Making sensible mode decisions

One thing the light-rail debate over the past couple of weeks has re-emphasised is there remains a lot of confusion about the different modes that can be used for rapid transit and also how we should make decisions about what mode to use for specific routes. Firstly, I think it’s important to put these discussions in the Auckland context and rapid transit is described in the 2018 ATAP as:

Rapid transit forms the backbone of Auckland’s public transport network, providing fast, frequent, high capacity services along corridors that are separated from general traffic and therefore unaffected by road congestion.

When thinking about the different modes there’s always plenty of debate to be had over different terminologies but there are generally four types of system we are discussing:

• Bus rapid transit – like the Northern Busway. I also classify trackless trams as one form of bus rapid transit as essentially they are just “big buses”.
• Light rail – like what was originally proposed for the City Centre to Mangere corridor, what’s being built in Sydney and what has been built in Seattle in recent years.
• Light-metro, like Vancouver’s SkyTrain – with modern versions usually being driverless. Wikipedia sometimes calls these “Medium capacity rail systems“.
• Heavy rail, which can be either a system fully dedicated to passengers like the New York Subway, or run on tracks that are shared with freight services like Auckland’s current rail system.

Each of those scale in terms of capacity, complexity and cost with the key differentiators being the size/type of vehicle and the extent of the priority they have.

For quite some time now, Auckland has had a broadly agreed “rapid transit network” This was outlined in the 2016 Auckland Transport Alignment Project – although at the time it was called the “Strategic Public Transport Network”. This network of different corridors shows where high quality, high capacity public transport is required to make Auckland work successfully over the next few decades. There are no big surprises here – essentially the map links up all the major centres of Auckland and ensure that most major parts of the city have some proximity to rapid transit (accessing by walking, cycling or feeder bus).

Aside from the rail network (and the Devonport Ferry) this map didn’t specify mode for the other corridors (although one could assume that they were generally thought of as “not heavy rail”). ATAP also set some direction about how mode decisions should be made through future work:

Mode choice for strategic network improvements should be driven by capacity requirements to meet forecast demand, integration with the wider network and achieving value for money.

These three factors – demand, network integration and value for money – provide some useful guidance for how sensible mode choice decisions should be made.

The 2018 version of ATAP was pretty similar to this network, although it did provide some guidance about “likely future mode” of the different corridors – matching the proposals in our own Congestion Free Network to use light-rail for the North Shore and Northwest corridors:

There’s a bit of discussion in the ATAP document around the preference for light-rail for the Northwest and North Shore corridors – mainly focusing on the point that light-rail provides a higher level of capacity than bus and therefore better meets future demand. It’s also noteworthy that rail of some form or another is proposed for all the corridors feeding into the city centre – with bus rapid transit focusing on “cross town” routes. This is most likely due to the higher demand on these ‘radial corridors’ and the lack of space in the city centre for terminating large numbers of buses.

What the map above illustrates is that a rapid transit network can be made up of a number of different modes, because the characteristics of the corridors are all pretty different from each other. For example, the demand from Akoranga and the City Centre is going to be much higher than say the New Lynn to Onehunga line. Put simply, many more people want and need to travel on some of these corridors than others.

Another important difference between the corridors is about space and level of priority – how easy or difficult it is to add a dedicated high capacity public transport corridor. The distinction here comes between the modes that can run ‘on street’ – like light-rail or bus rapid transit – and those that cannot, such as light-metro and heavy rail. Modes that are able to run on street allow for a corridor to be delivered at surface level where it would otherwise be very difficult/expensive to provide for e.g. the Northern Busway buses use the motorway and local roads when outside of the dedicated busway. This can be especially useful for stageability as if you go for the fully ‘off-street’ modes then the corridor needs to be either:

• At surface level in their own corridor (expensive and next to a motorway or cutting a swathe through an area by demolishing houses).
• Elevated (very expensive, ugly and possibly un-consentable in some areas)
• Underground (extremely expensive)

You’ll notice that many of the future rapid transit corridors are next to motorways but others, such as Dominion Rd, aren’t.

