On-line estimation of wind farm transient stability

Abstract

In the paper the probabilistic evaluation of wind farm transient stability is discussed and an on-line statistical inference approach for this evaluation is proposed. More specifically, the paper deals with the transient stability aspect of wind generation integration into power system, starting from the observation that the impact of distributed generation is able to affect the overall dynamic behavior; in this framework, the paper is devoted to the definition and evaluation of the transient stability probability (TSP). TSP is defined and evaluated by taking into account the random nature of both the prime mover power and the fault clearing time. Once that the probabilistic definition and assessment are illustrated, the paper aim is to establish a proper methodology for an investigation of the transient stability probability and its statistical estimation based upon the fact that the parameters involved are not known, but have to be estimated. Without loss of generality, as motivated by physical and mathematical reasons in the paper, the lognormal distribution assumption is adopted for both fault clearing time (FCT) and the critical clearing time (CCT), which allows a suitable way to handle the stochastic nature of the various parameters and to assess the TSP, once that the statistical parameters (typically, the means and variances of FCT and CCT). Some considerations are effected in order to justify this assumption with respect to the various technologies for wind farm. Hence, point and interval estimations of the TSP can be derived, based upon a suitable Beta probability distribution. In the final part of the paper, a numerical application is provided in order to show the easy feasibility of the proposed methodology.