Recent Changes
Recent Changes · Search:


Light Rail
A semi-rapid transit system, segregated from general traffic as much as possible, faster, with wide spacing between stops and priority at intersections.
A local system, mainly running with general traffic, slower, with closely-spaced stops. Wellington’s original tram system was a streetcar system.
This is a generic term for a light rail vehicle or a streetcar vehicle.

FIT is proposing a light rail system for Wellington, not a streetcar system.


PDF settings (show)


The length of a first line should be at least 5–8 km (European rule of thumb; Wellington Railway Station to the Hospital is 4km—too short for a first line). Time-savings on a short run cannot offset transfer delays or generate substantial revenue.

Light rail is generally economically viable if peak-hour patronage on the busiest section is over 3000 passengers per hour (European rule of thumb).

The optimum (not operational) capacity of a quality bus route is about 60 buses per hour (2011 Wellington Bus Review), which for conventional buses gives a capacity of about 3000 passengers per hour.

A high-capacity public transport route should be aligned with land use planning to maximise patronage. This suggests an initial Railway Station, Wellington Hospital, Newtown, Kilbirnie, Wellington Airport route. Apart from the airport and Newtown, this is the same as the PTSS route.

Other viable and feasible routes are likely, such as Miramar, Island Bay, Karori and Lower Hutt.

Aim for an average speed of 30 km/hr, giving a travel time from the Railway Station to Airport of 17 minutes or less. Allowing for a short layover at each end gives a round trip time of 40 minutes.

Good quality public transport in Wellington is very likely to generate strong patronage growth, as shown by international experience and Auckland’s North Shore Busway and electrified passenger rail (and Waikanae rail?). (Passenger rail patronage in Auckland was 5 million/yr in 2006, 10 million in 2011 and 18 million in 2016: transportblog Tuesday, 10 January 2017)

The estimated potential peak demand for light rail on a line from the Railway Station to the Airport is up to 8000 pass/hr, based on:

  • approximate population along the corridor: 50,000
  • 50% of population are commuters (adults and students): 25,000
  • light rail attracts 50% of commuters: 12,500
  • a 2-hour commuting peak: 6250 pass/hour
  • plus 20% for the airport and hospital: 7500 pass/hour
  • a further 3000 pass/hour on buses during peak periods

If the peak demand accounts for half the total daily demand, this gives a potential ridership of about 60,000 passengers per day.

A crucial problem in Wellington is a heavily overloaded two-lane bus route through the central city, with 130–140 buses per hour, carrying up to 6000 pass/hr. Dominion Road in Auckland carries a peak of 120 buses/hr and Auckland Transport deems this sufficient to justify light rail.

Bus Rapid Transit in Wellington would need four lanes. Light rail would need only two, but light rail without buses would demand too many transfers, so no-one is suggesting this. Light rail and buses need separate routes, with different stop capacity and frequency, to avoid the slower buses holding up the faster trams. This means a good quality rapid transit solution needs four lanes through the central city—“Four lanes to the trains.”

Costs can be kept to a reasonable minimum using a no-frills approach, with all intersections at-grade.

Practical operating capacity for an at-grade light rail system is about 20–24 trams/hr (2½ or 3 min headway). Higher tram frequencies potentially create traffic congestion at busy intersections and planners are tempted to hold up the trams so the cars can clear.

Ultimate route capacity is 40 tram/hr with grade separation (1½ min headway—fewer than for buses because trams can rarely double-berth at stops). One result is that shared bus/tram lanes are rarely acceptable, and never acceptable if the buses stop on the tram tracks: trams have much shorter dwell times at stops.

An extensive tram-train system (also running on KiwiRail tracks) is unlikely to be cost-effective in Wellington because of greater costs than light rail for limited additional benefits (tram-train would let some passengers avoid a transfer between modes at the Railway Station).

