Genetic algorithm model to control peak demand to defer capacity investment

Abstract

This paper formulates and develops a peak demand control tool for electric systems within the framework of direct load control plans. This tool allows defining a load dispatch centre for central air conditioning systems in commercial buildings, hence allowing a measured control of peak demand for such pieces of equipment, which are known for their important influence in the end customers' consumption and for the correlation their demand curve has with the system demand curve during summer months. Traditionally, this type of application has been developed in the field of demand management; however, the high energy consumption growth rates have taken electric firms to analyze their application on the system expansion planning, hence deferring, or even preventing, the need to invest in capacity to supply the demand during peak periods. The generic model presented herein is evaluated in an actual urban substation, characterized by a predominant commercial consumption, by the contribution of the air conditioning systems in the substation loads, and by the problems present in its capacity to expand; model that is solved through advanced genetic algorithm techniques.