So here's the post in a single paragraph: dispatchable hydropower is a massive and mostly
unused power storage solution available today to solve the problem of
power variability from wind and solar. The claim that power storage is
technically infeasible is wrong. There would be an economic cost but
it's manageable and getting smaller. Environmental issues could also be
addressed, especially because power can be dispatched without turning
rivers off. Maybe I'm missing something, but pumped hydro storage
seems like just a small part of a bigger solution.
My blogpost headline is a tiny bit exaggerated for effect. I have nothing against
pumped hydro storage, it's currently the biggest and most cost-efficient form of energy storage, and it'll be some years before electric batteries will overtake them.
Pumped storage might be biggest current storage of power, but it's barely a footnote compared to
total hydropower generation. Part of the problem for pumped hydro is that it's difficult to scale because you need a place to put decent sized reservoirs (or maybe two reservoirs, one uphill and one downhill) and you need to construct those reservoirs.
The other reason the description of biggest for current pumped storage might belong in scare quotes is that it's a footnote compared to the need for storage in a sustainable system that doesn't use coal or natural gas.
So what isn't a footnote? Hydropower, generating 16% of the world's electricity. If we stored and released hydropower to make up for the variability of wind and solar that will be the predominant energy sources in 20 years or so, then we'd have a large part of the variability problem solved.
The problem is that hydropower is currently used almost exclusively for baseload and high-demand power instead of dispatch, something that is done almost exclusively for economic reasons. The reasons are understandable - hydropower is some of the cheapest available power and the vast majority of the cost is initial construction while the fuel source is free. So the more power you produce as quickly as possible, the more quickly you can pay back the loans you took out to construct the dam and start earning a profit. To the extent you hold back on power generation, you only do that so you can maximize production during parts of the day when demand and price is the highest.
So okay, but if we have other concerns like not frying the planet, then maybe that should direct when we use the most hydropower and have it happen when wind and solar are not enough. When wind and solar are 50% - 70% of your annual power mix, you still call on those energy sources first on a daily basis and let the water get stored in your reservoir. At night and other low-wind time periods, you let the water out. The storage is so immense I believe it could even cover seasonal issues like the low availability of solar power in high latitude winters.
The economic cost AFAICT is substituting solar and wind power for your very cheap hydropower for baseline and some high-demand power. You still would be able to sell most (not all) of that hydropower but maybe not at as good a price. Yes, there's a cost differential, but it's getting smaller all the time as renewables constantly get cheaper, and again it shows that the storage issue isn't technologically impossible.
Obviously you can't turn a river off and on below a dam, but the flow level already varies quite a bit on a daily basis just for power generation reasons, on the order of 50% or more. Dispatchable hydropower would change why daily flow levels change, but not the fact that daily flow levels already change. Afterbays and dams discharging into still water sections also keep the river from running dry.
Add long distance transmission, less-variable offshore wind, other sources like geothermal and biomass, electric battery, and maybe a little natural gas plus CCS, and it's a sustainable system. Biomass plus CCS gets us to negative carbon emissions.
Maybe I'm missing something. One reason we're not doing this now is we don't need to - there's not enough solar and wind power to make variability a real issue. It will be someday though. Maybe the experts assume hydropower will be dispatchable instead of baseload, but that's not clear to me, nor is it clear why pumped hydropower would get the attention it does.
I do see hybrid systems between pumped storage and traditional hydropower currently happening, like
pump-back hydroelectric dams where water released below a dam is pumped back up during periods of extra or low-cost electricity. In Southern California, two existing dams uphill and downhill from each other
allow for pumped storage. I imagine this could happen in a lot of places, although it might be even easier to just not release water from the upper reservoir rather than pump it back up, assuming the upper reservoir is not an off-stream reservoir. Maybe these hybrid systems are a transition that will get us to using hydropower more consistently as a backup for wind and solar.