I regularly keep track of a number of statistics about transport and one of those is traffic volumes from the NZTA. Recently I noticed an anomaly with the figures for the Auckland Harbour Bridge. Previously volumes were reported as:

  • Centre Span
  • Left Clip-On
  • Right Clip-On

The monthly data for March and the annual data for 2014 (released in March) was different, instead reporting just Northbound and Southbound traffic volumes – the annual data also included the clip-ons but not the centre span. That in itself isn’t such an issue however the total traffic volumes were quite different, even for previous months/years. An example of the difference is shown in the chart below of annual traffic volumes. You’ll also notice that the volumes are up slightly – although they are still less than they were in 2005 and in percentage terms is low considering Rapid Transit services like the busway are growing by double digit figures. The chart also includes the traffic volume predictions found in the most recent business case for another road harbour crossing.

AHB Annual Volumes to 2014

So seeking an answer for discrepancy I asked the NZTA why the figures were different. The answer is below.

The original site was a National Telemetry Site with loop detectors on the two clip-on sides and an infra-red detector over the four centre lanes. This equipment used on the centre span could not determine directionality and loops could not be used due to the steel deck (the clip-on counters are on the concrete deck north of the main span).

The Auckland Motorway Alliance (AMA) established a count site just north of the bridge some years ago to collect directional data, but it was noticed that the AMA counts and the Telemetry site counts were drifting apart. The problem was with the centre span equipment, which was missing more vehicles as time went by. Therefore, it has been decided that the data from the centre span counter was too unreliable to use.

The Telemetry site was life expired anyway, so the AMA site will become the new Telemetry site. I am told that the clip-on counters are still providing reliable data, so there is no need to decommission them.

That seems a pretty reasonable explanation however as the monthly data released so far only extends back to March 2014 I asked if any further data was available. What I received back surprised me. I did receive some extra monthly data but far more interestingly I also got two years of hourly data by direction – from 1 January 2013 to 31 December 2014. Below are some of the insights we gained from that.

First up results by the day of the week. I was quite surprised to see that traffic builds up over the course of the week with an average difference between a Monday and a Friday of over 15,000 vehicles per day. The busiest single day over that two year period was December 19, 2014 when over 200,000 vehicles crossed.

AHB Traffic by Day

Breaking that down further by time the table below shows that while Fridays have the highest overall volumes, the strongest peaks occur earlier in the week which I can only guess is due to a lot of people rushing to get home whereas on a Friday the peak is smoothed a little, perhaps from people leaving work earlier or staying at work longer socialising. You can also notice that the late night/early morning volumes over the bridge are much higher than other days of the week from people ou

AHB Traffic by Time and Day
Click to enlarge

Showing traffic volumes over the course of the average weekday we get the chart below. I was quite surprised to see that the afternoon peak was stronger than the morning peak.

AHB Traffic by Hour

The data allows us to break that down further including by direction

AHB Traffic by Hour by Direction

While volumes peak in the morning and afternoon I was interested to see how things compare on a per lane basis as the moveable barrier on the bridge means that in the peak direction there is an extra lane available. It is often stated that a single motorway lane can move about 2,000 vehicles per hour. As you can see the volumes on the Harbour bridge fall short of that and peak at around 1,700 per lane. It’s also interesting that at times when the bridge is in a 5-3 configuration that lane volumes are similar.

Note: I’ve estimated the times that the barrier is moved as I’m not 100% certain.

Lane Capacities

I suspect it will be very hard for the bridge to hit any maximum capacity as it is limited by the motorways either side of it. That is also one of the major flaws of any plans to build and additional harbour crossing. You’d have duplicate or at least widen much of SH1 to either cope with the volumes or allow the connections to be used to their potential.

Lastly it’s worth considering the role that buses now play in the Harbour Bridge. Over the two hour morning peak (7-9am) around 200 buses cross the bridge southbound yet they carry around 9,000 passengers which is well more than the bridge carries in an a single morning peak hour. That points to one of the big benefits of PT investment, it’s capacity abilities. By having a strong, congestion free route it allows us to take the edge off volumes and move many more people at a time they want to travel. Imagine the impact there would be if tomorrow all the PT users who currently cross the bridge by bus instead tried to do so by car.

Overall fascinating data so thanks to the NZTA for providing it.

Share this


  1. “Imagine the impact there would be if tomorrow all the PT users who currently cross the bridge by bus instead tried to do so by car.”

