Improved algorithms for reservoir capacity calculation incorporating storage-dependent losses and reliability norm

Two algorithms that improve upon the sequent‐peak procedure for reservoir capacity calculation are presented. The first incorporates storage‐dependent losses (like evaporation losses) exactly as the standard linear programming formulation does. The second extends the first so as to enable designing with less than maximum reliability even when allowable shortfall in any failure year is also specified. Together, the algorithms provide a more accurate, flexible and yet fast method of calculating the storage capacity requirement in preliminary screening and optimization models.