Auckland Transport’s decision to not build Newton Station as part of the CRL project, instead upgrading Mt Eden, potentially has some impacts on future train operating patterns that are worth analysing and discussing further. Our preference for operating patterns, prior to the change, is shown in the Congestion Free Network map for 2020 – the immediate post-CRL period:


There are numerous small variations you could make to which lines link with other lines – should the western join with the southern instead of the eastern? Should the Onehunga Line go via Parnell instead of Grafton? Should Mt Roskill to Panmure be the “short runner”?

However, there are some pretty basic concepts here – most importantly that this is a Metro style operating pattern with our assumption being all trains stop at all stations and that we’ve tried to keep the patterns as simple as possible. For example, all trains on the inner part of the western line go straight into the CRL tunnel with journeys to Newmarket either being provided for by staying on the red line right around the city centre, or transferring services at Newton.

After seeing the response to the CRL changes on Friday we gave AT a bit of a push to publish just what their thinking was around how they would run trains after the CRL was built. To their credit they published it that afternoon.

At first glance the plan looks fairly simply however as I’ve looked at it more it’s raised a number of questions for me. These include

  • Do we really need a west-south service?
  • Can the Onehunga line cope with that number of trains?
  • Do we really need so much service between Newmarket and Penrose?
  • Can Newmarket cope with that number of trains?
  • How confusing will it be with the Red line doubling back on itself?
  • Does Manukau have enough service?
  • Does the Western Line past Henderson have enough service?
  • Does the Eastern Line have enough service?

Do we really need a west-south service?

So of course, not having Newton any longer – and having Mt Eden station removed from the “Grafton to K Road” direct rail connection, means that west to Newmarket trips which AT say currently make up about one third of western line trips would need to be provided for in one of four ways.

  1. Staying on the train right through the CRL and Parnell to Newmarket
  2. Changing trains at Karangahape Road from an inbound western line train to an outbound train travelling from the city to Grafton and Newmarket
  3. Providing a direct rail service from the west to Newmarket
  4. Running a shuttle between Mt Eden and Newmarket

Auckland Transport has seemingly chosen option 3 of the above, with their preferred train operating pattern showing a direct service between the west and Newmarket – then onto Otahuhu. This is the purple line on the map. At first glance the purple service seems potentially quite useful – providing that one seat ride from the west to Grafton and Newmarket, as well as recognising “not everyone wants to go to the city centre”. The question is though, “will the level of demand for the purple line be high enough to justify reasonable frequencies?”. And if not, for example if it is only run at three trains per hour, then it’s not good enough to expect people to just turn up and go and therefore won’t it be faster to just transfer at K Road?

It’s also worth remembering that just because 1/3rd of trips on the line are currently to Grafton/Newmarket, it’s not a massive number overall and won’t grow as much as trips to the city centre. I’d also point out that it doesn’t serve all West-Newmarket trips, my wife for example is now travelling to Grafton but as we catch the train from Sturges Rd it would still mean a transfer for her.

Other considerations include:

  • How many additional trains are necessary to operate the purple line and is there value for money from purchasing those trains solely for this line?
  • Are additional turnaround facilities required at Henderson and Otahuhu for these services, and how much will they cost?
  • How much will it actually cost to operate the purple line and how might that compare to say changing the New North Road buses so they directly serve Newmarket?

I don’t know the answer to these questions but they are ones that AT really need to answer.

I think my preference is to serve the west to Newmarket trips through a combination of transfers at K Road (which isn’t that much further out of the way than Newton would have been) and perhaps a Mt Eden to Newmarket shuttle. It still means a transfer but without the time penalty of going all the way to K Rd first.

The next three questions are all interconnected.

Can the Onehunga line cope with that number of trains?
Do we really need so much service between Newmarket and Penrose?
Can Newmarket cope with that number of trains?

It’s not clear from the announcement whether the Onehunga line is being duplicated. Based on discussions I’ve had I’m going to assume that because it’s not core to the CRL so it probably isn’t being done. We currently only run two trains per hour on the line however I understand that with a little work it’s likely we could run three trains per hour on it. Unless AT are planning worse frequencies than now it would mean some of the blue western line trains would have to terminate somewhere and the most likely place for that is at Newmarket. With both the red and purple lines travelling through the station would there really be enough capacity for that to happen. This is exasperated by the fact that a terminating train would have a period of layover longer than what Western Line trains currently do for the driver end change. It seems that if you do terminate trains at Newmarket you get capacity problems at the station but if you send them on down the line to terminate elsewhere the line south of Newmarket may have their own capacity constraints and you end up with the situation of it seeming like stations between Penrose and Newmarket would be serviced by many more trains than the stations on the CRL. With the exception of Ellerslie most of those stations don’t do that well for patronage despite the level of service they have.

