Slowness of Sydney trains

 
  tazzer96 Deputy Commissioner

The leppington line is similar in nature to perth's new lines, along with brisbanes Springfield line and the Gold coast and Rosewood segments of their respected lines.  The Leppington line is around 9km long with 1 intermediate station and is timetabled to take 9-10 minutes.   So around 60km/h average.   Not that great by others standards, however it should be noted that it is let down the the speed of which the trains can obtain, only 115 max.  Only yellow boards which mean that a white speed board would be minimum 125.    

The express services in sydney are let down by not being able to travel faster than 115 km/h, while the waratahs, Oscars and M-sets are easily capable of doing 130.

Sponsored advertisement

  simstrain Chief Commissioner

I do desperatly want to see a quad between revesby and macarthur.   Only having station platforms at macarthur, campbelltown, and glenfield.  Would do alot to speed up long distance services.
tazzer96
If quadding occurs past revesby it will most likely only be to east hills. The long distance services lose time (not much) between Glenfield and Macarthur and not between Revesby and Glenfield.

There is no room for 6 suburban tracks at Glenfield but there is room between Glenfield and Macarthur for 4 tracks.
  simstrain Chief Commissioner

Lets do a comparison

13 km segment:


Benchmark:
Perth-Mosman Park (13.5 km): 11 stops, 21 minutes, 3 trains an hour, 20 minute headway
Perth-Beckenham (13.8 km): 9 stops, 19 minutes, 5 trains an hour, 12 minute headway.
Perth-East Guildford (14 km): 11 stops, 20 minutes, 3 trains an hour, 20 minute headway.

Sydney:
Central-Homebush (12.7 km): 12 stops, 27 minutes - 4 trains an hour but up to 11 trains an hour along the inner west line. 5.4 minute headway and interconnectivity with the old main south and western line at homebush. 22 minutes is possible with interchange at Strathfield from express western line service which gets to Strathfield (11th stop on Inner West line) in only 13 minutes.
Sydenham-Bankstown (13.4 km): 10 stops, 25 minutes, 8 trains an hour, 7.5 minute headway, trains pathed to take into account sefton triangle and cabramatta connection with old main south. Connectivity with other lines around the city circle.
(Sydney Metro projected time for 13 km: Sydenham-Bankstown: 10 stops, 21 minutes) 15 trains per hour, 4 minute headways, separate line with ATO, platform screen doors for safety allowing trains to approach and leave at faster speeds. Level boarding for easy access for disabled and people with prams and shopping trolleys.

I say when perth reaches the same frequencies as Sydney then you can start comparing why our system is slower.
  djf01 Chief Commissioner

If they were serious about speeding things up, they would just add a tunnel from granville to the cbd for western line services.
tazzer96

Only it wouldn't be fast because it would have to go via and stop at Lidcombe and Strathjfield, because they are hubs, and Olympic Park and Rozelle.

As I argued in the previous thread, IMHO Sector 3 should operate non-stop Parra-Central in ~15 minutes.  Can a metro do that?

I'm not against metros, what I'm against is building metros to "fix" problems with existing rail.  Metros are hugely expensive, and IMHO should be focused on areas of high concentration with tight station spacing.  

A Syd-Parra makes sense, passing through high density (and growing) greenfields areas with two solid anchor tennants at each end.  But few or no stations to make up for poor running times on the surface heavy rail is just silly.  It's not solving the core problem, just ensuring we don;t get the full value from the new line.

