This is a guest post by sustainable transport and accessibility advocate Tim Adriaansen
Following the Liberate the Lane rally on May 30th last year, Transport Minister Michael Wood wrote to Waka Kotahi Chair Brian Roche, requesting the transport agency investigate options to trial walking and cycling across the Auckland Harbour Bridge.
We continue to recognise the importance of better pedestrian and cycling access across the Waitematā. Given any trials going ahead are a decision for the Waka Kotahi Board, I’d like [the] Board to seriously consider if a temporary trial could occur over the quiet summer holiday months or a long weekend if it can be done safely.
Recently Waka Kotahi released a number of documents showing the work they’ve been doing on providing walking and cycling access over the Harbour. While subsequent media coverage focussed on the price tag of potential one-off events and an inadequate ferry proposal, the information also included Waka Kotahi’s conclusion that:
Undertaking a “trial” as the Minister has requested, could set unrealistic expectations around the likelihood of a dedicated lane becoming available in the short term. Previous analysis demonstrates that unless a significant reduction in traffic demand occurs, the network is unable to support dedicated walking and cycling access on the bridge.
This post takes a look at some of those documents to assess Waka Kotahi’s claim.
Averting Risks or Hunting Ghosts?
The Auckland Harbour Bridge Shared Path Options Assessment produced by Resolve Group in August 2021 – and referred to in this post as “the Resolve report” – contains a preliminary investigation of different options for providing active travel across the Harbour Bridge. These options included 11 potential configurations of the bridge, including single and double lane reallocation, as well as “twin paths” at the outer edges of each clipon and even central path options which would be accessed via the existing Sulphur Beach underpass and Fanshawe Street.
The Resolve report sets out to understand the risks of bridge space reallocation:
Understanding the residual risks associated with each option was paramount in the production of this report, and every option has effects on AHB users, motorway and local road networks and changes to existing loading patterns of the AHB. This report did not aim to mitigate these risks, but rather, understand them for future option consideration and development.
The Resolve report needs to be interpreted with this in mind: its purpose was simply to identify the problems, not to investigate solutions. Further, the conclusions later clarify:
Due to the high level analysis performed by this assessment, all consequences and risks presented portray a “worst case scenario” for each option.
Most of the risks identified can be dismissed or easily mitigated. These include:
- the risk of litter being dropped off the bridge (clearly this risk exists in cities all around the world, and already exists in the context of the Auckland Harbour Bridge, as vehicle passengers could discard litter out of their windows too);
- lane sizes of differing widths on the bridge (which is already the case) and;
- privacy concerns around CCTV cameras (long since overcome at public transport stations and other public venues).
No risks or worst case scenarios are presented for Option 11 (no change to the existing set up), which ignores that the lack of active travel options over the bridge poses climate, health, environment, resilience and safety risks, as well as accessibility and equity issues.
The Resolve report concludes
All options carry some residual risks, however most of these risks can be mitigated to reduce likelihood and/or consequence to acceptable levels.
Despite these conclusions, an internal Waka Kotahi paper subsequently presented to the organisation’s Investment and Delivery Committee on November 23 (“the I&D Paper”), states that
the only viable option, that would not create operational or structural issues or cause significant traffic congestion (without a large mode shift) would be temporary access arrangements
that is to say, Sunday-only events.
It’s worth highlighting here that both Mode Shift and Transport Choice have been absolute top priorities for Waka Kotahi under the Government Policy Statement on Transport since 2018.
It’s unclear why the Resolve report concludes that all options are possible with risk mitigation, but the I&D Paper states that only weekend options are viable. As all of the risks could be mitigated, which were considered unacceptable? The I&D paper suggests three that are worthy of a closer look:
- Network impacts of traffic congestion, owing to a reduction in bridge motor vehicle capacity;
- “Operational issues”, including health and safety of motorway and shared path users;
- Load bearing capability of the clipons and the weight of suitable crash barriers.
For today’s post, we’ll look at the first of these, and explore the traffic demand modelling.
Lies, Damn Lies and Modelling
The Resolve report models traffic impacts for each possible lane configuration. A number of different models were applied, to assess the AM and PM peaks.
We need to know the inputs – the numbers assumed for each mode of travel – and how the modelling assumed people would change their behaviour if the configuration of the bridge changes. What mode shift and traffic evaporation assumptions were used?
The report suggests these numbers are included in the appendix, but this portion of the document is missing – the numbers are nowhere to be found.
Thankfully, the documents also include a June 2021 Traffic Impact Assessment (“the TIA”) conducted for NZTA by Auckland Systems Management Alliance (ASM), titled Auckland Harbour Bridge Active Mode Provision. It is in the TIA that we can review some of the numbers that fed into the modelling which underpins both reports.
The figures for mode shift to walking and cycling are derived from three reference points.
The first reference point comes from modelling performed by Flow Transportation Specialists (“the Flow report”) to investigate demand for walking and cycling across a potential future harbour crossing. The latest modelling suggests 5,000 pedestrians and cyclists would cross a newly constructed bridge set to open in 2028.
