Not what you might expect: we were building small distributed solar thermal plants that stored energy in a tank of supercritical water (basically: inject steam into tank of water -> it condenses; open a valve on the top and it flashes out as steam).
Since the pressure varies depending on solar flux and how much was in the tank we couldn't use a turbine (the shape of the turbine is tuned to a constant flow) so instead used multi-stage steam engines (the first new steam engine designed probably in 75 years. These were ordinary container-sized pressure vessels and the steam engines were built into a container too. Easy to repair; you could make your own parts, and last forever.
Oh that sounds like a ton of fun. In the 90's I was working on solar concentrators because the panels were so bloody expensive. Funny how that got entirely obsoleted by economies of scale.
Cloudy weather is harder, not easier when using concentrators (because it concentrates parallel rays of sunlight, not ambient light). Restricted space would still need to be covered with something and here you'll find that covering it directly with photovoltaic panels is more cost effective and will get you more power for a given area. As far as I'm aware concentrators are pretty much dead except for when you need lots of localized heat, for instance a solar oven. There they definitely work better than anything that is based off photovoltaics and because they tend to be totally passive probably more rugged and reliable too.
To my surprise there seems to be a small revival, though I don't quite understand the applications (not that I follow it closely).
In general land is less expensive than the cost of trackers or concentrators (capital cost incl installation, plus op ex of maintenance & power) so you deal with your cosine loss by just installing more panels.
Definitely. Though being a 21st century mechanical device it was heavily instrumented and had an electronic control system. Amusingly, our electronic cylinder indicator showed a problem in one of the stages but the old-school mechanical one did not. The ME's told me "your instrument is bad" but when the stage was taken apart...sure enough the head gasket had a hole.
The advantage of the mechanical system is that it would last indefinitely and could be repaired on site. People in India could make most of the spare parts themselves and do the maintenance themselves.
But the advantages of the semiconductor + battery systems were: 1 - no moving parts and 2 - massive economies of scale due to mass factory production of highly uniform parts that could be used in lots of applications. Makes sense that it won -- our approach was superior only in certain applications.
Ha, Thanks! I just ran into the other room and asked my wife who is a retired OR nurse what "positive Q sign" meant and she stuck her tongue out of the side of her mouth. I love learning new terms :)