Leveraging Machine Learning to Effectively Evaluate Petrophysical Properties in a Deep Saline Aquifer CO2 Storage Project

The document discusses the potential of carbon dioxide (CO2) storage in deep saline aquifers, highlighting the challenges of evaluating these sites due to unreliable geological data. It emphasizes the importance of quickly assessing undiscovered areas to accelerate storage site readiness. The paper showcases the use of artificial intelligence (AI) to effectively evaluate carbon capture and storage (CCS) projects, speeding up reservoir characterization and feasibility analysis.

The conventional approach to identifying potential storage sites involves field-based properties and static modelling, which is time-consuming and challenging. In contrast, AI models use basic open hole raw data to predict reservoir properties such as porosity, permeability, saturation, and lithologies. AI predictions conducted on 200 wells across various fields in Malaysia demonstrated a remarkable accuracy rate of approximately 97%.

By leveraging AI, the modelling process has been significantly accelerated, reducing the time required for petrophysical evaluations related to storage capacity and confinement assessments. This innovative approach ensures dependable subsurface characterization, guiding storage screening and containment evaluations.