Abstract
The recovery of residual and previously unmined ore reserves near caving and active-displacement zones is complicated by increased ore loss, dilution, development requirements, support demand, and economic uncertainty. This study presents an exploratory economic–mathematical framework for scenario-based assessment and local ranking of mining alternatives at the Annenskaya Mine of the East Zhezkazgan deposit. The analysis uses 54 modelled scenarios representing distinct combinations of mine-specific geological, technological, and cost parameters rather than independently observed production blocks. The framework integrates material-and-value balance, a quadratic penalty-based scenario model, and a standardised second-order polynomial fitted using Lasso regularisation. The variables include orebody dip, thickness as a stratification factor, ore loss, dilution, displacement-zone conditions, and principal development, extraction, and support costs. Five-fold cross-validation selected λ = 10. The model achieved R2 = 0.41 on the training set, whereas cross-validation yielded R2cv = −0.17, RMSE = USD 553,800, and MAE = USD 461,200. These results restrict the model to exploratory sensitivity analysis and local comparison. The positive-profit domain was widest at dip angles of 30–45° and narrowed at 50–51°. Hessian eigenvalues confirmed a saddle-shaped response surface without an unconstrained global maximum. Broader application requires recalibration, stronger baselines, continuous geomechanical indicators, and independent operational verification.