NSW Future Rail Plan - NWRL/Metro/Harbour Crossing

 
  bowralcommuter Chief Commissioner

Location: Asleep on a Manly Ferry

Sector 2 Plan
by Watson374



So I've managed to make this thing work, except that to make it work, the patterns are not spread out evenly. This is going to be a wall of text, representing today's more polished edition of last night's working out.

This edition is based on the City Circle being able to cope with 20tph, which I believe it can, and Sector 2 being able to handle the occasional two-minute headway. It also relies on the use of both the Lidcombe and Homebush turnbacks as well as the Macdonaldtown sidings.

With that declared, this is how it works. Coming out via Museum, we have five patterns of 4tph - two Bankstown, two Airport & East Hills and one East Hills. Coming out via Town Hall, we have four patterns of 4tph. The 4tph that 'vanishes' tips out at Central and reverses Macdonaldtown.

Out via Museum is pretty straightforward, but the headways are not equally spaced out. This is where that occasional two-minute separation is required, because gaps have to be built in to allow an Inner West stopper coming out the other side, forcing the uneven headways onto this side.

This allows us to work with repeated fifteen-minute blocks. Patterns are stated in the down direction because my local station is Central, so I think better in terms of down services, rather than up services.

The patterns are:
- Revesby via Airport
(2)
- Liverpool via Bankstown
(5)
- Campbelltown via Sydenham
(2)
- Macarthur via Airport
(3)
- Lidcombe via Bankstown
(3)

in that order. Spacing in minutes (leaving Central) is indicated in the numbers in the brackets between patterns.

Out via Town Hall is a little more complicated. Firstly, the Campbelltown via Sydenham expresses are detrained at Central and reversed Macdonaldtown. This opens a seven-minute gap, which allows a stopping train to lag by five minutes; thus, the stopper can take an additional five stops.

The patterns are:
- Glenfield via Granville
(2, 7)
- Homebush via Ashfield (this is the stopper)
(7, 2)
- Campbelltown via Granville
(3)
- Liverpool via Regents Park
(3)

in that order. The first value for the first two separations indicates the gaps at Redfern, and the second value the gaps at Ashfield.

By giving the three semi-fast patterns the same number of stops as far as Ashfield (five less than the stopper), they can travel as a 'flight' of three consecutive services, separated by two three-minute gaps. The result is the stopper just being able to fit in, while still permitting the majority of services on the Locals to be semi-fast.

(I have, however, distributed the four remaining less-major stops past Ashfield - Croydon, Burwood, Homebush and Flemington - evenly across the four patterns; this is due to there only being five stops to skip outright. This made for some interesting iterative working-out, but I've managed to make them just fit. The Glenfield service takes an additional stop at Clyde to even its spacing with the Campbelltown service, giving a nearly-even 8tph on the Fairfield branch.)

The specific patterns are:

Out via Musuem

Revesby via Airport
- Central, Green Sq, all to Wolli Ck, all to Revesby

Liverpool via Bankstown
- Central, Redfern, St Peters, Sydenham, Campsie, all to Bankstown, Birrong, all to Liverpool


Campbelltown via Sydenham
- Central, Redfern, Sydenham, Revesby, all to Glenfield, all to Campbelltown


Macarthur via Airport
- Central, Green Sq, all to Wolli Ck, Kingsgrove, Riverwood, all to Glenfield, all to Macarthur

Lidcombe via Bankstown
- Central, Redfern, Erskineville, Sydenham, all to Campsie, Bankstown, Yagoona, Regents Park, Berala, Lidcombe


Out via Town Hall

Glenfield via Granville
- Central, Redfern, Newtown, Ashfield, Strathfield, Flemington, Lidcombe, Auburn, Clyde, Granville, all to Glenfield


Homebush via Ashfield
- Central, Redfern, all to Ashfield, Strathfield, Homebush


Campbelltown via Granville
- Central, Redfern, Newtown, Ashfield, Croydon, Strathfield, Lidcombe, Auburn, Granville, all to Liverpool, all to Campbelltown

Liverpool via Regents Park
- Central, Redfern, Newtown, Ashfield, Burwood, Strathfield, Lidcombe, Berala, Regents Park, Chester Hill, Villawood, Cabramatta, Warwick Farm, Liverpool

I think I've got something here, but please feel free to criticise constructively.

This is a very tight set of patterns - I freely admit this is running it on the edge - but it appears to achieve the goal of nesting a stopper between flights of semi-fast trains, as well as the goal of a more direct service for Liverpool. It achieves this with minimal collateral damage to existing services, and with absolutely no new infrastructure other than that already built as part of Clearways.

It is, however, nearly saturated to begin with and leaves very little 'wiggle room', especially between Sydenham and Homebush via the City Circle.
"Watson374"


Congrats on the attempt itself, can be pretty frustrating I imagine. Off-peak perhaps I'd have:
C'town via Sydenham express: Town Hall, Wynyard, Circular Quay, St. James, Museum, Central, Redfern, Sydenham, Beverly Hills, Revesby, Glenfield, all to C'town.
Liverpool via Regents Park express: Museum, St. James, Circular Quay, Wynyard, Town Hall, Central, Redfern, Ashield, Strathfield, Lidcombe, Regents Park, Chester Hill, Cabramatta, Warwick Farm, Liverpool.

Sponsored advertisement

  Watson374 Chief Commissioner

Location: Fully reclined at the pointy end.
Congrats on the attempt itself, can be pretty frustrating I imagine.
"bowralcommuter"


Drove myself insane doing it.

Off-peak perhaps I'd have:
C'town via Sydenham express: Town Hall, Wynyard, Circular Quay, St. James, Museum, Central, Redfern, Sydenham, Beverly Hills, Revesby, Glenfield, all to C'town.
Liverpool via Regents Park express: Museum, St. James, Circular Quay, Wynyard, Town Hall, Central, Redfern, Ashield, Strathfield, Lidcombe, Regents Park, Chester Hill, Cabramatta, Warwick Farm, Liverpool.
"bowralcommuter"


Absolutely - the idea should be to maximise the amount of skipping withing the gaps available. Off-peak, gaps should be large enough to run decent expresses, again in 'flights'.

The problem with peak operation is that trains are so close that there is very, very little room for me to do anything with them. The stops taken after Ashfield were carefully selected so as to prevent the gaps going below two minutes, and to widen gaps as soon as possible.

