Short-Term Forecast of Vertical Grid Load with High Renewables Penetration: A Case-Study from Croatian Power System

Transmission System Operators (TSOs) have long been facing challenges related to the integration of renewable energy sources (RES) into the power system. The increasing penetration of RES within distribution networks has fundamentally altered the operational characteristics of these systems. In some areas, this has led to situations where local generation temporarily surpasses consumption, resulting in negative Vertical Grid Load (VGL) values—indicating energy is being exported from the distribution system to the transmission system. This paradigm shift, where certain network nodes are increasingly acting as generation rather than consumption points, can significantly impact the operational state of the transmission system. Given that this phenomenon is driven by RES installations within the distribution network, a systematic approach to collecting measurement data specifically from these generating units is essential for accurate VGL forecasting. To enable timely and effective operational responses, TSOs are exploring various forecasting methodologies tailored to the level and type of data available from Distribution System Operators (DSOs). This paper presents and evaluates several such models using data from a real-world energy system characterized by a high share of RES on the DSO side. Preliminary results suggest that specifically tailored forecasting models outperform the current business process.