We present the results of a feasibility study for seismic monitoring using conventional surface seismic experiments at the CaMI Field Research Station, Alberta, Canada, where a small volume of gas-phase CO2 is being injected into a sandstone reservoir at a depth of 300 m. We first apply a careful fluid substitution procedure to the results of reservoir gas saturation and pressure responses obtained from fluid flow simulations. We test different methods to compute the bulk modulus of the fluid for different fluid saturation models. Assuming a semi-patchy model and considering only the replacement of brine with a maximum saturation of 50% CO2, we estimate the reduction in P-wave velocity to be 20%. Adding an increase in pore pressure of 2.7 MPa increases the P-wave velocity reduction to 32%. After including a field-based signal-to-noise ratio of 5% to the synthetic seismic data, the time-lapse seismic anomaly should be detectable after one year of injection (266 tonnes of CO2).