It’s quite easy to forget that a couple of years ago Auckland’s regional and local government structures were turned upside down.

Since then Auckland Council seem to have settled into their work with relative aplomb. Nowhere is this more apparent than in their strategic documents, such as The Auckland Plan (TAP) and the City Centre Masterplan (CCMP). Both these plans identify a number of fascinating projects that Auckland Council considers necessary to transforming Auckland into one of the world’s most-loved cities (NB: In my heart that title is currently held by Istanbul).

View of Istanbul (Katakoy) from a ferry on the Bosphorus.

View of Istanbul (Katakoy) from a ferry on the Bosphorus.

Having delivered a number of strategic plans the Council must now consider how we get from where we are today to where we want to be in the future. Council itself controls land use outcomes and its thoughts on these issues will be outlined in the forthcoming draft Unitary Plan. Council also set budgets for each of their CCOs, such as Watercare and Auckland Transport, which will tend to define the rate at which they can make changes to the city.

To me this is where things get interesting: The interface between strategic outcomes, available budgets, and immediate priorities. Most of us will readily agree on the need to transform Auckland from a car-dependent, anti-pedestrian city and into a more diverse and dynamic urban environment that balances the needs of all transport modes. Where opinions are likely to diverge is on 1) the best path to get there and 2) how much money we should spend along the way.

From where I’m sitting (Brisbane, in case you’re wondering) few of the ideas advanced by the Auckland Council have caused such stark differences in opinion as their proposal to two-way Nelson and Hobson Streets. These one-way arterial roads are effectively tentacles for the octopus known as the Central Motorway Junction (CMJ). Hobson/Nelson streets allow vehicles exiting/entering the CMJ to gain rapid access to most parts of the city centre. In serving this purpose, however, they have effectively killed vast areas of the city centre. If one was trying to be level-headed and non-sensationalist about these things then you could describe them as nightmarish abominations that blight the entire western edge of the city centre, as illustrated below.

Nelson Street – from off-ramp
 Hobson Street - approaching on-ramp

Hobson Street – approaching on-ramp

Both streets are obviously designed to carry shed-loads of cars. But how many cars do they actually carry? Well, if one considers Auckland Transport’s count data then the answer would be “not many if any”. For example, the graph below shows the number of vehicles using these streets in the AM, midday, and PM periods in 2005 (source data). When looking at these graphs it’s worth keeping in mind that the capacity of one vehicle lane is approximately 2,000 vehicles per hour.

Traffic volumes

The first thing to note is holy flaming wombats; both Nelson and Hobson Street appear to be vastly under-utilised. These 4-6 lane monsters are shifting only 650 vehicles per lane-hour in peak periods, i.e. less than one third of their theoretical capacity.

But on the other hand, one of the main advantages of a one-way street is that it increases intersection capacity by reducing the number of conflicting movements. The signal cycles that result are more efficient, insofar as they are simpler and shorter – with more of the available time assigned to green phases (which could possibly include pedestrians, even though this is not currently the case). So even though the street capacity is under-utilised, the one-way system may still reduce delays at intersections. Another complicating factor is that any change to these streets may impact on the CMJ and thus on the wider state highway network. So whatever happens to Nelson/Hobson must first pass muster with NZTA.

So what might we do about all this? Well, first let’s define the problem more carefully: The main problem with Nelson/Hobson Streets, from what I can tell, is that the current configuration destroys amenity and undermines surrounding land uses. Vehicles on these streets really are moving too fast and too close to the footpath. It’s a classic case of the medicine (transport capacity) killing the patient (the city centre). Therein lies the strategic issue Auckland Council wants to address.

But does a lack of amenity mean that should immediately move to two-waying those streets? Or are there more incremental steps that might address some of these issues? One incremental option that I think is worth investigating in more detail is retaining the one-way streets (for now) but changing the allocation of space within the streets, or at least key segments of the streets. This could look, for example, to develop low-speed access lanes separated from the primary through-lanes by a hard berm. The purpose of the access lanes would be just that: To provide improved access to adjacent land-uses, while reducing vehicle speeds close to the edge of the street.

Such access lanes are relatively common in European cities. For example, the image below shows Rooseveltlaan, Amsterdam. This street was just down the street from where I used to live and it provided access to the E19, which is a major highway linking much of the Netherlands, Belgium, and the north of France, i.e. it’s a pretty important arterial road.

