Where's the difference that makes it feasible for powering the grid, but not for powering vehicles?
To power the grid from excess solar, you need a hydrogen-burning generator at the site of the electrolysis plant.
Why does this keep coming up? Who cares about this specific point? If there is demand, people will build the fuel stations, like we did with gas and oil?
I get there are other harder problems with Hydrogen, but worrying about infrastructure seems strange?
Hydrogen infrastructure, if what we're talking about is on the scale of current ICE cars, can't "organically" compete, and won't "organically" develop. It has no chance without a massive government investment/buy-in on a decade long scale. It needs huge investment for each of the big IFs I outlined above.
Look at the cost curves of solar / wind / batteries for the last 10 years, and consider that there are numerous technologies about to be rolled out that will improve those at the same or better rate and the overall industrial scaleout / economies of scale isn't close to being done.
Hydrogen is marketed these days with "hey it will cost X and that is (theoretically / lies damn lies and accounting) competitive with wind/solar/batteries (which exist and we KNOW what they cost) price today.
But there is absolutely no chance of hydrogen being cost competitive in general power transmission/carrier, transportation, etc in 10-15 years. Heck in 5 years it likely won't.
So the hydrogen people are praying that their lobbying + fortune would produce some boondoggle subsidy to build this out, and then the government would limp it along due to the sunk cost fallacy.
Energy density. For the fuel alone, volumetrically. Altogether, massively as well.
By weight, hydrogen and oxygen creates 20x more chemical energy per kilogram than lithium-ion batteries [1]. (It's comparable to gasoline and air.) By volume, uncompressed, it has far less. Once compressed, we have to take into account the weight of the containment tanks, the weight of the fuel cell and the energy lost in converting hydrogen to electricity, an output batteries directly provide, a combination which negates that specific energy advantage.
[1] https://en.wikipedia.org/wiki/Energy_density_Extended_Refere...
* Hydrogen is very light but takes up quite a lot of volume so it needs to be either liquified or compressed to very high pressures (and even then it's considerably bulkier than, say, natural gas). * The resulting storage tanks are heavy and add bulk. * The fuel cell to convert the hydrogen to electricity adds weight that a battery system doesn't have (the storage is the converter). * The energy density of lithium-ion batteries doesn't have to match hydrogen or gasoline, it just has to be good enough. For most light vehicle applications, we're getting pretty close to that point (though possibly not if you're planning to use a truck to tow things).
For heavy transport, the balance changes because big tanks are more mass-efficient than smaller ones, and the sheer mass of batteries currently required for long-haul trucking seriously cuts into the cargo that can be legally carried with road mass limits. That's why there's interest in hydrogen-fuelled trucks.
For hydrogen-fuelled planes, a similar argument applies with current and reasonably foreseeable battery technology you can't build an airliner with a useful carrying capacity and range. However, the bulk of hydrogen tanks required for a plane with intercontinental range is still a big problem. That's why you see all these unconventional body design concepts for hydrogen-fuelled planes - you need lots of room to store the hydrogen and still give a useful passenger load.