Probabilistic load flow with correlated wind power injections

Julio Usaola

DOI: https://doi.org/10.1016/j.epsr.2009.10.023

IF: 3.818

2010-05-01

Electric Power Systems Research

Abstract:The non-dispatchable nature of wind generation implies that system operation depends on wind power prediction programs that forecast wind farms production with high levels of uncertainty. This means that probabilistic power analysis tools become more and more necessary in systems with high wind penetration. Probabilistic load flow becomes especially difficult when wind generation is considered. The high uncertainty of the production, the non-Gaussian probability density function (PDF) and the clear dependence among the wind farms poses a challenge for conventional tools. The paper proposes an approximation that makes use of the properties of statistical moments and Cornish–Fisher expansion to tackle these new problems.

engineering, electrical & electronic