How confusing will it be with the Red line doubling back on itself?

I realise we already kind of have this issue so perhaps it’s nothing to worry about but it just seems to me that having a line double back on it’s like this one does risks creating confusion, especially for those trying the system for the first time. Going by the current way we label services – which lists the final destination – it means a person at a station between Westfield and Puhinui wanting to go to the city would actually have to catch a train destined for Manukau or Papakura which seems a bit counter-intuitive.

On a slightly related note AT talking of closing Westfield and Te Mahia. I assume they’re on the map because the final decision hasn’t been made rather than because they’re necessarily staying.

Again another similar set of questions.

Does Manukau have enough service?
Does the Western Line past Henderson have enough service?
Does the Eastern Line have enough service?

While some lines have double ups that significantly boost capacity it seems that Manukau, the Eastern Line and the Outer Western Line miss out. Does this plan deliver enough capacity for those stations? Because we are now sticking with the same number of trains we will have by the end of next year/early 2016 that suggests they might not get more than the six trains per hour they will have post electrification. Bringing it back to the west-south service what if we provided those trips by one of the other methods suggested. Could we use them to boost frequencies on the lines mentioned that at this stage don’t seem to get any frequency benefit from the CRL. This situation would change when we order an additional batch of new trains – now scheduled for around 2025.

All up the plan does answer a number of questions but also creates a lot of questions.

Share this


  1. What would help the analysis of the west-south idea is HOP data.

    We know Western liners alight at Grafton and Newmarket, in the case of the former many are likely ‘downhillers’ whose ultimate destiny will be better served by a CRL station in the future, not all of course. And in the case of Newmarket it would be good to know how many are terminating there and how many are transferring to Southern line services.

    1. Speaking directly on the Grafton train station that I use daily (it appears to me that about a ¼ of passengers exit off the peak flow western train. This is made up of – Med School, Hospital and School students along with some Business people. This is only going to increase with the movement of Fine Arts and Engineering out to new Newmarket Campus. There is no way a shuttle could cope with this. This is the correct decision from AT.

  2. One possible solution would be to have the Henderson short runner go to Onehunga. That line can have 2-3tph while running in parallel with a 6tph Swanson-CRL-Manukau service via the Eastern Line and a 6tph Papakura-CRL-Papakura via the Southern Line in both directions (still looping back on to itself but should be easier to understand).

    1. The issue of the southern line legibility is simply a labelling one. All northbound trains get called City via East, or City via Newmaket and southbound either Eastern Line, Southern Line, or MC via GI, and Papakura via NM…. hardly a problem that requires rerouting. Doubling back on the same route reduces the number of direct destinations and is to be avoided.

      1. In London there is at least one service which has a loop at the end and the train carries on round back to the start 2 hours later. It just happens to “change” its destination at a key point so avoids confusion but still runs round the loop.

        Very silly question but I ask what are the main reasons for not looping services back to the start? I confess I’ve never seen any comment on it – thanks.

        1. A couple of reasons, been discussed back when looping patterns were proposed. –Geometrically speaking a circle is the least direct way to get between two points, a straight line is the fastest.
          -Looping back to the line you just came from is actually useless for the passenger, they’ve literally just ridden through those exact stations, however running through to another line opens up a whole bunch of extra potential destination stops.
          -Looping means either you run all trains on a line around one way, or do funny things like alternating directions. In the first instance that means you are guaranteed to ‘go the long way around’ on each return trip, in the second you effectively halve the frequency to or from any given station, and have confusing situation of the same train departing on alternating sides of the platform. The scheme above takes people both in and out by the most direct path.
          -Auckland’s ‘loop’ is far too big to work as a loop anyway, running the wrong way around to the last station would take like 15 minutes.
          -Looping is inefficient for track utilization it creates extra movements over junctions you don’t need. In Auckland’s case both Newmarket and Mt Eden are off the loop, so you’d either have to skip them or have trains double back and change ends twice per loop.

          1. Southern and Onehunga line trains could realistically complete a loop, and would only miss Mt Eden (not really a city station anyway). It would be too difficult however for any train from the east or the west as they would need to do a reversal at Newmarket, like what happens on the western line now. That would not be a desirable outcome.

