The article suggests that mounting a 150 Wpeak panel (nominal surface area of 1.2 m2 at 12.5 % efficiency) on a car will have a much better financial payback because rather than produce cheap grid electricity it replaces heavily taxed motor fuel. The confusion regarding the plug-in capacity of the vehicle becomes important at this stage: doesn't it make more financial sense to fill up the car with cheap coal electricity from the grid and just skip the solar altogether?
The first problem that we encounter is a claim that a Prius can travel 7 miles on 1 kWh of electrical charge. This number equates to 8.88 kWh/100 km, which is not realistic. Ben at theWatt.com has compiled some of the Environmental Protection Agency numbers and the best is for the Nissan Altra EV at 17 kWh/100 km.
The next problem is that the author assumes that the capacity factor for a car mounted solar panel will be the same for a fixed mount on a building. The obvious problem is that a panel mounted horizontally on the roof isn't going to be at an idea angle to collect insolation. Furthermore, cars spend the vast majority of their lives parked. What happens if you happen to park in the shade of a high-rise or a tree; park it in a garage or parkade? Photovoltaics don't do well if they are even partially shaded, since it leads to reverse current in the shaded sections that degrade the performance of the system as a whole. Basic trigonometry suggests that if you're in Los Angeles (33 ° N latitude) then a horizontal panel should only see cos( 30 ° ) = 0.866 as much radiation. In actual fact, due to the higher insolation in the summer, RETScreen says that the horizontal panel in LA will produce 271.5 kWh/year versus 297 kWh/year for one fixed at 33 ° to the horizontal. These correspond to a capacity factor of 0.2055 and 0.2260 respectively whereas the author assumed a capacity factor of 0.25 for both systems.
The real issue here is that a car-mounted small solar panel can't contribute even a small amount of the energy needed to actually move the car. Using HOMER (which allows hourly rather than monthly resolution) and solar data for LA in 2003 I found the total power generated by a horizontal PV panel to be 245 kWh/year (even lower than from RETScreen).
Let's try another tack: maybe the power provided by grid-tied panels couldn't be used while the vehicle is actually driving. E.g. the grid has no storage capacity and everyone unplugs for and at the exact same time. Let's presume that a vehicle drives an average of 20 km on electric power every day of the week. Using reasonable figures for consumption (20 kWh/ 100 km) that's 4 kWh a day. We'll presume that the battery can store 2.4 kWh and has a peak output of 75 kW (105 hp). Let's also presume that the car only moves between 8:00-9:00 am and 5:00-6:00 pm. We'll be nice and assume the car is not grid connected for the entirety of those hours.