Requiem for a motorway tunnel

At first glance the NZTA proposals for a new harbour crossing are quite encouraging… that is if we assume the people of Auckland would not settle for a hideous motorway bridge destroying their new waterfront precinct and demand a tunnel instead.

A harbour tunnel certainly has it’s appeal: it would take all that state highway traffic out of St Mary’s Bay and Victoria Park and send it underground on a long invisible bypass of the city centre. We could separate peaky city commuter traffic from traffic going nowhere near downtown. It would allow us to wind back the harbour bridge to something more like a local arterial, probably with walking and cycling lanes too. We could pull down the much despised Victoria Park viaduct and remove half the lanes from St Mary’s Bay, perhaps even renovating it to act something like a western version of Tamaki Drive.

Those would be some great outcomes, but on closer inspection there are several huge issues with the harbour tunnel plan:

The approach motorway to Sydney's harbour bridge and tunnel. Do we want this in Freemans Bay and Northcote?

•  First and foremost, it would cost around five billion dollars. That is an absolutely huge cost, how can we fund that? What else would we forgo if we did fund it, or rather what better use could we find for several billion bucks? How many intersection improvements, bus lanes and cycleways would that fund? On five billion dollars the cost of capital alone comes in at $750,000 a day!
•  Secondly, do we actually ‘need’ a second motorway crossing in that same corridor? Do we need six more lanes of motorway when traffic on the existing bridge has been trending in reverse for the last half decade? After all, it only goes from the area around Onewa Rd to the Central Motorway Junction. Beyond that, do we actually ‘want’ a brand new route with plenty of capacity feeding into Spaghetti junction, something that might simply encourage more people to drive more often and create even more traffic and car dependence.
•  Thirdly, this five billion dollar proposal is for a motorway tunnel only, there is no public transport component. Certainly if a motorway tunnel was built this would allow a pair of lanes on the bridge to be marked for the busway, but if you think about it that wouldn’t be much improvement over the way the busway works already. Same route, same vehicles and capacity, same constraints through downtown, just a little less impact from congestion on the bridge.
• Finally, would there actually be much improvement to the capacity of the transport system? A six lane tunnel would provide three lanes each way, so in the peak it could move an extra 6,000 vehicles per hour. At our occupancy levels equates to less than 8,000 people per hour. That’s less capacity than the busway, at about twelve times the price!

If we look at it again we really need to go back to the drawing board. Five billion dollars to tidy up the waterfront and duplicate a few kilometres of motorway to move only 8,000 people an hour, I don’t think so. The BCR on a motorway tunnel must be abysmally small given such a huge cost and minimal benefits.

If not a motorway, then what? Are trains an affordable option either?

What we need is something more affordable, something that will reduce traffic rather than generate more, something that has wider reaching benefits and will actually reduce travel times in the long run. Given that we already have an eight lane motorway across our harbour (plus second motorway bridge across the upper harbour), surely the next crossing should be a high quality rapid transit link. One that is cheap, compact and relatively simple to build, but can shuttle tens of thousands of people to where they need to be each day completely independent of traffic congestion.

What we really need is a crossing that can move several times as many people for half the cost. This should be possible with rapid transit: a two lane public transport tunnel would be far cheaper to build than a six lane motorway tunnel (not to mention all the associated interchanges and linkages), yet two lanes of rapid transit could carry at least twice as many people per hour than six lanes of motorway.

If we want a good cost-benefit return then it has to be public transport, the question is which form gives us the most benefit for an affordable cost.

We can probably discount a busway tunnel from the start. A bus tunnel would be relatively expensive due to the demands of ventilation and fire safety (although still miles cheaper than a motorway tunnel), yet the capacity, speed and level of service offered by a busway extension isn’t game changing. The same can be said for ‘light rail’ tramway. A electrified tram tunnel would be cheaper to build than a bus one and the capacity and service level would be better, but it’s probably still not going to give enough bang for buck. To be honest when dealing with public transport in Auckland we’re going to need a huge bang from a small buck to get one over the motorway lobby.

If we want a quantum leap in capacity, speed and service then it seems our harbour crossing needs to be based around a proper ‘heavy’ railway. However the issue once again returns to one of cost. The logical route for a North Shore rail line is to convert and extend the busway, however the grades and curves of the busway aren’t suitable for heavy rail design characteristics. So much of the busway would need to be completely rebuilt if it were to carry suburban trains, possibly with long sections in expensive tunnels. NZTA suggests the entire busway would need to be widened by three metres. The alternative of not using the busway corridor would probably mean building a new line entirely in tunnel. So constructing the train tunnel under the harbour would be relatively cheap (around $1.5 billion according to NZTA estimates), but once we add in the city side connections and North Shore extensions we can start ticking off the billions.

