The Golden Gift is a recently discovered deep deposit below the main operating areas of the Stawell Gold Mine. At an early stage of the project, diamond drilling was used to define the mineralisation on sections approximately 100 - 150 m apart. The project economics were sensitive to both grade and tonnage. There were consequently two significant risks on resource quantification: risk on tonnage (geological risk'); and risk on grade (grade risk'). The degrees of freedom inherent in the geological model were considered to be potentially high because of wide drill spacing and likely structural complexity. To quantify geological risk three geologists familiar with Stawell mineralisation independently interpreted the deposit. This generated three plausible, different geological volume models. Conditional simulation was then used to quantify grade risk. Conditional simulation requires the user to define input statistics (histogram and variogram model) and a geological envelope. Based on global mean grade of an ordered set of simulations, pessimistic', median' and optimistic' simulation cases were defined._x000D_
The result was a 3 x 3 risk matrix with geological risk (the three interpretations) on one axis and grade risk (also three cases) on the other axis. An interesting result was that geological and grade risks were of a similar order of magnitude. Therefore a risk analysis of grade by geostatistical methods within a fixed geological model is of limited value, especially if the drill spacing is wide. After completion of this study, with additional drilling, the geological degrees of freedom at Golden Gift reduced substantially. A second Golden Gift risk study used a refined geological model to build a set of simulations to be used in preliminary mine design._x000D_
A third simulation study in the deeper parts of the Magdala ore deposit, a well-drilled part of the mine with production history, demonstrates that the geostatistical simulation method used at Golden Gift is a robust and useful tool to quantify grade risk.