Electric Power Systems Research 86 (2012) 17–27 Contents lists available at SciVerse ScienceDirect Electric Power Systems Research jou rn al h om epage: www.elsevier.com/locate/epsr Influence of load alterations to optimal network configuration for loss reduction Aggelos S. Bouhouras, Dimitris P. Labridis ∗ Power Systems Laboratory, Dept. of Electrical and Computer Engineering, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece a r t i c l e i n f o Article history: Received 3 March 2010 Received in revised form 17 October 2011 Accepted 21 November 2011 Available online 3 January 2012 Keywords: Load variability Loss reduction Network reconfiguration Real time a b s t r a c t The paper investigates how load alterations in distribution systems influence optimal configurations for loss minimization. In the proposed methodology network reconfigurations are implemented utilizing heuristics techniques while load variations are simulated by stochastic procedures. For the examined topologies initial available load data are considered as mean values and new altered load values are produced using uniform distribution. Various scenarios examined are assumed to simulate actual load conditions in order to examine how load variability may change the optimal configuration derived from the initial mean load values. The proposed algorithm was applied in three well known distribution net- works from published literature and to a real urban distribution network. The results indicate that for altered load conditions, groups of adjacent sectionalizing switches participate in all the configurations procedures. The work concludes that real management of the distribution networks for loss reduction could rely on a realistic approach which considers limited reconfigurations of the network, derived for the mean load values of the assumed time period. Divergences from optimal solutions are shown to be insignificant compared to the reduction of switching operations. © 2011 Elsevier B.V. All rights reserved. 1. Introduction Loss reduction in power systems has constituted one of the most important objectives for researchers and engineers. The constant growth of energy demand along with the polluting conventional power plants has forced engineers in searching methods to reduce losses in all three stages of a power systems’ operation; genera- tion, transmission and distribution. It is estimated that the largest proportion of losses in power networks corresponds to distribution networks; for a typical system in a developing country, distribution losses account approximately 13% of the total energy produced [1]. Over the past three decades considerable research has been con- ducted for loss minimization in the area of distribution systems. The basic concept for loss reduction, developed by Merlin and Back [2], aimed to take advantage of the distribution networks’ structure. Although distribution systems are designed as meshed networks, they operate as radial ones due to reliability and short circuit issues. The existence of tie switches that interconnect feeders and permit load transfer among them has lead to the idea of network recon- figuration for loss reduction. Changes of the network topology are performed by opening sectionalizing (normally closed) switches and closing tie (normally open) switches. All the needed switch- ing operations are implemented in such a way that a number of ∗ Corresponding author. E-mail address: labridis@auth.gr (D.P. Labridis). constraints, i.e. voltage and current limits, radial structure of the network, etc., are not violated. Network reconfiguration for loss reduction has been treated by many researchers and through a great number of different approaches. Although Merlin and Back [2] were the first who intro- duced the concept of distribution system reconfiguration (DSR), Civanlar et al. [3] proposed a purely heuristic algorithm based on a branch exchange method. By this approach they proposed an approximate formula in order to estimate whether a particular switching operation would increase or reduce losses. Shirmoham- madi and Hong [4] based their algorithm on the approach of Merlin and Back including optimal power flow as the basic criterion for the switches that should open. Baran and Wu [5] attempted to improve Civanlars’ method by introducing two approximation formulas for power flow. Moreover, in [6] the methods concerning loss mini- mization algorithms published in IEEE transactions between years 1988 and 2002 are presented. The reconfiguration algorithms may be classified by their solution methods in three basic categories; mathematical optimization methods, heuristics, and those based on Artificial Intelligence. In [7] the authors present a mathemat- ical model for loss minimization which consists in introducing non-conventional group of variables instead of the classical bus complex voltages. The main idea is to simplify the mathematical optimization problem by eliminating continuous and binary vari- ables. The result is to formalize the minimization problem with a linear objective function. Heuristics have kept being proposed by researchers for loss minimization due to their simplicity. In [8] a heuristic algorithm is proposed based on the direction of the branch 0378-7796/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.epsr.2011.11.023