Ok, lets go back to the start.
What is the No.1 target of the NSW govt and previous in providing a line to some where in NW (former inner NW project or current NWRL project). AUTOMATION. Why? Because Legacy systems have legacy's including higher operating costs.

I think you'll find that someone or someones in the NSW bureaucracy (Rod Staples is the public face of this) has been pushing an "metro" agenda for some time in NSW.  They just adapt it to suit the ideology of the government in power at the time, in this case putting one over those pinkos in the union by not having any staff.

I simple don't believe for one second the introduction of the NWRL format is going to reduce operating costs - at least not for the government.  In large part this is an exercise in cost shifting.

For instance, the new NWRL fleet will be cheaper to operate per vehicle - probably quite a lot - primarily because it is new.  Yet CityRail's fleet age profile will continue to deteriorate (despite the A sets it's still on the wrong side of the curve).  If the systems had a common fleet then both systems would benefit proportionately from fleet modernisation.   But as it stands all that benefit (all at taxpayer's expense I might add) goes to the privatised NWRL, while the cost of the older fleet profile is left with poor old inefficient CityRail and the NSW taxpayer.

In terms of efficiencies, the one genuine efficiency the new fully automated system will offer is zero crewing.  This is offset to some degree by:
- the need to run 50% more trains, and associated maintenance costs
- the extra costs imposed on CityRail to support the isolated NWRL.  (The biggest one is the Chatswood interchange - RailCorp will need to either build a new high capacity turnback at Chatswood - and there isn't really space to do it after they sold the air rights over the station - or run a lot of empty trains to the various turnbacks up the north shore.)
- the bringing forward of the 2nd harbour crossing, and the inefficient allocations of pathing resources imposed by the new format.

Another point I think worth considering is CityRail/RailCorp would not be allowed to run the the NWRL in it's proposed (a bit late to use that word) format purely on safety grounds.  They simply wouldn't be allowed to pump 1000 up a tunnel without there being some member of staff there to co-ordinate an evacuation.  Nor would they be allowed to run trains sufficiently light (ie lack of crash survivability) to achieve the specified speeds up those 1:20 grades.  That's not to say we do or don't necessarily need this much safety.  It's just pointing out that these "efficiencies" are being achieved by lowering safety standards.  It may well be appropriate to lower excessive safety standards to reduce costs.  But it makes no sense to do it on just one new line alone.

I actually remain to be convinced they will be able to fully implement crewless operations on the NWRL.  There are a lot of risks and very few (well patronised) systems around the world that have done it effectively.  How well those trains go up the 1:20 grades on a wet day with a robot deploying sand remains to be seen.  Similarly with boarding, particularly at Chatswood and perhaps Epping.  On it's own it's a laudable goal, but the benefits need to kept in perspective.  Most "fully" automated systems around the world do retain a single crew member to be on-site to deal with whatever it is the automation can't.  For high volume applications the benefits of having a human at the pointy end can outweigh the costs.  Where automation seems to work is with low volume systems: like Docklands LR or the Vancouver SkyTrain, where there just aren't that many PAX to introduce the human element.  

For mine, whatever real net advantages the NWRL format may have, these are dwarfed by additional and unnecessary the problems it's causing.

I think you'll find that someone or someones in the NSW bureaucracy (Rod Staples is the public face of this) has been pushing an "metro" agenda for some time in NSW. They just adapt it to suit the ideology of the government in power at the time, in this case putting one over those pinkos in the union by not having any staff.

I simple don't believe for one second the introduction of the NWRL format is going to reduce operating costs - at least not for the government. In large part this is an exercise in cost shifting.

For instance, the new NWRL fleet will be cheaper to operate per vehicle - probably quite a lot - primarily because it is new. Yet CityRail's fleet age profile will continue to deteriorate (despite the A sets it's still on the wrong side of the curve). If the systems had a common fleet then both systems would benefit proportionately from fleet modernisation. But as it stands all that benefit (all at taxpayer's expense I might add) goes to the privatised NWRL, while the cost of the older fleet profile is left with poor old inefficient CityRail and the NSW taxpayer.