If anything, this exercise shows just how congested the Locals are.
  simonl Chief Commissioner

Location: Brisbane
I don't really. It was only in reply to someone else believing that I forgot the people on that line, by leaving them without a service. But also as a link to strathfield so that A5 can get to wondabye if need be!!!!!!!!
"jedimasterc"

And I still think that you did indeed forget them.

I'd support such shuttles off peak to increase frequency on the Sefton-Carramar stretch.  Surely part of the reason for its poor PT use is the mediocre/poor service!

I don't think it is going to happen in peak hour though.  If you have 4tph reversing at Lidcombe, it will be hard to add 4tph additional services as it would allow less than 8 minutes to turn around, probably less than 6.

Watson, I think the idea of the K2RQ is for 8tph Revesby via Airport and all the cambos in peak going via Sydenham.  Which would be sensible.  You shouldn't mix and match with Cambo via Airport because then a problem on one part of the network isn't isolated.  Similar with Cambo via Granville.

no glenfield won't be viable but liverpool will be with the 4th platform and 2x small stabling facilities.
"jedimasterc"


I was under the impression that Glenfield reversers would become Leppington reversers.
"Watson374"

Absolutely.

Congrats on the attempt itself, can be pretty frustrating I imagine. Off-peak perhaps I'd have:
C'town via Sydenham express: Town Hall, Wynyard, Circular Quay, St. James, Museum, Central, Redfern, Sydenham, Beverly Hills, Revesby, Glenfield, all to C'town.
Liverpool via Regents Park express: Museum, St. James, Circular Quay, Wynyard, Town Hall, Central, Redfern, Ashield, Strathfield, Lidcombe, Regents Park, Chester Hill, Cabramatta, Warwick Farm, Liverpool.
"bowralcommuter"

I think off peak Cambo should be via Airport.  It's more important to increase frequency at Green Square and Mascot than save a few minutes from the Cambo trains.
  Watson374 Chief Commissioner

Location: Fully reclined at the pointy end.
And I still think that you did indeed forget them.

I'd support such shuttles off peak to increase frequency on the Sefton-Carramar stretch.  Surely part of the reason for its poor PT use is the mediocre/poor service!

I don't think it is going to happen in peak hour though.  If you have 4tph reversing at Lidcombe, it will be hard to add 4tph additional services as it would allow less than 8 minutes to turn around, probably less than 6.
"simonl"


I honestly fail to see any point in a Liverpool-Lidcombe via Regents Park shuttle, even off-peak - I consider the Liverpool via Regents Park service I crammed in a better alternative, as it offers a semi-fast direct connection to the CBD. This service should overlap with the existing Liverpool and Lidcombe via Bankstown services to cover the area with enough service.

During peak, it's even less viable. One potential benefit is the ability to transfer passengers onto the Western line expresses into the City, but during peaks these trains would shame a fish cannery (or a gulag jailer, if you prefer that analogy).

The direct service can fit (seeing as I inserted it as part of a 'flight'), negates the need for the shuttle, and allows for better use of the Lidcombe turnback.

Watson, I think the idea of the K2RQ is for 8tph Revesby via Airport and all the cambos in peak going via Sydenham.  Which would be sensible.  You shouldn't mix and match with Cambo via Airport because then a problem on one part of the network isn't isolated.  Similar with Cambo via Granville.
"simonl"


That makes sense as well, but I'm apprehensive about denying expresses for Padstow, Riverwood and Kingsgrove. The problem with such a long express run is that you're forcing everyone east of Revesby to take the local train. Is this a worthwhile culling for the separation of service?

I suppose it could be, but I'll need to recalculate the full Monty to account for the East Hills line taking only two patterns rather than three - unless we run a limited-stops Revesby starter?

I think off peak Cambo should be via Airport.  It's more important to increase frequency at Green Square and Mascot than save a few minutes from the Cambo trains.
"simonl"


Absolutely, especially the bit I've underlined. I would contend that increasing frequency to Domestic and International is also important, as good frequency would make a good impression on travellers arriving and using the Airport line. So as a whole, the Airport line would benefit with increased frequency - in any case, I'm fairly sure crossovers exist to allow expresses to exit Wolli Ck and then move out to the fast tracks.

Going back to that wall of text I spewed yesterday, with Clyde added back to the Glenfield via Granville service, would it be viable to shift the responsibility for Clyde over to the South line, dropping it entirely from the Western line, allowing the latter to run better expresses?
  djf01 Chief Commissioner


I think I've got something here, but please feel free to criticise constructively.
"Watson374"

... you asked for it ... Smile ...

Before I start though: good work.  Really good work.  You've covered a lot of bases.  There are a few things I'd still like to check before I really sink the boot in, but lets start with this lot .... Smile

OK.  2min separations out of the City just aren't going to work.  2.5min maybe.  The signal headway is is up to 100/110 seconds on sections of the line (but not on the approaches to the platforms).  The sections where the signal headway is long are where trains are supposed to be dwelling, but AIUI, it doesn't quite work that way anymore.  Train A has to be well clear of the platform before train B is allowed to enter under caution.  Dwells are up to and sometimes beyond 60 seconds.  For a train be able to maintain a 2 min separation right around the City Circle is ... shall we say, optimistic.

CityRail's timetables (I think) work on a 90 second stopping penalty, and that's what I used for most of my timetables.  It's a bit too long, but equally 60 seconds is a bit too short IMHO.

Given an accelleration/decelleration of 1m/s^2 (which is what the website says the A sets are rated at, and a really nice nuimber to make the arithmentic on the equations of motion simple): to get from 0 to 60kpoh (or 60kph to 0) takes ~20 seconds.  120kph is ~40 seconds.  If the stations are on a 1:33 grade the stopping/starting times are longer because
a = (1 -  g * sine(theta))
  = 1 -  10 * 1/33
 ~= 2/3.

So 30 seconds to get to/from 60kph, 60 for 120.

If you take 60 seconds as the worst case scenario for the accelleration/decelleration time, and a "typical" suburban dwell of say 30 seconds, that gives you a 90 second stopping penalty.  60 seconds to slow down, 30 seconds stopped, 60 seconds to speed up again vs 30 seconds and 30 seconds to cover the same ground at top speed.  Now that's the worst case.  But it's probably about right for stations like Werrington on the Western line.

For closely spaced stations like on the Inner West and Bankstown line, the train is never going to reach 100kph between stops, so even a worst case scenario you are looking at 30+30+30 vs 15+15 or a 60 second stopping penalty.  That is reasonable, but it doesn't leave much room for variance.

