5 ms·
> Think of it as like the relationship between cache memory and slower RAM: the combination performs better than either alone. Seems to me like upcoming long t
by danans 2y ago
> Think of it as like the relationship between cache memory and slower RAM: the combination performs better than either alone.
Seems to me like upcoming long term storage battery chemistries chemistries might be more up to the job analogous to RAM than e-fuels. E-fuels seem more useful for heavy transport and flights than for feid power.
- pfdietz 2y agoThe capex for underground hydrogen storage capacity can be very low (like $1/kWh of storage capacity; keep in mind this is not including per-power costs). This is a very challenging price point for batteries to hit. The key point here is that, like flow batteries, e-fuels can separate per-energy and per-power costs. Ordinary batteries can't do that, at least to the same extent. Things like iron-air batteries might be good for an intermediate timescale, like ~1 week. They would be intermediate on the efficiency-capex tradeoff. They might not be good for rare event backup (the dreaded Dunkelflaute) or seasonal leveling, but could be a good match for normal wind variability, which tends to be on a somewhat longer timescale than solar. BTW, electrolysers from China have been on a downward price curve, similar to that seen in other renewable technologies. Low cost electrolysers (at perhaps mediocre efficiency) are the pacing technology for green hydrogen from intermittent renewables. One would like to be able to use them when renewables are providing excess power, not have to run them 24/7 to amortize high capex (as the old nuclear-powered hydrogen economy idea had it.)