Friday certainly turned out to be interesting following on from our post from that morning that the NZTA now supported Light Rail along Dominion Rd and to the Airport in the future. The council pulled down the document our post was based on, which was a waste of time given we’d already published our post. Later in the day, Transport Minister Simon Bridges confirmed it all, followed by a joint release from Auckland Transport and the NZTA. While it’s great that we’ve now got the government saying that a rail connection to the airport will happen, the idea of it being 30 years away has been met with ridicule from a lot of people.
One of the first things I found interesting was the different way in which the agencies described the decision compared to the one from the minister. Most notably Bridges seemed to make more of an effort to talk up the recent Advanced Bus Solution study (ABS) and the ability of buses to cope with demand for the next 30 years.
The study found an advanced bus option could provide a credible solution over the next 30 years that could progress from the current bus-based system to a long term solution.
The study also demonstrated the advanced bus option has the potential to deliver on forecast demand depending on the rate of growth of the city.
A lot of the timing so far seems to be dependent on just how much capacity can be eeked out of buses before an upgrade to light rail is needed and that’s where the ABS study comes in.
As I mentioned on Friday, the ABS was commissioned by the NZTA following on from the Auckland Transport Alignment Project (ATAP) which only referred to some key public transport projects as ‘Mass Transit’. The NZTA describe the objective of the ABS as coming up with the preferred bus based option for between the city and the Airport and serving the central isthmus, utilising new and emerging technologies, and that is able to be compared with some of the previous studies that have ultimately recommended light rail as the solution.
The ABS study was also released publicly on Friday finally giving us the opportunity to look through it. As mentioned on Friday, the recommended the same route down Dominion Rd then along SH20 as the previous studies that recommended light rail have. It also recommends a rapid transit corridor for buses of the same width and location within the road corridor as is needed for light rail. This is why the announcements on Friday afternoon talked about getting on with route protection. Unfortunately, it didn’t take long to find some gaping flaws and major concerns.
The ABS also has another option that uses a second corridor between K Rd and Onehunga, going via Newmarket and Manukau Rd. For the purposes of this post though I’ll stick to just the Dominion Rd corridor.
To service the route from the airport to the city the ABS suggests 4 separate service patterns that they say will need to operate. Each service would operate at four minute frequencies and calls for a mix of new, 18m long articulated buses and double-deckers. The four routes are
A – an 18m articulated ‘all stops’ service from Mt Roskill Junction to Mt Roskill Junction via Britomart
B – an 18m articulated ‘all stops’ service from the Airport to Airport via Britomart
C & D – two double-decker ‘express’ services from the Airport to Airport via Britomart, only stopping at the express ABS stations
Four different services at four minute frequencies, that’s 60 buses an hour down Dominion Rd and Queen St and something that would present a number of challenges including:
- Express buses would need to be able to pass all stops buses if they are to have any time advantages. Unlike the Northern Busway stations which has four lanes at stations so buses travelling straight through can pass. The ABS envisages that the express buses will pass all stops buses by driving onto the other side of the busway. Buses at high frequencies in each direction weaving in and out of lanes sounds like a safety case nightmare.
- There are many key intersections along the route. Major intersections, such as at Balmoral Rd, can have signal phase cycles of over 2 minutes meaning that these buses are ultimately going to bunch up and become less reliable. To help counter this, the ABS assumes there will be significant levels of signal pre-emption to prevent buses from falling behind schedule. Signal pre-emption can work well when frequencies are a bit lower, like as is proposed with light rail, but the frequencies suggested for the ABS would create significantly less opportunities for alternative traffic movements, including cross town buses.
- This is even more critical in the City Centre, especially where it intersects with Wellesley St, which is expected in the future to be a 4-lane busway moving over 140 buses an hour. The ABS glosses over this and only suggests other isthmus buses could be turned into feeders for the Dominion Rd services.
- Things get even more tricky at the Northern end of the route where the ABS expects buses will turn around by looping around Customs, Lower Albert, Quay and Commerce Streets before heading back down Queen St. As a reminder, they’re suggesting 60 buses and hour would be doing this, half of which are double-deckers and the other half longer, articulated buses. It’s unclear what impact this would have on buses like the Northern Express which uses Lower Albert St or the busy Mt Eden Rd buses which start from Commerce St. It would also require retaining at least four lanes for traffic on Quay St, working against council plans to make the area more pedestrian friendly.
- The ABS also suggests using Commerce St and Quay St as a place where services could layover between runs. As above this would impact on Mt Eden Rd buses and it would also appear to block access to the shops directly above Britomart.
- Speaking of pedestrians, compared to the light rail option, with 60 buses and hour in each direction it will make it much harder for pedestrians to cross roads. This would be both in the city centre or on Dominion Rd as on average there will always be a bus in some direction 30 seconds away.
For a lot of the points above, they seem to have just explained away the challenges by suggesting technology will solve it. Some of what they’re suggesting sounds reasonable but I’m not aware of having been fully implemented anywhere in the world yet.
All up, it seems the authors of this report, and their clients at the NZTA, completely forgot/ignored ATAP when it said we can’t continue to stuff more buses in the city centre and proposes to solve problems by stuffing more, only slightly larger buses in the city centre.
Another important element in all of this is the cost. The ABS Appendix suggests that the standard Dominion Rd option, with contingency, would cost $1.132 billion to build. On top of that there would need to be an additional $70 million for the buses – 47 articulated buses at $800,000 each and 51 double-deckers at $600,000 each. That brings the total to about $1.2 billion.
That is cheaper than Light Rail, which from the city to the airport is expected to exceed $2.4 billion but it’s also important to take into account the operating costs(OPEX) – and 60 buses an hour would require a significant number of drivers compared to LRT options.
Interestingly the ABS study is only assessed over a 20-year period while most studies these days cover a 40 year timeframe. Even so, the OPEX is significant at almost $1.5 billion.
By comparison, the SMART study that previously recommended light rail to the airport suggests OPEX costs of $790 million but that was assessed over 40 years. Given SMART only looked from Mt Roskill to the Airport, assuming those costs increased by 40% to account for getting to the city, it will likely be cheaper over the long run to build LRT from the start.
