Five essays in energy system analysis

To mitigate climate change, fossil fuels have to be replaced with renewable energy sources. Since the potential of firm renewable energy is limited in most countries, future energy systems will heavily rely on variable wind and solar power, exposing these systems to weather variability. Balancing demand with variable supply requires temporal and spatial system flexibility, including different types of electricity storage and geographical balancing via transmission. The first part of this thesis focuses on understanding the characteristics and power system implications of extreme periods of low wind and solar availability, also referred to as “variable renewable energy droughts” or the German term “Dunkelflaute”. It provides a comprehensive overview of how to define and quantify such drought periods and characterizes historical wind and solar droughts in Europe. Results reveal that extreme drought events comprise multiple contiguous drought events of varying severity that occur within long-lasting periods of renewable availability well below average. Using a European power sector model, the analysis further investigates the long-duration storage needs required to cope with extreme droughts. To facilitate the rapid expansion of renewable energy sources, governments often commit to binding renewable shares in gross electricity consumption or generation for specific target years. The second part of this thesis explores unintended effects in cost-minimizing energy modeling related to such targets, particularly the phenomenon of “Unintended Storage Cycling”. This modeling artifact can be detected in the case of simultaneous storage charging and discharging, which leads to flawed model outcomes. The analysis examines the mechanisms, drivers, and modeling implications of unintended storage cycling, and discusses various options for removing the artifact. Additionally, the analysis is extended to model settings without binding renewable targets, where unspecified techno-economic trade-offs that such models optimize may cause unintended storage cycling. The insights from the first part of this thesis provide policymakers and system planners with a deeper understanding of the characteristics of variable renewable energy droughts and the need for system flexibility for coping with such extreme periods. The second part advances energy system modeling by identifying and addressing unintended effects, thereby enhancing the reliability of model-based policy recommendations used to guide the transition to renewable energy systems.