Speaker
Description
In recent years, the pre-salt reservoirs have gained visibility due to their large hydrocarbon reserves, currently representing the main source of Brazil’s oil production. These reservoirs are predominantly composed of carbonate rocks, which are highly heterogeneous and exhibit a wide range of pore types and pore sizes. Such complexity makes the estimation of petrophysical properties challenging when relying on a single scale of analysis, thereby requiring a multiscale approach. With advances in X-ray micro-computed tomography and computational capacity, petrophysical properties can be obtained through numerical simulations on 3D images, for example. However, this technique is limited by the relationship between sample size and resolution: higher resolutions provide better detail of the sample but require a smaller physical sample size. Additionally, high-resolution images typically generate heavy datasets, resulting in long processing times. Therefore, it is often necessary to select a representative elementary volume (REV), taking porosity and permeability values into account. In this context, this work analyses the impact of choosing different REVs on global porosity and permeability values, integrating macro and micro-scale data. To achieve this, one micro-scale and one macro-scale image from the Digital Rocks Portal were selected. These images were processed and segmented into three phases at the macro scale (pores, unresolved phase and matrix) and two phases at the micro scale (pores and matrix). Subsequently, several subvolumes were extracted from the micro scale, and porosity and permeability values were estimated and used for REV determination statistically. Two distributions were fitted for the subvolume identified as the REV, and the corresponding micro-scale porosity and permeability values were used in the unresolved phase of the macro scale in the Brinkman equation to obtain the permeability of the full image. For the micro-scale, connected porosity and permeability estimates were obtained using the PNM method available in the open source software GeoSlicer (developed by Ltrace, Equinor, and Petrobras). For permeability estimation at the macro scale, the Brinkman model was used after adapting the SimpleFOAM solver available in OpenFOAM. Preliminary results showed that fluctuations in micro-scale permeability have more influence than the porosity on the permeability estimates, suggesting that a rigorous REV selection is crucial for obtaining properties with a better agreement against experimental data when using resolved information from higher-resolution images in the Brinkmann model.
| Country | Brazil |
|---|---|
| Acceptance of the Terms & Conditions | Click here to agree |