To top it off, light rail could have considerably larger capacity. AT’s plans for light rail would see two 33m long vehicles carrying 225 people combined to carry 450 people vehicle every four minutes in each direction which is higher than the ABS option. However, many cities with modern light rail systems, like Seattle, are starting to go further and so it’s possible we could ultimately end up with 99m long vehicles carrying 675 people. That’s potentially over 10,000 people an hour, 40% more than the ABS option.
Overall this seems to be a report that’s high on theoretical solutions but that is unworkable in the real world. I guess that’s why the AT/NZTA boards say further work is needed to determine the best point to transition from bus to light rail. There are of course some sensible improvements that can and should be made now to improve buses on this route but it would be foolish to spend billions on CAPEX and OPEX just to have to do much of it again to upgrade to light rail in the future.
A briefing to Councillors (from page 251) included in an attachment for the Council’s Planning Committee meeting for next week says that both the NZTA and Auckland Transport boards have now agreed on a way forward for the city to airport corridor. And that decision will see the route upgraded to light rail, eventually.
The Auckland Transport Alignment Project (ATAP) shied away from saying light rail was needed and instead just referred to a number of strategic public transport routes as “Mass Transit”. We know that following ATAP, the NZTA commissioned a study to see if buses could meet the expected demand on the City to Airport route instead of Auckland Transport’s preferred light rail option. This was called the Advanced Bus Study (ABS) – as it happens I have an OIA request in for this report which was due back today but the NZTA have extended it by another 3-weeks. They even went as far as getting overseas consultants to do the study so it was completely independent to AT’s previous investigations on the issue.
The presentation states that like other recent studies, the ABS confirmed the proposed Queen St/Dominion/SH20/SH20A route which has helped “provide confidence and a degree of investment certainty” that they should get on with protecting it for mass transit. They also say it highlighted the need for better PT access from the East.
Here is a map of the route planned by the ABS showing it almost identical to the route AT planned for light rail.
The agreed way forward appears to be sensible. I read it as essentially being to make a range of short term improvements both north to the city and east to Manukau to help buses move around easier, which would likely take some of the current pressure off, while working to protect the route to the city for an eventual upgrade to light rail.
The short term improvements include infrastructure changes and it appears, additional services.
The changes to services are shown on the map below and would entail a number of new routes or changes to existing routes.
One interesting slide shows what the unitary plan allows for development in a rough walking distance around the proposed stations of both the North and East route. One aspect that stands out is the amount of almost white (single house zone) around some of the Dominion Rd stations. There is some mixed use density allowed close to Dominion Rd that is hard to see in these maps but the density does drop of quickly.
Light rail certainly feels like it’s going through the same sort of obstructive process as the CRL did before the government finally agreed to help pay for it and that it’s needed now. Let’s hope we don’t have to wait as long as the CRL did to get to that stage. The biggest unknown of course is the government who have been hostile to the project so far. I can’t understand what they would lose by supporting it, even if it was built in a decade or more, as the numerous reports back up the need for light rail and it’s a project that would be popular with voters.
Overall, it’s good to see there’s progress being made on this project and even better news that modern light rail continues to be the plan.
At the Auckland Transport Board meeting earlier this week, I did a presentation on behalf of the Campaign for Better Transport on airport rail, making the following points in a “one-pager” to the Board.
1. In our view the Jacobs “SMART Indicative Business Case | PDF” report underestimates the potential catchment of heavy rail, we assume because of the arbitrary requirement for a single seat journey to the airport.
On this point, the following from p.83 of the report shows the catchment for the heavy rail option. It clearly does miss out stations on the Western line, as well as the yet-to-be-built K Rd and Parnell stations.
2. We consider that some of the costs of heavy rail attributed to the airport heavy rail option will most likely be incurred anyway – in particular work required around level crossings.
3. We consider there is a high risk that the predicted Dominion Road journey times for light rail are overly optimistic, depending on the degree of separation from general traffic.
4. Implementation of either heavy rail or light rail from the north of the Airport is likely to be decades away and very costly.
5. Putting aside the report’s assessment of heavy rail vs light rail, we note that the three key problems identified in the Jacobs report do not have to be addressed by a single solution:
a) Constrained access to the Auckland Airport will limit economic growth and productivity;
b) Limited transport choice undermines liveability and economic prosperity for the Māngere-Ōtāhuhu area; and
c) Unaffordable and inflexible planned transport investment constrains access to the Auckland Airport and surrounding business districts and Māngere-Ōtāhuhu area
As is so often the case with any project, defining the problems you are trying to solve is paramount. The SMART study has some useful points, but it is flawed as it is implicit that a single solution must meet all three problems. By redefining the problem, the Puhinui solution emerges as an option to be considered.
6. We ask the Board to take the same approach as ATAP in measuring transport effectiveness. In the context of Auckland Airport, the measure would be the potential catchment of public transport users within a 45 minute radius of the Airport. This should not preclude transfers between modes to meet this target and should therefore necessarily examine the option of a transfer at Papatoetoe or Puhinui.
7. We note that the Jacobs report identified that 7,350 daily commuters originate from Manukau and the east, twice as many than that originating from the north and central Auckland.
This was the point that Patrick raised in this post back in August. The Jacobs report helpfully included this map on p. 36.
8. The current Airport 380 bus service connecting at Papatoetoe to rail services yields a fastest possible PT journey time of about 49 minutes from Auckland Airport to Britomart. However, there are a number of issues associated with transferring at Papatoetoe: frequency of service; ease and legibility of transfers, and the lack of a RTN quality right-of-way.
49 minutes is my own personal best for a trip from Auckland Airport to the CBD. It was a bit of a fluke as the 380 arrived at Papatoetoe about 1 minute before the train arrived. “Legibility of transfers” is a reference to the same bus stop being used for both Manukau-bound and Airport-bound directions of the same service. Moving the transfer point to Puhinui would have a positive impact on reducing the CBD – Auckland Airport journey time, but it will be absolutely critical for any new service to be much more frequent than the current half hourly service and it would have to be in its own right-of-way to avoid the ever increasing congestion along 20B.
9. It is timely to bring to the attention of the Board that NZTA is currently planning a widening of 20B along the Puhinui Rd alignment for general traffic.
