Wellington Scoop

Too many buses, too many queues

by Kerry Wood
Bus Rapid Transit was to run between three interchanges, at the Railway Station, Wellington Regional Hospital and Kilbirnie. Now it looks too light for the job, which probably means light rail.

Wellington’s old trams ran to Wadestown, Northland, Karori, Brooklyn and so on. They cost a mint of money, with two tunnels and four new roads for those four routes alone. The Council then had no real alternative, but later used buses for new suburbs.

Building tramways is still expensive: around $50 million a kilometre all-up. There is little point in spending that kind of money if buses will do the job, and trolley buses bring most of the environmental benefits of trams.

The small version of BRT chosen by the Greater Wellington Regional Council has a capacity of around 3000 passengers an hour. The economic minimum for light rail is also somewhere around 3000, on the busiest section. An important difference is that BRT buses can go beyond the interchanges on some suburban routes but light rail can’t. The rails must end somewhere and transfers are unpopular.

The Regional Council has a dozen or so existing and planned interchanges but has done done little to explain the benefits. Two hypothetical survey questions illustrate the dilemma:

Do you want to break your existing journey by making a transfer? (and perhaps wait a long time, wait in the rain or pay another fare)

Do you want faster, cheaper and much more reliable trips, anywhere-to-anywhere? (and usually save more time on the faster bus or tram than you lose on the transfer, with more reliable services, most transfers under cover and main routes more frequent)

To see the possibilities, imagine a town built around two main roads, with six bus routes, It is convenient for passengers but needs a lot of buses.
A much cheaper and simpler layout is an interchange where the straight-through blue and red routes cross. Buses from all four directions arrive at the same time, passengers transfer to whichever bus they want and the buses go on their way. It is quick and reliable, unless traffic congestion makes the timetables too unreliable. Quality services need a high priority to keep the transfers working.

other routes
Other routes can be added, making the diagram look more like a star. This is how the Christchurch bus system works. All trips are easy and quick, and passengers have much wider public transport options.

two halfstars
At present Wellington is similar, with two half-stars linked by a spine route: a binary system. In principle it can be extra buses on the blue route. But today, too many buses delay each other. The traffic lights group them into queues too long for the stops unless the traffic signals let them through a few at a time. It all sounds very familiar.

If BRT and bendy-buses cannot cut the mustard, the next step is light rail, but the blue route is still there.
brt and bendy The reason is service: transferring the Airport Flyer passengers to light rail would be a backward step, unless trams ran to the airport. Routes such as Brooklyn must also run through the city to reach an interchange.

The Railway Station is the obvious place for a northern interchange. Buses and trams could go above the rail tracks, with escalators down to the KiwiRail platforms and street level. The southern approach for buses and trams could be a ramp over Thorndon Quay, down through the existing Bus Exchange and ending near Molesworth Street.

Courtenay Place is no good for the southern interchange. First, it is too close to the Railway Station, making the trip too short to recover the time lost in transfers. It is also too close to the point where many passengers are getting off anyway. They will not want to make a transfer to go one stop further, such as the St James Theatre. All interchanges have this problem: the trick is to minimize the number of passengers trapped in the ‘one-stop-short’ problem. And at Courtenay Place there is too little space for a good interchange, with no good options for turning buses from the south.

Wellington Hospital is still a bit close (4.3 km) but much better. Few passengers are going only one stop further and many passengers—staff, day-patients and visitors—can go straight into the hospital. The Brisbane busway even has a stop inside the main hospital. Shops at an interchange are another good move: one Wellington supermarket is already on the ball.

Kilbirnie is a good spot for a third interchange, for passengers from the eastern suburbs. It might be built at the same time or as a second phase. A disadvantage is that passengers from Island Bay to Miramar would have to change twice, at both Wellington Hospital and Kilbirnie. How much that matters depends on how long the interchanges take and how many passengers want to make the trip. Another option might be a bus route around the coast.

Looking again at the theoretical layout, most of those coloured routes lose their city section. Even with no other changes, they keep clear of the most congested areas and run with fewer buses and better timekeeping. They are more reliable and cheaper to run, or some of the savings might go into more frequent services. These routes can be timed to meet light rail reasonably accurately, cutting down transfer times. Sorting out all the delay points along each route will make trips and transfers quicker still, but will take time. GWRC is already recording all the delays, through its real-time display system, and the next stage is putting the traffic engineers onto sorting out the worst spots.

GWRC estimates an average transfer will take 8 minutes (plus walking time) in the morning peak, the only time studied. It will take longer in the evening peak (the transfers are to less frequent services) and much longer at weekends. An average is in any case no way to handle transfers. Eight minutes is simply a fudge-factor plus the average of just catching and just missing the connection, but real passengers like to know which. Good practice brings reliable transfers in under five minutes, and best practice is 95% of transfers in under two minutes. Eight minutes doesn’t cut the mustard. However, GWRC does have a problem: the NZTA’s Project Evaluation Manual specifies a minimum fudge-factor of five minutes. It would be interesting to know why.

In summary:
— Transfers give passengers far more options for public transport trips.
— Transfers allow more frequent services with fewer buses and lower subsidies.
— The timetables must be right (it isn’t easy).
— Transfer time is manageable: less than five minutes should be possible.
— Nothing succeeds like success: more services make the system more effective and attract more passengers.
— There are important co-benefits in public transport, including commercial co-benefits.

Kerry Wood is a retired engineer with a long-standing interest in transport


  1. Ross Clark, 4. October 2013, 3:21

    Two thoughts:

    * My guess is that the region has been told that Central Government will give them $200m for the work if they go for BRT; if they want LRT, they won’t get any more for it, despite the extra cost. On past form, that will have confirmed the decision to go for BRT – the unwillingness to pay the extra local share for LRT.

    * The logic of most local politicians in terms of transport matters is that they want ‘more of everything’. So, in this area, they will say that they want both roads and rail – but if they could only have one or the other (not both), they would grimace and admit that they would prefer more roads to more rail. Because that’s what the voters really want.