          2. “Looping back to the line you just came from is actually useless for the passenger, they’ve literally just ridden through those exact stations”

            Why would any passenger do that? Looping is just an easier method of the current practice of having the same train go back to where it came from. It’s already done, right now. I don’t see anyone claiming it’s silly to have Western Line trains go back to the Western Line because it means passengers will go back to where they came from. It’s a nonsense reasoning.

            AT appear to be planning for looping post-CRL anyway – Papakura trains are going to go back to Papakura during the peak.

          3. Precisely Geoff, no passenger would ever do that so why plan a service pattern that way? It is silly to send western line trains back to the west, but operationally we are stuck with that as the best option for now. If we already had a through tunnel rather than a dead end terminus then wouldn’t send western line trains back west.

            You are incorrect on the last point by the way, Papakura trains will go to manukau before bouncing back.

          4. “Precisely Geoff, no passenger would ever do that so why plan a service pattern that way?”

            To avoid doing what AT were at one point planning, of having 12tph to Henderson for the sake of it because there are three routes south and one west. Looping some back makes sense, and this is what AT are planning to do, by sending Papakura trains back south, to Manukau, and some even back to Papakura during peak hours. It makes sense, as the alternative would be to just keep everything going west for no good reason.

            I agree through-routing is best if the number of through passengers is greater than the number of passengers being missed by not looping. But since the stations being missed are CBD-area ones I’m not convinced. I like to think of the physical loop created by the CRL as being a spoke of CBD stations (Britomart, Aotea, Mt Eden, Grafton, Newmarket and Parnell). Directly delivering people from all routes to as many of these stations quickly and conveniently is important.

            “You are incorrect on the last point by the way, Papakura trains will go to manukau before bouncing back”

            The AT operating map appears to shows peak services going back to Papakura. But Manukau or Papakura, it’s still south-city-south, which makes sense.

          5. “The AT operating map appears to shows peak services going back to Papakura. But Manukau or Papakura, it’s still south-city-south, which makes sense.”
            No Geoff it’s not. What they are trying to indicate (and not that well) is that at peak times there will be a unique running pattern that travels from Papakura via the waterfront. It runs through the CRL and then terminates at Grafton before heading to the stabling yard at the Strand.

          6. Wait, you want a loop to avoid something AT previously considered but did away with?

            What is wrong with the plan above, excluding the overlaid purple line there is one line west and three south, just as you point out. There is no ‘everything going west’.

            I agree the east-city-south-city looks good, so why split that up into two routes with big one way loops through town?

          7. No, I explained why I’m not convinced that having everything run though may not be better than a mix of through or loops. Loops guarantee you the train you hop on will get you to the CBD spoke station you want, without having to worry about transfers or getting on the wrong train. This is the benefit of looping. The benefit of through-running is for cross town travellers, so the potential benefits of each method should be weighed up.

          8. Ok, but loops also guarantee that you have to go the long way round on every return trip. For example west to K Rd is direct in and out as above, with a loop it would be direct in but incredibly indirect out. That would add like 15 minutes to the trip for K Rd to the western line. Same goes for Britomart on the eastern line. Who wants to go from Britomart to Orakei via the CBD, Grafton and Parnell?

            Also given the tracks, the western and eastern loops would need to skip Newmarket, the junction could not handle all that doubling back, whereas above they both run through Newmarket. Likewise with Onehunga, looping would require it to skip Mt Eden. So it’s not actually the case that a loop accesses all central stations, in some cases like the eastern line you’d be doing things link dropping Newmarket (and points south) only to pick up Parnell at the very end of a long slow circuit of town. Anyone from Parnell to the east would transfer at Britomart anyway.

        2. They do it in Melbourne too and it’s a complete mess, furthermore they are about to unravel their tricky loops, turn them into through routes. As I say above doubling back reduces the number and variety of direct destinations on any route, and the interaction with other services- possible connections.

          Transfers are useful, and people will happily make them if the frequencies are high enough, or they are perfectly timed, and if the process is clear and convenient. But you still want a system where the routes with the highest demand are as direct as possible, basically you want to avoid mass transfers where it is efficient to do so. So given that very, very, few people wish to sit on a train all through the city and end up where they started without getting off why would you make that your direct one seat ride? Many more people are interested in getting to the city, and another number again through to the other side.

          After all the return journey is already being met with the matching service in the other direction. By doubling a service back you aren’t adding to coverage at all. But you are adding to complication and confusing.

          Schedule design inevitably attracts puzzle solvers who can get lost in tricky solutions.

