Search

AbstractDistribution systems are undergoing significant changes as they evolve toward the grids of the future, which are known as smart grids (SGs). The perspective of SGs is to facilitate large-scale penetration of distributed generation using renewable energy sources (RESs), encourage the efficient use of energy, reduce systems’ losses, and improve the quality of power. Photovoltaic (PV) systems have become one of the most promising RESs due to the expected cost reduction and the increased efficiency of PV panels and interfacing converters. The ability to forecast power-production information accurately and reliably is of primary importance for the appropriate management of an SG and for making decisions relative to the energy market. Several forecasting methods have been proposed, and many indices have been used to quantify the accuracy of the forecasts of PV power production. Unfortunately, the indices that have been used have deficiencies and usually do not directly account for the economic consequences of forecasting errors in the framework of liberalized electricity markets. In this paper, advanced, more accurate indices are proposed that account directly for the economic consequences of forecasting errors. The proposed indices also were compared to the most frequently used indices in order to demonstrate their different, improved capability. The comparisons were based on the results obtained using a forecasting method based on an artificial neural network. This method was chosen because it was deemed to be one of the most promising methods available due to its capability for forecasting PV power. Numerical applications also are presented that considered an actual PV plant to provide evidence of the forecasting performances of all of the indices that were considered.

AbstractIn recent years, modern distribution networks have rapidly evolved toward complex systems due to the increasing level of penetration of distributed generation units, storage systems, and information and communication technologies. In this framework, power quality disturbances such as waveform distortions should be minimized to guarantee optimal system behavior. This article formulates the planning problem of passive filtering systems in a multi-convertor electrical distribution system as a probabilistic multi-objective optimization problem whose input random variables are characterized with probability density functions. A heuristic simplified approach including trade-off analysis issues is applied to solve the planning problem with the aim of optimizing several objectives and meeting proper probabilistic equality and inequality constraints. The approach is able to quickly find solutions on the Pareto frontier that can help the decision-maker to select the final planning alternative for practical operation. The proposed approach is applied to a 17-busbar distribution test system to evidence its effectiveness.

This paper analyzes the influence of technology uncertainties on the sizing of battery systems. The sizing is based on the minimization of the costs incurred by the end customer and is performed considering demand response applications in the frame of time of use tariffs. The randomness of i) battery round trip efficiency, ii) life time duration, iii) unit costs related to battery capacity, power conversion system, operation and maintenance and replacement is taken into account in order to identify the most convenient solution from an economic and technical point of view. Based on the load requests of actual industrial and residential loads, numerical applications have been performed. The results provided useful information regarding the influence uncertainties have in the choice of a battery energy storage system.

AbstractIn this paper, we propose a novel scheme that jointly employs sliding-window (SW) Prony and the discrete Fourier transform (DFT) for estimating power-quality indices based on the time-frequency analysis of non-stationary waveforms in power-system disturbances. These indices, recently proposed in the relevant literature, require the separation of the waveform in two parts, i.e., the fundamental component and the disturbance waveform. In the proposed scheme, Prony’s method was applied to calculate only the fundamental component from low-pass, filtered data, while the disturbance waveform is obtained by using DFT. The application of Prony’s method only to the fundamental component reduces the computational effort required by this method significantly and allows the size of the window of the DFT to be set as an integer multiple of the estimated fundamental period, which minimizes the spectral leakage of the DFT. The main advantages of the proposed, new approach include high resolution and accuracy in the estimation of indices and significant reduction of the computational cost. Experiments were conducted and comparisons were made using both synthetic and measured data. The results confirmed the effectiveness and efficiency of the proposed scheme.

In this paper, a probabilistic method is proposed to analyze the very short-term steady-state performance of an unbalanced distribution electrical system characterized by the presence of wind farms. This method, which can take into account the uncertainties of loads and wind productions, is based on a Monte Carlo simulation procedure applied to the non-linear three-phase load flow equations, including wind farm models. Bayesian time series models are used to predict the next hour's wind speed probability density functions, making possible a predictive evaluation of the very short-term system steady-state behavior. Numerical applications are presented and discussed with reference to the three-phase unbalanced IEEE 34-bus test distribution system in the presence of wind farms connected at different busbars.

In this paper, a probabilistic method is proposed to analyze the very short-term steady-state performance of an unbalanced distribution electrical system characterized by the presence of wind farms. This method, which can take into account the uncertainties of loads and wind productions, is based on a Monte Carlo simulation procedure applied to the non-linear three-phase load flow equations, including wind farm models. Bayesian time series models are used to predict the next hour's wind speed probability density functions, making possible a predictive evaluation of the very short-term system steady-state behavior. Numerical applications are presented and discussed with reference to the three-phase unbalanced IEEE 34-bus test distribution system in the presence of wind farms connected at different busbars.

The steady state thermal rating of overhead transmission lines is limited by the conductor`s maximum design temperature, which is related to the maximum sag and the loss of tensile strength of the conductors. Traditionally, this overhead transmission lines rating is computed using a deterministic approach, with reference to severe weather conditions. Thus, the application of this method leads to conservative results resulting in under-utilization of conductors. In this paper, a new method based on an hourly probabilistic index is proposed to predict the line thermal rating for each hour of the day; this index is evaluated using the conductor current limit probability density function (pdf). The method uses Bayesian time series models for the weather parameters (ambient temperature, wind speed and wind direction) and calculates the conductor current limit pdf using a Monte Carlo simulation. The probabilistic index is applied by considering measured weather data of both hot and cold seasons; the corresponding lines ratings are reported and analyzed.

The steady state thermal rating of overhead transmission lines is limited by the conductor`s maximum design temperature, which is related to the maximum sag and the loss of tensile strength of the conductors. Traditionally, this overhead transmission lines rating is computed using a deterministic approach, with reference to severe weather conditions. Thus, the application of this method leads to conservative results resulting in under-utilization of conductors. In this paper, a new method based on an hourly probabilistic index is proposed to predict the line thermal rating for each hour of the day; this index is evaluated using the conductor current limit probability density function (pdf). The method uses Bayesian time series models for the weather parameters (ambient temperature, wind speed and wind direction) and calculates the conductor current limit pdf using a Monte Carlo simulation. The probabilistic index is applied by considering measured weather data of both hot and cold seasons; the corresponding lines ratings are reported and analyzed.

This paper investigates the problem of contemporaneously choosing optimal locations and sizes for both shunt capacitors and series voltage regulators in three-phase unbalanced distribution systems. The sizing and placement procedure not only minimizes the power losses along distribution feeders but also makes sure that both capacitors and series regulators will have the minimum possible impact on the harmonic distortion of bus voltages in the system. The fact that distribution systems can operate under unbalanced loading conditions means that the optimization will have to account for any unbalances in the system. A genetic algorithm, which successfully solves the above-described problem, is developed and tested. This paper presents the results of simulations for a 34-bus IEEE distribution test system.

This paper investigates the problem of contemporaneously choosing optimal locations and sizes for both shunt capacitors and series voltage regulators in three-phase unbalanced distribution systems. The sizing and placement procedure not only minimizes the power losses along distribution feeders but also makes sure that both capacitors and series regulators will have the minimum possible impact on the harmonic distortion of bus voltages in the system. The fact that distribution systems can operate under unbalanced loading conditions means that the optimization will have to account for any unbalances in the system. A genetic algorithm, which successfully solves the above-described problem, is developed and tested. This paper presents the results of simulations for a 34-bus IEEE distribution test system.