficiency issues described above, but then I have also ignored the increased space needed for the disabled loo ...)

With an further increase of 50% in the number of vehicles on the tracks in the next 20 years, and (we hope) with each of those vehicles spending more time running than at present, the questions of track, platform, station, and indeed access to station capacities arise.

In places, there are some simple solutions to be had to increase the capacity. 2 car trains run between stations which can all accommodate 4 or 5 car trains for starters. And you, dear Mr Rail Delivery Group, don't even have any increase driver or conductor costs. For sure, longer trains take a bit longer as they snake their way through speed restrictions - the head can't accellerate while the tail is still on a speed limit, but we're not talking huge long trains here.

In some places, lines are already at or near capacity, or so we're told. But with a standardisation of trains to the same performance envelope and stopping patter, and a lengthening of them all to the line's maximum length, you can get more people along the flow.  I'm thinking of Paddington to Reading, where you could routinely run 16 x 10 coach trains every hour. Heathrow served from Crossrail and the relief lines (they're not "slow" lines, are they?) and then in every 15 minutes:

A 10 car train via Bristol Parkway - to Swansea, or to Temple Meads

A 10 car train via Reading West - in each hour
   one train express to Plymouth and Penzance
   two trains dividing and serving Basingtoke, and different stopping patternd to Westbury / Bedwyn
   one train fast to Westbury, then stations to Newton Abbott dividing to Paignton and Plymouth

A 10 car train via Didcot
   two trains on to Oxford with 5 coaches forward alternately to Banbury and Hereford via Worcester
   two trains dividing at Didoct, 5 coaches forward to Oxford ...
   ... one portion on to Swindon, Gloucester, Cheltenham, Worcester, Bromsgrove and Birmingham
   ... one portion on to Swindon, Trowbridge, Westbury, Yeovil, Dorchester and Weymouth

A 10 car train via Bath Spa to Bristol Temple Meads
   five cars forward off two of the trains to Weston-super-Mare and Taunton

Each train to make a single intermediate stop between London and Reading - each dropping back one "path" with the first train stopping furthest out. This would give a 15 minute service to Slough, Maidenhaed and Twyford ... and I expect one of our members here would suggest Taplow for the fourth stop ;-).

Progressive, planned electrification of the routes covered by these trains - being bimodes - would in a long term and controlled manner move us towards greater electric haulage and the efficencies brought, and some of the line (you'll note) have London portions of trains replacing local services; an encouragement to passeners to use those services, and by splitting and joining trains, you're no longer routinely running packed services from London to Reading which remain the same length at present and thin out as they get further west.

It's getting late ... so I'll come back (if there's any interest / follow up) on track, station and depot capacities, and on that question of how we get people to the station [[ some station car parks have issues already ... not seen that addressed in the report, but then it's a long report and I may have missed it ... ]] .  And if this whole post looks a bit mad, I can turn around in two days time and say it was just last year's ideas ...

Very interesting article, and not one mention of the lost opportunity with GW electrification to increase track gauge out of Paddington to allow for Double Deck trains.

The big issue with the GW is the limited capacity at Paddington, for a major terminal station it does not have enough platforms.

Crossrail will release at least a couple though, more if in due course the Heathrow Express withers on the vine as passengers prefer the convenience of Crossrail services.

Every article I have read on double deck trains has come out saying they do not increase capacity that much as there is so much lost space for the door vestibules and the dwell times are increased substantially. 

Indeed. Can you imagine the dwell times as pax extricate themselves from a seat in the middle of the top deck, for example? Almost double I reckon at busiest stations to around 5mins a stop. Reducing capacity, not increasing it

Increased dwell time was certainly the problem with the Southern's 4DD trains - 20 seats per single width slam door exit. 

I am surprised that no moves have been made to extend the capacity by utilising the old parcels platforms at the end of platform 1.  A station car park is a bit of a luxurious concept within a major city like London.  There's room at least for two platforms in there, if not three.

Would need a bit of adjustment to the station throat on the low numbered side and the Royal Oak sidings.

