Factors Influencing the Expert System Profitability

"The technology and equipment design of the Twenty-first Century permit us to exploit low-grade ore bodies with milled tonnages reaching 300,000 tpd. Because of their substantial operational costs (energy, wear parts, etc) and precarious economic situations, the operations of today must maximize their metallurgical performance.The implementation of an Expert System becomes a logical avenue to pursue in view of increasing plant performance.Many operations, following the implementation of an Expert System, meet the objectives they determined before its integration; such as increased tonnage, stability, decreased costs, etc. However, they often cannot develop the full potential of the system due to certain factors that have been neglected or omitted during the implementation of the system.Three main elements have been identified during the implementation phase at INMET Mining Troilus Concentrator: 1- Instrumentation & Control 2- Software 3- Integration. Even though they have been clearly identified, these elements are often neglected due to lack of resources or lack of comprehension on the part of the operations personnel. This document will demonstrate, based upon our experience at Troilus and that of Metso Minerals, that we can identify and handle many factors which influence the profitability of the Expert System, for both short and long-term process control improvements.INTRODUCTIONThe Inmet Troilus Mine is an open pit mine, situated 175 km north of Chibougamau in the province of Québec, Canada. The mill, which started production in November of 1996, was originally designed to treat a gold – copper ore at a rate of 10,000 tpd.In early 1998, the mine plan called for a pit pushback and decreasing head grades for the following two years. With the price of gold nearing decade lows, the Troilus project was facing a major economical setback. There was a sense of urgency to reduce operating costs below the market price of gold. The mill team presented a project to management calling for an increase in mill throughput while minimizing capital expenditures and operating costs. This idea was supported by the existing mine plan, which called for a mine cut-off of 0.7 g/t Au and stockpiling low grade 0.5 g/t material for future milling."