Wellington Scoop

Three years of chaos: Wellington’s public transport plan, and how to fix it

by Kerry Wood
Since the Transport Agency’s Basin Reserve flyover plan was cancelled, there’ve been claims that it has thrown Greater Wellington’s Public Transport Plan into chaos. In fact the plan has been in chaos much longer: at least three years.

The 2014 plan has been approved but is in reality a train-wreck. Dramatic improvements in public transport usually need institutional change, which is now clearly essential in Wellington.

In recent years the Regional Council has made and commissioned a series of very bad public transport studies. Serious errors, noted in submissions, have been ignored. Probably the worst were the trolleybus debacle and the final stage of the Spine Study (AECOM/GWRC, 2013), both so bad that reasonable conclusions are impossible.

The 2011 Bus Review

A very good study of public transport in Wellington was the 2011 Bus Review (MRCagney). An important recommendation was an upper limit of 60 buses an hour each way on the golden mile. The reasoning was clearly explained (page 54). In summary, minor bus stop delays of up to about a minute are routine and, with only a single lane each way, buses must leave a stop in the order they arrived. Delays longer than the time-interval between buses begin to accumulate. Buses leave the stop in a queue, and with little bus priority at traffic signals, short queues are assembled into the cascading delays familiar in Wellington.

This is why Bus Rapid Transit usually has four lanes at stops and a high priority at junctions. If this is impractical a rapid-transit junction needs a flyover.

The form of rapid transit proposed for Wellington is two-lane and at-grade, as at present, and faces the same 60 buses per hour limit as the existing system. The Bus Review proposed a maximum 73 buses per hour on the corridor, to be reduced to 57 buses an hour by using a secondary route for peak-only services, on the busiest section: Taranaki Street to Whitmore Street. This was less than half the maximum at the time, 133 buses an hour.

It didn’t happen. Critical features of the MRCagney study proved unpopular, were dropped and the study benefits were lost. The unpopular features were:

• Nine interchanges (including Wellington Hospital; Kilbirnie and Johnsonville.
• The secondary route, between Whitmore St and Taranaki St. Officers made it worse before consultation: from Kent and Cambridge Terraces to Whitmore Street.
• Some golden mile buses to be diverted onto The Terrace.

No later studies have achieved less than 60 buses an hour, or delivered credible alternatives for a single lane in each direction.

The Bus Review even allowed for retaining the trolley buses, which would have given valuable options for the coming revolution in electric vehicles.

Inadequate Project Information

Decisions involving hundreds of millions of dollars were taken without detailed route plans. The objective was only preliminary design, of course, but even that has to be taken far enough to confirm feasibility. That means being confident that everything will fit. Working without drawings is an extraordinarily high-risk approach, contrary to the terms of reference, and it created problems. The only justification was (oral, briefing meeting on 13 August 2013):

…because there is not a sufficient level of work done.

This was a better reason for properly completing the project than withholding publication.

A complex project like this needs whole-of-route plans at a scale of say 1:1000 or larger, available well before final decision-making. They are needed to assess route curvature and speed (for modelling); to check that everything will fit and clearances are safe (including trams and bendy-buses); to check junction layouts (fit and modelling capacity); to check stop lengths and layouts; to check provision for all road uses; to cost any land purchases; to plan rerouting of underground services; and to provide public information.

Because there are no plans there is no indication, for example, of how Bus Rapid Transit or trams will negotiate the Basin Reserve (the consultant doesn’t know either), or turn around at Riddiford Street (it needs buses with doors on both sides). Another result is this gem (consultant’s report, p 39):

Concept designs and vehicle tracking curves have confirmed that there is a feasible route that can be implemented but it will be necessary at a detailed design stage to understand how Light Rail Transit will transition to/from the Basin Reserve.

The Train Wreck

A great many factors contributed to the train-wreck, not all within Regional Council control:

Outdated thinking

New Zealand’s transport thinking is stuck in the 1960s, with action on climate change imperative and many other countries taking a more balanced approach. Reducing car numbers is more a solution than a problem, with massive gains in public health (exercise, pollution, crashes), road capacity (all alternatives to cars use road space much more efficiently), cost (imported petrol, vehicles and spares) and quality of life.