Very interested background Graham to the subject of accommodating growth.

One comment I would make it that it strikes me that whilst we have some areas where there is very little capacity for easy growth - you are not going to get many more passengers per hour into Euston or Waterloo without some hugely expensive engineering work- there are still plenty of other places on the network where a doubling of numbers could be relatively easily accommodated.  If you replaced the 4 car Voyagers with 6 car IEPs you would doubt capacity with very little need for infrastructure spend at all.  If you doubled the frequency and therefore capacity through Melksham the only major cost would be in finding an extra unit and staffing for it.

So an important strategic question is:

Do we direct our spending to areas where capacity and demand is already sky high and therefore extra capacity tends to be very expensive to purchase (essentially London), or do we direct our spending to areas where there is still slack in the system and an upping of frequency and lengthening of trains can be provided at little more cost than the provision of some extra units?

I would venture that whilst we probably need to do a bit of the former, that spending on the latter is where the focus should be because the money goes much further and because it has the side effect of rebalancing the economy away from London and the SE.

Agree, at a station with huge capacity problems it amazes me that no real effort has put into restoring to use these facilities.
The disused platform 1 extension could handle very much longer trains, this facility may be viewed from any train departing platform one.
A handful of  18/19/20 coach IEPs departing platform one could significantly improve capacity.

And which other stations on the GWR network could handle such long trains? Are you suggesting 'For Reading and Newbury/Didcot(or Swindon) use the rear 9 coaches only' with the front 9 coaches going on to farther destinations?

Why not? As long as they split well down the line (say BPW for the SW/Wales)....selective door opening can be got used to.

Other than platform signalling and lengths of platform loops, I would agree. If you decide, say, 'rear 9 cars only for x,y,z' then will the front 9 cars stop beyond signals or would either end potentially foul points where platforms are on loops?

.
IIRC, Taunton could handle a train of that length, at which point the train could divide.

"the 18-03 service will depart from platform 1. Join any part of the train for Taunton where the train will divide. Join the front 9 cars for Plymouth and then all intermediate stations to Penzance, join the rear 9 cars for intermediate stations to Plymouth. The rear portion terminates at Plymouth."

on.

Yup, but I suspect EXD will be the split point as there's staff facilities there.

Apart from the fact that EXD could not accommodate an 18 coach train without fouling either the Red Cow level crossing or access to Platforms 1 and 3 from the south (having overshot the signal at the end of the platform).

fair point, but would you need to take 2x9cars west of EXD? Maybe a fast 9car plus a slow 5car?

Having been sad enough to go to EXD for the first run of the Class 800, I was surprised just how long the 9 car version was. Pulling right up to the southern extreme of platform 4, there was just about enough room for another two or maybe three coaches before it fouled the level crossing at the rear of the train, so I think you'd even be pushed to get a 9+5 train into EXD.

Yup - talking reality, I suspect we're looking at 2x5cars....

In the near term I agree, but in the longer term some significantly longer trains are going to be needed to handle rising passenger numbers.

An obvious way to achieve this is the running of very much longer trains from platform 1 at Paddington to someplace well outside London where the train may be divided.

The merit of so doing is no extra path is taken up in the congested London area.

Just three such trains would move an extra couple of thousand passengers.

It is surprising that all the upgrades to London mainline stations, and to Birmingham New Street appear to be improved current issues and not address longterm growth issues. A limit of 16 Intercity trains an hour from Paddington is to low.

Whilst I accept that Double Deck trains are slower at stations, we must start to think of radical solutions to our rail infrastructure issues.

One of the ways to address Paddington would be to great a 'Gateway' station at Old Oak Common (part of HS2) to allow the better handling handling of trains. As Paddington is not really the destination for most people, this would not be an issue and could address a lot of capacity issues. Trains could terminate/start at Old Oak Common, and visit Paddington before starting their main journey.

Capacity limits move from the running lines to the stations. At stations with significant passenger numbers off and on, perhaps alternate trains into alternate platforms?

