Physical property analysis, numerical and scale modeling for planning of surface seismic surveys: Voisey's Bay, Labrador

The Voisey's Bay area is located on the northeast coast of Labrador and hosts one of the most important recent mineral discoveries in Canada---the Voisey's Bay Ni-Cu-Co deposit. The Voisey's Bay ore bodies consist of massive sulphides and breccias of variable sulphide content associated with a geometrically complex troctolitic intrusion hosted in gneiss of variable composition. In general, the physical properties data indicate that velocity-sensitive techniques are more likely to be effective for direct detection of ore bodies at Voisey's Bay and impedance-sensitive techniques more effective for imaging the magmatic system and structural mapping. However, the data support significant potential for impedance driven ore body detection depending upon the specific setting of the ore body.; The Voisey's Bay area offers a wide range of viable seismic targets of differing complexity that can be used to develop suitable acquisition and processing techniques for minerals exploration. A 2-D forward model was designed that incorporated both the geometry and geometric complexity of the Eastern Deeps zone in the Voisey's Bay area according to the working model for the Voisey's Bay deposit suggested by Cruden et al. (2000). By starting off with the simplest form of this model (i.e. constant velocities) and then progressing to greater and greater complexities (i.e. heterogeneity and velocity gradients) it was possible to fine-tune the processing parameters for a 2-D seismic survey at Voisey's Bay and to provide a basic template for interpretation of the reflection data.; Evaluation of the modeled data determined that there were many processing and interpretation challenges such as the absence of stratified reflectivity, complex scattering, inconsistent stacking velocities, important near-critical events and incoherent arrivals. In order to image the events in the model it was necessary to deal with both reflected and scattered events. Limited offsets of 0-2000 m were used for stacks and migrations because the long offsets did not contribute constructive information. A pre-stack Kirchhoff migration algorithm was preferred for imaging, as opposed to the more commonly used post-stack Kirchhoff migration algorithm, because it allowed more control over which velocities and events were stacked. This was necessary because of the conflicting velocities for diffractions and specular reflections that were present in this typical mineral exploration data-set.; Although 3-D seismic techniques have been demonstrated to be effective for imaging ore bodies, the cost of such surveys is often prohibitive. We are developing new approaches based on dense receiver arrays and sparse source arrays that should decrease the cost of 3-D seismic and make the technology more cost-effective for mineral exploration. Forward modeling, aimed at a first evaluation of the technique, demonstrates the potential of using sparsely illuminated seismic volumes and animated time-slicing to detect the distinctive scattering pattern associated with ore body sized targets. Time-slices of the individual sources reveal that sparse illumination effectively displays the characteristic 'bulls-eye' pattern of an individual scattering body. Stacking of the individual sources provides illumination from all sides of the scatterer and demonstrates that scattering occurs from multiple source points on the scattering body. 3-D migration of the stacked sections illustrates that the diffractions are essentially collapsed to small area located at the apex of the diffraction hyperbola as expected. However, comparison between unmigrated and migrated time-slices reveals that the unmigrated data are more effective for diffraction detection while the migrated data provide better localization of the scatterer. This approach of using few sources but many receivers for 3-D land seismic acquisition has the potential to be a cost-effective exploration and development tool.; Analysis of the extensive data-set...