In terms of efficiencies, the one genuine efficiency the new fully automated system will offer is zero crewing. This is offset to some degree by:
- the need to run 50% more trains, and associated maintenance costs
- the extra costs imposed on CityRail to support the isolated NWRL. (The biggest one is the Chatswood interchange - RailCorp will need to either build a new high capacity turnback at Chatswood - and there isn't really space to do it after they sold the air rights over the station - or run a lot of empty trains to the various turnbacks up the north shore.)
- the bringing forward of the 2nd harbour crossing, and the inefficient allocations of pathing resources imposed by the new format.

Another point I think worth considering is CityRail/RailCorp would not be allowed to run the the NWRL in it's proposed (a bit late to use that word) format purely on safety grounds. They simply wouldn't be allowed to pump 1000 up a tunnel without there being some member of staff there to co-ordinate an evacuation. Nor would they be allowed to run trains sufficiently light (ie lack of crash survivability) to achieve the specified speeds up those 1:20 grades. That's not to say we do or don't necessarily need this much safety. It's just pointing out that these "efficiencies" are being achieved by lowering safety standards. It may well be appropriate to lower excessive safety standards to reduce costs. But it makes no sense to do it on just one new line alone.

I actually remain to be convinced they will be able to fully implement crewless operations on the NWRL. There are a lot of risks and very few (well patronised) systems around the world that have done it effectively. How well those trains go up the 1:20 grades on a wet day with a robot deploying sand remains to be seen. Similarly with boarding, particularly at Chatswood and perhaps Epping. On it's own it's a laudable goal, but the benefits need to kept in perspective. Most "fully" automated systems around the world do retain a single crew member to be on-site to deal with whatever it is the automation can't. For high volume applications the benefits of having a human at the pointy end can outweigh the costs. Where automation seems to work is with low volume systems: like Docklands LR or the Vancouver SkyTrain, where there just aren't that many PAX to introduce the human element.

For mine, whatever real net advantages the NWRL format may have, these are dwarfed by additional and unnecessary the problems it's causing.

I have to call you on the low volume systems. The Victoria line is a fully automated line it handles 210 Million Passenger Journeys a year, It is the most intensely used line in the UK. The point is though that a lot of automated systems still need to retain staff in case something goes wrong.

Michael

I have to call you on the low volume systems. The Victoria line is a fully automated line it handles 210 Million Passenger Journeys a year, It is the most intensely used line in the UK. The point is though that a lot of automated systems still need to retain staff in case something goes wrong.

Michael

Called.  I used the wrong phrase: "automatic" instead of "crewless".

Rod Staples has probably done a bit of research on Automated systems and hence realised significant costs can be saved.


Power Supply
You mentioned in another thread why use the 1500VDC overhead?
- Ground based are more expensive to install
- Singapore's NE line used 1500VDC for its 100% underground NE line because of cost and no need to have the visual issues
- Hanging a wire from the roof is alot cheaper than building the track side 3rd rail and uses twice the voltage
- Madrid when one step further, ridgid O/H 1500VDC
- Why not use 25kVA AC, simple (equivalent to about 33,000V DC in arcing across air), you need a certain level of clearance for such high voltages. They may not have this?

Costs of Auto
You mentioned you find it hard to believe its cheaper?
- To run a 15min timetable from the city to the end of the NWRL (roughly 1hr up and 2hr return including layover), you need 8 crews (note I'm using DOO). To run from 5am to 1am 7 days aweek you need 3.5 x 8 = 28 people + leave + sick + training + support for these people, pushing 40 people. x $100k a year to include O/T, Super, Training, uniforms etc etc = $4mpa as a a starter but probably over $5m when other factors are added like cost of maintaining the drivers cabs, higher costs in train design. A bank of PLC's takes up less space than a drivers seat and doesn't have an extra door. If you have two man crews, double the cost.

