Lithium-ion batteries (LiBs) are currently the dominant energy storage system across various scales: from portable electronic devices to urban and aerospace vehicles, and they will likely play a major role in the construction of net-zero energy buildings. Current mathematical models of battery charge and discharge often simplify electrochemical behavior due to the complexity and computational expense of electrode morphology characterization. This is especially true on the anode side, where graphite particles have a flake shape that is difficult to model accurately. This variability in particle shape has a significant impact on the electrochemical process, making accurate in-silico reconstruction of the microscopic electrode geometry fundamental for understanding charge and discharge processes at the macroscopic level. To address this, we focused on geometrical characterization of the electrode in a graphitic half-cell. We represented graphite flakes as ellipsoids, using a Python-based script to generate realistic particle geometries based on experimental particle size distributions (PSD). This approach ensured accurate electrode representation while maintaining computational efficiency. These geometries were used as the basis for detailed, but computationally expensive, COMSOL simulations of the electrode. Next, we applied asymptotic homogenization to upscale the microscopic behavior, significantly speeding up simulations while retaining accuracy. This faster model allows us to explore different electrode morphologies and real-size distributions. It serves as a digital twin for rapid, resource-efficient simulations and will provide training data for a neural network model. This work has received funding from the EU's Horizon Europe research and innovation programme underGA no. 101137725 (BatCAT)