Note—Nested Decomposition of a Dynamic Energy Model

Abstract

This paper reports computational experience in using a nested decomposition (staircase) algorithm to solve a version of Manne's linear programming model of U. S. energy options. Nested decomposition is found to be 28% to 52% faster than a direct simplex approach for our test problems. Effects of various computational strategies are also investigated. Our results indicate the applicability of nested decomposition to the important class of staircase linear programs arising from dynamic energy models.