OFFICIAL blast outcomes the root causes are examined, and actions are put in place to reduce the risks. Although the discussion here is focused on the causes of flyrock, it is worth mentioning in most of the cases the actions put in place also address other poor blasting outcomes like fragmentation and floor control. Importantly, the framework’s management component reframes the perception of flyrock: rather than seeing it as random bad luck events, flyrock are treated as measurable, manageable outcomes that can be improved. This has started to shift organizational culture towards proactive risk mitigation. Moreover, establishing a structured flyrock knowledge base opens the door for advanced analytical techniques. For instance, data mining and machine learning could be applied to the growing dataset to uncover deeper insights. While such analytics are earmarked for future work, the current management practices ensure the data collected is sufficiently rich and organized to support those efforts. 4. RESULTS AND OBSERVATIONS Results obtained so far through the implementation of the framework moved flyrock awareness away from being a blind spot for the company. As an example, Figure 11 shows maximum results extracted from 6 blasts in a specific material domain at mine site 2. The historic default exclusion zone for equipment was 300m. Estimations using the implemented prediction model indicated a significant flyrock risk (light blue bars). Equipment clearances were adjusted. The red line indicated the actual projectile distance observed during the monitoring and measuring of the blast. Figure 11. Maximum results extracted from FRED for a specific domain As part of the management component where the outcomes were discussed it was also found that for one of the blasts (grey bar) in Figure 11, data discrepancies in the captured stemming length were observed that resulted in the underestimation by the prediction model in FRED. This information was then also used to improve QA/QC activities.
RkJQdWJsaXNoZXIy MTM0Mzk2