It is not clear, therefore, why the TIA and Resolve report modelling has used a significantly lower figure of only 1,600 daily pedestrian and cyclist trips. (The report refers to “commuter” trips but models them as happening throughout the day, not just at commuter peak hour).
The Flow report also models annual average daily trips, where (as we’ll see later on) the TIA compares weekday traffic.
When we look at existing cycling corridors like the Northwestern Path, we can see that weekday cycle counts are much higher than weekend cycle counts – by a factor of 2 or more. An annual average daily count for the bridge, then, will be much lower than a weekday average.
The TIA contains two other reference points.
An Auckland Forecasting Centre estimate of 2,000 cyclists and 500 pedestrians per day (2,500 active modes trips/day); and numbers of cyclists from the Burrard Bridge in Vancouver – 1,000 per day when it opened, and 3,300 per day by 2017.
But again, these numbers are average annual daily counts, and likely underestimate weekday usage. There’s a figure missing, too: Burrard Bridge has always had pedestrian access, which in 2018 moved approximately 2,000-3,000 people per day – giving us a reference range of 3,000-6,300 active modes trips per day from Burrard Bridge.
This means the lowest reference estimate for Auckland is 2,500; and the highest is 6,300.
It is unclear how the TIA model arrives at 1,000 to 3,000 walking and cycling trips per day as inputs for modelling.
For Public Transport modelling, it appears as though an estimate was calculated with the Macro Strategic Model, the output was regarded as “counterintuitive”, and the numbers subsequently used bear little relevance to the source material.
The MSM output for 7-lane options (presumably in response to differences in congestion) suggests that nearly six times as many people will catch public transport home in the evening as will have taken it into work that morning.
The discrepancy is noted, but no other modelling or working is presented, and four different mode shift figures are eventually utilised: 2,000; 4,000; 7,000 and 11,000 people shift from their car to public transport daily.
Changing Trips: New Routes, New Times of Day, and Traffic Evaporation
The Resolve report identifies that
some customers affected will choose to modify their trip behaviour to avoid congestion and delays. This could include choosing an alternative route… retiming their trip… cycling or walking… utilising public transport, undertaking a different trip… or cancelling their trip altogether.
These changes can have positive impacts on the wider network.
The TIA uses two different models – MSM and NCI SATURN – to estimate the number of vehicles which will take an alternative route, and arrives at a vehicles per hour range from fewer than 100 (based on a 7-lane configuration) through to 650 (based on a 6-lane configuration).
However, no estimates were given for how many trips across the harbour people shifted to other times of day, replaced with more local trips, or didn’t make altogether.
The Results of the Modelling
The Traffic Impact Assessment contains a table representing 4 different demand models:
- 7-lane option, low demand reduction (i.e. a low level of mode-shift)
- 7-lane option, high demand reduction (i.e. a high level of mode-shift)
- 6-lane option, low demand reduction
- 6-lane option, high demand reduction
Of particular note here are the peak hour numbers. The table shows that of 1,000 active modes trips, only 42 (or 1/24th) will be made at peak hour. This number suggests the model assumes roughly as many people will be cycling the bridge at each hour of the day – including in the middle of the night (this applies to all levels except level 2).
The vehicle re-routing is interesting, too, as the majority of vehicle re-routing takes place outside of peak hour. However, outside of peak times the bridge is unlikely to be at capacity and rerouting would be unnecessary.
A Preferred Option for Walking and Cycling
Of the 11 options modelled, one appears particularly realistic: Option 3 – a 7-lane configuration with the easternmost lane converted into a shared use path.
Option 3 produces low and high demand reduction models which look like this:
And when the report overlays weekday traffic demand to evaluate impacts (with the level 1 – low demand shift model), we get this:
All in all, the modelling suggests the difference between the available bridge capacity and remaining traffic volumes under Option 3 is fewer than 1,000 vehicles per hour.
Is this significant?
Well, if we consider that the model:
- Significantly underestimates active modes demand;
- Compares average annual daily walking and cycling trips with weekday traffic demand;
- Assumes active modes trips are equally distributed across 24 hours;
- Does not account for trips made at different times, more locally, or not taken at all;
Then it is entirely possible that a more realistic model might not only fit comfortably within the available bridge capacity, but that it may result in better traffic flow than occurs on the network presently, owing to an overall reduction in network traffic.
All models are wrong, but some are useless
There is, of course, a very robust way to get real-world data of what would happen if a single lane on the Auckland Harbour Bridge was made accessible for walking and cycling: Waka Kotahi could fulfill the Minister’s request and trial it – and with current low traffic volumes, now would be the ideal time.
With the release of the latest IPCC report into the ongoing and worsening effects of climate change, by far the greatest risk would be failing to take advantage of every opportunity to promote mode shift to sustainable, healthy transport.
Header image credit: Nabulen Photographer