Lattice-Boltzmann modelling of oolitic grainstones : effects of cementation and dissolution

• Main Author: Felce, Graham Peter
• Published: University of Bristol 2011
• Subjects: 552.58
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.551300

Using the lattice-Boltzmann method, simulations of fluid flow have allowed the effects of moldic dissolution and cementation on porosity and permeability to be studied. Moldic dissolution was simulated by the removal of whole and partial spheres to generate vugs and uniform cement by increasing the radius of the spheres so they coalesce. For an uncemented domain, the evolution effects of increasing moldic dissolution showed that the permeability is initially controlled by the intergranular porosity and the moldic pores are essentially isolated. As porosity increases, with approximately 9% dissolution the moldic vugs start to form connections across the domain (percolation) which agrees with existing empirical relationships. The permeability upon percolation is dominated by the connected pore-space and controlled by the moldic porosity which agrees with other work on field and simulation studies of carbonate rock fabrics. With increasing cementation, the sphere packs show consistent decreases in permeability as porosity is occluded and correlates well to existing analytical solutions of coalesced sphere packs and empirical relationships. With the addition of moldic porosity, which leaves a cement shell intact, initial permeability is controlled solely by the intergranular porosity until percolation occurs and then connected moldic porosity dominates. It was found that connected moldic porosity and intergranular permeability can be treated as separate additive contributions an allow data sets of contrasting sphere arrangements to be correlated. Partial moldic dissolution was found to represent an earlier stage of dissolution that leads to the porosity and permeability of values seen for the whole-sphere dissolution. Modification of the cement by adding in holes reconnects the two pore systems and gives an automatic 54% permeability increase that occurs when a small connection is made for just a 1 % removal of cement. This study highlights the importance of the controlling pore type and the petrophysical property of pore-throat diameter. An additional field study of carbonate rock samples collected from Eleuthera, Bahamas allowed rocks to be reconstructed using micro-CT. Simulations of fluid flow of these samples and a quantitative image analysis of the micro-CT samples indicates that permeability can be effectively described and characterised by percentage contributions of pore-length to porosity, pore-area, a macropore shape factor and the number of data points measured in the sample.