This is a Guest Post by regular commenter Nick R. If anyone else wishes to submit a guest post please email it to jarbury[at]yahoo[dot]com. It is written as the executive summary of a (as yet draft) proposal to ARTA/KiwiRail. Once the proposal has been finalised it will be linked to from here.
Analysis of the proposed Auckland CBD rail tunnel project, and a potential modification to increase capacity and operational performance
The Auckland rail system is currently approaching capacity due to the limit on the number of trains that can fit through the twin-track approach to Britomart station. A CBD tunnel project is being investigated to connect Britomart with the Western Line at Mt Eden via the central city. This will provide two more tracks accessing Britomart allowing an increase in system capacity as well as shortening the route of the western line while providing greater accessibility to the rail system via three new stations under the CBD. The project represents the most significant investment in the history of Auckland passenger rail and will form the centre piece of the network for generations to come.
However, it is identified that the proposal connection to Britomart has some unnecessary drawbacks despite the substantial capital investment, namely that the expansion in capacity will be less than double the current limited capacity, and that the central capacity would still be only two-thirds of the capacity of the existing suburban system. Furthermore conflict between through-running and terminating trains for slots in the existing Britomart approach tunnel would greatly limit the peak capacity of the Auckland rail network, as terminating trains must necessarily occupy two slots on existing approach tunnel as they enter and exit via the same route. Other drawbacks include the inter-mixing of operationally different ‘rapid transit’ and locomotive hauled trains at Britomart, the reduction in the potential for Britomart to function as a regional terminus, and the need to retrofit Britomart to support rapid transit style services on platforms 1 and 5 at great expense. There is also a strategic drawback of the proposal. While the system is approaching capacity now, no new capacity would be available until the entirety of the project was complete at a lead time of 7 to 10 years and a cost of approximately $1.5 billion dollars. This represents a significant period of unrealisable potential patronage growth on the network.
A modification to the CBD tunnel concept is proposed to address these drawbacks. It is proposed that the CBD tunnel be extended a further 900m eastward connect directly to Quay Park junction via a cut-and-cover tunnel under Quay Street. The existing Britomart station and approach tracks would be bypassed, while a new pair of tunnel platforms would be built adjacent to Britomart and connected to the existing station concourses via short underpasses.
The existing Britomart station and its approach tunnel would be untouched and continue to operate in their current capacity as terminal station for diesel and electric powered services, while all services using the CBD tunnel would use the new purpose-built ‘rapid transit’ platforms adjacent. Connecting the eastern end of the tunnel to Quay Park junction via a new pair of tracks (rather than the existing Britomart approach tracks) would provide triple the current system capacity (rather than double), and would therefore match the capacity at the core of the system to that at the periphery. This would allow the CBD tunnel to be utilised to its fullest capacity for electric suburban rapid transit services while allowing the existing station to accommodate various diesel or electric express, regional and intercity services.
The modified proposal also allows for capital expenditure and capacity increases to be introduced in stages. As a first stage the initial 900m cut-and-cover tunnel and new platforms adjacent to Britomart could be constructed and opened first, allowing a doubling of system capacity with a 3-4 year lead time and a cost of approximately $250 million dollars. This would allow rail patronage to continue to grow while the second stage extension through the CBD to Mt Eden was being constructed.
The modification represents the addition of 900m of new tunnel and an fourth pair of underground platforms to the existing proposal at an estimated increase in total project cost of 15-20%. However, the modification would provide 100% more capacity over the existing proposal plus considerable benefits in terms of in operational flexibility and project staging. Therefore the relatively minor additional expenditure of the modified proposal may result in greatly improved outcomes in cost-benefit analysis, particularly if long term growth over the lifetime of the infrastructure is taken into account.
Green & Yellow Lines – existing tracks into Britomart
Blue Line – track from Newmarket & Orakei heading into CBD rail tunnel
Red Line – track from CBD rail tunnel heading to Newmarket & Orakei
Black lines – indications of pedestrian access points to new station & station outline
To be honest I don’t think it is a great idea. If it were that easy to increase patronage then I think this would kill the tunnel as that is one of the main reasons for building it. I think it could be designed and built in a way that midtown comes into use while the rest of the tunnel is still being built, this would also help with patronage. I would also want us to look at the option of changing Britomart so that the tunnel platforms are in the middle of the station with the terminating platforms on the outsides.
