Abstract
Gas-bearing clay-rich clastic reservoir characterization presents challenges in accurately predicting reservoir properties, particularly in giant fields. Despite two decades of production, log-based predictions remain uncertain, and this study addresses the challenges by proposing a comprehensive workflow that integrates various data sources and methodologies.
The main hurdle lies in the substantial permeability variation, up to two orders of magnitude for a given porosity value, primarily attributed to dispersed clay. Incorporating clay into petrophysical models becomes imperative but increases the complexity of Log evaluation, especially considering variable invasion profiles and drilling practices influencing porosity related Log responses. Moreover, uncertainty persists due to the variability of clay volume across the wells and challenges associated with tilted contacts and incomplete water level data for their appropriate 3D characterization.
To overcome these challenges, an inversion-style workflow is proposed, leveraging the consistency of two independent water saturation (Sw) estimation methods: the Saturation Height function (SHF) and the Resistivity Log. The approach builds upon establishing 2 dimensional correlations based on core measurements which incorporate dispersed clay volume. Bulk rock porosity and clay volume are estimated using neutron-density logs corrected for gas effect applying Gaymard & Poupon (1968) and Segesman & Liu (1971), with a revisited clay correction method accommodating calibration variations across neutron tool types. Addressing the underdetermined nature of solutions, ensemble-style interdependent formation models are prepopulated for fixed increments of invaded zone saturation. These models extend to include clay-dependent permeability, SHF and Sw derived from resistivity logs. Moreover, the Thomas-Stieber- Tyurin thin bed reservoir model (Tyurin & Davenport, 2023) is applied to accurately evaluate dispersed clay, laminar shale volumes and sand texture porosity. The Workflow is concluded by addressing the free water level (FWL) plane uncertainty through iterative calibration of the Log estimates to the relevant core measurements and selection of formation property ensembles which minimize discrepancies between the independent Sw methods.
The results highlight the significant influence of uncertainty related to invasion or gas effect correction on reservoir properties, emphasizing the importance of including clay volume in petrophysical models. By achieving near-zero calibration errors for key reservoir properties such as porosity, permeability, and Sw, the proposed workflow enhances reservoir characterization accuracy. Integration of independent methods with constraints, namely sufficiency of gas correction and fluid volumes balance, considerably reduces uncertainty of the predictions.
In conclusion, this study presents a comprehensive workflow that ensures consistency between core, log, and field data, thereby enhancing the reliability of reservoir models through an implicitly defined dependency between reservoir properties.
