It is now widely believed that human CO₂ emissions have thrown the natural carbon cycle, where CO₂ resulting from all natural and anthropogenic carbon sources is taken up by CO₂ sinks, out of balance. As a consequence, a large fraction of the approximately 39 Gt of CO₂ that are emitted on an annual basis resides in the atmosphere (Friedlingstein et al., 2022). A recent report published by the Intergovernmental Panel on Climate Change (IPCC) concluded unequivocally that human influence resulted in unprecedented climate warming, and that the warming is driven by greenhouse gas emissions (IPCC, 2021). Figure 1A illustrates the observed and modeled warming since 1850 and clearly demonstrates the temperature response to human and natural factors. Figure 1B shows the perturbation of the global carbon cycle caused by anthropogenic activities resulting in the addition of about 19 GtCO₂ to the atmosphere per year (Friedlingstein et al., 2022).

This article provides a current snapshot of geologic carbon storage (GCS) options. What is carbon dioxide (CO₂)? As is commonly known, CO₂ is a colorless, odorless, and incombustible gas at atmospheric pressures and room temperature. CO₂ results from the oxidation of carbon and is produced during respiration by all animals, fungi, and microorganisms. It is used in a variety of industries, for example as a solvent (Valluri et al., 2022), an anesthetic, a food packaging gas (Lee et al., 2016), and a refrigerant (Bansal et al., 2012). Most relevant to this study, CO₂ forms by burning of fossil fuels and it enters the atmosphere as an important greenhouse gas.

First, we summarize the defining characteristics of a successful geological CO₂ storage site. Subsequently, five proven methods of geological CO₂ storage are briefly discussed: (1) Depleted oil and gas reservoirs, (2) Use of CO₂ in enhanced hydrocarbon recovery, (3) Saline formations/aquifers, (4) Coal seams, and (5) In-situ/ex-situ carbon mineralization.

Before examining the geological CO₂ storage site requirements, the elements of a successful petroleum system are presented in a simplified manner. The genesis of a petroleum system begins with the deposition of a source rock rich in organic matter. Afterwards, a reservoir rock with the potential to preserve appropriate porosity and permeability upon burial is deposited. The sedimentation sequence ends with a seal unit, commonly referred to as cap rock, with very low porosity and permeability. Figure 2A captures this sedimentary sequence. The next step in the development is an event that creates a petroleum trap. For example, folding associated with horizontal compression can create anticline shapes, which often act as traps. Sometimes, however, seals are deposited after trap formation or even after partial erosion have taken place. Successive deposition of sediment layers leads to progressively deeper burial of the source rock. Due to the geothermal gradient, at a certain burial depth, temperatures are sufficiently high to convert kerogen in the source rock to hydrocarbons. In order to fill the trap with oil and/or gas, a migration pathway has to be present. Figure 2B illustrates these elements. Lastly, the relative timing of all of the above steps needs to be favorable for a successful petroleum system to form. For example, if burial is not deep enough for maturation to occur, or if a seal rock is deposited after migration, a trap will not be filled because petroleum would not have formed or it would have escaped, respectively.

A successful CO₂ storage site necessitates the same elements as a petroleum system, except a source rock and a migration pathway. In other words, a CO₂ storage site requires a reservoir rock, a seal rock, and a trapping configuration.

Another key difference with hydrocarbon projects is the requirement to characterize and monitor seal performance because seal efficacy is unknown unless it has been geologically proven by the presence of hydrocarbons or some other trapped fluids, for example via pore pressure measurements. In addition to data acquisition at the site characterization stage, long-term monitoring should be put in place to prove that any CO₂ leakage from the sequestration site is zero or within allowable tolerance (Roberts and Stalker, 2020).

