Efficient models for Microbially Induced CAlcite Precipitation as a seal for CO2 storage

Storage of CO2 in North Sea formations can make a significant contribution to mitigate climate change if several gigatonnes per year are stored. These high injection rates will be linked to large pressure buildup, especially near injection wells, leading to a concern for leakages. A promising method to stop leakages is to produce biofilms that clog the top of the storage formation in the tens of meters around the injection well that experience the highest pressure buildup. The activity of some bacteria will also lead to formation of calcite, which further seals the rock. This is called microbially induced calcite precipitation. So far, laboratory measurements and simulation tools have mainly focused on the sub-meter scale. For field-scale application of this method, it is important to be able to design a strategy for injection of microbes and other components that favors a seal at the top of the formation while maintaining the possibility to inject CO2 further down. To be able to answer questions such as these, current models must be upscaled. It is well known that use of subsurface flow models that include reactions can produce large errors if not properly upscaled. Small-scale variability in concentration of microbes and solutes due to variable rock properties is a key factor here. Thus, the main goal in project MICAP is developing a robust and efficient model that includes small-scale variability within a field-scale simulation tool. As a first task, we performed a literature review of the most important microbial processes for large-scale injection. From the study we developed two promising injection approaches for efficient calcite precipitation: (I) injection of pre-cultivated microorganisms and cementation solution for calcite precipitation from suspended bacteria; and (II) injection of bacteria, growth solution, and cementation solution to cultivate biofilm (i.e., a community of bacteria) and subsequent calcite precipitation from the biofilm. Numerical simulations showed that both injection strategies produce significant sealing of the rock, where we chose (II) as our main approach in future studies. Based on learnings from the literature review, we developed a field-scale simulation model where only the necessary approximations and simplifications of small-scale biochemical reactions were included. The simulation model ran at a low computational cost compared to previous pore-scale and core-scale simulators, enabling simulations of 3D systems with leakage path explicitly included. Numerical investigations with the simulation model on a well-establish 3D CO2 leakage benchmark showed that we were able to close the leakage path completely. The simulation model was implemented in Matlab Reservoir Simulation Toolbox (ad-micp module) and the Open Porous Media (OPM) Flow simulator, and is a part of the official releases as of versions 2021b and 2021.10, respectively. Part of the Post Doc work was to complete a core-scale model from an earlier NFR project (IMMENS), which contributed to the development of the field-scale MICP model by advancing the knowledge about biofilm growth. An important task in project MICAP was the investigation of formal upscaling methods from pore- to field-scale. The biochemical reactions in MICP leads to geometry changes on the pore scale, which in turn leads porosity/permeability changes. We have proved, using mathematical arguments, that these changes are monotonic. The results have been extended to effective permeability and the diffusion tensor. We have also studied the reactive transport in fracture geometry, and derived upscaling procedures in certain regimes of porosity and permeability scaling. Further we have worked on developing an optimization tool to achieve optimal sealing of leakage paths. Using a novel statistical method where we combined of a proxy model of the simulator and an efficient Monte Carlo sampling procedure, we were able to consider various uncertainties that exist in the reservoir, f. ex., rock (permeability, porosity, etc.), biological (biofilm growth, etc.), and physical parameters (injection rates and periods, etc.). Numerical results showed that we could get robust values for, for example, injection start for one of the chemicals under any realization of a highly sensitive biological parameter. The method could thus aid in finding optimal injection strategies that works regardless of the uncertainty present. To focus on maximizing calcite precipitation at the leakage paths, we developed an optimization procedure with a novel injection strategy where components were injected at separate parts of the well. Numerical results on 3D systems showed that we were able to close various leakage paths, reducing the CO2 leakage by more than 99.8% in all cases.

The main outcome from the MICAP project is a field-scale numerical model for microbially induced calcite precipitation (MICP) implemented in open-source simulators OPM Flow and MRST. This enables academic, as well as industry, practitioners of MICP to easily run field-scale simulations and further develop the models. The rigorous numerical analysis and derivations of upscaling methods done in project MICAP will increase the robustness of MICP simulation models. The numerical results from the simulation and optimization studies performed in project MICAP advances the understanding of MICP for field-scale application. Together with field-scale demonstrations, the technology can be elevated to the level of safe utilization. With robust and efficient simulation tools for MICP we believe that the public confidence in the sealing technology will increase and, as a result, the social acceptance of geological CCS will increase.

Project MICAP will develop robust and efficient simulation tools to improve strategies that prevent leakage of CO2 in subsurface storage operations. Large-scale CO2 storage is pending on the regulatory framework that requires measures to mitigate and plug possible leakages. The project focuses on sealing by biomineralization, which was chosen because of its significant sealing capabilities and the potential for laterally extensive seals of biofilms and calcite. The ability to design injection strategies for this method is imperative to ensure good sealing properties while also maintaining the ability to inject CO2. This requires simulation tools that describe relevant processes at the field scale. Current models apply to much smaller scales, and the direct application of these models outside of their range of validity could lead to disastrous consequences. Accordingly, Project MICAP develops methodology to upscale proven existing models to the appropriate scales with verification at each step. The resulting methods will be used to develop strategies to seal potential leakage pathways in CO2 storage and will thus be essential for the implementation of large-scale CO2 storage as a method to limit global warming.