Hydrogen is simply less efficient than batteries. It is only useful in circumstances where batteries are inconveniently heavy and bulky. Such as earth moving machinery in remote areas.
@ninalanyon Only less efficient if part of that fundamentally inefficient national-grid system shown by that diagram, which appears to be an anti-solar power campaign image.
The British earth-moving and agricultural machinery manufacturer JCB is developing, in fact I think now producing, hydrogen-fuelled versions. I don't know, but guess, the gas is used in internal-combustion engines rather than via fuel-cells. The market is still likely to be fairly small because hydrogen in far less readily available than Diesel oil; but this will change especially if methane/hydrogen blends or staright hydrogen replaces methane-only as "mains" gas..
There are also trials under way of hydrogen-fuelled railway locomotives or multiple-unit trains; particularly for use on non-electrified lines.
"Remote areas" do not really come into it, because you still need take the fuel to the machines' locations. It could even be argued that the hydrogen would be better used in non-remote locations such as town-centres.
It is not right though to say that the exhaust from i.c. engines burning hydrogen is "only" water-vapour, as so often claimed. It will contain some nitrous-oxide too, though presumably that can be dealt with by urea solution as on modern Diesel vehicles. I don't know how it is done, but do know the reaction produces nitrogen and oxygen. The reagent is the pale blue fluid sold in most filling-stations, in the UK at least, as 'Ad-Blue'.
@ArishMell The end to end efficiency of a hydrogen fuelled internal combustion engine is less than the end to end efficiency of modern batteries. When you burn the hydrogen the efficiency is limited by Carnot's Theorem [1] [2]:
The autoignition temperature of hydrogen is 535 C, 808 K [3]. If we assume that the exhaust temperature is 50 C, 323 K then the maximum possible efficiency is
1 - 323 / 808 = 0.6
Lithium ion batteries have an efficiency of about 0.9 [4]
So the battery vehicle is more efficient than the H2 ICE vehicle. This analysis also omits the efficiency of the H2 production process which further reduces the end to end efficiency of hydrogen. Such a high flame temperature also increases the production of NOx to greater degree than for a diesel engine so that more urea is required.
That mathematics is remarkably simple for something so profound; but I find the concepts of entropy (and enthalpy) hard to understand.
I think for practical purposes the question of relative efficiency, is not sufficient. We'd need consider all those other factors such as production, transport or supply of electricity, etc. Environmentally too, it would come to which is the more sustainable, or more accurately perhaps the less unsustainable.
It's also worth considering that the battery manufacturers are all trying to make ever better batteries, especially searching for active ingredients better than lithium.
The JCB earth-movers might use i.c. engines but I think the locomotives being developed use fuel-cells. I have wondered if electric multiple unit trains using batteries and control systems similar to those in road vehicles could be effective, especially on relatively short routes like the Leeds-Settle-Carlisle that climb over big fells but the descent can partially re-charge the batteries. (Even over the fjells North of Trondheim?) It may need some trade-off between sustained speed, so lengthening journey time, and electricity consumption. One option may be use the overhead lines as far as they go (presently Leeds to Skipton in that first example), then switch to the batteries.
However, the OP shows a diagram of a (presumably national) electricity supply system with a complicated hydrogen "middle-man" stage that must make the overall system, from sunlight or wind-force to 13A mains socket, a lot less efficient overall than it should be.
What that proposed system does do, is remove the need for huge battery-banks; but whether that would make the system relatively more desirable on environmental grounds is anyone's guess.
In the end, the old saw about the "free lunch being a myth" is right!.
We'd need consider all those other factors such as production, transport or supply of electricity, etc.
That's what end to end is intended to convey.
Assuming solar power the process for H2 ICE is something like this
Solar panels -> inverter -> transformer -> transmission lines -> transformer -> electrolysis -> compression -> road and/or rail transport -> transfer to static tanks -> transfer to vehicle
The are more steps in the H2 process all of which introduce additional losses. Each step requires more land and creates more pollution.
The OP's diagram simply replaces grid connected batteries used for peak power supply and frequency control with electrolysis and fuel cells. The only advantage there is that the fuel cells and H2 store probably occupy less space than a battery of equivalent capacity. But remember that they also need a supply of water. For every molecule of hydrogen produced you need a molecule of water, H2O. The atomic mass of oxygen is 16 so a molecular mass of water is 18, that's 9 times as much as a molecule of hydrogen, H2. So to produce a ton of hydrogen you need 9 tons of water.
The Hornsdale Power Reserve battery supplied by Tesla is a 150 MW (194 MWh) grid-connected energy storage system [1].
To run a 200 kW fuel cell for 8 hours, (200 kW x 8 hours x 3600s) / (51% x 120,000 kJ / kg) = 94 kg of hydrogen is required. [2]
That's a capacity of 1.6 MWh. To get the same capacity as the Hornsdale battery you will need 194 / 1.6 * 95 kg = 11.5 tonne of hydrogen and therefore 104 tonne of water.
I haven't been able to find any information on the average actual output of the Hornsdale installation so i don't know how big an impact the water requirement would be but as it is in Australia I imagine that it would be at least worth considering. I don't know how fast the electrolysis cells work either but that must also be a consideration as the entire capacity of the battery can be depleted in less than 90 minutes. I imagine that the electrolysis cells would have to be significantly over dimensioned in order to cope with the peak demand.
