Agent‐Based Discrete Element Modeling of Microbial‐Induced Carbonate Precipitation

ABSTRACT Biomineralization is a promising approach for producing composite materials that contain living organisms, biopolymers, and minerals. Brittle biomineralized materials can be potentially toughened by more deformable microorganisms and biopolymers. Microbially‐induced carbonate precipitation (MICP) is a common type of biomineralization that occurs when urease‐producing bacteria hydrolyze urea to form calcium carbonate (). MICP is often used in porous materials to improve their mechanical properties. The morphology of the precipitation is strongly influenced by the bacteria and the morphology and porosity of the host medium. To explore their effects on MICP, we constructed an agent‐based discrete element model to investigate clustering and morphology in tandem with the growth of bacterial cells and the secretion of extracellular polymers. The model parameters are validated and optimized via our experimental measurements. We further addressed parametric uncertainties and quantitatively evaluated the relative importance of enzyme secretion, catalytic kinetics, and diffusion dynamics parameters by employing Monte Carlo simulations combined with data‐driven variogram analysis. Based on this model, we also examine the influence of bacterial concentration, bacterial distribution, urea concentration, and the geometry of porous media on the morphology. Unlike prior field‐scale or particle‐based methods that primarily focus on structures, our model will pave the way toward the study of the kinetics of urease‐producing bacteria and the mechanically relevant microstructural features of MICP‐treated composite materials.