When it comes to public transport network planning, the vexed issue of transfers always comes up. Inherently, public transport routes are fairly inflexible while travel patterns tend to be very flexible. This means that your trip isn’t always catered for by the route the runs past where you are – so therefore you might need to catch two buses to complete your trip, or a bus and a train. Switching between the two is a “transfer”, and how attractive (or unattractive) that transfer is tends to have a large effect on the success of public transport in being able to service trips beyond those who work in the CBD.
If we look at the Auckland approach, as can be seen in the maps attached to my post yesterday, all of Auckland’s half-decent bus routes operate between the suburbs and the CBD – effectively being useful for those who work or study in the CBD, but generally not likely to be useful for anyone else. Although Auckland does have a number of cross-town bus routes, generally these run at low frequencies and have low patronage in any case – making it extra-difficult to justify additional investment in such services. Furthermore, parking is generally provided free of charge outside the CBD (ie. it’s subsidised by a little bit added onto what you pay for stuff, a little bit chopped off your pay-packet and a little bit added to your mortgage repayments or rent) which makes it extra difficult for public transport to be preferable to driving. However, at the same time I don’t think we can simply “write off” bothering to improve public transport for those who work outside the CBD, as in Auckland 88% of jobs are located in “the suburbs”.
So how can we provide a public transport service that will actually be useful for people other than those who work in the CBD, and also how can we make our public transport system useful for trips other than the “home to work commute”? I explored part of the answer in my post a few days ago on “Urban Transit versus Commuter Transit“, although for this post I want to delve into a bit more detail about networks, timetables and fares – which really are the nuts and bolts of a public transport system. And, a critical part of this is the issue of transfers.
Again, I will use Paul Mees’s ‘A Very Public Solution‘ as a source for some useful quotes on topic. Mees’s whole book is based around exploring ways to improve public transport for dispersed cities – both in terms of low residential densities and also in terms of dispersed employment patterns, so it explores this aspect in quite a lot of detail. Mees argues:
The best ‘regular’ (urban) public transport systems, like Zurich’s, have net- or grid-like structures, comprised of both radial and cross-suburban routes, with high levels of service on each kind of route. The busiest routes (which in strong-centred cities are usually the radial ones) may be operated by higher-capacity modes (rail or tram), but the essence of the concept is high service levels for travel in all directions. Because it is not possible to directly link all trip origins and destinations, flexible public transport systems rely on a high rate of interchange between routes.
This ‘interchange’ is what we call the ‘transfer’ – when you get off one bus and onto another, or off a bus and onto a train (or vice-versa). Transfers are annoying, but in many ways they are a necessary evil as they provide the system with enormously increase efficiency and flexibility – as a ‘hub and spoke’ system can be created, or the kind of ‘net’ that Mees talks about. Transport planner Robert Cervero looks at the balance between the annoyance of transfers and the advantages they can bring:
Transportation experts have long regarded the ease of physically accessing transit facilities, along with the maintenance of frequent, reliable schedules, to be key determinants of whether travellers opt for transit or not. Commuters particularly abhor the hassle of transferring or anything that disrupts the process of making a trip. Where densities are low, trip ends dispersed, and networks laid out on a grid, transferring becomes an unavoidable way of life. Near effortless connections of modes thus becomes imperative if motorists are to be won over to transit.
Now if we look at Auckland’s system, it seems as though great care has been taken to avoid the need to transfer – to as great an extent as possible. However, this has come at the cost of making the system pretty inefficient and running many of the routes at incredibly poor frequencies. So it’s clear that a balance needs to be found – where the ‘cost’ of transfers is minimised, yet the advantages that transfers create are captured. The key to this is making the transfer easy, which I’m afraid is about as far from the Auckland situation as humanly possible.
Mees explores this matter further:
Easy transfers require excellent facilities at transfer points, which in a grid network means at every intersection of the main streets along which buses or trams run. They also require special attention to timetables and fares.
If transfers are to be promoted, timetables must be structure to minimise waiting for transferring passengers, which in turn requires either rigorous co-ordination on the pulse-timetable model, or frequent services on all routes.
Making transfers simple, in terms of waiting time, is really only half the story though. The other half of the story is transfer-friendly fares. This means integrated ticketing between companies, but also what is known as “free transfers” – or ‘time-based’ rather than ‘trip-based’ ticketing. The idea is that you buy a ticket that covers unlimited trips within a certain amount of time – either two hours, a day, a week or a month. This is the kind of system that I envisage for Auckland, and Mees explains it a bit further:
A fare system based on free transfers fits logically with an emphasis on periodical tickets to reduce boarding delays on buses, as well as to encourage passenger loyalty to public transport and off-peak travel. In Zurich, periodical tickets are heavily discounted and promoted as an environmental measure. By contrast, the typical pattern in free enterprise public transport is for cumbersome fare systems designed to maximise operator revenue regardless of the inconvenience to customers. The acme of this approach is the ‘section’ or ‘stage’ fare systems used on buses in Britain and in some Australian cities [and in Auckland obviously], in which passengers making unfamiliar trips rarely know in advance what the fare for their trip will be, and periodical tickets, when they exist, can only be used on a single operator’s services.
