Wellington Scoop

Sorting out the chaos: time to stop ignoring the problems with public transport

by Kerry Wood
The Regional Council’s Public Transport Plan (PTP) does nothing to solve Wellington’s biggest transport problem: the cascading bus delays of Wellington’s golden mile. The proposed Bus Rapid Transit (BRT) solution is little more than a name-change, with some extraordinarily bad route choices. Now, with the Basin flyover rejected, is the time for a much-needed change of direction.

    Two Neglected Problems

Officers and councillors have spent three years tiptoeing around two politically difficult problems. Public transport in Wellington needs more central-area road space, or more passengers transferring. Or a bit of both. Ignoring one problem makes the other worse, and ignoring both virtually guarantees failure.

Reallocating road space and junction time to public transport is unpopular with car users but very effective. It relieves bus congestion, cascading bus delays and unreliable timetables. Faster and more reliable public transport attracts passengers away from cars, and fast peak-hour services set a limit on congestion: the ‘Downs-Thompson effect’. Under these conditions mode-switching to quality public transport is far stronger than NZTA modelling admits: potentially over 30%.

More transfers are even less popular than bus lanes because they delay passengers and slow trips. This is the perception, but the reality is that properly designed transfers make many or most trips faster, and the system better adapted to flexible passenger demands.

Morning-peak trip-speed estimates in the 2011 Bus Review (MRCagney, Table 14) show savings generally greater than losses, even before timetables are optimised:

  •  40% no change or faster: 35% up to 5 minutes faster, and 25% over 10 minutes faster.
  • 60% slower: 20% up to 5 minutes slower and 12% over 10 minutes slower.

Many of the slower trips are because of new arrangements at Johnsonville.

Several factors contribute to faster trips:

  •  Bus utilisation improves because cascading delays are rare, some routes are shorter and new route design minimises duplication and self-competition with other routes.
  • . Costs fall because of better utilisation, and faster and more reliable services generate more revenue.
  •   Core services can run more frequently because better utilisation frees buses for more effective redeployment.
  •   Waiting times are shorter because core services run more frequently and all services are more reliable.
  •   Trips and transfers are faster because interchanges and timetables are designed to minimize delays.

More integrated services give passengers more options. They can easily travel to new destinations, or change plans at short notice, knowing that public transport can deliver. A common result is fewer cars, because families can sell the second or third car.

Nobody likes transfers but everybody likes quality systems relying on transfers. Examples include London, Portland, Freiburg and Schaffhausen (population 35,000). The secret is quality: transfers as quick and pleasant as possible, well designed, under cover and preferably cross-platform. Just as important is locating interchanges in the optimum positions to minimise unnecessary transfer.

Two hypothetical questions illustrate the attitudes:

Do you want to break your existing journey by making a transfer? No
You may wait a long time, wait in the rain or pay another fare

Do you want more frequent services and faster, cheaper and much more reliable trips, anywhere-to-anywhere? Yes
You will usually save more time than lost on the transfer because buses are more frequent and reliable. Most transfers are under cover. More effective passenger transport may allow you to sell the second or third car.

Delivering the second question needs transfers, so that frequent services can be adequately utilised. Transfers are much more acceptable when properly evaluated and understood, and passengers encouraged to see the big-picture benefits. And simple, reliable transfers in quality interchanges help too.

    The 2011 Bus Review

The 2011 Bus Review, by consultants MRCagney, is Wellington’s most important public transport study that I can remember. It recognises and addresses the problems, and highlights the difficulties (p 25):

The politically hardest part of a high-frequency network is the reliance on connections.

Throughout this proposal, we suggest replacing some services that now run every 30-60 minutes with connections [to] services that run every 15 minutes or better. The connection is a disincentive, but the alternative is usually self-competition, along with poor frequencies that are not attractive for travel between key destinations in the city. Strategic investment in interchange points is an important consequence of this proposal…

The need for transfers had been clear since 2009, when Opus completed a review of bus operations in Wellington’s central area (p 4):

Analysis indicated that reducing the number of buses using the Golden Mile would improve the efficiency and reliability of the bus operations. The creation of hubs outside or at the periphery of the Golden Mile has been identified as the preferred long term solution…

The Opus study (Figure 3.7) shows that average peak-hour bus occupancy on the golden mile did not exceed 27% in 2009. Too many under-used buses are delaying each other and clogging up the golden mile.