That’s is the one potential downside of my proposal, it would be all too easy to build stage one and then forget about the rest of it once the capacity crisis was averted. I guess the idea would be to lock in the whole project through contracts and laws like they did with Britomart, and have stage one as only the first operational phase of the ongoing project.
Regardless of what is done with the position of the tunnel platforms at Britomart or bringing midtown into use early, the key problem still remains. All trains in Auckland will still have to pass through the two track throat tunnel to the east of Britomart at some point, whether they are running through or terminating. Three tracks in from the suburbs and three tracks out, yet only two through the middle. Sure it’s better than the one through the middle that we effectively have now, but it’s not good for the future. Auckland is planning to spend a billion and a half bucks to shoot itself in the foot capacity-wise.
Consider it like this, say in the nearish future there are four main lines in Auckland (the Western, Southern, Eastern and an Onehunga-Airport line) and we want to run a train on each of these every ten minutes during the peaks. That gives us 24 trains an hour we need to run. We can pair these up so that an incoming Western line train passes through the tunnel and becomes an outgoing Southern line train and vice versa (or whatever the best routing is) and likewise with the other two. This means to get a ten minute service on our four main lines we need 12 trains an hour passing through the CBD tunnel and Britomart in one direction and another 12 going the opposite direction. That’s all cool because each track should be able to handle 18-20 trains an hour.
But the problem comes when we want to add in other trains. With 12 trains an hour each way for the main lines this leaves 6 or 8 slots each way for other things. That sounds fine until you realise that any train that terminates at Britomart uses two slots, one to come in and one to head out again. All diesel powered trains will have to do this, so we only have room left for 6 to 8 terminating trains overall.
When you consider the possibilities the spare slots start to disappear real quick. There are two diesel trains an hour proposed for Pukekohe beyond the wires, plus another two an hour out west past the wires to Waitakere/Huapai. That’s four gone already.
And then there is the Overlander, the proposed Waikato connection, and in the future possibly regular trains to Tauranga and Rotorua. Together these might account for three or four trains an hour that need to get in and out of Britomart at the peak. So we won’t be able to do them all.
Plus so far we haven’t considered the possibility that the Manukau branch would have it’s own services, and this is before we even start thinking about a line out to Botany or one along the Avondale-Southdown route or any other new route. Furthermore going to anything better than a train every ten minutes on the main lines would be simply out of the question, it would be physically impossible even without any other trains running at all.
A key thing to remember is that what we build now will still be the backbone of the system eighty or even a hundred years in the future. We need to allow for a bit more capacity that merely double what the meagre capacity is now. All the problems like sitting at Quay Park waiting for a slot to get in to Briomart will still happen, that end of the line will be exactly the same. I also think it is essential that Auckland maintains some form of terminal station and builds the CBD tunnel specifically for rapid transit services, rather than mixing the two together at Britomart to achieve poorly on both counts.
It is a great idea, construction would be a nightmare but luckily they don’t include congestion created during construction in BCRs…
Ahhh – nice idea Nick, but by far the most effective way of increasing the capacity of Britomart in the short term would be to tunnel the start of the CBD Loop on the already designated route onward from Britomart. Incoming trains in morning peak hour could discharge passengers and then move West, freeing up a slot for another incoming train. Both trains(even three if the tunnel is dug far enough) could be joined and leave together. Obviously it would be vice versa in the afternoon.
Nick – For those services that are beyond the wires we should only run shuttles to the nearest station that has electrics running to it(Papakura/Swanson), that means that the extra tunnel slots are only taken up by inter regional trains.
As for Manukau, I think these should become the eastern line trains. That way all trains from south of Manukau go through Newmarket.
If we are going to do something about Quay park then it should be something to remove those nasty curves that would also speed up services through the area, to me this should go up the strand on a bridge and join in back at Parnell rise.
Lastly if we are going to spend any extra money it should be on building the junction for the North Shore line so that doesn’t need to interfere with the tunnel when it is also built.
I’m pretty sure the plan once Manukau opens is for all eastern line train to terminate there, and for all trains south of Manukau to go via Newmarket.
I’m not necessarily a huge fan of that idea, as I think a lot of the demand for Southern Line services comes from south of Manukau and I don’t know whether a train every 10 minutes will be enough for that, although it would certainly simplify the timetables, which is probably a good thing.