Admin has proposed one solution to this conundrum, suggesting that we could build the harbour rail tunnel and a heavy rail extension to Akoranga and Takapuna while leaving the busway as is. The idea is that bus passengers would continue to use the busway proper but transfer to a fast train at Akoranga for the remainder of the trip into the city, presumably until such time as we can afford to rebuild the busway as a rail line. This idea certainly has it’s merits but I doubt it could ever really work politically or garner much public support. In terms of a radio sound-bite, it is a plan to spend two billion dollars to add one new station at Takapuna. I can hear the words ‘boondoggle trainset’ already.

Driverless light-metro, ticking all the boxes at an affordable price?

What we really need is a rapid transit rail system that can run though a harbour tunnel, but also be cheaply retrofitted to the busway without any major reconstruction. It needs to provide top notch capacity and service with low operating costs, and ideally we should be able to build a whole North Shore network for less than the cost of a motorway crossing if we are really going to win over the public.

Readers of my previous post will know where I am going with this: Driverless light metro could be just the right combination for the North Shore. It’s cheap to build, cheap to run, yet fast, frequent and high quality. I’ve gone into the merits of this form of railway in a previous post, but I’ll quickly recap on what we’re talking about:

1.  It’s driverless: Computerised operation removes the need for human drivers. This means the trains can run reliably at very fast headways without worrying about drivers missing signals. More importantly the lack of staff massively reduces marginal operating costs, and therefore allows high frequency service to be maintained all day and all night, seven days a week. I cannot stress enough this benefit, in Vancouver for example their Skytrain actually turns a small operational profit despite running every couple of minutes twenty hours a day.
2.  It’s ‘light’: These systems are specifically designed for urban rapid transit only, so the tracks aren’t limited to what heavy rail can handle. The system used in Vancouver and Kuala Lumpur can handle curves as tight as 35m radius and hills as steep as 1 in 10, or in other words tracks about four times as tight or steep as our regular railways. The vehicles themselves are relatively compact and use third rail power supply rather than overhead line, so tunnels and underpasses can be quite a bit smaller. This all makes it ‘light’ on infrastructure and ‘light’ on cost, but not light on performance. This is a huge plus in the North Shore context, tracks could be laid straight onto the busway without modification and new branches and extensions could be built easily in and around the existing urban fabric.
3.  It’s metro: Again these systems are custom designed just to move people, providing high frequencies, high speed and comfortable capacious trains without delays or interference from freight or anything else. With a train arriving every few minutes at every station on the line it would provide as good a service as the metros of London, Paris or New York.

In summary, a light metro system on the North Shore could be as cheap to construct as a tramway, cheaper to operate each day than buses, yet provide greater capacity and service than even a full blown suburban railway. For well less than the cost of a motorway tunnel under the harbour we could have a whole metro network for the North Shore. Indeed it could also be the perfect mode for other areas of Auckland that have no rapid transit and similar constraints to building it, in particular the northwestern corridor, the upper harbour and southeast through Howick, Botany and Flatbush.

What would a North Shore light-metro cost?

As a benchmark for costs I will use the recent Canada line light-metro that was recently built in Vancouver (which despite the name is actually two lines, a main one and a branch to the airport). The total cost of this project was $2.054 billion in 2009 Canadian dollars, which equates to about NZ$2.95 billion today. This line is actually totally independent of the rest of the Vancouver Skytrain system as it was built using Korean technology that is slightly different to the rest of the network. As such it is a good representation of a complete ‘turnkey’ network like Auckland would have to build.

This three billion dollar sum bought a total of 18.4km of double-track line (comprising 9,080m in tunnel, 7,349m on elevated viaduct, 1,386m at grade and a bridge 614m long), one major junction, 16 stations (8 underground, 6 elevated, two at grade), a operations and maintenance facility, and twenty two-car automatic trains to run on it.

So this represents a cost of NZ$160 million per kilometre for all the track, trains, stations, tunnels, bridges and viaducts needed to build and run the line. As you can see most of the Canada Line was built in tunnel or elevated, so it really represents the top end of what we would pay in Auckland given that we already have most of the corridor available at-grade. Using this rough guide we can get a ball park figure of what light metro might cost on the North Shore.

Lets start with the harbour tunnel itself, a 3.2km link from Wynyard wharf to the vicinity of Onewa Rd interchange. NZTA have estimated this would cost about $1.5 billion to construct to heavy rail standards. For the purposes of this exercise I’m going to drop this back to $1 billion to account for the fact that light-metro can handle steeper grades, tighter curves and would have a much smaller cross section so would require substantially smaller diameter tunnel tubes.

