SEISMIC ANALYSIS AND RESERVOIR ASSESSMENT FOR CARBON STORAGE USING MACHINE LEARNING APPROACHES IN QADIRPUR GAS FIELD, CENTRAL INDUS BASIN, PAKISTAN

Abstract

The research addresses critical global issues such as climate change, energy transition, and environmental protection. By evaluating the Qadirpur Gas Field for carbon storage potential and employing cutting-edge machine learning techniques, this study offers practical solutions and insights that can shape both policy decisions and industry practices while promoting a more sustainable and responsible approach to energy and resource management. The key goals involve evaluating the Qadirpur gas field's potential for carbon dioxide (CO2) storage through well log and seismic data. The evaluation involved petrophysical analysis, forecasting lithofacies, estimating shear logs, petro-elastic analysis, delineating horizons, and identifying faults using seismic data. The reservoir assessment included analyzing the spatial distribution of impedance and porosity, conducting thermodynamic analysis, and estimating the capacity for secure containment of CO2. Petrophysical analysis reveals that CO2 can be securely stored within a 30-meter reservoir zone (depth range of 1336 to 1366 meters) in the Sui Main Limestone (SML) Formation. Porosity values within this zone are favorable, with average values for total porosity of 21.09% and effective porosity of 18.86%. Lithofacies predictions through Self Organize Maps (SOM) successfully distinguished various lithologies and confirmed that the identified zone is hydrocarbon-bearing limestone. Moreover, a machine learning algorithm namely, Multi-linear Regression was used to predict the shear log, enabling the calculation of petro-elastic properties which are in turn used in the stress state within the reservoir and caprock. Caprock integrity was validated through assessments of permeability and effective porosity variations with lithology, as obtained through the SOM. Seismic analysis revealed that the study area lying in the Middle Indus Basin is relatively stable in terms of tectonic activity. Spatial distribution analysis of reservoir properties, including impedance and porosity, highlights a promising reservoir with low impedance values and high porosity values, demonstrating its suitability for CO2 storage. Furthermore, initial temperature and pressure conditions within the SML Formation indicate the feasibility of storing CO2 in a supercritical state. The developmental history of the field is marked by sequential capacity enhancements, vi resulting in continuously diminishing production. This makes the field suitable for Enhanced Oil Recovery, subsequent CO2 storage and its significance in promoting sustainable energy practices.