Thermal energy storage (TES) plays a pivotal role in meeting growing heating demands while integrating increasing shares of variable renewable energy sources. However, challenges in optimizing system design, performance, and cost-effectiveness remain critical to their widespread adoption. This thesis develops, analyzes, and optimizes macro-encapsulation-based hybrid TES solutions, combining sensible heat storage in water with latent heat storage (using phase change materials), leveraging numerical simulations, experimental validation, and machine-learning tools to balance energy density, heat transfer performance, and efficient operation. Chapter 1 positions TES as essential for the energy transition, highlighting limitations of purely sensible or latent configurations and emphasizing the merits of hybrid TES. This motivation frames the objectives and introduces the in-depth studies on capsule geometry and system integration presented in the chapters that follow. Chapter 2 takes a step back and explores the packed bed system independent of the TES application, offering a detailed investigation of particle shape for packed-bed configurations, encompassing both spherical and non-spherical geometries. Using rigid-body simulations and computed tomography (CT) scans, the study validates how particle arrangements and border effects affect packing density (PD), effective surface-to-volume ratio SVeff and other key performance indicators. The results provide general trends that highlight the importance of particle shape and border effects in applications requiring both high effective surface and compactness. Chapter 3 builds on earlier fundamental insights by developing and validating a system-level model that holistically integrates sensible-latent TES, PV arrays, and heat pumps. A multi-objective optimization routine investigates how design and operating parameters affect cost, global warming potential (GWP), and self-sufficiency. Targeting 70-80% thermal self-sufficiency balances moderate costs (0.27-0.35 CHF/kWh) with large GWP reductions (over 70% relative to fossil heating). However, pushing self-sufficiency above 85% leads to steep increases in both storage volume and system expense, delivering diminishing sustainability returns. Chapter 4 explores how operational and design parameters interact to shape overall system performance. Multiple control strategies, from off-peak charging to dynamic solar-threshold approaches, are tested and show that real-time solar coverage thresholds substantially increase PV self-consumption and self-sufficiency. Particle shape is re-examined within a complete system context, showing moderate effects on charging rates but relatively small annual gains. In contrast, the role of PCM in the low-temperature (SH-TES) tank grows more significant with larger volumes, enabling more effective use of midday PV surpluses. Further sizing analyses confirm that aligning tank volumes and solar capacity is critical for occupant comfort and system efficiency. By integrating capsule geometry analysis, system-level optimization, and control strategy considerations, this thesis contributes practical recommendations for deploying high-performance hybrid TES systems. While the present work focuses on thermal energy storage, the underlying particle-scale modeling and shape parametric methods can be extended to a broader range of heat and mass transfer applications where balancing compactness and surface availability are key.