In actual fact Auckland Transport officials were already aware of this, but in the past Auckland Transport have had to play catch-up with New Zealand Transport Agency. Hopefully there will come a day where Auckland Transport advance public transport projects ahead of the NZTA’s road building exploits. AT have even gone as far as looking at catchments and alignments of what could be the Botany Line, which are shown in these two illustrations that were supplied to us.
1. As a matter of urgency, AT should work with the NZTA to designate a rail corridor east of Auckland Airport on the 20B alignment with a connection to the main trunk line. This designation work should also consider extending further east to include Botany.
2. Immediately establish a bus shuttle service between Puhinui Station and Auckland Airport, preferably with bus priority measures.
3. Auckland Transport should continue with designating a rail corridor between Onehunga and Auckland Airport.
That final point is important. The residents of Mangere and surrounding areas deserve decent rapid transit as much as anywhere else in Auckland, and they really have been short-changed by successive organisations failing to plan a rapid transit corridor. Perhaps if the main CBD – Airport connection is decided to be via Puhinui, then alternative alignments could be looked at between Onehunga and Mangere that have greater catchments and, potentially, could be a bit cheaper and quicker to implement too.
The presentation was received by the AT Board without much in the way of comment. It will be very interesting to see how AT evaluate and prioritise a Botany – Puhinui – Airport Line against all the other transport projects going on, including Dominion Rd LRT. When you look at the potential catchment of the Botany Line and consider that it will probably be cheaper to build, it wouldn’t surprise me if it ranked higher than Onehunga to Auckland Airport rail. The simple service pattern that would also result from a transfer at Puhinui is also extremely compelling – every Southern or Eastern line train connects to Auckland Airport, both from the north and from the south. We will have to wait and see where this heads now.
The ATAP final report includes a 30 year vision for Auckland’s strategic public transport network. It is a substantial expansion of what we have today and quite closely resembles our “Congestion Free Network” developed in 2013:
ATAP generally goes out of its way to avoid making a call on the specific mode of new strategic public transport projects, instead using the phrase “mass transit”. However, it does show CRL as the only expansion of the heavy rail network (in red) with all other new strategic PT routes presumably being something other than heavy rail. Elsewhere, ATAP notes the need for ongoing investment in upgrading the existing heavy rail network over time to provide for growth in passenger and freight services – but not an “expansion” of that network.
This is quite a change from the 2012 Auckland Plan, which envisaged heavy rail to the Airport, the Avondale-Southdown Line and, in the longer term, rail to the North Shore. At times we have also seen the Mt Roskill rail spur being considered as another useful (if relatively small) expansion of the heavy rail network.
This change appears to have occurred on a relatively ad hoc project by project basis, rather than as part of an overall strategic plan, which I think sits behind much of the discomfort that people have felt about Auckland Transport decision to progress light-rail, rather than heavy rail, as their preferred strategic public transport mode to the Airport. It is worth thinking about this shift at a network level, in particular at the question of whether further expansion of the heavy rail network is likely. If not, it seems that CRL may actually be the end of the heavy rail network – rather than a key catalyst for its expansion.
Compared to other PT modes, heavy rail has some advantages and disadvantages:
- Very high capacity
- High speed
- Can leverage off existing network
- Very demanding geometry leading to high construction costs
- Creates severance when at surface level
For Airport rail, the capacity requirement of heavy rail wasn’t really a factor, due to relatively low projected passenger volumes – around 2,000 southbound trips in the morning peak in 2046 (compared to around 10,000 peak trips coming over the Harbour Bridge today in the morning peak):
While I think actual use will be much higher than this (models tend to substantially under-estimate future public transport use) it will still be well within the capacity capabilities of other modes like light rail. Therefore, the comparison really came down to a speed vs cost trade-off, with the high cost of serving heavy rail’s much more demanding geometry making this trade-off clearly fall in favour of light-rail.
The high costs of serving heavy rail’s demanding geometry means that heavy rail is most likely to “stack up” as the best option when we’re looking at a corridor with extremely high demand (i.e. beyond what might be able to be catered for through other modes) or where we can utilise the existing network.
North Shore rapid transit is potentially an example of a corridor which is likely to have very high demand in the future – because it is the only connection between a very large part of Auckland to the north, and the rest of the region. Early work a few years back suggested heavy rail as the preferred option, but more recently this appears to have shifted – illustrated by ATAP’s strategic PT network map linking North Shore rapid transit into the proposed Dominion Road LRT line. I know Auckland Transport are currently looking at different rapid transit options to serve the North Shore once the Northern Busway hits its capacity limits. I suspect the main question will be the trade-off between the extra capacity you get from heavy rail against the much higher costs of having to regrade the busway, along with the challenge of how it would link into the rest of the public transport system.
Importantly, even if the CRL does “complete” the heavy rail network and we don’t see major new lines in the future, there’s major upgrade of the network we have that will be required over time. Most obviously this is to separate passenger and freight services, but over time I see the need to separate local and express passenger trains – especially as the southern greenfield area grows. Therefore, ATAP’s $3 billion 30 year rail programme is almost certainly on the light-side of likely future investment in the heavy rail network in Auckland.
Auckland Transport have published a new version of their airport rail video, essentially stripping out the heavy rail parts while also adding a little bit more detail about the airport.
Perhaps one of the more interesting aspects is it shows a bit of how light rail would get through Onehunga. It appears the plan is to elevate the light rail line over Neilson St right where AT are about to remove the bridge that lifts the road over the rail corridor.
Every time rail to the airport is discussed, here or in other places, there are a number of people who question AT’s decision to use Light Rail to connect to the airport. The biggest complaints/misconceptions I’ve seen against the idea of using light rail to the airport are:
- Light rail is slower – especially on Dominion Rd with lots of stops
- That it will be like Melbourne mixed traffic
- Light rail doesn’t have enough capacity
- It will mix with trips on Dominion Rd
- It’s light and so heavy must equal bigger and better
So let’s step through some of these and to do so, I’m going to use Seattle as an example. The reason for using Seattle is that its Link Light Rail has many characteristics that appear to be almost identical to what Auckland Transport is proposing.
