The level of support amongst the general public, coupled with the recent reopening of the Onehunga Branch Line, has really put the idea of rail to Auckland Airport back on the agenda in a big way. Mayoral frontrunner Len Brown, as well as Mike Lee (who seems likely to end up with one of the two council seats on the board of the Transport CCO), have thrown their full support behind completing the Airport Line as soon as possible. In a recent post I had a look at what we know about the Airport Line from past studies – in term of its cost, alignment and potential patronage. In this post I want to have a brief look through what our options might be if we wanted to fast-track construction of this line.
The main thing holding back the speedy construction of this line (aside from the money and the lack of a designation) are the capacity constraints at Britomart. The map below shows the number of trains from each of the lines that will be arriving at Britomart during the peak hour from early next year – once Western Line trains are bumped up from four per hour to six per hour.
20 trains per hour into Britomart is pretty much its capacity until we get the CBD Rail Tunnel built – which would allow trains to enter the CBD from two points: Mt Eden and Quay Park – thereby at least doubling the capacity of the system. So if we were to try and build the Airport Line without first constructing the CBD rail tunnel, we would effectively have two options: extend the two Onehunga peak hour services or do something further south linking up the airport with the Southern Line via Puhinui. We could do both, as per the preferred option of ARTA’s 2008 report, although with constrained CBD capacity that might not be the best idea in the world.
The first option – what I call the “Onehunga option” is shown below: 
This option extends the Onehunga Line another 9km southwards – over the Mangere Bridge and then largely next to SH20 and SH20A to the Airport. There could be up to four stations on the way (as per the suggestion in ARTA’s 2008 report,) although I personally struggle to see why both Rimu Road and Walmsley Road stations would be necessary. The cost of this option, assuming that we ran trains at a low enough frequency to avoid having the double-track the Onehunga Line – and grade separate many of its level crossings, was estimated to be around $707 million. I’m not sure whether this includes the work required within the airport designation, although it seems likely the airport might be called upon to fund that section of the track – plus its station.
This option would have the benefits of providing a direct route between the Airport and the CBD, and would also provide for good links between the isthmus and the various Mangere stations that would end up being constructed. Its disadvantages are that without upgrading the Onehunga Line you could end up with some capacity problems through that section, while also the access from this track to Manukau City would be poor – and you wouldn’t create an alternative to the Southern Line between Manukau and Penrose, potentially useful if something goes wrong.
The second option is a simple link from the Airport to Puhinui and the Manukau Station. 
According to ARTA’s report, this section would cost around $470 million to build – which seems surprisingly high to me considering it would largely be running across undeveloped fields, or along the currently under construction Manukau Branch. The advantages of this option are its relatively low cost, its access to Manukau City and the fact that it would probably be the easiest option to build quickly. The disadvantages are that it wouldn’t improve public transport to the southwest corner of Auckland around Mangere, it is a very long and indirect way between the Airport and Auckland’s CBD and you wouldn’t create the handy loop that a full link between Manukau and Onehunga would create.
In short, this would most probably be the cheapest and fastest option. Unfortunately, it would also generate the lowest benefit and be of little use to anyone except air travellers (and probably not even that attractive to them because of its overly long route.)
The third option is the full loop between Manukau and Onehunga – what ARTA’s report recommended as the preferred route. This is shown below:
Without spending the $271 million to double-track and upgrade the Onehunga-Penrose section of track, this option was estimated by ARTA to cost around $1.178 billion. It would offer significantly greater benefits than either of the earlier options – although its potential popularity may place great strain on bottlenecks on the network: such as Onehunga-Penrose (if that was not upgraded) and certainly on Britomart. I also wonder whether only being able to run two trains an hour at peak times would generate enough patronage to justify the $1.2 billion spent on its construction. But it’s certainly possible.
There is a reasonably good argument out there that public transport proponents should start learning from how motorway builders are so successful in getting their projects prioritised. By that I mean perhaps instead of shying away from building a project that will generate a bottleneck, we should actually embrace that situation – because it will make that next project needed so much more desperately. After all, this seems to be how most of our motorway projects end up being justified.
However, personally my feelings are that it would not be the best idea to complete the Airport Line before completing the CBD Rail Tunnel. For a start, planning on the CBD Tunnel is more developed – and I would hate for that process to go on hold while we wait for the Airport Line to “catch up”. Secondly, I truly do believe that the CBD tunnel is a more necessary project for Auckland in the next 10 years – and is a project that would likely have a better business case. Thirdly, I think that to get the best out of what would be at minimum a $700 million investment in Airport Rail (I don’t think the Puhinui option would bring enough benefits to really warrant serious consideration) we will need to run peak hour trains at frequencies much higher than two trains per hour. That means we will need a CBD rail tunnel first – although if we didn’t, chances are we would probably end up with one pretty quickly after the Airport Line opened!
