Follow us on Twitter

Speed versus throughput

Somewhat picking up on the point I was trying to make with yesterday’s post regarding how the biggest benefit of the City Rail Link is how it simply massively increases the capacity of Auckland’s transport network, particularly the capacity of the network for trips to the city centre, here’s an interesting article that analyses the issue of whether transport network improvements should be designed around making trips faster (time savings benefits) or whether they should be based around simply increasing throughput:

There’s an old joke in computer science circles: never underestimate the bandwidth of a truck full of tapes. That is, if you want to move a large amount of data from one point to another, but you don’t care how long it takes to get there, the most efficient way to do it is to put the data on a removable media (magnetic tape in the old days, hard drives or DVD-R’s today) and drive them across the country. The freeway system becomes a high-throughput, albeit high-latency, data network.

Exactly the same point applies when you’re comparing a freeway to a grid of city streets. A couple of commenters replied to Sunday’s post by claiming that freeways are essential for moving people and goods into the city center. This argument confuses speed with throughput.

For throughput purposes, what you care about most is the number of lanes connecting any two points. And for a city grid, this number is huge. Central Philadelphia’s surface streets have around a dozen lanes in each direction. This is enough lanes that city streets tend not to be major bottlenecks during rush hour. You’re more likely to find congestion on the freeways themselves, or at the points where the freeways intersect with the street grid.

So urban grids provide plenty of throughput. What they don’t provide is speed. At non-rush-hour times, freeways shave 10 or 20 minutes off the time it takes to get from an outer suburb into the city center. But from an economic perspective, it’s not clear how important this is. First, as already noted, those time savings come at the expense of users of other modes of travel who suffer from a less-compact (and therefore less walkable and bikeable) urban core. And second, these time savings almost disappear completely during rush hour, when freeways don’t move much faster than city streets.

Finally, while we’re talking about throughput, it’s important to remember that a subway line offers dramatically higher throughput—between 5 and 15 times, depending on your assumptions—than a freeway lane. It would be essentially impossible to have a city the size of New York rely primarily on freeways to get people in and out of downtown.

Freeways are a great way to move people around the suburbs, and to move people from one metropolitan area to another. But they’re a poor way to move people into, around, or through the urban core. And it was a huge mistake to destroy thousands of homes and businesses in cities like St. Louis and Minneapolis to make room for urban freeways.

Essentially, we have an existing rail network that has heaps of capacity – just constrained by the Britomart bottleneck. Unless we fully utilise the full capacity of the rail network then we’ll place a huge burden on our roading network to “feed” the city centre in the years and decades to come. Building motorways, widening roads, adding super-expensive road tunnels and so forth can’t really provide more throughput to the city centre, unless we plan on demolishing half the place to widen roads and build more parking lots. Even the “speed advantages” that may arise from constructing something like a second harbour crossing would, as the article above notes, achieve little at peak times when inevitably the road would become congested (thanks to induced demand).

Whether or not the City Rail Link “takes cars off the road” is actually irrelevant. What matters is that it dramatically increases the transport ‘throughput’ of the city – particularly (although certainly not solely) along the most congested part of the network: to and from the city centre. Due to this increase in ‘throughput’ it suddenly becomes possible for the city centre to grow and develop, just as (on a different scale of course) New York City’s rail network has enabled that city to grow to such an extent.

16 comments to Speed versus throughput

  • Matt L

    Except in the case of the CRL it will actually boos both speed and throughput from the West, even at peak times :-)

  • This analysis gets to the heart of the disconnect between the proponents and opponents of this project: For those whose assumptions and hunches lead them to feel that the CRL won’t be used much, especially won’t get anyone out of a car this simply reflects their conception of Auckland as an entirely suburban place, either in fact or as an ideal. Other countries have cities, are urban, but that’s not who we are. They disregard the CBD as either unimportant or at least as a largely undesirable anomaly that is to be suffered but not in anyway valued or encouraged, and especially could never be something to be proud of. This is what people mean when they say the AK doesn’t have the population, density, or form to support a real transit system.