The two neglected problems were, well, neglected and the study tossed out. Someone must have assumed a better solution would show up. It didn’t, hasn’t and under present policies won’t. No later studies have either faced these problems or delivered adequate results.

The least popular features of the 2011 study are in reality just what is needed:

  •  Nine well-placed interchanges, including Johnsonville Railway Station; Wellington Railway Station; Wellington Hospital and Kilbirnie (needed to manage bus numbers).
  •   A secondary bus route between Whitmore St and Taranaki St (also needed to manage bus numbers).

This proposal was altered, before consultation, to a longer and less satisfactory route using the Quays, Cable and Wakefield Sts and Kent and Cambridge Terraces. Two particularly bad features were ‘Courtenay Place’ stops in Kent and Cambridge Terraces, and a southbound route on the harbour side of the Quays and Cable Street: too far from the city, too difficult to cross the road and too exposed in bad weather.

  • Fewer Johnsonville buses to the city. Passengers would either take the train or an indirect bus route by Newlands, with an improved interchange at Johnsonville.

The intention was lower costs: it is difficult to to justify a subsidised bus route running parallel to a subsidised, under-used railway. A further problem was that rail is slower than road, by about 6 minutes. The consultant preferred rail, for future access to a growing centre and to bypass congestion in the Ngauranga Gorge. Also noted was the option of conversion to light rail. Recent timetable experiments on the Johnsonville Line might help to speed up the trains.

  • Not quite ignored was a maximum of sixty buses an hour each way on the golden mile (or any similar bus route), described by officers as an ‘aspirational target’ and never achieved since.

The 60 bus/hr limit is emphasised by the consultant and explained very clearly. In summary, the reasoning is that minor stop delays of up to about a minute are routine and inevitable. If there is only one lane each way, buses cannot overtake and delays accumulate. With very little bus priority at signals, minor queues are marshaled into long queues and the cascading delays familiar in Wellington.

The Bus Review proposed a maximum 73 bus/hr on the corridor, further reduced to 57 bus/hr by a secondary route for peak-only services. The secondary route covered only the busiest section, Taranaki Street to Whitmore Street, but the combination of routes achieved the target, less than 60 bus/hr, a creditable 57% reduction.


Passengers will not accept more frequent transfers unless the reasons and benefits are clear. Satisfying them is much easier if interchange design is done well, with concept drawings available at the consultation stage. Without drawings, passengers will naturally think of what they know: unacceptable existing interchanges such as Johnsonville and Kilbirnie.

Large interchanges must be close to major destinations, and even small interchanges are best placed near passenger ‘attractors’ such as local shops. The worst possible place is one stop away from a major destination, as proposed for Wellington Hospital.

Probably the best interchange in the Wellington Region is Waikanae, and better examples are appearing in Auckland. Perhaps the worst in Wellington is Johnsonville. The Bus Review (p 49) says:

A major rail hub in a regional activity centre will always need a substantial number of bus lines terminating there, because the buses need to both serve the activity centre and connect with the trains. The current configuration in this area, which requires buses to circulate through a mall car park, is unacceptable and needs to be studied to create an appropriate facility.

The Bus Exchange at Wellington Railway Station is better but a long walk from trains, and probably inadequate as a spine route terminus. There are no published studies. New facilities will need parking and turning space for bendy-buses or trams (which usually reverse), and turning space for buses from the north. This will probably need more space: an new bus terminal above the existing passenger rail platforms would be a good choice.

A Kilbirnie interchange could be near Bay or Onepu Roads, and either Rongotai Road for connection to bus routes, or Coutts Street for spine access to Wellington Airport. In either case the interchange must be laid out so that passengers running to catch a bus do not have to cross a busy junction, and all buses running in the same general direction depart from the same platform. It will have to go on private land.

    Limited Route Choices

Wellington has limited options for quality public transport, because the city is hemmed in by hills and harbour:

  • Limited layout choices.
  • Limited route options, particularly in the Old Bank/Willeston Street area.

A self-imposed limit should be keeping principal routes away from the Basin Reserve. There is no particular reason for going that way and congestion is inevitable without grade separation for buses. A secondary route will be enough.