Clearly, the existence of petroleum traps around the world demonstrates that buoyant fluids such as oil or gas can be entrapped for millions of years in the subsurface. Besides entrapping petroleum fluids, geologic structures can also trap CO₂. As a result of magma degassing and metamorphism of carbonates, naturally occurring CO₂ fields are present on nearly all continents (Holloway et al., 2005; Miocic et al., 2016). Some of these fields have stored CO₂ for up to 70 million years (Stevens and Tye, 2007), proving that CO₂ storage in geologic formations is not only technically viable but it is in fact a natural process. Several natural CO₂ fields are commercially exploited, for instance, to enable CO₂-enhanced oil recovery (EOR) in petroleum fields. For example, the McElmo Dome in Colorado, United States, contains approximately 1.6 billion tons of nearly pure CO₂ (96%–98%) and produces about 15 million tons per year from 41 wells (Zabel, 1955; Thomaidis, 1978; Gerling, 1983; Cappa and Rice, 1995; Stevens et al., 2005). CO₂ from McElmo Dome has been re-injected for EOR in the Permian Basin, Texas, for more than 30 years (Holloway et al., 2005; Stevens et al., 2005). McElmo Dome is generally viewed as the best analog for geologic CO₂ storage sites for depleted oil and gas fields (Stevens et al., 2005; Stevens and Tye, 2007).

CO₂ at McElmo Dome is stored in a 75–90 m thick lower Carboniferous carbonate reservoir at about 2,300 m depth. The overlying caprock consists of 400 m thick halite and there is no evidence of CO₂ leakage into overlying strata (Stevens et al., 2005). Several faults perturb the reservoir, but they do not extend across the caprock. Figure 3 illustrates the geological key aspects of McElmo Dome. Although McElmo Dome may be the most commonly cited example of a natural CO₂ field, there are several others that are well documented, such as the St. Johns field, Arizona/New Mexico (Holloway et al., 2005), and the Bravo Dome in New Mexico (Sathaye et al., 2014).

Furthermore, natural gas has been stored in depleted oil and gas reservoirs (e.g., Stuart, 1991), saline formations, and salt caverns for almost 100 years and in many parts of the world to provide for peak demand and to balance seasonal fluctuations in gas supply (IPCC, 2005). The great majority of storage projects are safe and successful, though a few have leaked gas, due to well integrity issues or leaky faults (IPCC, 2005). Gas storage projects are typically more complex than CO₂ storage projects because of the highly flammable and explosive properties of gas, and there are highly variable but sometimes frequent cycles of injection and extraction. CO₂ leakage risk is high if concentrations are high (Holloway et al., 2007), or if pressures are high, for instance in pipeline and injection systems (Huang et al., 2014). Also, a number of these facilities, such as the Berlin Natural Gas Storage Facility (IPCC, 2005), are in areas of high population density, which would not be the case for a remote onshore or an offshore CO₂ sequestration site. Therefore, in addition to natural CO₂ fields, existing natural gas storage facilities in geologic formations offer excellent opportunities to study and design CO₂ sequestration sites. In addition, much can be learned about seal capacity and storage potential from ongoing research devoted to hydrogen storage in salt caverns and porous formations (e.g., Matos et al., 2019; Lankof et al., 2022). In the following section, five methods of geologic carbon storage are discussed briefly.

Five methods of geologic carbon storage have been shown to be technically viable, and some are employed as pilot, demonstration, or even full-scale projects (the majority of active projects in the 25 years to date are reviewed by Zhang et al., 2022): (1) storage in depleted oil and gas fields, (2) use of CO₂ in enhanced hydrocarbon recovery, (3) storage in saline formations/aquifers, (4) injection into deep unmineable coal seams, and (5) in-situ/ex-situ carbon mineralization. Each of these methods is briefly described below.

The principal purpose of structural geological and geomechanical characterization of potential CO₂ storage sites is to ensure hydraulic integrity of the geological formations into which CO₂ is injected and the ones around it (Hawkes et al., 2005). This can be achieved by describing the structural geometry, for example through identifying potential leak points, spill points, as well as potential pathways along which CO₂ may escape to undesired formations and/or the atmosphere, and establishing the hydrodynamic conditions in the reservoir (Larkin, 2010). In addition, CO₂ injection increases the pore pressure in the reservoir, which alters the state of stress in the near and far fields, which can trigger fault reactivation, create new fractures, and thus induce seismicity (Burghardt, 2017).

