I'm an advocate of solar power -- when it makes sense. For energy harvesting in data loggers, site-based power, spacecraft, and large fixed installations such as houses, for example, it often works out fairly well, or is the only viable solution.
At the same time, I get annoyed when the promise of free solar power is used as an attention-getting gimmick, regardless of how impractical or nearly useless that promise is. The latest example of this is a piece I just saw in The Wall Street Journal, "Ford to Show Solar-Powered Hybrid at CES." (Sorry, it may be behind the paywall.) In brief, Ford is equipping some of their hybrid C-Max concept-car models (based on a commercially available vehicle) with solar panels on the roof, along with some sort of sun-tracking system. It's not clear. It looks like it may have steerable Fresnel lenses in order to provide both basic and booster free power.
What I like to do when confronted with this sort of technology claim is perform a quick back-of-the-envelope, order-of-magnitude sanity check. I have a notepad I made just for this function, to use at concept meetings. It's made entirely of envelopes. (See photo.) My analysis: I don't care how good that solar power system is, it won't collect enough power to make it even remotely worthwhile.
A few rough numbers tell the solar tale. On the source side, the maximum solar radiation reaching Earth's upper atmosphere is about 1,000 W/m2, according to various reputable sources. That's the maximum. Just factor in atmospheric losses, clouds, the angle of the sun at different latitudes, and seasonal effects, and that number drops down rapidly.
On the conversion side, you have the efficiency of the solar cells reaching say, 15% at best, minus losses in the dc/dc power system of around 50%. (Remember, this is all an estimate. We are just trying to see where things stand.)
A car roof is perhaps four square meters -- let's say five square meters, to be generous. If the car roof captures full-intensity solar radiation, you've received 5,000 watts per surface, less the capture/conversion losses. Work those losses in and you may garner 375 W for the car battery to store, assuming you got the full 1,000 W/m2 at the cell surface, which won’t happen.
To make a long story short, you'll be lucky to collect a few hundred watts. Now look at the load: One horsepower is about 750 W, and you've collected less than half that amount. How far can you go, then? Although it depends on how long you've been able to collect that energy, the real answer is "not far at all." Well, perhaps you can use it to supplement the car's internal power source? Again, not much help. A car today can use it up easily for all the infotainment subsystems. The bottom line is you are putting a lot of cost and stuff into the car for very little useful return.
The power-harvesting reality is simple, determined by the laws of physics. You collect energy over time, when it is available, but you spend it quickly, as power, to meet the much higher needs of the load. Just because you are collecting it at a slow rate doesn't mean you can get away with spending it at that same rate, since most real-world loads have specific minimum-power requirements to make them functional. (See Energy Availability vs. Power Needs if you need a refresher.)
Ford's announcement worked. It got them attention. I'll give them credit for that. But the solar-powered car story is an oldie but goodie. It always amazes me that it still works. (Check Popular Science from the 1950s and you'll see the same story!)
Have you seen any free-power-harvesting schemes that made little or no sense? Have you ever tried to point this out to proponents?