    I’ve often thought a PT free day or week would end the debate around whether PT was value for money. You hear endlessly people complaining about the subsidy that PT receives and how fares should be full recovery, without thinking of the benefit to traffic those subsidies provide.

    1. On a similar note, I think a “parking warden free day” might get people to realise that they actually do a useful job.

    2. Drivers’ strikes are very useful for this sort of analysis! Of course it is the busway that enables the Bridge and its approaches function at all. And so cheaply.

      The quality of the alternative Rapid Transit service sets the driving speed: http://www.smh.com.au/national/new-motorway-will-derail-commuters-20140217-32hvs

      In a rational world every step would be being taken to optimise the functioning of buses on this route before even one penny was spent was on looking at added more traffic lanes. The busway would be being extended, street priority for those buses both in the city and on Shore streets would be prioritised, ped and cycle access to stations would be direct, lavish, and complete.

      And then, of course, the third crossing would be being planned in he obvious form of the missing Rapid Transit route, leaving the bridge for what it was built for; road vehicles, and removing the buses. An efficient electric rail system connecting Takapuna and the current busway route to Wynyard and the CRL.

  2. Slightly off-topic, but weren’t heavy vehicles banned from using the clip-ons? I’ve seen heavy trucks using whatever lane they want.

    1. My understanding is that going north (i.e. laden – moving away from the port) heavy vehicles must stay in the centre lanes. Coming south (i.e. unladen – coming back to the port) they can use any lane.

      1. I thought I saw a sign forbidding trucks to use the center lane (3rd) when the bridge is in 5-3 configuration. I appreciate a truck breaking down in that lane precludes any attempts to remove it until the barrier is moved.

        Is that sign there as one heads southbound over the bridge with 3 lanes available?

    2. There are no restrictions on trucks on the harbour bridge, the clip ons were strengthened in 2010.

    1. A good idea but even now, unless there has been a crash, even at peak, the bridge flows pretty well.

    2. There almost is southbound. Buses use the left lane southbound, with a bus lane at either end. Fewer cars want to be in that lane. Northbound is not so good as there are no bus lanes northbound. I suspect this was to allow people to get to work on time, but less important they get home on time. Similarly Onewa Rd had a city-bound bus lane for years, but the return journey could easily take twice as long…. And the rest. The homeward bound bus lane on Onewa Rd had only just gone in.

      1. The biggest bottleneck for buses around the bridge at the moment is northbound Fanshawe St onramp to Curran St Onramp. The almost-never-needed southbound Bus lane on the other side of the motorway must be the stupidest thing ever. Why didn’t they place that bus lane on the northbound side instead, puzzles me every day I go past.

        1. I wish. I will write to NZTA and ask why it’s been done like that … i’m not holding my breath for any reasonable answer though.

  3. Would both spans be two-way or would one be for north and the other for south? If the latter, there would likely be about 100 years of capacity on each. And you do run into the problem of capacity of the motorways which would lead to huge back-ups. But putting the new bridge approaches roughly where the current ones are now would have the same effect. The perception would be that the new bridge was a couple billion dollars to make things worse (and it would be true).

    You could make one bridge PT-only but would volumes justify it?

    I agree that the current bridge should be maximised for transit. The fifth lane at the peaks would allow one to be dedicated PT and still four available for cars. It would be worth a try.

    1. No. This is a road bridge, and should remain so. Transit use, and therefore its relief to traffic demand, will be optimised by making the next route a dedicated Rapid Transit one. Rail tunnels. Much cheaper to build too.

      1. Aucklands busway has smashed the rail system in terms of performance. Why must it be replaced?

        Busses are faster than the trains we use, they can branch out from the busway (greater catchment w/o transfers )and can take detours.

        1. That’s not really true. The busway performs about as well as the southern line for patronage, and about the same on speed.

  4. Great post and I love that heat map.

    As you point out, Onewa is the southbound bottleneck and the whole CBD is the northbound bottleneck. Oneway is naturally holding vehicles back from reaching the bridge so I don’t really see that much growth occurring on the harbour bridge in future. I would just point out that with a very steep incline you will never get max cars per lane because aren’t travelling as fast but are also not bunched up as much.

    I do think they need a northbound bus lane though.

  5. Excellent graphs. The last one shows the available useful capacity between the morning and evening peaks. As demand grows from around 1300vph to 1600vph in those hours we will see why more capacity will be needed or why the existing capacity will need to be reallocated either through occupancy restrictions or pricing.

