This morning, I found myself mentioned on Jarrett Walker's excellent transport planning blog Human Transit, with reference to my thoughts on London's complicated bus network. I have a lot of thoughts about this, and I intend to eventually draw them together into a convincing report on quite why London should adopt a much better bus network, but for now, I'll set out the basis of the small amount of work that I've done so far.
My concern with London's bus network started from three points:
- The complaints of cyclists about articulated buses. I like them, as free boarding and alighting through three doors means they don't spend very long at stops, and I'd rather stand for a short journey than have to go up and down steps on a double-deck bus. But, so it's explained, the relatively difficult-to-control movement of the rear section creates problems when the buses are making turns; both right turns and left turns. Aside from questions about whether curb-hugging cycle lanes are the best idea, this rases one question to me more than any other: Why are buses making so many turns on London's torturous narrow streets? This slows them down, as well as making them problematic for other road users.
- The fact that most people are intimidated from using the bus network, because of how complicated it is. Tourists generally stick with the tube network, even where it's not appropriate, and tourists are famous for leaving London with little knowledge of the structure of London beyond the tube map - it's popular belief that London doesn't have any real structure to it street network, which it actually does. Meanwhile, when around professional and business visitors, no one but me will ever have used a bus; these sort of visitors stick to taxis, which is terrible in congestion terms.
But the tourists and the business visitors have a point. having got to grips quickly with networks in New York City, DC, Helsinki and other places, I still find myself in London getting buses and finding that they make a turn that I didn't expect, and start taking me away from my destination. If you're in a rush, or don't know the city too well, you won't take that chance. If I knew my bus was staying on one important street, and that I'd have to get a different bus to change direction, then this wouldn't happen.
- Money has to be thrown at the buses in order to maintain their frequency, which tends to be around every ten minutes for individual routes. Buses in London, in my experience, often aren't that full, and as consequence, the network swallows up more subsidy than the rest of Britain combined, costing £574 million over the 2009/10 financial year. If cutbacks are to be made, which fiscal austerity seems set to require, it can come in two ways: Reduction in frequency over every route, or simplification so that fewer routes run just as frequently, or even more frequently. It is similarly helpful not to sell off articulated buses. Because they don't spend long of bus driver's paid time at bus stops and they carry a lot of people per bus, they are very efficient buses. To build new double deckers to take their place would be much less sensible than to put them on suitable routes with few difficult turns.
All three factors point to one solution: Running a relatively small number of bus routes, which run in straight lines. But how am I to make a journey that isn't in a straight line? It's quite simple, and well explained by the aforementioned Jarrett in his article on grids and on the benefits of networks involving interchanges- you will make, ordinarily, one change, in which you won't have to wait very long at all.
For the sake of argument, I've drawn a map of what it might look like in Central London. Central London is the main problem, as buses in the suburbs tend to be happier to stick to one main arterial road, and the routes I've drawn would more or less continue down the arterials on which they leave Central London. I would expect the result of a serious analysis of how you'd implement such a network to be pretty similar.
Google Maps Link
Note that routes are x-series, e.g. 5 series, meaning that I'd envisage buses on a common route all having the same number to start with, with another number representing different routes they might take much further out in the suburbs - 50, 51, 52, etc.
So, supposing that you wanted to get from King's Cross station (The cluster of National Rail stations in the top-centre) to Hyde Park Corner (in the centre-left where Green Park and Hyde Park almost touch). You'd take the Bright green 17 series (170, 171, etc.) until Theobalds Road, where you'd take the 3 series (30, 31, etc.) to Hyde Park Corner. That, I'm sure you will agree, was really easy to work out from the map. You could simultaneously build up an idea of the layout of main roads in London and the buses that run along each, so that in no time at all you could spontaneously work out how to get anywhere in the city.
Yes, this means you have to change instead of going direct. However, rather than waiting perhaps ten minutes for a direct bus, the simplified lines could run about every two minutes, replacing the huge numbers of routes the map showed running on each road. This means a maximum four minute wait in total, making for less time spent waiting even if buses found themselves bunched together and you had to wait slightly longer.
So, in summary, London clearly demonstrates that simplicity has a universal appeal, even when you are providing high frequency on your complicated network:
- Frequency is always better on a simple network - up to the point at which you don't really have to think about waiting at all, which London would be able to provide.
- Particularly on the constrained road networks in European cities, avoiding turns is useful.
- Simple networks are easy to understand, good for both newcomers and locals that want to expand the range of places they go to.
- Potentially, simple networks could help avoid the downward spiral of usage that starts when you cut frequency and thus make public transport less convenient.
