That's it. I studied this issue a few years ago for my work. Below is a rough translation of some salient points I wrote then.
Knowing that the
carbon previously fixed and stored by photosynthesis in the one and half kilo of oil used to produce one liter of gasoline comes from the decomposition and fossilization of nearly 25 tons of organic matter (which is the payload of a semi-trailer of 40 tons) it is easily understood why the deposits of fossil fuels are a providential irreplaceable gift.
Science and technology have limitations and can’t accomplish miracles. The most viable substitute to oil currently available is natural gas (methane) but the automotive industry is a gigantic thing with a considerable inertia. Even if it could immediately switch to produce vehicles running on methane, it takes on average about 15 years to renew the world’s
one billion motor vehicles.
Like electricity, hydrogen is not an energy resource, but only an energy carrier, since there is almost no natural hydrogen usable as fuel on the planet. There are two hydrogen atoms in each molecule of water, but the
extraction of this H2 requires much more energy than its combustion can produce: this is a fundamental law of thermodynamics nothing can change. With the inevitable loss in storage, transportation and final mechanical or thermal applications, the question of the overall efficiency of the conversion chain, "well to wheel" arises then, as in any purposes.
Hydrogen production by electrolysis using electricity has currently an astronomical cost, efficiency does not exceed 70% — although this is the most efficient way. It seems affordable on a large scale only by generating this electricity in additional nuclear power plants: to completely replace oil, we would have to multiply the number of nuclear power plants by a factor 10!
97% of hydrogen is currently produced from fossil fuels. This approach consumes proportionally more fuel than when burned directly, as is the case today. Reforming of methane is the shorter-term option, but why not burn the gas directly in the engine? If hydrogen is produced by exploiting fossil fuels, efficiency losses inherent in the process at least double CO2 emissions compared to the direct use of methane.
While the existing infrastructure can deliver almost any kind of liquid hydrocarbon and electricity as well, not only the large-scale production of hydrogen is problematic, but the infrastructure necessary for the generalization of its use in transport and distribution is also to be built from scratch. Using the pipeline network designed for natural gas is impractical. First, for the same amount of energy, the volume of hydrogen would be tripled compared to methane, which therefore require a tripling of speed in the pipelines. There should be 4 times more powerful compressors, valves, seals and even pipes of different specifications to prevent leakage; losses would be almost doubled. Second, as the network is still used for natural gas, it can not be converted (at great expense) for hydrogen — although it is possible to mix up to 10% or 20% hydrogen in the natural gas network (Hythane).