  6. If that new green line continues at the same rate (no guarantee of course) it looks like around 2045+ that a new crossing is needed – maybe later. So in reality we shouldn’t even begin building until 2040. And if we build rail to the shore then probably delays it even further if not all together.

    1. Exactly. The costs of any additional crossing are huge and can be delayed by finessing the buses so much more. However it is clear that the missing trick across the harbour (after the Active modes which are about to be added) is a dedicated Transit route not additional traffic lanes which just cause huge additional costs and burdens on the road networks and land use on either side.

      And a pair of rail tunnels will provide a service so direct and fast between key city and Shore points as to be irrestable for thousands, keeping the bridge at an optimal utility indefinitely.

      1. Plus, half way between the end of the green line and the edge of the graph, we will see the effect of the Waterview connection to the second harbour road crossing (that we already have)

  7. How much would it cost to convert one of the clip-ons to a shared tramway/bus lane with trams running along water to Takapuna and up Onewa road to Birkenhead/Beach haven. All up cost including trams into city and to B’head and Takapuna? Would be interesting to know how much traffic this would remove from bridge. Tram network could then be progressively expanded up the northern busway.

    1. Please don’t say trams when you mean light rail/metro. It would be furiously expensive because politically it won’t happen without the tunnel being built.

        1. If you’re running it up the middle out of traffic then generally it’s light rail, and on the busway would be metro according to most definitions.

          It’s important because people think of trams as slow and in traffic.

      1. I guess I am thinking that it would be a lot cheaper than building a tunnel. I don’t think we can afford a 3rd harbour crossing of any type as a city without sacrificing too much else in the budget. When I drive over the bridge every day and it is by far the least congested part of the motorway network (80km/h every morning with no problems), I think we should look at turning over some of the lanes to Rapid Transit which would reduce number of people in cars even further so shouldn’t cause too much pain. Would basically go to 4+2 lanes in peak instead of 5+3.

        Rapid transit could be a mix of trams for local places like Takapuna and Northcote/Birkenhead and more high speed light rail for up the busway.

      2. What is the difference between a tram and a light rail? My understanding was that it just depends on the frequency of stops. Many stops in the central city, it’s a tram, heads over the harbour bridge not many stops, it’s light rail, many stops in takapuna, it’s a tram again.

        1. No hard and fast definition, but generally “trams” are smaller roughly bus sized vehicles rolling along the street in mixed traffic. They tend to be slow, fairly low capacity and unreliable.

          Light rail tends to be much longer vehicles (roughly train sized) that run either entirely off street, or if they are on street they have dedicated separate running way, priority or grade separation at intersections, etc.

  8. I was wondering about the capacity of the northern busway. You mention 200 buses southbound during peak hour. How many of those come from the busway?

    On apps like track-my-bus, you see about one bus every 2 minutes passing by Akoranga station. 30 buses in one hour with 45 passengers means about 1350 people per hour using the busway (one way). That’s way less than the capacity of a general traffic lane. Maybe is that app not showing all services? Since the frequency is so high during peak anyway so at that point it’s not much a problem if it misses a few buses.

    Or is each motorway lanes indeed carrying more people than the busway?

    1. If you read it closely, it’s 200 buses over two hours. So more people on buses in the peak direction in two hours than in cars in the peak direction in one hour. But still more in cars (across five lanes!) per hour than the buses take in substantially less than one lane.

      1. I was referring to the busway between Akoranga and Constellation Road.

        Suppose you’re standing on the Aroranga Station platform. Or on the overpass to the AUT campus.

        You’ll see only the buses coming from the north via the busway. You also won’t see the buses coming from Takapuna via Esmonde road. Further north some of the buses also go off the busway to Takapuna.

        So over 2 hours:
        around 60 buses pass by. Those can carry around 2700 people. Is 45 people the estimated capacity of a full bus, or a packed-like-a-sardine-can bus?
        On the motorway, a lane can carry around 4000 vehicles in theory, carrying around 4800 people. Probably it’s not possible to reach that capacity, but it’s a large difference.

        So standing on the overpass, the general traffic lanes you see are carrying more people than the bus lanes next to it. Or am I still missing something?

        Of course the buses are going way faster than the cars, having that option already makes it worth it. And there’s still lots of room for more buses.

