An Optimization Model for the Design and Operation of Reliable Power Generation Systems

Abstract

Abstract This paper aims to develop a new optimization model for the design and operation of reliable power generation systems. This work optimizes the selection of redundant or backup units and operating units to maximize the reliability and to minimize the cost. In particular, every possible failure state that the power generation systems can have is investigated to evaluate the system reliability. To achieve this goal, we develop an optimization model that minimizes the total cost using Generalized Disjunctive Programming (GDP). The GDP model includes two decision variables: the first is a selection of redundant units with different sizes to increase the reliability of systems, and the second is a selection of operating units to satisfy the power demand. Specifically, the model determines the system reliability and corresponding expected power production by considering the number of redundant and operating units, and possible failure states under each design and operation mode. The model imposes a penalty when the demand is not satisfied, and the system has a low reliability. We have applied the proposed model in a small power plant (one stage with up to three generators) and verified through a sensitivity analysis that the model installs larger and more units to improve the system reliability as penalty rates increase.