Modeling Options to Answer Practical Questions for CO2 Sequestration Operations

Abstract

Mitigating climate change requires addressing both the CO2 atmospheric concentration and thus coal dominant share of baseload-electricity production. This necessitates a worldwide ramping up of CO2 capture and sequestration implementation in the next decades. This will come with several challenges. One of them is CO2 sequestration reliability which is impaired by very numerous leakage pathways. Traditional complex numerical tools are useful to provide physical insights of the CO2 behavior. But they are often inappropriate to investigate risk associated with leakage, especially when significant uncertainty leads to the need for computationally intensive probabilistic assessments. A range of models may be developed which become sequentially simpler as more assumptions are applied to the system. The models, and the assumptions, are often associated with specific scales of resolution in the model, both in space and in time. A clear accounting of these models, and their associated assumptions as well as the length and time scales, allows models to be chosen that are best suited to answer the questions being asked. Almost all sequestrations operations will involve three questions about the size of the CO2 and pressure-perturbation footprints, the possibility of leakage of fluids out of the injection formation, and the long-term fate of the injected CO2. These ubiquitous questions are all associated with large space and time scales, and models to answer these questions should be associated with those same scales. This leads to a set of accurate simplified models that can provide meaningful answers to all three of these questions, with a broader multi-scale framework proposed that can accommodate other scales of importance and can combine different models into a so-called hybrid modeling approach. Examples include simplified and CPU-efficient analytical and semi-analytical models for multiple formation (10+) with multiple wells (1000+) within the context of many thousands of simulations required for a probabilistic of leakage risk along old wells; to more complex vertically-integrated numerical models solving injection and post-injection CO2 and brine migration, incorporating heterogeneity and complex topography. Such a multi-scale hybrid modeling approach represents a very promising direction that has evolved in the geological sequestration field since the SRCCS publication in 2005, and it provides a broad platform for on-going and future work across a wide range of modeling approaches.