There are many similarities between the Waterview motorway project the City Rail Link. Both are extremely large projects involving considerable amounts of tunnelling but the comparisons don’t stop there. While there are very clearly a lot of other motorway and rail projects being bandied about, both projects effectively complete their respective networks allowing us to get the most out of our existing investments in them. With Waterview now well under way, I thought it might be a good idea to see if there was anything we can learn from it to help with the CRL.

The first area where Waterview could provide some useful benefits is in the area of tunnelling. While the massive TBM that is being used on Waterview is way too big for the CRL, my guess is that the engineering and construction knowledge gained by local agencies and companies will be invaluable when it comes time to build the CRL. This is especially the case as the two projects almost perfectly dovetail into each other. Waterview is under construction now with the tunnelling itself expected to start around October of this year and completed in 2016. If the CRL sticks to its current schedule of being operational in 2020, construction on the project is likely to start in 2015 but really ramping up in the 2016-2019 period as shown in the graph below from the original business case.

CRL construction cost timeline

When it comes to promoting the project, the cost is perhaps another area that we could learn a lot from. Back in 2008, when the previous Labour government were starting to get serious about Waterview, the cost of the project was reported at $1.89 billion for a pair of twin lane tunnels. To build it wide enough for three lanes, as was being pushed by the roading lobbies was projected to cost $2.14 billion.

Also interesting to note that back then then National Party spokesperson, Maurice Williamson, said he doubted we could afford $1.9b on a single project through traditional financing methods as it would deprive the rest of the country of investment yet less than a year later the party embarked on the RoNS programme, which included Waterview and that it is set to cost the country ~$10 billion over a decade.

“I don’t think the Government could ever fund a $1.9 billion road from just straight land transport funding,” Mr Williamson said. “That would mean the rest of the country would get nothing for nearly three years, so you have go to find an alternative source of funding.”

Back to Waterview, six months later in early 2009, following more work done on the project by treasury and the MoT the cost had ballooned to $2.77 billion for the two lane option and $3.16 billion for the three lane option. This caused then transport minister Steven Joyce to send the NZTA back to the drawing board and look at other options. Also worth noting that he quite clearly stated that he wanted to see the tunnels with three lanes each.

Serious doubt has engulfed Auckland’s Waterview motorway tunnels project – the vital last link in the western ring route – after a cost blowout to between $2.77 billion and $3.16 billion.

The Government has ordered an urgent review of route options after the Treasury and Ministry of Transport added financing costs of more than $500 million and an upgrade of the nearby Northwestern Motorway for $240 million to the main project.

Previous estimates of $1.89 billion for two-lane tunnels each way along the 4.5km Waterview route or $2.14 billion for three-lane links – as sought by the Automobile Association and business groups – did not include any of those costs. The new estimates are $2.77 billion for a 3.2km pair of two-lane tunnels and $3.16 billion for three lanes in each direction.

Transport Minister Steven Joyce announced yesterday that he had given officials until April to review all options for a connection of State Highway 20 to the Northwestern at Waterview, including a potentially disruptive surface route through Mt Albert and previously discarded “cut and cover” proposals.

Going from a proposal of two lane tunnels at $1.9 billion to three lane tunnels at a cost of $3.2 billion represents an absolutely massive price increase. By late 2010, despite being a three lane tunnel option, the cost of the project was back down below $2 billion at $1.75 billion. Fast forward to today and the project is being built $1.4 billion with the causeway project coming in at an extra $220 million. That means that all up both the Waterview project, and the causeway are costing ~$1.6 billion,  almost $300 million less than just what the two lane tunnels were expected to cost roughly 5 years ago.

Why have the costs come down so much being almost half of what they were at their peak? I believe much of it relates to how we estimate these types of projects. As a project moves through the stages of investigation, costs tend to increase as all of the potential issues/risks start to get thought through and these start to get factored in to the cost of the project. As the knowledge of the project improves, many of the risks can be addressed and this can give some certainty about how much the project will ultimately cost. Such a big project should hopefully also cause construction companies to be extremely competitive in the tendering process.

So how does this relate to the CRL? Well it’s going through exactly the same process. Back in 2009 it was estimated to cost $1-$1.5 billion. The November 2010 business case, came in well over that figure at $2.4 billion, as seen in the construction cost summary. Most interesting is that the construction costs alone come in at less than $1 billion but the rest of the $2.4 billion is made up of contractor costs, design and planning costs as well as factoring for various risks. The cost of the project was the one area that both Auckland Transport and the Government agreed on when the latter reviewed the business case in 2011.

For planning documents Auckland Transport then inflation adjusted the price out to when it would be built which saw the cost increase further to $2.86 billion. However critically it also emerged that Auckland Transport had already managed to find over $150 million in savings off the base cost as shown below.


