HR: 0800h
AN: MR11A-0915    [Abstracts]
TI: Relation between Transport Properties and Heterogeneities from Grain and Sample Scale in Fontainebleau Sandstone
AU: * Song, I
EM: Insun.Song@ruhr-uni-bochum.de
AF: Inst for Geophysics, Ruhr-Univ Bochum, Bochum, 44780 Germany
AU: Graphchikov, A
EM: agraph@iem.ac.ru
AF: IEM, Russian Academy of Science, Moscow, 142432 Russian Federation
AU: Renner, J
EM: renner@geophysik.ruhr-uni-bochum.de
AF: Inst for Geophysics, Ruhr-Univ Bochum, Bochum, 44780 Germany
AB: We carried out an experimental study of Fontainebleau sandstone for correlation between pore structure and transport properties. Eight blocks of the sandstone were available from different locations. Several cored samples, 30 mm in diameter and 60 mm in length, were prepared from each block. The available blocks yield a range in porosity from about 3 to 11% with no difference between connected and total porosity. Scanning Electronic Microscope analysis revealed a clayless-pure ($>$99$%$ quartz), well sorted sandstone with different porosity and pore geometry for different blocks. The pore texture varies from ­risolated pores­_ to ­risolated grains­_ in two-dimensional images. To further constrain the effect of heterogeneities in sample scale on bulk fluid transport properties we tested samples with up to six sawcuts parallel or perpendicular to the flow direction. The planar faces of the saw cuts were machined to a smooth finish. We also stacked pieces alternating between two sandstone varieties of 6 and 9$%$ porosity to model inhomogeneity for one-dimensional fluid transport varying in scale. On intact sample, we measured ultrasonic velocity, electrical conductivity, hydraulic permeability and specific storage. Ultrasonic velocity and electrical conductivity were measured on as-is, oven-dry, and water-saturated samples at room temperature and atmospheric pressure. Permeability and specific storage were determined from steady state, linear injection and oscillatory tests at different effective pressures ranging up to 200 MPa at room temperature. The variation of the ultrasonic velocity covers the full range between the lower and the upper Hashin-Shtrickman bound. The velocity ranges relate to the pore geometry rather than porosity; e.g. high velocity in samples with ­risolated pores­_ and low velocity in samples with ­risolated grains­_. The electric impedance and hydraulic permeability of the sandstone correlates strongly with the connected porosity. The permeability depends negatively on effective stress only for lower porosity samples ($<$5$%$. Elastic moduli and transport properties of our porous sandstone samples are not only sensitive to the volume fraction of pore but also to the grain-scale heterogeneity of the pore space, i.e., the presence of pore-like and crack-like conduits. Cuts perpendicular to the main flow direction constitute a barrier in high porosity samples because the fluid channels are not perfectly matched between the two polished surfaces. Cuts parallel to the main flow direction perturb the flow field yielding increased scatter in permeability. The apparent specific storage and permeability of inhomogeneous samples composed of six discs alternating between the two sandstone varieties show a positive linear relationship with pore pressure but little variation with oscillation period. In contrast, the hydraulic parameters of samples assembled of two pieces, one of low and the other of high porosity, depend on oscillation period in a similar way as the homogeneous low-porosity sample, i.e., both hydraulic parameters increase with a decreasing rate with increasing period, and asymptotically approach a maximum value. The variations in specific storage and permeability result in a decrease of hydraulic diffusivity with increasing oscillation period suggesting that pore pressure diffusion becomes less efficient owing to increasing storage in ­rdead-end­_ parts of the pore network. The low porosity sample is closer to the percolation threshold and the percentage of the network constituting dead storage is relatively larger than in the more porous sample.
DE: 5114 Permeability and porosity
DE: 5139 Transport properties
SC: Mineral and Rock Physics [MR]
MN: 2004 AGU Fall Meeting