ExplainabilityTransparencyAccountabilitycs.AI
Analysing drivers and interdependencies in European electricity markets using XAI
Antoine Pesenti, Aidan O'Sullivan
June 17, 2026
arXiv.org
Key Takeaways
-
Renewable energy has outsized influence on electricity prices: Solar and other renewable sources disproportionately drive price formation across European markets despite representing a smaller share of total power generation, suggesting that even modest changes in renewable output can significantly impact prices.
-
Gas prices remain the most consistent market driver: Across all 39 European bidding zones analyzed, gas prices emerged as a dominant and stable determinant of electricity prices, indicating that fossil fuel markets continue to be a critical factor in price formation even as renewable capacity grows.
-
Regional electricity markets are deeply interconnected: The analysis reveals that interconnections between regions significantly shape price dynamics, demonstrating that European electricity systems are highly interdependent—meaning price changes in one area ripple across multiple markets.
-
XAI techniques can unlock insights from complex AI models: By applying explainable AI methods (SHAP) to deep neural networks, researchers were able to identify specific drivers of electricity prices that would otherwise remain hidden in a "black box" model, showing how XAI can make powerful AI systems more useful for policy and business decisions.
-
A fully integrated EU electricity market would function differently: The paper's counterfactual analysis of a single EU-wide market suggests that current fragmented regional markets produce different price dynamics than a hypothetical fully integrated system, with implications for future market design and policy.