Specifications for track and trams “the same as Auckland” if at all possible, probably with standard gauge track (1435 mm). A likely deviation is trams 2.4 m wide, instead of the more usual 2.65 m, probably necessary in Wellington’s narrow streets (such as Hunter, Constable, and Riddiford).

Where necessary, remove on-street car parks to make room for light rail. In congested inner-city streets, on-street car parks are a low-value use of scarce space.


Modern low-floor trams with level boarding from low platforms (~350 mm). This effectively means tram-train use of these vehicles would not be possible, as train systems are high platform.

Modular construction for cost-effectiveness (Siemens Avenio used as a representative example). The vehicles are modular and can be purchased in short sections, then lengthened cheaply as patronage grows by adding new modules. This is cheaper than coupling multiple trams together because the most costly components are the cabs and control system (Avenio is available in 2 to 8 sections, 18 to 72 m).

Capacity about 420 passengers for a 2.4 m wide or 470 passengers for a 2.65 m wide, 7-section vehicle (63 m long). Siemens give the limiting capacity (football match etc) as about 30% higher than this, or 540. Standing room is 4 persons per m2. Maximum passengers per hour for 2.4 m tram shows the estimated route capacity for different headways.

For an estimated initial peak demand of 4000 pass/hr, plan for a 5–6 minute headway during the morning and afternoon peaks, 10–12 minute headway off-peak and on weekends; or preferably use shorter vehicles off-peak and keep the peak period frequency from 7 am to 7 pm every day.

A 2.4 m 7-section tram can comfortably carry 10,000 pass/hr at a 2½ min headway. Replacing 2.4 m width trams with 2.65 m trams increases capacity to almost 11,300 pass/hr.

Optimise vehicle performance settings for fairly steep gradients (6–8%).

Vehicle numbers are (round trip + layover + recovery) times for each scheduled trip, plus 10% spare, plus an additional tram for orders of less than 12 trams.

Autonomous (self-driving) trams may well be commonplace by the time light rail comes to fruition in Wellington. Such vehicles will dramatically lower the marginal cost of providing high frequency off-peak service, thereby attracting new passengers.

Route capacity for a range of tram sizes and headways
LengthCapacity3 min6 min12 min2½ min5 min10 min
18 m1092180109054526161308654
27 m17234401720860412820641032
36 m236472023601180566428321416
45 m296592029601480714035521776
54 m362724036201810868843442172
63 m42084004200210010,08050402520
72 m48697204860243011,66458322916


Initial route from the Railway Station to Wellington Hospital, then by Constable St or Wellington Zoo to Kilbirnie and Wellington Airport.

A high priority on maximizing congestion-free running, and average speed.

Stop-spacing generally around 800 m, adjusted to suitable locations at major destinations.

Minimum curve radius 25 m, generally about 28–29 m for the outer track, including end-throw. Approximate curve speed = sqrt(radius) × 3.6 km/hr).

Nominal stop length 63 m (7-section Avenio). With trams 2.4 m wide this gives a comfortable capacity of about 420. For 2.65 m wide trams, the capacity is about 470.

Route capacity is taken as (tram capacity × tram/hr) or nominally 10,000 pass/hr for an at-grade route, ignoring “football match” crush loading capacities (potentially up to 16,800 pass/hr with grade-separated intersections).

Minimum lane width based on UK regulations, which seem conservative, generally 3.1–3.3 m. But the GSDG gives 3.0 m (Auckland assumes 3.5 m). Pole location, parallel parking, and truck wing-mirrors are also issues. See the GSDG.

Minimum side platform clearance (passenger standing area) to any obstructions 1.5 m, use 2.5 m (including 0.5 m for ticketing machines, seating etc. GSDG gives 2.5–3.0 including canopy supports, or more at busy stops. Island platforms an assumed 4.5 m min (1.5 m for shelter, seating, ticketing), although Auckland assumes 3.0 m.

Buses are probably best on the golden mile. The existing stop pattern gives good CBD access for people unable to walk far, and the buildings provide relatively good shelter. Such people will tend to prefer avoiding a transfer, and buses will offer them a slower but more direct service to many destinations.