- Most Automated trains still have roving security for revenue and asset protection, they are mobile, active and far more visual. Some networks have them on every train, some like Singapore and Vancouver random. Dubai has an attendant who is there to protect the usage of the women's and 1st class cars because the bulk of the users from the sub continent think they are better than everyone else and ignore clearly marked signs. UAE is also very protective of woman's isolation from males when they want it. I was told this is very much for cultural reasons more than safety and security but yes they can also drive the train and I have seem them lift the panel at times to review something.

- As automated trains often tend to be smaller because they run more frequent, operating costs can be less. In Sydney you have a train capable of hauling 1000 people outside peak usually running around with a bus load or in some areas and late at night a maxi taxi load. Using a 350t train with 2 crew is hardly efficient. This is clearly evident when trains run around with 50% of the cars lights off. Dragging around 200t of dead wait is hardly efficient or low cost maintenance practices. Labour costs are two high to justify splitting the train for 6hr of operation. Auto trains you run as required, responding to turnstyle volumes live (obviously to within operational agreements on frequency). The comments I found on the Singapore NW line was the system being auto had lower maintenance costs, lower energy costs, greater flexibility and more reliable. Trains on an auto system are more to live user demands, than crewing needs planned months in advance. In peak running more trains to cater for smaller capacity is same as running bigger trains. The higher frequency also has the ability to attract more users. We all know the greater the frequency the more attractive a system is.


EDIT: About half of this post got deleted some how. Anyway, the NWRL will be what it is and hopefully expanded quickly to the city and south as indicated and as well in the inner west and Epping to Paramatta. I think those who doubt the technology are best served by hoping on a plane and seeing for yourself what can be and is done elsewhere more successfully, more reliability and with higher safety outcomes than most manual driven systems achieve and why more and more are being built and more are being converted to automation. If you don't want to or cannot afford a plane ticket, then in 5 years you will get your chance to see how it will work and why we have this. Or in mean time read up on Madrid, Barcelona, Paris, London, Stockholm, Nuremberg, Vancouver, Singapore, Dubai... there is plenty of info in WIKI as a starter.

There's little point in continuing to debate the merits of the auto rapid transit system for the NWRL as it's been done to death.  It is now a done deal and there's no going back.  

However, its future extension across the harbour and conversion of the existing Bankstown Line IS still open for debate.  It's a question of finding the best outcome.  The initial scheme to convert the line to Cabramatta and Lidcombe has already been pulled back to Bankstown without any explanation why.  Have they belatedly realised that it just wasn't feasible because of the disruption it would cause to the existing South Line network?  The previous rapid transit conversion to Hurstville is also absent in the latest proposal.  If it was supposed to be part of a long term plan, then why would they neglect to mention it?  

Putting aside the NWRL issue, my biggest criticism is this ideological obsession with privatising everything the government can lay its hands on by confiscating parts of the existing network and neglecting to factor in the cost and disruption in sidelining Sydney Trains.  I'd like to see a cost/benefit analysis of the relative merits of converting parts of the existing network to rapid transit, including the cost of lengthy shutdowns and the disruption it causes, compared with the upgrading of the existing network with improved signalling and track enhancements permitting greater frequencies.

An earlier option to convert the North Shore Line across the harbour bridge to single deck operation with increased frequencies and connect it with the Illawarra Line was abandoned because it was found to be totally impracticable with the massive cost and complexity of reconstructing the flying junctions at Central and the unacceptable disruption it would create on the busiest part of the network.  It was also found that Town Hall and Wynyard stations wouldn't be capable of handling the (alleged) increased passenger flow. How do we know whether there may be other constraints similarly placed on the Bankstown Line conversion?

There are still other options for the government to reduce the operating costs of the existing network, such as dispensing with guards, difficult as it may be, but nonetheless something which sooner or later will have to be faced.  The building of a new rapid transit network in greenfield locations is another matter entirely which is what I have been advocating.