    Including; the PM ['it will only take people who are currently using buses']; The Herald’s John Roughan ['sprawling suburban paradise']; Joyce and clearly the Wellington based MOT; and many ‘right thinking’ Aucklanders. For these people it is obvious that it is a waste of money to invest in something that they personally don’t see themselves using and that only would interest those they either frankly disregard [the PM's bus users] or are actively suspicious of [Roughan's 'planners'].

    This is further proven when if you can get any engagement with these people they insist that the only terms of evaluation is whether this project might get some current road users off the road network and out of the way of those good people who will of course only ever drive. And even when this is proven they still won’t really budge as they can’t imagine using it themselves and it still doesn’t fit with their suburban ideals.

    Many of these suburban minded opponents are in positions of power but, interestingly, are increasingly in the minority. This is a fascinating battle for the future of the quality of Auckland both as a place and as an idea.

  • Cam

    Well said Patrcik. It is indeed a battle, an inceasingly exhausting one. Let’s hope it’s not the people you mentioned above who come out on top.

  • Peter in Sydney

    Perhaps the question of whether the CRL will get more Aucklanders out of their cars can already be seen by looking at the current statistics. Can any of the regulars to this site say if the Auckland public transport usage overall is rising faster than the population? If so then there has to be a mode shift out of cars. Hammering this point to all those who expect a clear road to wherever they wish then will reduce their opposition to the CRL. Time savings on the western line along with greater frequency on all lines after the CRL is built must encourage an even greater mode shift to PT.

    • Not quite my point but yes, it is clear that the PM cannot do simple arithmetic- or rather doesn’t feel the need to bother with the facts on this issue- as BOTH bus and rail use is consistently up, and in the case of both rail and the one route of grade separate bus way dramatically so, therefore it can not simply be a case of people who used to ride buses now using trains. More people are using both more often, even though neither yet offer a very high quality of service.

      With regard to population rises, well not much of that now…. lots more heading your way Peter.

    • Matt

      Patronage growth is absolutely in excess of that of the population. We’re consistently seeing double-digit percentage increases on a “this month last year to this month this year” basis.

    • Matt L

      The last few years have seen overall patronage increase by between the 4% and 8% per annum with it at the top of that range at the moment. By comparison the population has kept increasing but at around 1.6% per annum and has been pretty steady at around that rate for a while. Because PT has been coming off such a low base then it is taking a while for it to have a large impact however the last few years have seen the average trips per person increase from around 41 to over 44.

      The key will be to see if we can sustain these patronage increases for long period of time, if we can then it will likely make quite a difference to roads.

  • Nick R

    Just saw a letter to the editor in the Herald claiming that “even if every western line passenger heading to the city used the rail tunnel it would still cost $24 per person for thirty years”.

    Kinda shows the public misconception about the CRL* , with the same idea that its only about taking one small bunch existing trips and making them a bit faster, rather than massively increasing throughput.

    *Five major issues there:
    1) the idea that it is only for the western line, even though the other four lines/branches will use it too.
    2) that it only for existing lines, no concept of it being necessary for new lines to be built.
    3) that it is only for improving existing trips, rather than allowing/generating patronage growth (and they are talking about 30 years FFS!).
    4) that it only benefits the CBD, rather than allowing a massive boost of frequencies right across the network.
    5) that spending on public transport infrastructure is a ‘subsidy’ while spending on other infrastructure is presumably an ‘investment’.

    • Feijoa

      Hit send to editor yet? It would be interesting to point out the cost of alternatives to add the same capacity into the city. How much would it cost per existing commuter to build motorway tunnels out everywhere?

  • Ari

    The same thing could be considered in regards to this whole fibre broadband project. What is the point in having 20Mbit connections to the outside world, when we still have to pay huge sums to use data. It’s like having a 10 lane motorway but with ramp signals that only let on 1 car a minute. Overseas they have speed and capacity. Here we have neither and we are investing into speed and doing very little about capacity.

