Digital Twins for energy systems must reconcile high-fidelity physics with the scalability required for rapid, large-area decision support. We propose a physics-informed Digital Twin for photovoltaic modules coupled with phase change materials (PV-PCM) and aluminum fins to enhance electricity generation [1]. The Digital Twin is built via a variational inverse methodology that transfers information from high-fidelity finite-volume method (FVM) simulations to a fast, deployable surrogate [2]. High-fidelity FVM data are used to calibrate an effective low-dimensional thermal model through inverse methods, targeting robustness across operating regimes and meteorological variability [3]. Because direct high-fidelity simulations are computationally prohibitive for territorial-scale planning due to long runtimes, the resulting Digital Twin provides a fast and accurate alternative suitable for industrial deployment. The Digital Twin is coupled with hourly atmospheric forcing to enable territorial-scale assessments and is readily extensible to climate-change-driven scenarios through modified boundary conditions and input parameters. As a representative case study, PV-PCM performance is evaluated over 1840 locations in Chile under historical summer and winter conditions. For each site, the PCM melting temperature is optimized to maximize daily electricity generation using a temperature- and irradiance-dependent efficiency model. To assess climate-change impacts without sacrificing scalability, a subset of critical locations (highest generation gains) is further analyzed under future climate scenarios. Outputs are synthesized into spatial maps of cell temperature, energy yield, relative energy gain versus a base PV reference system, and optimal melting temperature by climate region. Results show a robust linear dependence of the optimal melting temperature on mean air temperature, with wind velocity providing a secondary correction. Annual average electricity generation gains reach up to 6.5%, with the highest gains in arid zones. Avoided emissions reach up to 115.1 tCO_2e MWp^-1 year^-1, valued at 5.75 kUSD MWp^-1 year^-1.