But is it usefully dispatchable? Nuclear can be made dispatchable but it's not usefully dispatchable because the costs are fairly similar whether the plant is on or off.

Like nuclear, I believe geothermal has high capital cost and low running costs, suggesting that it isn't usefully dispatchable.

But that's too simplistic. A big limitation of geothermal is that rock has poor thermal conductivity. So once you remove heat it takes a while for it to warm up again. If you're running it 100% then you need a large area to compensate. OTOH, if you're running it at a lower duty cycle you likely need less area.

So if you know the duty cycle in advance, then you can likely significantly reduce costs. Yay!

But that also means that you likely can't run a plant built for low duty cycles continuously for 2 weeks during a dankelflaute. It's likely great for smoothing out daily cycles, but not as good for smoothing out annual cycles. That means it's competing against batteries, which are also great for smoothing out daily cycles, and are very inexpensive.

> I believe geothermal has high capital cost and low running costs

Higher capital costs, but not nuclear high capital costs.

> That means it's competing against batteries, which are also great for smoothing out daily cycles, and are very inexpensive.

It likely would supplement batteries rather than compete against them. A battery buffer would allow a geothermal plant to slowly rise to load and fall as that load goes away.

A very large battery can store 200MWh worth of energy. The largest geothermal plant produces 1.5GW. (A lot of the large plants look like they are in the range of 100->200MW). Presumably those plants can run for more than a few hours which ultimately decreases the amount of batteries needed to smooth out the demand curve.

That’s a lot of words to admit that geothermal has its place.

Nuclear produces very dangerous substances. The long term cost to guard us from them for a million years and the risk that something gets out of control are extemly high.