    In the same manner, the CRL will dramatically increase the capacity of the network as a whole to pick up the growing demand of users. I doubt it will affect travel times anywhere off the rail network and it won’t take any vehicles off the road, but over the long term it will stop travel times from getting worse and give people other options.

    • Ari, I’m certain it will take vehicles off the road, but more importantly allow the roads to function for much longer at pre-critical capacity. So its’s not like it will clear roads of traffic but it will keep the levels to good levels. It is absolutely essential for keeping the widespread road network functioning well. Or to put it another way; to enable the city and its economy to function and grow better than it otherwise could. Those new train and bus riders have come from somewhere….

      • Nick R

        I think it would be a case of taking the edge off congestion by lowering the ‘equilibrium’. Assuming many people will shift to rail if it is substantially faster than driving then road speeds will improve a little, as you effectively lower the comparative level of delay that people are willing to put up with before they don’t bother travelling. However as long as driving is cheap and freely accessible then there will always be congestion at peak times. This is visible on the Northern Motorway, traffic levels are basically flat but bus patronage has skyrocketed.

        Patrick, while new train and bus riders do have to come from somewhere there is always the possibility that people will simply travel more frequently, longer distances or make trips they otherwise wouldn’t do at peak time because congestion was bad, so it’s likely that most of those vacated spots on the motorway will be taken up again.

        So maybe a mild easing of the worst of peak traffic congestion, although at the end of the day we are talking about move tens of thousands extra people during peak times without going anywhere near traffic. So indeed who cares if it doesn’t take any drivers off the road, the real value is much improved non-road travel and preventing an increase in demand for more road travel.

        • Hi Nick, yes as I noted in my first comment I see the ‘taking cars off the road’ obsession as largely coming from the those that just don’t get cities at all anyway and what makes them work, and I am not particularly fussed, but that doesn’t mean there won’t be impacts for road users. Yes you are right, the CRL, on top of the new trains, integrated ticketing, and higher frequencies will mean increasing numbers of people choosing to take the train. It is hard to quantify how many of these people would have stayed at home instead, but this seems to be the point. If the choice, especially for travel for work, is between PT or not getting to that job, then it is an absolute matter of urgency that we improve the effectiveness, regularity, and ubiquity of quality PT for the economy’s sake. So these trips are even more valuable than those that substitute existing road journeys. Yet another way that the anti lobby has the world backwards.

          Remember the cost disincentives to driving are seriously high for many and constantly rising, for me it isn’t just speed, it’s frequency and reliability but also spread and network reach.

          • But… the low levels of PT use in AK are not ‘cultural’ but rather the result of the absence of much good service- this we know because of the increased uptake everytime physical and service improvements come online, so I really do expect a not insignificant number of people to leave there cars when given the opportunity. Me for example. Especially as there is such a long way to go to reach per capita levels that are common in comparable cities. And- this is clear- this is the way the zeitgeist is heading…. Only thing that can stop it is arresting the improvements.

          • Nick R

            Indeed, what I would see happening as we improve PT is traffic levels staying much the same while new and future travel demand is met with public transport trips. If we take fifty thousand people off the road and onto rapid transit then that is fantastic, 50k people going about their day efficiently without exposure to congestion. 50k more people who can work or live in a given area. But I just don’t think that the roads would then be clear and free flowing, as the large demand to drive at peak time would soon mean traffic hits its equllibrium state.

            The ‘natural’ state of peak traffic isn’t free flowing with spare capacity and reliable trip times, it is at a level of congestion that is enough to start dissuading people from driving. As long as we have mass ownership of cheap imported cars, relatively cheap fuel, unpriced roads and easy registration/licensing/ownership there is (for the foreseeable future) always going to be huge demand for driving at peak times, and folks are going to keep doing that until congestion starts to stop them (or some other factor like fuel prices does).

            As far as I’m concerned the CRL can still be an extrememly worthwhile piece of infrastructure if it doesn’t take a single car of the road. They need to look at what it affords to travellers, not the remedial benefits to existing road users.

  • Patrick R

    Freedom from Mandatory Motoring.

Leave a Reply