Abstract
The first stage of the extraction process in mining is rock fragmentation. Rock fragmentation involves breaking down hard rock ore bodies by mechanical excavation, drilling, and blasting. Effective fragmentation is critical for ensuring efficiency in downstream operations such as loading and hauling. This study aimed to investigate the control and optimization of blasted rock fragmentation at A diamond mine, an open-pit operation in Mokhotlong, Lesotho, with the goal of improving the understanding of key variables affecting fragment size and their impact on overall mining efficiency and production costs.
To achieve this, five core objectives were pursued: (1) assess the degree of alignment between drill and blast designs and actual blast layouts, (2) optimize blasting parameters using empirical fragmentation modeling (Kuz-Ram), (3) validate fragmentation analysis through Split-Desktop image analysis software, (4) evaluate the impact of fragmentation on downstream processes such as excavator efficiency and loading cycle times, and (5) provide practical recommendations for fragmentation control in hard-rock open-pit mining.
An experimental research design was adopted, manipulating blast parameters as independent variables and assessing their impact on fragmentation size distribution and downstream productivity. Thirteen blast blocks were analyzed for compliance with planned drilling and charging designs. Only six blocks met the compliance threshold of 90%, indicating substantial misalignment in drill execution. Regression analysis revealed a strong relationship between drilling accuracy and fragmentation outcomes, with a coefficient of determination (R2) of 0.78 for X50 and 0.72 for X80, indicating that acceptable compliance with blast design results in finer and optimized fragmentation.
The Kuz-Ram model was used to calibrate burden and spacing, and results were validated using Split-Desktop image analysis. While X80 predictions closely matched between the two methods, X50 was consistently overestimated by the Kuz-Ram model. Error analysis indicated that the accuracy of the Kuz-Ram model differs across blocks. The alignment between empirical model predictions and Split-Desktop was strong in well-fragmented blocks, confirming the model’s reliability under stable geotechnical conditions, to optimize fragmentation. However, deviations in error metrics across poorly fragmented blocks, especially higher RMSE values, signal the
iv
influence of outliers and the need for refinement in predictive accuracy under complex ground conditions.
Optimized fragmentation significantly improved operational performance, with regression analysis showing a strong influence of mean fragment size on loading cycle time (R² = 0.82). Correlation coefficients of R = 0.93 for X50, 0.82 for X80, and 0.71 for top-size confirmed that coarser fragmentation was associated with longer loading times.
The study concluded that drilling accuracy is critical to achieving optimal rock fragmentation, with deviations leading to larger fragment sizes and reduced loading efficiency. While empirical models such as Kuz-Ram showed strong predictive alignment with actual fragmentation results, their reliability varied across blocks, particularly in heterogeneous geological conditions. The findings emphasized that fixed blast designs are risky without site-specific adjustments, as oversized fragments can increase secondary breakage and cycle times. Overall, the research highlighted the need for adaptive blasting strategies, continuous model calibration, and close monitoring of fragmentation to improve operational productivity.
Despite these findings, the study faced limitations related to geological variability and excluded variables such as groundwater presence, timing delays, and firing sequences, all of which can affect blast outcomes. The research concludes that enhancing quality control in blast execution and refining empirical models with detailed geological input can significantly improve fragmentation outcomes. Additionally, the study did not explore the interdependencies between different mining stages—drilling, blasting, loading, hauling, and crushing—which in reality function as an integrated system where inefficiencies in one area can impact overall productivity. Future studies should investigate swell factors, bucket fill efficiency, and floor elevation variance to further enhance loader performance and overall operational efficiency.