“Assuming that power could evenly be bought at €20 (US$30) per megawatt-hour during peak production and sold at €70 (US$100) per MWh during peak consumption, and assuming that the system had 70 percent conversion efficiency, Heindl expects a return on investment in four years for a system with a 500 meter radius and 1 kilometer depth.”

That’s a $0.03/kWh to $0.10/kWh spread. If the system can operate at a profit and store for 7 cents per kWh then it has promise.

There are utility scale batteries (Eos Systems) that are going on the grid in the next few months and claim to be able to store for 10 cents.

I’d bet that batteries can hit the 7 cent price point. And they will have significant advantages in the ease at which they can be sited and distributed around the grid.

Pump-up hydro costs about the same, but it harder to site as you point out.

It’s going to be interesting to watch the storage puzzle being solved….

There is a deal on the issue to use waterstorage capacity in norway.
Another ‘just an Idea’ for the future are Hydraulic Hydro Storages.

http://blogs.worldwatch.org/revolt/unconventional-%E2%80%9Chydraulic-hydro-storage%E2%80%9D-system-offers-energy-storage-for-the-grid-on-a-grand-scale/

This part is interesting.

The overall math is going to be even more interesting. With a 50% inefficiency in production and some other level of inefficiency in conversion of hydrogen to electricity (50% in a fuel cell?) the input costs for an output kWh of electricity are going to be high. Ten cents in the front end might be worth only 2-3 cents out the other end.

Pump-up hydro, running about 85% efficient has a big head start. Hydro can be used for very short term grid smoothing and daily time shifting electricity. The hydrogen approach is too lossy to make sense for those uses. And by storing a “three/whatever day full output” reserve of water in the reservoir the turbines can be used in the infrequent period of very low renewable input.

The more frequently a system gets used the cheaper it is. You can spread the fixed costs over more units of sales.

BTW, latest Gen 3+ nuclear reactors can reduce the increase / decrease the load in 5 minutes. When the electricity / hydrogen storage comes up, nuclear power will also benefit.

So if the electricity costs negative to use, why not the French buy some batteries to get charged at that time, later when the home needs more, that electricity can be drawn from battery.

I think long term it is a mixture.
– Fast/short time: UltraCaps or flywheels (real time balancing)
– Time of day shifting: Batteries, compressed air, pump storage, heat/cold storage (for home, commerical, industry)
– Shifting of days/weeks is not yet clear to me. Maybe it is cheaper to use time of day. Build to 110% for time of year at lowest production. Then when time of day shifting systems full crank up CO2 capture machine use the extra juice that way.
But then we are still a ways off from that problem, and better solutions will likely come along.

However, you are right that some stuff needed the nozzels and the pressure limiters changed to adjust the flow. Real modern gas heating might not care at all because it is just measuring the oxygen content in the exhaust gas and will be readjusting the ratios. But there will be problems if H2 is mixed in the gasnetwork above a certain %tage. Especially with pressurized storage tanks that keep the gas pressure adequate in the network. The H2 would leak out and damage ferrous metals in the process. So they are looking at transforming that stuff to CH4.

Still I think they should rather make NH3 from it instead of sourcing the H2 needed for that from natgas.