However, I don't think that the Parramatta to Epping Rail Link will be one of them, particularly if the Parramatta to Macquarie Park Light Rail Line goes ahead which looks likely.  The forecast patronage just doesn't warrant an expensive rapid transit proposal at more the 3 times the cost of the light rail for a similar benefit.  It's been stated on another forum that the previously proposed stub tunnels towards Parramatta at Epping on the new NWRL have been quietly dropped.  Now you wouldn't want to cause any disruption to a rapid transit service to construct them long after the line opens, would you?

Costs of Auto
You mentioned you find it hard to believe its cheaper?
- To run a 15min timetable from the city to the end of the NWRL (roughly 1hr up and 2hr return including layover), you need 8 crews (note I'm using DOO). To run from 5am to 1am 7 days aweek you need 3.5 x 8 = 28 people + leave + sick + training + support for these people, pushing 40 people. x $100k a year to include O/T, Super, Training, uniforms etc etc = $4mpa as a a starter but probably over $5m when other factors are added like cost of maintaining the drivers cabs, higher costs in train design. A bank of PLC's takes up less space than a drivers seat and doesn't have an extra door. If you have two man crews, double the cost.

I'm not for one second suggesting CityRail is somehow efficient, or that two man crewing is a good idea, far from it.

When I last worked some numbers to try and figure out the cost of running a CityRail train, the cost I came up with was about $4mil pa per train set, of which *about* $1mil was crewing. Very rough because the actual data isn't available and it also depends on a lot of factors such as how much the set runs off peak, it's age and so on. But that $4mil included everything, including a reasonable amount for provisioning (ie capital cost/depreciation/however you want to account for it).

To provide it's 4min frequency, the NWRL will need to run 15 trains peak, and probably ~7 off peak (10 min frequency).

CityRail would service the line need to operate 8 train sets peak (8tpj, 7/8min freq) and 4 off peak (15 min freq).

If you just look at crewing costs and just on the NWRL there is, no comparison. NWRL: $0, CityRail $8mil pa.

If you look at all variable costs (and I'll crudely say $1mil/train for NWRL, $2.5mil for CityRail) then it's NWRL: ~$10mil, CityRail: ~$15mil. (CityRail closes the gap by running fewer but larger trains).

If you include total cost of operation and ownership, then (@$2mil per NWRL train vs $4mil CityRail train) then it's NWRL: ~$20mil, CityRail: ~$24mil. This metric is where the double decker's reveal their productivity: you just don't need as many trains, which is why we went that way in the first place.

*If* you have to crew the NWRL trains then even with 1/3rd crewing costs of CityRail, the costs work out about even.

That's only if you consider just the NWRL. When you bring in how CityRail will have to change to support the NWRL in either format you get a very different picture. (There will necessarily be a lot of supposition here, and I'll ignore the effect of closing the ECL on the basis the impact will be the same for both scenarios).

If CityRail were to operate the NWRL, they would need 16 extra train sets to provide the peak service of 8tph RH to the City. However, almost certainly there would be a reduction in frequency Chatswood - Hornsby by 2tph (because of slot availability if nothing else), meaning the net number of extra trains they need to run is 12.

With the isolated NWRL they will need tun an extra 8tph Chatswood to the City, or 8 train sets at 8tph. Again, lack of slots means they won't be able to do this, so it'll probably be 6. However, there is no turnback at Chatswood. These trains will need to be turned at Linfdield, Gordon or (more likely) Hornsby. On balance, will probably be an *increase* in service frequency on the upper north shore of 2tph, meaning the total number of "new" train sets CityRail will need top run is 10.


So by comparison, 10 "new" CityRail trains to support the "efficient" NWRL. 12 "new" trains to operate the NWRL directly.

Now matter how efficiently the NWRL is (and inefficient CityRail is), the whole system is not going to cost less than 2 CityRail trains no-matter how you account for it. Hence my argument that the system is going to be more expensive than had they just left it with RailCorp.

This inefficiency is created by the topology (ie over-servicing the branches while under-servicing the trunk), and that is created by the system incompatibility. And the solution? $25bil for a new harbour crossing we don't need.