          Here’s Jarrett:

          Keeping it simple seems to be the hardest thing to get.

          1. Actually, loops seem rather simple to most Melburnians. You get on a train, it gets to the loop. If you want to go to other places, you get on another train that goes to those places.

            I get that there are disadvantages, but Jarrett has exaggerated the difficulty to passengers to make those disadvantages seem more compelling.

          2. Yes the Austro-Hungarians knew how to build cities.

            The best PT system I have ever used day to day is Prague. The Prague metro (most used per capita in the world) has no loops and works really well:

            As you can see all the lines just go straight through the centre and it leads to a great network. It does mean transfers are needed but the frequency means that is not a problem.

          3. Now that is ideal. Exactly what you’d design when starting from scratch and not adapting an inherited half-built combined freight-passenger system on the surface with tricky topography.

            But still, with the CRL, and especially once the other extensions and optimisations are done Auckland is going to have a really efficient and even elegant system that will make for a great core to the whole city’s movement networks. And, dare I say it, one that will be the envy of other cities, and will come to redefine the city as future focused and dynamic.

            We just need to get on with it, it is THE opportunity for step change.

          4. That three lines, three interchanges, six termini arrangement is very close to the geometrically ideal arrangement. No station is more than one transfer away, every trip can be made without doubling back or going tangentially. Unfortunately in Auckland there is bend the lines a bit through the CRL, but the scheme above isn’t too far way from a two line version of that.

            I’ll be using that system myself next week, I look forward to it!

          5. George, I think you’ll find Jarrett’s views on loops are more nuanced than you’ve presented them here.

            The disadvantages of loops are not only that they provide an indirect way between any two points, but also that they are difficult to operate reliably. You can see this on the inner link and the outer link, where passengers are forced to sit on the bus at timing points just so the buses keep to schedule. Both these disadvantages increase with loop size.

            The advantage of loops is that they provide a one seat journey between any two destinations on the loop, albeit somewhat indirect. Which incidentally is why I’ve seen Jarrett personally advocating for retaining the inner link ;).

          6. To be fair George isn’t talking about two way complete loops like the Inner Link, but rather rail lines with one way balloon loops at the city end like most Melbourne trains. Those issues still stand with regard to the balloon section, except that a ballon loop is necessarily one way only so they *don’t* provide a one seat ride between all points because you can’t travel in one direction.

            Case in point is Melbourne where in the afternoon peak you can get a train from Southern Cross to Parliament station, but you can’t get one from Parliament to Southern Cross! (With Southern Cross being the terminal for regional trains, this is especially infuriating).

    2. The one issue with that solution is that Onehunga would lose its direct to city service, and all who wish to go to the city would be forced to transfer.

      1. Agree.

        Also losing direct Panmure to West is not ideal. Given Panmure’s emerging importance as a hub, and the fact that many people joining the train there from further East have already just transferred, using the train for cross-town journeys just got less attractive.

  3. What do those grey arrows represent? The one going over the harbour bridge is obviously the northern busway, but the one going northwest from henderson?

  4. I was under the impression the duplication of the Onehunga line was included in the CRL cost? I could swear I read it in there somewhere. So if it was double tracked, it could cope with all those blue line trains terminating there. I would still rather the blue line go via the Eastern line to Manukau or via the Southern line to Papakura, but I realise it’s tricky to get that balance right.
    I think the ultimate solution is that one of the lines may have to complete a ‘loop’ around the city and terminate back on itself (maybe alternating in which direction it loops), similar to what happens in Melbourne.
    Realistically there are only 3 main lines in Auckland, and 1 branch line. This is where the difficulty lays in through routing. It would be a whole other story and much simpler if there was 4 main lines.

      1. Yes, I don’t think loops are ideal either. But I think if in Auckland it was only one line that looped (probably the Southern line via Newmarket), then that would be manageable.
        I don’t think they’re getting rid of all loops in Melbourne, only some of them.

    1. From what I understand they are trying to decouple those other improvements from the cost of the CRL itself, a dedicated Onehunga line isn’t necessary for the CRL to work amazingly, so if the cost isn’t already out It surely will be soon. Like those extra trains, would be good, but still huge benefits without them so they can be deferred until really needed.

      As a former Melburnian I can tell you the City Loop is shit and nothing to emulate, nothing worse than going the wrong way around the loop for fifteen minutes to go one station over… And Auckland’s ‘loop’ would be some 50% longer again. Alternating loops would be even more disastrous and confusing than Melbourne. FYI PTV have a new plan to remove most of the loop services from Melbourne and to run them through like a metro instead. Failed experiment with the 70s commuter rail model.