First a little bit about the system.
Seattle has only recently started building its light rail system and the first section opened in July 2009. Since then there have been a couple of extensions, to the airport in the south (six months later) and just in March this year, a 5km, 2 station extension to the north. Further extensions in each direction are already under construction with other lines and extensions planned.
As of now the entire light rail system is just over 30km in length which is almost identical to the distance between Papakura and Britomart. It does have fewer stations though and outside of the city, much wider station spacing. The route is a mix of grade separated right of way with a mix of tunnels and bridges, median running and in the city centre it shares one tunnel with buses. Below is an image from Streetview showing Martin Luther King Jr Way which a 6.2km long section of median running and is similar to what we can expect along Dominion Rd. As you can see it is not mixed with traffic and the rail is separated from the road by a small kerb. Access across the tracks at intersections is controlled by lights.
The light rail vehicles used in Seattle are capable of speeds up to 105km/h which at maximum is only 5km slower than our heavy rail trains are capable of, and which ours don’t often get close to achieving in normal service. Seattle has some fairly lengthy sections which over which I imagine it is able to make the most of it’s speed. That means it only takes about 44 minutes to travel the 30km for an impressive average speed of just over 40km/h. That is about the same average speed as the Eastern line from Manukau but considerably faster on average than the Southern, Onehunga and Western lines, the latter two average less than 30km.h.
Even if you exclude the section from the Airport to Rainier Beach and from Westlake to the University of Washington, the system achieves similar average speeds to our network.
Obviously our existing trains need to be faster but that is a discussion for a separate post. What is clear is that at the very least, it is possible to get light rail up to a similar speed as what we’re achieving now with our rail network.
To achieve the times Auckland Transport claim, LRT would only have to average 30km per hour, the same as being achieved on the Rainier Beach to Westlake section. With AT planning to create a corridor like shown above (but with a single traffic lane instead of two), that should be possible. There’ll be no light rail mixing with cars and also no stopping ever few hundred metres like many buses and many traditional tram networks such as Melbourne do.
It’s all very well saying that heavy rail has more capacity but just because you can build a rail line capable of running trains with a capacity for 1,000 people every 90 seconds, it doesn’t mean you should. It is very expensive both to build and run so most cities only do it if they absolutely need to. Better to build enough capacity for what you’re going to need (plus a bit of redundancy).
As we know, AT are planning on using up to 66m long light rail vehicles (two 33m coupled together) that can carry up to 450 people running every 5 minutes. Looking over at Seattle, they have 29m long vehicles that can carry around 200 people running at up to every 6 minutes in the peak. They too can couple vehicles together and until recently were limited to joining two trains together but their system allows for up to four to be coupled. Four vehicles connected together would be around 116m long and carry up to 800 people – more than one of our 6-car trains are designed to carry (ours carry 750 people). Given the technology is obviously already available, there doesn’t seem to be a technical reason why we eventually couldn’t see longer light rail trains here – assuming we designed for the possibility.
Another way of looking at capacity to see how it’s performing. Sound Transit who run the system publish ridership data monthly. The opening of the extension to the University of Washington in March has seen ridership soar at up to a staggering 83% compared to the same month a year prior. That means over the last few months, this single LRT line is carrying more than Auckland’s entire rail network combined. The results suggest that by the time the extension has been operative for a year that their system will be carrying 20 million+ trips a year. Seattle’s weekday numbers are about the same as what we have but they do much better on weekends, something we’ll hopefully see the new network improve.
On both speed and capacity, the example for Seattle shows that Light Rail can be every bit as good as our current heavy rail system. For me the key is not the name of the mode but how it’s designed. The pressure that needs to be applied to Auckland Transport, the council and the government is to provide the funding needed as soon as possible and to ensure that it’s implemented to the level advertised in the video above (or better). Light or heavy, it’s still rail to the airport.
Within the Auckland Transport Alignment Process Interim Report is the news that the six lane highway across the harbour as currently planned, the Additional Harbour Crossing, just doesn’t work, and is now at least mortally wounded, if not actually dead.
Clearly a total re-think of ways to serve the growing movement demand between the city and the north east is required. In January the Waterview connection and the supersizing of the North Western will open, and later a new interchange between SH 18 and SH 1 will complete the next road connections between the upper harbour and the wider city.
Additionally the SkyPath, SeaPath, and improvements to ferry services will also be added to the mix for the inner harbour, for the Active modes.
The great missing piece in the movement jigsaw for this part of the city [as elsewhere] is a Rapid Transit connection across the harbour and through the Shore. The case for the next major connection across the harbour utilising the high capacity, smaller foot print, and traffic reduction outcomes of Rapid Transit is now all but made.
Here I want to explore the options for an actually successful harbour crossing, particularly in light of recent announcements about the Mangere/Airport RTN, but also because of the need for a viable near term system that fits into the longer term needs of our growing city. I can’t emphasise this enough; for a system to actually get funded and built it surely must fit both those criteria, and that is not easy.
What follows is an exploration of possibilities done with a fairly broad brush; a high altitude view, with a lot of scope for variation in detail.
In 2012 I wrote a couple of speculative posts proposing the Shore Line as an extension of our current rail network. Here and here. I was interested in looking beyond the CRL to examine future improvements to the Shore and through Mangere to the Airport. In particular in offering an Albany to Airport one seat ride through an east-west CRL II type project, like this:
CITY CENTRE ‘The Cross’
While there is some elegance to these proposals [and some problems] it is clear that I started from an unexpressed assumption that all future rail systems would be extensions of the current network. This is no longer my view primarily because our current network has very firm structural upper limits for train movements, that means at some point it becomes limiting to try to add ever more arms to this body. Adding an additional high demand area like the North Shore will bring those limits forward, especially to hard to expand pinch points like Newmarket Station and Junction. Remembering too that freight movements are growing on parts of the network too. This isn’t to criticise our re-born urban rail system, it will remain the vital core of Auckland’s RTN, growing quickly post CRL to 50m trips pa. and beyond. But rather to recognise that it needs to be free to serve these core roles, supported by other existing and enhanced networks, and some new ones.
