The salt-flat experiment: a case for 15km car-free urban growth boundaries

Since no one has the means of conducting this experiment, I will ask you to use your imagination. You (my reader) and I (Dr. Behooving) are standing at two randomly selected points on a vast plane, roughly the size of a pre-WW2 city; let’s say it is a 15km diameter circle. Beside you is a car and a bike, both ready to roll. You decide to come join me.


How far apart are we? So close that the logical thing to do is just walk? Far enough that the logical thing to do is kick the bike stand and ride? Or are you far enough away that it would be worth your while to get in the car and drive to my position?

For the sake of this argument, you’re not an enthusiast, either for walking, cycling or driving. You just want to get there, by any means. Neither do you have a physical impairment, like obesity, that makes walking or cycling unpleasant.

If you like, you might leave a comment at the end of this post and let me know how your guesses compare to those that I posit here. For a trip on a salt plane of less than 30 meters, I would be surprised if you didn’t walk. For trips of between 30 and 100 meters, I would expect you would cycle. For trips of 100 meters or more, I would think you would drive. Given the salt plane is 15km across, you would be driving on most occasions.


Next imagine you are not the only person wanting to move across this salt plane at this moment. There are as many people wanting to make similar trips as there are at any busy time in a low-density city. It’s still just a 15km diameter plane, but with thousands of people moving between random points. Can most of those people still choose to drive? 


From all our existing car-centric cities, we know reasonable flows can be maintained by herding the cars and using roundabouts or traffic lights to help them cross paths. At the risk of impeding the prioritised traffic we might even add some pedestrian crossings and bike lanes so these modes can still be used safely. After 60 years of living with cities like these the general consensus is the road deaths, obesity rates, road wear and energy consumption aren’t really deal breakers.

The pact only unravels when a city’s population grows beyond 200,000—the low density capacity of a city that is yet to burst beyond the approximate 15km diameter of a city-proper.

A hundred years ago the solution was to build up and add subways. In any 15km diameter piece of New York there are roughly two million people. The subway moves people around. The roads are given over to delivery vans and anyone with enough patience to sit at the lights in a car or a taxi.

The post-WW2 answer has been to build out and add motorways. Taking this approach the diameter of a city of two million people balloons to almost 50km.

In both cases the economic advantages of living in a city of millions begins to loose a lot of its shine for inhabitants, who can be losing 10 hours per week just getting to work, and as much again running errands. They become monetarily rich, but time poor.

The automatic response in the machine age has been to seek technological solutions to the problem of urban mobility, hence the array of personal, public, shared, driverless and composite mobility solutions now in development. A few companies will become extraordinarily rich by their efforts. The logical solution would not make any one company or inventor any money at all. The spoils would be shared evenly between all of the residents of whichever city had the wisdom to see the solution, and the courage to choose it.


The logical solution is there to be seen on the salt flat. When your own trips were the only trips there, the long trips generally suited the car and the short trips generally suited your feet. There would have been so few trips of 30 to 100 meters that you would have discarded the bike as a toy. But what if, when the population had started to thicken, everyone had agreed to banish all cars? You may still have built subways, but as densities rose, you would have had the whole ground plane for cycling. You would have discovered an overlooked fact, that bikes average 15kph without any more effort than walking, 30kph with maximum effort, and somewhere in between as a rule. The average time taken to bike between two randomly selected points on a 15km diameter plane would be 20 minutes, better than the subway, and far better than a car in any kind of city of one or two million.

The logical solution to urban mobility is to put 15km diameter limits on urban expansion, maximise density using perimeter blocks and the likes, and completely ban cars from the whole metropolitan zone. Due to the corrosive impact that each and every motorised machine has on cycling—cycling being the mode that promises everyone safe, free and healthy trips of around 20 minutes—no vehicles large enough to cause jams should be allowed within the metropolitan zone unless every other alternative has been exhausted. Even then they should not be capable of greater speeds than a bike (unless they are vehicles with the right to use sirens). Such a city could sell itself as having the advantages of a two million strong population, yet with average trip times of 20 minutes.

New York could achieve this tomorrow. All freight could be handled using cyclelogistics, pedicabs could replace motorised taxis and unavoidable vehicles could be speed restricted to 20kph. Each borough could have one or two of its transit hubs made into an express station with mountains of mechanised parking for bikes.

Then development incentives could be put in place to allow through-site connections to shorten the blocks, and gradually replace the walk-up building stock with ramp/bike access buildings like the ones I have in development.


  1. Jimm Pratt says:

    “… bikes average 15kph without any more effort than walking, 30kph with maximum effort…” If you accept and incorporate other variations of cycling, such as e-bikes (average 25kph), e-cycletaxies (average 20, max 30), and velomobiles (all weather, 20-25 average, 60-70 kph with maximum effort), then this 15km utopia should become even more appealing to more types of people. Yet still keep the car/truck/van relegated to outside the perimeter.

    • Steven says:

      Thanks Jimm, all true, and I’ll be first to bring a velomobile to velotopia. For design purposes, I think we should focus on the slower among us who rank highly in terms of productivity, eg, a parent with a child in a bakfiets.

  2. Jimm Pratt says:

    “…we should focus on the slower among us…” And we do – at least in Europe. For several of the big cities it’s been determined over the last few years that 20 kph is the average cycle commuter speed – from mom doing the school run in her bakfiets, to college students, to men in business suits. Using Copenhagen as an example (and this might be adopted in other cities in the near future), we have a couple of “Green Waves” on arterial roads heading into and out of the city center where the traffic lights are timed for a cyclist travelling at 20 kph. If you enter the wave in the right time-frame, you can cycle for 5-6 kilometers non-stop – saving lots of time by not having to stop and start repeatedly. For a couple of years this has worked well, so it’s apparent that 20 kph is a good average. Only the cargo transports (utility, postal workers, food bikes) go slower – although many are using electric motors more and more lately.

    …And now that I think about it, this whole reply is kind of moot, since a proper ‘cycle-city’ design would likely not use traffic controls based on motor vehicles *anyway*… we can cater to any level of cyclist skill and speed with just the simple roundabout/traffic circle. Although I think we could eliminate most car-eccentric road models and use more of the sailboat/boating style of navigation across the city – bikes should give way to other bikes approaching from the right (for example).

    • Steven says:

      First, thanks for playing ball!
      It’s not really a moot line of thinking if we think about city limits, and the attendant issue of density. We need to think of that 15-20kph rider in calculating average trip times from the periphery to civic core, then determining how big the city can be area-wise. A 15km diameter city promises 30 minute commutes from the edge to the centre, for the slower among us.
      But of course, roundabouts would keep flows of unmatched cyclists all doing their own personal maximum speeds, up to the speed of a velomobile. I like the idea of raising roundabouts on mounds to slow cyclists at these points without making them use their brakes. There is a fascinating range of ethical and design challenges in this exercise.

  3. Matthew says:

    Umm, sorry, but, plain, not plane. Maybe too many hours on 3D modelling programs?

  4. 7homask says:

    it is plane – as in a euclidean geometrical surface defined by the area connecting 3 points.

    • Steven says:

      Having published a book that referred twice to “lugged and braised” bike frames, I will trust you both.

Leave a Reply