      The plan above is far better, it has the four lines joined up into two main patterns which looks like it will work great. Don’t see the point in looping two lines around rather than linking them up.

      1. Are you referring to the AT Plan or the Congestion Free Network plan? Last I checked the Mt Roskill spur has never been officially announced. I definitely prefer the CFN plan to AT’s.

          1. This better not be official – having to wait until 2041 for North Shore rail is scandalous! If it’s true then a pretty serious shake-up is needed at Auckland Transport. Not to mention some rates refunds.

  5. Personally I think the best and most balanced option (given no extra trains) would be for the Western Line to go to Manukau via the Eastern Line. The short runner from Henderson (Although I would extend this to Swanson) to go to Onehunga and the Southern line does a loop of around the city going back on to itself. Half to go via Grafton and half via Parnell. Biggest negative with this one is no Onehunga-City direct service, and all those going Onehunga to city would need to transfer at either Penrose or Newmarket, but sometimes sacrifices like these need to be made.

  6. After trying to figure out a better operating pattern, I can only come to the conclusion that the CRL design change creates more operational problems than it solves. That purple line, especially if we don’t buy any more trains yet, sucks capacity off all the other lines.

    Assuming Onehunga still gets duplicated and grade-separated as part of this project, I would just drop the purple line in favour of reallocating those trains to higher frequencies on the other lines, creating easier transfers network-wide.

    1. Exactly. And in a better world the O-Line would be doubled or at least Te Papapa Station would be two-tracked and in a trench through the May Rd intersection. And the Mt R Branch would be underway now too. However these, like electrification to Papatoetoe, and other network improvments are separate projects and not necessary to the main benefits of the CRL and can come later.

      And they are a separate budget issue. You will never see the work at St Lukes or Lincoln Rd intersection included in the cost of the Waterview project, but they do help it function better too.

    2. Andrew and Patrick, how much would putting in the Mt Eden East station solve this?

      Wouldn’t it remove the need for the Purple line for Western to Southern services and thus increase capacity elsehwere in the network, and so it it avoids the need to buy more trains sooner, will save the thick end of $100m eventually for what will be a simple addition to the mix?

      1. Yes it would, but also slow those southern travellers that AT wants to bank a 2 minute gain for. But also just sitting on the train to K Rd and there’s a real easy cross platform, always undercover switch right there…. Which looks pretty good unless there are zillions doing it. And in the morning they’ll be switching to a counter peak service. That’s the cheapest option. Furthermore then all those purple line trains and slots can be used to increase frequency and capacity on the main routes, making everyone’s life better, including, of course, those transferring.

  7. Can’t they just do a bit of operational analysis on these networks and find an optimised solution? Seems to me like there’s a lot of opinion pieces there, and very little data in support of this or that change. Model it, justify it! and then start doing the same with the highway projects. If you find you can’t justify them, you will have saved us all a lot of unnecessary expense, and people will be thankful.

    1. You show a touching faith in models Jacques. Because the CRL changes so much current data is not a great deal of help to future patterns, current patterns are shaped by current services. And models that use current patterns as inputs simply reflect back current service patterns. Furthermore we have serious doubts about the models used in Auckland for predicting Transit demand as they have consistently under predicted it all this century.

      But still, as I say above, how many are now travelling west to south beyond Newmarket would be good to know.

      1. It isn’t modelling, it’s analytics, give me every trip origin and destination, and I can write code that will optimise for travel time for all users. Could even give constraints such as ‘no loops’ etc.

  8. I think that the purple line is totally superfluous at the moment.
    Customers on the Eastern line also don’t have a direct link to Newmarket in the current scheme – they must either transfer at Britomart or stay on the train and go the slow way via all the city stations.
    Why should the Western Line be any different? Just because that trip can be made now without a transfer is no reason to keep it that way. It serves a minority of passengers who can make an easy cross-platform transfer at K-Road.
    Keep it a simple, metro style network. Convenience comes from high frequencies so transfers are seamless.

    This begs the question: does Mt Eden Station need a platform for Newmarket-bound trains? My opinion would be “Yes” because it already exists and allows operational flexibility for future running patterns, service disruptions and event trains. Just don’t spend too much money on upgrading this platofrm until it has a regular timetabled service.

    1. Yes. Also during construction the western line will need to function as well as possible through here, and of course there still are freight services from Northland to consider (till the Nats cut that branch next). Also for system resilience and special services (big events at Eden Park for example) west/ connection and a Mt Eden station is handy and being at grade is not expensive to upgrade and maintain.