A useful way to help illustrate the concept of structural geometry for laypersons is to conduct a thought experiment with a hypothetical marble run (Figure 9). Imagine placing a marble on the top ramp and trying to predict which path the marble will take until it ultimately ends up on the bottom ramp (Figure 9A). A special condition in this marble run is a thin paper membrane on the third ramp (from top) that cannot hold the weight of the marble. The marble will break the membrane and drop down to the next lower ramp. This represents a transmissibility barrier. Now turn the marble run upside down (Figure 9B). Repeat the thought experiment and imagine that the marble is buoyant so that it will move upwards. The upside-down marble run is an excellent analog for petroleum and CO₂ trap analysis. Now imagine that the marble represents a buoyant hydrocarbon or CO₂ fluid molecule. Forecasting the upward path of the marble is akin to predicting the migration pathway of a fluid in the subsurface. The migration pathway can be along permeable sandstone or limestone units that are separated by impermeable shale or evaporite seal units (analogous to the black ramps in this experiment). Faults may juxtapose either sandstone/limestone on seal creating a sealing configuration, or sandstone/limestone on itself or another sandstone/limestone unit, creating a leaking configuration. The paper membrane in Figure 9B may represent a leak point in the subsurface shaped by fault juxtaposition. Upward flow will continue until an ultimate geometric trapping configuration is encountered.

A number of potential leak points can exist in geological CO₂ storage sites (Figure 10). For example, faults can juxtapose reservoir on reservoir or act as fluid conduits themselves, hydraulic fractures can affect and compromise the seal above a storage reservoir, and there may be cement issues in a well connecting two adjacent or stacked reservoirs (Hawkes et al., 2005). These and similar issues can cause inadvertent CO₂ migration from an intended storage reservoir to an unintended porous formation, such as an aquifer, and ultimately to the surface and into the atmosphere.

When CO₂ and formation water coexist as separate phases during structural, stratigraphic, or residual trapping in a geologic formation (Figure 8), capillary processes determine the column height of the CO₂ body (Ringrose, 2020). The physics of capillary trapping is well understood and the height of a gas column that can be held against gravity by the capillary entry pressure of the seal rock can be calculated according to (Berg, 1975): z g = 2 γ ⁡ cos ⁡ θ 1 r c a p − 1 r r e s g ρ w − ρ g (6) where r _cap and r _res are the pore throat radii in the cap rock and reservoir, respectively, γ is the interfacial tension, θ is the fluid contact angle, and ρ _w and ρ _g are the water and gas densities, respectively. CO₂ column height is one important input for determining the CO₂ storage volume of a potential site. As can be seen from Eq. 6, a large difference between the pore throat radii of the cap rock and reservoir and a small density contrast between water and gas yield a large gas column, whereas a small difference in pore throat radii and a large density contrast achieve a small column.

The risk for fault reactivation can be assessed using simple Mohr-Coulomb theory. The slip tendency T _s of an existing fault in the subsurface can be calculated if the shear and normal stresses on the fault plane, τ and σ _n , as well as the pore pressure, P _f , are known, according to (Streit and Hillis, 2004): T s = τ σ n − P f (7)

Eqs. 6, 7 are useful for determining CO₂ column height and fault slip tendency for screening purposes and to help identify and rank the most viable and lowest risk candidates among a portfolio of potential storage sites.

It is important to understand that the in-situ stress state exerts a first-order control on the geomechanical behavior of rocks during CO₂ injection. While the approaches summarized above provide first-order answers, they are limited to one-dimensional analysis and assume spatially homogeneous as well as uniform rock properties and behaviors. Quasi-one-dimensional simplifications, as often used in the petroleum industry, are inaccurate (Burghardt, 2017). Stress is a three-dimensional tensor, and thus characterizing a potential CO₂ storage site more comprehensively and accurately requires three-dimensional geomechanical modeling. Three-dimensional geomechanical modeling provides the ability to incorporate fault plane orientation, in-situ stress orientations and magnitudes, pore pressure, spatial variation in reservoir and seal parameters, and wellbores in the investigation. Seamless integration of geomechanical modeling with reservoir modeling and simulation is highly desirable to capture the production history as well as the spatial distribution of porosity and permeability.