Rationale: the short term problem with the CBD tunnel is that you don’t get any real capacity increase until it makes it all the way to the Western Line at Mt Eden (seven to ten years after we start building the thing). Sure building the first section of the tunnel from Britomart would help a little operationally as you could effectively store empty trains up there to manage peak flows better. But it still doesn’t overcome the fact that they will all have to come in or out via the same two track tunnel as now.
Matt, yes the shuttles is one way to do it (expresses might be better in my opinion tho), Manukau will probably become the terminus for the eastern line etc. So yeah the CBD tunnel would be great to begin with and we should be able to fit all our current trains in just fine, but what about ten years down the track after ten years of sustained patronage growth? What about in twenty? One thing to consider is that this investment has a functional life of maybe 80 or more years, we need to think a little further along the line than what the trains would be like if we had this tomorrow. A case in point is Britomart itself. Back less than ten years ago when they were designing the thing it was being called excessively expensive, a huge hole in the ground, way bigger station than Auckland would ever need. But now it is still a new station by any standards, but it is basically full already. I don’t want to do muck with the CBD tunnel concept too much, but this fairly small extension would give twice as much new capacity for well less than twice as much cost. Its about getting a really efficient outcome from limited funds.
I do have a concept for how the new tunnel would connect to the existing line at Quay Park, the new links for the tunnel services are grade separated with lesser curves and grades than the existing approach while the existing link would remain for the terminating services. I had a look at a connection to the tracks near the Parnell overbridge, it does get rid of the curve around Vector Arena but you need a significantly longer tunnel due to the height changes involved and more complicated trackworks to connect the eastern line to it also. I wrote that off for the first step because it would add too much cost (my goal is a low cost, big return modification to the proposal), but it is a possibility that would remain open for future expansions (like maybe the North Shore Line).
I agree any connection from the North Shore needs to be grade separated at the least, and ideally avoid the CBD tunnel tracks altogether by accessing Britomart or midtown on it’s own tunnel. You can see the value of having this separate, but this project would be a couple of decades down the track. So what about the current tunnel project, why not have that separate too?
It seems like it would be cheaper and just as effective to widen the existing tunnel into Britomart to 4 tracks as a first stage.
Ok, just noticed Josh has posted my concept for Quay Park above. So that is stage one, the yellow and green lines are the existing tracks to Britomart that stay the same as they are now. The blue and red lines are the new tunnel links that go via a cheaply constructed cut-and-cover tunnel to a new set of platforms under Quay St. The black lines at Britomart are some suggestions about how to connect the new platforms to the existing station.
At Quay Park the black lines indicate the position of the ramps to and from the new tunnel. They each are 200m long giving a grade of 1 in 40 or better, and the curves are less tight than the existing one. It all fits within the existing rail corridor so no land purchases are required and no building foundations are in the way. The lighter coloured lines are in tunnel.
Stage two is to continue the blue and red lines along the proposed CBD tunnel alignment.
Perhaps the key thing is you can see two pairs of lines coming in from the suburbs (one from Newmarket and one from the east) and also two pairs of lines through to the main station. After stage two you get two more lines out to the western line, so that the capacity through the CBD is identical to that through the suburbs.
Matthew: that was the idea I started with, but looking into it I’m almost certain you can’t widen the existing tunnel due to it being built out on both sides by now right along its length. So the next best thing is to put the extra tracks under the closest road corridor where you can be free of building foundations. Either Quay St or Customs St, and to me Quay St seems the easiest of the two.
Another thing to consider if you get four tracks into Britomart it would probably be a bit small if you are trying to terminate services there, i.e. four tracks leading to just five platforms. There is some scope to widen Britomart to six or maybe seven platforms, but that would be a hell of a job to do on an existing underground station that would need to remain in use during construction. My conclusion was that the cheapest and easiest thing would be to just leave the existing station and approach tunnel alone and build the new tunnel and platforms alongside.
Hi Nick,
Great idea, I can see how the existing plan might not provide the capacity. I would have to say that there is another alternative. Simply turn Britomart into a pure metro station and build a new station for terminus trains somewhere else. Near say the old railway station on Beach Street or maybe closer to Parnell. People arriving at the terminus could then transfer to the metro system for the final part of their journey. This would also have the benefit of providing Parnell with a station not to mention providing an alternative station for Vector Arena and the tennis centre. It is also close to the northen end of the university and the port motorway. I would be interested to hear your thoughts on this.
