The wealth of documents the Ministry of Transport has published on its website in relation to the review of the CBD Rail Link project provide a unique insight into the process that has gone on throughout the review process. If only the development of all business cases (yes, I’m looking at you Puhoi-Wellsford) were able to be so openly analysed.
Much of the information is very technical, and while outside of the immediate teams involved in the review process I may have read more about this project than most people, at times the technical aspects of it can confuse me. Modelling systems like ART2, ART3, APT, the differences between them, the assumptions made in modelling that accompanied the 2005 RLTS versus the 2010 RLTS all seem pretty complicated – but it would seem that it is these very technical issues that sit behind the different outputs of the review undertaken by the Ministry of Transport and the review undertaken by Auckland Council.
One particularly important element of the whole review process was perhaps the most technical part of it: the analysis of the traffic modelling itself. Two very interesting papers to read, that form part of this review process were a February 11 Memo from the Ministry of Transport – outlining their concerns about many of the ‘assumptions’ in the business case; and then a March 2 reply to that memo from APB&B: the group of consultants that put together the original business case.
The MoT memo raises a number of concerns about some of the inputs into the traffic modelling that was undertaken in the business case. These concerns are summarised below:Reading through the Memo is more detail, it would appear as though APB&B’s assumptions in the project’s original business case were sourced from a variety of areas that did not completely match the assumptions of the 2010 Regional Land Transport Strategy.
- The 2010 RLTS predicted a lower level of congestion in 2041 than the 2005 RLTS did: I assume largely because the 2010 RLTS was more visionary and included the completion of a lot more projects.
- The 2010 RLTS assumed a real price of $16 a day for parking in 2041 (inflation adjusted), which APB&B thought was unrealistically low – so they used the 2005 RLTS assumption of $30 a day.
- The two diagrams below compare the 2005 RLTS ‘bus lane map’ (on the left) with the 2010 RLTS ‘bus lane map’. The reality of what can be implemented by 2041 probably sits somewhere between the two: but MoT insisted that the 2010 RLTS assumption be used.
Changes to these assumptions seem to have contributed to the MoT’s work showing that the rail tunnel would be much less popular than anticipated by the business case: pretty obvious if MoT are saying that the streets would be less congested, parking would be cheaper and there would be more bus lanes. Whether those are realistic assumptions is what deserves further analysis – and that’s what APB&B stressed in their response to the points raised by MoT:This reinforces my key criticism of MoT’s review: that they’ve never considered whether the CBD can actually handle all those extra cars and buses. This ‘street network constraint’ was clearly in the mind of APB&B when they put together the business case input assumptions – so while they may not have been consistent with the requirements of the 2010 RLTS, I think they’re logical and justified.
In terms of parking charges, as was outlined above another key difference between MoT’s work and the original business case is whether a charge of $16 or $30 a day for 2041 (in real terms) is more realistic. Of course nobody knows the answer to this question for sure, but APB&B put up some pretty compelling arguments for using the higher figure: APB&B also note how illogical it would be for parking charges to effectively remain at their current levels for the next 30 years:Once again it seems that APB&B made some very logical and sensible decisions when choosing their assumptions. One does wonder why the modelling behind the 2010 RLTS was done with such silly assumptions.
The difference in opinion between bus lane networks surprisingly has a relatively low impact on patronage levels for the rail network. Perhaps more than anything else, that highlights the overly narrow focus of the way the project’s benefits have been calculated: largely as a “how do we get more people into the CBD” issue rather than a “what impact will this have on travel patterns right across the region” issue.
The final document produced by MoT as part of the ‘technical stream’ of the business case review highlights the various parts of the assessment of traditional transport benefits which have led to such a wide difference opening up between the original business case and their review:As you can see, the big differences do not relate to the first row – the “model outputs”, but rather to how those model outputs responded to the ‘application’ of various other factors like capacity constraints, congestion assumptions and annualisation factors. My understanding is that the parties generally accepted the annualisation issue was a technical error in the original business case, but in regards to the other two matters, there seems to still be ground for lively debate over which is more realistic.
This is what MoT say about the ‘capacity constraints’ issue: While Auckland’s new electric trains will clearly add capacity to the rail network – by being longer than our current trains – I must say I find it very difficult to believe that in 30 years time we won’t have any capacity issues on the Eastern Line and we won’t have significant capacity issues on the other two lines. Post electrification we’re only going to be able to operate six car trains every 10 minutes on the three main lines, plus two trains an hour from Onehunga. That’s not a huge boost in capacity from what we’re running now.Critically, MoT’s estimates for rail patronage with the CBD Tunnel in place that inform the figures above are lower than what was estimated in the business case. This seems to have been informed by their unrealistic optimism about the capacity of the rail network without the tunnel and their unrealistic optimism about the ability of the CBD to cope with an enormous increase in buses and car.
The raw difference in numbers is detailed below:
The final factor in the table earlier in this post that leads to dramatically different transport benefits is the “Application of CBD congestion assumptions”. This is another significant area of difference, resulting from the extent to which the modelling outcomes are adjusted to reflect existing levels of congestion in the CBD and the ability to get more people into the city at peak times by measures other than the CBD Tunnel. As I’ve noted many times before, this is a fundamental difference between the Auckland position and the Central Government position: Auckland understands that it’s impossible to double the number of buses into the CBD and pile 10,000 more cars in at peak times. For some reason, the Ministry of Transport doesn’t get that.
Ultimately, it seems that the key point of difference relates to the question of “when does the rail network hit capacity?” MoT’s lower estimates of rail patronage, their enormously optimistic estimates of bus patronage and their somewhat illogical assessment of rail capacity (no capacity issues on the Eastern Line in 2041 – seriously?) have resulted in the vast difference between the level of transport benefit the business case says the project will generate and the level of transport benefit the MoT’s review reckons the project will generate.
Because many of the wider economic benefits (something I’ll look at in a future post) are simply calculated as being multipliers of the transport benefits, any differences keep on being exaggerated over time as a wider variety of benefits keep getting considered. Debates over how many additional employees will locate in the CBD due to the project obviously also has an impact on the differences between the positions.
Auckland Council and Auckland Transport’s response to these issues points out many of the same concerns about MoT’s assumptions that I have noted. They say:
Well, these criticisms seem fairly obvious. If the modelling tools assume unlimited street space for additional buses and cars, of course the calculated benefits for a rail project like this are going to be significantly lower. But the fact is that the city doesn’t have unlimited street space for additional vehicles – which is one of the main points of this project: it’s the only feasible way to significantly add transport capacity to the city centre.
Could we please have a proper review of the business case that doesn’t ignore the elephant in the room – our inability to increase CBD road capacity? Oh that’s right, that’s what Auckland Council and Auckland Transport ended up doing.