Proposed light metro lines on the North Shore (black). Stars indicate station locations and purple lines are major bus corridors.

Next up is the brand new parts of the line. For the city side connection we’ll assume a 1.4km cut and cover tunnel from the corner of Wellesley and Albert St to the start of our harbour tunnel at Wynyard wharf. This includes two stations, one at Aotea and one at Wynyard. As an aside, the site we dig out for the Wynyard station would be the perfect spot to launch the machine that bores the harbour tunnel. From the northern portal of the harbour tunnel we have a line from Onewa up to Akoranga, then from Akoranga let’s continue across Barry’s Point and the adjacent inlet to terminate our branch at an underground station under Huron St in Takapuna. So that’s an extra 3.4km of track (mostly just widening the existing motorway causeway, but with some viaduct and underground) and two new stations at Onewa and Takapuna. Altogether our brand new track requires 4.8km of track with four new stations, applying the Canadian costing gives us a rough figure of $768 million for this section. Once again I will point out this is the average cost of Vancouver’s mostly tunnelled and elevated line, so probably well above the maximum we could expect running it along the motorway in Auckland.

After this we need to look at the busway. From Akoranga to Constellation is bang on 6km long, with four existing stations that would need some level of modification. To account for the fact that most of our infrastructure already exists I’m going to (somewhat arbitrarily) halve the cost of this section. $80 million per km should be sufficient to install track, power delivery and control systems and modify the station platforms. So my guestimate is that it would cost $480 million to refit the busway proper as a light metro line.

Next would have to be an extension of the line to Albany. For this I’m going to assume a 4km route through Albany to the existing park-n-ride station, mostly elevated with short sections at grade and perhaps a tunnelled section in Albany itself. I’m also assuming two new stations: one at Rosedale Rd, the other central to the Mega Centre/University/Mall. Furthermore we will probably locate our stabling facility in the industrial area somewhere near the new Rosedale station. Once more applying our costing figure gives us a price of $640 million for this extension.

Where next? Well the obvious route would be a branch from the vicinity of Constellation station along the SH18 corridor. For the moment we’ll stop at Greenhithe Rd, but eventually this branch could reach right across the upper harbour to Henderson and the Western Line. So here we’re looking at 4.5km of line, mostly elevated, with three new stations at Unsworth/Albany Industrial estate, Albany Highway and Greenhithe respectively. Using our reference figure this comes in at $720 million. A touch pricey for those few stations but I guess the real value would come with the subsequent extension over to west Auckland.

Right, to wrap that all up we are looking at a system of a three metro lines on the North Shore running through a harbour tunnel from the CBD to Takapuna, Albany and Greenhithe respectively. This is a total of 22.5km of double track metro rail (comprised of a 3.2km harbour tunnel, 13.3km of new route and 6km of refitted busway), with five upgraded interchange stations and ten brand new ones. That’s quite a system really, it should work fantastically with a combination of decent bus feeders, the odd park-n-ride and a little intensification around stations.

But the bottom line, how much would this cost? Well to add up these simple estimates we arrive at a maximum figure of $3.6 billion to cover everything, track, stations, tunnels, trains the lot. I realise this is a very basic analysis, but using these figures that’s only 70% of what is proposed for just a motorway tunnel from the lower North Shore to Spaghetti junction. So instead of a motorway tunnel we might be able to build this whole metro system and still have $1.4 billion left in the budget to upgrade the harbour bridge or extend our metro elsewhere! Of course three-point-six-billion is still a huge amount of money, so we could obviously start with the basics first. If we exclude the Takapuna and Greenhithe branches we get a figure of roughly $2.7 billion for the metro line from central Auckland to Albany, and just over two billion if we stopped at Constellation.

So how would it operate, what would it be like to use?

The figures for Bombardier’s ART light-metro trains show that under normal conditions they operate at a top speed of 80km/h and accelerate and brake at a rate of 1.0ms-1 (the can actually brake much quicker in an emergency, and if they are running behind they can boost speed to 90km/h in catch up mode). If we plug these figures and the spacing of the stations along our proposed lines into a little model we can work out what sort of travel times we could expect.

The main line from Albany to Aotea in the central city would take just 21 minutes from end to end. That’s a full 12 minutes faster than the current timetable of the Northern Express bus to Britomart, which doesn’t even take into account the effects of major traffic congestion in the city. It would be about the same time as driving off-peak, and much faster than driving during rush hour.

What a North Shore light metro network map might look like.

