My comments on the video :
a lot of exageration. But it's still a valid potential tech for the future grid storage. Not sure if it would be economically useful, tho.
Don't forget there's a lot of thermal management involved.
@4:33 : "CdTe solar cells won't scale" Exagerated. You need approx. 50x more Si. than Te to make the same amount of solar output in a PV cell. So the cost of a trace material like Te is not that critical. As with Lithium mining, supply slowly scales with demand.
@5:55 "platinum is OK for the jewlery market, but not for widespread use in automobiles" : LOL. That's completely wrong. Nearly half of the Pt mined today is eaten up by the automobile industry, to make catalythic converters ! His point on cost of FCEV still is valid, but exagerated.
@26:12 : less cells wiring complexity : yeah, Lithium exists in bigger cell sizes, achieving the same.
@2:58 : the graph is probably not up to date on Liion storage.
he claims about "5k kWh" -> 5MWh, which is obviously far too low!
That's obviously wrong by a few orders of magnitudes, as the tesla grid storage in Australia has 130MWh installed, and there are others, growing very fast.
The total number seems to be in the 2-3 GWh installed now, according to :
https://www.greentechmedia.com/content/images/articles/storage-forecast-2016yir.pngwhich seems plausible, considering that the largest 10 grid storage batteries in operation total 1.146 GWh aggording to :
https://en.wikipedia.org/wiki/Battery_storage_power_station#Largest_grid_batteriesPumped hydro is ~130 GWh, which should be much closer to the "100M" mark on his graph.
So storage batteries for grid buffering, mostly Lithium based, are already at about 2% of pumped hydro, and growing fast. In a few years this will become significant.