My guess is that the cycle efficiency would be a simple Carnot cycle, and since the delta(T) is small the efficiency would be minuscule and you never measure the difference, never mind pay back the glass.

Your better off putting floating solar panels on top

Not Carnot, it's mechanical. https://en.wikipedia.org/wiki/Pumped-storage_hydroelectricit...

It's not 100%, but it's not bad.

The comment you're replying to is discussing using heat for pressure, not electrical pumps.

Ah, you're absolutely right, that was my mistake. hp48fan seems correct that it would make a minuscule difference.

Are floating solar panels a thing? My naive assumption is that (most places) land is abundant enough that you can put solar panels everywhere. Dealing with the nature-hates-you-engineering of working with water seems like a method of last resort.

You get co-benefits of reducing water evaporation and plant growth (some places just use plastic balls for this same purpose) on the water side and on the solar side you get some cooling which increases efficiency.

For existing dams the ability to share the existing connections in an area that probably isn't flat except for the water is neat.

Pumped hydro energy storage is already low specific energy. Compressed air and lifting large masses are even less economical schemes.

It may be cheaper to just offer time of use pricing to incentivize demand response. EV charging, washing machines, and dishwashers use timers to shift demand to off-peak hours. HVAC and water heaters can be connected via smart grid switches or thermostats. Precooling or preheating can be exploited. During the highest demand days temperatures can be relaxed or units can be duty cycled by splitting customers' units into four groups A B C D and running each for 15 min. to smooth out a soften a sharp spike in total demand.

They pump during the day when solar is strong. (The nuclear plants are long decommissioned.) The pumps would fight the pressure.