Why is all of this so important? Well based on what we saw with Waterview, it is likely that the actual cost of the project will end up coming in much less than the $2.86 billion that opponents (and AT) like to throw around. A figure somewhere in the $1.5 – $2 billion mark is perhaps much more likely and would have an absolutely massive impact on the viability of the project. In fact it seems that despite the governments continued opposition to the project, we may have missed an extremely important milestone. Subtly the conversation has actually shifted from “if” to “when” with the government and MoT now seemingly focusing on the timing of when it should happen rather than if it should happen at all. If costs continue to come down, like they did with Waterview then it only help to further justify the project.

Share this


  1. That construction cost profile is interesting. Wasn’t Waterview supposed to be OPEN by 2017? What are they spending all that money on in 2018 and 2019 and even later?

    1. I guess the problem for the CRL is that one has less ability to construct parts and consider them beneficial in distinct pieces. Building half the tunnel has few benefits, building half the Western Ring Route has – because that is really just a massive upgrade of an existing network (i.e. works, even if some parts of it are sub-standard compared to the “best” parts – whereas the the CRL creates a new route, and trains can’t exactly deviate onto local roads for any missing bits…

      I am still interested in what kind of staging and value engineering is possible for the CRL. Can we do without the eastern connection in Mt Eden initially, for example (i.e. designate it, allow for it with the construction design, but don’t build it)? The lack of those grade separated crossings that wouldn’t be required must run into several hundreds of millions, while the downsides (in my admittedly rather layman’s view) would seem be much less (reduced service pattern flexibility). It certainly seems more sensible to save money there than by initially having fewer stations.

      1. I sort of compare the various smaller projects of the WRR (Mt Roskill, Hobsonville, Manukau etc.) as similar to double tracking and electrification. They were the “easy” bits that could be broken out of the overall vision leaving just the final but most expensive piece to the end.

        Yes there is the option for staging some parts of the CRL. The K Rd and Newton stations for example could be left to be built later which based on the current costings, would save around $300m each. But then you also miss out on the benefit that they provide. Also it is quite likely that we are going to expect people to transfer to get to Newmarket and based on that we would need at least Newton to allow for that to happen. While the stations are designed such that they could be built later, the eastern link would have to be done at the same time as the rest of the CRL.

        1. > the eastern link would have to be done at the same time as the rest of the CRL.

          Why? I can understand that – seen over the total cost – it is more financially sensible to build the eastern link while you are on site anyway. However, I can’t see the argument that it would HAVE to be done immediately. More sensible, yes. But some hundred million less is a tempting reduction, if one doesn’t do any DESIGNATION or DESIGN shortcuts (i.e. makes sure it can be done later).

          1. The junction has to be dug out and the disruption of doing that at a later stage would be too severe as it would likely require a prolonged closure for the CRL itself. The two deep bore stations have been designed in a way that they could be built separately (which is why the tubes through those stations are quite far apart). The attempt to drop the eastern link is why the Inner West Interchange was proposed in the first place, one or the other is needed. The problem though is that the IWI ended up costing a similar amount.

          2. Clearly there are economies of scale in doing all the work at the portal area at once, well you could form all the grades and sort the bridges etc and not lay the track, but then you would be spending most of the money and only getting some of the benefit, hardly makes sense. Clearly it would be way more expensive to reopen the site later and have to work around a then functioning rail system. I guess you could do most of the work, initially open the CRL for western line services only, then proceed to finish the eastern connection, and that may happen if it stacks up. There will be a build up of services anyway I’m sure. But the sums saved by this may be lost in the complications of working round rail ops, and losing the ability to do earthworks etc across the whole project at once.

          3. Apaologies if this in all likelihood has been chewed over a few times: Can someone remind me why we need the IWE or the Eastern Link at all – as in NEED it, not as a sensible benefit? Couldn’t trains run through Britomart (or passengers transfer there) if they wanted to go to Newmarket or onto the Southern Line? Sure, Grafton Station would become a bit of a dead dog and Newmarket-bound passengers would suffer delays compared to now, but is that all? Or am I missing something.

          4. Don’t need the IWI in our view. But Grafton is already important and certain to become ever more so as the University develops the old brewery site. The East Facing Connection is vital for the best running patterns for the whole network, for balance; sending trains in the CRL both ways from Newmarket. And for running a Metro pattern, for increased frequency, legibility, and integration with buses. By being able to access the highest demand area CBD in more flexible ways we can actually serve the whole city better, ie leverage off that demand in a greater variety of ways.

          5. Without the IWI or the eastern link, all trains would need to be sent further out west to Swanson to turn around. That would leave a very unbalanced system where you end up with 20+ trains an hour in each direction out west but only 1/3 – 1/2 of that on the rest of the network. While those of us out west would love the ultra high frequency, it does result in a lot of unnecessary running which would have severe impacts on opex and capex costs (capex costs as we would need more trains to run such a pattern).
            The IWI was intended to be used to turn some of those trains around so they could get back out south/east but would have resulted in a lot of trains dead running through the CRL taking up valuable space. The Eastern link allows for all lines to be through routed.

  2. Yes this is the problem, staging, or nibbling, has been the motorway builders best friend. Especially useful in making huge expensive projects appear cheaper.