Rooseveltlaan, Amsterdam

The image below shows a cross-section of Rooseveltlaan. Starting from the right, we have the access lane running in front of the adjacent buildings. This lane is separated from the through traffic by a hard berm. In typical awesome Dutch style they have also included some gardens and trees, a good shared pedestrian/cycle lane, and a median bus/tram lane. One of the great advantages of this type of street is that it separates all of the slow-moving traffic that needs to access the adjacent properties from the faster moving through traffic. By removing this side friction the capacity of the central lanes is higher than what it would be in a situation where all the traffic was mixed together.


NB: If you are interested in transport/land use integration then you may want to have a look at this recently released NZTA research report (I was one of the authors). Page 37 onwards considers the role of street networks and a range of opportunities for better transport/land use integration.

Using Rooseveltlaan as my inspiration, I then wondered whether we could adapt these ideas to create a new cross-section for Nelson/Hobson Streets. One possible cross-section I landed on, which seemed to meet the necessary capacity while greatly increasing space for access, is summarised below (where the elements are listed from one side of the street to the other). 

Nelson cross-section

By way of comparison, a few quick measurements on Google Earth suggests that Nelson/Hobson Streets are about 27-30m wide. When you consider the above configuration then you start to realise how much space is currently been wasted (yes, I think it’s fair to use wasted in this context) on Nelson and Hobson Streets.

Of course the access lanes would have to re-join the primary traffic lanes at some point. The Dutch manage this in various ways, of which a common approach is to connect the access lane to the side street via a priority intersection, which is then in turn connected to the primary street via a signalised intersection, as shown below. I particularly like this image because it shows 1) how the access lanes are useful for providing access for service vehicles, such as rubbish trucks, and 2) how Amsterdam has set differentiated speed environments (in this case 30 versus 50 km/hr) between the access lanes and the through lanes. It also shows some of the general attention to detail that the Dutch bring to their streets and urban places, such as using raised pedestrian tables on side streets, which have been discussed at length in other posts.

Rooseveltlaan side street interaction

But my key point is that before we start changing the underlying structure of the network, i.e. two-waying Nelson/Hobson Streets, maybe we should look at re-allocating space within those streets. After all, the one-way street system does provide greater intersection capacity (some of which could be reinvested in catering for pedestrians and cyclists). Also, Hobson/Nelson Streets somewhat naturally lead onto “one-way” state highway connections, i.e. on-ramps and off-ramps.

Ultimately I think Auckland Council and Auckland Transport have a tough choice:

  1. All-or-nothing – this option involves going straight to a two-way street. While this would provide greater amenity, it would also raise the hackles of NZTA and motorists in general. Personally, I think that if push came to shove then Auckland Transport would have their way – they are after all the relevant road-controlling authority. Perhaps most problematic is the fact that such a major change will take ages to implement and is likely to be tagged to the completion of the CRL. This in turn means the western side of the city centre would be left to rot for another decade; or
  2. Incremental steps – this option would seek to retain the one-way streets (for now) but commit to progressively re-allocating space within those streets so they better balance the competing need for access and mobility. The benefit of this option is that we could start immediately and hence realise some of the potential benefits earlier. It also would not hinge less on the CRL, which after all may be delayed. This is my preferred option, provided it is done in a way that supports, or at least does not compromise, the end-goal that Auckland Council has identified, i.e. two-waying these streets.

Probably time for me to step down off the soap-box so that some of you can have your (equally valuable) two-cents worth. Enjoy.

Share this


  1. I can’t see for the life of me how 2 two way streets tie back into the existing motorway on and off ramps short of rebuilding sections of the concrete collar or tunnels. Maybe have a gyratory for the block nearest the motorway – cheap and easy to do. Something for the short term?
    Longer term once CRL is in make the leap, shut the ramps totally (except maybe for cyclists and busses).

    1. From memory that’s been discussed know the past quite a bit. Basically Hobson & Nelson each link with the on and off ramps from SH1 & SH16 rather than one being the onramp for both motorways and the other the offramp for both motorways.

      As Hobson seems to get way more congested than Nelson there could even be traffic benefits of separating the onramp traffic across two roads.

  2. First of all, I’m not sure why on street parking needs to be retained along these roads but putting that to the side, I do think it would be useful to separate out the local access and even turning traffic from what is the ‘main flow’.

    My preference is for two way streets for a couple of reasons I will touch on later but If you want to keep the one way system then I think it is perhaps worthwhile looking at how the streets work. Firstly looking at Nelson St what is interesting is that both the SH1 and SH16 off ramps are only three lanes wide but yet are fed by a single lane off ramp so it is no surprise that in total they aren’t achieving much more than traffic movement than the theoretical capacity of a single lane. The three lanes at the end is purely there as a system to be able to store lots of cars so that they don’t end up backing up onto the main motorway flow. Yet as soon as those lanes hit Nelson St they move to a 5 lane monster which seems primarily about giving one lane dedicated lane for every type of movement. Perhaps it should just be left at three lanes (reducing to two as it gets further north) and have the turning lanes separated by a median like your examples above which also acts as the local access (in fact Nelson St already kind of has this on the western side of its southern most section).

