The aim of all modelling is to replicate geological reality as closely as
possible using available data. This paper is aimed at one end of the
spectrum, where very limited data is combined with intuitive geological
ideas to synthesise models for either target definition or for determining
minimum viable mining scenarios. Interactive 3D modelling allows a satisfactory hypothetical geometry
of surfaces to be assembled, and provides a fast quantitative method for
generating targets. As data flows from exploration, an evolving
geological model assists with the progressive evaluation of a prospect. Two case histories are described where this approach has been used
successfully, the Mosquito Hill Prospect in Queensland and the
Longwood Ranges Platinum Prospect in New Zealand. The possibility is
also discussed of hypothetically modelling an orebody which has yet to be
discovered. The proper description of an orebody is the foundation upon which all
mine design and planning is undertaken. The volumetric accuracy of the
model is crucial in valuing an orebody. Geologists must form a geometry
which matches the conceptual view of the ore occurrence and ensure that
they understand how the computer system used calculates volumes. If the
underlying assumptions inherent in the method conflict in any way with
the conceptual model visualised for the ore occurrence, then that method
must not be used.