3 ms·
> There are better solutions That's a good euphemism for "this solution is a spectacularly bad one". Storing electricity in lithium batteries is the last thing
by jhgb 4y ago
> There are better solutions
That's a good euphemism for "this solution is a spectacularly bad one". Storing electricity in lithium batteries is the last thing you want to do when balancing the grid. You want to start with demand response, then continue with overgeneration, then continue with pumped hydro. In the future (when we will have replaced natural gas by synthetic hydrogen as part of the demand response) then possibly continue with expending stored synthetic gas in a CCGT plant (this would be especially suitable for the "dunkelflaute" events).
Grid storage lithium batteries for large-scale storage (not things like frequency regulation) are somewhere at the end of the cost scale and it's not something you really want to do since you need those batteries primarily for mobile applications. Even if you wanted lots of stationary batteries, there might be better candidates -- I still want to see some company to cut their teeth on advanced NiFe batteries, for example.
One thing you could do as part of the demand response step, however, is modulating the charging rate of BEVs. But personally I consider this to be more a part of the transportation sector than energy generation as such.
- webmaven 4y ago> You want to start with demand response, then continue with overgeneration, then continue with pumped hydro. Pumped hydro is in the class of solutions I had in mind. An example would be a refrigerated warehouse using off-peak power to freeze large blocks of ice to be used later as heat sinks. > Even if you wanted lots of stationary batteries, there might be better candidates For some applications (an orbital LINAC/railgun, for example) you would want large banks of capacitors rather than batteries.