The storage problem
Some attention has been given to the role of hydrogen to provide grid energy storage for unpredictable energy sources, like wind power.
The most obvious competitor is pumped storage.
The primary difficulty, with using hydrogen for grid energy storage, is that such use assumes that converting power to hydrogen and back is cheap, which it presently is not.
Water turbines and electric wires are currently more economical than electrolysis plants, fuel cells, and hydrogen pipelines.
Pumped storage is presently more efficient and cost-effective than hydrogen storage.
Although it is generally not referred to as load balancing, the varying rates of supply, refining, and consumption of hydrocarbons are balanced by storing liquid hydrocarbons, generally in the familiar tank farms around refineries.
Such storage is practical because the economic value of the hydrocarbons stored is very large compared to the cost of the tank.
Natural gas is also stored in tanks, but these are much less common because natural gas is much more expensive to store, as it is a low energy density per volume gas which requires a large expensive container.
Typically, if the gas would be stored for longer than a few months and a gas pipeline connection is available, it is cheaper to flare it off and buy more when needed. As a result, the primary form of storage for natural gas is within the distribution pipelines themselves.
Hydrocarbons are stored extensively at the point of use, be it in the gasoline tanks of automobiles or propane tanks hung on the side of barbecue grills. Hydrogen, in comparison, is quite expensive to store or transport with current technology.
Hydrogen gas has good energy density per weight, but poor energy density per volume versus hydrocarbons, hence it requires a larger tank to store.
A large hydrogen tank will be heavier than the small hydrocarbon tank used to store the same amount of energy, all other factors remaining equal.
Increasing gas pressure would improve the energy density per volume, making for smaller, but not lighter container tanks (see pressure vessel).
Compressing a gas will require energy to power the compressor. Higher compression will mean more energy lost to the compression step.
Alternatively, higher volumetric energy density liquid hydrogen may be used as in the Space Shuttle.
However liquid hydrogen is cryogenic and boils around 20.268 K (–253 °C or -423 °F). Hence its liquefaction imposes a large energy loss, used to cool it down to that temperature.
The tanks must also be well insulated to prevent boil off. Ice may form around the tank and help corrode it further if the insulation fails. Insulation for liquid hydrogen tanks is usually expensive and delicate.
Recently there have been some concerns over possible problems related to hydrogen gas leakage.
One issue, that may present itself with widespread hydrogen usage, is permanent hydrogen loss.
Molecular hydrogen is light enough to escape into space. With a continuous cycle of hydrogen being liberated and then combined with oxygen, some will leak from containment.
If significant amounts escape, it has been hypothesized that this may eventually cause an abundance of oxygen and lack of water.
However, it would take a lot of leakage to engender an appreciable and permanent loss-related effect.
Another issue is that hydrogen gas (H2) may form water vapor as it reacts with oxygen and cool, or form free radicals (H) due to ultraviolet radiation, in the stratosphere.
These free radicals can then act as a catalyst for ozone depletion. A large enough increase in stratospheric hydrogen from leaked H2 could exacerbate the depletion process.
However, the issues associated with hydrogen leakage may not really be as much of a problem for various reasons.
The amount of hydrogen which leaks today is much lower (by a factor of 10-100) than the estimated 10%-20% figure conjectured by some researchers; in Germany, for example, the leakage rate is only 0.1% (less than the natural gas leak rate of 0.7%).
At most, such leakage would likely be no more than 1-2% even with widespread hydrogen use, using present technology.
Furthermore, it would take at least 50 years for a mature hydrogen economy to develop, and new technology could likely reduce the leakage rate even more.
Also, less NOx, a major smog contributor and ozone depletor, will be produced than today, if fuel cells are used in the future instead of internal combustion engines.
Assuming all of that is solvable the density problem remains.
Even liquid hydrogen has worse energy density per volume than hydrocarbon fuels such as gasoline by approximately a factor of four.
Ammonia (NH3) can be used to store hydrogen chemically and release it in a catalytic reformer.