Bad modelling

The Spine Study used road-traffic models adapted to study public transport. This is standard Transport Agency practice but it’s always risky if quality public transport is the objective. The results were nonsense:

• Patronage growth was supposedly well below population growth, compared with 18% per year for Britomart.
• Modelled interchange delays were an average eight minutes in the morning peak: this will be one reason why predicted patronage was so low. No other times of day were modelled (a weakness of motor-traffic thinking, not the model) but passengers would clearly wait longer. And eight minutes is already well above best practice, which is 2–5 minutes. The figures suggest fairly frequent delays of over 20 minutes — say once a week for many commuters — because the modelling assumed no synchronizing of timetables.
• One extraordinarily bad outcome is an interchange located 300 m south of Wellington Hospital. Main interchanges should be at or close to main passenger destinations, and a regional hospital is a classic example. The Brisbane Busway has a stop inside the main hospital. A public transport model would have ‘assigned’ people flows to major destinations like this, but traffic models only assign over broad areas. Important steps in modelling public transport were left out.

In the latest PTP the interchange serves a different function and has been moved 500 metres further north, to another wrong place.

Routes and timetables were not modelled and not optimised, and there was no real attempt to manage bus numbers.

The modellers estimated bus speeds simply by adopting average speeds from supposedly comparable systems. They did not check that route design was comparable and could be negotiated at the chosen speed (how could they, with no route plans?).

Conflicting objectives

Two standard objectives of public transport are patronage and social service: competing with cars and helping those without cars. An often-ignored problem is that these objectives pull in opposite directions. On the one hand frequent, direct routes for those who value their time, and on the other hand infrequent, indirect routes for those having no other choices. The only clear exception in Wellington is the MRCagney Bus Review, where the conflicts are recognised and managed.


The Spine Study in particular has seen large-scale muddle. The final report was three months late, and when it emerged it showed every sign of a desperate rush. Quality control suffered. Some documents show alarmingly late amendment dates and other control sheets are blank. How well were these studies integrated into the project, and what undetected errors do they contain?

There are large inconsistencies in the patronage estimates (consultants’ report, Tables 6, 7 & 31), as well as inconsistent units.

The Regional Transport Committee received officers’ route option proposals for the final evaluation stage on 13 February 2013, barely ten weeks before the final report was due. Nobody acknowledged that the information was late or would delay final completion: hence the rush. The chosen option was very odd, never satisfactorily explained, and was itself rushed. Three versions of the plan all differ; one presented to Councillors, another in the consultants’ report and a third in the Regional Council’s own Modelling Report.

The trolley bus studies were no better: one consultant was given information so confused that it merged two types of trolleybus into one: the vehicle described would never have been allowed on public streets. Trolleybus upgrade costs are so bad that the cheapest option is unknown, but almost certainly not the outcome chosen.

A recurring double-muddle is interchanges, transfers and route capacity.

Interchanges were recommended in the Opus operational review (2009), locations chosen in the Bus Review (2011), then rejected. But they remained in vogue: the 2011 PTP shows nine interchanges, region-wide, including Wellington Railway Station, Wellington Hospital and Kilbirnie. Interchange studies were explicit in the Spine Study terms of reference, retained in the final report (2013) and since dumped. BRT will avoid the need for interchanges by using buses, which can continue beyond the interchange and don’t need one.

MRCagney notes that interchanges are ‘politically difficult.’

Route capacity is the reverse of the interchange coin: you may be able to ignore one, but not both. In 2013, MRCagney proposed a secondary bus route for the most congested part of the golden mile. It was rejected. Now the Spine Study has a ‘secondary spine’ with some Cheshire Cat-like characteristics. It is in the Modelling Report, with a small-scale plan showing either alternative routes or separate routes for each direction; it is mentioned in passing by the consultant; but it doesn’t appear at all in planning by WCC, who will have to implement it.