And same with Liverpool to Sydney.  Liverpool suffers because there are too many routes to the city with too many intermediate stations that all need to be serviced.  It needs it's routes rationalised (Syd Metro sort of does this) and sped up (Syd Metro impedes this).  It does not need *another* competing rail service that will inevitably offer another sub-standard service.
  simstrain Chief Commissioner

In finalising I would like to state that Perth's frequencies are significantly down on Sydney's and the Sydney system is significantly more complex vs the dedicated lines that Perth offers. Sure those systems are faster but Sydney's system has always been about frequency and capacity over speed. Please remember that Sydney's train system moves Adelaide's population every day on a legacy system that first serviced passengers in 1855.
  simstrain Chief Commissioner

As I argued in the previous thread, IMHO Sector 3 should operate non-stop Parra-Central in ~15 minutes.  Can a metro do that?
djf01
This is laughable. Sector 3 can only currently be done in 25 minutes with a single stop at Strathfield. The only way to do 15 minutes Central to Parramatta would be on a new dedicated line with trains travelling at 130km/h at the very least.
  tonyp Chief Commissioner

Location: Shoalhaven
Lets do a comparison

13 km segment:


Benchmark:
Perth-Mosman Park (13.5 km): 11 stops, 21 minutes, 3 trains an hour, 20 minute headway
Perth-Beckenham (13.8 km): 9 stops, 19 minutes, 5 trains an hour, 12 minute headway.
Perth-East Guildford (14 km): 11 stops, 20 minutes, 3 trains an hour, 20 minute headway.

Sydney:
Central-Homebush (12.7 km): 12 stops, 27 minutes - 4 trains an hour but up to 11 trains an hour along the inner west line. 5.4 minute headway and interconnectivity with the old main south and western line at homebush. 22 minutes is possible with interchange at Strathfield from express western line service which gets to Strathfield (11th stop on Inner West line) in only 13 minutes.
Sydenham-Bankstown (13.4 km): 10 stops, 25 minutes, 8 trains an hour, 7.5 minute headway, trains pathed to take into account sefton triangle and cabramatta connection with old main south. Connectivity with other lines around the city circle.
(Sydney Metro projected time for 13 km: Sydenham-Bankstown: 10 stops, 21 minutes) 15 trains per hour, 4 minute headways, separate line with ATO, platform screen doors for safety allowing trains to approach and leave at faster speeds. Level boarding for easy access for disabled and people with prams and shopping trolleys.

I say when perth reaches the same frequencies as Sydney then you can start comparing why our system is slower.
simstrain
You must be looking at an old timetable or something. All lines including Fremantle and Midland lines are typically 15 min headways off peak and Armadale 15 mins at outer end and 7 mins at inner end off peak. The Armidale line to Cannington is average 12 trains per hour in peak as are similar sections of the Mandurah and Joondalup lines. Sydney metro has an ultimate capacity up to 30 tph.

So far I'm seeing a whole lot of arguments for wholesale separation and conversion/newbuilding of Sydney lines to metro which is something I didn't originally believe in but, from the information presented, there doesn't seem to be any long-term alternative. That leaves the problem of the interurban lines which can't be solved by metro/S Bahn - although considering Mandurah is the same distance from the centre as Thirroul, Douglas Park, Blaxland and Woy Woy it gives you food for thought! 51 minutes in one of the Transperth suburbans is certainly more comfortable over 70 km than a much longer time in an Oscar or Endeavour, not only because the seats are more comfortable but, most importantly, because the journey is over so quickly. This is what we sorely lack in Sydney and region:



https://www.youtube.com/watch?v=8hGgVC_rOi4
(That's between Murdoch and Cockburn Central which is 6.7 km, similar to Holsworthy-Glenfield which is 6.3 km; Murdoch-Cockburn Central taking 3 minutes compared with Holsworthy-Glenfield taking 5 minutes)

The slowness of Sydney and interurban trains actually has serious implications for urban planning because it inhibits the development of satellite commuting to ease the pressure on the main Sydney urban area. Satellite commuting is basically left to the motor vehicle through construction of motorways because rail can't deliver the journey times. It's actually a very serious long-term issue.
  djf01 Chief Commissioner

As I argued in the previous thread, IMHO Sector 3 should operate non-stop Parra-Central in ~15 minutes.  Can a metro do that?
This is laughable. Sector 3 can only currently be done in 25 minutes with a single stop at Strathfield. The only way to do 15 minutes Central to Parramatta would be on a new dedicated line with trains travelling at 130km/h at the very least.
simstrain

It's a joke alright.  Central to Parramatta is 26km.  

