I love backofenvelope calculations, and I often throw them into
my public lectures.
Here is one:
Question:
If we powered all the cars on a road using biofuels grown on the verge of the road,
how wide would the verge have to be?
Assumptions:  Comments 

one lane of cars 
60 miles per hour  (typical speed on the open road) 
30 miles per imperial gallon  (average for new cars in Europe today) 
1200 litres of biofuel per hectare per year  (typical for biofuels grown in Europe) 
80 metres carspacing  A reasonablysafe stopping distance? 
Let's do another calculation.
Imagine that we switch all the cars over from today's liquidpowered cars to today's
electric cars, and power them with wind turbines.
How far apart would the wind turbines be?
Assumptions:  Comments 
one lane of cars 
60 miles per hour 
20 kWh per 100 km  (the Tesla can do 15 kWh per 100 km, so this may be a pessimistic figure) 
0.5 MW per turbine  (assuming a 2MW turbine with a load factor of 25%, which is typical in Europe) 
80 metres carspacing 
The power required per unit length of road is
(60 miles per hour) * (20 kWh per 100 km) / (80 metres )
= 0.24 MW per km.
(you can choose the units in which google calculator returns its answer)
So the spacing between turbines, to power one perpetual lane of cars, would be
( 0.5 MW per turbine ) / (0.24 MW per km) = 2 km per turbine.

If you would prefer to express this in terms of the 'effective width of the verge', assuming the verge
were 'all windfarm', we can use my estimate of the power per area of wind farms of 2.5 W/m^{2}.
(See figure 3 of my paper 'Could energy intensive industries be powered by carbonfree electricity?' for data supporting this number.)
We get:
Windfarm verge width = (0.24 MW per km) / (2.5 W/m**2) = 100 metres,

which is 80 times less land
than the biofuel solution; and most of that 100 metre strip would not be occupied by the wind turbines,
so it would still be available for other nonwindharvesting uses such as agriculture.
The windfarm solution uses less land because,
under the above assumptions,
electric vehicles are more energyefficient than liquid vehicles, and because the power per unit area of
biofuels in Europe is very small, even compared to the powerperarea of windfarms, which, at 2.5 W/m^{2},
is not huge.
What about solar panels?

Solar panels on a roof alongside a motorway in Belgium.
Picture: AFP/GETTY


If the verge were occupied by solar panels, how wide would the verge need to be to power all the cars?
We have got most of the numbers already, from the wind/electriccar assumptions.
We need to introduce just one more number, namely the average power per unit
area of solar parks in Europe.
This can be found in my paper
'Solar energy in the context of energy use, energy transportation, and energy storage': 5 W/m^{2}, for real solar parks
in Germany and England. (Higher powers per unit area, such as 10 W/m^{2}, are achieved
by solar parks in more sensible sunny locations such as Spain and California.)
Solar park verge width = (0.24 MW per km) / (5 W/m**2) = 50 metres.
For comparison, a typical single lane on a motorway is 3.65 metres wide.

[Note however that solar parks in Europe produce seasonally intermittent power
that would not be well matched to transport demand, which is quite steady.
In UK and Germany, the average midwinter power output
of a solar park is roughly ten times less than the midsummer output.]

Summary: how big would the verge be? 
Biofuel verge  liquidfuelled cars  8000 m 
Wind park verge  electric cars  100 m  (or one 2MW turbine every 2km) 
Solar park verge  electric cars  50 m 
All the standard disclaimers apply — answers depend on assumptions; assumptions could be
changed; lots of innovations are possible in crops / crop processing / vehicle design, such
that the assumptions might be changed.
Yes, I know! Indeed all those possibilities are one motivation for doing backofenvelope calculations,
to make clear which assumptions are the one we really want to make progress on, to make a material
difference.
David MacKay FRS is a member of the 2014 Longitude Committee. He was
the Chief Scientific Advisor at the Department of Energy and Climate
Change from 2009 to 2014, and is Regius Professor of Engineering at the University of
Cambridge. He is well known as author of the popular science book,
Sustainable Energy —
without the hot air.