Processing...
With the Onehunga branch as is four trains an hour should be easy to achieve, they just need to bring the Te Papapa crossing loop into play. If they can handle two each way an hour without the loop, four each way an hour with a loop would be easy.
That would give 15 minute frequencies on an airport line which I think would be acceptable for a ‘rapid transit’ service (although ten minute frequencies would be better).
So it could just be a case of squeezing two more trains an hour into Britomart, or at worst taking two trains an hour from another line. For example if the southern line were sharing tracks with the airport line on the isthmus and the eastern line further out, then it might be reasonably to run the southern line only four times and hour. Or perhaps each of the four lines could run at five trains an hour. That doesn’t give a nice clock face time table but with one train every twelve minutes that wouldn’t be much of a problem.
I do have a couple of issues with the suggestion that an airport line cannot be built unless the CBD tunnel is first.
Firstly it obviously means if the tunnel project doesn’t get funded then the line is off the cards too.
Secondly, it is sort of an excuse to do nothing in the meantime. So we wait ten years for the tunnel to be funded, another ten for it to be built and opened, then some time in the 2030s we start talking about an airport line… and maybe in the 2040s it is funded and built.
Thirdly, perhaps it is more important to increase rail access and capacity in suburban corridors first. The CBD tunnel does have manifold benefits but arguably the greater need is getting rapid transit to more areas that the three rail lines and the busway currently cover.
Ideally I’d like them to be progressed together, such that the airport rail line can open a year or two before the CBD tunnel to really prime it for success, then once the tunnel is opened the second tranche of EMUs start to arrive to alleviate capacity constraints (one can dream). That would require up to four billion over ten years, although four hundred million a year invested into a real rapid transit system would be perfectly affordable if they could sort out the political issues around transport funding. It is after all only one third of what Banksy has pledged for his motorway programme!
I don’t see why trains can’t be run at higher frequencies and simply run every second one out West, that way other parts of town have a direct airport connection and we avoid overloading Britomart. I do feel however, that both projects need to keep getting pushed forward and if the airport line gets built first so be it, but to let it become an either or, one after the other type thing means we’ll be waiting 20-30 years before they’re completed.
Well it’s clear we need them both, as Nick says the absence of the CBD loop can’t be allowed to be used as an excuse to delay the Southwestern. Direct Western Line -Southwestern services would be popular, and not only for airport users. We’ve got to push for all of it, and as you’ve noted building capacity that highlights constraints that need sorting is good tactics…. sometimes Admin, perhaps, you play too nice; as we know the real bottleneck is the one Joyce’s head is sitting on……
One fear is that he could deal with the increasing clamour for rail in AK [in an election year] by agreeing to say the CBD Loop but on terms that penalise Aucklanders financially [for voting for the wrong council] and a schedule that pushes it way out into the never-never land- while furiously paving everything he can….
Remember this is the ‘can do’ Minister, so if he ain’t doin’ it, it’s cos he don’t want it.
How about Western Line trains via Newmarket to the Airport
Eastern Line Trains terminating at Manukau
Southern Line Trains unchanged.
Not ideal for Western Line travellers (especially if they needed to transfer to the Eastern Line) but it would remove the bottleneck and allow decent Airport line frequencies.
It’s $700,000,000 via Mangere without any upgrade to Onehunga, right..?
Seems that constructing that in conjunction with the CBD tunnel is the best idea…
It will be much cheaper and easier to construct the line on a 7 year time frame than on a three year one. Starting early is an advantage.
It seems pretty wasteful electrifying the Onehunga line when it is likely it will need upgrading in next few years. (particularly as it will need upgrades for EMU’s to fit anyway)
James, taking the Western Line trains out of Britomart would be a major no-no. You don’t want to massively disadvantage EXISTING passengers to support a new line.
The idea of doubling the airport frequency by running every airport train on a common route to Newmarket, but then continuining every second train on the Western Line (or if necessary, stopping it at Newmarket!) has a lot of sense though.
Apart from getting us 15-minute airport frequencies we also get (if we continue out west) a direct “cross-isthmus” train from out west to south. It would be feasible and useful (because not everyone needs to continue to the CBD) AND it would immediately increase the pressure to build the CBD loop, so making it a bit more sensible in case it is built first.
