Investigating the effect of top surface morphology on CO2 storage using numerical simulation

    Student thesis: Doctoral ThesisDoctor of Philosophy


    Geological carbon storage is a promising technology for reducing CO2 emissions released into the atmosphere as one contribution to the Paris Agreement goals to limit the global temperature increase to 1.5oC (UNFCCC 2015). Among possible storage sites, deep saline aquifers have the largest storage capacity. Once injected into the aquifer, due to its smaller density compared to that of formation brine, the CO2 in the free phase tends to migrate upwards and forms a thin layer below the caprock (a low permeable formation that prevents the free CO2 plume from further upwards migration).

    This work investigated the impact of the caprock morphology on CO2 plume migration in the saline aquifer storage sites, especially in relation to elevation changes below the seismic detection range. Several mathematical tools and methods have been employed throughout this work, such as analytical calculation, numerical simulation, vertical equilibrium (VE) modelling (MRST and EVE), and data analysis and optimisation. The impact of caprock morphology and aquifer boundary on the CO2 storage process through 3D numerical simulation and analytical calculations were subsequently investigated. The effect of boundary conditions on the storage process shows that CO2 dissolution in aquifers with one closed end (due to faults, salt walls, etc.) is higher than an open aquifer.

    Moreover, the analytical approach shows promising performance for estimating the CO2 storage capacity (and possibility), making it a suitable site-screening tool before performing numerical simulations. The results suggest that dissolution is strongly correlated with formation dip (the acute angle with a horizontal plane). In models with low vertical permeability, however, increasing the tilt angle resulted in a lower dissolution (opposite to what was observed in previous studies). The plume outline in VE and 3D models was found to be similar, and in terms of computational cost, MRST was found to outperform the rest significantly. The conventional 3D simulators could be computationally intractable for long­term geological CO2 storage problems. Therefore, the feasibility of simplified, computationally inexpensive VE models in studying the CO2 storage process was investigated over a wide range of scenarios with various caprocks. The results were compared with several full 3D simulation methods and show that the VE approach is an effective method to model the relevant physical effects of geological COstorage.

    An optimisation tool implemented using VE formulation was seen to improve the match between the observed and simulated plume outline in a synthetic model and the operational Sleipner field storage site in Norway. The results show an improvement of around 8% in the Sleipnerplume match resulting from an average absolute elevation change of 3.23 metres. Calibrating the porosity, permeability, CO2 density and injection rate results in a 5% improvement in the match, and once caprock morphology is included in the optimisation process, the match improvement increases by 16%. Subsequently, and for the first time, the importance of caprock topography variations has been quantified in the Sleipner model in the presence of other uncertain parameters, including porosity, permeability, CO2 density, injection rate and temperature, using data analysis tools. The results show that caprock morphology is the second most important parameter (after density) in controlling the CO2 plume migration in the Sleipner field.

    This work raises the scientific understanding of the complexity of the impact of the caprock morphology on CO2 plume migration in a real field model for safe sequestration, such as the most recent Sleipner Benchmark simulation grid and implies that its impact on model predictions has previously been underestimated.

    Date of AwardAug 2021
    Original languageEnglish
    Awarding Institution
    • Coventry University
    SupervisorSeyed Shariatipour (Supervisor) & Odd Andersen (Supervisor)

    Cite this