In contrast, light rail performs best on fast routes with relatively infrequent stops (although stops may be slightly closer together in the city centre and further apart in the suburbs). People wanting fast trips will tend to be willing to walk further to reach a stop (up to 10 minutes).

FIT has considered three options for the best section of route through the CBD:

  • Buses on the golden mile, light rail on the waterfront (likely the seaward side, clear of most turning traffic), with stops at the Railway Station, Frank Kitts Park and perhaps Kumutoto (mid-way between the station and Frank Kitts).
  • A hybrid option. Light rail shares the golden mile with buses on Lambton Quay, with an interchange at Midland Park, then continues down Lambton Q to Panama St, the waterfront and a stop at Frank Kitts Park. Northbound, light rail travels on Hunter St between the waterfront and Lambton Q. Buses mostly on the golden mile but deviating to the east between Hunter and Johnston Streets:
    • southbound, Lambton Q, Johnston St, Featherston St, Panama St, Customhouse Q, Willis St
    • northbound, Lambton Q.
  • Buses on the golden mile, light rail “somewhere else,” taking advantage of space made available from narrowing existing traffic lanes or removing on-street parking. No specific route has been identified.

Characteristics of light rail track envisaged:

  • Fully reserved route off-road (Johnsonville Line, Mt Albert Tunnel).
  • On-street, centre-running is preferred (gutter running tends to attract delays from turning or parked vehicles).
  • Near-absolute priority at traffic signals, anywhere. This needs two priority signals: priority in x seconds followed roughly x seconds later by priority now. This gives the traffic signal computer time to organize signal phases effectively.
  • May require barrier arms at street crossings (cf. the rail crossings on Waterloo Quay).
  • Grassed track where suitable (practical where no other vehicles use the light rail lane).
  • Semi-reserved route separated by raised trackbed (UK: kerbs would be cheaper), roughened to further discourage motor vehicles (Constable St).
  • Motor vehicles permitted when turning right at traffic signals at some intersections, timed to clear the traffic queue before a tram arrives (Constable St, Riddiford St).
  • Controlled right turns needed at most intersections, making other vehicles go around the block (G and Q turns) rather than turning directly.
  • Motor vehicles permitted to cross tracks to access private property where this is unavoidable (Constable, Riddiford, Wallace Sts).
  • Some sharing with local motor vehicles but barriers to prevent through motor traffic (Wallace St, although removing on-street parking to make room for dedicated light rail lanes would be preferable).

Note that a lot of the FIT route has not been planned in detail and may change with more information. Detailed design includes identifying route bottlenecks that would slow the system down (increase the travel time), and find ways to eliminate them.

Possible future routes include:

  • Miramar (requiring peninsular passengers to transfer at Kilbirnie may discourage patronage from what is a high-growth area).
  • Karori (large potential demand, difficult and costly, but the alternatives may well be worse).
  • Island Bay (Rintoul St tricky and residents appear reluctant to reallocate road space from cars to more efficient modes).
  • Ferries and Hutt Valley (explore route to Petone, then Melling Line and to Waterloo).
  • Johnsonville long-term (when the existing line runs out of capacity).


A 5–6 minute headway on a route optimised for speed will minimise the perceived transfer penalty.

The golden mile needs good hubs for transfers at each end.

The Railway Station hub, for buses and trams, will probably go above and across the tracks, with escalators down to rail and street level. Configuration will depend on whether light rail runs on the waterfront, via Stout St to Lambton Quay at Midland Park, or from the existing bus hub to Lambton Quay.

The southern hub could be at Te Aro Park (Taranaki St route) or Courtenay Pl (Basin Reserve route).

Midland Park is a possible intermediate hub.

A Kilbirnie hub for transfers to and from the Miramar Peninsular.

Timetables must be arranged to minimise transfer times (this only becomes an issue off-peak, when the headway is greater than 5 mins).

Page last modified 14 September 2017 at 11:36 AM