    Hobson St is a different beast though and what I find interesting about it is how it often acts when busy. In the afternoons the traffic heading up the hill is often backed up over all lanes yet those going west will often reach the motorway and find it empty (for a while anyway). It seems that the major issue is that people turn out of the side streets or car parks and take up what ever space they can find and often that is the western most lanes. As they get closer to the motorway they then try and force their way into the stalled queue for the southern motorway and in the process block the lanes for anyone going west which quickly sees all lanes grind to a halt. In the case of Hobson there might be value in a median to separate out the motorway flows but I think the biggest issue in this case will be a political one as many people simply won’t understand how you could justify ‘removing busy lanes’.

    1. Drop the on-street parking if you like – I was simply trying to include a little “local” accessibility into the mix and activate the edge of the street. Even without parking you’re probably still going to want loading bays at various points. Interesting observations on Hobson.

      1. I was thinking about on-street parking when I was in Wellington over Christmas. There has been some controversy there over the new bus lanes through what used to be Manners Mall. People are getting hit by buses and this (or the resulting publicity driven by the anti-bus lane protesters) is apparently alarming the city council. I had a look at the new bus lanes and I think the fault lies with people stepping out in front of the buses. Pedestrians have to take some responsibility for their own safety and not walking out in front of fast moving buses, cars, or trains has to be part of that process. But… there is no on-street parking in Manners Street and the bus lanes are quite narrow. I think maybe people in city centers aren’t used to coming to a complete stand still before crossing the street. I think they take a quick look around for traffic and just step on to the street without pause. My theory is that a row of parked cars proves a barrier that buffers people from the traffic… a safety barrier, but also a barrier against noise and fumes and the wind suction created by buses and other vehicles. And the need to walk between parked cars before encountering moving traffic also gives people another three or four steps to check properly for oncoming hazards without needing to stop walking.

        There are other ways to achieve the same buffering and road-checking opportunity. But on-street parking pays a dividend to the council and may be a good thing.

        1. Agree with this. It’s not so important that it’s a car- but having a physical barrier and distance between peds and moving vehicles is critical for walking comfort. The main design rationale for the multi-way boulevard is that it concentrates faster moving vehicles in the central lanes and allows more nuanced and much slower access movements to occur next to the footpaths.

    2. I like the idea Stu. Parking in the access lanes makes sense to enable revitalisation of the area. As per the Netherlands, these could be 30 km/h shared lanes (not Woonerf or shared space – 5km/h) where they are for access by car but as part of a through route for cyclists?

  3. One point to clarify, the capacity of a motorway lane is around 2,000 vpd per hour. However a lane of city street is substantially less where speeds are lower and there are traffic lights and grade intersections.

    Hobson and Nelson would truly have to be motorway extensions to carry 2,000vph on each lane.

    1. Fair point Nick but doesn’t that just go to show how we currently have the worst of both worlds on these two ‘streets’: They have the quality and inhumanity of motorways but not the utility. Arrrrg.

    2. The theoretical free flow capacity is independent of traffic lights and intersections. So even in low speed environments mid-block capacity still approaches 1,500-2,000 vehicles per hours.

      Delays induced by intersections quickly drop this – but 650 per lane is certainly achievable, especially on a one-way street with low side friction (i.e. no on-street parking) and little local traffic (on the access lanes).

  4. these streets need to be considered from a strategic viewpoint, when Albert Street is out of commission while the CRL cut and cover is underway, Nelson, Hobson and probably Quay Sts will have to take some of the traffic that would have used Albert and Customs, so any substantial changes to these streets should be postponed until after the CRL work is substantially complete

    at the same time, the CRL work presents an opportunity to develop TDM and traffic management packages for during construction that would shift more CBD travel from car to PT, as such, it’s critical that interim traffic measures during CRL construction do not unessessarily inhibit bus travel to, from and within the CBD

    one of the unintended consequences of the busway construction was that the Shore’s highly successful bus shoulder lanes were closed and buses had to sit in traffic with all the SOV commuters, consequently the entirety of planning for the CRL, from beginning to end MUST be PT supportive

    1. As part of being strategic we should be expecting “evaporated traffic”.
      Here’s a headline from the opening of Octavia Blvd formerly an elevated freeway- “Traffic
      Planners Baffled by Success: No Central Freeway, No Gridlock, and No Explanation”