Most people's destination is beyond Paddington: in the City, the West End, Docklands or some other part of London, which they'll probably access by tube, bus or taxi. So this just shoves the problems onto the Tube. Paddington doesn't really have particularly good Tube connections IMO – it has lots of lines but the access between mainline platforms and tunnels is not good – so possibly Old Oak Common could help if it has better Tube (and bus) connections. Although it seems to be the wrong side... Hopefully Crossrail should help.

Agreed - and that dilutes the "we must have Plymouth to London (Paddington?) in three hours" mantra somewhat. Quantitative comparators are indeed useful, but timing to a stop at a station platform that's an intermediate changing point on people's door to door journey and from where they have to walk considerably to their ongoing transport makes the comparator somewhat artificial.

The Crossrail interchange an Paddington - assuming a much better change from inbound expresses to the Betty line - should save several minutes, as would an interchange and new measuring "London" point at Old Oak.

Once OOC is built & up & running, I foresee more pax changing there for onward connections via Crossriail than at Paddington. OOC will be built to cope with this. All trains will call, I understand, adding at least 3 mins to journeys.

This at a point where the HEX contract expires. I can't see it continuing either, as pax will connect at OOC with Crossrail to Heathrow.

It would be really useful if when OOC is rebuilt, it could allow ALL trains to terminate there, empty, clean and then start their journey and head to Paddington.

This would probably require a lot of platforms/bays at OOC to handle this, but if done properly the current restriction of 16 trains per hour could be lifted, for a lot less than adding another another

• platforms at Paddington.

rebuilt?

I don't think the space is there for what you are suggesting, together with the infrastructure to clean (staff accommodation, waste disposal, etc), nor is there the ability to convert PAD into a station where services arrive, pick up & depart, rather than terminate & restart as now. The throat just doesn't have enough lines to achieve this. OOC is already being seen as a station where nothing can terminate....cf Chiltern's proposed services, which are being developed on a loop basis, with no reversals.

Depends which stations.  At Paddington a double decker isn't slower because the train is already sat at the platform for more than just a minute or two.  They are only slower at intermediate stations and for trains with few intermediate stops that is not much of an issue.  The problem is that when you think about possible trains with few intermediate stops you come up with the idea of running Paddington to Taunton or similar and that is a hell of a lot of miles to lift the gauge (and the lovely new electrification) on.  Increasing the loading gauge is only likely to be affordable for relatively short heavily used routes and those routes will not be the ones you want to run a limited stop service on.

Double Deckers would of course be very suitable for the limited stop services on HS2.  Please let them build the infrastructure to allow that even if the initial batch of trains are single deckers.

At the moment we seem to be at somewhat of a stalemate regarding passenger capacity to the southwest.
"it is not worth building significantly longer trains, or extra multiple units to significantly lengthen existing trains, because most busy stations cant accept longer trains"

Followed by "it is not worth lengthening platforms significantly because we have not got/can not afford much longer trains to utilise these platforms"

I, and others have suggested use of the extended platform 1 at Paddington to provide significantly increased capacity by running much longer trains, rather than by more trains.

An alternative might be to run some services from London Waterloo out to traditional GWR territory. Waterloo is arguably as short of paths as is Paddington so the opportunity for extra trains is limited, but Waterloo does have the merit of the very long former international platforms.
How many IET vehicles could fit into the former Eurostar platforms at Waterloo ? Would 9+9 fit ?

At a time of steadily rising passenger numbers, I do not feel it reasonable for the rail industry to simply state that the network is full and nothing beyond higher density trains and more standing can be done.

Network Rail are currently in the process of shortening the old International platforms at both ends, to make more space for milling about (by passengers) and swapping tracks (by trains). I'm not sure how short they'll be - maybe 12 cars, maybe a bit more?  Which kind of answers that question - Waterloo is, after all, the terminal for SWT/SWR services.

I came across a picture showing them doing just that a few days ago, but I can't refind it just now.