Cheers,
Nick, my question is whether this Quay Street tunnel could be “retrofitted”. By that I mean the following timetable of events:
2020ish – currently proposed CBD rail tunnel opens
2030ish – currently proposed CBD rail tunnel starts hitting capacity through that Quay Park junction area
2035ish – realign CBD rail tunnel to new alignment (as per your suggestion), with the bonus being that as a backup the tracks could link with Britomart as well.
Possible?
Nick’s full idea: http://transportblog.co.nz/wp-content/uploads/2010/03/CBD-tunnel-concept.pdf
One thing I will say is I think your costing is too low, not only is this a cut and cover but given it’s proximity to the harbour and the fact it is on reclaimed land you are basically building an embankment which is not cheap…
James, yes that would be an alternative option, make Britomart just a metro station and put the terminal somewhere else. Who knows if the old Auckland station is viable anymore but maybe the new Parnell station could have enough room for four or six platforms… but you would still have the same problems that prompted them to move the old Auckland station in the first place, the fact it is on the edge of town rather than where people want to go. This might not be such an issue with intercity trains, but for regional ones and expresses and the like the best place for them to terminate would be Britomart.
But doesn’t is seem backward, to turn our purpose designed five-platform terminus into a two-track metro station, then build a new terminus elsewhere? Why not just build a new two-track metro station and leave the terminus as it is!
Jarbury, it would be possible to build the current CBD tunnel first then the Quay Park section at a later date when it was needed… but they would need to design the tunnel alignment with that in mind so the grades and curves were all possible. The downside of this is you don’t get that first increase in capacity within a few years, only after the whole tunnel is open, plus you have the added expense of the tricky tunneling under the CPO building and lengthening the platforms and adding escalators etc at Britomart so it can handle tunnel EMUs. All this is basically wasted cash if you go and move them to some new platforms five or ten years later. The extra tunnel links could give some useful operational flexibility, particularly in emergencies and planned closures… but there wouldn’t be any extra capacity to be had Generally keeping through running ‘metro’ trains and terminating loco/diesel services separate is going to have the best performance.
Jeremy, the existing Britomart 500m cut-and-cover tunnel was built for $30 million nine years ago. What we are looking at here is a 900m tunnel in the same soil/water conditions and a new station box built by cut and cover also.
The beauty of this section is that it can be all built by shallow cut-and-cover, the rest of the CBD tunnel must be bored with stations probably constructed by mining outward from the bored shaft (this is how the built Melbourne deep stations at least). Generally speaking cut and cover tunnels cost about a quarter of the cost of bored tunnels.
Plus there is plenty of space on the waterfront to manage the construction. Just over the red fence from Quay St is a strip of largely unused wharf land almost twice as wide as Quay St itself. So the road lanes could be diverted onto this space while the street is used for the construction site. Another thing is they main CBD tunnel will need a shaft somewhere in the region to start the TBM from anyway (QEII square and lower Queen St have been suggested), so thing could be the ideal spot to avoid disruption.
$250 million is just a guesstimate on my part based on a brief analysis of a similar project in the 2003 URS report, but I think it is the right ballpark. It’s basically a third more tunnel (1km or so) and a third more stations (one more station), but much easier to build than the rest (due to cut-and-cover being ideal) and there are savings by not spending any money connecting to or modifying Britomart.
Yes it probably is backwards, but I guess that’s what happens when you build a new station with no thought to the future needs of the network. I would point out that London’s main stations are all on the periphery of the city and they seem to do fine (except when you have to transit through from one station to another, but we don’t have to worry about that yet). I guess my concern is that, with a transport minister who is seemingly hostile to rail, Auckland has to do all that is practical to cut the cost and make this business case as strong as possible. If your proposal worked out cheaper and allowed for a stronger business case then that would be a far better idea, but it may be a case of having to put up with a less than ideal outcome. I also think that any proposal should take into account the idea put forward of having a boulevard along Quay Street and potentially undergrounding a section of road. Maybe we could do a two level cut and cover tunnel with the top level for road and the bottom level for rail. Although this might make the gradiant too steep up Albert Street.
Hmmm… I think my official position is going to be to promote the original alignment but to ensure it’s future proofed to link with a parallel line under Quay Street. That way we don’t risk the project as it’s currently proposed going ahead, but also we can get that extra capacity of the Quay Street tunnel/station when we need it.
