The impact of pore-throat shape evolution during dissolution on carbonate rock permeability: Pore network modeling and experiments

•Calcite dissolution can generate a spectrum of pore shape geometries.•In existing PNMs, conductance relations ignore pore shape evolution.•New conductance relations were derived from single pore scale simulations.•These relations are a function of volume and imposed dissolution regime.•Permeability may be overestimated by up to 27% if pore shape change is neglected. Pore network model simulation (PNM) is an important method to simulate reactive transport processes in porous media and to investigate constitutive relationships between permeability and porosity that can be implemented in continuum-scale reactive-transport modeling. The existing reactive transport pore network models (rtPNMs) assume that the initially cylindrical pore throats maintain their shape and pore throat conductance is updated using a form of Hagen-Poiseuille relation. However, in the context of calcite dissolution, earlier studies have shown that during dissolution, pore throats can attain a spectrum of shapes, depending upon the imposed reactive-flow conditions (Agrawal et al., 2020). In the current study, we derived new constitutive relations for the calculation of conductance as a function of pore throat volume and shape evolution for a range of imposed flow and reaction conditions. These relations were used to build animproved new reactive pore network model (nrtPNM). Using the new model, the porosity-permeability changes were simulated and compared against the existing pore network models. In order to validate the reactive transport pore network model, we conducted two sets of flow-through experiments on two Ketton limestone samples. Acidic solutions (pH 3.0) were injected at two Darcy velocities i.e., 7.3 × 10−4 and 1.5 × 10−4 m.s − 1 while performing X-ray micro-CT scanning. Experimental values of the changes in sample permeability were estimated in two independent ways: through PNM flow simulation and through Direct Numerical Simulation. Both approaches used images of the samples from the beginning and the end of experiments. Extracted pore networks, obtained from the micro-CT images of the sample from the beginning of the experiment, were used for reactive transport PNMs (rtPNM and nrtPNM). We observed that for the experimental conditions, most of the pore throats maintained the initially prescribed cylindrical shape such that both rtPNM and nrtPNM showed a similar evolution of porosity and permeability. This was found to be in reasonable agreement with the porosity and perm