Reliability of Interconnected Power Systems With Correlated Demands

Abstract

The rationale for system interconnections due to the associated reliability improvement and reserve benefits is well understood. In calculations to obtain the loss of load probability (LOLP) of interconnected systems, the correlation between system loads is generally ignored. This paper proposes a method for the calculation of the LOLP of two interconnected systems incorporating the correlation between demands. This has been done by using the bivariate Gram-Charlier expansion. The LOLPs for different tie-line capacities are evaluated by using the hourly loads of the two interconnected systems. The reported work is an extension of the work of Baleriaux and Booth and the notion of equivalent load of two areas with correlated loads. The proposed method is computationally efficient since convolutions are performed by using the appropriate statistical moments and cumulants. The results of a study of two interconnected systems using the proposed method has been outlined. The accuracy of the bivariate Gram-Charlier's expansion for this class of problems has also been discussed. Although the method has been applied to two interconnected systems, there is no conceptual difficulty in extending it to three or more areas by using a multivariate Gram-Charlier expansion.