      1. to a degree you’re right Kent and the concept of dissapearing traffic has been well known and documented for nearly 20 years (notably “Cairns, Sally; Hass-Klau, Carmen; Goodwin, Phil (1998). Traffic Impact of Highway Capacity Reductions: Assessment of the Evidence”),

        nevertheless, there is a lot of legitimate goods and serives related traffic which has to get to its destination to support the economic activity of the CBD

        all I’m saying is don’t take any irrevocable steps like demolishing the Hobson St viaduct until a CRL construction mitigation plan has been fully worked through

    1. that kind of configuration can be found on most Southern European cities. Usually the lateral “access roads” become defacto shared spaces with provision for pedestrians, byciclists and local traffic.
      Sometimes though restaurants and bars put tables everywhere, legally or not, so cyclists have to go back to the main road.

  5. Thanks Stuart for your insightful and detailed commentary on Nelson & Hobson Sts.

    We (that is, Pippa Coom, Rob Thomas, and myself) have long advocated for what we call ‘two-waying’ of these streets, in order to create more livable streets. This idea has some backing from officers, but your post is a good reminder to see if we can push a little more on this idea. One of the sticking points has been NZTA – they have no money (their words), and would need some millions to change the layouts at the top of each street to accommodate connections to the motorways. Once we can resolve this point, I imagine the rest becomes relatively straightforward – essentially a re-assignment of road space, including bike lanes, and an upgrade of pedestrian facilities.

    Christopher Dempsey
    Waitemata LB

  6. Have AT or the Council actually assessed what would be required for the ramps? Are we talking millions, 10s of millions or hundreds of millions? No feasibility assessment, can’t take this seriously. If there has been nothing more than talk and concepts then how can we know this is nothing more than window dressing from the City?

    1. As far as I know, nothing more than an eyeball estimation. I will see if I can pin it down further.


  7. I have no idea what it would cost, not an engineer, are you? Looks though like it would have to be a tunnel not a bridge. The climb up then down, with the lower motorway than surface street would make the ramps steep and the crest of the bridge to pointy.
    That’s unless there could be a set of traffic lights at grade for the crossover bit. A computer model can tell us that.
    So, a tunnel starting from half way down neilson street then.
    Maybe an assessment is the way to go.

    1. The bridge would be an extension of the existing bridge joining to nelson. The link from Nelson would pass under it, as it has to drop lower to join the motorway already. In fact there might be an opportunity to simply go under the existing structure with the new on ramp, if the geometry were acceptable.

  8. Can I just ask why we need a direct link from Nelson St to one if the on-ramps? There is already queuing space on Union St that turns on to the ramps. Couldn’t Nelson St traffic just use that? With a bit of inventive phasing, the traffic could make it through there to the motorway on a single set if green lights, perhaps. Then there’s no need for expensive bridges at all.

  9. I think you’ve short-changed the pedestrians there Stu with 2m-wide footpaths but we can haggle over the details (a contraflow bikeway would be nice too). The “boulevard” concept has a lot going for it.

    You’ve picked up on the issue of one-way vs two-way, and it’s definitely not a simplistic “one good, one bad” issue as often painted by politicians/public; very much depends on the street environment and what objectives you are trying to achieve. Just FYI, we have a Masters student currently developing an assessment framework for considering one-way vs two-way street options; the aim is to produce a useful template that any Council could use to evaluate different options. Final report should be produced in a couple of months and will also be presented at the IPENZ Transportation Conf in April; happy to forward more details.

    1. Technically, he suggested four pedestrian spaces, as he noted that there should be a ped/cycle space next to each side of the central road. Though I think we should consider making that narrower but cycle-only. Give the extra space to the pedestrians, and allow faster cycling with less conflict with walkers near the central lanes. Copenhagen lanes of course, one-way, and one of them being contraflow.

    2. Yes on second thoughts I think you’re right. Take another 1m or so from the berm and whack it onto the side. Or you may find that you can remove on-street parking in narrow sections to provide more footpath width. Your contra flow bike lane idea is very a good one and yes please do forward details on your student’s research – I might even be attending the conference in Q’town so can catch up over an asparagus roll in person.

      1. If you’re in Q’town you might be alone – it’s in Dunedin this year! But yes, would be good to touch base if you are there; otherwise I can pass on the report/paper later.

  10. Why restrict ourselves to paralell arrangements?

    Why not have cycling on one side with access and angled parking on the other?

Leave a Reply

Your email address will not be published. Required fields are marked *