An Evolutionary Algorithm for Reactive Power Compensation in Radial Distribution Networks Carlos Henggeler Antunes (1,2), Carlos Barrico (2,4), Álvaro Gomes (1,2) , Dulce Fernão Pires (3), António Gomes Martins (1,2) (1) Dept. of Electrical Engineering and Computers, Polo II, University of Coimbra, 3030 Coimbra, Portugal. (2) INESC Coimbra, Rua Antero de Quental 199, 3030 Coimbra, Portugal (3) Escola Superior de Tecnologia de Setúbal, Instituto Politécnico de Setúbal, 2910-761 Setúbal, Portugal (4) Dept. of Informatics, University of Beira Interior, 6201 Covilhã, Portugal Abstract In this paper, the problem of locating and sizing capacitors for reactive power compensation in electric radial distribution networks is modeled as a multi-objective programming problem. An evolutionary approach consisting of an elitist genetic algorithm with secondary population is used to characterize the Pareto Optimal (non-dominated) frontier and provide decision-makers and planning engineers insightful information about the tradeoffs to be made. Two objective functions of technical and economical nature are explicitly considered in this model: minimization of system losses and minimization of capacitor installation costs. Constraints refer to quality of service (related with the node voltage profile), power flow (associated with physical laws in electric networks), and technical requirements. Keywords Genetic algorithms, Multi-objective models, Power factor compensation, capacitor location, Distribution networks 1. Introduction Reactive power compensation is an important issue in electric power systems, involving operational, economical and quality of service aspects. Consumer loads (residential, industrial, service sector, etc.) impose active and reactive power demand, depending on their characteristics. Active power is converted into “useful” energy, such as light or heat. Reactive power must be compensated to guarantee an efficient delivery of active power to loads, thus releasing system capacity, reducing system losses, and improving system power factor and bus