I also suspect that the required crosstown tunnel would be prohibitively expensive, especially when combined with a cross harbour one. Together these costs could be sufficiently high to kill the plan.
Instead, adding a new and complimentary network, making connections by transfer at interchange stations, adds resilience as well as more capacity, in that problems on one network won’t affect the other. This also allows us to tailor the next network to our current demands and utilise all the latest technology and thinking without needing to accommodate it to the physical parameters of the existing system [freedom from the ‘happenstance of yore’].
On the assumption that the current Busway can be relatively easily converted to Light Rail, the obvious opportunity is leverage off AT’s Light Rail plans for a Queen St-Dominion Rd system through Mangere to the Airport so that this:
becomes this, an idea we have explored in the past:
So the return of the highly desirable one-seat ride from Albany to the Airport, on a high catchment spine; the A-Train, but this time via Light Rail, on a combination of higher speed grade separate paths, and high access street running. The advantages and disadvantages of these conditions have been debated at length on previous posts, what I want to examine here are future network possibilities of such a system, because networks are always greater than their parts. However it is worth a visit to this analysis of the recent addition of Light Rail to Paris’ Transit mix on this very issue; degrees of separation, I guess you could call it. Basically it concludes that while it is always better to strive for as much separate running as possible, this needs to be balanced against both capital cost and quality of access. And especially the quantity and quality of the transfer nodes with other major Transit Networks.
It takes just a few seconds looking at the RATP map to see why the Paris trams are so useful. In Paris’s hub-and-spoke transit network, they are the rim of the wheel, connecting the ends of Metro and RER lines in far-flung parts of the region. All nine lines offer at least two stations that connect to other modes of transit. Some offer many more:
||No. of transfers
||No. of total stations
||Pct. of stations with transfers
So returning to Auckland and the proposed A Line we can see that while it would be a great complement to our existing fast growing Rail Network, and likely further RTN extensions including AMETI, and the North-Western, it still only connects with it at the four stations indicated with blue dots above. However the possibilities for leveraging off this system to create a second route with a rich abundance of connectedness:
The western section carries along the grade separate SH20 alignment to a new station at Owairaka, then continues to Mt Albert Station, Unitec, and to the future RTN station at Pt Chev on the North-Western line, on-street [The RTN connections are why I prefer this possible extension over one to the metro-centre of New Lynn]. The southern to Puhinui Station and the Manukau City Interchange Station down Lambie Dr, again mostly on its own alignment.
Which of course can be extended to include AMETI:
The simplest idea would be to run these as two overlapping lines A and B, giving the Airport and Mangere great connectivity west, south, north, and east. Direct to the City Centre and great connectivity with every branch of the RTN, including of course the Onehunga line. Or say three lines all converging on the Airport. Such a system is also highly stageable, and you all can haggle over your favourite technology for each part….
The key principle though must be future proofing for upgrades. I think it is vital, for example, that the harbour crossing, if it is to be Light Rail, is built so it can take Light Metro for the time in the future that the demand from the Shore is high enough to justify a tunnel from Wynyard to Aotea Station and the option of implementing a fully driverless system as then it would be 100% grade separate. I’m sure some would like to start straight off with such a system, but I think it is clear that designing systems that can grow with the city is the only viable way forward.
Without trying to put a date on it, below is a pretty good integrated RTN future to aim for:
Rail running at 5min frequencies on the outer lines so a train every 2.5 mins in the CRL and other places [Red and Green]
LRT also at varying frequencies depending on place, perhaps even two routes from Wynyard to the city
More Ferries, Rapid Bus on Gt North Rd and the NW and across the Upper Harbour, although especially from Pt Chev to the city could be LRT too.
When or whether parts of this are Buses or Light Rail, are not so much my focus but rather getting the coverage optimised, and the routes protected.
No doubt it will change but here’s a potential version. Discuss:
A lot of debate over the last week has focused on rail to the airport and Auckland Transport/NZTA’s decision to dump heavy rail as an option, primarily due to costs but also because they believe light rail could deliver similar benefits. As has happened pretty much every time in the past when discussing this topic, the focus of many commentators here and across the media spectrum has been squarely on connecting the airport and the CBD. That’s somewhat understandable given both are significant destinations but when focusing on a singular use outcome it naturally results in a misunderstanding of what is trying to be achieved and people trying to come up with alternative ways to achieve that.
The two prime examples of this is the suggestion that we need to have a non-stop express service between the two main destinations and that we could save money but just building a connection across open land from Puhinui to the airport. The latter was even the subject of the herald editorial on Friday.
When most of us look at a map of Auckland’s railways, a spur line to the airport appears obvious and easy. The main line south runs through Wiri and its Puhinui Station is about 5km from the air terminals. A track could be laid through largely open land to the airport perimeter. What could be simpler?
Alas, simple solutions seem not to be welcome in the organisations charged with planning Auckland’s transport.
They are looking at the wrong route
While a faster and/or a cheaper connection would be nice for those just going to the CBD, the numbers doing that exact trip are never likely to be higher enough to justify the scale of investment that rail requires. More importantly, the issue with both of those positions is they ignore one of the key strategic goals that are trying to be achieved, to improve public transport for those that live and work in the Southwest and this is a goal regardless of the mode used.
Currently the Southwest is estimated to be home to close to 50,000 people and growing to an estimated 66,000 by 2043. Some of that growth is already underway with developments like the Walmsley Rd SHA set to deliver around 1,600 new dwellings which is likely enough to house another 5,000 people. There are also around 31,000 jobs in this area of which about 12,000 are at the airport itself with the rest in the industrial areas to the north or in and around Mangere and Mangere Bridge. The airport company expect to employment numbers around the airport will increase significantly as more land is developed. That will help to make
Whichever way you look at it, that’s a lot of people and a lot of jobs and it would be short-sighted and unfair not to give them some form of quality PT option – this equally applies to other parts of the Auckland urban area too.
Looking at Stats NZ commuter view for two of the area units included in the figures above, we can see that the biggest single destination (outside of working within the same area unit) was working in the Mangere South area which includes the Airport and the Ascot/Montgomerie Industrial area. But while not to the same individual level, there are a number of people who also travel north to work in other parts of the city.