I thought about that and realised that it would complicate the timetable if you only had half. They’d be no ‘turn up and ride’, you’d be back to looking at the timetable to determine whether the next train goes via Newmarket or the Airport. I’d like to modify this slightly. It would also increase the efficiency of the Western Line trains because their would be no need for the reversal from Newmarket. Peak hour there would still be capacity for two trains to bypass Newmarket and go straight to Britomart. Eventually we have to get used to transfering, may as well get people into the habit of it as soon as possible.
Thinking about you wouldn’t want to transfer with big bags so maybe make it Western Line to Southern Line and Airport line straight through to Britomart.
its absolutely critical that any major transport initiatives are considered as part of the new auckland spatial plan. there still seems to be separate discussions between a spatial plan and a transport plan, and that transport problems can be solved with transport solutions. yet most of our transport problems are caused by where all our origins / destinations are located. if we want more a more compact, poly-centric city city, then these sort of transport initiatives are essential and the cost becomes relatively superfluous. if you don’t do it, you won’t be able to shape the city, it will become more spread out, and transport problems will become worse, requiring even more expensive solutions.
a city the size of auckland must have a more comprehensive rapid transit system, whether that be rail, tram or bus rapid transit. the question is not IF, its WHEN. its so expensive now, simply because we’ve put it off for 50 years. it will only get more expensive. we cannot keep putting it off.
The Airport handles about ten million passengers per year.
1. If a tenth of those used the airport link, that works out to three thousand passengers/day in both directions.
2. This, logically, would be 1,500 passengers per day in one direction.
3. This is an average of a hundred passengers per working hour per direction (15-hour working day). In practice, I would expect about 150 passengers per hour in the peaks, less in the rest of the day (international loads are much more balanced across the day tham domestic loads).
4. So, for a fifteen-minute frequency, we are talking about twenty-five passengers per service across the day, up to forty in the peaks.
5. Seen in that light, managing this extra load would not be difficult. The main issue, is that you would have to kit the trains out to allow for the luggage (which can take up an awful lot of room).
6. The airport bus probably manages about 5 percent of the passenger flows, ditto in Wellington. Airport bus services manage better in the UK, but not overmuch so. Edinburgh Airport gets well over 25 percent of its passengers arriving or leaving by the bus service (8-10 minute frequency through the day) – 2.5m pax per year – but that is exceptional.
Have I forgotten anything?
Ross Clark – you forgot two things: Mangere Bridge and Mangere Township residents who would ALSO use the rail service. If a station is also somewhere near the industrial area north of the airport then possibly even more passengers for the line!
No, the point I was trying to make was that the passenger load specific to the airport (Mangere Bridge is a separate issue), would in fact be a fairly steady load through the day – and a fairly light one by the standards of the demand elsewhere in the system. Ergo, the other parts of the system should have the higher priority. At /this/ stage, I would still suggest the cheap solution of a free bus link through to Papatoetoe railway station, and then see if there is any takeup.
You have however forgotten the 22,000 people that work at the airport precinct, which I believe makes it second only to the CBD as the largest employment zone in greater Auckland. To put that in perspective twice as many people work at the airport than in Newmarket, or four times as many than work in either Albany or Takapuna. To put that another way all of Waitakere city has only twice as many jobs as just the airport alone!
The bulk of those will be peak hour commuters although there will be a significant amount of shift workers also.
Yeah, but those 22,000 workers have plenty of free parking, and I know from having worked in passenger rail that free parking will trump provision of rail service for the commute, nearly every time. Simply put, it will be faster under most circumstances, even with a 15-minute frequency.
Also, the airport campus is about a mile from end to end and a kilometre from north to south, as the bottom of page 2 of this map shows:
http://www.maxx.co.nz/assets/timetable%20south/S11_Airport%20to%20Manukau_July%202010%20web.pdf
A bus can serve this, and does, but rail would be another matter. The jobs are simply too spread out, from what I can see, for a single station to make much of a difference.
In the third map, why is the airport station a terminating one? Wouldn’t it make more sense to loop through the airport so that trains can run from the city (and Mangere) through to Manukau while avoiding a Newmarket reversing situation?
That was what was shown in the report from 2008 but remember we haven’t actually sorted out any designations yet so don’t get to worried. In one sence you would probably want a long dwell time at the airport anyway to make it easy for people with luggage to get on and off so might not be a major issue.
