This article was published in 2017, in Newsletter 135.
There are a few things that matter for public transportation networks. Speed is important, but frequency and capacity are much more important.
Take, for example, the proposals for the ‘Affordable Very Rapid Transport’ (AVRT) system being proposed by the Greater Cambridge Partnership along with Cambridge Ahead and the University of Cambridge. They quote travel times like just 3.6 minutes from Cambourne to the West Cambridge Site, and then another 2 minutes to the Science Park, or 2.1 minutes to Addenbrooke’s. This is incredible.
No really, this is incredible. I’m going to assume that the laws of physics will apply, and that simple GCSE maths will be used along with equations such as s=ut+½at2. Now, the distance from Cambourne to West Cambridge is about 10km, line of sight. To travel this distance in 3.6 minutes takes some effort, and a little bit of speed. You have to accelerate at 1.3 m/s2, the typical ‘fastest tram’ acceleration speed, for about 50 seconds. Then once you are travelling at a slightly uncomfortable 135mph, you then decelerate back down to stationary in the final minute, hoping that the brakes work. After that, you have to walk over to another stop, and wait for almost 5 minutes for the next shuttle to take you the rest of the way.
You could make vehicles that will accelerate faster, but that also would constrain the vehicles to be seated only, probably with the requirement to wear seatbelts, or those over-shoulder bars you have on amusement park rides. This would also restrict the capacity of the system. But this is all probably a small typo in the big picture of things. Of course, if a more realistic acceleration profile was used, and a more realistic maximum speed, and intermediate stops, then we’d have a much more realistic journey time of about 7 minutes or more.
The shuttle approach also allows for cheaper infrastructure by having single-track sections between stops. So the vehicles can only go in each direction once every 15 minutes. Of course, they could build routes with bi-directional traffic, but then that would cost a lot more. About twice as much more.
No, the real problem is that frequency is what matters. You see, if you ride a cycle, or drive a car, or walk, there is never any wait. Perhaps sometimes you have to manoeuvre the vehicle out of a parking place but there is no structurally imposed limit on when you can do that. The shuttle vehicles will, on the other hand, only travel infrequently. This means the capacity is very low. Forty people per vehicle four times an hour is very low compared with over 10,000 people per hour per direction that a typical light rail system could support. Oh, and light rail systems can have intermediate stops, like a stop at Comberton.
A show-up-and-go public transport system requires not only good frequency but also sufficient capacity so that you won’t be standing in a queue for 30 minutes just to get on-board.
What we need is public transport that works with other modes: that connects with P&R sites that people can cycle to; that supports having intermediate stops; and has sufficient capacity to move everybody and is frequent enough to just turn up and go. What we need is a multi-modal transportation network.