Personally, I don't think we're going to see the second crossing, because the patronage to justify it's price tag won't be there. More likely Sector 3 will be converted to "Rapid Transit" and those inner city suburbs like Penrith and Richmond will benefit from those higher frequencies.

An earlier option to convert the North Shore Line across the harbour bridge to single deck operation with increased frequencies and connect it with the Illawarra Line was abandoned because it was found to be totally impracticable with the massive cost and complexity of reconstructing the flying junctions at Central and the unacceptable disruption it would create on the busiest part of the network. It was also found that Town Hall and Wynyard stations wouldn't be capable of handling the (alleged) increased passenger flow. How do we know whether there may be other constraints similarly placed on the Bankstown Line conversion?

I think you'll find the 2nd crossing is a $20bil+ flyover.

There's little point in continuing to debate the merits of the auto rapid transit system for the NWRL as it's been done to death. It is now a done deal and there's no going back.

However, its future extension across the harbour and conversion of the existing Bankstown Line IS still open for debate. It's a question of finding the best outcome. The initial scheme to convert the line to Cabramatta and Lidcombe has already been pulled back to Bankstown without any explanation why. Have they belatedly realised that it just wasn't feasible because of the disruption it would cause to the existing South Line network? The previous rapid transit conversion to Hurstville is also absent in the latest proposal. If it was supposed to be part of a long term plan, then why would they neglect to mention it?

I still maintain my prediction that this is how our system, will look ~2030: The North Shore and the suburban tracks on Sector 3 converted to "Rapid Transit".

On the costs, this is the size of the train chosen for Dubai, built from 2006, 75 km of track moves 400,000 per day. By 2020 Expo with another 50km of extensions this technology is projected to move 1,000,000/day (same as Cityrail now) and the Red line will itself be approaching 75km long.


This is the size of the train used to run Skytrain Vancouver, built since the late 80's adding one new line per decade. Only slightly shorter in route km moving a similar number of bodies per day.

There must be a reason the newer systems with lines, travel times and people movements similar to the NWRL are going this way. Part of its frequency, but when you don't need to worry about manning, the whole dynamic of the operations and design changes.

I wouldn't say that the operations cost of running these above train equals that of a Cityrail A set (ignoring the driver and guard costs) on a train of train basis.

I don't know the detail costs of any system, but I look at trends. One or two cities could get it wrong, but not the numbers out there now and growing.

Timetable post NWRL opening.

My understanding there is 18 trains per hour through the city at the peak of morning peak, meaning there is potentially 2 slots free. 7:54 and 8:09 at Town Hall.

I don't see why you would have to use Hornsby as a Turn-back when you have two suitable locations at Lindfield and Gordon. They are used now, why not more?

There is basically now
- From Hornsby 4t/hr all stops, 2t/hr from Berwora
- From Hornsby 4t/hr limited express
- From Gordon 4t/hr
- 2t/hr Express from CC
- 4t/hr from ECRL

Causing a mix of express, limited express and all stoppers on the upper NSL converging to all stoppers from Chatswood

For my mind what I see could happen post NWRL

- 20t/hr post Chatswood to City, all stops, made up of following

- 4-6t/hr from Lindfield, all stops
- 4-6t/hr from Gordon, all stops
- 4-6t/hr from Hornsby, all stops
- 4t/hr from Hornsby, limited express to Gordon, probably a mix of CC and Berwora trains

A max of 20t/hr from Chatswood. Note some of the capacity used south of Chatswood is currently users from upper Nth main.

The Govts people movement study says that 50% or so of NWRL will actually continue past Chatswood.

Not sure how this will balance out with Nth main and Main West. I assume trains to/from nth main will start/finish Epping/Hornsby. For those on NWRL travelling to Central/TH, surely it will be very attractive to change at Epping for a ride that takes a few min longer and grab a seat.

The sooner the NWRL to the city s built the better.

I still maintain my prediction that this is how our system, will look ~2030: The North Shore and the suburban tracks on Sector 3 converted to "Rapid Transit".

