The paper “Reconciliation of wind power forecasts in spatial hierarchies” was recognised as a top-cited article in the Wiley publication!

In this paper, the authors consider an application of reconciliation to state-of-the-art hierarchical forecasts of wind power production intended for the day-ahead market. The focus is on reconciling wind power production forecasts on horizons beyond 24 h, which presents a different challenge from previous studies.

“1 | INTRODUCTION

The development of renewable energy sources is increasing all over the world. In Denmark, the majority of electricity already comes from renewable energy sources. The Danish transmission system operator (TSO) Energinet expects wind power alone to account for 59% of the electricity production in Denmark in 2022.1 In order to manage a power system where more than half of the power generation is dependent on the weather, the TSO requires accurate and reliable production forecasts. Spodniak2 showed that there is a direct link between wind power forecast errors and spreads in electricity prices in Denmark. Consequently, the development of new methods to improve power production forecasts is necessary to ensure affordable and clean energy.

An efficient implementation of the future low-carbon energy system requires electricity demand to follow the weather-driven energy production at all scales of the power system. This implies that forecasting techniques will play an important role for both market participants and system operators. The trend toward more decentralized production, which is often integrated at low-voltage levels, means that distribution system operators (DSOs) need to have more focus on forecasting renewable production, for example, to keep the voltage within appropriate limits or to ensure that the temperature of transformers do not exceed critical limits. The future calls for more coordination between the low- and high-voltage system operators, and consequently, there is a need for coherence between actions taken by TSO and DSOs, who operate at different spatial scales. ”

Read the full paper HERE.