Key points are not available for this paper at this time.
Depleted oil and gas reservoirs have the potential to store an estimated 675 to 900 gigatones of carbon dioxide (CO2). Such reservoirs represent a significant resource for long-term geologic CO2 storage (GCS), which is a key mitigation strategy to reduce atmospheric CO2 (Metz et al., 2005). However, the relatively high density of wells drilled into oil and gas reservoirs means that there exists significant potential pathways for unwanted fluid migration into overlying receptors in a GCS scenario. Wells, especially in fields that span many decades of exploration and production, are often known to have issues with their integrity. The reasons for loss of well integrity comprise a wide range of technological, operational, and environmental causes. In GCS, the buoyant CO2 plume and region of elevated pressure expand from the injectors and contact a large number of the existing and potentially leaky wells. This contact can occur both during injection and for significant time during the post-injection site care. Understanding the long-term potential of these wells to serve as unwanted fluid migration pathways is, therefore, critical to ensuring the viability of these resources as storage reservoirs. The National Risk Assessment Partnership (NRAP), a United States Department of Energy led initiative, is developing a set of tools and methodologies that use a science-based understanding to characterize risks associated with GCS. One such product is the Well Leakage Analysis Tool (WLAT), which is a tool that estimates well leakage in the form of a response to different scenarios representing initial well states and configurations. This tool contains a set of reduced-complexity models (e.g., reduced-order models and reduced-physics models) that give a rapid assessment of leakage for several different scenarios and well conditions. In this study, the WLAT is used to estimate CO2 and brine leakage along existing wells in a hypothetical storage scenario in Natrona County, Wyoming. We identified several key parameters that in comparison with others can significantly affect the amount of CO2 and brine leaked. Among those parameters are distance from injection site, well age, leak path length and permeability. Using the WLAT, the wells with the highest likelihood of excessive leakage were estimated, and the rate of CO2 and brine flux quantified. A workflow is outlined in which candidate fields for GCS can be assessed to identify wells that represent the greatest impact of leakage. Additionally, the proposed approach can be used to test possible linking relationships (e.g., if leak-path permeability is correlated with well age or a documented history of well integrity issues) and their impact on the leak behaviour. This information can then be used to rank wells with the highest priority for recompletion before CO2 injection is initiated. Additionally, predicting the potential breakthrough time and flow rates of fluids from the existing wells provides valuable information to help estimate where and when monitoring will be most effective and aids in decision making about injector placement at a storage site. Operators can then make decisions regarding the location of the injection site based on proximity to the wells with the longest breakthrough time and minimum flow rate. The ability of the WLAT to predict potential leakage before any CO2 injection occurs will ultimately reduce the resources (time and money) spent on injection into sites ill-suited for long-term geologic CO2 storage.
Doherty et al. (Sat,) studied this question.