Wellbore Leakage Model for Above-Zone Monitoring at Cranfield, MS

摘要

Geologic storage of CO2 for atmospheric emissions reductions imposes unique requirements to document containment. Monitoring pressure in strata above the injection interval is a fit-to-purpose technique to document performance of confining system and degree of isolation provided by existing wellbore completions. Field data are collected over two-and-a-half year period during a continuous industrial-scale CO2 injection at an enhanced oil recovery (EOR) site at Cranfield Field, Mississippi. Continuous downhole high-precision pressure and temperature data were collected at a monitoring well at two depths: at the injection interval and at a selected above zone monitoring interval (AZMI). The AZMI is a prevalent non- productive sandstone above the injection zone and a thick confining system. Pressure data show a perturbation in above zone contemporaneously with pressure elevation in injection zone, which suggests a possible interformational fluid communication via wellbore. Meanwhile temperature data maintain a linear correlation between zones with a consistent differential, which indicates negligible volumes of injection interval fluid being introduced into the AZMI. Interpretation of the data requires a physics-based transport model to illustrate the possibility of wellbore leakage and quantify the rate if leakage exists. We model the wellbore leakage by coupling the flow in wellbore and a diffusion model in the above zone sand layer. Matching the pressure data yields an effective wellbore permeability in order of tens of darcies. This corresponds to a large flow rate along the pathway which would very likely raise the temperature in the above zone. To gain insight about the temperature response, we model the heat transfer between the fluid in wellbore and the surroundings. The heat transfer coefficient is tuned and justified by modeling the heat conduction in the formation rock. In order for the temperature in above zone to remain unaffected by that in injection zone, the flow rate should be no more than 10 g/s and the corresponding wellbore permeability not exceed a few darcies. This value is at least an order of magnitude smaller than that estimated from the pressure response. Only if the sand layer in above zone is assumed to have a closed boundary within a few hundred feet of the monitoring well can the pressure data and temperature data be made consistent. However the assumption of closed boundary is not very feasible since there is no evidence of the sand layer being closed by faults locally. We conclude that leakage from the injection zone is very small. The observed pressure increases in the monitoring well are attributed to larger-scale geomechanical phenomena.