A heavy rail train can't run at 104kph?  

I'm sorry to pick on you @simstrain, but this mentality typifies the cultural problem we have.  A blanket statement "it can't done", when it's blindingly obvious it can be done.  It's just there are just a million neigh on trivial little reasons why not, and no-one arguing the case to do it.  Instead, people want to re-create the problem with a parallel metro.
  tazzer96 Deputy Commissioner

As I argued in the previous thread, IMHO Sector 3 should operate non-stop Parra-Central in ~15 minutes.  Can a metro do that?
This is laughable. Sector 3 can only currently be done in 25 minutes with a single stop at Strathfield. The only way to do 15 minutes Central to Parramatta would be on a new dedicated line with trains travelling at 130km/h at the very least.

It's a joke alright.  Central to Parramatta is 26km.  

A heavy rail train can't run at 104kph?  

I'm sorry to pick on you @simstrain, but this mentality typifies the cultural problem we have.  A blanket statement "it can't done", when it's blindingly obvious it can be done.  It's just there are just a million neigh on trivial little reasons why not, and no-one arguing the case to do it.  Instead, people want to re-create the problem with a parallel metro.
djf01
Its not a case of can it be done.  Almost anything can be done.  It's can we do it without bankrupting the nation.
  djf01 Chief Commissioner



So far I'm seeing a whole lot of arguments for wholesale separation and conversion/newbuilding of Sydney lines to metro which is something I didn't originally believe in but, from the information presented, there doesn't seem to be any long-term alternative. That leaves the problem of the interuban lines which can't be solved by metro/S Bahn - although considering Mandurah is the same distance from the centre as Thirroul, Douglas Park, Blaxland and Woy Woy it gives you food for thought! 51 minutes in one of the Transperth suburbans is certainly more comfortable over 70 km than a much longer time in an Oscar or Endeavour, not only because the seats are more comfortable but, most importantly, because the journey is over so quickly.
tonyp

The fundamental problem with Sydney's network is terminal facilities.  There are too few paths through the city, and those that exist have to cater for all types of traffic.  In 50s and 60s we addressed this problem with higher capacity hybrid trains capable of fulfilling a dual role and avoided building more paths.  In the 90s and 00s we just avoided building new paths.  

The main problem I have with the new Sydney metro system is - being incompatible - doesn't do enough to redress the imbalances - and arguably makes them worse ie, removes 10 stations from the lightly used Bankstown line, adds 5 from the upper northern for little if any net benefit.

A new path should be taking 20-30 stations off the existing network.  But it's not.  It should also allow better route length segregation, but it's not doing that either (much).

This is also at the heart of the speed problem.  There are too few paths, which imposes too few stopping patterns, which means each stopping pattern has too many stations and thus too long a run time.

But ... as @tonyp has already pointed out ... that's just one of many many issues.  

My argument is the real cause of the speed problem is no-one wants to solve it, they're all too busy saying "it can't be done".
  djf01 Chief Commissioner

Its not a case of can it be done.  Almost anything can be done.  It's can we do it without bankrupting the nation.
tazzer96

The nation is going to go bankrupt if the 8:15 from Penrith doesn't stop at Lidcombe?  Give me a break.

Speed is productivity.  Faster services attract more patronage and revenue (there are well established multiples for this), it also reduces operating costs.  If Perth ran as slow as Sydney, they'd have to put on more trains just to maintain their frequencies.
  simstrain Chief Commissioner

As I argued in the previous thread, IMHO Sector 3 should operate non-stop Parra-Central in ~15 minutes.  Can a metro do that?
This is laughable. Sector 3 can only currently be done in 25 minutes with a single stop at Strathfield. The only way to do 15 minutes Central to Parramatta would be on a new dedicated line with trains travelling at 130km/h at the very least.

