A Hybrid Framework for Short-Term Risk Assessment of Wind-Integrated Composite Power Systems

• Published in: IEEE transactions on power systems Vol. 34; no. 3; pp. 2334 - 2344
• Main Authors: Ansari, Osama Aslam, Chung, C. Y.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.05.2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Component reliability; Computation; Computational modeling; Computer simulation; Cross entropy; Health risk assessment; Monte Carlo simulation; Optimization; power system operation; Power system reliability; Probability density function; Reliability; Reliability analysis; reliability modeling; Risk assessment; Risk management; Short term; Wind speed
• ISSN: 0885-8950; 1558-0679
• DOI: 10.1109/TPWRS.2018.2881250

This paper proposes a new framework for the short-term risk assessment of wind-integrated composite power systems via a combination of an analytical approach and a simulation technique. The proposed hybrid framework first employs the area risk method-an analytical approach, to include the detailed reliability models of different components of a power system. In this regard, a novel reliability modeling approach for wind generation for short-term risk assessment is also proposed. Thereafter, a non-sequential Monte-Carlo simulation technique is adopted to calculate the partial risks of the area risk method. As a result, the proposed framework is also capable of including the contingencies and constraints of the transmission system that are customarily neglected in the area risk method. The computational performance of the proposed framework is greatly enhanced by adopting the importance of sampling technique, whose parameters are obtained using the cross entropy optimization. Case studies performed on a modified 24-bus IEEE Reliability Test System validate that the detailed reliability modeling of wind generation and consideration of the transmission system are necessary to obtain more accurate short-term risk indices. Furthermore, the computational performance of the proposed framework is many orders higher than any other comparable methods.