440 3. MATERIALS AND METHODS 3.1. Dataset Description 3.1.1. Description of Variables Considered in the ODJIT System The variables considered within the ODJIT system for cluster construction correspond to lithological and alteration classifications that describe the geological composition of the material delivered to the processing plant. These variables reflect the dominant alteration assemblages and host rock types associated with the Toromocho deposit. The following categories were included in the clustering framework: Lithological and Alteration Variables: ● INTR|USIVO A: Characterized by potassic alteration, predominantly represented by biotite and/or potassium feldspar assemblages. ● INTRUSIVO B: Associated with propylitic, phyllic, and argillic alteration, including chloritic alteration, sericite, clay minerals, and zones of silicification. ● SKARN SERPENTINE–MAGNETITE: Comprising serpentinite–magnetite skarn and magnetiterich skarn assemblages. ● SKARN ACTINOLITE: Represented by tremolite–actinolite skarn with associated chlorite alteration. ● HORNFELS: Including green hornfels (diopside hornfels), biotite hornfels, sandstones, calcareous sediments (marble), and shales. Geochemical Variables: Fe, MgO: Major element grades representative of mineral type and alteration intensity, providing chemical indicators of ore composition and metallurgical behavior. Ore Competence Parameters: BI, MIC, HARD, DWI, BOND: Parameters describing ore resistance to fragmentation, including hardness, toughness, and abrasiveness indices. These variables directly control comminution efficiency and influence ore response during primary and secondary grinding stages. The analysis is conducted using data aggregated at a 15-minute granularity over the period from January 1, 2023 to August 26, 2025. 3.1.2. Residence Time Considered for Time-Series Transformation The residence time of the ore material is defined from its extraction at the mining face to its discharge at the Primary Crusher. These time intervals (start and end timestamps) are recorded through the MineStar system, which provides detailed information on shovel loading time, truck hauling time, and unloading time at the Primary Crusher. For the estimation of residence time within the processing plant, the analysis begins at the Primary Crushing stage and extends through the transport pathway from the Primary Crusher to the Stockpile, and subsequently from the Stockpile to the SAG mill. Residence time calculations were performed based on design layouts and mechanical drawings, including distance measurements and elevation profiles, to ensure accurate modeling of material transport across each stage and between unit operations. (Albarracin, 2021) This approach enables proper alignment of the material timestamp with its effective arrival at the SAG milling stage (Phase I), allowing the transformation of operational data into consistent time-series representations associated with the actual ore processed.
RkJQdWJsaXNoZXIy MTM0Mzk2