Microstructure Characterization For Sandstone Pores And Fluid Flow Simulation Based On Micro X-ray CT

Sandstone is an important oil and gas reservoir.The three-dimensional distribution of pores and fluid transport mechanisms at different scales in sandstone have important effects on oil and gas storage and transportation.It has become an important research method to establish the 3D pore network of sandstone samples based on X-ray CT imaging technology and numerically simulate the fluid transport phenomena in them.The current mainstream CT image threshold segmentation algorithm identifies a single CT voxel as complete solid or pore,which can result in the loss of structural information smaller than the CT voxel size.Increasing the CT imaging resolution will reduce the impact of this factor,but the significantly lower sample size that can be characterized by a high CT resolution will make the characterization area less representative.Quantitative acquisition of multi-scale pore distribution in sandstone reservoirs and the development of multi-scale percolating methods are important for understanding fluid transport mechanisms.In this paper,micro X-ray CT imaging was performed on an artificially sandstone sample.The three-dimensional distribution of the sample compositions was obtained using the least-squares image threshold segmentation method and the data-constrained modeling（DCM）respectively.The CT images in the same area of the sample at different resolutions were calculated by DCM.Numerical comparison was made with the calculated results.The single-phase water flow in the sample was simulated by partially percolating lattice Boltzmann method（PP-LBM）based on the flow fields generated by the two algorithms.Comparing to the least-squares image threshold segmentation algorithm,the DCM calculation results included not only the large pores between quartz particles but also the small pores that usually coexist with calcium carbonate.The porosity that the DCM calculated is similar to the results obtained by helium adsorption method.The DCM can obtain total porosity values and multi-scale pores based on lower resolution CT images similar to those of higher resolution CT images.The multi-scale flow field was established with the results of DCM calculations,and the single-phase flow of water in the sample was numerically simulated using the PP-LBM.The results showed that the permeability calculated based on DCM generated flow field was significantly higher than the flow field obtained by the threshold segmentation method.The pores smaller than the CT voxel size have a significant contribution to the permeability.The fluid flow velocity distribution in the artificial sandstone samples is not proportional to the porosity distribution,while the pore distribution at the fluid inlet surface has a greater influence on the fluid flow velocity distribution.The method was extended to the multi-scale structural characterization and seepage simulation of three natural sandstone samples from different regions of Ordos Basin.A standard aluminum sample was imaged with the sandstone sample,and the X-ray equivalent energy was determined by the standard aluminum sample.The corresponding data-constrained model was established based on the X-ray linear absorption coefficient characteristics of each mineral compositions of the samples.The total porosity of the sandstone samples was calculated using the digital terrain model（DTM）.The pore self-energy parameters in the data-constrained model were determined by the total porosity value calculated by DTM.The multi-scale pore structures of the three sandstone samples were calculated and the multi-scale flow fields of the samples were established.Basing on the PP-LBM,the single-phase flow of water in the three natural sandstone samples were simulated,and the three-dimensional distributions of flow velocity were obtained.The histogram of porosity distribution and three-dimensional distribution of pore clusters of the samples were calculated.The relationships between the three-dimensional structure of pores and fluid flow velocity distribution of different samples were compared and analyzed.The results show that for individual sample,high flow velocity regions usually exist high porosity.There is no direct relationship between sample permeability and porosity.The porosity distribution,flow velocity distribution,permeability and the relationship between the pore network and the permeability of the three natural sandstone samples show different characteristics.The results of this paper provide new ideas for the development of corresponding multi-scale structural characterizations and numerical simulation methods for fluid transportation in sandstone reservoirs.It is also a data basis for reservoir evaluation,fracturing modification of oil and gas reservoirs,and CO₂replacement and sequestration.