Future Nordic power systems with high shares of variable renewable energy (VRE) require new approaches for characterising imbalance. In systems with high shares of weather-dependent generation, forecast uncertainty is increasingly recognised as an important driver of balancing needs, particularly in future system configurations where conventional thermal generation is expected to be progressively displaced by VRE.

This paper proposes a novel framework for analysing VRE forecast-error-driven imbalance statistics in future Nordic power-system scenarios. A large-scale long-term multi-sector energy system optimisation model is run with lead-time-aware, spatiotemporally correlated day-ahead wind forecasts across multiple weather years, so that the resulting dispatch reflects a forecast-based day-ahead schedule rather than a perfect-foresight schedule. Wind imbalances are then derived as the mismatch between scheduled and realised wind generation, thereby capturing how forecast uncertainty shapes imbalance after scheduling, network constraints and curtailment have already influenced the system state. The study examines the statistics of these imbalances across future sector-coupling scenarios, while considering the role of intraday-market updating in shaping the share of forecast-error-driven imbalance that ultimately reaches real time.

The imbalance is decomposed into two components: forecast updates from day-ahead to gate closure (GC–DA), and from gate closure to real time (RT–GC). This decomposition provides a structured basis for interpreting the role of intraday markets in reducing the imbalance that reaches real time. The resulting imbalance statistics are analysed through regional positive and negative imbalance duration curves for Nordic bidding zones.

To examine how future system structure reshapes these imbalance statistics, the analysis is carried out under three alternative sector-coupling scenarios: (i) a fully sector-coupled system including electricity, heat, hydrogen and transport, (ii) electricity, heat and transport, and (iii) electricity, heat and hydrogen. These alternatives are used as different renewable-integration pathways, leading to different scheduling conditions and consequently different wind-imbalance characteristics.

The analysis shows that both the magnitude and the asymmetry of regional wind-imbalance duration curves depend on the renewable-integration pathway. It also shows how the effectiveness of intraday-market updating influences the share of forecast-error-driven imbalance that ultimately reaches real time.