Coupled Geochemistry Modelling of Underground CO2 Storage from Lab to Field Scale

Abstract Carbon sequestration is the process of capturing and storing atmospheric carbon dioxide. The objective of any carbon sequestration project is to store CO2 safely for hundreds or thousands of years with a goal of mitigating global climate change. Based on literature, geochemistry simulation modelling is a critical component of this process. It involves the use of mathematical models to simulate and predict the chemical reactions and transport processes that occur when CO2 is injected into geological formations. These models can help us understand how CO2 will behave over time and assess the long-term stability and safety of CCS projects. Static and dynamic aging are two techniques used in the laboratory experiments to measure geochemical reaction kinetics which were then used in dynamic simulation modelling. The static ageing technique ages the rock (core plug) sample saturated with brine by keeping the sample in the CO2 chamber under reservoir pressure and temperature for a certain amount of time. Usually this is between 30-90 days. Cores and effluent collected at the end of ageing days are analyzed to measure the changes on the key geochemical parameters (i.e., rock porosity, permeability, mineralogy) per and post-ageing and pre-ageing. However, the results of this lab experiment cannot predict the amount of CO2 mineralization, as the mineralization is very slow process and it takes place over hundreds of years, so long-term dynamic simulation modelling is required to forecast the mineralization. In numerical dynamic modelling there is a procedure to match the lab measurement and upscale the geochemical reaction kinetic. The question which this paper is going to answer is how to use the geochemical lab measurement data and adopt them to be used in the coupled geochemistry dynamic flow modelling of CCS. To answer the question, sample data from static and dynamic aging are used for batch model of the process in a small model with only one grid block. History matching process of mineral changes have been done for the batch model for every single mineral and reaction, and then the reaction kinetics were upscaled to be used in a full field. A sector model from an aquifer with one CO2 injection well was extracted and coupled geochemistry with compositional simulation modelling was performed. The model included CO2 solubility and geochemistry with five minerals (Kaolinite, Dolomite, Muscovite, Siderite, and Smectite) and their respective geochemical reactions. In the lab model ‘Muscovite’ and ‘Dolomite’ dissolved whereas ‘Kaolinite’ and Smectite were precipitated, and ‘Siderite’ initially is dissolved and then precipitated. Similar behaviors are observed throughout the simulation model. The upscaling workflow is explained in detail, and the results showed that the amount of mineralization in the model is about 10% of the injected CO2 after 1,000 years from the start of injection, while in the model without geochemistry is almost zero throughout the simulation. The results of this study prove that excluding geochemistry can cause the model to miss some of the active geochemistry processes, leading to less accurate results in the simulation. The geochemistry simulation should be based on the lab mineralogy data (the amount of each mineral in the core) and the static/dynamic ageing experiments and upscaling to the field should be considered for all CCS projects.