3-D Pore-Scale Modelling of Sandstones and Flow Simulations in the Pore Networks
- Stig Bakke (Statoil Research Centre) | Pål-Eric Øren (Statoil Research Centre)
- Document ID
- Society of Petroleum Engineers
- SPE Journal
- Publication Date
- June 1997
- Document Type
- Journal Paper
- 136 - 149
- 1997. Society of Petroleum Engineers
- 5.4.2 Gas Injection Methods, 5.3.4 Integration of geomechanics in models, 5.3.2 Multiphase Flow, 1.2.3 Rock properties, 5.5.2 Core Analysis, 2.4.3 Sand/Solids Control, 1.14 Casing and Cementing, 4.3.4 Scale, 5.3.1 Flow in Porous Media, 5.1 Reservoir Characterisation
- 9 in the last 30 days
- 1,561 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 10.00|
|SPE Non-Member Price:||USD 30.00|
A new method for generating realistic homogenous and heterogeneous 3-Dpore-scale sandstone models is presented. The essence of our method is to buildsandstone models which are analogs of actual sandstones by numerically modelingthe results of the main sandstone-forming geological processes - sandgrainsedimentation, compaction, and diagenesis. The input data for the modeling areobtained from image analyses of thin section images of the actualsandstone.
The spatial continuity of the sandstone model in the X, Y, and Z directionsis determined using a scale-independent invasion percolation based algorithm.The resulting spatial continuity function, which is an ellipsoid, may be usedas a heterogeneity descriptor for the sandstone model. Heterogeneity analysesshow that compaction reduces the spatial continuity in the horizontal directionmore rapidly than in the vertical one.
The architecture and geometry of the network representation of the porespace are determined by applying various 3-D image analysis algorithms directlyon the fully characterised sandstone model. A 3-D pore network which wasgenerated from thin section data from a strongly water wet Bentheimer sandstoneis used as input to a two-phase network flow simulator. Simulated transportproperties for the sandstone model are in good agreement with those determinedexperimentally.
The existence of a highly-permeable void or pore space distinguishesreservoir rocks from other rocks. The pore space of reservoir rocks is highlychaotic, consisting of a spatial network of pores in which larger pores (porebodies) are connected via narrower pores (pore throats). The architecture andgeometry of the pore network and its complementary grain matrix determineseveral macroscopic properties of the rock such as absolute permeability,relative permeability, capillary pressure, formation factor, and resistivityindex1 .
The problem of predicting macroscopic rock properties from the underlyingmicroscopic structure and pore-scale physics has been the subject of extensiveinvestigation. One commonly applied tool in this investigation has been the useof network flow simulators (network models). This approach requires a detailedunderstanding of the physical processes occurring on the pore-scale and acomplete description of the morphology of the pore space. The procedure hasbeen applied, with success, to two-phase flow in simple or idealized porousmedia2-6 using pore-scale physics identified inmicromodels7-9. Recently, much of the pore-scale physics forthree-phase flow have been unveiled10-16, leading to the developmentof network models for three-phase flow17-19. The extension of thesetechniques to real porous media has been complicated by the difficulty ofadequately describing the complex 3-D pore structure of real porous rocks.
Information about the pore structure of reservoir rocks are usually obtainedfrom mercury injection data and image analysis of thin section images.Excellent overviews of methods for characterizing pore structures are given byYanuka et al.20 and Wardlaw1. Mercury injectiondata provide statistical information about the pore throat size distribution,or more correctly, the distribution of volumes which may be invaded withinspecified pore throat sizes. Traditional analyses of pore geometries by imageanalysis of thin sections are normally based on Delesse'sprinciple21 which states that 2-D observations in thin sections arerepresentative for the 3-D sample. This is, however, only valid for scalars,i.e. for areal and volumetrical considerations. The principle is thus valid forporosity and mineralogical measurements, but not for geometrical measurementsof pore bodies and throats since these are vectors.
|File Size||7 MB||Number of Pages||14|