In this paper we focus our attention on the long-term load forecasting problem, that is the prediction of energy consumption for several months ahead (up to one or more years), useful in order to ease the proper scheduling of operative conditions (such as the planning of fuel supply). While several effective techniques are available in the short-term framework, no reliable methods have been proposed for long-term predictions. For this purpose, we describe in this work a new procedure, which exploits the Empirical Mode Decomposition method to disaggregate a time series into two sets of components, respectively describing the trend and the local oscillations of the energy consumption values. These sets are then used for training Support Vector Regression models. The experimental results, obtained both on a public-domain and on an office building dataset, allow to validate the effectiveness of the proposed method.