Reactive rock-fluid dynamics under impurity-bearing CO₂ injection: toward predictive geological CCS

Geological carbon storage must simultaneously address two practical challenges: (i) realistic CO₂ streams that contain impurities from capture and transport, and (ii) highly nonlinear subsurface reactions that can modify porosity, permeability, and stress, ultimately controlling injectivity and long-term containment. We present an integrated modeling strategy that links multicomponent thermodynamic evaluation of CO₂-impurity mixtures with fully coupled hydro-mechanical-chemical (HMC) simulations of reactive flow in deformable rocks. 

Our framework resolves the mutual feedbacks between fluid migration, chemical re-equilibration, and evolving pore space, capturing localized mineral alteration and the emergence of sharp reaction fronts. A key outcome is that mineral trapping and pH evolution are strongly localized: carbonation and dissolution tend to occur within a narrow front whose migration speed is dictated by the interplay of pressure gradients and evolving permeability. Importantly, we find that simplified chemistry can severely over-acidify predicted fluids, whereas explicit treatment of aqueous speciation and host-rock buffering stabilizes pH at realistic values, even in the presence of acid-forming impurities. 

Accurate prediction of rock-fluid equilibria and pH fields demands shock-resolving spatial resolution together with full aqueous speciation. Under-resolved meshes and simplified reaction networks artificially diffuse concentration discontinuities, leading to large errors in acidity, mineral alteration extent, and permeability evolution. Benchmarking against laboratory-scale observations and illustrative field-scale scenarios confirms that GPU-accelerated, fully coupled HMC simulations are essential to capture extreme localization and front propagation dynamics. 

In summary, impurity-bearing CO₂ storage is controlled by sharply localized reactive fronts, where pH buffering, mineral alteration, and porosity–permeability evolution are tightly coupled. Only conservative, fully coupled high-performance HMC simulations with explicit multicomponent speciation (neutral and charged species) can resolve this localization and provide robust guidance for impurity-tolerant injection design and long-term containment, delivering the predictive capability required for reliable geological CCS.