Wellington Scoop

Why Wellington’s buses need better timetables and connections

by Kerry Wood
The Regional Council has timekeeping problems on Wellington’s buses (25 May) but what about the timetables themselves? Good timetables and connections are vital to the quality system the council is seeking, and will be even more important when new timetables are introduced next year.

Good timetables need good timekeeping. Passengers win because of shorter wait times, more reliable connections and less need to catch an earlier bus ‘just in case’. The council and ratepayers win because more passengers reduce subsidies, and car users win because good bus services reduce congestion.

Good timetables are ‘clock-face’, running past your stop at the same times past each hour. If the first bus runs at say five past six in the morning, it is good to have a bus at five past every hour, say until five past eleven in the evening.

The pattern may need adjusting for peak-hour delays, although a better approach is giving buses enough priority to keep time. ‘Slack time’ is also important, to cover unexpected delays. If drivers can usually manage a trip in 33 minutes, setting the formal run time at 36 minutes helps a lot. If timekeeping is bad it might have to be 40 minutes: good timekeeping brings better productivity.

How well does the Regional Council do in practice?

Route 2, for example, runs clock-face timetables for part of every day, but not all day and with no consistency of either hours or days. Ignoring the random stuff, clock-face departures from Miramar are:

– 9.06 to 15.36: 06, 21, 36, 51; 15 min headway
– 16.15 to 19.55: 15, 35, 55; 20 min
– 20.00 to 23.45: 15, 45; 30 min
– 8.30 to 12.30: – 00, 15, 30, 45; 15 min, different times
– 13.00 to 17.20: 05, 20, 35, 50; 15 min, 5 min later
– 20.00 to 24.00: 00, 30; 30 min, same times as AM
– 7.00 to 11.30: 00, 30 Same as Sat eve
– 12.00 to 18.00: 00, 15, 30, 45 Same as Sat morn
– 19.05 to 23.35: 05, 35 Unique

This is clock-face in name only, and switching between different numbers of buses an hour, as well as different times past each hour, will ensure bad connections.

The only fully clock-face timetable I know of in Wellington is the Johnsonville Line, which runs every fifteen minutes at peak hours and every 30 minutes at other times, including weekends. It was introduced eighteen months ago and I find it a big improvement. Start times at Johnsonville are:

00, (15), 30, (45) min (bracketed times are peak-hour only).

The disruptive effect of inconsistent timetables can be shown by imagining a Johnsonville bus route with the same timetable as Route 2. The bus would deliver passengers to a citybound train at these times (brought forward two or three minutes, to meet at least some trains):

Weekday AM
– 6 min late; Wait 9 min at peak hours, otherwise 24 min
Weekday PM
– Meet 1 train/hr Peak timetable only (16.00 to 17.45), otherwise wait 15, 10 or 5 min
Weekday eve
– 15 min late Wait 15 min
Saturday AM
– Meet 2 train/hr OK
Saturday PM
– 5 min late, Wait 25 min
Saturday eve
– Meet OK
Sunday AM
– Meet OK
Sunday PM
– Meet 2 train /hr OK
Sunday eve
– 5 min late Wait 25 min

If connections matter, timetables must be a lot better than this.

A clock-face timetable for the new Route 2 (the existing Route 2 will disappear) might be something like this:

– First bus 5.30 then every 15 minutes, last bus 24.00
– Morning peak 07.30 to 08.30, every 7.5 min
– Evening peak 16.30 to 18.00, every 7.5 min
– First bus 6.30 then every 15 minutes, last bus 23.00

No need for a paper timetable the size of an umbrella: a card in your wallet is enough.

Better still, the council can cut out this gem of bureaucratese:

A Sunday timetable applies on all public holidays except when Waitangi Day and ANZAC Day fall on a Saturday when a Saturday timetable will operate.

Of course, timetable quirks are there for a reason, and one reason will be bus scheduling to minimise costs. But how real are the cost-savings? How has the passenger inconvenience of ‘savings’ such as these been costed?

Another problem is connections between bus routes. I tried (notionally) catching a bus at random weekday start-times, taking directions from Metlink, starting opposite Wellington Hospital and ending in Norwich St, near Wilton Rd. The route was chosen for its multiple connection options. A single option usually works best (route a to route x, not route a, b or c to route x, y or z) because fewer routes allow more buses on each.

Most connections were timed at 5-7 minutes. How reliable are they, when ten minutes late is officially ‘on-time’?

