This is a Guest Post by commenter Peter, who has written previous posts here and here

There has been a lot of discussion in recent months about the need to improve Auckland’s bus network so that it integrates with, rather than duplicates, the rail network and also the Northern Busway. Auckland Transport’s board papers for their meeting this coming week seem to indicate that this process is finally happening – as Matt pointed out in his post earlier this week, we see the following:

The PTNP will define a new connected and integrated network for public transport in Auckland of:

  •  Rapid Transit Network (RTN) of electrified and modern rail services and Busway services.
  • A connected network of high frequency bus services along key arterial road corridors, supporting and integrating with the RTN – 15 minute worst case frequency seven days a week between 7am and 7pm – plus some high frequency ferry services
  • Secondary routes of lesser frequency ferry and local feeder bus services connecting to the RTN and high frequency bus network, school bus services and peak only services.
  • The main benefits of the new PTNP will be to simplify the network, remove the “spaghetti” A-B bus routes, remove duplication of services and provide high frequency services that are convenient and reliable. This in turn will create resource efficiency “savings‟ that can be used to focus bus resources where the demand and congestion relief is most needed – on key arterial corridors.

It seems fairly likely that turning a number of existing “long-haul” bus routes into feeder services to both the rail network and the Northern Busway will be a key part of this plan. Logically it should, because as many people have pointed out earlier it’s just silly for us to run bus routes that completely duplicate our rail corridors doing the same “long-haul” job that we’ve spent billions of dollars on the rail network to do.

However, this “trunk and feeder” approach generally hasn’t worked well in Auckland so far. We have a few feeder buses that run around the North Shore at the moment, and the general conclusion that I’ve come to from discussions with various people is that they’re rubbish. And because they are rubbish, nobody uses them. Here are two of the feeder services into Albany and Constellation stations: The 880 route (the red one) takes almost an hour to complete its giant circuit of this part of the North Shore, and only runs at 30 minute frequencies – fine if you’re heading into the city where you can transfer onto the much higher frequency Northern Express: rubbish if you’re heading out of the city and just miss your connecting feeder bus to take you home. Furthermore, Constellation and Mairangi Bay are the 3 stage boundaries, i.e. an express from Mairangi to the city is three stages. However for some infuriating reason all the 880 route along Maxwellton Dr falls into the ‘upper zone’ of the North shore, while Constellation is still in the lower zone. The end result is a longer, slower and less direct 839 from Mairangi to town on a lower zone pass, but using the feeder and transfer to the NEX for a quick trip needs the more expensive upper and lower zone pass. The 887 bus (the blue route) takes 40 minutes to complete a pretty incomprehensible route, and also only runs every half-hour – making it hopeless for returning trips.

So we haven’t got feeder buses right on the North Shore. Does that mean feeder buses in general are always going to be almost completely empty and utterly useless, or is there a better way we can do them? Looking overseas, we find that in cities like Perth, Vancouver and Toronto, the majority of people catching the train have actually arrived at their station on the bus – suggesting that feeder buses to rapid transit is quite possible. But how?

Taking a look at Perth’s bus network gives us some clues about how to run better feeder services – here’s a section in the very northern area that feeds into Clarkson Railway Station.  Tracing the individual routes takes a bit of work (and there’s a typo which doesn’t help), but generally they follow fairly logical routes feeding into Clarkson Station in the bottom-right of the image. With some exceptions they take the fastest trips while still offering a fairly decent coverage of the area. Looking at a timetable, we can also see that some care and thought has gone into integration with trains: The connections also work in the outbound direction: What’s interesting in comparing Auckland with Perth is a fundamental difference between a busway and a rail line – that being that you’ll probably always be running much higher service frequencies along a busway than a railway line (simply because you can’t fit anywhere near as many people on a bus). The lower train frequencies make it a bit easier for connecting feeder buses, because you can time your feeder to connect to your train (and vice-versa), even at fairly low frequencies. With a busway, it’s likely to be much more difficult to match the frequency of your feeder with that of your trunk route. The Northern Express runs at 5 minute frequencies or better for much of the day these days, and although you wouldn’t need any feeders to match that frequency, you would need them to be coming at worst every 15 minutes to ensure that non-timed transfers didn’t result in particularly long waits.

In future posts I’ll try to look at how Vancouver and Toronto operate feeder buses to their rail networks, and perhaps look at how Brisbane operates its southeast busway in a bit more detail – to really explore how we can make feeder buses to rapid transit work.

Share this

14 comments

  1. Perth has real time train information available and one use that is made of it is to give a display to bus operators at stops at rail stations showing how long until the train arrives.

    Tony Bailey
    Editor
    Transit Australia

  2. Additionally-

    For those sample bus timetables that you used the train rums far more frequently then the buses.

    Perth headwayss are the model for electrified rail services in Auckland and Adelaide.