I don’t need to rehash the problems Auckland experiences through a lack of integrated ticketing, but I certainly think that for transfers to ‘work’, we really need to ensure that they are as painless as possible – and ticketing plays a crucial role in that.
All up, there are clearly pros and cons to systems based more around ‘transfers’ than what we have at the moment. The advantages are that you can use the current infrastructure more efficiently (through feeder buses and the grid-like system detailed above) and you can provide for a greater number of trips. The cons are, well, that’s it’s damn annoying. So the real focus needs to be on making transfering as ‘un-annoying’ as possible. That involves integrated ticketing, that involves an overhaul of the fare system, that involves better frequencies on cross-town routes and that involves better transfer points between services. It’s a pretty big task, but the benefits could be quite enormous for Auckland’s public transport system if we went down this path further.
Processing...
That’s an interesting consideration — that transfers are actually a fact in a flexible infrastructure and that they need to be made nice and easy.
I think through routing and integrated timetabling would reduce the hassle of transfer immeasurably…
There are transfers and there are transfers. I am constantly amazed when ever I go back and visit Perth just how seamless their bus train transfers are. Any inconvenience of transferring is more than offset by the travel time benefit of the fast trunk rail service and the increased bus frequency made possible by not running every bus into the city. But then bus stands are often just an escalator ride away from a train platforms again making the transfer seamless.
Electronic displays are provided at the bus stands for bus drivers so they know if the train they are connecting with has arrived yet or not and whether they need to wait.
The on bus driver consoles in some European networks actually tell the driver if connecting buses have arrived and whether they need to wait for a late service.
It all acts to give passengers absolute certainty that they won’t be stranded or delayed along the way.
This is a world away from trying to transfer in Auckland where the network is just not designed for transfers. Services don’t generally connect or wait for each other if one is late. If passengers don’t have absolute confidence in a transfer then it becomes a drama. No one wants to be stranded in the middle of nowhere waiting for a connecting bus that may or may not come.
A reasonably transfer doesn’t need to be complex or infrastructure intensive, especially where there is an obivious main traffic flow (from a feeder bus to a main line train for example). In my case the transfer I use to get to work each day is very low tech. There is simply a bus stop on the street closest to the train station, and the bus driver waits there until the next train comes. Because the buses run at roughly the same frequency as the trains, 9 times out 10 there is a bus waiting for you at the station. The bus doesn’t leave until the passengers have stepped off the train and made their way over and are all on board. If the driver sees a train coming from the other direction he will occasionally stay there an extra minute or two to pick up those people also.
While it is far from ideal and occasionally frustrating, it generally works quite well and requires zero investment. All it needs is a little management and cooperation between the companies running the train and bus. This approach would tend to work only at the ends of a bus route or at major interchanges. You couldn’t have the bus stopping and waiting at every single interchange point on a grid without some serious co-ordination of timetables, otherwise it would spend more time waiting than moving. Also it only works well in one direction, as the trains do not wait for the bus on the trip home.
But still, in Auckland it could be very useful for bus routes that ended at a suburban rail station, busway station or ferry terminal.
I think transfers are needed for better public transport but while once transfer is okay, two or more is not that good. I think of it that every time you transfer you lose 50% of potential passengers. So its a trade off between whether its worthwhile providing them a direct service or not
I really think it depends on the quality of the transfer Louis. With a really seamless interchange you might not lose any potential travellers, and multiple transfers can likewise be fine if they are quick and easy.
For example, I’m about to finish work and head into the city to meet a friend. This involves a feeder bus to the station, a train into the city loop then a short tram to my final destination. Because the feeder bus works well and the trams at the other end are super frequent there is basically no delay involved in transferring.
Really, if connections can be made really efficient then a PT trip involving transfers is no more difficult that a car trip involving more than one road.
I agree Nick, that it really really does depend on the transfer.
Let’s say we turned the Northern Busway into a railway line, and ran feeder buses to it. This would mean a lot more people would have to transfer than do at the moment, but if we ran trains at 6 minute frequencies that would mean an average waiting time of around 3 minutes. If the train trip was say 5 minutes faster than your old bus trip, and the transfer was very easy, then you would actually be better off time-wise by making the transfer.
The current cost-benefit analysis of public transport projects in New Zealand ignores this possibility, and just slams on a 50% transfer loss arbitrarily.