Even if it is a terminating station, it only takes a couple of minutes for the driver to walk to the other end of the train to take it back out again. Airport stops will have to be a lot longer than stops at other stations to allow for the extra (un)loading time associated with international travellers and all their luggage, which gives plenty of time for a train to come in forwards and go out backwards, as it were.
The ARTA report has a ‘pinched’ terminus coming in from the east, the airport masterplan shows the line looping around the end of the new runway to the far west and terminating on that side. I guess the only conclusion is that there isn’t any firm plan yet.
Personally I think the line should come in from the north (perhaps under the new runway) rise to the first floor level, have a two-way station in between the two terminal wings (i.e. above what is currently the taxi/shuttle queueing area), then follow the curve of the terminal building to head east to Manukau, with a second station in the vicinity of Geoffrey Roberts Rd to serve all the people working on the eastern part of the precinct.
A terminal station will mean single direction trains, such as excursion and long distance services will not be able to service this stop.
Well DMUs like the silver ferns would be ok, and loco trains could with some messy shunting around.. but generally I agree it should be a through station.
I wonder what the change of having four platforms at the airport station is, so that intercity trains could stop at the station for a couple of minutes while suburban whizz in and out. Snowballs chance I guess….
Problem is, like the just hopeless enough siting of the Manukau City station, someone high up will cut the funding just enough to make whatever does get done is sufficiently compromised to be really effective… Like the talk here of keeping the Onehunga line single track, don’t even think about it, methinks there’s bit of Stockholm syndrome amongst us PT advocates sometimes…
One of our roles here should definitely be devising the ideal solution, then lobbying for it. There will be plenty of pressure for compromise soon enough from outside.
I am one who thinks we should take a leaf out of the motorway builders book. They have been extremely successful in getting more and more projects planned and built, even ones that were never on any plans and there is no sign of it even slowing down.
For the airport line I think the major problem is the debate has only been framed as needing to have rail to the airport and not at looking to improve connectivity in the south-west part of the city and that makes it easy for anti rail people to argue against and use examples like Sydney and Melbourne. The only reason we even have the Puhunui link rearing its head is because the debate has been framed this way and that the airport is the only consideration and reason for building it.
Personally I think it should be built from Onehunga but perhaps leave out the Puhunui link for the time being for it to be completed at a later date as a “network completion” job. Additional stations at Montgomerie Rd and Mangere Town centre are a no brainer but I think only one extra station not the two suggested in the report. Going for stations at both Rimu Rd and Walmsley Rd would slow the trip down quite a bit making it a bit like the inner part of the Western line. I also wonder if there is a mental block for many people caused by the Harbour crossing and that once we get that part built it will be much easier for people to see a line working along the corridor.
As for the CBD tunnel vs Airport one, I do think we could make it work without the tunnel and it would definitely increase the pressure for it. I would prefer them both at a similar time though.
I think that four new stations on the Onehunga to airport section is warranted, given that it is just under ten kilometres long. The inner west has stations spaced at 1,150m on average, where as the airport section would be every 1,960m. Thats pushing double the stop spacing as it is.
Only one station between Onehunga and Mangere Town Centre would mean gaps between the stations of up to 3km. I think a Rimu Rd station is needed to service all the people living in Mangere Bridge plus all the workplaces along Mahunga Dr, plus a station at Walmsely is needed to service the huge workplaces in Favona. Both of these are well placed for bus connections too.
While the airport is a big trip generator, it can’t support a line alone. The line needs to be designed as a suburban rapid transit line that stops at the airport, not an airport shuttle that might stop at some suburbs on the way.
Nick, it is about 2.5k from the Airport to Montgomerie Rd and then about 2k to Bader Dr, that is half of the journey to Onehunga already meaning the next three sections would only be 1.5k apart. Trains will spend much of their time accelerating and braking. Instead having only one station between Bader Dr and Onehunga would just stretch out those distances possibly saving 3-5 minutes of journey time.
With electric trains, the acceleration thing isn’t such a big problem. Also, the aim should be to encourage as many people as possible to use the train, which means maximising the population of the 800m pedestrian catchment radius around stations. Two stations in that space would greatly increase the number of people living inside the walking radius of a station, which has direct bearing on the non-airport viability and utility of the line.
1.5km is already a lot wider than the norm in Auckland, add that together with the long sections closer to the airport and the line will be much faster than anything else in Auckland anyway.
This line needs to work for the people and jobs of western south Auckland, if you are really concerned about skipping stations to save time then it would be a lot more logical to skip the three stations between Penrose and Newmarket that are already served by another line. There is no point running a line through Mangare bridge but then depriving the 10,000 residents of that suburb the chance to use it.
