Distinguishing recoverable from non-recoverable metal-bearing mineral phases is critical for defining recoverable resources and forecasting metallurgical performance. However, comprehensive mineralogical characterisation is often constrained by analytical cost and limited sample availability, limiting the ability to map element deportment at the scale required for mine planning, processing strategy selection and revenue estimation. To address this limitation, a Python-based element-to-mineral model (EMM) workflow was developed to predict modal mineralogy from geochemical assays, using quantitative X-ray diffraction (QXRD) mineralogy and scanning electron microscopy (SEM)-QEMSCAN-derived mineral chemistry as calibration anchors. The workflow was developed for the Black Star Open Cut (BSOC) Feasibility Study in Mount Isa, Queensland, and targets orebodies where metal deportment varies between sulfide and non-sulfide hosts and where chemical signatures overlap across mineral groups, meaning that simple stoichiometric approaches can produce non-unique or unstable mineral estimates.