Each path is 3 min wide.  If you assume a 90 second stopping penalty that means there can only be 2 stations difference in the stopping patterns.  Even if you assume a 60 second stopping penalty, that's only 3 stops difference.

If you have 2 patterns on a track pair, a stopper and an express, the biggest difference you can have in run time over a segment is 9 minutes.  Which is 6 stops using cautious assumptions, and 9 using aggressive ones. 

When you get 4 of the 5 possible paths filled, you can have at absolute most 3 stops difference in the patterns, but for the timetable to have stability it probably needs to be only 2.  To some extent it depends on the stops.  Stopping: Stanmore, Petersham, Lewisham would probably only cost 3 minutes, or the 1 slot that's available.

You're timetable is working on a difference of 5 stops, and making it work using the 2 min separation leaving Central.  You could claw back some of that by having the train after the terminator skip stopping Redfern.  I *think* Redfern gets a high enough frequency at the moment Smile.




  bowralcommuter Chief Commissioner

Location: Asleep on a Manly Ferry

I think off peak Cambo should be via Airport.  It's more important to increase frequency at Green Square and Mascot than save a few minutes from the Cambo trains.
"simonl"


I see your point, I suppose that's what the Kingsgrove terminators do in peak hour?
  Watson374 Chief Commissioner

Location: Fully reclined at the pointy end.
Before I start though: good work.  Really good work.  You've covered a lot of bases.  There are a few things I'd still like to check before I really sink the boot in, but lets start with this lot .... Smile
"djf01"


Thank you. Very Happy

OK.  2min separations out of the City just aren't going to work.  2.5min maybe.  The signal headway is is up to 100/110 seconds on sections of the line (but not on the approaches to the platforms).  The sections where the signal headway is long are where trains are supposed to be dwelling, but AIUI, it doesn't quite work that way anymore.  Train A has to be well clear of the platform before train B is allowed to enter under caution.  Dwells are up to and sometimes beyond 60 seconds.  For a train be able to maintain a 2 min separation right around the City Circle is ... shall we say, optimistic.
"djf01"


I had the nagging feeling that it was being optimistic. Unfortunately, it seems my gut feeling was right, and that it seemed a little too good.

I suppose I could try to shuffle the City Circle run times so as to keep them three minutes apart, but even that is a tough call.

CityRail's timetables (I think) work on a 90 second stopping penalty, and that's what I used for most of my timetables.  It's a bit too long, but equally 60 seconds is a bit too short IMHO.
"djf01"


Ouch. I've checked the timetables, though, and they tend to indicate a minute saved per skip - even for longer series of skips where the 90 seconds would show. Could this be related to the skipped stops having a lower timetabled dwell?

Given an accelleration/decelleration of 1m/s^2 (which is what the website says the A sets are rated at, and a really nice nuimber to make the arithmentic on the equations of motion simple): to get from 0 to 60kpoh (or 60kph to 0) takes ~20 seconds.  120kph is ~40 seconds.  If the stations are on a 1:33 grade the stopping/starting times are longer because
a = (1 -  g * sine(theta))
  = 1 -  10 * 1/33
 ~= 2/3.

So 30 seconds to get to/from 60kph, 60 for 120.
"djf01"


Seems alright. I didn't go into this - although in hindsight I should have - because I was working off the current timetables.

If you take 60 seconds as the worst case scenario for the accelleration/decelleration time, and a "typical" suburban dwell of say 30 seconds, that gives you a 90 second stopping penalty.  60 seconds to slow down, 30 seconds stopped, 60 seconds to speed up again vs 30 seconds and 30 seconds to cover the same ground at top speed.  Now that's the worst case.  But it's probably about right for stations like Werrington on the Western line.
"djf01"


That works for me. However, as far as I know, the critical parts of the areas we're dealing with don't have this kind of line speed.

For closely spaced stations like on the Inner West and Bankstown line, the train is never going to reach 100kph between stops, so even a worst case scenario you are looking at 30+30+30 vs 15+15 or a 60 second stopping penalty.  That is reasonable, but it doesn't leave much room for variance.
"djf01"


No, it doesn't. The room is tiny. The alternating skip-stop thing I came up with for the Bankstown line (which I believe was inspired by you) already puts considerable pressure, since 7.5 minute gaps contract to 3.5 minutes and expand to 11.5 minutes at Campsie. This basically proved that running a super-express down the Bankstown wasn't going to work.

Each path is 3 min wide.  If you assume a 90 second stopping penalty that means there can only be 2 stations difference in the stopping patterns.  Even if you assume a 60 second stopping penalty, that's only 3 stops difference.

If you have 2 patterns on a track pair, a stopper and an express, the biggest difference you can have in run time over a segment is 9 minutes.  Which is 6 stops using cautious assumptions, and 9 using aggressive ones.

When you get 4 of the 5 possible paths filled, you can have at absolute most 3 stops difference in the patterns, but for the timetable to have stability it probably needs to be only 2.  To some extent it depends on the stops.  Stopping: Stanmore, Petersham, Lewisham would probably only cost 3 minutes, or the 1 slot that's available.
"djf01"


Very, very tight. This is a very congested railway, isn't it.

You're timetable is working on a difference of 5 stops, and making it work using the 2 min separation leaving Central.  You could claw back some of that by having the train after the terminator skip stopping Redfern.  I *think* Redfern gets a high enough frequency at the moment Smile.
"djf01"


I think I could do that - but it'll have knock-on effects down the line (although down the line is where they can be better absorbed). Might need to use the Clyde trick to even out gaps.

Well - it's certainly proved how stuffed we are! Thanks for the feedback, djf01.

In a few days, I might take another shot at attempting to construct a timetable plan.

EDIT: Hang on, can we resignal the City Circle to reduce speed and thus safe distance, allowing denser packing?
  grog Train Controller

Before I start though: good work.  Really good work.  You've covered a lot of bases.  There are a few things I'd still like to check before I really sink the boot in, but lets start with this lot .... Smile

OK.  2min separations out of the City just aren't going to work.  2.5min maybe.  The signal headway is is up to 100/110 seconds on sections of the line (but not on the approaches to the platforms).  The sections where the signal headway is long are where trains are supposed to be dwelling, but AIUI, it doesn't quite work that way anymore.  Train A has to be well clear of the platform before train B is allowed to enter under caution.  Dwells are up to and sometimes beyond 60 seconds.  For a train be able to maintain a 2 min separation right around the City Circle is ... shall we say, optimistic.

