Medium-term electricity load forecasting and climate change in arid cities

Electric utilities need to consider how potential changes in climate patterns will affect their peak loads. This study incorporates weather and socio-economic variables into a medium-term load forecasting model to consider potential climate change effects on the challenging summer peak season for utilities in the arid southwestern US. Our 'average hourly load by month' model shows marked improvement over a purely autoregressive approach to load forecasting used by some electric utilities. In light of climate change, electric utilities and society can benefit from minimizing inaccuracies in load predictions. Decision-making based on more climate-sensitive forecasts will reduce the water and carbon footprint of electric utilities and improve their investment strategies for renewable energy technologies.