Workers at the Airport itself come from all over the region, a lot from the east but also a significant numbers the north. The numbers from the east will be partially why AT have said they’ll start the early stages of investigation into an RTN route between Botany, Manukau and the airport.
So rail to the airport is actually about serving three separate markets
- Travellers themselves
- People working around the Airport and nearby employment centres (this is likely to be the biggest share of potential users)
- People living in Mangere and its surrounding suburbs
Spending the kind of money needed to build a rail line to serve any one of these uses is almost certainly not going to be able to stack up but with all three uses combined it can. What’s more from a PT perspective a Southwest line likely has a couple of big advantages that could make it one of the busiest on the network. Compared to our existing RTN lines, having very strong anchors at each end of the line in the form of the city and the airport will help get good bi-directional usage. Added to that the inherent nature of airport arrivals and departures throughout the day would help drive off-peak usage. As such frequencies are likely to need to be kept fairly high which also benefits others on the route.
Like how some people have incorrectly assumed that because the CRL is in the city, it’s all about trains going around the city, rail to the airport is assumed just to be about people going to the airport but as I’ve discussed it’s really about serving the entire Southwest. Formally the project is called Southwest Multimodal Airport Rail Transit (SMART) but that’s a name that will never catch on. People understand the term Airport Rail and also like the CRL, some people probably won’t realise/understand just what it is trying to do till it actually opens.
Following on from my post a few days ago on light rail being preferred to the airport – which has been one of the most commented on that we’ve ever had – by far the most common concern is around the speed of light rail, particularly on the section along Dominion Rd.
I’ve taken a look in the past at what other cities who have built similar light rail systems to what AT are proposing and I found that a number of them are capable of the speeds suggested. But of course they aren’t Auckland so I thought I’d take a quick look at some of the things we know that AT are proposing that will help ensure light rail will be quick and reliable along this important corridor.
Fewer stops and more direct route
Buses along Dominion Rd between the Southwestern motorway and the centre town have almost 30 stops to pass along the way. There are of course bus lanes along a lot of the route but they stop short at key intersections which can cause delays. Before light rail came along, AT had planned to lengthen these bus lanes right to the intersection to help speed up buses further. The current timetable shows travel times between Denbigh Ave close to the motorway and the city taking between about 23 minutes and 41 minutes depending on the time of day.
With light rail Auckland Transport plan to significantly reduce the number of stops along the route with just eight on Dominion Rd itself and only 10 to get to the same point in town. Of course one down side to some residents is it means they may have to walk further to get to a stop. For light rail it appears they’ve effectively picked one stop in the main centres along the route with intermediate stops between them.
You can see the difference in the two in the map below. Fewer stops means vehicles can travel faster as they’re not constantly accelerating or braking.
Buses can also suffer from dwell time issues from boarding and alighting passengers. It is possible to address some of these, such as restricting cash payments, allowing all door boarding which could speed up buses a bit but off line ticketing and fast boarding are a fairly standard feature of light rail so provided that AT implement it properly that will help compared to the current situation people compare to.
In addition to fewer stops, you can see from the map above that light rail takes a more direct route to the city and so has fewer intersections to negotiate.
Buses currently are at the mercy of traffic lights and at the big intersections with long cycle times this can slow them down and cause buses to bunch. When bus frequencies get too high it results in there almost always being buses waiting to get through lights, which is often a leading cause of them bunching and is not good for operations or passengers.
One reasons we heard AT were for looking at light rail on Dominion Rd in the first place was that it allowed fewer but much larger vehicles to run. The advantage of that is it allows for signal priority to be effective and so depending on how it is set up, it could mean the light rail never has to stop at Intersections other than at its stations. Again that could end up being a significant time saving.
On a related note, we know that light rail will be able to skip one major intersection with AT have suggesting that they will tunnel under the K Rd ridgeline as part of building that part of the line. I believe this is mainly for operational reasons – removing a steep part of the hill – but it will also benefit travel times.
Centre running protected route
Running light rail down the centre of Dominion Rd means it won’t get held up by cars turning left into side streets or driveways, parking to run into the shops or drop someone off/pick them up. Cars would only be able to turn right at specific intersections and not when a light rail vehicle is around. It is also common for the light rail corridor to be protected by a raised kerb to prevent cars from accessing the tracks or using them as a median, in much the same way that kerbs are used to stop people driving in cycle lanes. Examples include the Gold Coast as shown below and Seattle.
A raised kerb can also be seen in some of the concept images AT have released in the past such as this on of Ian McKinnion Dr
When you combine all of these elements it means that light rail is able to travel down Dominion Rd pretty much completely unimpeded by other traffic and only needing to stop a few times along the way. The speed limit for light rail would almost certainly still be 50km/h but even if it could average half that, it could travel the 7km between the centre of town and SH20 in just 17 minutes. As a quick comparison, in the same timeframe leaving from Britomart would be at about Ellerslie – although admittedly we expect AT go get that time sped up a bit.
Once the route reaches SH20 the LRT route appears to be almost completely separate from traffic with the exception of a small section in Onehunga and so for most of it could easily travel at up to 100km/h if the vehicles and tracks were designed to do so – and that has been suggested by AT in the documents they’ve produced so far.
This gives me confidence that for LRT at least, the times AT have suggested are realistic but I’d add that a lot depends on just how well they implement everything.
A big day yesterday with Auckland Transport officially deciding to drop the option of heavy rail to the airport in favour of light rail (LRT) or perhaps even buses as had been foreshadowed in the Herald in the morning but also hinted at for some time from previous information released by AT.
Yesterday afternoon they released the previously confidential paper that was used to justify the decision which has thrown up a few interesting details.
In the end the biggest nail in the coffin for heavy rail has ended up being the cost which is now estimated at $2.6 to $3 billion. Being even more expensive than the CRL and with fewer benefits – after-all the CRL improves the entire rail network – it is always going to be hard sell and in the end AT and the NZTA have said it simply offered “low value for money” with a Benefit Cost Ratio of 0.37-0.64. As a comparison they estimate LRT could have similar or even greater overall benefits – I’ll get to that soon – but come in at less than half the price at an estimated $1.2-1.3 billion giving it a BCR of 1.11-1.72. That figure seems to be an improvement on earlier information such as the video that was released at the beginning of the year.
