The efficiency argument is a huge canard for the applications for which hydrogen makes sense. The "cost of inefficiency" is proportional to the number of charge-discharge cycles. For annual seasonal leveling, the cost is very small compared to diurnal storage.
Hydrogen critics also say "just overbuild the renewables instead". Apparently, dropping excess power on the floor giving an efficiency of its use of ZERO is to be preferred to making hydrogen with it at efficiency great than zero.
Being 25% efficient means you need massive surplus to offset any kind of seasonal deficits, but by building those massive surpluses you reduce those same seasonal deficits. On top of this you don’t want a grid that breaks down due to unusual weather or really any predictable issues so you want to build in excess production, and again that safety net further reduces seasonal deficits.
Hydrogen might work as a kind of black swan protection, but storing hydrogen for very long periods isn’t cheap or easy. Hydrogen embrittlement is a problem.
At best it might fit the edge case that current reserve natural gas power plants do. But, that’s such a tiny percentage of annual demand we could just use natural gas without significant global warming concerns. I am fine with a 99.95% green grid, at that point many other issues need to be addressed.
Storing hydrogen underground for long periods is actually shockingly cheap. It's the same technology used for storing natural gas underground. The cost is as little as $1/kWh of storage capacity. This is already proven technology, so no whining about embrittlement please.
Yes, it fits edge cases -- but it has an outsized effect on the cost of a 100% renewable grid. Trying to cover black swan events with batteries or overprovisioning would be much more expensive. Optimizing baseload output in Germany using historical weather data and 2030 estimated cost figures for PV, wind, batteries and hydrogen, including hydrogen cuts the cost of the system in half.
PV, wind, batteries and hydrogen ignores hydroelectric generation making such numbers pure fantasy. Of there where any need to do so we can use them for seasonal storage simply by not using them when doing so is unnecessary.
I don't think hydro can displace hydrogen for these applications. In particular, I don't think hydro has enough energy storage capacity to provide seasonal load leveling in many places.
Hydrogen critics also say "just overbuild the renewables instead". Apparently, dropping excess power on the floor giving an efficiency of its use of ZERO is to be preferred to making hydrogen with it at efficiency great than zero.