Evaluation of the Effectiveness of Injection Termination and Hydraulic Controls for Leakage Containment

Abstract

One of the greatest concerns that the public has with large-scale adoption of CCS technology is the potential risk of carbon dioxide leakage from sequestration reservoirs. Among other things, addressing these concerns will require developing intervention and remediation strategies that can be quickly and effectively implemented should leakage occur. The three main methods proposed for remediation of leakage from carbon storage reservoirs are: (1) termination of CO₂ injection activities, (2) injection of a chemical or biological sealant to act as a flow barrier [1] or (3) the development of hydraulic barriers [2], [3], [4]. Creating hydraulic barriers relies on the injection of water in the overlying aquifer with or without injection or production of fluid from the injection reservoir in order to manipulate the pressure field to stop or reverse leakage. Here we first investigate the effectiveness of passive remediation (stopping CO₂ injection), and then determine its effectiveness in combination with a variety of different hydraulic controls such as injection of water above the fault, injection of water below the fault and production of brine in the lower reservoir.

Regardless of when the leak is detected, simulation results show that passive remediation (stopping injection) almost immediately reduces the leakage rate by an order of magnitude. Depending on the degree of residual trapping, leakage can be even further reduced. For example, with a maximum residual CO₂ saturation of greater than 20%, leakage rates drop by another factor of 4 or more. In many cases stopping injection may reduce the leakage to the extent that is so small that it reaches an acceptable level. However in the case where further reductions are desired, for example, in order to completely stop leakage, the implementation of hydraulic controls may be necessary. The most effective method for completely stopping leakage is to inject water into the overlying aquifer near the breach in the caprock. Drawbacks of using this method alone is that while it stops leakage during water injection, leakage resumes after the cessation of water injection. In order to prevent leakage from continuing after water injection ends, the CO₂ plume in the injection reservoir must be displaced away from the bottom of the fault zone. Displacing the CO₂ plume away from the fault is accomplished by injection of water below the fault while injecting water above the fault, and by producing reservoir brine on the opposite side of the CO₂ plume. Reservoir brine production helps pull mobile CO₂ away from the bottom of the fault. In this study, production rates are established by balancing water injection rates, therefore reducing costs and surface storage issues associated with leakage remediation. The most effective hydraulic controls for stopping CO₂ leakage combine brine extraction from the storage reservoir with water injection into the overlying aquifer and storage reservoir near the fault.

Overall this study demonstrates that temporally limited, multi-stage remediation strategies using a combination of stopping injection and hydraulic controls can permanently terminate leakage while having the additional benefit of dissolving most of the CO₂ in the overlying aquifer into the resident brine. This finding should provide assurances to industry, policy makers, and the public that intervention measures can quickly and effectively mitigate the risks of leakage should leakage occur.