Petrophysical Joint Inversion of Multi-Geophysical Attributes and Measurements for Reservoir Characterization

摘要

Interpretation of petrophysical parameters such as porosity and saturation is very important for reservoir characterization and resources evaluation. Petrophysical inversion is typically done using well log data to characterize the reservoirs. Seismic and electromagnetic (EM) measurements are sensitive to different rock properties and their joint interpretation can improve the characterization results. We address the task of improving the interpretation of porosity and fluid saturation in the reservoir rocks by means of the joint inversion of seismic, gravity and EM data subject to geological constraints. Simultaneous inversion is expected to provide higher sensitivity, better resolution, and an overall more robust estimation of rock parameters than conventional log analysis. We developed in-house petrophysical joint inversion (PJI) codes and tested them on simulated and field data. We also designed a rigorous workflow for reservoir characterization through PJI of multi-geophysical attributes and measurements. First, we conduct analysis and inversion of the available well log data to understand the petrophysics of the reservoir and to choose the constitutive equations linking petrophysical attributes to geophysical ones (calibration). Second, we implement sequential PJI of geophysical attribute volumes such as seismic velocity, density and resistivity. Finally, we produce a full PJI of geophysical measurements such as seismic, gravity and EM data. Each of these steps requires the integration of geophysical, geological and petrophysical knowledge about the specific reservoir. The first step of the workflow was applied to well log data from an important reservoir in Saudi Arabia. Mineralogical information from the well logs indicated mainly calcite with some dolomite at the reservoir depth. The rock properties were adjusted with depth according to the known stratigraphy. The recalculated log resistivity, density and sonic data via joint inversion fit the observed data well. The estimated porosity and fluid saturations were the petrophysical parameters in the joint inversion while gas saturation was ignored. The joint inversion estimated porosity was similar to neutron log-derived porosity. Improved oil saturation estimates are expected to enhance the interpretation of the resource. To test the other two steps of the workflow, we created a simulated 3D multi-geophysical model by lateral interpolation and rescaling of well log data from available wells. Then, we generated the related geophysical data through forward modeling. Finally, we tested the sequential and full PJI algorithms on this dataset. We are currently working on the application of the described method and workflow to a measured dataset from a Saudi Arabian reservoir. If successful, this approach could lead to an overall more robust estimation of the reserves.