Fragmentation-based assessment of blasting response in a data-scarce underground mine: A sublevel caving case study

Blast fragmentation is a critical consideration in sublevel caving, particularly during preliminary design when production-scale blasting data are unavailable. This study evaluates two complementary fragmentation-modelling approaches for the Sechahoon XII iron ore deposit. The Kuz–Ram model, using available rock, explosive, and blast-design parameters, yielded a rock factor of A = 3.709, a uniformity index of n =  2.04, and a median fragment size (x ₅₀ ) of approximately 82.7 mm. The second approach used JKSimBlast/2DRing with Swebrec fitting and produced x ₅₀ ≈202 mm. The two archived modelling pathways therefore showed an approximately 2.4-fold numerical difference in median fragment size. However, an audit of the archived inputs identified inconsistencies in Young's modulus and P-wave velocity, while JKSimBlast also represented individual fan-hole geometry. Accordingly, the observed difference cannot be attributed uniquely to model formulation and instead reflects combined model, geometric, and input-data effects. The results indicate that a single deterministic blastability value cannot be established reliably at the preliminary-design stage without field calibration. Comparison of complementary archived fragmentation predictions can nevertheless help identify uncertainty and guide subsequent blast monitoring, model calibration, and underground ring-blast optimisation.