Taking a typical 7 days, I would love to know
a) How many seats are provided into (say) Paddington?
b) How many of those seats are occupied?
c) How many people have to stand because there are no seats available on the train they're on?
d) How many people stand although there's seating available (e.g. elsewhere on the train, or the middle seat of 3)

Departures (ORR) 18.2 million. Spread over 363 days = 50,000 per day.   I count 393 trains leaving in passenger service (that's counting Real Time trains for next Wednesday) which is 127 passengers per train.  So if evenly spread, there would be a seat for everyone if every train was 2 coaches.   But of course the 18:03 to Penzance is rather busier than the 11:36 to West Ruislip, the 06:27 to Oxford or the 23:03 or 23:10 services to Heathrow ...

I very nearly posted "someone had better protect those long platforms at Waterloo, before network rail shorten them so as to limit future capacity" But I felt that such a remark might be unduly cynical or negative, even by my standards !

Looks to me like a very bad decision to actively shorten platforms at a time of considerable overcrowding.
Is this a new policy "12 car maximum ever, and be glad you are getting that much"

When you look at the cost of extending platforms along a route, then I can see why the view might be that 12 is the likely maximum for the foreseeable future.  Though stuving was only speculating as to their reduced length, so at this stage we don't know for sure what they will be.

Class 373 is / was 387 metres long.  9 IEP + 5 IEP is 364 metres; 9+9 would be 468 metres.

Gloucester is 602 metres if you're looking for a destination for your super long trains. There are also a couple of rather long platforms at Bristol Temple Meads, one at Colchester and one at Cambridge.

The thing with Waterloo is that they want to run most of the Windsor side service group out of the Waterloo International platforms.  By doing that they free up existing central platforms for main line side services. 

To run about 18 tph into that side requires a completely different layout in the immediate approaches to cater for the sort of frequencies required.  P20-24 saw about as many Eurostar trains per day as they now want in the high peak, and IIRC did not allow for simultaneous arrivals and departures.

So as well as shortening at the buffers by about 50m to fit adequate circulation space in, they need to shorten the other end to fit in all the new points and crossings.

I think photos suggest the crossovers will be well inside the international "shed", and with the circulation space I'd be surprised if they are much longer than about 260m now.

Found a view showing the way the crossovers have been positioned inside the shed on skyscraper city:
http://www.skyscrapercity.com/showpost.php?p=137556304&postcount=147
Possibly like the photo stuving had seen somewhere?

Theres a photograph showing progress on the new concourse area at the bottom of the page here: http://www.kentrail.org.uk/waterloo_international_3.htm

That photo is dated July 2015?

No its not.  Is dated 03 October 2016.  You are looking at the photograph at the bottom of the page are you not....... ::)

nope....but still 18 months ago.

Er, 6 months ago, it is only 2017 now...

Errrr.....yup, you're right. Just checked :-)

Took ages to find again, but this page on a sub-contractor's website (Kelly Group) includes a video showing the concourse end of the Waterloo International platforms looking towards the buffers.  There seem to be fairly massive concrete stop blocks being cast where the tracks are cut back:

http://www.kellygroup.co.uk/projects/waterloo-international-terminal/s3806/

Paul

Thanks for finding that, Paul.  Looks like its going to be a fairly big concourse area.

Hopefully it will be fairly future proof, the various NR videos seem to show a pretty much full width barrier line to go in.  It should also remove a great many waiting passengers from that end of the main concourse, in the vicinity of the LU escalators.   I think one of the good ideas is that there will be additional stairs down to the lower level which will provide a more direct route to the Bakerloo and  Northern line concourse, (I'm assuming they'll reopen the existing mothballed LU gateline) and that will avoid the current need to exit the barriers and do a 'U-turn' back towards the LU escalators. 

I did hear that P19, the one that runs alongside but outside of the international P20 would be part of the Windsor side platform group, but retain its own oddly positioned barrier line behind the LU escalator box;  it might have been better in the long run to raise it up to the same level as P20 and open up a few gaps in the wall...

Paul