The availability of cost-effective energy storage technologies with durations from 10 to 100 h is key for intermittent renewable energies, like wind or solar, to become a large share of the electrical grid power. Battery prices forecasted for the upcoming years are still too expensive; and storing the energy as heat instead of electricity, arises as a promising cheaper solution. Even if there is an efficiency penalty when converting heat back to electricity, the low cost of thermal energy storage (TES) systems is an important advantage. Besides, not always the heat stored in a TES system needs to be converted to electricity, as heat corresponds to about 50% of the global energy demand. In this work, the potential of Ultra-High Temperature Latent Heat Thermal Energy Storage (UH-LHTES), which can reach energy capacity costs below 10 € well beyond 1000 ¿ /kWh by storing heat at temperatures C, is presented with the help of a Computational Fluid Dynamics (CFD) model. Modelling results, together with real materials performance and cost data, allow to estimate cost-efficiency boundaries for UH-LHTES systems for a range of sizes from 10 kWh th to 10 MWh th . The CFD model has been validated reproducing the real size and materials of a UH-LHTES prototype developed in the framework of the AMADEUS project. The CFD model is proven to be a reliable tool to reproduce the operation of a UH-LHTES system and it predicts quite precisely UH-LHTES prototype discharge rates and heat losses