Faruk Omer Alpak, Shell International Exploration and Production Inc.; Vivek Jain, Shell India Markets Pvt. Ltd.
Carbon dioxide (CO₂) capture and storage (CCS) through geological sequestration technology involves durable removal of CO₂ from the atmosphere by technological, biological, geological, or other means, recycle for further usage, and injecting into deep underground geological formations for permanent storage or for additional recovery of hydrocarbons. The continued need for fossil fuels across the world suggests that the amount of emitted undesirable greenhouse gases will remain on the increase. CCS offers an integrated solution for meeting climate change targets by facilitating the net removal of CO₂ from atmosphere, decarbonize various industrial sectors, recycle and reuse CO₂ for marketable products (including hydrogen fuel), enhanced oil recovery, and safe sequestration of the transported CO₂.
Most of the ongoing/planned CO₂ sequestration projects take place in two types of subsurface settings: (1) Depleted oil/gas reservoirs with a structural closure, i.e., a proven trap with legacy wells: CO₂ plume is not allowed to reach the spillpoint in this setting. (2) Saline aquifers with few legacy wells, which are typically not structurally closed: Capillarity (residual gas saturation), dissolution in water and mineralization stabilize the plume in this setting, but potential conduits in the caprock such as conductive fault/fracture zones are more uncertain requiring probabilistic static and dynamic modeling. Besides CO₂ containment threats (driven by CO₂ plume location) there are also pressure-driven threats such as fault-reactivation-induced seismicity and leakage of saline brine to groundwater/surface through wells/faults. These constraints drive model-based field-development activities during project design and constitute the focus of monitoring and forecasting activities with calibrated models during operations.
Optimal well-placement under subsurface uncertainties is thus a crucial factor for CO₂ sequestration project design to balance multiple competing objectives. We have developed a versatile computational framework for subsurface field-development optimization (SFDO) under uncertainty. SFDO framework utilizes state-of-the-art compositional flow simulators as forward model(s) to evaluate objective function(s) that steer the optimization to desirable manifolds in the uncertainty space and identify alternative solutions that balance multiple potentially conflicting objectives. SFDO combines high-performance parallel computing with state-of-the-art optimization and machine learning algorithms to examine 1000s of well configurations effectively and proposes multiple optimal configurations for consideration. In this work, we demonstrate the application of SFDO on a set of realistic CO₂ storage models. Results demonstrate that the storage efficiency can be significantly improved by optimally placing the injection wells in the domain while keeping pressure-driven threats under control and honoring geological and engineering constraints.