Hydrogen Energy Storage for Long Time Scale Energy Production Fluctuations

The possibility of using hydrogen as major energy carrier in a post fossil fuel world has been condemned as economically impractical by a number of a number of technical/economic analysts. Among the economic/technical problems associated with this energy carrier are:

  1. The costs associated with electrocatalytic or photocatalytic decomposition of water hydrogen a very high expensive fuel compared to natural gas.
  2. The cost of transporting hydrogen over long distances is much higher than that of natural gas and potentially involves large energy losses (e.g. in the process of liquefaction) which further increases the cost of a unit of energy delivered to end users.
  3. Hydrogen fuel cells which are the primary means being developed to convert the chemical potential energy of hydrogen to electrical energy are expensive and short lived compared to the mature technologies (e.g. Otto cycle engines, diesel engines, Rankine cycle steam turbines, and combined cycle gas turbines) used to convert the energy of fossil hydrocarbons to electrical and mechanical energy.
  4. The round trip efficiency of electricity storage by means of an electrolyzer/fuel cell combinantion is less than 50%

In spite of these condemnations, interest in hydrogen as an energy storage medium continues. Can this interest be justified, particularly in the light of the falling prices of lithium ion batteries which is currently making them popular as a grid storage medium? I think that this answer to this question is 'yes'. Current lithium ion battery grid applications supply energy for time periods in the range of four to six hours. However, renewable energy flows experience variation on much larger time scales than this. In the case of summer to winter insolation difference the time scale of the fluctuation is six months.

Suppose that we wish use stored energy over a time period of 60 hours (2.5 day) rather than for six hours. The up front storage cost will increase by a factor of 10. Further let us suppose such long time period draw downs of stored energy occur on average once every ten days. A thousand cycles of the energy storage system will then occupy 27 years of time. If a battery has a cycle life of 3000 cycles, then 81 years would be required to obtain full value from the battery system. Even if the batteries were designed to last for such a long time the excess costs during the transition period before the power system has reached long term equilibrium would be substantial. For the six month summer/winter solar variation the economic barriers to using battery energy storage would be even more formidable.

On the other hand suppose we wish supply power for 60 hour periods once every 10 days using hydrogen produced by electrolysis. An electrolyzer that is capable of producing six hour of hydrogen on an average day could be run continuously for ten days and the hydrogen could be drawn down over a 60 hour period to run a fuel cell. Of course the storage tank would need to be ten times larger than if the maximum run time of the fuel cell was six hours. However, hydrogen fuel tanks are relatively much cheaper than the energy storage electrodes of batteries. Even longer periods of power production would be possible using the same six hours per day electrolyzer. The electrolyzer could be run continuously during the half year with the highest solar input and the hydrogen supply could be drawn down during the part of the year with lower incident solar energy flux. Again, this longer draw down time would require another substantial increase in the size of the hydrogen storage reservoir. However, large storage caches for hydrogen in underground caverns are likely to be much much cheaper than huge banks of energy storage electrodes holding several months worth of energy.

Aug 29, 2017

rogerkb at energystoragenews dot com