Implementing Stochastic Response Surface Method and Copula in the Presence of Data-Driven PV Source Models

As PV penetration in the power distribution network is growing on an unprecedented scale, probabilistic power flow analysis is becoming crucial to assess the health of network operation. State-of-the-art probabilistic analysis relies on Stochastic Response Surface Method and Gaussian Copula for including PV sources correlation. However, it has been observed how in the presence of complex statistical distributions of injected PV powers, such a standard approach can provide inaccurate evaluations of the output variable distributions. In this article, we carefully investigate the origins of such a drawback and propose a novel implementation flow that overcomes the problem. In order to check the correctness of the proposed methodology, the results obtained with the Stochastic Response Surface Method are compared with Monte Carlo simulations. Several investigations, such as voltage uncertainties, network health and probability of violating quality constraints are conducted on two test networks.