Track 1: AI and Data-Driven Decision Making

enhance the operational management of copper SX/EW circuits. The digital twin provided physically grounded and plant‑realistic predictions across a wide range of conditions, including variations in PLS acidity, Fe loading, extractant strength, and hydrodynamic behavior. Coupled with an edge‑based Operator Assistant, the system enabled real‑time analysis, scenario evaluation, and generation of constraint‑compliant operating recommendations. Evaluation across historical and simulated datasets showed that the hybrid model consistently outperformed standalone first principle ‑based and purely data‑driven approaches, particularly in forecasting copper recovery, SX efficiency, and Fe transfer to the rich electrolyte. The optimizer’s recommended pH and O/A setpoints closely tracked plant-applied values and provided smoother operating trajectories that aligned with hydrodynamic stability requirements. The optimized conditions yielded measurable metallurgical improvements including 2-4 % gains in copper throughput, reduced Fe co‑extraction, and improved current efficiency in EW because of reduction in Fe co-extraction. These improvements were achieved while maintaining chemistry and equipment constraints, demonstrating the practical deployability of the system in industrial environments. The work validates the feasibility of deploying first principle ‑guided hybrid models at the edge to support real‑time decision making in hydrometallurgical circuits. Future extensions include coupling the SX digital twin with upstream leaching dynamics, incorporating rate‑based mass‑transfer and spatial hydrodynamic models, and linking downstream predictions to electrowinning current‑efficiency behavior. Broader multi‑site deployment and continuous online calibration using incoming assay streams will further strengthen robustness and generalizability. Taken together, this framework provides a scalable path toward more stable, adaptive, and energy‑efficient copper production under increasingly variable ore and process conditions. REFERENCES [1] Ritcey, G. M., & Ashbrook, A. W. (1984). Solvent Extraction: Principles and Applications to Process Metallurgy (Vol. 1 & 2). Elsevier. [2] Sole, K. C., & Hiskey, J. B. (1995). Solvent extraction of copper by hydroxy oximes. Hydrometallurgy, 37(2), 129–147. [3] Flett, D. S. (2004). Solvent extraction in hydrometallurgy: The role of organophosphorus extractants. Journal of Inorganic Chemistry, 96(1), 223–234. [4] Habashi, F. (1999). Handbook of Extractive Metallurgy. Wiley-VCH. [5] IUPAC (1997). Compendium of Chemical Terminology: The Gold Book (2nd ed.). [6] Atkins, P., & de Paula, J. (2018). Physical Chemistry (11th ed.). Oxford University Press. [7] Sole, K. C. (2002). Acid balance in copper solvent extraction circuits. International Solvent Extraction Conference (ISEC), Cape Town. [8] Levenspiel, O. (1999). Chemical Reaction Engineering (3rd ed.). Wiley.

RkJQdWJsaXNoZXIy MTM0Mzk2