It's a joke alright.  Central to Parramatta is 26km.  

A heavy rail train can't run at 104kph?  

I'm sorry to pick on you @simstrain, but this mentality typifies the cultural problem we have.  A blanket statement "it can't done", when it's blindingly obvious it can be done.  It's just there are just a million neigh on trivial little reasons why not, and no-one arguing the case to do it.  Instead, people want to re-create the problem with a parallel metro.
djf01

No. I'm saying it can't be done with our current rail system. You are comparing an 1855 rail system that has had bits and pieces added on over the years with interconnected components vs a new 1990's system in perth that has dedicated lines that are mostly straight. The old system is going to be slower and there isn't anything that can be done about it which is why people like myself see the need for a new rail system in Sydney that isn't connected with the existing rail system. I would love to have a 15 minute rail journey from Liverpool but I know it isn't going to happen on the current Sydney system.

25 minutes from Parramatta is not too bad when you consider how long that journey takes by car. The bathurst bullet takes 24 minutes for the journey without any stops which shows you that there is no capabilities for faster trains along this corridor.
  RTT_Rules Oliver Bullied, CME

Location: Dubai UAE
Having travelled on the Perth system a lot lately and on metros overseas and then looking at the specifications of the Sydney metro, I don't see a lot of practical difference between the Transperth operation and Sydney metro, apart from the Sydney automation and platform screens (and Perth is apparently looking at adopting ATC). Technically, one is an S Bahn, one a U Bahn, but they morph into each other and the boundaries are blurred. Trains are similar, journey times are similar, comme ci, comme ça. I think the Sydney project should borrow a bit from Perth and give the trains a higher maximum speed so that it can perform even better and be more S Bahn-like over longer distances (those 130 km/h leaps that Perth can do at the outer ends).

Because it's called a metro, critics are thinking of it too much within traditional definitional boundaries rather than accepting that it has elements of S Bahn as well. Even the arguments about seats and longer distances, critics argue on a car by car comparison whereas, because the metro is capable of throughputting more trains per hour, the difference between the potential number of seats delivered per hour closes up - iirc at maximum potential capacity 15,000 seats per hour per direction for metro vs 18,000 seats per hour for the double deck system. Not that great a difference - although the double deckers obviously become more essential in terms of seats towards the outer ends of the metropolitan area and in the interurban area.

What I'm saying basically is that in the debate we shouldn't get too hung up on the "metro" aspect of the Sydney metro. The definitional edges are quite soft and rubbery in Sydney's case. It's more like an S Bahn using metro technology.
tonyp
Its unbelievable how many people are getting hung up on the name and the first and only comparison they can make is the old (not even the new) Paris Metro lines.
  RTT_Rules Oliver Bullied, CME

Location: Dubai UAE
Yes the Sydney system is slow but that is because when you have 20 trains an hour doing vastly different runs on the main trunk routes of an interconnected network (that clearways did very little to actually fix) what else can you expect. A third track pair between Strathfield and Lidcombe would help the situation vastly. Quadding to epping on the north, Sutherland on the south coast and between Glenfield and Macarthur on the main south is what is needed to speed things up.


The liverpool metro extension does make sense to me as well RTT but since our mayor is now labor instead of liberal the extension has been cancelled from what I understand. Liverpool has traditionally been a safe labor seat and former liberal mayor Manoun is something that I never thought I would ever see in Liverpool. Since his departure all discussion has ceased with the current state government and this silly new express which really isn't an express but just a move of the 55 minute via granville service to via bankstown is not an actual express in anything other then the amount of stops it takes. The new metro would be better off going underground to Bankstown freeing up the existing line to run a proper express.