Ammonia would provide exceptionally high hydrogen storage densities (greater then liquid hydrogen) as a liquid with mild pressurization constraints, or very mild cryogenic constraint, it can also be stored as a liquid at room temperature and pressure when mixed with water.
Ammonia is the second most commonly produce chemical in the world and a large infrastructure for making, transporting and distributing ammonia already exists.
Ammonia production usually involves raw hydrogen, most of which is produced from natural gas though some of it is produced from water in electrolysis.
Ammonia can be reformed to produce hydrogen with no harmful waste, or can mix with existing fuels and burned efficiently.
Pure ammonia burns poorly and is not a suitable fuel for most combustion engines.
Ammonia is very energy expensive to make.
Existing infrastructure would have to be greatly enlarged to handle replacing transportation energy needs.
Ammonia is a toxic gas at normal pressure and temperature and has a potent odor.
There are proposals to use metal hydrides as the carrier for hydrogen instead of pure hydrogen.
Hydrides can be coerced, in varying degrees of ease, into releasing and absorbing hydrogen.
Some are easy to fuel liquids at ambient temperature and pressure, others are solids which could be turned into pellets.
Proposed hydrides for use in a hydrogen economy include boron and lithium hydrides.
These have good energy density per volume, although their energy density per weight is often worse than the leading hydrocarbon fuels.
Solid hydride storage is a leading contender for automotive storage. A hydride tank is about three times larger and four times heavier than a gasoline tank holding the same energy.
For a standard car, that's about 45 US gallons (0.17 m³) of space and 600 pounds (270 kg) versus 15 US gallons (0.057 m³) and 150 pounds (70 kg).
A standard gasoline tank weighs a few dozen pounds (tens of kilograms) and is made of steel costing less than a dollar a pound ($2.20/kg). Lithium, the primary constituent by weight of a hydride storage vessel, currently costs over $40 a pound ($90/kg). Any hydride will need to be recycled or recharged with hydrogen, either on board the automobile or at a recycling plant.
Often hydrides react by combusting rather violently upon exposure to moist air, and are quite toxic to humans in contact with the skin or eyes, hence cumbersome to handle (see borane, lithium aluminium hydride).
This is why such fuels, despite being proposed and vigorously researched by the space launch industry, have never been used in any actual launch vehicle.
Few hydrides provide low reactivity (high safety) and high hydrogen storage densities (above 10% per mass) Leading candidates are Sodium borohydride and Lithium hydride.
Sodium borohydride can be stored as a liquid when mixed with water, but must be stored at very high concentrations to produce desirable hydrogen densities, thus required complicated water recycling systems in a fuel cell.
As a liquid Sodium borohydride does provide the advantage of being able to react directly in a fuel cell, allowing the production of cheaper fuels cells that does not need platinum catalyses for hydrogen.
Recycling sodium borohydride is energy expensive and would require a recycling plants.
An alternative to hydrides is to use regular hydrocarbon fuels as the hydrogen carrier. Then a small hydrogen reformer would extract the hydrogen as needed by the fuel cell.
The problem is reformers are slow and given the energy losses involved plus the extra cost of the fuel cell you were probably better off burning it in a cheap internal combustion engine to begin with.
Direct Alcohol Fuel Cells can circumnavigate the problem of reforms, but provide lower efficiencies and power densities compared to convention fuel cells, although this could be counter balanced with the much better energy densities of ethanol and methanol over hydrogen. Also alcohol fuel is independent from fossil fuels.
Solid-oxide_fuel_cells can run on light hydrocarbons such as propane and methane with reforms, can run on higher hydrocarbons with only partial reforming, but the high temperatures and slow startup times of these fuel cells makes then prohibitive for automobiles.
More exotic hydrogen carriers based on nanotechnology have been proposed, such as carbon buckyballs and nanotubes, but these are still in the early research stage.
Storage is the main technological problem of a viable hydrogen economy.