2057 3. ENERGY USE The mining industry still relies heavily on fossil fuels for the majority of the energy consumed during the extraction and processing phases. This is due to the often remote locations without connection to the grid where mining operations are situated, preventing electrification. Although renewable energy generation on-site is tested, the adoption is slow, given the intermittent nature of the renewable energy sources, high prices of battery storage, and the need for suitable conditions (daylight or persistent winds) year-round (Issa et al. 2023). In light of the fact that the sector is estimated to account for between 1.7 to 3.5% of the global energy use (Aramendia et al. 2023; Engeco 2022), any efforts directed towards more energy-efficient extraction are worth investigating. Energy efficiency is often indirectly addressed in optimization studies, despite not being the primary objective. For example, a more selective extraction schedule may lead to a lower amount of unnecessary waste introduced into the processing circuit, in turn reducing energy demand (Levinson, Dimitrakopoulos, and Keutchayan 2023). Similarly, fleet optimization, although typically focused on cost reduction, often results in lower trucking demand, reducing the need for active material movement, fleet size, and fuel consumption (Rahnema, Amirmoeini, and Moradi Afrapoli 2023). Due to the multiple objectives of the presented studies, in this work, they were broadly categorized into energy reduction efforts related to material considerations and fleet management. 3.1 Material considerations Similar to the inclusion of geochemical variables into mine planning and optimization, geometallurgical considerations are gaining traction in the research community. Along with the shift to optimize mining complexes as a whole instead of considering the extraction stage exclusively, the possibility to optimize processing feed characteristics is more commonly investigated. Utilization of existing synergies between the two stages appears to be a promising research direction, as comminution can account for up to 80% of the energy used on an individual mine site (Abouzeid and Fuerstenau 2009). Kumar and Dimitrakopoulos (2019) introduced an adaptation of the simultaneous stochastic optimization framework developed by Goodfellow and Dimitrakopoulos (2016) that incorporates grindability as one of the optimized parameters of the objective function. Here, two hardness indices are used, namely the SAG power index (SPI) and bond work index (BWI), representing the grindability of the material in SAG and ball mills, respectively. This geometallurgical variable is then used to create a blending solution, ensuring consistent throughput of material in the milling facilities. The novel formulation, solved using simulated annealing with adaptive neighbourhood search, is then tested in a large-scale copper-gold mining complex. The geometallurgical uncertainty is assessed by sampling each of the 10 lithologies present in the deposit. In comparison with the base case, the stochastic solution resulted in reduced underutilization of the individual mills due to joint optimization of the destination policy of all underlying facilities. The optimized schedule also results in 19.3% higher NPV due to more efficient use of synergies between various components of the mining complex. Morales et al. (2019) proposed a two-step methodology based on direct block scheduling that integrates throughput rate and metallurgical recovery as uncertain variables into the stochastic mine planning framework. This is a notable addition to the stochastic optimization problem, as it addresses the performance of the downstream recovery processes directly during the extraction planning. The geometallurgical variables are simulated separately from the grade, as no clear correlation was found between these two categories, and are based on tests performed on site. Evaluation of the method in a case study performed on a coppermolybdenum mining complex resulted in a 9.4% increase in NPV with significantly lower probability of deviations from established production targets. A different approach, leveraging material behaviour is introduced by Levinson et al. (2023). Preconcentration relies on separating mined-out material before it enters the comminution circuit. This effectively leads to more selective extraction, with a lower amount of waste being unnecessarily processed, therefore lowering the energy demand of the comminution and milling stages. The researchers proposed a simultaneous stochastic optimization framework that incorporates a preconcentration facility into the
RkJQdWJsaXNoZXIy MTM0Mzk2