The biggest issue with the LRT option though is it assumes that LRT will already be in place along Dominion Rd and that this is therefore just an extension. The issue with that is so far there is no agreement from the NZTA or the government that the Dominion Rd route will be supported – although the ATAP report last week seems to confirm something more than just more buses will be needed. But even if the cost of LRT along Dominion Rd (estimated at $1 billion) was included in, it still comes in cheaper than heavy rail and would have even higher benefits. The LRT option also benefits from providing new connections on the isthmus to the South West so represents a greater expansion of the rapid transit network. A comparison between the accessibility of the two is below.
The estimated costs and benefits discounted to 2015 $ are shown below.
On the costs, one expensive part of heavy rail is the need to also upgrade the Onehunga line and deal with the level crossings. AT say they considered three options for this, a long New Lynn style trench, realignment via Mt Smart and elevated rail with costs ranging from $458-578 million. Later it appears a low cost option of closing two of the level crossings and barriers which would cost $155 million and that was used in the Indicative Business Case but AT are also concerned not grade separating the crossings would cause traffic issues.
Within the airport they also eventually came up with a more direct route shown in blue but that requires bored tunnels ~20-25m deep and adds to the costs listed above and is what pushes the cost to $3 billion. There are other issues too such as with the terminal station under the airport, AT would have to pay for it when the Airport wanted to upgrade the terminal which is likely to be sooner than when AT have LRT scheduled and that would add $100 million to ATs already tight budget requirements.
The paper also addresses many of the issues that have been raised on here and in other places in the past.
One such issue is the travel time and people have in the past questioned suggestions that LRT could be the same or even faster than heavy rail. They say travel times were worked out using several models taking into account issues such as LRT speed down Dominion Rd, the acceleration/deceleration possible, station dwell times, specific track lengths, curves and gradients along with how fast vehicles could travel over them. They say that because of the heavy rail geometries required there is only a few locations where HR can hit top speed.
All of that resulted in the range of times shown below from both Aotea and Britomart with LRT coming out on top from Aotea which is centre of the CBD.
One particular area where LRT is much faster is in and around Onehunga, presumably the section over the harbour where the line has to be elevated above Neilson St then again over the East-West mega road before dropping under the motorway bridge as shown below.
The big concern for me with travel times is that it comes down to how well they’re implemented and so far AT haven’t proven themselves good at at that. This is evidenced by how poor AT have seemingly been so far at addressing our slow trains.
Another issue that has been raised before is the capacity of LRT. What needs to be remembered is that even before this idea came along, AT aren’t planning on dinky little streetcars running around the suburbs but are large, long and potentially quite fast. As AT have said in the past they could be up to 66m long carrying over 400 passengers each which is on par with our current electric trains.
So over a two-hour morning peak with services every five minutes LRT could be moving up to 10,000 people an hour in each direction which is about the same number of people who currently arrive at Britomart each morning and nothing to sneeze at. They say the model estimates that with LRT around 3,500-7,500 people would cross the Manakau Harbour each morning with around 5,300-6,900 using the section from Onehunga to Dominion Rd. The biggest concern would be that if it was too popular it might restrict capacity on Dominion Rd but in many ways that would be a nice problem to have.
One interesting aspect of the paper is a discussion on the opportunity cost of Heavy Rail. While it doesn’t happen in reality, they say the hypothetical situation of taking the money saved by using LRT would be enough to also build a light rail link from the Airport all the way to Botany – whereby it could link with the proposed AMETI busway. I don’t think that this route is a high priority to sink another $1b+ into just yet compared to other PT routes such as the NW-busway but it is interesting to ponder long term as part of a wider LRT network.
Here are the recommendations the board were provided which I assume they accepted. Given the capacity that has been suggested is needed and that we know there are already issues with bus capacity in the city I can’t see a bus option stacking up other than for some short term improvements.
- That Management discount heavy rail to the airport from any further option development due to its poor value for money proposition;
- Instructs Management to:
- a) Develop a bus based high capacity mode to the same level of detail as the LRT option to allow a value for money comparison with the LRT option and submit this to ATAP for consideration;
- b) Refine the LRT option further to address the high risk issues as articulated in this paper;
- c) Report back to the Board on the findings of the bus based high capacity mode and LRT comparison.
- d) Progress with route protection for bus / light rail, not heavy rail;
- e) Align the SMART and CAP business cases to enable the consideration of an integrated public transport system between the city centre and the airport
- f) Progress the business case development of the RTN route between Botany, Manukau and the airport and align this with NZTA’s business case development for SH20B.
The decision between light and heavy rail will never please everyone but personally I’d rather a light rail connection that actually happened than a heavy rail one that never did. A bit of a case of don’t let perfect be the enemy of good but in this case it’s not clear that heavy rail is perfect. I also don’t think it’s realistic to think that it’s just a political change away from the decision changing like some have suggested. I personally can’t see other political parties agreeing to fund something with such a poor business case given the alternative option that now exists. We’ve been critical of road projects that have poor economic cases so it’s only fair that we do the same with PT projects (although do wish these road projects were subject to greater levels of scrutiny – looking at you East-West).
What is needed though is to simply get on with things. We can’t afford to wait another 15+ years for this to be built, we need to be getting on with it. Compared to some of the motorway projects which now get accelerated rapidly, major PT projects like this seem to languish in the back of the planning departments for years, if not decades. This needs to change if Auckland is to become a much more liveable city.
One of the main concerns I’ve seen raised about the idea of Light Rail to the airport has been the speed. In particular, that light rail is too slow in comparison to heavy rail, especially along the Dominion Rd section where it is also suggested it could also be held up by traffic. After I wrote about it last week Auckland Transport updated their website with some more details of the project – mostly with details from the video they’ve created but also with new a new travel time comparison as they say the times in the video are now out of date. The updated travel times are shown in the table below and as you can see are even more favourable to light rail, putting it just five minutes slower than heavy rail from Britomart and equal in travel time from Aotea.
|Britomart (downtown) to airport
|Aotea (new City Rail Link station) to airport
In the blog and other places where discussion about the idea has occurred I’ve seen people questioning the travel times claimed by Auckland Transport. If these timings are accurate I think it makes a significant difference as to the viability of light rail as an option, not just to the airport but potentially for other applications such as to the North Shore, East Auckland and the North West. With this post I thought I would examine the light rail timings in more detail to see if they stack up.