James: London’s main stations are on the periphery of the city, but they are also linked by one of the world’s most extensive metro systems. I might also add that London is currently planning tunnels to link its main overground stations via the central city. A better comparison might be the Australian systems which are the same hybrid suburban/metro that Auckland has. In each case like Sydney Central, Melbourne Southern Cross, Brisbane Roma St and Perth Station, the main terminus is in the CBD proper.
I don’t think is would come in cheaper than the current proposal but it might only be marginally more expensive when all is taken into account. While it does involve an extra station and 900m more tunnel, the extra tunnelling is a lot cheaper to do than section under the city and you save money by not having to build the tricky connection to the existing platforms.
You have brought up two good points. If it is simply a case of making things as cheap as possible my modification will obviously lose out. But if it is a case of building a stronger business case then mine might be superior. Remember it should allow twice as much new capacity on the rail system (or three times what we have now) for well less than twice the cost of the existing proposal. Like I’ve said above it might only add 15% cost to gain 100% more capacity plus a variety of other benefits. That might do wonders for the cost-benefit analysis and the business case…. which might make it more attractive to the Minister for Transport who seems to think in such terms.
Admin: That’s about as much as one could ask for however I would just point out two things. It *is* the original alignment, just an extension to it. All the work that had been done so far is still perfectly valid, I’m saying we take it a little bit further from where the current alignment stops. I’m not sure just how much consideration has been paid to the Britomart end so far to be honest, they have been focussing on just getting a preferred route through the CBD. Secondly, I would argue we need the extra capacity of the CBD or Quay St tunnel now, not in ten years or so.
Nick, the flip-side to needing the additional capacity NOW is that we have more pressure to build it ASAP. If Joyce was a huge fan of this project we could have it completed by 2017 I reckon.
There is also plenty of capacity in the tunnel if we were to lengthen our existing trains, 6 or 8 car sets would almost double out existing capacity without taking extra slots in the tunnel and most stations are already big enough to handle the 6 car sets with some already big enough for 8.
Yes I agree that your proposal could have a better cost benefit analysis. Unfortunately the political thinking may be skewed towards the less cost/less benefit side of the scales. Convincing me is easy, convincing Steven Joyce may not be quite as easy. Given the minister’s statements that he doesn’t think the whole project is necessary and you have a situation where trying to get him to open the purse strings wide enough to produce anything other than the cheapest option may be impossible.
One other point that I think would need to be cleared up is getting Ports of Auckland to open up the land on the other side of the red fence for the temporary road that you propose. Given the Port’s attitude to opening up that land before schedule it might be difficult to get them to come to the party.
Matt L, one problem with lengthening the trains is that Britomart cannot handle anything longer than a six car set. This is why building new tunnel platforms would be handy, you could build them all to accommodate eight car sets. Otherwise the new CBD tunnel stations will probably get built to the six-car maximum that Britomart has. No point having an eight car long platform at midtown if Britomart can only hold six.
I’ve assumed the trains would be run in six or eight car sets. While having longer trains does boost passenger carrying capacity it does nothing for train capacity… when it comes to having several lines running at quick headways you need to be able to get a lot of trains through, regardless of how long they are. I’m assuming here that it is desirable to have as frequent headways as possible. Obviously in the long run having the longest trains possible running and frequently as possible would provided the greatest capacity, that is what I have gone for here. My modification would allow twenty eight-car sets a hour each way in the tunnel. It looks like the current proposal would be limited to twelve to sixteen six-car sets an hour (once you account for terminating trains).
James, the ARC owns the Ports of Auckland so I’m sure they can manage some temporary arrangements. Most of that will be ‘party central’ before construction would start anyway. The port has indicated it’s intentions to vacate the finger wharves and consolidate at the container terminal. In terms of the political will and funding levels I’d put my money on what always happens in Auckland. A good project gets cut back to a shadow of its former self in order to save ten or twenty million bucks… then a few years later it is revealed to be inadequate and poor quality.
This is facinating Nick. I certainly think that it needs close attention from the people responsible for finalizing the CBD rail project decide on the full scope of things.
Thanks Will, it is interesting to note that there has been no discussion or evaluation so far regarding the northern end of the tunnel… it has been assumed from the start that it would be connected to Britomart without and evaluation of any other options.
I guess it is easy to to get blinded by the mere possibility Auckland gets a rail tunnel at all, without thinking about whether the proposal is the most efficient or effective possibility.
Great stuff!