Scheduling long term energy storage

Eventually, society will need to be powered almost entirely by renewable energy sources. This will require very large scale storage for long time periods, such as between seasons or even between years. On these timescales, issues such as gradual leakage of energy and price rises must be considered. This paper considers optimal storage management schedules on these long time scales, from the point of view of minimizing the purchase price of electricity while meeting an inelastic demand. We show that the two factors (leakage and price rises) have complementary effects on the optimal storage level, which implies that peak generation is non-monotonic when either is varied by itself.