If Richard can only get about 1.5 hours with a wind generator backing up the battery then you would need a very big battery to get 8 hours. I therefore assume that the transponders get interrogated a lot more than the above examples assume. The guy who did the “science” part of the LAA submission to the CAA certainly thought it was a non starter to power an existing transponder with a battery, but I do not have a copy of this assumptions and calculations. Unless you could build an entirely portable box it would be of little use in most aircraft anyway.
Rod1