      There are a lot of schools around Newmarket and Grafton, and down the southern line that currently use this service. With the southern line heading Newmarket-Grafton-KRd, those students will exchange direct services on the western for direct services to the southern and CRL stations, and one transfer for the west. So? Still an improvement.

      Incidentally a big winner from the CRL is Auckland Girls Grammar, a fantastic school with a wide catchment but with some difficult connections. Currently you see a lot of pupils from there at Britomart, they will now have all sorts of much faster and direct connectivity to the K Rd station.

    2. The fact trains go south on the Western line is irrelevant for anyone west of Henderson – they need to make a change from blue to purple. The “outline train plan” needs further thought.

      1. Which ever pattern is used there will be some transferring required for some journeys. This is why the key issue is frequency. As on any pattern with high frequency transfers become a trivial factor.

        Remember the CRL is designed to handle 2.5 minute frequencies each way, which means we can build to having 3 lines with 8tph serving that high demand core of the system facilitating simple painless transferring without over supplying the outer areas.

  9. There is a lot of possibility here.

    That is a very good thing. We can make of it what we want and need, as busways and rapid/frequent transit services are expanded and developed, and living patterns develop and change. This is the half of the reason to build the Rail Link – to create a flexible and open system.

  10. I think the western line joining the eastern line is important as it seems to be the only through routed route that could compete or beat cars.

    1. See above, there will be some transferring somewhere. We favour a direct west/east service, but if that doesn’t fit the overall pattern a transfer at Britomart or one another of the CRL stations is unlikely to be burdensome or involve hardly any kind of delay.

    1. Well it’s basically unchanged, except with Mt Eden replacing Newton, for west south change at K Rd, or to buses at Mt Eden [or possibly a Mt Eden-Newmarket shuttle?], so is:

      South to Panmure via Grafton [we agree that this is best route for the main line]
      Swanson to MC via Eastern line
      Onehunga to Mt Roskill via Parnell

      So for disambiguation here’s the full map, it’s important to keep in mind what we’re building towards. Note the third line adds frequency where it’s much more likely to be needed; at its core. So the only tripled lines are in the CRL, the heart of the network. Before Mt Roskill is built run the O-Line to Henderson, and if there’s a need, cutting or adding services at Britomart/Quay Park as necessary. We believe there is space to stop those southern services south of the Panmure station and crew exchange etc, otherwise they can continue to Otahuhu, depending on the traffic through that key interchange…. and so on, but basically there’s a clear, high frequency, model to work on. And apologies to those who are affected for the red/green combo.

      1. This layout of track (CFN 2030) even without Newton station looks really good / much better than the AT plan.

        What about adding a Southeast to West line via Grafton and Mt Eden?
        Penrose / Ellerslie to Sunnyvale / Sturges Rd stations?
        In the Southeast the line could use the existing freight spur at Southdown Reserve or an extra line the length of the longest required train could be laid here.
        There doesn’t need to be any kind of station it is just a place to keep the train off the main South line.
        First / last station on the line could be Penrose or Ellerslie.
        In the West the line could run a third line the length of the longest required train to the North of either Sunnyvale station or past Sturges Rd station to the west of Waitemata Rugby Park and north of the ponds.
        Probably the best route would be Ellerslie – Grafton – Mt Eden – Henderson.
        Depending on demand it could extend and begin / finish at Penrose – Sturges Rd.
        This service could be used only at weekday peak times, outside of 6-9am, 4-7pm M-F the ‘shuttle’ / link is not used so as to avoid confusion on the Western line.
        If there is no demand for this direct cross town service, easy to discontinue its use with little cost having been spent.
        Also as it would never go into the CRL it could still be run potentially with diesels if the number of new electric trains were stretched across the rest of the network.

      2. What about creating a rail triangle to the East of Southdown Reserve to run trains Northeast – Northwest?

        You could then run an East – West line, Panmure to New Lynn / Henderson via Grafton, Mt Eden.

        From Panmure the new Eastern interchange hub:
        4 stops to Britomart, Red / Green line.
        6 stops to Newmarket, new East – West line.
        7 stops to Grafton each way, Red or East / West line.

      3. Another East – West option I can think of:

        Parnell to New Lynn / Henderson via Grafton, Mt Eden.
        Service starts / ends at Parnell but the trains could turn around at Quay St, a possible future extension / interchange with the Red / Green line?

        This would mean the inner core stations would all now be served by 2 or more lines.

