Here at the Auckland Transport Blog we have finally found a road project that seems more unwarranted than either Puhoi-to-Wellsford or Transmission Gully. This is a road project that costs more than both of those two premature white elephants combined. Yes, we’re talking about the “Additional Waitemata Harbour Crossing Project“.
We often hear how important and urgent it is for Auckland to get another crossing over the Waitemata. This message is usually swaddled in one or more of the following soothing spoken reasons:
- Congestion is bad, i.e. your valuable blogging time is being wasted.
- The clip-ons are going to fall off, i.e. you better learn to swim.
- It’s always been in the plans, i.e. you can’t stop it now.
Let’s consider each of these arguments in turn and see whether they hold much sway.
First, the sweet smell of congestion – that addictive road building tonic that both demonstrates the folly of designing cities around private vehicles, while simultaneously providing the justification for continued investment in private vehicles. This is the same circular logic that entrapped the entire city of Los Angeles for decades, and which still seems to prevail within the MoT and NZTA. The thing with congestion is that you can never beat it by building more roads – investing in the latter usually simply enables congestion to grow back, it’s like NZTA want to spray water onto the mould that covers the bathroom ceiling.
Moving along – the particularly interesting thing about congestion in this corridor is that we as a society has just invested in:
- The Northern Busway, which provides an extraordinarily successful (given the prevailing lack of development around stations) and relatively congestion-free public transport corridor that runs in parallel with SH1. From what I can tell (back of envelope numbers) the Northern Express is now operating at close to full cost recovery. Further only being a fraction of the vehicles that cross the bridge, around a third of all people heading southbound in the morning peak do so on a bus. This graph from ARTA is a couple of years old but shows how quickly the situation is changing thanks to developments like the busway. There is still a lot we can do to further improve the bus experience for Aucklanders, especially on the city side through improved bus lanes and facilities.
- The Western Ring Route, which (when complete) will provide an alternative route for some vehicle trips. From what I know, transport modelling predicts a drop of about 5,000-10,000 vpd on the bridge when Waterview is complete. As such, the WRR could reduce vehicle volumes on the existing Harbour Bridge by around 5%. Not huge, but probably enough to noticeably reduce congestion.
Conclusion #1: Together the Northern Busway and the Western Ring Route are the main congestion safety valves that Auckland needs across the Waitemata.
Second, we accept that the clip-ons may fall off when a continuous line of fully laden 40+ tonne trucks comes to a grinding halt on the bridge while being buffeted by hurricane force winds. Yes, the next time we have a hurricane that prompts some kind of lemming like heavy vehicle pilgrimage (which for some reason ends up stopping on the Harbour Bridge) then the clip-ons may become pontoons. But does it necessarily follow that we should spend $5 billion on a new crossing? Not really, if you consider the following options as being alternatives to a new crossing:
- Ban heavy vehicles from using the bridge in high winds; or
- Restrict heavy vehicles to using the main span of the bridge; or
- Require heavy vehicles to divert via the aforementioned Western Ring Route.
Some of these management techniques have already been used in the past so they wouldn’t be something new. Conclusion #2: Ultimately there seems to be at least three more effective ways of extending the life of the clip-ons (perhaps indefinitely) that helps us to avoid spending $5 billion on another crossing.
Third but not least, we come to the suggestion that “it’s always been in the plans.” To understand this argument we tried having a look at the plans. The latest ones came out in 2010, when NZTA last considered the merits of the project. This included preparing some detailed drawings of how it could work and a Business Case that analysed the costs and benefits of a bridge option and a tunnel option. The cost-benefit ratio for the two options are shown in the box below, which is found on page 65 of the business case document:
With a capital cost of close to $5 billion for the tunnel option, this would mean a return of around $1.5 billion excluding agglomeration benefits or $2 billion including them (using undiscounted figures). So pretty much the same as getting $3-3.5 billion and burning it. Now you don’t have to be an economist to know that this analysis is suggesting that this project is morse code for “absurd waste of money”.
Conclusion #3: If the additional Waitemata Harbour Crossing has always been “in the plans” then we’d just respond that it damn well shouldn’t be. Or perhaps more accurately: Just because something was a bad idea in the past, doesn’t mean it will be a good idea in the future.
At this stage some of you may be thinking “case closed” – let’s strike that project off the list and move on. But wait there’s even more to this post: When reading the NZTA’s Business Case we found, how shall we say this, “questionable” assumptions about future growth in vehicle volumes. But before going into them its time for a little pop quiz, if you are modelling traffic flows in 2010 to predict future vehicle volumes do you:
a) base it off the currently available numbers
b) ignore what actually happened and use a model to predict what the numbers were
Here are the numbers that were used which comes from page 10 of the business case:
The important thing is the 2008 numbers seem to have been generated by a traffic model, because they do not match what was actually observed. For example, the NZTA’s own numbers for 2008 say there were only a average of 154,925 vehicles per day that crossed the bridge, a difference of over 13,000. The difference in the volumes can be easily seen from the graph below, which shows actual traffic volumes (green) and predicted traffic volumes (red).
There’s a couple of interesting things about this graph: The first (and most obvious) is that the modelled traffic volumes are approximately 10% higher than the actual volumes, even before the Business Case was released in 2010. Basically, the Business Case appears to be re-writing history by using traffic volumes that are higher than what was actually observed. Bizarre eh?
Personally, I would have thought that where you had a transport model that was predicting volumes that were 10% higher than the actual observed volumes then that would be reason to re-calibrate the model so that it more closely matched actual volumes, certainly before you did absolutely anything else with the outputs – let alone argue for us to spend $5 billion. Nevertheless, it seems (from what we can tell reading the Business Case) that the red line above was the one used in calculations of benefit-cost ratios, despite actual data showing that it was 10% off the mark from the outset (NB: A 10% difference in vehicle volumes makes a huge difference to congestion reduction benefits).
The next interesting thing is that based on the red line NZTA has (somehow) concluded that the next Waitemata Harbour Crossing is needed between 2020-2030, at which point they were expecting between 188,000-200,000 vehicles per day (vpd) over the bridge. If we take the mid-point of this range as defining the “critical threshold” then we can conclude that the current crossing arrangements is maxed out around 194,000 vpd. The figure below illustrates this critical threshold as the black dashed line. We have also extended the actual volumes (green) using the modeled traffic volumes (dashed green line). The latter sort of shows what you might expect to happen to vehicle volumes in the event that we returned to the modeled “trend” for the next 30 years.
We can see that even by 2041 the dashed green line stays below the dashed black line, i.e. we have not come close to hitting the “critical threshold” even by 2041. Stated simply, the reduction in vehicle volumes observed on the Harbour Bridge in the last 5 years or so seems to have bought us at least another 20 years when it comes to developing an additional Waitemata Harbour Crossing. Why so much time? Well, because it will take us ten years to get back to the level we were 5-10 years ago. That’s why we now have so much time and why this project, if it’s needed at all, does not seem to be “urgent”.
With all this in mind I can’t understand why some people at NZTA (and certain politicians on the North Shore) seem to be pushing for an additional Waitemata Harbour Crossing to be constructed soon, like between 2020-2030. Surely the smartest thing to do before pushing this project along is to:
- Wait until we have completed a few projects that may impact on the need for this crossing; and
- Recalibrate the model and issue an updated business case incorporating revised traffic growth assumptions?
At this stage, not only does another road-based crossing of the Waitemata Harbour seem like an ineffective way to address the issues put forward, it also seems like it’s nowhere near as urgent as some people make out. I have no problem with long term planning for another crossing, but let’s not kid ourselves that it’s needed in the next 30 years. By crikey.