The Basin Reserve & Mt Victoria Tunnel

The original spine route was from the Railway Station to Wellington Hospital, but with an extension to Kilbirnie mooted for the final stage. The final options appeared as a split route, dividing at the Basin Reserve and running to Kilbirnie and the Hospital. It requires additional capital and operating costs and presents a series of operating disadvantages, all for no obvious benefit. Despite these problems there was no move to reconsider any single-route option.

The proposed Mt Victoria layout was twin two-lane road tunnels. Buses would have shared them with general traffic but trams would have needed a separate tunnel for ‘safety reasons’. Buses have fuel tanks but trams have circuit breakers, on both trams and overhead, just like trolleybuses.

An intelligent Martian might conclude that the primary objective was ensuring an unaffordable tram option, and might note other highly undesirable features:

• Public transport provision at the Basin Reserve needs a reserved route where there is little road space, which might have helped to justify a ‘Basin Bridge.’
• Buses will be delayed in the road tunnels, making them less competitive with private vehicles. Like most urban road tunnels, the layout is a sitting duck for rapidly increasing congestion because of ‘triple convergence.’
• Low-density, residential Hataitai is over-provided with services but under-provided with access, on the edge of the catchment area and with city-bound buses on the wrong side of SH 1.
• High density, mixed-use Newtown is under-provided with services and access. The spine route would terminate in Riddiford St, just north of Constable St, at a very difficult site for an interchange. With no drawings, its working is opaque; many passengers would have to interchange even for local trips.
• The Basin Reserve would be a de-facto interchange, regardless of design intent, and transferring would presumably involve a busy, possibly dangerous, road crossing. Again, no drawings.
• Passengers going to or from Wellington Hospital from either Newtown or the eastern suburbs would face an interchange at either Constable Street or the Basin Reserve.

Back in the real world, the Modelling Report gives the reason for the odd layout: the Hataitai route is faster than Constable and Riddiford Streets. ‘Up to six minutes’ is claimed (p 38) but the table referred to shows 4 minutes and the data accuracy is strictly wet-finger. With no drawings, speed estimates are bound to be dubious A more reasonable claim would be ‘probably faster outside peak hours.’ The Modelling Report is cautious here (p 38):

If looking solely at potential travel times and excluding other factors such as cost, a rapid transit route from Kilbirnie to Courtenay Place via Mt Victoria tunnel would provide considerably more benefits for passengers from the eastern suburbs than a route from Kilbirnie to Courtenay Place via Newtown.

This seems to be the only justification for a split route: a minor time advantage, regardless of cost, the operating problems noted by the Martian or damage to Wellington’s Town Belt.

Institutional Change

Something needs to be done, and some of these points might help:

• Hire an independent mentor, advising officers but reporting directly to councillors. The mentor would be a consultant, hired to keep an eye on what is going on, comment on reports and above all make suggestions. The ideal might be a recently retired european engineer with wide public transport operational experience.

• Use evaluation software designed for public transport (hopefully recommended by the mentor).

• Have a management consultant review earlier studies and recommend institutional changes to manage the identified errors.

• Focus on quality of results, not unrealistic deadlines.

• Insist on clear terms of reference and clearly compliant reports.

• Encourage all staff and councillors to use public transport regularly. It might usefully be a requirement for senior managers.

• Send all Regional Council staff on a writing course. Councillors cannot be expected to approve unclear reports and should reject them. What, for example, can busy councillors be expected to make of this (quoted by PGGM on Wellington.Scoop, 18 August). What did the author think it meant, or the manager who approved it, or was it even approved?

Amend the PT Plan to clarify that the proposed network will not reduce service levels, where the services are well-patronised and that corrections be made in the final plan to make clear that the proposed hours of operation do indeed have no change where this is the case.

• Come to a clear view of which decisions should be made by officers and which by councillors (see next bullet point).

• Buy 50 copies of Jarrett Walker’s book, Human Transit: how clear thinking about public transit can enrich our communities and our lives (2012), for councillors and staff.

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

Also by Kerry Wood: Riddled with errors and contradictions