AI ENABLED OPERATOR ASSISTANT TO ENHANCE COPPER THROUGHPUT AND REDUCE ENERGY CONSUMPTION IN SOLVENT EXTRACTION AND ELECTROWINNING *Venkateswara Rao Kottana1, Dinesh Gondhi2 , Satish Kumar Lekkala1, Juan Manuel Pardal Arnejo2, Bhaskar Sinha1, Srisuhasini Gottumukkala1, Padmaja Bodanapu1 (1) Schneider Electric Systems India Pvt. Limited, India | (2) Schneider Electric Systems USA Inc, USA *Presenting author: venkateswararao.kottana@se.com, dinesh.gondhi@se.com ABSTRACT Copper Solvent Extraction/Electro Winning plants increasingly face operational challenges due to declining ore grades, ageing infrastructure, and poor process control. Elevated iron levels in the pregnant leach solution (PLS) disrupt copper extraction efficiency (~80 – 85%), contaminate the organic phase, and lowers current efficiency (~60 - 75%), leading to sub optimal throughput (reduction by 5-10%), higher energy consumption and increase in lower grade copper cathodes (2- 6%). Traditional solvent flow adjustments depends on lab analysis of metal composition in the PLS solution, which delays the corrective actions. Without real-time visibility and insights, operators cannot respond promptly to process deviations. During this lag, fluctuations in PLS composition degrade SX performance and cause inconsistent EW outcomes like reduced metal recovery and operational efficiency. Fragmented integration between plant instrumentation and analytics, due to incompatible communication interfaces hinders real-time data use. Industry needs an intelligent solution which enables real-time process visibility, deeper insights and recommendations on real time adjustment of the process variables (like extraction flow adjustment rates) to improve the copper loading kinetics into the organic phase, reduces the iron & impurities co-loading and optimizing the current efficiency. This approach significantly improves extraction efficiency, reduces operational costs, and supports energy-optimization, sustainable copper production. This paper proposes an AI enabled operator assistant solution that combines First principle-based models with Machine learning-algorithms, enabling real-time process visibility, Insights, recommendations and control at the Edge layer. This solution captures live plant data— including sensor inputs, lab samples, historical trends—and runs what-if scenarios locally to deliver insights, actionable recommendations and performance alerts to the operator in real-time. It leverages Open Automation Platform to interface with diverse industrial protocols, ensuring seamless connectivity and edge-level computing. With this AI enabled operator assistant solution, operator gains process visibility, insights, and gets recommendations tailored to ore conditions and Fe separation. It dynamically adapts to ore grade variability, improving Fe removal in SX and potential to optimize Copper Extraction efficiency and current efficiency in EW leading to improvement in Copper Throughput by 5-10% and reduction in Chemicals consumptions such as Extractants and Acid (H2SO4), and reducing the percentage of lower grade copper cathodes.
RkJQdWJsaXNoZXIy MTM0Mzk2