  3. there are two ways of approaching PT supply, opeartionally driven and customer driven and particularly in the case of the Busway, a tension between the two (obviously the following does not apply to rail!)

    while operational efficiency in the peak may suggest feeders to busway stations and a transfer to a linehaul bus, a customer focus suggests the opposite; thoughts like “I’ve got my seat and my book out and I don’t want to get out of this bus because it’s raining” and “there may not be a seat on the next bus through” and possibly most important on the trip home “this bus is running late and I’m going to miss my connection” work against feeders when the nice warm car is sitting in the garage

    the planning concept for the Busway was to run expresses in the peak with reasonably direct routes to the Busway and then the quick ride into town, with the trunk/feeder concept in the interpeak and evenings, sadly ARTA didn’t appear to understand the Busway concept and the opportunities it offered and did a minimalist rehash of existing services, they really didn’t plan ahead to implement anything different and lost a huge opportunity

    clearly, rail is a different case as you can’t run buses on a train track and several of the objections don’t apply, trains tend (should) have better on time performance for example, but even then you don’t want transfers too close in as you are carrying too many empty seats for short distance passengers and the psychological transfer penalty gets more severe the closer you are

    1. I don’t disagree with what you say there Steve, bu the reality is that we need the ‘operationally driven’ to provide the sorts of frequencies people can rely on for usin public transport in a normal, flexible and spontaneous life.

      As an example, a single bus and driver could do the trip from Mairangi Bay to Constellation and return in twenty minutes comfortably. As a comparison the same bus and driver would do the trip from Mairangi Bay to Constellation and down the busway to downtown (I.e the 86x) in about 40 minutes, or 80 minutes return.

      So with the resources of two buses and drivers we can operate the feeder to constellation every ten minutes, or the full route to town every forty minutes.

      What is more likely to put people off, a bus every ten minutes connecting to the NEX which runs every five minutes, or a bus every forty minutes that duplicates the NEX for 90% of its route? Who is going to get in the car, the one who just missed the bus and has to wait ten minutes, or the one that has to wait forty?

      Sure there is still a place for peak hour expresses that are full by the time they leave the suburbs, but in the case of Mairangi and the like that is perhaps three or four each way per weekday.
      The general network has to be built around transfers, otherwise we’ll never be able to afford anything close to the required frequency.

  4. Keep in mind that bus companies in NZ are allowed to operate their own commercial urban services, and so they do this in addition to council contracted services, and operate in competition with trains. For a feeder system to be effective, you would need to introduce legislation that outlaws commercial urban bus operations, to prevent the feeder system from being undermined. However, I can’t see the government doing that. Preventing private enterprise from growing its business is a big ask.

    1. Actually that has been more or less addressed under the new PTOM contracting model that is being implemented as we speak. From now on bus services will be broken up into area units, and operators must run the whole unit to an agreed minimum standard of all day every day service. Services in each unit will be planned by local government in partnership with operators, with a focus on whole-system network connectivity. It will still be private enterprise and competitive tenders, but from now on operators will have to take the feeders with the trunk routes, the peaks with the Sunday night runs. There is still the potential to register an entire units as commercial, which I suppose would be the ideal if we can get a complete integrated chuck of the all day network running to a good standard entirely subsidy free.

      The only remaining stumbling block is integrated fares, which seems to be progressing slowly. Shortly we will have an integrated *ticketing* system with flag fall discounts on transfers, and there are ongoing concepts about fully integrated zone systems that will hopefully become a reality soon.

  5. Thanks for the feedack everybody.

    Good to hear that Perth is seen across Australia as “best practice” – although my reading of the timetable is that bus feeder frequencies match train frequencies. I’ll double check there.

    The mentality with busway feeders of “but this bus could just continue on down the busway” will be an interesting one to explore over time. Nick makes a good point though that getting high frequencies throughout the suburbs is dependent upon ensuring those trips are as quick to complete as possible, which clearly means feeders rather than long-haul routes into town.

    Right, now to make sense out of Toronto’s bus network.

    1. With regards to the train frequency – the extracts here are from the bus timetable, which only shows trains that connect with the buses on the timetable (there’s a footnote on the timetable cover explaining this.) The full train timetable is published on a different timetable, and is more frequent than the buses here – every fifteen minutes off-peak for example.

  6. I tend to agree with Nick re that feeders do worse than a bus which works as a feeder then immediately continue down the busway to the city. I live a stones throw from Torbay shops so can catch any Torbay / Long bay headed bus. I rarely use the feeders but at times have used the expresses (87x and 881) as feeders home (After ending up at Britomart and catching an Northern Express bus to the busway)
    Yes a bus going in to the city will be empty nearly half the time as it travels back for a second load but so will the feeder. Trains work better with feeders than busways as the boarding time for trains is so much faster than buses- buses at best take 3-4 seconds per person but overall I would allow 5 seconds. The Northern busway can roughly fit 3 buses so can load 3x5x1/5=3 people per second. – 180 per minute or 10,800 per hour. Obviously there would be times when no buses would be loading so the actual number will be less. This is less than the 19,000 that optimal busways can get per hour.

  7. oops – my calculation was not fully explained. 3 buses loading at the 5 Northern busway stops times by 1/5 as only a fifth of a person can get on each second to get 3x5x1/5 = 3 per second

  8. Here in Ottawa, Canada we are planning to run a 12 km LRT through the downtown core, with various feeder buses at each end. During morning rush hour, about 100 people/min need to exit buses and board LRT cars. Vice-versa in the afternoon.

    I’m trying to find an example of a system where this has been tried, as I’m very skeptical that 100 people/minute can exit train cars and find a specific bus heading home. Does this even have a hope of working?

Leave a Reply

Your email address will not be published. Required fields are marked *