Maybe, but I doubt it. Time will tell.

I think the plan has been scaled back to Bankstwon for now because its simpler. To go to Lidcombe and Liverpool they have to deal with the freights. Stop at Bankstown and that problem goes away, for now and reduces the cost of the 2nd phase of the project, same applies for Hurtsville.

Hurtsville is also on Eastern Suburbs line and perhaps the benefits of conversion at this time for next 10 years or so is limited, if any. I thought the idea of doing this enabled more slots for trains from south and Cronulla.

The main issue I believe is the tunnels. Getting the Bankstown trains out of the tunnels opens up more paths for others from west. But I'm sure there are other bottle necks to get them there.

I still maintain my prediction that this is how our system, will look ~2030: The North Shore and the suburban tracks on Sector 3 converted to "Rapid Transit".

And pigs might fly.

My understanding there is 18 trains per hour through the city at the peak of morning peak, meaning there is potentially 2 slots free. 7:54 and 8:09 at Town Hall.

I don't see why you would have to use Hornsby as a Turn-back when you have two suitable locations at Lindfield and Gordon. They are used now, why not more?

It's quite technical, but it's to do with the ratio of turnback separations to the timetable cycle.

To start with the window to turn a (CityRail) train at these simple turnbacks is quite small. The crew need to secure the train and change ends. It *can* be done in under 3 minutes, but you can't timetable such a quick turn because there are going to be variances. 5min is reasonable. I understand CityRail schedule a minimum of 6.

You also need a modest buffer on arrival, so late running on the Down does now prevent the turned train making it's path on the UP. Equally, you need a similar buffer on the other side of the turn so that the late arrival of the UP train's slot does not mean there is a train in the turnback blocking the arrival of a DOWN train. These two buffers can't be the same.

Also, unless you specifically engineer the timetable so that the UP and DOWN are synchronised at the turnback station in question, there needs to be some extra time allocated to allow the turned train to "switch phases" so to speak.

Realistically, a max of 4tph is possible at these simple turnbacks, but with at most 5min timetable buffer (ie not enough for the system to be reliable) that if exceeded will see the Turnback propagate delays. They can do it at Gordon because of the ECL: it doesn't matter if a DOWN train is delayed entering the turnback because the preceding train is heading to the ECL.

There is another problem though, and that is the phase shift issues between the UP and DOWN timetables. If you have them matching at Gordon say, then they will be completely out of phase at Lindfield because Lindfield is (about - it's 4min out of a 15min cycle) 1/4 of a 15min timetable cycle.

Chatswood on the other hand is close to exactly half a timetable cycle, so up and back is a full cycle. You could turn trains at Gordon and Chatswood comfortably without them interfering with each other - if there was a turnback at Chastwood.

If they do manage to get 24tph working then things become even more complicated, because the turnback windows are even shorter, but also the turnbacks are less favourably distributed, but it does become viable to turn 3 at each of Gordon and Lindfield.

There are quite a lot of possible solutions to these issues. Better turnbacks is one. Automation is another (for faster turns). Or "simpler" (read more expensive) stopping patterns with less turns and more trains running the full route.

For my simple cost comparison analysis I've assumed 2 of the 6 trains per hour run the full route to Chatswood to Hornsby, and the other 4 "new" trains instantly teleport from the DOWN to UP. But however they solve these problems, the fact is CityRail will have to run more trains to haul fewer passengers post the NWRL, and in my view advocates of the isolated NWRL aren't properly accounting for this cost.

Not sure how this will balance out with Nth main and Main West.

Basically they lose 4 slots.

For those on NWRL travelling to Central/TH, surely it will be very attractive to change at Epping for a ride that takes a few min longer and grab a seat.

It could potentially be faster from Epping via Strathfield to the city, depending on the operating pattern adopted for express and all stations services.  Empty trains starting from Epping will be a big incentive to change even it does take slightly longer.  Intercity express trains with only 2 stops at Eastwood and Strathfield will also be attractive.  Regrettably, this will only add to the congestion and overcrowding on the Lower Northern Line which doesn't have much extra capacity.