CityRail's timetables (I think) work on a 90 second stopping penalty, and that's what I used for most of my timetables.  It's a bit too long, but equally 60 seconds is a bit too short IMHO.

Given an accelleration/decelleration of 1m/s^2 (which is what the website says the A sets are rated at, and a really nice nuimber to make the arithmentic on the equations of motion simple): to get from 0 to 60kpoh (or 60kph to 0) takes ~20 seconds.  120kph is ~40 seconds.  If the stations are on a 1:33 grade the stopping/starting times are longer because
a = (1 -  g * sine(theta))
  = 1 -  10 * 1/33
 ~= 2/3.

So 30 seconds to get to/from 60kph, 60 for 120.

If you take 60 seconds as the worst case scenario for the accelleration/decelleration time, and a "typical" suburban dwell of say 30 seconds, that gives you a 90 second stopping penalty.  60 seconds to slow down, 30 seconds stopped, 60 seconds to speed up again vs 30 seconds and 30 seconds to cover the same ground at top speed.  Now that's the worst case.  But it's probably about right for stations like Werrington on the Western line.

For closely spaced stations like on the Inner West and Bankstown line, the train is never going to reach 100kph between stops, so even a worst case scenario you are looking at 30+30+30 vs 15+15 or a 60 second stopping penalty.  That is reasonable, but it doesn't leave much room for variance.

Each path is 3 min wide.  If you assume a 90 second stopping penalty that means there can only be 2 stations difference in the stopping patterns.  Even if you assume a 60 second stopping penalty, that's only 3 stops difference.

If you have 2 patterns on a track pair, a stopper and an express, the biggest difference you can have in run time over a segment is 9 minutes.  Which is 6 stops using cautious assumptions, and 9 using aggressive ones. 

When you get 4 of the 5 possible paths filled, you can have at absolute most 3 stops difference in the patterns, but for the timetable to have stability it probably needs to be only 2.  To some extent it depends on the stops.  Stopping: Stanmore, Petersham, Lewisham would probably only cost 3 minutes, or the 1 slot that's available.

You're timetable is working on a difference of 5 stops, and making it work using the 2 min separation leaving Central.  You could claw back some of that by having the train after the terminator skip stopping Redfern.  I *think* Redfern gets a high enough frequency at the moment Smile.
"djf01"


Good technical post, this has inspired me to look at what could be done for South Line travel times and capacity, as well as travel times on the Rapid Transit lines. The big mistake everyone seems to be making is making the system too complex, which is the opposite of what needs to happen.

South Line

Once again I will probably be going against the grain here, but I think travel time saving could be made on the Inner West / South Line by moving to a 100% A set and M set timetable, simplifying stopping pattern and fixing track where it is not up to scratch.

Someone else can tell me what the speed restrictions are along the line , but my working is based on an average TOP speed of 80km/h.

So the basics are:

  • All DD stock having the same performance characteristic and being capable of accelerating at 1m/s^2

  • All trains stop at all stations, reducing the need for padding in the timetable due to uncertainty

  • 38 second average station dwell time (should be less than this - this is the padding in the timetable)



My calculations are that it will take 22 seconds to get from 0 - 80km/h, and that the train will have covered 244m in that time, so 44 seconds to decelerate and accelerate over a space of 488m, which would normally take 22 seconds to cover.

This gives a stopping penalty of 60 seconds including a bit of padding in the dwell time.

Liverpool - Central is 36km distance via Granville. Presuming Macdonaldtown is closed, there are a total of 24 stations to stop at not including Central itself.

Travel time over 36km @ 80km/h is 27 minutes, plus a 24 minute stopping penalty, equals a total travel time of 51 minutes. This compares to a current travel time of 56 minutes on the current stopping pattern with a mixed stopping pattern and mixed fleet.

For this plan to work (which would be post new harbour crossing and Bankstown/Hurstville becoming Rapid Transit) what remains of sector 2 would need to be 100% A set/M set. This is also going to be needed for the Western line to get speeds up there. So stock type will have to look like this:

Sector 1: All T Sets
Sector 2 (remaining): All A/M sets
Sector 3 (main): All A/M sets
Cumberland: All A/M sets
Hurstville/Bankstown/NW: SD Rapid Transit
Northern line into Central: Whatever is left

(Remember that this is likely to be needed in 2030 at the earliest so fleet mix should be at that point by then).

This would allow you to run a full 20tph through the inner west, giving a 3 minute service at the Inner West stations. Pattern would look something like:

8tph SWRL, all to CBD via Granville
Times past hour (15 minute pattern): 00,06
Advertised as: "Trains every 6 - 9 minutes"

4tph Liverpool, all to CBD via Granville
Time past hour: 09 + 00,06 from above
Advertised as: "Trains every 3 - 6 minutes"

4tph Homebush, all to CBD
Time past hour: 03 + 00,06,09 from above
Advertised as: "Trains every 3 - 6 minutes"

4tph Ashfield, all to CBD
Time past hour: 12 + 00,03,06,09 from above

With the above we have a maximum of 4tph being turned at any one platform and are getting the maximum throughput of trains without mixing stopping patterns, and still getting decent travel times through the use of high performance rolling stock.

Rapid Transit

Using the same methology, but with SD rapid transit stock, I have come up with potential travel times from Cabramatta, Bankstown, Hurstville, Rouse Hill, Castle Hill and Epping to the CBD for the rapid transit line.

The stock characteristics are based on the S Stock recently delivered for the London Underground (http://www.bombardier.com/en/transportation/products-services/rail-vehicles/metros/london--united-kingdom?docID=0901260d8000d19e#) which is capable of 1.3m/s^2 acceleration and has a capped top speed of 100km/h. I have once again assumes an average top speed over the line of 80km/h which will likely require track improvements on existing lines.

With this stock, the acceleration time to 80km/h is 17 seconds, covering a distance of 188m in that time. Deceleration and acceleration combined take 34second over a distance of 376m, which would normally take 17 seconds in total. Assuming an average dwell time of 28 seconds (once again allowing some padding) you have a stop penalty of 45 seconds.

