Geological sequestration of CO2 in the pore space of subsurface rock formations offers a safe and permanent carbon storage solution. In this work, we present the application of a pore-scale flow simulator to the study of CO2 storage in geological formations. We model the rock pore space geometry, extracted from high-resolution X-ray microtomography images of suitable rocks as a network of connected capillaries. Assuming piston-like flow within each capillary and conservation of mass at each network node, a large system of equations can be solved to compute properties like pressure distribution or flow rate at each point in the network. Multi-phase flow simulations track the displacement in time of the fluid interface within each capillary. These dynamic simulations on the high-resolution capillary network representation of the rock are very computationally costly. Alternatively, analysis is carried out on the aggregate results of multiple two-phase flow simulations on several statistically equivalent capillary network models of the rock sample, which retain topological properties of the original at a significantly lower computational cost. We performed a sensitivity analysis with respect to multiple fluid parameters, such as viscosity, interfacial tension, contact angle, pressure, and temperature, and quantify their influence on the infiltration and retention of CO2 inside a capillary network that is representative of an actual rock.