189 1. INTRODUCTION Unlike conventional heap leaching where blasted ore is moved to the crusher circuit on the surface, In-Place Recovery (IPR) uses high-intensity blasting to generate permeable broken stock within a stope. Then, lixiviant is applied to the blasted ore, dissolving target minerals as it percolates through the stope, and the pregnant leach solution (PLS) is collected for subsequent processing, as shown in Figure 1 (Ghorbani et al., 2016; Hassanvand et al., 2025a). IPR can potentially be an effective method to extract minerals such as copper from low-grade and hardrock deposits with minimal ore transportation and surface infrastructure (Dare-Bryan and Hassanvand, 2023; Estay et al., 2023).To achieve successful IPR implementation, it is important to be able to model accurately the operation process to assess project feasibility, mitigate risks, optimise designs, and forecast recovery. For the past few decades, substantial modelling works have been done to predict copper recovery for heap leaching (Ferrier et al., 2016; Marsden and Botz, 2017). In these works, column test results are often fitted with empirical or semi-empirical models based on the shrinking core concept to capture the total copper recovery change with time. The fitted models are then used to predict recovery for the actual heap leaching operations with the same model parameters. This approach can be quite effective for heap leaching, as the column tests are designed to represent the actual operation conditions, such as with the same ore particle size distribution (PSD), irrigation rate, and feed lixiviant concentration. Moreover, the heaps are often engineered to have uniform and homogenous material properties through ore crushing, grinding, and often agglomeration, which further warrants the direct upscaling of the column test results. For IPR, however, the blasted stope contains a much wider range of ore sizes (Wang et al., 2025). This requires a lot more column tests with larger column sizes to generate representative results for model fitting, which can take significantly more time with a higher cost (Hassanvand et al., 2025b). Figure 1 – Illustration of In-Place Recovery (IPR) with (a) blasted stopes, (b) lixiviant
RkJQdWJsaXNoZXIy MTM0Mzk2