• copulas;
  • multivariate analysis;
  • uncertainty analysis;
  • microgrid


This paper discusses the conjecture that Archimedean copulas—mainly Gumbel copulas—provide better stochastic models than Gaussian copulas for multivariate analysis of small wind energy generation clusters (focused on the analysis of microgrid viability). The paper provides guidance on how to model the multivariate Gumbel copula, thus allowing to follow up some recently published results that show that the correlation structure in bivariate models (generator pairing) is best defined by bivariate Gumbel copulas rather than by their Gaussian counterpart. However, it is shown in this paper that the higher the dimension (the larger the number of microgrid generators) the more probably the Gaussian copulas outperform the Gumbel copulas. Copyright © 2013 John Wiley & Sons, Ltd.