Travel time from key stations:

Cabramatta
Distance to Central: 31.5km
Stations: 22
Travel time: 24 minutes
Stopping penalty: 16.5 minutes
Total travel time: 40.5 minutes

Bankstown
Distance to Central: 19km
Stations: 14
Travel time: 14.5 minutes
Stopping penalty: 10.5 minutes
Total travel time: 25 minutes

Hurstville
Distance to Central: 15km
Stations: 11
Travel time: 11.5 minutes
Stopping penalty: 8.5 minutes
Total travel time: 20 minutes

Rouse Hill
Distance to Wynyard: 43km
Stations: 16
Travel time: 32.5 minutes
Stopping penalty: 12 minutes
Total travel time: 44.5 minutes

Castle Hill
Distance to Wynyard: 30km
Stations: 10
Travel time: 22.5 minutes
Stopping penalty: 7.5 minutes
Total travel time: 30 minutes

Epping
Distance to Wynyard: 24km
Stations: 8
Travel time: 18 minutes
Stopping penalty: 6 minutes
Total travel time: 24 minutes

Conclusion

I guess the above is a long winded way to say that we need to be looking beyond what people SAY they want to what people REALLY actually want.

What people SAY they want:

  • Express trains that skip as many stations as possible

  • Trains that run on time



What people REALLY want:

  • Frequent trains that are not delayed (certainty)

  • Fast journey times (more time at home)



Sometimes it is the case that you need to give people something other than what they think the want or say they want to give them what they fundamentally are asking for when they ask for these things. Simplifying the system, reducing the need for timetables and getting rid of skip/stop patterns helps achieve what people really want.
  Watson374 Chief Commissioner

Location: Fully reclined at the pointy end.
Speed, frequency, and a reduction in complexity. That's what people really want.

You've delivered it all, grog. I'm convinced - you've made the numbers work, and I can't fault them.

I suppose the future lies with something like the London Underground S8...

EDIT: I suspect not everyone will rejoice at the idea of an all-Tangara Sector 1 (which under grog's extrapolation would lose the locals, retaining only the mains for only fast runs, I take it?), but hey, it makes sense from the strictly operational, no-nonsense point of view.
  grog Train Controller

Speed, frequency, and a reduction in complexity. That's what people really want.

You've delivered it all, grog. I'm convinced - you've made the numbers work, and I can't fault them.

I suppose the future lies with something like the London Underground S8...

EDIT: I suspect not everyone will rejoice at the idea of an all-Tangara Sector 1 (which under grog's extrapolation would lose the locals, retaining only the mains for only fast runs, I take it?), but hey, it makes sense from the strictly operational, no-nonsense point of view.
"Watson374"


By the time the above is in place, there is a good chance T sets will be retired anyway.

Another addition to the above - you could perhaps drop Petersham to save an extra minute (I am on the fence about this), and Clyde could also go if you did something decent with the Carlingford line. This would bring the time from Liverpool down to 49 minutes.
  Watson374 Chief Commissioner

Location: Fully reclined at the pointy end.
By the time the above is in place, there is a good chance T sets will be retired anyway.

Another addition to the above - you could perhaps drop Petersham to save an extra minute (I am on the fence about this), and Clyde could also go if you did something decent with the Carlingford line. This would bring the time from Liverpool down to 49 minutes.
"grog"


Behold the Illawarra Decker!

Many of these stops could go. Clyde would go if the Carlingford line was used for the PERL or summat. Some of the Inner West stops might be better off consolidated.
  grog Train Controller

By the time the above is in place, there is a good chance T sets will be retired anyway.

Another addition to the above - you could perhaps drop Petersham to save an extra minute (I am on the fence about this), and Clyde could also go if you did something decent with the Carlingford line. This would bring the time from Liverpool down to 49 minutes.
"grog"


Behold the Illawarra Decker!

Many of these stops could go. Clyde would go if the Carlingford line was used for the PERL or summat. Some of the Inner West stops might be better off consolidated.
"Watson374"


I think the most you could lose in the Inner West would be Petersham and Macdonaldtown – anything else and you are disturbing the existing development pattern too much. These 2 stations are only candidates as they are each very close to TWO other stations which reduces the impact of closure on the local area. I say this as a previously regular user of Macdonaldtown station, so you can be assured that this is a hard-headed view of the situation.


  djf01 Chief Commissioner


Good technical post, this has inspired me to look at what could be done for South Line travel times and capacity, as well as travel times on the Rapid Transit lines. The big mistake everyone seems to be making is making the system too complex, which is the opposite of what needs to happen.

South Line

Once again I will probably be going against the grain here, but I think travel time saving could be made on the Inner West / South Line by moving to a 100% A set and M set timetable, simplifying stopping pattern and fixing track where it is not up to scratch.

Someone else can tell me what the speed restrictions are along the line , but my working is based on an average TOP speed of 80km/h.

So the basics are:

  • All DD stock having the same performance characteristic and being capable of accelerating at 1m/s^2

  • All trains stop at all stations, reducing the need for padding in the timetable due to uncertainty

  • 38 second average station dwell time (should be less than this - this is the padding in the timetable)



My calculations are that it will take 22 seconds to get from 0 - 80km/h, and that the train will have covered 244m in that time, so 44 seconds to decelerate and accelerate over a space of 488m, which would normally take 22 seconds to cover.

This gives a stopping penalty of 60 seconds including a bit of padding in the dwell time.

Liverpool - Central is 36km distance via Granville. Presuming Macdonaldtown is closed, there are a total of 24 stations to stop at not including Central itself.

Travel time over 36km @ 80km/h is 27 minutes, plus a 24 minute stopping penalty, equals a total travel time of 51 minutes. This compares to a current travel time of 56 minutes on the current stopping pattern with a mixed stopping pattern and mixed fleet.

For this plan to work (which would be post new harbour crossing and Bankstown/Hurstville becoming Rapid Transit) what remains of sector 2 would need to be 100% A set/M set. This is also going to be needed for the Western line to get speeds up there. So stock type will have to look like this:

Sector 1: All T Sets
Sector 2 (remaining): All A/M sets
Sector 3 (main): All A/M sets
Cumberland: All A/M sets
Hurstville/Bankstown/NW: SD Rapid Transit
Northern line into Central: Whatever is left

(Remember that this is likely to be needed in 2030 at the earliest so fleet mix should be at that point by then).

