There’s an interesting section in the CBD Rail Tunnel business case that looks at a couple of different options for how trains should run around Auckland’s rail network once the CBD Tunnel is completed (pages 44-47 if you’re interested in taking a full look). The business case proposes the “base option”, which my understanding is what they want to do, but which requires a reasonable number of additional trains for it to work. It also proposes a “minimal option”, which only requires four additional trains.
The key constraint seems to be that our current SA Trains (the ones that are hauled by diesel locomotives and will be hauled by electric locomotives post-electrification) won’t be able to use the CBD Tunnel. One reason for this is that they have an inadequate fire-rating (which one assumes can be fixed), but another reason is likely to be that the tunnel is going to simply be too steep for such trains – which is something more of a permanent problem. I must say I’m a little worried about how steep the rail tunnel will have to be, and I think it would be a good idea to look into options that would lower the Western Line at Mt Eden to a greater extent, so that the tunnel’s gradient can be eased a bit.
But anyway, getting back to operating patterns, the diagram below shows the “base option” – which would require 24 extra 3-car EMUs to be operating on the network at the time of the tunnel opening:
I’m sure it’s not just me thinking this, but the above operating patterns do look very very complicated. Do we really need half of the trains from Papakura to go via the Southern Line and half via the Eastern Line, do we also really need to split the trains from Manukau the same way? Another thing I find odd is the proposal to run 12 trains per hour along the Western Line between Henderson and Britomart. While it’s definitely fantastic to consider running trains at 5 minute frequencies along the Western Line, how essential are such frequencies? How many extra trains are going to be needed to operate such high frequencies that probably won’t be needed at that stage? The final strange thing is the way the Onehunga Line will be operated. I guess half the trains will go around the rail link clockwise and half anti-clockwise, which once again sounds rather confusing.
The “Plan B” option, which only requires four additional EMUs (as opposed to 24), is obviously a bit crappier and would eliminate a direct link between Onehunga and the CBD:
It is interesting that going from this operating pattern to the one in the first image (and they’re not that different) requires 20 additional EMU trains – which certainly wouldn’t come cheap.
I’m not completely sure of the rolling stock requirements of my proposed operating pattern, which is outlined below – but I certainly know that it’s far simpler and easier to understand than either of the patterns shown above:
I think one of the greatest strengths of my system is its simplicity: we simply have two lines. One from Manukau to Onehunga via the Eastern line and the CBD Rail Link, and the other from Swanson to Papakura via the CBD Rail Link and the Southern Line. You also manage to avoid inefficient uses of the infrastructure – with 6 trains per hour on pretty much all parts of the system, and 12 trains per hour on a few very busy sections – like the CBD tunnel.
There’s still an interesting remaining question of whether we should have a direct link between the Western Line and Newmarket. The issue of “how to serve Pukekohe & Waitakere/Huapai” is not addressed either – as I think the current proposal is for rail shuttles to be run. In the end, I suppose the infrastructure will enable a mixture of service patterns to be run according to particular demand patterns – I just like the idea of creating a system that’s more like a “Metro” in the way it works.
Processing...
Yours is certainly easier to understand. For us, and also for users of the system. You don’t want people standing around at stations scratching their heads while they try and figure out how it all works, which is my experience of Sydney’s complicated mess. And, unless I’m being really stupid your scheme eliminates the need to reverse trains out of Newmarket and that’s got to be a good thing.
Thanks Obi, yes the idea is definitely to make things as simple as possible.
I think all the schemes eliminate the Newmarket reverse – which is very very good.
I guess 6tph running to Onehunga would require a full upgrade to that line, think the airport study costed it at $270 million. Guess that would include double tracking and major grade seperation to remove all the level crossings.
On one hand this would be a great thing as it would be stage 1 of the airport rail line. However I’m not sure that this extra cost can be squeezed in as soon as the CBD loop tunnel opens.
With Pukekohe could an extra couple of diesel hauled SA sets fit in Britomart at peak times? These would need to be limited stop services so they wouldnt hold up the faster electric trains.
Double tracking and upgrading the Onehunga Branch was actually included in the costing of the CBD rail link, for that very reason.
It looks like they have set up the first system to try to minimise transfers.
I think so too Swan, but is that what we want? If people need to switch from the southern to the eastern then they’ll only have to wait 5 minutes at Otahuhu (or a nicer transfer station). If people want to switch between the Western Line and some line to get to Newmarket then the wait would be similarly short (or we could run west-south services).
The beauty of simplicity is that it both makes the system easier to understand, but also easier to run. Your two lines operate fairly independently of one another, with just some care taken at the points where they share the tracks to ensure they’re both not on the same piece of track at the same time.
I also think that it would be possible to boost service capacity to 12 tph each way on both lines without too many problems.
Admin,
I agree. I would be surprised if there is another example as strange as the first system. All metro rail type systems I have seen tend to have a set of simple linear routes.
