Data Fusion Workflow to Characterize Reservoir Water Distribution and Tortuosity

Abstract Formation water exists naturally in conventional hydrocarbon reservoirs and is associated with oil and gas in the pore space. Characterizing formation water and its connectivity can help not only to evaluate hydrocarbon volume but also understand oil and gas migration path. In CO2 storage projects, the information is also valuable for optimization of CO2 injectivity. For heterogeneous reservoirs and shaly formations, it is challenging since porosity and fluid distributions vary both vertically and laterally. Conventional resistivity-based imaging provides resistivity information, not formation water nor water tortuosity. Resistivity to water saturation conversion requires independent inputs that are often unknown or not easy to obtain. On the other hand, conventional multi-frequency dielectrics measurement provides water volume information, only in front of the measurement sensors without azimuthal coverage. In this paper, we provide a new workflow to derive images of water-filled porosity and water-phase tortuosity by data fusion of measurements from the latest generation OBM imager and an unidirectional multi-frequency dielectrics logging tool. The challenges of the data fusion of the above two measurements include the unknown bias of the processed imaging data and the data compatibility between the two tools. A fusion strategy is required to combine data from both tools, while preserving the maximum azimuthal and depth variations in the data. In this paper, we adopted methods based on an established model and also laboratory measurement to investigate the bias of the imaging data processing and identify relationships between rock parameters and the inversion process of water-filled porosity. The above workflow is applied first to an outcrop rock measurement in the laboratory, then to an example of downhole logs. The laboratory data acquired in a controlled environment validates the compatibility of the two types of data as well as the accuracy of the fused water-filled porosity. The data fusion method developed in this study provides an accurate image of water-filled porosity with no assumptions on the rock types and rock parameters. For the downhole log data, the Archie-based water-filled porosity image is also derived as a comparison. Results indicate that in the sections containing interbedded sub-inch layers, where the water tortuosity is high, the Archie-based image of water-filled porosity tends to underestimate water volume, most likely due to the smaller value of cementation factor used. In heterogeneous rocks with complex pore systems, the fusion of the two measurements of OBM imager and unidirectional multifrequency dielectric log can generate a new reservoir characterization answer product. The separation of water-filled porosity and water phase tortuosity from rock conductivity and permittivity measurements can not only improve the accuracy of reservoir characterization but also provide additional information on reservoir quality.