The line from Greenhithe to Aotea would take only 23 minutes all up. Right now the best option is the 956 bus using Upper Harbour Drive and the busway, that takes 49 minutes. So we’ve saved an amazing 26 minutes on this route, and again this is much faster than driving if there is any sort of congestion.

The last line between Takapuna and Aotea would take only 11 minutes from end to end. This is a massive improvement over existing bus links like the 839 and 875 that actually take 30 to 35 minutes to make the short trip! Slashing travel times between Takapuna and the CBD like this would have one very good outcome: it would allow the two centres to effectively operate as a single business district. Getting from Queen St to Takapuna by light metro would take you no longer than walking up to the university or catching the bus up to K Rd.

Fast travel times are all well and good, but not if you have to wait for ages to get a train in the first place. So what are the frequencies we could expect? Well if we again assume an equivalent number of trains as used in the costings we got from Vancouver’s Canada line we arrive at a figure of 24 two-carriage sets included in the price of our network.

Based on the travel times for the three lines above we can work out that a single set can make 1.4 return trips an hour to Albany, 1.3 per hour to Greenhithe and 2.7 to Takapuna. So our 24 sets are enough to provide a train every six minutes on each line, plus have a couple of sets in reserve for operations and maintenance.

A train every six minutes on those three lines is itself is a fantastic level of service, however it gets better. Because the lines overlap there would actually be a train every three minutes between Constellation and the city, and a train every two minutes through Akoranga, Onewa and Wynyard stations! That sort of frequency makes transfers a complete breeze. With computer control maintaining regular spacing you would never wait more than three minutes to transfer between any of the three lines. And if we recall the driverless operation allows us to affordably run the system at these headways all the time, these are the same frequencies and quick transfers you’d get at any time on any day of the week. Transferring to get from Albany to Takapuna would be just as painless at 2am on a Sunday morning as it would be on a weekday at peak hour.

But what about capacity? Could a light-metro system really move more people than a huge motorway?

In a nutshell, hell yes. A motorway lane hits the wall at approximately 2,000 vehicles per hour, so our motorway tunnel would have the capacity to carry only 6,000 vehicles per hour in the peak direction. At the usual levels of vehicle occupancy that’s a maximum of just 8,000 people per hour each way through the motorway tunnel.

So what of the metro? As we worked out above our light-metro system could easily operate under the harbour at one train every two minutes each way. With a comfortable capacity of 342 passengers per two-carriage train that works out to be 10,260 people per hour each way (and quite a bit more if we are happy to crush load people in like sardines).

So just using little two-carriage train sets we can carry more people than the motorway crossing, but as patronage increases we could very simply couple more pairs of carriages together to make longer trains. With four-carriage sets the peak hourly capacity would go up to 20,520 people, and with six-carriage sets we could move 30,720 people per hour. That’s almost four times as many people as the proposed motorway tunnel.

In other words a cheap twin track light-metro tunnel could move as many people as a motorway tunnel twenty-three lanes wide!

But there’s an even bigger gulf to consider. With a motorway crossing all those 6,000 vehicles per hour have to use the same old motorways and streets either side of the tunnel. All that extra traffic will still need to funnel down either the northern motorway, Esmonde Rd and Onewa Rds at one end, and through the to the southern and north-western motorways at the other. On the other hand our light-metro system includes the cost of new tracks right up to Albany, Takapuna and Greenhithe, so we could move tens of thousands more people per hour right across the Shore and the harbour without a single extra car on the motorway. In reality we’d probably see less considerably less cars on the motorway if it were so easy to get around without driving, plus all the buses would be redeployed to feed the local stations so there would be far fewer of them in congestion on the motorway (and some arterial routes) too.

In conclusion: huge benefits at more affordable price

So there we have it, a broad indication that a truly world class metro rail system could indeed be possible right across the North Shore for the sorts of costs that have been proposed for a harbour crossing.

NZTA really should look at realistic alternatives to a hugely expensive motorway tunnel under the harbour, given that a motorway that would only further entrench Auckland into a spiral of traffic congestion and parking issues. If we do want to spend billions of dollars on transport under the harbour then why not spend it on a light-metro system that will have far greater benefits and a lower cost?

It is currently being trialled in a pond beside SH1 near Silverdale and if successful will be rolled out to ponds in other parts of the city. The trial will last for two years and is also involving the University of Auckland as well as the council. The pond itself has been split in two with one side getting the treatment and the other side left as it is so the scientists can see exactly how effective the system is.

Here are some images of what these floating islands look like and you can read the full press release here

The NZTA often gets a bad reputation from some on this site but I do think it is important that we commend them when they work to reduce the impacts of the roading system.