There is no point in comparing Sydney to the other state capitals because of the complexity and the high frequencies of the network and things like the gaps at certain stations rather then the lack of doors is why dwell times tend to blow out. Getting rid of those close stations is also pointless because those close station gaps is what makes out train system popular and putting more people on buses is not the answer.
simstrain
With the exception of Clyde, removal of close stations just leaves a large gap, you need to adjust the station spacing for 1-2 stations either side or redo the track alignment such as Wollestoncraft and Waverton to merge to one. But most of the time these are $1B projects.

Bankstown fine be surface. You wouldn't build the Metro out there UG, may as well go to Hurtsville or Inner West if going to do that.
  RTT_Rules Oliver Bullied, CME

Location: Dubai UAE
A few years back, On Time Performance (OTP) was suffering badly.

Solution - New TT was introduced to slow running times - OTP suddenly back to acceptable %.
mikesyd
Yes but wasn't that because the original long running TT was a work of fiction only obtained by using trains without speedo's and no automated doors or door closing announcements. As soon as trains had speedo's and worse, blackbox recording the speed, the TT became unworkable.
  djf01 Chief Commissioner

The old system is going to be slower and there isn't anything that can be done about it ...  I would love to have a 15 minute rail journey from Liverpool but I know it isn't going to happen on the current Sydney system.
simstrain

Rubbish - apart from the bit that it's not going to happen.

There is plenty that could be done about it.

This 1855 argument is total BS.  Is the line still single track steam hauled on wrought iron rails?  If anything, it's harder and more expensive to re-do things because you're building greenfields in an urban environment.
  RTT_Rules Oliver Bullied, CME

Location: Dubai UAE
As I argued in the previous thread, IMHO Sector 3 should operate non-stop Parra-Central in ~15 minutes.  Can a metro do that?
This is laughable. Sector 3 can only currently be done in 25 minutes with a single stop at Strathfield. The only way to do 15 minutes Central to Parramatta would be on a new dedicated line with trains travelling at 130km/h at the very least.

It's a joke alright.  Central to Parramatta is 26km.  

A heavy rail train can't run at 104kph?  

I'm sorry to pick on you @simstrain, but this mentality typifies the cultural problem we have.  A blanket statement "it can't done", when it's blindingly obvious it can be done.  It's just there are just a million neigh on trivial little reasons why not, and no-one arguing the case to do it.  Instead, people want to re-create the problem with a parallel metro.
djf01
You will not get 140 (or even 104)km/hr out of the existing ROW without removing 1-2 tracks or buying up at least +15m of the adjacent corridor to enable proper 140km/hr alignment. The corridor looks straight but its riddled with lots of little curves and S bends to move around station platforms and everything else they needed to do to jam 6 tracks in there.

The Parrellel but UG solution avoids all of this and likely for same or lower cost. The realestate along side the railway line allong the inner west is now +$3m a house and probably only 8m each, do the math, I get nearly $10B plus doing all the civils, moving services etc etc. May as well dig a hole!!!
  RTT_Rules Oliver Bullied, CME

Location: Dubai UAE
The old system is going to be slower and there isn't anything that can be done about it ...  I would love to have a 15 minute rail journey from Liverpool but I know it isn't going to happen on the current Sydney system.

Rubbish - apart from the bit that it's not going to happen.

There is plenty that could be done about it.

This 1855 argument is total BS.  Is the line still single track steam hauled on wrought iron rails?  If anything, it's harder and more expensive to re-do things because you're building greenfields in an urban environment.
djf01


In the built up areas of Sydney your options are limited and hence they are now going underground.
  RTT_Rules Oliver Bullied, CME

Location: Dubai UAE


So far I'm seeing a whole lot of arguments for wholesale separation and conversion/newbuilding of Sydney lines to metro which is something I didn't originally believe in but, from the information presented, there doesn't seem to be any long-term alternative. That leaves the problem of the interuban lines which can't be solved by metro/S Bahn - although considering Mandurah is the same distance from the centre as Thirroul, Douglas Park, Blaxland and Woy Woy it gives you food for thought! 51 minutes in one of the Transperth suburbans is certainly more comfortable over 70 km than a much longer time in an Oscar or Endeavour, not only because the seats are more comfortable but, most importantly, because the journey is over so quickly.
The fundamental problem with Sydney's network is terminal facilities.  There are too few paths through the city, and those that exist have to cater for all types of traffic.  In 50s and 60s we addressed this problem with higher capacity hybrid trains capable of fulfilling a dual role and avoided building more paths.  In the 90s and 00s we just avoided building new paths.  

