ABSTRACT. In this study, an inexact inventory nonlinear programming (IINP) model is proposed for supporting electric power system planning under multiple unit prices and uncertain demands. The proposed IINP can deal with uncertainties presented as intervals and address nonlinearities in the objective function. It can also help to solve material supply problem with diverse unit prices as well as what, where, when and how much material should be purchased under uncertainty. Then, the IINP is applied to a case study of energy resources supply planning for an electric power system. Results obtained are useful for supporting (a) determination of reasonable energy resources supply scheme with global solutions, and elimination of step by step comparisons among many purchase schemes, (b) adjustment or justification purchase batches of energy resources supply and facility expansion for power-conversion technologies under different demand levels, and (c) integration of policies regarding energy resources supply, economy objective and environmental protection for in-depth analysis of tradeoff between purchase batch and unit price as well as system cost and risk.