        1. Why bother? What’s the point of building extra track and infrastructure and buying extra trains to run all and sundry rambling routes between far flung suburbs. Run the most simple pattern possible at high frequency and people can connect across anywhere to get anywhere on the network.

          1. Well the why bother is because the axing of the Newton station has been identified in this post as a ‘perceived’ problem for East / South – West connectivity.
            I’m not saying they’re optimal just possible.

            I thought I saw somewhere in the blog a call for different options to address this… that’s what I did.

            Pretty minimal track required for the 3 options, in fact no extra track required for the 3rd option.

            “This layout of track (CFN 2030) even without Newton station looks really good / much better than the AT plan.”
            This was my very 1st sentence.

            Still when you look at the inner core of stations on the CFN Grafton and Parnell are only served by one line, that I don’t believe is ideal

            1. If there is no demand for this direct cross town service, easy to discontinue its use with little cost having been spent.
            2. As it would never go into the CRL it could still be run potentially with diesels if the number of new electric trains were stretched across the rest of the network.

            low cost experimentation to actually test demand / no extra trains.

          2. Parnell and Grafton are fine with one line with high frequency, they are city fringe stations but also just a quick transfer away from every other route. Grafton also is well integrated with busy bus routes.

            There is nothing cheap about diesels, but also no space to terminate trains at Parnell.

            Keep it simple, lots of service, and all points within one transfer and it’ll be not only a fast, legible, and attractive network, but also a cost effective one.

  11. Would like to be positive about this but am struggling! Too many lines with unnecessary confusion. Agree with the sentiment that loops are loopy and ‘doubling back’ should be avoided. Personally I think the best and simplest solution is
    Line A: Swanson to Papakura via CRL and Newmarket.
    Line B: Manukau to Onehunga via Eastern line, CRL, Grafton, Newmarket.

    By keeping the network down to two lines you have excellent legibility and can also run a good frequency. Given that the next most likely upgrade to the rail network is an Airport link, it would be a case of extending Line B and no changes to the line structure would be required.

    The two problems with my suggestion that I can see are:
    1/ Lack of direct west to south trains.
    2/ Too many trains terminating at Onehunga

    1/ is only a problem without Newton station. If Newton was built then transferring would be extremely easy, and a high frequency and cross platform transfer would make connecting between Line A and B easy. With no Newton station, then the time wasted is more significant.

    2/ Is only a problem until the Airport line gets opened. As Matt touches on, it’s not ideal to be terminating trains at Newmarket because operations become much trickier once you start changing directions at such a busy station.

    1. I also thought of that pattern, but I think 6tph to Onehunga may be overkill. I would also think with 6tph going to Papakura, then you would need some kind of grade separation at Penrose to avoid conflict (I could be wrong on this) as that would see 12tph in each direction.

  12. I can understand a need to reduce cost and not build Newton but to preclude it from being built later seems daft to me.

  13. In regards to potential confusion over destinations: what would be wrong with trains’ displays continuing to change as they pass through Britomart?

  14. AT have excellent station to station (line to line) detail in their AT HOP data. In terms of Western to Southern Line this is extremely low currently. In terms of Western Line customers wishing to use Grafton this was reasonable and is in the Top 10 Stations for patronage in the overall network (hospital + schools + some workers). I think the balance of every 3rd Western L:ine train having access to Newmarket is about right on patronage demand – or wait from the long trip through the CRL to loop back around to it.
    I wonder what the turnaround times might be like at Otahuhu (or are they using Westfield? Huge frequency in the Penrose to Newmarket section which could grow a lot of patronage, especially if bus customers could easily hop off at Ellerslie and jump on a high frequency train service (speed/time is much better). If this was like a Panmure situation you could see some signficant overall PT growth. One of the biggest losers I can see is the highly talked about Parnell Station. If it is only being serviced by one line the frequency will be poor – and with this being heralded as potentially the 2nd or 3rd highest use station in the near future, this will not happen. Shame for all of the uni students as potential future PT converts.

  15. “How confusing will it be with the Red line doubling back on itself?”

    I think you’ve misunderstood – the map provided by AT shows operational patterns. Trains going back to where they came from already happens right now, so it’s no different. Public timetables and network maps will look different, probably still yellow for Eastern and red for Southern.

    1. The operational plan is different again as it contains peak only services, some of which short run. They wouldn’t let me take that plan out of the office though.

        1. That’s a shame. I’d always thought one of the benefits of CRL was to free up the terminating platforms at Britomart for potential regional/intercity services.