The variety of routes was extraordinary, the sample probably too small to pick up all options:

• At least eight different first-leg routes.
• Connections in at least five different places.
• At least four different second-leg routes.
• Two departure stops: one option began with a seven minute walk to route 47.
• Two destination stops, a couple of hundred metres apart: confusing for visitors.

The solution is coordinated timetables, with all headways (time between buses) on all routes chosen from one of these series: choose the series, then design the timetables:

(120), 60, 30, 15, 7.5, 3.75 minutes (used on the Johnsonville Line), or
(120), 60, 30, 10, 5, 2.5 minutes, or
Run so frequently that headway doesn’t matter: effective in large cities but costly.

Another series of possible headways uses 20 minutes instead of 30. However, note that 12 minutes is an unsatisfactory headway, despite being clock-face, because there are no clock-face options between 12 and 60 minutes: how would feeder services be timetabled?
– (120), 60, 12, 6, 3 minutes

Some conclusions:

• Wellington’s bus planners have been tinkering for many years, with no overall plan.
• The inconsistencies have been patched with a computer trip-guidance system, intended to be helpful but in practice often sowing confusion and delay.
• A fundamental problem is too many routes with too few buses on each. This is a major source of delay: frequent services speed trips. The omens are worrying. Some of them are visible on a recent schematic route map (December 2016):
• No timekeeping improvements attempted yet.
• No sign of coordinated timetables: ‘High-frequency’ services are to run every 10-15 minutes, ‘standard services every 30-60 minutes: much like Route 2 today.
• A myriad of routes: eight each through the Karori Tunnel and on Riddiford St.
• Peak-only express services, which don’t work in the evenings. They are popular because they are fast, and the people who rely on them to reach outlying stops often can’t get on.


  1. TrevorH, 13. June 2017, 8:41

    Thanks for this interesting analysis. I have tried to use buses but I have found them unreliable (i.e. they simply don’t turn up) and time-consuming because of their circuitous routing. Is it really true that commuters from Miramar will need to change at Kilbirnie from next year? If so that’s madness.

  2. Fiona, 13. June 2017, 13:49

    No one would want to change buses! Hubs might be alright for the operator but not for the customer.
    If buses are not on time is it usually due to poor resources and planning. Nothing to do with circuitous routing. Everything has to start at A and finish at B.

  3. H., 15. June 2017, 17:16

    I agree on not wanting to change buses – but that’s better than routes being totally stopped, A corollary of fewer-routes serviced more frequently is that the servicing of a lot of houses could drop substantially. Wellington’s streets can be windy (hmm, both kinds – that that blows & that that bends) and not always feeling safe or pleasant to walk at night. On leaving a movie/restaurant/theatre at 10 or 11pm a 5 minute walk home from a bus that you’ve waited 15-25 minutes for in a well-lit area can be much better than a 25 minute walk where you’ve waited 5 minutes.
    Dropping routes will reduce night-time use of public transport.

  4. Henry Filth, 16. June 2017, 15:26

    “A fundamental problem is too many routes. . . ” What? For someone to catch a bus, there has to BE a bus.

  5. Kerry, 17. June 2017, 15:18

    It is counterintuitive but true, Wellington has too many bus routes with too few buses on each. The choice is not between bus and no bus, but between a short walk to an infrequent bus or a longer walk to a more frequent and often faster bus. Nobody wants a new change on their existing route unless there are other advantages.
    The whole point is that there usually are other advantages, if the system is well designed. That needs better timetables, hubs and timekeeping, or in Wellington much better all round.

  6. m, 18. June 2017, 19:17

    I can catch a 44/43 or an 11. Pretty much wherever I catch it, the 44/43 and the 11 come within minutes of each other. It’s very frustrating to see endless “Lyall Bay 3s” going past before two buses I can catch come at once. The 44/43s are on clock time. They leave from the Strathmore terminus at :55 and :25 weekend/weekday.

    The thing is that the transport people are wedded to the idea that a bus route goes from one place, follows the same path and gets to the same destination. An 11 leaves Seatoun, goes through Newtown and gets to the Railway Station. A 2 leaves Miramar, goes through Hataitai and gets to the Railway Station. But there could be 11a and 11b and 2a and 2b – the a’s going through Newtown and the b’s through Hataitai. Rather than all this get off at Kilbirnie and swap buses stuff.

    I think most people who use buses would rather the buses came on time with minimum walking, than having more scheduled buses that you have to walk in the wind and rain to get to or having to go to hubs to swap buses. Bus patronage falls away in the winter because people don’t like having to walk to the bus stop in the cold/wind/rain.