    Delaying the Eastern Connection won’t either achieve much saving nor certainly make it cheaper in the long run. However building the tunnel, track, and Aotea Station as a stage one then moving the station construction team to the other two stations progressively would certainly spread the spending and the skills demand. Each of the two deep stations are priced at 300m.

    One problem with that is the need for Newton Station for west south transfers on the running models we’ve been looking at. Not insurmountable, with temporary running patterns.

    1. That certainly makes sense – though it won’t reduce the “exorbitant costs!” the central government can moan about to the media (even if stretched over a longer time, the station costs will still be in the budget), yet it will mean the benefits from them will come later.

      If any station should be delayed, it would be K’Road I think. Newton is more important for city revitalisation I think, and Aotea – well, that would be built anyway as part of the cut-and-cover.

  3. Just a curious question; why is the TBM used for the Waterview project too big for the CRL? Is it due to the surrounding structures in the City? Would it not be feasible to bore one deeper tunnel with this machine to carry both lines? And from a cynical point of view: was this by design by the goverment so that it could not be used for this project afterwards due to their adversion to rail?

    1. No. Two smaller tunnels is more economic. TBMs wear out anyway. I can assure you the gov was not thinking anything about the CRL when taking advice on Waterview. But as Matt says above there will be a whole lot learned from the Waterview project that should benefit the CRL, by the project managers, by the construction companies, hopefully by NZTA.

    2. The TBM proposed for the CRL is 7m in diameter, the TBM being used for Waterview is over 14m. In even drilling two tunnels with the 7m TBM will see half the amount of rock needing to be removed compared to using one large tunnel.

      As Patrick says, not any kind of conspiracy, the Waterview TBM needs to be big enough to drill a hole that allows for three lanes of traffic.

      1. Couldn’t the TBM that is used for the CRL then be used to create a tunnel under the harbour to Takapuna?
        As the CRL tunnels are only several km’s long in each direction wouldn’t have plenty of life left in it?

        1. The CRL is longer than Waterview. If you want to use a TBM that after having finished the CLR then can go more than TWICE that distance again, you’d be likely looking into doing a massively costly overhaul or a higher-spec machine costing many hundreds of millions more. And building a tunnel is more difficult than just flicking on a machine and walking away. Its a close-by thought to re-use a TBM, but it doesn’t seem to stack up, whatever way you turn it.

    3. Also, the popular perception of “we can re-use it” seems to be false in any case as well. NZTA have clarified that after its done with Waterview, the TBM will pretty much be trashed (think of a car that has run a couple hundred of thousands kilometres), and much of it will only be worth for recycling, not for actual re-use. The manufacturer has already been contracted to buy back the Waterview TBM, but will apparently only be using some parts of it in future TBMs.

      Lastly, boring just one tunnel for the CRL would mean quite a difficulty in meeting the fire / evacuation needs. With two tunnels, you can do cross-passages, and use one tunnel as the escape route from the other. With a single tunnel, you essentially would have to fire-rate it much more, AND then would still have to build a fire-rated “enclosed tunnel” within your first tunnel as an evacuation route – or dig surface emergency access routes every hundred meters or so, at significant extra cost.

  4. I really don’t think budgeting for projects based on what you hope contractors will tender for when you estimate the price to be much higher is a very good practice.

    1. I’m not suggesting that we budget for a lower price but that it is quite possible the price will much less than what is currently being bandied about. We have to go though the various project stages to get to that point though.

    2. At Waterview, the tenderers were paid (a steal, at some hundreds of thousands of dollars only 😉 for the IP of their bids. I.e. they were asked how they would do the job best AND cheapest, and then they selected one winner, but he was able to pick and chose from the crop in terms of what of the suggestions would work best for whatever part of the project. It isn’t sure yet whether that will save a lot of money, but it seems a good way to go for huge projects like this, where creative ways of going about issues can save a lot of dough. Plus it makes it at least a little less painful for the losing tenderer who has just wasted several man-years preparing a bid…

      1. Yes the contractors really bid on process and cost. This is great, it rewards creativity and innovation. NZTA have some real good IP around running these big builds. We want them on the job with the CRL too.

  5. Considering staging of the CRL, Aotea and the cut & cover section from Britomart could be constructed first. Construction of the rest doesn’t rely on access through Aotea so can be independant and not cost any more.
    It should also be easier to consent as except for the downtown centre, it’s all below the road.
    So surely construction could start next week?
    At least the stormwater diversion?

    1. The Aotea station will be an important point where the TBM is taken out of the ground, something that wouldn’t be possible if we had it working as a live station.

      1. Are they going to angle it upwards so it can pop out of the ground with its cutting head still spinning? I’d buy a ticket to see that!

        On a more sober note… I hope they build a right angle stub tunnel under Aotea so that a future Shore Line can be joined up to it without disrupting station operations, or causing a major redesign to dig a new tunnel, re-jig the elevators and escalators, etc. Otherwise it’ll end up like some of the London Underground stations that have had extra lines added in a less-than-optimal fashion.

Leave a Reply