The main problem I have with the new Sydney metro system is - being incompatible - doesn't do enough to redress the imbalances - and arguably makes them worse ie, removes 10 stations from the lightly used Bankstown line, adds 5 from the upper northern for little if any net benefit.

A new path should be taking 20-30 stations off the existing network.  But it's not.  It should also allow better route length segregation, but it's not doing that either (much).

This is also at the heart of the speed problem.  There are too few paths, which imposes too few stopping patterns, which means each stopping pattern has too many stations and thus too long a run time.

But ... as @tonyp has already pointed out ... that's just one of many many issues.  

My argument is the real cause of the speed problem is no-one wants to solve it, they're all too busy saying "it can't be done".
djf01
Perth is a sand flat, Sydney is not especially on Nth Sth alignment.

Sydney was running 3min terminating services into St James when Perth barely new what a train was. Perth is now basically a radial of just three lines criss crossing the city with little more than 1 or 2 stations common. Yes engineering and operationally wise Perth is easier and more modern not even being sparked until 1990's and hence you would think things are better.

The Bankstown line is 8 trains per hour, packed to plus 130%. Its far from lightly used but the Metro will certainly make it far more attractive than the current sardine status and see more bodies gravitate too it. Going to Liverpool was the icing.

Nothing wrong with the train incompatibility issue, how does it help? Most major cities have dedicated trains per line these days such as Paris and London as even if you wanted to run a train on another line, you cannot as no capacity to do so. Sydney is now there as well.

Other comments that it doesn't need to be a Metro as stations are too far apart, actually the reverse is more true. The stations are designed to be high volume stations where boarding time will be longer, hence rolling stock that does this better will be faster. Does the Metro need to be 140km/hr capable, hardly, the stations are not that far apart. Sydney's issue is not that trains cannot do 140km/h, rather they spend too much time below 80km/h. The Metro design helps solve that. 1.5km apart a train with a top speed of 90-100km/h will only hold that speed briefly between stations.

Its damn near impossible to remove 20-30 stations off the existing network with the total Metro route km. This would also require 100% parallel running of the existing lines. Obviously over 15km is new to rail territory, reclaimed ECRL, one new location stations on way to City plus Nth Sydney, three new stations in the city new to rail or supporting Town Hall. South of city its about removing one of the branch lines from the CBD tunnels. The choices were Inner West, Banktsown or Hurtsville. Hurtsville needed to be UG to work, Inner West most likely the same. Only Bankstown could be surgically removed with minimal impact on the network which if they ran to bloody Liverpool would have solved the main issue.

As you said the heart of the problem is multiple stopping patterns which are designed to allow multiple tier operation on a single pair of tracks. But growth is forcing this out and additional expansion of track capacity by adding more tracks on certain corridors. However simply making the corridor wider in most locations is now becoming extremely expensive due to land costs and in most cases no spare or easy land to take.
  simstrain Chief Commissioner

I will concede that trains on the new main south line should be doing 130km/h or more to reduce these intermediate times and reduce overall travel time from Campbelltown since that was the purpose of the quad to revesby. The problem is still in the inner city where the lines all merge into one another.

The rest of the network however has significant issues with conflicting moves where lines cross at Homebush and Strathfield for instance. The points at Macdonaldtown are another example. Nsw intercity and regional trainlink services interfere with Illawarra and South Coast services between Erskineville and Sydenham is another example. Removing Bankstown services could allow this conflict to be removed with Illawarra services in and out of the ESR.

