This article was published in 2015, in Newsletter 119.
Which has the higher capacity – your average bus lane or your average bicycle lane?
Recently, I made a slightly flippant remark on Twitter that caused me to think long and hard about a few things. The original posts from @AsEasyAsRidingwere:
Wow. 66% of all bus trips in London are under 1.8 miles (3km). How many of those could be cycled, if the conditions were attractive?
[A]nd nearly a half (46%) of all London bus trips are under 2km. Staggering.
My reply was:
Does that imply that good bicycle infrastructure may reduce bus use, and make even more efficient use of road space?
Somebody else responded with:
Surely a full bus is more space efficient than same number of bikes? What is average occupancy?
Are false assumptions driving the desire to increase bus priority when other modes of transport might be better value for money?
To answer this, I decided to calculate the space efficiency of various forms of transport, from the humble bicycle to the car and bus. By space efficiency, I mean the number of people that can be transported along a typically-sized route each hour per metre of space, using the units of people per hour per metre.
For buses, I assumed a double-decker bus, something like the new Routemaster, with a crush load of 87 people. Typically, local buses have stops every 400 metres and on a high capacity route will have on average eight people get on at each stop, although some stops may be more or less popular. We also have to factor in the distance required to walk to or from these bus stops as, unlike a bicycle, buses do not go door to door, so let’s assume this is on average 200 metres at each end.
Finally, let’s make an assumption that it takes about 10 seconds for a single bus driver to take a fare, give a ticket and possibly provide change, and then greet the next passenger. Every stop lasts on average 1 minute and 20 seconds to process those eight people, although some stops may require over 2 minutes and 30 seconds dwell time to cope if there are many passengers waiting.
With the above assumptions, the bus doesn’t do very well. A journey of 3km takes about 15 minutes, giving an average total journey speed of about 7.4mph. Yes, it is quicker than walking, but it is a lot slower than riding a bicycle. Essentially, we are limited by the time required to get passengers onto and off the bus. We could fix that problem by introducing prepaid tickets or contactless payments, as has been done in London.
By radically reducing the time required to process each passenger to just 2 seconds, each stop could be reduced to a frantic 16 second dash. The time required to cover the 3km is reduced to 7 minutes and 30 seconds, with an average moving speed of about 14.8mph. This looks good, except we have to also consider the time required to walk to the bus stop and wait for the next bus. Even with smart ticketing, this increases journey time by about 5 minutes, reducing the whole journey speed to 9.7mph. It is not looking so good for buses.
This is not yet capacity. To calculate that we have to determine how many buses we could fit down a single lane. For this, we have to work out the minimum time between buses arriving at each stop. This is effectively the maximum time that one bus is stationary and the bus behind is still moving, such that this stationary bus can safely move off to be quickly replaced by the bus behind.
Let us not assume that a whole busload gets on at a single stop, but perhaps a quarter of the busload does. Given this, buses would be stopped for a maximum of 45 seconds at a time, and with an additional 15 second safety window, we can get a maximum of one bus every minute along the bus lane, giving a total capacity of 60 buses an hour. If each carried their maximum 87 passengers, this gives a carrying capacity of 5,220 people per hour per bus lane. A bus lane being a minimum of 3m wide, means that a bus lane carries 1,740 people per hour per metre of lane.
A car lane can carry approximately 1,800 vehicles an hour, and in Cambridge they carry on average 1.2 people, so within the 3m-wide lane, they have a total capacity of 2,160 people per hour, or just 720 people per hour per metre of lane. This is why cars cause congestion; they are most inefficient users of space. They are great in rural areas, but are terrible in dense urban areas. Data from Cambridge also suggest that the average car speed is just 11.4mph. I think even I can move faster than that on a bicycle.
Well, the average speeds of bicycles are very similar to that of the most efficient bus. For example, Copenhagen assumes bicycles average 20kmph (12.4mph), and there are regular ‘our cyclists are faster than yours’ discussions on average cycling speeds between countries and cities. Of course, you can also usually cycle from your start to your destination, instead of having to walk at each end.
What about the capacity of a bicycle track? Well, opinions differ. Swedes think it is about 1,500 bicycles per hour within each 1.2m space, while the Dutch have observed 3,000 bicycles for every 0.78m of space. To help calculate this we have to consider the length of a bicycle and the gap between your bicycle and the person following behind. This gap is determined by the safe following time, and is observed to be about 1 second. A bicycle is on average about 1.8 metres in length, and that 1 second gap gives a space between bicycles of 5.6 metres. A whole bicycle/gap combination is therefore about 7.4 metres in length and moving at 5.6 metres per second. For one-third of a second, a bicycle will pass somebody observing from a fixed point, and then for another second there is nothing, followed by another bicycle, etc. This means that about 2,700 bicycles would pass that single point every hour. But, a 2.1m-wide cycle lane can support two lanes of bicycle traffic, so this would be almost doubled to about 5,400 bicycles per hour in that lane. This means that bicycles have a space efficiency of over 2,500 bicycles per hour per metre of lane, three and a half times more efficient than cars, and one and a half times more efficient than buses.
To answer the question posed: if we took all of our 5,220 people who were on the 60 buses travelling along our road and placed them all on bicycles, they would use significantly less space. In fact, they would fit into a 2.1m-wide bicycle lane without a problem (see table). A simple reallocation of space would even leave space for trees.
|Vehicles and people per vehicle per hour
|Minimum lane width
|(people / hour / lane)
|(people / hour / metre of lane)
|1,800 x 1.2
|60 x 87
|5,400 x 1
If we want to speed up buses, then we need to speed up the bits when the bus is not moving. Moving to contactless payments or pre-purchase of tickets, as the Busway does, would help speed up buses much more than removing every tree and widening every roadway.
Cars are quicker than buses. But, if you really value your time, you should be getting on that bicycle, for the bicycle wins at everything. Bicycles use the least space, can carry the most people, and have the highest average speeds in urban areas.