Modelling the non-linear response of Spanish electricity demand to temperature variations

The demand for electricity is a key variable because its links to economic activity and development; however, the electricity consumption also depends on other non-economic variables, notably the weather. The aim of this study is to analyse the effect of temperatures on the variability of the Spanish daily electricity demand, and especially to characterise the non-linearity of the response of demand to variations in temperature. In this article, we explore the ability of Smooth Transition (STR), Threshold Regression (TR), and Switching Regressions (SR) models, to handle both aspects. As we conclude, the use of LSTR approach offers two main advantages. First, it captures adequately the smooth response of electricity demand to temperature variations in intermediate ranges of temperatures. Second, it provides a method to analyse the validity of temperature thresholds used to build the "cooling degree days" (CDD) and "heating degree days" (HDD) variables traditionally employed in the literature. (c) 2005 Elsevier B.V All rights reserved.