I don’t know how we should operate the trains but this is what we should do with the stations!
http://artbns.com/swedish-subway-system/
Wow. That is seriously the coolest thing I have seen.
Ever.
I remember riding on the Stockholm underground in 1975 when all those stations were painted in primary colour monochrome paint. It looked amazing even then without the art works. What are the chances of having a future Modern Art Museum in the Auckland stations? I bet the advertising agencies will claim as much wall and train space as possible first.
I figure you could set up temporary exhibits in stations as part of an Auckland art week and then provide maps of the circuit. It would help to advertise the service and provide something interesting to see while you wait for your commute.
I’d never thought of a twp line option before… Really good idea, never more than one transfer – amazing frequencies…
The idea that transfers are undesirable comes from a belief that the car is still the best model, ie all transport must be point to point. This leads to support for poor systems like the the bus routes we now have. And in part explains the thinking that can’t see the value in rail’s big advantages. And how these advantages become dramatically unlocked within a coordinated network. When this often unexamined idea is linked with the self interest of private bus companies and the ideological bias that claims that private is always better than public you can see why it has been so hard to move forward with even the most important rail plans.
As I said earlier, I think Josh’s scheme is the best of the three presented here and transfers aren’t a problem. Simplicity and understandability more than makes up for some people having to change trains.
However… transfer-free rail can have its uses. London’s Northern Line should logically be two lines, with Edgware trains always going around the western central branch, and High Barnet trains always going around the eastern central branch. But they don’t, they swap over 50 percent of the time. The reason is to avoid 50 percent of passengers having to change trains at Camden Town. The station platforms just couldn’t handle that quantity of people transferring.
None of which applies to Auckland. It’s just a bit of trivia.
The first system is obviously set up to minimise conflicting movements at Newmarket as well as minimising transfers. If we look closely they only have one train every 20 minutes that would pass in front of another at the junction, but they have the thing where trains take different paths depending on which is the next one to come along. I think this is a pretty poor way to plan services though, likewise with planning routes based on minimising rolling stock.
Admin, your favoured routing is the same one I came up with when trying to devise a ‘metro’ style routing pattern that was most simple, most direct and most easy for people to use (I like to claim I discovered this arrangement first many years ago, however as it is the most elegant solution I wouldn’t be surprised if many people have discovered it).
This routing fulfils the following criteria:
-All lines run ‘through’ (provides wider coverage of the whole region rather than just the CBD, and improves interchange opportunities).
-Nothing terminates in the city (keeps things fast and efficient, avoids platforms being occupied for minutes at a stretch).
-Nothing does a full loop in the city, rather all lines do three-quarters of a loop (this ensures good central area coverage, yet removes the absolutely useless fourth-quarter that simply returns to where the train just came from).
-All lines take the most direct route to the nearest CBD tunnel portal. As Newmarket is equidistant from either tunnel entrance, one line goes one way, the other the other (ensure the train goes straight to the largest proportion of direct travellers will be headed).
-Nothing reverses at Newmarket, but all lines still pass through Newmarket (facilitates good interchange while taking the pressure of the junction).
-The system is an efficient hybrid between a radial commuter system and a grid-like metro system.
-No station is more than one transfer away from any other station in Auckland.
-Transfers can be made at any of the four city tunnel stations and Newmarket in the central area, but also at most stations on the southern line.
-Extensions to the airport do not require any change in operating pattern or more paths in the central area, just enough new trains to run the extension.
-It provides paths for regional/intercity diesel trains to get to Quay Park junction from the North Auckland line and the NIMT. We can trace a route through the western line, Newmarket and the Parnell Branch that has only one metro line on it (and thus has up to six slots an hour for diesels). On the southen-eastern line we have the same, although this assumes the proposed third track is built between southdown and the Manukau branch where both metro lines share the track. Where to terminate these diesels is still an issue, The Strand station is the easiest option.
As far as I can see it is the best solution on all accounts.
Yes this does look the best. Hey Nick: we desperately need your CBD modes analysis to be published somewhere… the Herald would be best, perhaps we should pass the hat round and bring you over as an ‘overseas expert’? How many words is it?
You mean my post comparing the requirements to meet the predicted CBD growth using the various modes? That’s about 1800 words, but it could be condensed down to a few paragraphs to get the main points across.
I’ll actually be in Auckland for two weeks over Christmas, no need for a hat. I’m perfectly happy to sign of an article as “Melbourne based transport policy analyst” lol.
how are we supposed to operate them when we don’t have them. lets work on that
Effectively how they are operated now, once Manukau is opened and electrification is finished then there isn’t likely to be any changes unless the tunnel is built as we are limited by Britomart and Newmarket which are both pretty much at capacity.