This would allow you to run a full 20tph through the inner west, giving a 3 minute service at the Inner West stations. Pattern would look something like:

8tph SWRL, all to CBD via Granville
Times past hour (15 minute pattern): 00,06
Advertised as: "Trains every 6 - 9 minutes"

4tph Liverpool, all to CBD via Granville
Time past hour: 09 + 00,06 from above
Advertised as: "Trains every 3 - 6 minutes"

4tph Homebush, all to CBD
Time past hour: 03 + 00,06,09 from above
Advertised as: "Trains every 3 - 6 minutes"

4tph Ashfield, all to CBD
Time past hour: 12 + 00,03,06,09 from above

With the above we have a maximum of 4tph being turned at any one platform and are getting the maximum throughput of trains without mixing stopping patterns, and still getting decent travel times through the use of high performance rolling stock.

Rapid Transit

Using the same methology, but with SD rapid transit stock, I have come up with potential travel times from Cabramatta, Bankstown, Hurstville, Rouse Hill, Castle Hill and Epping to the CBD for the rapid transit line.

The stock characteristics are based on the S Stock recently delivered for the London Underground (http://www.bombardier.com/en/transportation/products-services/rail-vehicles/metros/london--united-kingdom?docID=0901260d8000d19e#) which is capable of 1.3m/s^2 acceleration and has a capped top speed of 100km/h. I have once again assumes an average top speed over the line of 80km/h which will likely require track improvements on existing lines.

With this stock, the acceleration time to 80km/h is 17 seconds, covering a distance of 188m in that time. Deceleration and acceleration combined take 34second over a distance of 376m, which would normally take 17 seconds in total. Assuming an average dwell time of 28 seconds (once again allowing some padding) you have a stop penalty of 45 seconds.

Travel time from key stations:

...

Conclusion

I guess the above is a long winded way to say that we need to be looking beyond what people SAY they want to what people REALLY actually want.

What people SAY they want:

  • Express trains that skip as many stations as possible

  • Trains that run on time



What people REALLY want:

  • Frequent trains that are not delayed (certainty)

  • Fast journey times (more time at home)



Sometimes it is the case that you need to give people something other than what they think the want or say they want to give them what they fundamentally are asking for when they ask for these things. Simplifying the system, reducing the need for timetables and getting rid of skip/stop patterns helps achieve what people really want.
"grog"


Thanks for another really thought provoking post.  Making me rethink my SD vs DD arguments.

The run time performance gains of Rapid Transit are down to better acceleration/deceleration and I expect that is at least in part down to Aluminum construction.  Whether that's appropriate or not in a Sydney context is possibly debatable.

But what it really highlights is that we are using the benefits of DDs terribly ineffectively.  Our second tier really needs way fewer stations and faster track speeds.  The sort of service Liverpool *should* be getting is a frequent (5 min) DD with a 120kph average run speed making 3 or 4 stops (Lidcombe, Burwood, Ashfield).  Over 35km that's a run time of 18 min with say 4 90second stopping penalties: 24min from Central.  Under half an hour from the city.  Instead we're struggling to "optimise" to under an hour.

As for what people *really* want.  I can tell you what they *won't* like: watching 8 western lines trains overtaking them while *their* train makes every stop on the inner west.  A fair assessment of absolute performance goes out the window when your neighbour is getting a better *relative* performance Smile.

Also, you've concinved me that 60 seconds is a reasonable stopping penalty assumption, at least on the inner west and probably on all of sector 2.  Only a tiny fraction of Sector 2 has speed limits above 80 kph (on the straight bits of the main south, old main south and East hills), but more places are 65/60/50. 

Being a bit picky I don't think you can just use the equations of motiuon to calculate acc/dec times.  The acc rating of trains is in optimal conditions, and doesn't apply over all speeds.  It drops off as the train approaches it's maximum speed, and there is a reduced efficiency of the motors at very low speeds too.  On top of that, you need a bit of a taper when stopping and starting or you throw the PAX around like their on a C set.  Performance also varies with wind, rain (potentially rail adhesion), train load and probably some other factors as well.  Particularly with braking, there needs to be some buffer, because any of these variances might see the train overshoot.  It really needs to be done empirically, and I just don't even know where to go about fining that kind of data.

Skilled drivers can get close to the optimal performance out of trains by picking the right spot to brake for the given conditions.  Humans are surprisingly good at this sort of thing compared to their electronic competitors, in particular good at anticipating potential problems i9n particular conditions.  IN computer programming terms, it's the "common sense" class all objects need to inherrit, but none actually do.
  djf01 Chief Commissioner


In a few days, I might take another shot at attempting to construct a timetable plan.

EDIT: Hang on, can we resignal the City Circle to reduce speed and thus safe distance, allowing denser packing?
"Watson374"


The Circle is *already* restricted to 40kph (as is the underground Nth Shore).  This is deal with the Bradfield ultra tight signal spacing on the approach to platforms.

You really should download the Drivers Route Knowledge Diagram PDF (sorry, can't find the link off hand) which details all of the signal spacings.  It used to be the case that trailing trains could approach within 40m of preceding train dwelling at a platform, and that after stopping, the trailing train was allowed to enter the platform before the preceding train had completely left it, protected by the automated stops.  I have boyhood memories of seeing this in action.

I don't know if this practice is still allowed, but I suspect not, and it would seem inconsistent with the general principles of signal safety the current regulator has laid down: ie A train should be able to be brought to a stop from maximum line speed from the first red signal.  Maybe one of us could post the question down on the Signals and Infrastructure board her at RP.  I had read someone here on RP make the comment that there had been changes to the way the circle was signalled, but it's clear the signal spacing hasn't changed since the Bradfield days.

To answer your question directly, the answer is absolutely yes.  In fact, it has already been done, and in the past much tighter signal headways were achieved.  That the current system could handle 24tph unmodified is far less certain.  It definitely can't handle 30 DD trains an hour (because of the dwell), but I'm yet to be completely convinced the issue with 24tph is more to do with RailCorp using full minute timetable increments rather than limits of the signalling system.  But I don't really know for sure.

In any case, you *could* rejig the signal spacing to achieve 24tph with the currently installed technology and modern safety standards.  But why would you?  ATC/ATP does away with the need to pre-determin/optimise block locations, but more importantly does away the need for visual signals altogether.  Resignalling using 21st century ATC/ATP would unquestionably be *cheaper* (unless you try and refit the S set fleet) and have a greater longevity than resignalling using 1950s technology.

ATC/ATP has many other benefits.  The "problem" is one of those benefits is it enables DOO (or perhaps better described as GOO: Guard Only Operation Smile).







  grog Train Controller

Thanks for another really thought provoking post.  Making me rethink my SD vs DD arguments.

The run time performance gains of Rapid Transit are down to better acceleration/deceleration and I expect that is at least in part down to Aluminum construction.  Whether that's appropriate or not in a Sydney context is possibly debatable.