As a reminder this is the route Auckland Transport suggest.
From Britomart to SH20 via Dominion Rd the light rail route is fairly straight and AT say it would travel down the centre of the road in dedicated lanes. They also say light rail would have priority at intersections along the way so vehicles would not often need to stop unless doing so at a station. AT reaching SH20 it would then follow the motorway down to Onehunga for a short section on street before a dedicated section from there to the Airport. In total this route is around 22.6km and at 44 minutes giving an average of 31km/h. As a comparison our current rail lines achieve the following speeds (although AT do need to get them faster):
- Western Line – 27.2km – 55 minutes travel time which is 30km/h
- Southern Line – 33.1km – 56 minutes travel time which is 33km/h
- Eastern Line – 25km – 37 minutes travel time which is 41km/h
- Onehunga Line – 12.8km – 27 minutes travel time which is 28km/h
To examine the speeds suggested by AT I thought the best option would be to conduct a few case studies to see what other cities manage to achieve. Below I look at five cities that are not too dissimilar to Auckland that have working light rail systems. The travel times they achieve are based on published timetables.
Calgary’s C-Train system is one of most used light rail networks in North America with over 330k trips on an average weekday, about six times what Auckland’s current rail network achieves. The system has two lines that share a central section though the CBD, the Red line is 33km and the blue line 25.7km.
The lines generally run down either in the centre of the road or on one side and are fenced off from traffic and people however they also cross through many intersections at grade. On some intersections where the light rail route changes direction/arterial it is running on they use short underpasses so it avoids the intersection completely. In short it’s a largely dedicated corridor which is not to dissimilar to what we would expect to see in Auckland.
Calgary’s light rail vehicles are capable of speeds up to 80km/h and below is how long each line is and how much time the timetable suggests a trip the entire length will take.
- Red Line – 33km – 59 minutes travel time which is 34km/h
- Blue Line – 25.7km – 46 minutes travel time which is 34km/h
Seattle has two light rail lines, one a short shuttle in Tecoma but the main one the Central Link, is a 25.1km line from the Airport to the city that opened in mid-2009. Much of the route is elevated, in tunnels or offline (alongside a freeway) however there are a few significant sections where the route travels down the centre of a road corridor separated from traffic only by a small concrete kerb.
There are 13 stations all up. In the city station spacing is about every 600m but as it gets into the suburbs it expands and becomes more like heavy rail. There is also a 9km section with no stations which would help speeds.
Seattle uses two 29m long light rail vehicles that are coupled together able to carry a combined 400 people – they eventually plan to couple up to four vehicles together. They have a top speed of 105km/h which would be useful on the long spaced out sections.
According to the timetable a trip from the along the line would takes around 37 minutes from end to end. Over the 25km that’s an impressive average of 41km/h – quite a bit faster than all but the Manukau line on our rail network. The single line carries close to 12 million trips a year – what Auckland’s rail system carried just over a year ago but remember it’s only been open since 2009.
Salt Lake City
Salt Lake City runs a three-line light rail system through city streets in dedicated lanes and from what I can tell there is no grade separation. The system first opened in 1999 and generally tracks run down the centre of streets separated from general traffic by a small kerb. Unlike the examples above the system doesn’t have any grade separation at intersections and many side roads cross the tracks along the way. It uses signal priority to achieve a mostly uninterrupted service along the routes. The Blue Line is 31.1km, Red Line 38.1km and Green Line 24.2km.
The system uses similar vehicles to what is used in Calgary and like Calgary they have a top speed of 80km/h. The system currently carries 18-19 million trips a year. According to the timetables the three lines take the following length of time to travel end to end and following that is their average speed.
- Blue Line – 31.1km – 51 minutes – 37km/h
- Red Line – 38.1 – 59 minutes – 39km/h
- Green Line – 24.2 – 46 minutes – 32km/h
So despite not having the grade separation that the systems above have the system still manages to achieve some pretty good speeds.
Houston maybe a famous for its sprawl and massive motorways but it also happens to have an increasingly used light rail line. Its Red Line opened in 2004 was extended in late 2013 to a total of 20.6km. Two new lines opened in May last year but for this I’ll focus on the red line which carries the vast majority of over 16 million light rail trips. The system runs largely on street level in dedicated lanes down the centre of the road separated from traffic with concrete kerbs. Following the line via Google Maps shows an extensive numbers of roads that cross the tracks at points along the journey and it’s not clear if there is any signal priority and this affects the speed.
Houston’s Light Rail system includes this section through a water feature in the CBD
The system uses light rail vehicles capable of 106km/h but given the station spacing and intersections I doubt they ever get close to that. Trips along the 20.6km line take around 55 minutes giving it an average of 22km/h – a step down from the cities above.
Lastly one of the closest to home is on the Gold Coast – which I experienced myself just a few months ago. The vehicles are 44m long and can carry 309 passengers and have a top speed of 70km/h
The system is 13km long and runs in a dedicated corridor which is mostly down the centre of streets and at most intersections it has signal priority. Importantly through areas such as through Surfers Paradise it seemed to be limited in speed to around 30-40km/h for extensive sections and that limits the overall speed of services. At 37 minutes from end to end it also happens to have the slowest average of the examples at just 21km/h.
There are of course many other systems that could be examined but what is clear from the ones above is that there are a range of systems and a range in how those systems are implemented. The systems that tend to have a mix of infrastructure such as dedicated corridors and sections of on street running seem to do fairly well and those systems are also likely to be the most similar to what Auckland Transport are proposing. As such, assuming AT design the system right then a 44-minute transit time from Britomart to the airport or 41 minutes from Aotea actually seems reasonable. Further I would have expected that they’ve calculated it far more accurately that I have.
Perhaps speed isn’t quite the issue that some have made out and we can save $1.2 billion by using light rail to the airport. That might also then let us extend the light rail somewhere else such as the North, Northwest or East.