3 ms·
If you want power at night, what do you do? If it's cloudier one day than another, what do you do? You need baseline production to flatten the peaks and valleys
by rland 5y ago
If you want power at night, what do you do? If it's cloudier one day than another, what do you do? You need baseline production to flatten the peaks and valleys.
If you want to run on solar, you need to provision far more supply of energy than demand (and use a ton of material/land to do so), and then melt rocks or something when there's too much supply. Or build a continental power distribution grid that wastes all the energy you've made to get a trickle of power from over-provisioned areas to under-provisioned ones. Or, you can store it all in batteries. OK? What are losses? How many batteries do we need to build for grid-wide ~12h of storage? The answer is, a lot. What are the costs of mining, building, installing, replacing all those batteries?
The key thing I'm driving at is efficiency. Losses. The efficiency in batteries is not good. The efficiency in moving energy 1000's of miles is not good. Those are required for a majority solar grid.
The conversation here is not about whether it's possible, but whether it is efficient. It is possible after all to power the whole grid with coal.
- pfdietz 5y agoNone of that argument justifies your claim. Yes, storage is needed. No, that doesn't mean solar cannot power civilzation. Simply raising questions doesn't prove your point. It's up to you to show that all the various storage technologies, alone or combination, along with transmission and dispatchable demand, cannot do the job. No, it IS about whether it's possible. You didn't claim it wasn't "efficient", you claimed it wasn't possible. "Can't." I'm not sure why efficiency matters -- is agriculture impossible because its average efficiency at converting sunlight to food energy is less than 1%?