But what it really highlights is that we are using the benefits of DDs terribly ineffectively.  Our second tier really needs way fewer stations and faster track speeds.  The sort of service Liverpool *should* be getting is a frequent (5 min) DD with a 120kph average run speed making 3 or 4 stops (Lidcombe, Burwood, Ashfield).  Over 35km that's a run time of 18 min with say 4 90second stopping penalties: 24min from Central.  Under half an hour from the city.  Instead we're struggling to "optimise" to under an hour.

As for what people *really* want.  I can tell you what they *won't* like: watching 8 western lines trains overtaking them while *their* train makes every stop on the inner west.  A fair assessment of absolute performance goes out the window when your neighbour is getting a better *relative* performance Smile.

Also, you've concinved me that 60 seconds is a reasonable stopping penalty assumption, at least on the inner west and probably on all of sector 2.  Only a tiny fraction of Sector 2 has speed limits above 80 kph (on the straight bits of the main south, old main south and East hills), but more places are 65/60/50. 

Being a bit picky I don't think you can just use the equations of motiuon to calculate acc/dec times.  The acc rating of trains is in optimal conditions, and doesn't apply over all speeds.  It drops off as the train approaches it's maximum speed, and there is a reduced efficiency of the motors at very low speeds too.  On top of that, you need a bit of a taper when stopping and starting or you throw the PAX around like their on a C set.  Performance also varies with wind, rain (potentially rail adhesion), train load and probably some other factors as well.  Particularly with braking, there needs to be some buffer, because any of these variances might see the train overshoot.  It really needs to be done empirically, and I just don't even know where to go about fining that kind of data.

Skilled drivers can get close to the optimal performance out of trains by picking the right spot to brake for the given conditions.  Humans are surprisingly good at this sort of thing compared to their electronic competitors, in particular good at anticipating potential problems i9n particular conditions.  IN computer programming terms, it's the "common sense" class all objects need to inherrit, but none actually do.
"djf01"


Yep some very good points raised as usual. I attempted to compensate to some extent for these variations to some extent using extra dwell time, but I have no idea if what I have added is enough - you may have noticed that they made nice round numbers!

Certainly agree regarding out use DD on runs with small station spacing - which is why I think that running the Western Line as a mostly express single stopping pattern will work so well.

As far as the Western Line trains overtaking, they always have the option to switch to these trains at the first available interchange station  Laughing

I do agree with a lot of what you are saying though - I guess I was trying to do is shift the debate a little so we stop trying to make the system everything to everyone because we feel we can, and start thinking about the realities of operating a system that is approaching capacity in the most efficient and reliable way.

We all know the main reason for the slow down in speeds in recent years - complexity, leads to unreliability, leads to the need to slow things down to compensate and look like you are on time. When we are running at capacity we no longer have the flexibility to be as complex as we have in the past, and there will be some losers as a result for this. The vast majority, however, will benefit.
  djf01 Chief Commissioner


I think the most you could lose in the Inner West would be Petersham and Macdonaldtown – anything else and you are disturbing the existing development pattern too much. These 2 stations are only candidates as they are each very close to TWO other stations which reduces the impact of closure on the local area. I say this as a previously regular user of Macdonaldtown station, so you can be assured that this is a hard-headed view of the situation.
"grog"


Other than Stanmore has a something of a small commercial precinct, I was thinking rationalising Lewisham/Petersham/Stanmore back to just Petersham.

As I see it, the big issue with a lot of these really tightly spaced stations is poor pedestrian access.  The station I think is the biggest candidate for rationalisation is Milson's Pt.  It's less than 800m from Nth Sydney.  But you need to walk >1200m as well as gain/lose 20m of altitude to get between the two on foot.

If McTown has a direct walkway to Newtown, and Newtown was serviced by at least 2 stopping patterns vs McTown's 1, no-one would use McTown.
  djf01 Chief Commissioner

Has anyone else picked up on how a seemingly benign administrative change can have a such a big impact on the planning process elsewhere in the network?

All previous incarnations of the 3 Tier plan have included the inner west on the metro.  But with the shift of the interurbans to NSW trains, there is now a need to retain the mains between Strathfield and Redfern for their use, and hence no room for the inner west on the metro, even though as a route it is the best candidate for this type of operation.

Other options were interurbans and western lines into the Wex, or using the HR harbour crossing (whichever one that was).

Under the old administrative arrangements, a 4tph (peak, 1tph off peak) stopping pattern on sector 3 running Springwood - Parra - City - Chatswood - Gosford over the bridge would have freed up the mains for exclusive use by sector 3, the suburbans for sector 2 and the inner west for rapid transit.  It would have trhe added benefit of eliminating the 20,000 peak movements at Central where 10K PAX/hr shift from Interurban to Suburban services, with half of those journey's having originated in the suburban zone anyway.  Not quite service differentiation for the outer tier (nor anything but a headache for the member for the Blue Mountains Smile), but certainly a much better allocation of the capacity of a second harbour crossing than to Bankstown and Hurstville.

The capacity benefits of the current proposed second harbour crossing to sector 2 is all of 8 slots (the Bankstown Line post 2013), with another 4 slots to Sector 1 which won't need them till 2040 anyway.

Exactly the same thing can be achieved with ATP/ATC for 1/100,000 of the cost.



  djf01 Chief Commissioner

Here is another issue for you guys: how well will the Airport line handle 20tph?

I recently used that line: and I had to unload 1 spouse, 2 small kids, 4 pieces of cabin baggage and 6 pieces of checked luggage (2 oversized) into that 38 second dwell!  That was off peak, weekend.

At the other end of the trip (Zurich Airport, a city 1/4 the size of Sydney) the airport station (in addition to the light rail and bus terminals) had a DD waiting for me with super comfie 2+2 seating on one of it's four 400m (that's right: 400m) bifurcated platforms for the 10 min trip to the city.

But the baggage wasn't a problem, because Swiss railways offered a checked baggage service (provided by trucks believe it or not) to my ultimate destination, 3 trains, 2 gauges and 2 companies away.
  KymN Assistant Commissioner

Location: Sydney
At the other end of the trip (Zurich Airport, a city 1/4 the size of Sydney) the airport station (in addition to the light rail and bus terminals) had a DD waiting for me with super comfie 2+2 seating on one of it's four 400m (that's right: 400m) bifurcated platforms for the 10 min trip to the city.
"djf01"



 the Swiss sure do it well.  In total contrast I once (never again) used the Tube to get to Heathrow.  The doors slammed shut as I was lifting my wife's case onto the train, leaving her on the platform.  Grubby horrible system it is.
  Watson374 Chief Commissioner

Location: Fully reclined at the pointy end.

