The high efficiency of electric traction can be also illustrated by calculating how many m2 of photovoltaic cells we would need to drive a certain distance. Of course the sun does not shine the same everywhere and it is not possible everywhere to locate solar cells optimally.
If we start again with a consumption of 130Wh/km and assume for battery and charger electronics an efficiency of 65%, the cells would have to provide about 200 Wh per km to be driven. If a car port roof measures 2.5x5 m = 12.5 m2, this roof could provide in a favourable southern climate about 12.5 kWh/day, or 4562 kWh in a year. The resulting driving range of 22'812 km exceeds current typical average distances driven in a year in much of Europe. Even if we reduce now the insolation figures for more northern locations, say to one half the values used above, we could cover average driving needs from the roof of a carport. Comparable calculations for other climatic conditions and less favourable locations have been made e.g. at the Technical University Vienna as reported here (in German) or can be derived from offers for solar panels and installations.
So far so good, but how much land would we need to drive the same distance using alcohol from sugar cane? Sugar cane has a very high yield, so should give us a favourable starting position. Sugar cane yields about 0.748 litres/m2 and year. Of course also in a favourable, southern location. Alcohol contains less energy than gasoline or diesel fuel. Remaining within the same ballpark number of 10 litres/100 km for petrol, a consumption of 15 l/100 km with alcohol could be indicative. To drive then the 22'812 km calculated above, we would need 228*15 = 3420 litres of alcohol. To produce this amount would require then a surface of 4572m2, or 366 times the surface area needed with photovoltaic cells.
This does naturally not prove that alcohol use is necessarily always out of the question, but the comparison shows clearly the large benefit we get, if we start with a high efficiency at the end point of use. This lessens, in our case here dramatically, the calls on the whole chain back to the original source of energy. Energy prices to the end user, excluding taxation, can give an indication of total capital and operating costs. For the examples calculated here we could take 36c/kWh (USD, photovoltaic) and about 1$ per litre of alcohol, or 15$/100 km with alcohol. The solar electricity would end up at 7$/100 km. So even when using one of the most expensive sources for electricity, electric traction can keep a relatively favourable position among potential alternatives to crude-oil derived fuels.
Fears that the additional electricity for electric cars would become a lot more expensive are in most places unfounded. There have been several studies analyzing such scenarios, in Europe and the US. Again it is the high efficiency that has the crucial effect. The additional call on electricity production is so modest, and expected to show up largely off-peak, that not even new generating capacity would have to be built e.g. in the US or Germany: Electricity supply is not a critical problem. Also the effects on CO2 production and other pollutants have been studied extensively. Links to several studies can be also found e.g. via Lynne Mason’s well done pages. The observations made above do agree well with the calculations made by Tom & Cathy Saxton with many data based on actual use of electric vehicles. Also the data for the all electric Nissan Leaf support these estimates. Interesting is that the far less powerful Nissan Leaf does not show a very notable advantage over the Tesla Roadster if both are driven in about the same manner: an effect quite unlike experiences with combustion engined vehicles.
We should also not forget that it is by no means the same if exhaust gases are blown directly in the faces of pedestrians in crowded cities, rather than emitted from a high rise smoke stack 100 km away - even if total emissions were comparable.
For CO2 e.g. the calculation shows that conventional cars emit more CO2 than a comparably powerful electric car, even if its electricity is produced from coal. An erroneous assumption is also that petrol reaches the fuel tank emissions free, when in reality it is produced with quite some energetic input in a refinery and transported over large distances. Basically the crude oil has to be cooked to get usable products for motor cars (via fractional distillation among other things). Some referenced numbers can be found here. Make your own estimate, on the basis that a coal fired power station emits about 1000g of CO2 per kWh generated (modern stations 20% less) and that a car burning 10liters/100km produces about 230g of CO2/km.
In addition the impact of real driving patterns is mostly underestimated. The lower efficiency of piston engines in city cycles is visible in the differences between city and highway EPA mileages. The city mileages reach about 60% of highway mileages. For the Tesla Roadster the opposite is happening. It will actually go further in slower traffic as the electric engine's efficiency remains unchanged and air and rolling resistances are reduced. In cooler climates piston engines do not reach their ideal operating temperatures for much of their often short trips, resulting in a consumption well beyond standard test results. These effects all amplify the advantages of electric vehicles in real use. Few studies do account for these effects and data on driving patterns are not easily found. Incidentally also hybrid cars benefit from these effects, particularly the Chevrolet Volt, essentially an electric car, capable of extending its range with a conventional engine. Until better data become available, supported by practical experience, I estimate that the energetic advantage of electrical traction is about an order of magnitude (10x) at the point of end use.
It is therefore likely that the impact of electric traction on oil products use could become very large, if not dramatic, amplified by actual driving patterns, where most of the fuel is used to cover short distances in stop and go traffic. This is exactly where the internal combustion engine is least efficient and where electric motors are as efficient as ever. Please note that the number of digits used above is not an indication of accuracy. I have used the numbers in this way just to facilitate verifying the calculations.