This paper proposes a fast and efficient strategy to determine the number and locations of fault indicators (FI) in medium voltage distribution networks. The objective function takes into account the most important characteristics of the network such as topology, non-uniform failure rates of the lines as well as the power demand and number of customers. For seeking the minimum of the objective function, a straightforward algorithm based on the Pattern Search (PS) is developed. The methodology is tested on a real distribution network, showing its great potential to improve the reliability indices at the lowest investment costs.

For development and testing of Fault Passage Indicators (FPIs) a laboratory model of a three-phase overhead line has to be constructed. For reasons of current source intensity and space limitations in laboratory conditions, the phase currents and support pole dimensions should be properly scaled in order to achieve same values of magnetic flux density that would appear in normal FPI operating conditions. Scaling laws for calculating the phase currents for different types of faults and various support pole dimensions of the three-phase line model are proposed in this paper. Scaled fault currents and dimensions of the three-phase line model are calculated using the proposed scaling equations and a laboratory model is constructed. The proposed methodology is verified by measurements on the laboratory model.

Distributed generation (DG) has the potential to bring respectable benefits to electricity customers, distribution utilities and community in general. Among the customer benefits, the most important are the electricity bill reduction, reliability improvement, use of recovered heat, and qualifying for financial incentives. In this paper, an integrated cost-benefit methodology for assessment of customer-driven DG is presented. Target customers are the industrial and commercial end-users that are critically dependent on electricity supply, due to high consumption, high power peak demand or high electricity supply reliability requirements. Stochastic inputs are represented by the appropriate probability models and then the Monte Carlo simulation is employed for each investment alternative. The obtained probability distributions for the prospective profit are used to assess the risk, compare the alternatives and make decisions.

In this paper we present a methodology for economic assessment of installing distributed generation (DG) for industrial/ commercial customers. We propose an approach where the benefits of improved reliability and optimal DG operation are taken into account, respecting the overall lifetime of investment and converting the monetary amounts into their present values. The method is applied by comparing the finite number of market available investment options, detecting the profitable ones, and in case of a positive answer, declaring the best solution for customer.

Abstract Distributed generation (DG) may represent a viable option for the consumers to decrease their energy bills. In order to achieve maximum savings, the generating units must be operated according to optimal dispatch schedules. Due to various electricity pricing tariffs, which often include the peak demand component, this task is different from solving the classical economic dispatch problem. There are several optimization algorithms presented in the literature intended to help the consumers to find the best dispatch solutions in order to minimize overall energy costs. In this paper, one of those optimization algorithms is used as an auxiliary tool for economic evaluation concerning installation of new generating facilities.

Distributed energy resources (DER), located close to end use loads, have the potential to meet consumer requirements more efficiently than the existing centralized grid. In this paper, a specific scenario is considered, where the consumer center uses distributed resources to meet a part of its energy requirements, while the rest of the needed electricity is purchased from the utility distribution company. In order to minimize total consumer costs, a robust optimization algorithm based on a search method has been developed. The main outputs of this algorithm are optimal dispatch between distributed resources generation and the electricity purchased from the utility as well as the particular contribution of every on-site generating unit in the dispatch schedule. The algorithm was tested using several characteristic simulations and two basic areas of its application have been discovered - optimal dispatching of existing distributed energy resources and economic evaluation concerning installation of new generating facilities.

1 University of Banja Luka, Department of Power Engineering, Patre 5, 78000 Banja Luka, Bosnia&Herzegovina, E-mail: cedomir@etfbl.net 2 University of Banja Luka, Department of Power Engineering, Patre 5, 78000 Banja Luka, Bosnia&Herzegovina, E-mail: bilchy@blic.net Abstract: Power-flow problems in electrical power networks have been investigated over the years and several calculation methods were introduced. In order to determine the work state of a power network, it is often necessary to solve a large system of non-linear transcendent equations. Therefore, the program routines are complex to develop, while specialized commercial software packages are expensive and not easily available. On the other hand, Microsoft Excel as a part of Microsoft Office software package is widely-spread in the world and relatively easily available. Excel possesses a plug-in named Solver which is a powerful tool for constrained optimization and solving of non-linear equations. Solver can be used in power network calculations for both practical and educational purposes. Solver is also applicable to the other problems concerning electrical theory and practice (circuit theory, electromagnetics, etc.).

—The electricity customers may use photovoltaic systems supported by batteries in order to fulfill a fraction of their energy requirements and to decrease the peak demand. The achievable savings primarily depend on a system control strategy. In this paper, one algorithm based on a threshold control is described and tested. The sensitivity on the most important input variables is analyzed by extensive set of numerical simulations.