18 th Power Systems Computation Conference Wroclaw, Poland – August 18-22, 2014 Probabilistic Assessment of the Impact of Wind Power Generation on Voltage Sags in Composite Systems João Eduardo Ribeiro Baptista* Anselmo Barbosa Rodrigues * Maria da Guia da Silva * joao.baptista89@hotmail.com anselmo.ufma@gmail.com guia@dee.ufma.br * Electrical Engineering Department Federal University of Maranhão (UFMA) São Luís-MA, Brazil Abstract— This paper presents a methodology to assess the impact of wind power generation on the voltage sags in generation and transmission composite systems. The proposed methodology is based on probabilistic techniques to model uncertainties associated with fault scenarios and wind speed fluctuations. The probabilistic technique used to represent the uncertainties is the non-sequential Monte Carlo Simulation. The voltage sags resulting from fault scenarios were evaluated using symmetrical components technique. The combination of these two techniques was used to estimate the nodal and systemic indices associated with the frequency of voltage sags. The proposed methodology to evaluate these indices was tested on the IEEE Reliability test system of 24 buses. The tests results demonstrate that the wind generation has a significant impact on the frequency indices related to voltage sags. Keywords—power quality; voltage sags; wind power generation, probabilistic methods; Monte Carlo simulation; Clustering. I. INTRODUCTION Currently, constraints associated with the greenhouse gas emissions have motivated the use of the renewable energy resources such as: wind, solar, biomass, tide, waves, small hydro stations, etc. The wind energy generators have achieved large penetration in power systems around the world due to the advances in the construction of large scale turbines and the low operational costs compared with conventional thermal generators. A common issues related to all renewable energy resources is the variability of the primary energy source (wind speed, solar irradiance, water inflows, etc.). In the wind power generation, the stochastic behaviour of the wind speed introduces significant uncertainties in several areas of the power system planning and operation, such as: voltage and reactive power control, optimal generation dispatch, reliability, short-circuit levels, etc. The short circuits are the main cause of voltage sags. That is, a reduction in the RMS voltage between 0.1 and 0.9 pu with duration in the interval from half cycle to one minute [1], [2]. Consequently, the fluctuations in the power output of the wind turbines have potential to cause power quality problems in sensitive loads such as power electronic devices and microprocessor-based controls. On the other hand, the severity of the voltage sags is also affected by the pre-fault conditions. Due to this, wind power generation has a double impact on the voltage sags profile of an electrical network: (i) pre-fault: changes in the base case voltage profile caused by random variations in the wind speed; (ii) post-fault state: variations in the fault currents. A suitable assessment of these two factors must be carried out in a probabilistic framework to accurately recognize the stochastic nature of the wind power generation. The main advantage of the probabilistic methods is the ability to combine severity and probability to provide a true assessment of the system risk [3], [4]. Probabilistic methods have been used for decades in reliability studies to evaluate indices such as: frequency, duration and expected load curtailments [3], [4]. The probabilistic methods have also been applied in power quality studies associated with voltage sags [5]-[12]. These applications are based on two methods: analytical [5], [6], [7], [9], [11] and Monte Carlo Simulation (MCS) [8], [10], [12]. The probabilistic assessment of voltage sags through the analytical method is usually based in two techniques: critical distances [5], [6] and fault positions [9]. Furthermore, other analytical techniques have been proposed to reduce the computational cost of the fault position method and to obtain the probability distributions related to the voltage sags in specified load points [7], [11]. These techniques are based on the proposed approach in [13] to evaluate post-fault voltages due to short circuits along lines. On the other hand, the application of Monte Carlo Simulation in the probabilistic voltage sag assessment has been based on state space representation, that is, non-sequential MCS [10], [12]. The computational effort of the analytical approaches to carry out a probabilistic analysis of voltage sags is usually lower than the one associated with MCS. However, the MCS is more flexible to represent system operations issues [3], [4]. In spite of the high penetration of wind power generation in electric networks and of the significant number of publications on the probabilistic voltage sag analysis, the impact of wind power generation on the voltage sag profile has not been assessed in a probabilistic framework. In this way, the main aim of this paper is to carry out a probabilistic assessment of voltage sags in composite system (generation and transmission) considering uncertainties associated with