variavel0=<b>Luís O. Emerich dos Santos</b> - emerich@lmpt.ufsc.br Universidade Federal de Santa Catarina
<b>Paulo César Philippi</b> - philippi@lmpt.ufsc.br Universidade Federal de Santa Catarina
<b>Celso Peres Fernandes</b> - celso@lmpt.ufsc.br Universidade Federal de Santa Catarina
<b>Henrique Cesar de Gaspari</b> - henrique@lmpt.ufsc.br Universidade Federal de Santa Catarina
<b>Abstract.</b> Macroscopical physical proprieties of materials are decurrent of their microstructural formation. In this manner, the permeability of an oil reservoir rock depends upon the size distribution and spatial organization of porous phase. To understand this influence and determine macroscopical proprieties, physical processes have been simulated in 3-D models of structures. The data needed to generate these models are frequently acquired by mercury intrusion, adsorption isotherms, and more recently by means of image analysis techniques applied on 2-D sections of the materials. Other techniques, like microtomography and serial sectioning, have been used permitting to obtain directly the 3-D structure. Nevertheless, these techniques are more expensive than the formers and, saying only about microtomography, this technique does not permit high spatial resolution. This article is concerning the 3-D modelization, commonly called 3-D reconstruction, of porous microstructures that conserves (or tries to conserve) geometrical proprieties measured on 2-D images of materials. The truncated Gaussian method have been widely used with this purpose, nevertheless, it does not seem to be suitable to preserve the geometry of strongly spatially correlated structures. The superposed spheres method presented in this article intend to supply this lacuna with the creation of 3-D models in which are imposed the conservation of the grain size distribution measured in 2-D images of the materials. Although it imposes the conservation of the grain distribution (modeled like spheres) the aim is the conservation of the porous size distribution and spatial organization of the porous phase, where the fluid invasion processes occur. The method is described and applied to 3-D reconstruction of various rocks. The geometry of the 3-D reconstructed medium is compared with the geometry of the 2-D images aiming the validation of the proposed method and also their applicability and limitations.
<b>Keywords.</b> oil reservoir rock, microestructure, 3-D reconstruction.