          1. They will operate through the station, but be terminating for the passengers. I.e., Swanson to the stabling yard at Quay Park, with Britomart being the last station for passengers.

          2. Arthur pay the fool no mind. A few services may terminate there as at other parts of the network, but not for capacity reasons only for operational ones, and they will be by far the minority of the trains that visit the station. Silly Geoff thinks he is being clever but is merely exaggerating a minor detail out of all proportion, as he always does.

            Furthermore no one knows what the eventual running patterns will be, but as one of the few points on the network with stabling there are bound to be some trains stored there, it’s hardly an amazing or interesting observation. And it certainly not one that undermines the general pattern of through-routing. Geoff just gets a little boyishly excited by imagining he’s the clever clogs that proves the world wrong.

  16. I don’t see any realistic operating route which gives every major destination access to every other major destination. Therefore, instead of duplicating resources, we should just accept this reality and focus on

    1) making every station accessible to every other station via only one transfer.
    2) make transferring short, easy, painless. For example, little wait times, free wifi in stations, comfortable seating, syncronized timetables.

    1. In general that’s the idea, but there will be no space to terminate services at Mt Eden. also without Newton the fastest way to Aotea from the south is via Grafton which is why we flip that route through to Panmure and the O-Line through Parnell. Also Grafton is almost certain to a much more important destination than Parnell because of the new Uni campus there [plus hospital and Med school], and the line around the Vector is worth avoiding where possible…..

  17. What’s wrong with the Strand railway line? Ah I see, I didn’t consider Grafton. Re space to terminate at Mt Eden, is it not possible to make space? There’s going to be a ton of land used for construction, wouldn’t it be a glimmer of a budget to add in a siding?

  18. The double back is a real mistake. What are they going to call the red line trains at Papatoetoe? Red line clockwise vs red line anticlockwise? Read line east then the west vs red line west then east? It reminds me of the district line in London that has lots of branches and is really confusing.
    Also the peak only bit isn’t needed, would be better to make people change than to have the added complexity.
    Simplicity is the key to success with public transport.

    1. Southern and Eastern trains are all on one Southern Line timetable right now, so it wouldn’t be any different? The Southern Line timetable already covers both routes, so really they should just drop the Eastern timetable (the yellow book). I suppose it might be handy to some of those living on the Eastern itself, who start and end their journies on the Eastern, but they would be a minority, and really it’s become something of an unecessary timetable product. Onehunga much the same, that’s 100% in the Southern timetable as well.

      1. Yes, it would be different Geoff as this isn’t two lines, it is one line going in opposite direction, so that instead of 2 choice of Eastern and 2 of Southern you have 4 red choices.

        1. Have you seen the Southern Line timetable? One timetable, one colour, covers Onehunga, Southern and Eastern.

  19. The problem is the NZ government and councils in general make silly cutbacks to public transport projects that are vital in an attempt to save money just like they did when building the Manukau Branch when they refused to spend anymore money extending the line where it actually needed to go and whats worse the way the station was built, was built in a way that it was almost impossible to extend it further without spending an obscene amount of money putting it right.

    Not building the Newton station is a wrong step in the wrong direction. Why can’t we cut the over-funded roading budget instead?

    1. My thoughts too, hobbling the system and the regeneration of Newton to save $150 million, a sum that for any motorway project is just a rounding error.

  20. I’m wondering whether the crl is worth the spend if it’s only going to benefit the Karangahape and not also the Newton areas. Is it now mainly making the western line trip to the cbd a shorter trip? Of course the government don’t seem to be factoring in the benefits to these areas.

  21. It’s also not fare because $150m is not a significant cost saving, because the CBD rail link even without Newton Station in the equation will still cost a lot of money to build and the fact that it will never get built means that the Newton misses a vital opportunity to redevelop and become an important area. It’s true that we want to see the rail link built soon but the reality is we will not see it built until after 2020, especially if National win the next election which is highly likely unfortunately, I’d rather see the CBD rail link get built properly than have a half baked solution like what happened with the Manukau rail link.

    If Newton station is to not be included it should at least be future proof so it can be built later on in the future if possible rather than leaving it out all together and making it almost impossible to build!. Not a good decision in my opinion on the Auckland Council!

  22. I think the Purple line would be much more effective if it went to the Panmure interchange. This would give a single transfer (bus-train) trip to the likes of Ellerslie, Greenlane, Newmarket, Mt Eden etc without having to go the long way. It just needs a small section of track and wiring. What was the Manukau south link pricing? $10M?

Leave a Reply