Online identification of evolved Takagi Sugeno fuzzy model for CO2 sequestration process

In recent years, carbon capture and storage (CCS) has been recognized as a promising technology to achieve a considerable reduction of greenhouse gas emissions from large local industries. Among different methods of CCS, carbon dioxide (CO₂) sequestration in underground saline aquifers has gained much attention due to its long-term storage and low cost benefits. This type of sequestration, however, poses over-pressurization as a potential risk. This paper aims at effective monitoring of critical parameters which directly impact the CO₂ sequestration performance due to over-pressurization and cap rock failure. A synthetic reservoir model is simulated in reservoir simulator (ECLIPSE-100) environment and an online fuzzy model is identified using an evolving Takagi Sugeno (eTS) algorithm. The approach recursively develops an evolving fuzzy rule-base model structure with linear rule antecedent parts using Recursive Least-Squares (RLS) parameter estimation to track reservoir dynamic changes during the CO₂ sequestration. Suitability of the presented adaptive identification approach in modeling CO₂ sequestration dynamic performance in an underground saline aquifer is verified via various test studies.