Basically they lose 4 slots.

ECRL was in part built so they could balance the trains travelling across the bridge to those entering the bridge and I thought reduce the wastage of terminating at Nth Sydney. Increases services on the Nth Shore to compensate was deemed feasible as they didn't have the numbers to justify, even when trains used to terminate at Chatswood.

So 4 trains per hour from Hornsby to Epping then via the NSL will again return to the North Main. Meanwhile a substantial number of cattle will be delivered to Chatswood thus requiring and justifying increased number of services across the bridge. As Chatswood is no longer a suitable terminating station, Lindfield makes the obvious choice.

However, 4 trains per hour back down via Straithfield into the city tunnels are competing with trains from the west and north. Clearly an issue.  

The current operation of the city tunnels is
- Main West and Nth Main and Bankstown enter tunnels and exit to NSL
- Inner West and South and Bankstown enter tunnels and exit to Airport, Bak and Glenfield

Bankstown occupies 8 trains per hour of the tunnels in peak, with Bankstown converted Automated and connected via a new tunnel from Sydnam to a new crossing in the city and under harbour. You not only get those 4 slots back, you gain another 4 slots for use by others.

What to do in the 10 years or so between closure of the ECRL and Bankstown tunnels being made available? Probably run a few trains (if capacity allows) to/from Central Terminal to/from Epping.

regards
Shane

ECRL was in part built so they could balance the trains travelling across the bridge to those entering the bridge and I thought reduce the wastage of terminating at Nth Sydney. Increases services on the Nth Shore to compensate was deemed feasible as they didn't have the numbers to justify, even when trains used to terminate at Chatswood.

So 4 trains per hour from Hornsby to Epping then via the NSL will again return to the North Main. Meanwhile a substantial number of cattle will be delivered to Chatswood thus requiring and justifying increased number of services across the bridge. As Chatswood is no longer a suitable terminating station, Lindfield makes the obvious choice.

However, 4 trains per hour back down via Straithfield into the city tunnels are competing with trains from the west and north. Clearly an issue.

The current operation of the city tunnels is
- Main West and Nth Main and Bankstown enter tunnels and exit to NSL
- Inner West and South and Bankstown enter tunnels and exit to Airport, Bak and Glenfield

Bankstown occupies 8 trains per hour of the tunnels in peak, with Bankstown converted Automated and connected via a new tunnel from Sydnam to a new crossing in the city and under harbour. You not only get those 4 slots back, you gain another 4 slots for use by others.

What to do in the 10 years or so between closure of the ECRL and Bankstown tunnels being made available? Probably run a few trains (if capacity allows) to/from Central Terminal to/from Epping.

regards
Shane

I agree with most of what you say and I assume that you inadvertently included the existing Bankstown Line as exiting to the NSL which it clearly doesn't.  It is split between the Inner and Outer City Circle.  

With the removal of the Bankstown Line services from the City Circle, that provides an additional 4 paths each way.  While the additional paths to the CC Inner could be readily taken up with services from the East Hills Line and/or Illawarra Local, the paths to the CC Outer would presumably be allocated to services from the Western and Northern Suburban tracks (T1) via the flying junctions.  This is not ideal and totally compromises the sectorisation concept.  It's just another example of how the Sydney Trains network will be buggered up to fit in the Rapid Transit concept without any thought of how it impacts on the overall network efficiency.  An additional 4 tph from the Western/Northern Lines falls way short of what is required for existing services, let alone what will be needed for future expansion.

You are correct in your last paragraph, but whether this is politically acceptable is another matter.  Even after the additional capacity through the CC becomes available, it still constrains future capacity enhancement from the West.

Has the operator of this rail link been announced formally?

http://www.smh.com.au/nsw/melbourne-train-operator-metro-trains-melbourne-to-run-sydneys-northwest-rail-link-20140624-zsjyr.html

http://www.smh.com.au/nsw/melbourne-train-operator-metro-trains-melbourne-to-run-sydneys-northwest-rail-link-20140624-zsjyr.html

Thanks. Found it on here too.