Pulling together some of the key factors, each mode has its strengths and weaknesses. I’ve elaborated on this in the table below:

Generally cost increases as you move from left to right, but you are also able to carry more people more efficiently as well. Importantly, speed of the corridor doesn’t really depend on mode, instead depending on the right-of-way and how many stations there are. Some of the most important distinctions are:

• Because light-rail can carry more than four times as many people per vehicle as bus rapid transit, you need at least four times as many buses (and drivers) to meet a certain level of demand as you’ll need light-rail vehicles. At low levels of demand this is fine, but as demand grows it can become increasingly difficult to maintain an efficient bus system – especially where all these buses end up terminating or picking up and dropping off large numbers of passengers.

• Because light-metro cannot run on street, it requires much more expensive infrastructure than light-rail if there is no existing corridor to run on. Vancouver’s early Skytrain lines used an old railway tunnel through downtown but outside of that is generally built on high viaducts (hence the name). Later lines have required extensive tunnelling that was costly and disruptive, as well as having short station platforms that will be extremely difficult to extend. Driverless operation does mean that these systems usually run very cost-effectively, which over time will offset the highest construction costs to some extent.
• Because heavy rail has much more demanding geometry than light-metro – a very low tolerance for sharp curves or steep gradients – it has the highest construction costs of all. But heavy rail also provides the highest amount of corridor capacity as very very large trains can be run. The Yamanote Line in Tokyo – for example – carries over 4 million journeys a day on trains that are 11 cars long and come every 2 minutes.

Putting all this together we can pull out some pretty basic questions that help point towards the right mode:

1. Can the corridor easily accommodate an “off street” mode, or do sections of it require challenging tunnelling or elevated sections?
2. Is demand high enough that bus rapid transit would not work effectively (without massive infrastructure like underground city centre stations)?
3. Is demand high enough that surface level light-rail would struggle to operate effectively, and tunnelling or elevated infrastructure is justified?
4. Is demand high enough to require heavy rail and justify its cost?

So let’s apply this approach to the City Centre to Mangere corridor and see where we get.

• The southern part of the corridor, between Mt Roskill and the Airport, can accommodate an ‘off street mode’ next to the motorway. There are some tricky sections, like getting through Onehunga, over the Mangere inlet and accessing the Airport, but generally an off-street mode is feasible in this section. The northern part of the corridor is a different story, with no available space between Mt Roskill and the city centre to ‘easily’ accommodate an off-street mode.
• The demand seems like it’s high enough for bus rapid transit to not work effectively, especially at the city centre end of the route. A few years back NZTA explored this option in detail before eventually agreeing with Auckland Transport that light-rail was a better long-term solution.
• It seems unlikely that demand is high enough for surface level light-rail to struggle, and therefore conversely for major tunnelling or elevated infrastructure to be justified. Light-rail can carry up to 500 people per vehicle and can run at frequencies of one vehicle every 2-2.5 minutes. This means around 10,000-13,000 people per hour can be carried – substantially higher than any existing rapid transit corridor in Auckland. Furthermore, while Dominion Road is one of Auckland’s busiest bus routes, it still carries far fewer people than the Northern busway or the main railway lines.
• In the long-run, especially if the ATAP rapid transit network map becomes a reality and all three light-rail lines plug into Queen Street, then it’s feasible to see light-rail no longer working and there might be a need for a different mode. But that feels like a 2050 issue that can be solved in a while variety of ways (more light-rail corridors through the city or keeping more North Shore bus services using the Harbour Bridge like we propose in the Congestion Free Network).

Many media articles over the past couple of weeks highlight that the Super Fund proposal seems like a ‘light metro’ option. This means that it will need to be tunnelled or elevated between Mt Roskill and the city, with some reports suggesting they’re looking at an elevated track above Dominion Road and then a tunnel under Queen Street. Given it’s pretty impossible to see elevated rail down the middle of a main street being consented, this probably means the whole Mt Roskill to city section would need to be tunnelled – which is over 7 kilometres! It seems unlikely demand will be high enough for this to be justified, even if the operating costs are reduced by having driverless trains.

Therefore, we come back to light-rail being the most sensible mode for the corridor. As has been pretty clear for many years now.

Share this