Optimization Under Geological Uncertainties for CO2 Injection in CCUS: A Case Study from the Lusitanian Basin

Abstract

This study presents a workflow for optimizing CO₂ injection under geological uncertainty for Carbon Capture, Utilization, and Storage (CCUS), using the Q4-TV1 prospect in the offshore Lusitanian Basin (Portugal) as a case study. The approach integrates a high-resolution 3D static geological model with dynamic reservoir simulation through the Big Loop™ framework and Bayesian optimization techniques. The objective was to maximize total CO₂ injection over a 30-year operational period while ensuring long-term containment and minimizing leakage risks associated with intersecting faults and a legacy well. Geological uncertainty was represented through 16 variable subsurface parameters, including porosity–permeability relationships and fault behaviour. A total of 948 simulation scenarios were evaluated, with unsuitable scenarios excluded where plume migration approached critical geological features. The selected optimal scenario identified a well location and perforation interval that achieved an estimated maximum injection capacity of approximately 24 million tonnes of CO₂, with a probabilistic median (P50) of 8.8 million tonnes. Extended simulations over 1,000 years demonstrated sustained plume containment away from faults and the legacy well. Sensitivity analysis showed that injection performance was primarily controlled by bottom-hole pressure targets and perforation depth. The results demonstrate that combining stochastic geological modelling with Bayesian optimization provides a robust and scalable framework for improving the safety, efficiency, and reliability of CO₂ storage projects under uncertainty.