        And the park & ride in Albany and Constellation Road have a total of 1470 places, enough to fill a substantial part of that capacity. So are half of the people on the buses you see passing by, coming from a park & ride? Or am I underestimating how much of those people go to other places on the Shore?

        1. First of all you are way off on your vehicle capacity estimate for the motorway. The theoretical limit is around 2,000 vehicles per hour but in practice it seems to top out a bit over 1,800.

          And the actual data: Over the two hours of the morning peak (7am to 9am), 167 buses leave Akoranga southbound on the busway (a handful of those from Takapuna wasy don’t actually stop at the station however) carrying 6,868 passengers (on average last year). A further 46 buses join from Onewa Rd heading south too, with a further 1,591 passengers.

          So the busway proper carried 3,434 people an hour in 84 buses, or 41 people per bus on average (average, some are a lot more crowded, others less). So around Akoranga the busway does the work of two lanes of motorway inbound (i.e. you would need to build two extra lanes each side to carry traffic equivalent to the number of bus passengers), while the motorway itself actually has three. So yes, the busway is “only” carrying 2/3rds of the number of people as the motorway next to it.The whole north shore citybound bus system crossing the bridge carried 4,230 people an hour on 107 buses, or 40 people per bus on average. So crossing the bridge the buses are doing the work of three lanes of traffic, while the bridge has five southbound. So it is doing 3/5ths of the motorway next to it.

          In rough terms busway patronage would have to grow by about 50% for the busway to carry more people than the motorway next to it. However at the current growth rate (pushing 10%pa) that could be achieved in three or four years, if traffic didn’t grow much.

          Heading south from Constellation on the busway we have 4,831 passengers before 9am. If the park and ride holds 1,470 vehicles and we assume about 1.2 people per vehicle, we have the park and ride supplying just over a third (36%) of those commuters who are heading to town before 9am from the upper busway. However across the whole day and all the busway, we have some 9,570 passengers heading over the bridge. So the park and ride supplies about 18% of busway passengers, while the other 82% arrive by bus, cycle, foot or get dropped off.

          Note that all of this excludes bus passengers who don’t head across the bridge, i.e. anyone using the busway system to get around within the North Shore only. It also assumes that people only park and ride to head to the city.

        2. Thanks for the info.

          So in a few years the benefits of the busway, in terms of extra capacity, will exceed the potential benefit of new extra harbour bridge.

          The bottleneck on the motorway southbound is the 3 lanes going under the Onewa Road bridge. There’s one additional lane coming from the Onewa Road onramp. So actually there’s only 4 lanes feeding into the bridge, which explains why the bridge itself (5 lanes) has a lower throughput per lane. Now, the really bold move would be making one of those 5 lanes a bus lane.

        3. Thanks Nick
          Do you have stats for the other Northern Busway stops? (Smales farm, Sunnynook and Constellation) I am most interested in Sunnynook as that is beside a two lane motorway.

    2. Lol, 45 people on a bus would be a dream come true, the NEX buses are pretty much all full at peak, which I think is over 75. All of the Onewa Road buses, and the 881 are the same.

      1. My own experience at Smales Farm also confirms the packed-like-a-sardine-can metric is more useful during peak hour.

        7500 people would handily beat the capacity of a motorway lane, more in line with what I was expecting. It also means the estimate of 9,000 passengers over the bridge (45 per bus) is probably too low.

    3. Also, I count 105 total timetable buses using the busway across AT’s timetables between 7 and 9.

  9. The link on line 1 of Matts connects to NZTA data. Some of the volumes of daily averages are for buses on the bus way. (pgs 22-25 of Data Booklet – 2010 to 2014 )These are (for 2014) Mclymonts Rd on ramp, 241, Constellation Bus on ramp Southbound, 241, Northern Busway both ways 1km south of Upper Harbour Drive, 502, Akoranga Bus Station on ramp SB 502, Esmonde Road on ramp SB 100, and Northern Busway at Onewa 502.

    Matt Land Nick R is it possible to find out how you got the one hour harbour bridge breakdown and Northern busway data respectively?

  10. Yesterday 2000+ cyclists rode across the bridge and back – it took me around an hour cruising. We used two lanes. Waka Kotahi “closed the lanes due to event…”. It seems that if WK wants to DOUBLE normal capacity – it should “close” more lanes more often. PS – i highly recommend riding the bridge. An Asset to Auckland.

Leave a Reply

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