Here is another issue for you guys: how well will the Airport line handle 20tph?
"djf01"


The rest of the line will cope just fine; it's the two airport stations we need to pinpoint. Admittedly, their use should be spread out across the day, so it might not be as peak-vulnerable as they first seem (unless we end up with foreign tourists attempting to carry four bags, each large enough to hide a schoolboy in, and board a Pocket Rocket in the middle of AM peak, whereupon we can reasonably expect dwell time to blow out).

Perhaps 'assisted boarding' could be provided by the Airport line company...

Exactly the same thing can be achieved with ATP/ATC for 1/100,000 of the cost.
"djf01"


DD/SD?

But I definitely support shifting to ATO/ATP/ATC to allow better use of what we currently have - it would certainly allow an insanely good Sector 1, with or without the stopping run to Hurstville.

It would certainly be advantageous to run everything in Coded, from what I know - balancing human optimisation with computer limitation.

Sudden n00b question: Would it be viable to run the Inner West stopper using SD stock, with better performance being exploited to reduce the lag, between flights of express DD trains?

At the other end of the trip (Zurich Airport, a city 1/4 the size of Sydney) the airport station (in addition to the light rail and bus terminals) had a DD waiting for me with super comfie 2+2 seating on one of it's four 400m (that's right: 400m) bifurcated platforms for the 10 min trip to the city.
"djf01"


the Swiss sure do it well. In total contrast I once (never again) used the Tube to get to Heathrow. The doors slammed shut as I was lifting my wife's case onto the train, leaving her on the platform. Grubby horrible system it is.
"KymN"


I much prefer the Singapore MRT's airport branch.

We all know the main reason for the slow down in speeds in recent years - complexity, leads to unreliability, leads to the need to slow things down to compensate and look like you are on time. When we are running at capacity we no longer have the flexibility to be as complex as we have in the past, and there will be some losers as a result for this. The vast majority, however, will benefit.
"grog"


This.

We have to sacrifice complexity and encourage connections. We have to emphasise a network of simpler, faster and more frequent services, rather then a plethora of one-seat rides.

  simonl Chief Commissioner

Location: Brisbane
I think acceleration drops off once the motor starts producing maximum power - at something like 25km/h.  Then acceleration is inversely proportional to speed.  I don't think the calcs above appreciate how severe the drop off is going to be.
  djf01 Chief Commissioner

I think acceleration drops off once the motor starts producing maximum power - at something like 25km/h.  Then acceleration is inversely proportional to speed.  I don't think the calcs above appreciate how severe the drop off is going to be.
"simonl"


Absolutely right Simonl.    Sorry I kicked this off now, just using teh calculus free high school equations of motion.

Motors are rated by power which is more or less conmstant, not acceleration.  Acceleration is not constant, but a function of speed.  E = 1/2mv^2 so the energy of a train at 50kph is 4 times that of one at 25kph.  It needs 3 times the energy to go from 25 to 50 as it does from 0 to 25.  If your power output is the same, then acceleration is going to decline with speed.  DOH !!!!
  jedimasterc Chief Commissioner

Location: Banned
I don't really. It was only in reply to someone else believing that I forgot the people on that line, by leaving them without a service. But also as a link to strathfield so that A5 can get to wondabye if need be!!!!!!!!
"jedimasterc"

And I still think that you did indeed forget them.

I'd support such shuttles off peak to increase frequency on the Sefton-Carramar stretch.  Surely part of the reason for its poor PT use is the mediocre/poor service!

I don't think it is going to happen in peak hour though.  If you have 4tph reversing at Lidcombe, it will be hard to add 4tph additional services as it would allow less than 8 minutes to turn around, probably less than 6.

Watson, I think the idea of the K2RQ is for 8tph Revesby via Airport and all the cambos in peak going via Sydenham.  Which would be sensible.  You shouldn't mix and match with Cambo via Airport because then a problem on one part of the network isn't isolated.  Similar with Cambo via Granville.

no glenfield won't be viable but liverpool will be with the 4th platform and 2x small stabling facilities.
"jedimasterc"


I was under the impression that Glenfield reversers would become Leppington reversers.
"Watson374"

Absolutely.

Congrats on the attempt itself, can be pretty frustrating I imagine. Off-peak perhaps I'd have:
C'town via Sydenham express: Town Hall, Wynyard, Circular Quay, St. James, Museum, Central, Redfern, Sydenham, Beverly Hills, Revesby, Glenfield, all to C'town.
Liverpool via Regents Park express: Museum, St. James, Circular Quay, Wynyard, Town Hall, Central, Redfern, Ashield, Strathfield, Lidcombe, Regents Park, Chester Hill, Cabramatta, Warwick Farm, Liverpool.
"bowralcommuter"

I think off peak Cambo should be via Airport.  It's more important to increase frequency at Green Square and Mascot than save a few minutes from the Cambo trains.
"simonl"


You have to realise that leightonfield is poor performing because of it's position. It is so close to villawood and it is surrounded by industrial areas and not residential and office areas which normally create the high PT usage. It doesn't have a carpark or bus interchange either which also normally creates patronage. carramar I could concede could be unskipped because after checking my usage statistics it seems to have as much patronage as villawood. But sefton has easy access to chester hill or regents park or birrong.



  simonl Chief Commissioner

Location: Brisbane
Well if the station adds negative value, then close it!

I'll bet you dollars to donuts that there will be no Liverpool-Lidcombe shuttles to serve stations which are missed on the stretch we are discussing.  Which makes it a bit of an unreasonable proposal to miss any stations from Carramar to Sefton.
  Watson374 Chief Commissioner

Location: Fully reclined at the pointy end.
Well if the station adds negative value, then close it!

I'll bet you dollars to donuts that there will be no Liverpool-Lidcombe shuttles to serve stations which are missed on the stretch we are discussing.  Which makes it a bit of an unreasonable proposal to miss any stations from Carramar to Sefton.
"simonl"


That's why I put it on the Liverpool via Regents Park service in that plan I came up with that broke because of pathing limitations.

Sponsored advertisement

Display from:   

Quick Reply

We've disabled Quick Reply for this thread as it was last updated more than six months ago.