This dissertation explores the challenge of decarbonization of the energy sector from different perspectives, applying various methods. Chapter 1 provides background and motivates the thesis. Chapter 2 assesses the relationship between geographical and temporal flexibility in a 100% renewable energy scenario across twelve central European countries. Applying a capacity expansion model and a factor separation method, it disentangles the impact of interconnection on optimal storage capacity. It can be shown that interconnection leads to a reduction of 30% in storage needs, primarily attributed to differences in wind power profiles between countries. Chapter 3 examines the integration of heating into the power sector, particularly the role of heat pumps. The power sector impacts of a substantial rollout of heat pumps in Germany by 2030 are assessed, considering buffer heat storage. The results indicate that even in scenarios with limited wind power expansion, heat pumps, accompanied by solar photovoltaics, can be deployed with limited additional costs. Importantly, heat storage proves effective in reducing the need for electricity storage and other generation capacities, while overall, a substantial reduction in natural gas consumption and CO2 emissions can be achieved. Chapter 4 expands the analysis to an international setting, studying a simultaneous rollout of heat pumps in several central European countries. Assessing the effects on electricity generation capacities, the chapter also explores the alignment of heating demand with renewable energy scarcities. Because of correlated heat demand between countries, geographical balancing does not substantially reduce the additional needed generation capacities. Confirming the results of the previous chapter, thermal energy storage capacities help reduce the need for additional generation capacities. The chapter also shows that results vary substantially between different weather years. Chapter 5 remains in the field of heating but takes an empirical perspective and studies behavioral gas savings in Germany during the 2022-23 heating season prompted by a potential gas supply shortage. Using open data and causal machine learning, significant behavioral gas savings by German households and businesses are quantified, contributing to closing the supply gap. Temperature-dependent saving dynamics are explored, emphasizing the importance of timely and accessible data for informing the public and policymakers. Continuing with the empirical perspective, Chapter 6 estimates the externalities of energy infrastructures, focusing on the potential health impacts of wind power plants. Data on German households of the Socio-Economic Panel (SOEP) is combined with geolocated data on wind power plants. Applying a staggered difference-in-difference estimation, the analysis finds no evidence of adverse health effects on nearby residents.