I do like the idea of a metro west but a tunnel from Strathfield to Lidcombe to add an extra track pair is still needed regardless of if the metro west gets built.
  djf01 Chief Commissioner

It's a joke alright.  Central to Parramatta is 26km.  

A heavy rail train can't run at 104kph?  
You will not get 140 (or even 104)km/hr out of the existing ROW without removing 1-2 tracks or buying up at least +15m of the adjacent corridor to enable proper 140km/hr alignment. The corridor looks straight but its riddled with lots of little curves and S bends to move around station platforms and everything else they needed to do to jam 6 tracks in there.
RTT_Rules


Who said 140kph?  104 is the average speed needed to achieve 15 minutes.

I've checked the curve & grade diagram: IMHO it should be 120 from Parra to Stanmore, and 80 from Stanmore to Central (with the option of a 100+ stretch between Newtown and the flying junctions).  

The curve & grade diagram has all the curve radii described in Chains Smile, but it's enough.  The curve near the Liberty St overpass at Stanmore is 20 chains (400m radius).  If I read the diagrams right there are some tight curves on the mains approaching the platforms at Flemington.  

It seems to me - without doing a complete engineering audit - the main impediment to faster speed boards (if any) are all the facing points, most of which simply aren't needed.

Even without any capital works at all, running at the speed boards would achieve ~19 minutes, and I'd be very surprised if 16 min wasn't achievable without any capital works beyond certifying faster speed boards.


The Parrellel but UG solution avoids all of this and likely for same or lower cost. The realestate along side the railway line along the inner west is now +$3m a house and probably only 8m each, do the math, I get nearly $10B plus doing all the civils, moving services etc etc. May as well dig a hole!!!

And project involving land requisition and curve straightening could achieve a min curve radius ~1000m, and a running speed of 160 - and hence a transit time of 10 minutes.  That's why I said 15min.  It's perfectly possible - and regularly/typically achieved elsewhere in the world (including Australia) using conventional rail technology within the existing corridor.  Just not in an S-Set.

The problem is no-one wants to do it.
  tonyp Chief Commissioner

Location: Shoalhaven
I've checked the curve & grade diagram: IMHO it should be 120 from Parra to Stanmore, and 80 from Stanmore to Central (with the option of a 100+ stretch between Newtown and the flying junctions).  

The curve & grade diagram has all the curve radii described in Chains Smile, but it's enough.  The curve near the Liberty St overpass at Stanmore is 20 chains (400m radius).  If I read the diagrams right there are some tight curves on the mains approaching the platforms at Flemington.  
djf01
Years ago when I was doing some railway work, I annotated my old C&G book to convert chains to metres but I also got from somewhere in the organisation some speeds for the curves (that is speeds according to physics, not speed boards). Those I noted include 80 km/h for 500 metres/25 chains and 128 km/h for 1,000 metres/50 chains. Browse through your diagrams and see how much actual restriction is imposed by curves on the Sydney system. The sort of gradients typical on the Sydney plain shouldn't be an issue for EMUs.

I share your frustration at people coming up with endless reasons why something can't be done rather than outline technical steps that need to be taken in order to solve the issue. I used to joke to myself that the Australian national motto must be "it can't be done mate!", so often it seemed that I heard it!
  RTT_Rules Oliver Bullied, CME

Location: Dubai UAE
If we accept what is currently being built as done and move on this is how I see the future options as I agree the average speed to get to the city is slow'ish and the average speed simply to move around the suburban area is far worse and hence why trains are not used for this on a large scale. Classic is Wollestoncraft and Waverton. In peak all trains stop because they have too due to frequency, but in off-peak there are expresses through this twisted outdated screw up.


