j0elc, hydrogen fuel cells are efficient at converting hydrogen to electricity. But if you use electrolysis to make the hydrogen and you use coal to make the electricity to do the electrolysis on a large scale you end up with worse efficiency and higher costs compared to just using diesel in the bus. In reality the hydrogen fuel cell bus is coal and natural gas powered in Australia but due to all of the conversions along the way is not as efficient overall (it is highly efficient in the final conversion of hydrogen to electrical energy)
As usual there are no free lunches. The full story for a fuel cell bus in Australia:
Mine coal, (using diesel)
transport to power plant, (using diesel)
convert coal to heat, (thermal losses)
convert heat to steam, (thermal losses)
use steam to drive turbine, (some thermal and turbine efficiency losses)
use turbine to turn an alternator to make electricity, (some losses due to alternator efficiency)
use transformer to step up voltage for transmission over long distances (transformer losses)
send electrical energy over the grid to the hydrogen factory (transmission losses)
step down the voltage using a transformer to practical voltage (transformer losses)
use electricity in hydrolysis plant to make hydrogen and oxygen.
Transport hydrogen to fuel cell vehicle refuel point
use hydrogen in fuel cell bus.
You could use a wind turbine to make the hydrogen but that still costs between $3.84 and $5.80 US per kg of hydrogen before subsidies. (see slide 15 here
http://energy.gov/sites/prod/files/2014/03/f11/webinarslides_wind_to_hydrogen_011713.pdf)
The buses currently use average 15.48 kg per 100km .
http://www.nrel.gov/docs/fy14osti/60603.pdf
If a city bus does 100,000 km a year it would need 15480kg of hydrogen and would require 30.96MW of additional electricity per bus. There is a fair bit of capital cost associated with hydrogen fuel storage because it is compressed to 10,000 PSI and has very low density kg/m3 and a lot of capital cost for wind turbines due mainly to land prices and linking wind farm to the grid. Its all feasible but you need expensive oil for this to be economically viable.
(20min delay)