Improving accuracy of energy forecasting through the presence of an electric vehicle fleet

Abstract

Accurate forecasting of energy is of pivotal importance for the Smart Grid vision. Today it is achieved through aggregation of numerous individual energy loads, and hence stochastic behaviour costs are distributed to the entire aggregated stakeholders. However, if one was able to accurately control the internal energy behaviour in order to meet an accurate self-forecast (as seen from external stakeholder's point of view), new business opportunities in the Smart Grid era could be considered. Static storage systems may be used to absorb forecast errors, however these are still costly, and therefore alternatives are sought. Stakeholders, however, can benefit from a plethora of alternative “storages” offered by properly utilizing their assets. We approach the absorption of forecast errors by utilizing a fleet of electric vehicles whose on-premise presence is used to compose a variable energy storage. An empirical assessment with real-world data is provided and results demonstrate the significance of electric vehicles towards helping stakeholders achieve and maintain the accuracy of their self-forecast.