Electric Power Systems Research 58 (2001) 45 – 51 A reliable approach for solving the transmission network expansion planning problem using genetic algorithms H.A. Gil *, E.L. da Silva Uniersidade Federal de Santa Catarina, CTC/Department of Electrical Engineering /LabPlan, CEP 88040 -900, Florianopolis, SC, Brazil Received 4 February 2000; accepted 4 April 2000 Abstract This paper presents a reliable approach for solving the transmission network expansion planning (TNEP) problem through a genetic algorithm (GA). GAs have demonstrated the ability to deal with non-convex, non-linear, integer-mixed optimization problems, such as the TNEP problem, better than a number of mathematical methodologies. The procedure presented consists on finding unfeasible solutions for the problem through the GA. These solutions are used for predicting the cost of the optimum solution using a ‘loss of load limit curve’, of the transmission system. Once this cost is estimated, the optimum solution can be found by performing a local search starting from the unfeasible solutions that have costs close to the estimated cost. This approach makes the GA more robust and reliable for solving the problem for different transmission systems. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Transmission network expansion planning; Genetic algorithms; Optimization www.elsevier.com/locate/epsr 1. Introduction The transmission network expansion planning (TNEP) problem consists of defining when and where new circuits should be installed to serve, in an optimal way, the growing electric energy market, subject to a set of electrical, economic, financial, social and environ- mental constraints. Strictly speaking, such a problem has a dynamic nature, since the requirements of trans- mission facilities (lines or power transformers) should be defined over time within a given horizon. In fact, this problem consists in the execution of two successive tasks, the first one corresponds to the defini- tion of the set of candidate transmission routes whereas the second one, corresponds to the selection of the best subset of circuits among the set of options found before. Our work is addressed to this second task. This complex problem is often simplified by the planners who use mathematical models to solve a ‘static’ TNEP problem (STNEP) problem, which con- sists of minimizing the investment costs of new trans- mission facilities, subject to operational constraints, to meet the power system requirements for a single future demand and generation configuration expected in some future year, which may be, for instance, 5, 10 or 20 years from now. In developing countries, the STNEP problem re- quires a careful evaluation. In Brazil, investments in transmission will represent about 40% of the total investment in the electric energy sector during the next five years, which is calculated at about US$ 30 billion [1]. Therefore, any effort for reducing the cost of the transmission system expansion by some fraction of a percent, allows savings of a significant amount of capital. Furthermore, under a new deregulated model, trans- mission systems play a neutral role in order to increase the competition in the generation and electricity retail- ing, so that the transmission system planning should be carried out in order to minimize the global costs of expansion and operation. Thus, we argue that the transmission projects should be planned and developed by a transmission planning body on behalf of all con- sumers within the market, such as is proposed by Ilic et al. [2]. The centralized transmission planning is also impor- tant because it incorporates a ‘long-run view’ into the * Corresponding author.. 0378-7796/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII:S0378-7796(01)00102-X