> Transit systems are usually sized for peak demand. Because you are paying for the vehicles and the infrastructure anyway, the marginal cost of offering better service outside peak hours is low. Especially if the drivers receive monthly wages and have semi-reasonable shifts
Right. Now remove drivers from the equation.
The ideal size for a bus goes down because you can effectively turn off parts of the bus outside peak hours. (By parking some of the fleet.) Moreover, you can induce new demand by adding (and moving) virtual stops wherever your customers are. Busses via rideshare doesn’t work with a driver; it does if you can spin up and down your fleet on a dime.
No, it doesn't, because if you have lots of small buses then high-traffic areas will get clogged with lots of little buses loading and unloading and just physically taking up space in bus lanes. If anything, you generally want _bigger_ buses; most large cities size buses as big as the road infra will take (with the exception that a lot of cities won't consider double-deckers even where they'd make sense, due to lack of familiarity). Now, some of that is drivers, but some of it is just capacity.
(Or sometimes slightly bigger than the road infrastructure will take. About once a decade, Dublin Bus tries articulated buses, which then get stuck on roundabouts, reminding Dublin Bus why it doesn't have articulated buses.)
The mass difference between a big bus and a small bus also isn't generally _that_ huge. A smallish single-decker is about 10 tonnes, these days a big double-decker is only about 12.
The ideal size does not go down. You want the bus as large as practical in all cases. The difference in cost between a small and large bus is not significant (a little more full, and the bus costs 10x as much - but both of these are insignificant over the life of the bus), and the large bus can handle it if you decide to take everyone bowling after your kids birthday party or some such.
The dominant constraint is usually road capacity during peak hours. You need a certain number of large buses to handle that, and those buses are cheap to use outside peak hours.
Ideal bus size may go down if peak demand also decreases. For example, if working from home becomes more common. Smaller buses may also become more viable when direct bus lines are replaced with rail lines and feeder buses.
But many transit systems use larger buses to increase passenger capacity.
Most city streets are intended only for local traffic, and it's often inconvenient to use local streets for longer trips. Which means buses usually drive on major streets, which may become congested during peak hours. Due to fixed overheads, large buses use less road space per passenger and less stop time per passenger getting on/off.
It's also common that there is not enough space for a bus to pass a stopped bus due to congestion. Which means buses may have to stop at every stop on the route, even if no passenger is getting on/off.
> But many transit systems use larger buses to increase passenger capacity
Non sequitur.
You said busses are constrained by congestion. That's simply not true. When congestion is reduced, due to demographics or a congestion charge, cities don't run out to buy more busses. If the release of a suspected constraint doesn't cause an increase in signal, it isn't a constraint.
The real constraints on buses are demand, capital budgets and operating budgets. The dominant component of the last is the driver. The dominating deterimant of the first is the route. If you can have more buses on more routes for the same capital and operating cost as fewer, big buses, you'll increae ridership.
> If you can have more buses on more routes for the same capital and operating cost as fewer, big buses, you'll increae ridership.
The problem is, you probably only have a limited number of suitable arteries, so those routes are going to have places where they overlap. There are bus stops near me which at peak times have multiple buses per minute arriving, on many different routes, and at peak times these sometimes pile up and have to wait to load and unload. That's with big hundred-person double-deckers; it would be far worse if they were smaller.
I said that the dominant constraint is road capacity. Which means that buses must be large enough to carry the required number of passengers on the streets that exist. When you have buses running on ordinary streets with dedicated bus lanes (but no bus rapid transit), peak capacity is around 120 vehicles/hour/lane. Even mid-sized cities can have bottlenecks where that limit is reached due to geographic constraints.
Right. Now remove drivers from the equation.
The ideal size for a bus goes down because you can effectively turn off parts of the bus outside peak hours. (By parking some of the fleet.) Moreover, you can induce new demand by adding (and moving) virtual stops wherever your customers are. Busses via rideshare doesn’t work with a driver; it does if you can spin up and down your fleet on a dime.