You wouldn’t save anything like 5 minutes by cutting one station. EMUs can acellerate and brake very rapidly, a Perth EMU is rated for accelleration at .87ms2 and braking at 1.05ms2 in service. That means they go from standstill to 100km/h in 30 seconds and brake to standstill again in 25 seconds. Adding in dwell time this would take about 90 seconds, or around 60 seconds longer than running through.
This is very similar to the line I catch each day. The express skips three stations and saves 2.5 minutes according to the timetable, or less than one minute saved per station.
Why not make the trains going to the airport longer so they can split up in Newmarket to go to Britomart and the other half west? Or join trains together at Penrose or Manukau so the number of movements into Britomart can remain at the maximum with more trains able to use the rest of the network?
Splitting trains sounds like a bit of a nightmare, particularly when many of the passengers would be tourists. You’d not only have to manage the coupling and uncoupling but you’d need to make sure people get in the right part of the train from where they are going. Sounds like a logistical lightmare
But by shifting to an all six-car fleet at peak times with twenty trains an hour, the system could handle about twice as many passengers as currently anyway.
So rather than moving to six trains an hour per line (using three car trains), pehaps it would be better to focus on four or five trains an hour per line with six car trains to handle the demand. That way there is room for a new line to open and everything can still go to Britomart. The only trade off is frequencies aren’t quite as quick as they should be, but it would still be the same number of carriages along each line per hour.
Nick R, when you are standing at the station waiting (especially when you are trying to get a “no timetables needed” system), you want shorter trains coming RIGHT ABOUT NOW, rather than longer trains shifting the same amount of people in the same overall time, but arriving later / less frequently 😉
So I don’t think that is useful. We should go to 10 minute frequencies – I’d say that may be more important (in the short run anyway) than airport rail. And WILL happen much quicker, because we get it from next year, and there’s no way we could build airport rail in less than 3-5 years anyway – and then once we do that, well, we don’t want to degrade existing frequencies for a new service! So we will have to find another way – any way, really, just not one which removes existing service quality. Maybe widen the Britomart approach tunnel to a third track.
I agree. The need for no timetable transit in Auckland is desperate. It’s not rapid transit unless you can turn up and ride. That’s why the CBD tunnel is so crucial as it effectively removes most road blocks towards achieving this. All future lines simply become plug and play.
I agree with you guys about no-timetable frequencies, but things like a third track into britomart just aren’t feasible and we have to work within the constraints. Do bear in mind that five trains an hour means a train every 12 minutes. Would that really be that much worse than one every ten minutes, especially if it meant the difference between having an airport line or not?
From a practical point of view no. However from a psychological one, breaking that ten minute barrier on all lines would be huge. Personally even 10 minute frequencies is on the upper edge of the turn up and ride. I think we need to get people used to the idea of transfering between lines.
A third approach would cost around $200 million, which is money better spent on the CBD tunnel than on a short term fix. I honestly think that insufficient rolling stock will be the bigger impediment in an airport line in the end that finding places to stick the trains.
I’m pretty sure widening the existing tunnel is impossible as it is now built in on both sides along it’s length. I’ve suggested we build two more tracks in parallel under quay st, initially as a short term fix but then going on to be extended as the cbd tunnel.
Rolling stock might b a problem, but again if they can run three lines at ten minute headways, they’ll have enought to run four trains at 12 minute headways.
“I’ve suggested we build two more tracks in parallel under quay st, initially as a short term fix but then going on to be extended as the cbd tunnel.”
That is what I was refrring to. Yes, it might be costly, but I am warming up to it. It would allow us to continue improving rail (and the case for more rail, such as the CBD tunnel) in the near term without breaking the bank under a tight-fisted National government unwilling to spend much. Heck, Auckland alone could “easily” pay 200 mil itself.
This report from 2004
“Britomart West Rail Extension Feasibility Study (URS Group, final report for Auckland City Rapid Transit Group, 2004)”
is the one claiming it will cost 140-200 million, I assume when it was written they took into account the planned and ongoing developments around the tunnel seeing as the land had been rezoned by that point.
It’s not clear in that report how the propose to quadruplicate the approach for that price, I had always assumed they were talking about widening the existing tunnel but I think these days it would have to be a new tunnel in parallel under a road corridor.
Anyway if building the 3km long tunnel under the CBD with three new deep level stations is going to cost about a billion, I imagine a 900m long shallow cut and cover tunnel with one station could be had for about $200 million.