Modeling forms an essential part of the workflow across all stages of an SCS project and informs the risk assessment of multiple hazards. However, many published pre-injection model-based predictions show either minimal leakage or no failure over very long-time scales. For example, the Figure below displays results from an offshore, well-regulated model of 12 gigatons of CO2 injected into saline aquifers. The amount of leakage over a period of 10,000 years is one-half of one percent, with most of this occurring between years 1,000 and 10,000.
Subsurface models are necessarily a simplification of a complex geological system; therefore, stress-testing models is a key due diligence step. Best practice recommends that multiple subsurface models should be generated. However, due to the time and computing power needed, there is a tendency to generate a “best case” model without fully exploring the impact of parameter uncertainty on the risks.
Independent critical evaluations of storage complexes have been recently undertaken and published. For example, several critical issues were identified in the HyNet project in Liverpool Bay, United Kingdom, including the shallow top of the main reservoir, a waste zone above the main reservoir, faults that may reach the seabed and concerns about seal integrity. Independent analyses such as these should form part of a comprehensive assessment of model inputs and risks.
Running failure cases with the models will help to identify and better assess the causes of failure. An ongoing evaluation of subsurface models in conjunction with the monitoring of injection and plume migration can inform model updates and identify which model result is the closest representation of reality. It is important to stress that being objective in this modeling work is paramount, which means setting your biases and desired outcomes aside.