I do think that that the long term solution is to pedestrianise Queen St with perhaps the exception of using it for some form of PT however it is going to be a number of years before we can get to that stage so my solution focuses on what we can do today to further reduce traffic which will help to make pedestrians feel safer in these areas.

My main concern with traffic is the ability of the streets to be used as rat runs due to how they were traditionally used and both the Elliott /Darby St and Fort St suffer from this as the diagrams below show. The solution is to simply change the direction through which traffic travels, this would would negate these streets as rat runs while still preserving vehicle access for vehicles that really need to use them i.e. delivery vehicles.

In the case of Elliott/Darby St traffic wanting to get from Wellesley St to Victoria St or even just wanting to bypass the Queen St and Victoria St intersection, which has a lot of pedestrian phases, can use these streets to do so. Elliot St is also the only way to access the car park in the currently empty site at the northern end and vehicles have to travel use it and possibly Darby St to access that car park (of course ideally that site would be developed but who knows when that will happen).

Elliott St Before

Elliott St After

Simply changing the direction of these streets and only allowing left turns on and off of them would remove that rat run and mean that anyone using them will result in a vehicle going around in a circle which means that only cars that had to use it would do so. In this case it would also mean that cars accessing the car park wouldn’t need to travel the length of the road but only one end of it.

The second example is Fort St which is probably even more of a rat run based on the number of vehicles I see using it. For anyone coming from Beech Rd and wanting to get to Queen St it is probably far easier to use Fort St than get through the three intersections along Customs St East. I think what makes it even worse is the straight section between Gore St and Commerce St allows people to accelerate and they don’t slow down as much which means they travel faster through the shared space section.

Fort St Before

Fort St After

Again changing the direction of the traffic would remove that rat run, also as cars would need to turn off Queen St to access the shared space they would be more likely to travel slower through the area. It is also worth noting that the section between Gore St and Beech Rd is also eventually meant to be a shared space, I am undecided about the best direction for this but have opted to leave it the same to avoid creating more confusion/phasing at the intersection. Commerce St, Gore St and the middle part of Fort St will remain a normal road but the footpaths are being upgraded and widened as well.

The thing I think is best about these changes are they are simple ones that could be done fairly quickly and easily as they only require a few changes to signs and a bit of education but they have the ability to remove a lot of traffic from these areas in which pedestrians have been given far more priority.

The citybound motorway lanes will be closed between the Brigham Creek roundabout and Lincoln Road off-ramp starting from 10pm on Friday (13 January) until midday on Sunday (15 January).  The westbound lanes will also be closed between Te Atatu Road off-ramp and Lincoln Road on-ramp.

“We want to cause as little disruption as possible, but this is a busy section of Auckland’s motorway network and delays and congestion will be unavoidable and our best advice to people is to stay away if possible,” says the NZTA’s acting State Highways Manager for Auckland and Northland, Steve Mutton.

“We’re taking advantage of the good weather and lighter traffic flows to enable us to demolish the old Selwood Bridge piece by piece. To ensure that the crews can do this safely and as quickly as possible, we need to close the motorway.”

The NZTA is recommending that people going west avoid using the Northwestern Motorway altogether from the causeway, as big delays are likely. And for those travelling to east to the city should use the new SH18 Hobsonville Motorway from Westgate, to access the Northern Motorway (SH1) and then travel south over the Auckland Harbour Bridge.

People travelling to events in the city are advised to plan their travel well ahead and allow extra time for their journeys. Detour routes will be in place as well as live traffic updates on radio and extra roadside electronic signs to advise drivers of conditions.

Mr Mutton says he is confident that the motorway lanes will be opened on Sunday, before most people return from their Christmas break to return to work.

“We’re aiming to cause the least amount of disruption to commuters and having our crews work around the clock means we can push on to the next stage of improvements work to complete the new bridge.

“Depending on progress and traffic volumes, we may be in a position to open one single lane westbound between Te Atatu and Lincoln Road during the day on Saturday. However, congestion will still be extremely heavy and our best advice to drivers is to stay away completely.” he says.

Two lanes on the new Selwood Road Bridge opened to traffic in December completing the first half of the bridge as part of the Lincoln Road Interchange improvements project. The bridge will be fully completed in 2013, with seven lanes for traffic and a shared path for walkers and cyclists. Ramp improvements will improve capacity, safety and trip reliability.

The $100M Lincoln Road improvements are part of a suite of projects in the Western Ring Route – Road of National Significance, prioritised by the Government to support growth and economic development in Auckland. The alternative route to SH1 will improve connections between Manukau in the south, Auckland International Airport, Waitakere and the North Shore, and ease congestion on the Auckland Harbour Bridge and the central city.