The search engine here works well.

https://www.railpage.com.au/news/article-14591

A few points from me. First of all, I'm not particularly fond of crew-less trains, prefer at least one staff member on board every train, maybe two in case of high capacity trains. I do however strongly support platform screen doors, preferably arbitrary position ones.
Consider the Scarborough RT in Toronto. It first opened in 1985 and to this day is still a single line not connected to any other, and like the NWRL is fully-automated. It has remained much the same since it was opened, and now there are planes to replace it by light rail.
In fact it was originally planned as an extension of the existing tramway network in Toronto.

I find the NWRL somewhat reminiscent of the Scarborough RT.

I agree with most of what you say and I assume that you inadvertently included the existing Bankstown Line as exiting to the NSL which it clearly doesn't.  It is split between the Inner and Outer City Circle.  

With the removal of the Bankstown Line services from the City Circle, that provides an additional 4 paths each way.  While the additional paths to the CC Inner could be readily taken up with services from the East Hills Line and/or Illawarra Local, the paths to the CC Outer would presumably be allocated to services from the Western and Northern Suburban tracks (T1) via the flying junctions.  This is not ideal and totally compromises the sectorisation concept.  It's just another example of how the Sydney Trains network will be buggered up to fit in the Rapid Transit concept without any thought of how it impacts on the overall network efficiency.  An additional 4 tph from the Western/Northern Lines falls way short of what is required for existing services, let alone what will be needed for future expansion.

You are correct in your last paragraph, but whether this is politically acceptable is another matter.  Even after the additional capacity through the CC becomes available, it still constrains future capacity enhancement from the West.

There are 8 trains per hour in the busy AM peak from Bankstown into city. 4 from each branch.

These can be offered to other Western or SW parts of the Network, in some cases additional work is required to make this happen efficiently. But once complete those lines capacity will increased.

The sectors boundries may need to be redrawn. This is not a negative and they were never intended to last forever.

Local to Hurtsville would be replaced with more services from further south, again removing the inner short services enables more longer haul trains and increases branch capacity further away from the city.

The issue still to be resolved is what to do with the outer branches of the Bankstown. Two options
1) extend Metro to Liverpool, or at least junction with SW line. Then start Homebush from Regents Park.

2) Or extend Metro to suitable junction station on the outer branch, lets say Regents Park. Trains from Liverpool now run through the Y to Lidcombe and into the city in same pattern as trains via Granville. Inner western trains that currently choke those two tracks from Homebush are removed to a new UG Metro connecting with NWRL in city. This also provides option for doing the inner NW Metro to Top Ryde.

Then start Homebush from Regents Park.

Isn't there a lack of tracks/land/road crossings between Hombush and Lidcolme?

Wasn't that the whole idea of terminating the Inner West line at Homebush?

Isn't there a lack of tracks/land/road crossings between Hombush and Lidcolme?

Wasn't that the whole idea of terminating the Inner West line at Homebush?

Yes two extra lines needs to be built and the area immediately west of Homebush is the worst bit for getting extra space. After then its not 100% straightforward but can be done as its mostly railway land and industrial.

You need 6 tracks to Lidcombe to enable the Western Main outer suburban, interurban and SW lines to all run more functionally. To Straigthfield you also need an inner west Metro which would be a tunnel. You could go cheap and stop earlier but the system is more functional at Straithfield.

If you have the Metro take over the entire Bankstown line, it still doesn't work as the SW and inner west and outer West compete for track capacity inefficiently between the Gladsville and Homebush.

Regardless what you do, Metro or not these extra two tracks are needed sooner or later as is inner west capacity.

1:20 grades

Would you have a web link to confirm the NWRL will have 1:20 grades?
It is absolutely pathetic and short sighted planning to build the tunnels too small for existing trains, and too steep as well.
I am sure the extra cost of keeping the gradient to 1:30 (adjusted) would be minimal. It is simply terrible planning.