  • Up to three mode choices.

All these options can solve the cascading delay problem and make public transport faster and much more reliable. However, easy, effective and cheap options are unlikely.

    Limited Layout Choices

There are only three or four route layout choices:

1. A single two-lane route through the central city, largely or wholly on the golden mile and probably from the Railway Station to Wellington Hospital and Kilbirnie. Other inner-city services would be limited to a few buses on other routes, with no priority.
Passenger numbers transferring would be much higher than today, some of them having to transfer twice, such as at Kilbirnie and the Railway Station.

2. Two two-lane, two-way routes through the central city, or three routes if the present route philosophy is retained. One route might run from the Railway Station to Kilbirnie.  The second might run from the Railway Station to Courtenay Place, or might be a shorter relief route.
Passenger numbers transferring would be would be higher than today, probably with some incoherence in route design, such as different stops at peak hours. A particular problem is how to manage inner suburb routes such as Kelburn and Brooklyn, with few alternatives to joining the golden mile at the busiest point.

3. Two two-lane, one-way routes through the central city, in opposite directions, probably running from the Railway Station to at least Taranaki Street.
Passenger numbers transferring would be roughly the same as today, but with substantial incoherence because of separated routes, perhaps widely separated.

4. A single four-lane route; probably impractical but included for completeness.

    Limited Route Options

The number of possible main routes is much wider than assumed in the spine study but still fairly small:

. The golden mile: Lambton Q; Willis St; Manners St; Courtenay Pl.
. The Terrace, to Salamanca Rd & Ghuznee St.
. Lambton Q; Whitmore, Ballance or Brandon Sts; Featherston St; Victoria St; Wakefield St to Taranaki St.
. Lambton Q; Whitmore, Ballance or Brandon Sts; The Quays; Wakefield St, with both bus lanes on the city side of The Quays and Wakefield St.
. The Bus Tunnel and Kilbirnie Cres.
. Taranaki St; Wallace St; Riddiford St; Constable St; Crawford Rd.
. Cuba St; Taranaki St; Wallace St.
. Tory St; Tasman St; Adelaide Rd.
. Kent & Cambridge Tcs; Adelaide Rd.

The single-lane Pirie Street Bus Tunnel has adequate capacity for much more frequent services than at present: down to about every three minutes each way.

Ballance Street might be a good alternative to Whitmore Street. At Customhouse Quay their junctions could be treated as a single junction for traffic signalling purposes.

Constable Street is perfectly adequate for trams, despite the narrow street, with at least six options practical. Buses need more width and might be trickier.

    Mode Choices

Mode choices are bus, tram or possibly BRT. The best solution is usually the mode that best matches the characteristics of the chosen route:

Bus – Suitable wherever a capacity of 60 buses an hour is sufficient. Passenger capacity is roughly 2500 passengers an hour, or a maximum of roughly 5000 pass/hr using double-articulated bendy-buses.

Tram – Suitable wherever a capacity of 40 trams/hr is sufficient (the difference is because trams cannot be allowed to queue at stops or traffic signals: they are too long and obstruct junctions). Passenger capacity in Wellington is roughly 5000–10 000 pass/hr, depending on seating arrangement and tram dimensions). Trams are needed for Option 1.

BRT – No advantage over buses unless the layout is high-capacity and designed to a recognised BRT standard, with two lanes in each direction (Options 3 or 4) and adequate junction capacity. Some junctions will probably need flyovers. Passenger capacity in the right conditions is even greater than a tramway but this looks improbable in Wellington.

In principle there is no objection to BRT minimizing transfers by fanning out onto conventional bus routes beyond a BRT interchange. The problem is managing the side-effects:

  •  Suburban routes suitable for articulated buses, including turning and layover space at the interchange and outer terminus.
  •   Timetable management to ensure buses consistently return to the BRT section at the right time, to maintain a consistent stream of peak-hour buses.

If this is done there would be a case for having a spare bus at the interchange, with driver, to substitute for a delayed bus on the BRT route, when needed.

But whatever is done will need new thinking.

Kerry Wood is a retired Wellington engineer with a long-standing interest in transport matters.  This article is a follow-up to his earlier article (16 September) which gave examples of Wellington’s chaotic public transport policy.