Abstract
<jats:title>Abstract</jats:title> <jats:p>The objective of this study is to update dynamic reservoir simulations through the integrated use of 4D seismic volumes and to enhance the reliability of CO2 plume prediction. The developed workflow was evaluated using the Sleipner CCUS project, where it demonstrated robust performance and effective integration of seismic monitoring and simulation data. In this method, time-lapse (4D) seismic surveys were analyzed to map the injected CO2 using Bayesian inversion followed by interactive deep-learning seismic interpretation tools for plume identification. The interpreted plume was imported into the reservoir model's mesh grid. A robust workflow was developed to history match the seismic and simulated CO2 plumes, using the difference between the numerical model's gas saturation and the seismic-derived plume as the metric for quantifying match quality. Sensitivity analysis was conducted to assess how uncertainties in key parameters affect the history match.</jats:p> <jats:p>At Sleipner, the CO2 plume became identifiable in seismic data by 1999. The plume's size and height increased until it reached the top of the formation; afterward, CO2 saturation increased and the plume expanded laterally. The upper layers continued to spread, whereas the lower layers stabilized in size. It was found that the Sleipner plume trends northeast and is strongly influenced by reservoir heterogeneities, geological architecture, and the vertical permeability of interbedded shales. The Area of Review (AoR) was calculated for each year and expanded from 1 km2 in 2001 to 3.3 km2 in 2010. Using deep-learning AI for seismic interpretation, the CO2 plume was identified and mapped across multiple seismic datasets with remarkable speed and accuracy using only 0.33% of the available data. By integrating seismic monitoring into the dynamic model and AoR calculation, this methodology supports continuous compliance with regulatory requirements for post-injection site monitoring and verification, helping ensure long-term containment integrity and environmental safety.</jats:p>