I have thought of an alternative routing and would like feedback on it. It consists of three lines ( Western, Southern and Airport) and requires the building of the Parnell station. The Southern and Western lines would each go around the CBD loop, one clock wise and the other anti-clockwise then back to their originating terminus. The “Airport” line would go from Britomart to Onehunga via Glen Innes. This line would be extended in the south to the airport and onto Manukau and in the north to Wynyard and under the harbour to Esmond Rd.
Western line loop;- Swanson — Mt Eden – Newton — Britomart – Parnell – Grafton – Mt Eden — Swanson. Transfer points Newton, Britomart and Parnell
Southern line loop:- Papakura — Newmarket – Parnell – Britomart — Newton – Grafton – Newmarket — Papakura. Transfer points Newton, Britomart, Parnell, Penrose and Wiri.
Airport line:- Esmond —Wynyard- Britomart — Glen Innes — Penrose — Onehunga — Airport — Wiri – Manukau. ( Italics represent future development) In meantime Wiri – Manukau on a shuttle line. Transfer points Britomart, Penrose and Wiri.
The big advantage of this routing is that the three lines are physically separated with the following exceptions:-
The Southern and Western lines have three cross over points near Newmarket ( if Southern line is on inner side of CBD loop). In the future these could be grade separated.
Current the eastern entrance to Britomart would have to be shared with the Airport line. A short section to grade separate in future.
Airport line will need to share section of track between Southdown and Penrose initially. In future divert Onehunga line down existing rail corridor that runs through Neilson Street and Southdown freight yard and grade separate where it will cross the Southern line at Southdown and Wiri.
Other advantages are.
All destinations available with a single transfer.
Separate lines allow greater frequencies in future.
Each platform only has one train travelling to one destination thus less passenger confusion.
CBD loop has train travelling both ways around the loop At 6tph in each direction if Southern and Western lines are at 6tph.
Future development will have little adverse effect on Western and Southern lines.
Easy to open Wynyard station without altering routes.
John that is probably the most sensible other operating pattern that I’ve come across. I think I prefer mine because it offers the opportunity of creating one-seat through trips – not that many people would necessarily take the opportunity I admit (it would be faster to transfer at Newton) for a New Lynn-Penrose trip for example).
I also think sending Onehunga/Airport trains via the Eastern line is a bit odd as the link is quicker via the Southern (plus it serves Newmarket).
One other thing to consider is that under your scheme Grafton station would be pretty inaccessible for inbound trains as everything would only reach it after travelling right around the “loop”.
Question: Josh… How would a hypothetical future north shore line fit in to your scheme? To me, Aotea looks like it will be a much more popular station than Britomart and I don’t see any useful way to visit both from a north shore line. Would you skip Britomart completely? If so, and assuming the tunnel comes ashore somewhere around the tank farm, then is there a use for a new station served only by the new line somewhere around Air NZ?
Probably something like this:
As a bonus we get to reopen the old Strand station as our inter-city rail terminus, with connections to the blue and red lines. You would need to four-track the eastern line – but that’s inevitable at some point anyway.
Obi, every official concept for a North Shore line has a station at Tank Farm somewhere (latest was a Halsey St) it’s pretty much a given.
As fo the routing, I would say not joining the CBD tunnel would be essential, instead having a new alignment with additional
capacity and new dedicated platforms at or alongside one of the CBD stations.
I think there are two feasible options:
1) A route via Britomart, joining the eastern line at Quay Park or the Southern line at Parnell.
2) a route via Aotea, joining the Southern line further up the Parnell Branch or perhaps tunnel all the way to new underground platforms at Newmarket.
The first benefits from a much shorter, cheaper tunnel and it maintains a single interchange point between all lines at Britomart. The second however brings Shoreites further uptown and allows the option of stations near the university and/or hospital.
As for routing, either way I would send the line down the southern to terminate at Newmarket or maybe the airport.
^^
I mean *manukau* or the airport.
I had two options for how a North Shore Line would blend in with the Tunnel.
1. Trains would go through Britomart/A future Parnell Station/Newmarket/Mt Eden/K Road/Midtown(Aotea) then back to Wynyard Quarter & back to Albany.
2. A junction on the CBD side of the Wynyard Station. Trains wanting to go South/East would go to Britomart. While Trains wanting to go West would go to Midtown(Aotea).
Also Wellington has 48 EMU’s coming and Auckland will have 38. But wouldn’t further expansion of Auckland’s network require like, what 30 additional EMU’s?
Not sure about option 1 there, that is a very long one-way loop. It would take you a good 20+ minutes via six or seven stations to get to midtown, plus it requires a direction change at Newmarket. Option 2 sounds better, just run the trains on a reasonably direct through route and let people transfer to get around within the CBD.
Also Anon, Aucklands EMUs will be three-car and Wellingtons are two-car, plus Auckland is getting some electric locomotives to run some of the existing carriage trains. But yes, the CBD tunnel and any new lines would most likely require more trains to be ordered.