1) North Shore Line
- As mentioned before, bypass both Wolls' and Waverton with a straight tunnel, build just one station, save about 5min for everyone travelling from St Leonards to potentially the Central Coast via NSL, not chicken feed and alot of complaining about noise by locals. Also enable Laverton Bay line to be closed and sold/green spaced and old alignment of NSL to be greenspaced and partly redeveloped.

2) Western Metro, Inner West Metro, Top Ryde Metro
- From St James, Pitt Street, Central, Syd Uni/hospital and then seperate into lines

- Inner West, replace most of the surface trains with a UG line as far as potentially Straithfield. Macdonaldtown not replaced and Newtown moved across Parramatta Road, rest pretty much close to existing station. Only Ashfiled and Burwood retained, rest demolished. Main west corridor rebuilt to enable +140km/h running west of Redfern. Speed up both frequency and speed of services from SW

- Western Metro, services Bays District, Rosehill and west to Paraamatta/Westmead

- Branch off around Five Dock and head via Gladsville to Ryde and connection to Main Nth

All three lines use same stations through the city, bi-bifurcated platforms will easily support 90sec intervals, saving significant cost.

3) Quad Revserby to East Hills, rebuild the line west to Glenfield to 140-160km/hr running

4) Quad to Macurthur or at least Campbelltown

4) Relocate South Coast services to terminate at unused platforms at Central 26 and 27

5) Extend Bankstown Metro to Liverpool direct.

6) Extend NW Metro to at least Schofields

7) Quad Hurtsville to Sutherland, but no need for additional platforms.

8) On western corridor, merge the Nth Main suburban and Central Coast services to same pair of tracks as same stopping frequency to Straithfield and then onto Main North, Quad to Epping

9) Western Surburban and Blue Mountains would use inner tracks to Paramatta

10) Remove many of the cross overs on western corridor as of less value than the past.

11) LR Carlingford to Parramatta va Rosehill, close Clyde Station

12) extend ESR to Nth Bondi and open the mystery station
  simstrain Chief Commissioner

So RTT here are my feelings to you point work.

1. No getting rid of the W's. The new Metro will more then solve the problems you talk about and the W's actually keep 4,000 people a day off the roads because of their handy closeness to locals. The locals don't even notice the noise since the stations are surrounded by greenery. I think that laverton bay could be built over (not apartments) without removing the trains as well.

2. a) Newtown station isn't anywhere near Parramatta road so I am not sure what you are getting at there.
b)Why are you demolishing all of the inner west stations for? They provide a great local service and if you demolish them it will only make road traffic worse. If inner west trains were all stations services and express trains actually used the suburbans instead of the locals then along with the tunnel from Strathfield to Lidcombe a fair chunk of additional capacity could be added.
c) What are you trying to achieve running a metro branch line to the main north line? I do not understand the purpose of this.

3. absolutely 160km/h running on the line from wolli creek all the way glenfield.

4. a) yes. b) Yes along with a whole new connection into central terminal to allow speedier ins and outs for nsw trainlink services to the main south.

5,6 and 7. yes

8. Quad to epping is a yes but the other part is not advisable because central coast services terminate at Sydney terminal and northern line services cross the bridge. Hence why there is a flyover at Strathfield for suburban services. This is one of the reasons of the slow travel time

9. Already occurs but the suburban services have to switch across at either homebush or macdonaldtown.

10. Those crossovers are of extreme importance because the western line is also used for freight outside of peak hours. A freight train breakdown is why those points are needed and why you can't have a 160km/h train on a line where a freight train could be doing only 80km/h.

11. Already happening

12. I'm not sure about that one.
  djf01 Chief Commissioner

If we accept what is currently being built as done and move on this is how I see the future options as I agree the average speed to get to the city is slow'ish and the average speed simply to move around the suburban area is far worse and hence why trains are not used for this on a large scale. Classic is Wollestoncraft and Waverton. In peak all trains stop because they have too due to frequency, but in off-peak there are expresses through this twisted outdated screw up.
RTT_Rules


My cheaper version of these projects ...