At a large Cu-Au deposit in the Philippines, mineralised porphyries occur below an erosional surface buried by post-mineralisation volcanic and lake-bed deposits (cover sequence); ore lies below the river stage in hilly terrane where rainfall totals 4.8 meters annually. Intervolcanic muds in the cover sequence, including a thick basal unit, caused the loss of numerous drillholes during exploration. Consequently, hydrologic drillholes were hastily “telescoped” to as deep as 700m into the basement, leaving few opportunities for hydraulic tests or monitoring wells in the upper units. Groundwater flow in the confined basement aquifer was deemed more crucial to dewatering studies than flow through overlying volcanics, based on a block-cave mine plan.
Three basement pumping tests (10L/s to 136L/s) were insufficient to define boundaries or resolve vertical recharge rates. As a result, early dewatering predictions for a laterally-unbounded, high transmissivity basement were very high. However, dewatering a basement decline, which affected water levels throughout the hydrologic program, was increased late in the program to 200-300L/s, thus providing an opportunity to passively monitor a large stress of the aquifer for a long period (10-months).
Boundary effects were eventually seen, which reduced numerical predictions of block-cave dewatering by two-thirds. Partial recovery in the decline allowed an estimate of groundwater storativity, and net recharge through leaky lateral and vertical boundaries.
An iterative approach was used to evaluate the robustness of model calibrations and to validate alternative conceptual models. In total, the groundwater model was reasonably well recalibrated five times to account for new data. This required revising the conceptual model with regard to transmissivity, vertical hydraulic conductivity, groundwater storage, recharge, and boundary conditions—in the basement only. Throughout the process, the cover sequence was assigned generic properties, to which model predictions of dewatering the block cave were largely insensitive.
Post pre-feasibility, the mine plan was changed to an open pit design. The existing data and block-cave model accurately defined inflows from basement rocks, but those flows now constituted less than a third of potential pit inflows, and would occur later in the mine development. The poorly-constrained volcanic sequence, however, now transmitted another third (direct precipitation the remainder) of potential mine inflow, immediately upon development, and posed significant slope-stability risks. Major data gaps relative to open-pit mining, included:
- Intermediate circulation of meteoric recharge through the cover aquifer
- Water levels and gradients within the volcanic sequence
- K values in pit-wall material, critical to both slope stability and inflows
- Layering and vertical K in volcanics, critical to dewatering strategies
An entirely new phase of field and numerical investigation was required to assess these gaps and the feasibility of the open pit mine.