Inventorying, screening, and selecting a site in the first place is a portfolio development and optimization process akin to any other kind of resource exploration. However, carbon sequestration is not a mature line of subsurface engineering and there are few documented cases where a full cycle of exploration through delineation and project execution has been achieved. There have been a number of portfolio development cases (Bentham et al., 2014; Gruson et al., 2015). Although free from the constraints of hydrocarbon source kitchen and migration pathways that are considerations in hydrocarbon portfolio development, there are several additional technical criteria relevant to CO₂ site portfolios. These criteria include acceptable depth range, proximity to other subsurface management projects such as hydrocarbon production, potable water aquifers, civil infrastructure, the CO₂ sources, as well as potential reservoir size, effectiveness and uncertainty. Various criteria and screening methodologies are available in guideline documents (ISO, 2017; NETL, 2017; Frailey et al., 2018) but these ultimately have to be tailored to local geopolitical and commercial requirements.

Large-scale CO₂ sequestration projects carry significant pressure management requirements due to the volumes in question, especially in the option of CO₂ solution injection (Bryant, 2013; Pool et al., 2013). These can be ameliorated to some degree by choosing to inject into depleted hydrocarbon fields (Jenkins et al., 2012), areas of natural underpressure such as actively deglaciating regions (Bekele and Rostron, 2003), or unconfined aquifers (Sigfusson et al., 2015). The storage volume potential of a prospective CO₂ sequestration site is difficult to estimate due to the absence of storage efficiency information in large-scale projects, and various estimating methodologies are available (De Silva and Ranjith, 2012).

One important aspect of carbon management to mitigate climate change is geologic carbon storage. Five methods of GCS have been validated through several demonstration and pilot projects around the world: (1) storage in depleted oil and gas fields, (2) use of CO₂ in enhanced hydrocarbon recovery, (3) storage in saline formations/aquifers, (4) injection into deep unmineable coal seams, and (5) in-situ/ex-situ carbon mineralization. The role of geoscience in carbon storage lies in bringing forward and characterizing prospective CO₂ storage sites. Once a site has been identified, basic subsurface work inevitably entails: (1) reservoir characterization to estimate pore volume and flow behavior, (2) describing the structural geometry to flag potential leak zones, spill points, and fluid pathways along which CO₂ may escape from the intended storage formation, and (3) seal characterization and reservoir pressure to ensure hydraulic integrity.

Selection of any of these five methods requires storage site screening and feasibility studies to ensure project viability. Elaborating on these efforts is beyond the scope of this article. It is useful, however, to distinguish between CO₂ storage options from a site selection and project management perspective as well as considering trapping mechanisms, which are subject to naturally occurring fundamental chemical and geomechanical processes. Once a storage site has been selected for one of the five options we summarized, multiple trapping mechanisms may occur simultaneously or sequentially over time (Figure 8). For example, carbon storage in a saline aquifer or a depleted hydrocarbon field typically begins by injecting CO₂ as a supercritical fluid into a suitable subsurface containment. Initially, structural and stratigraphic trapping of CO₂ as a single-phase fluid occurs as the dominant trapping mechanism. Over time, residual and solubility trapping start contributing as well, and after hundreds to thousands of years, mineral trapping may set in, permanently locking CO₂ in the subsurface by producing new carbon minerals. Not surprisingly, some trapping mechanisms can occur during multiple storage methods as exemplified by in-situ carbon mineralization/mineral trapping. However, practical and operational factors during project execution would obviously favor in-situ mineralization in suitably mafic rocks and environments where reaction rates are fast (on the order of years) rather than relying on mineral trapping as a minor contributor in mainly subsurface CO₂ fluid storage with considerable uncertainties regarding the timing and amounts of in-situ carbon mineralization.

Another way to think about temporal transitions amongst different carbon storage methods is to conduct a simple thought experiment and track a single CO₂ molecule which transitions from one storage option to another. For example, a CO₂ molecule may start off in an enhanced oil or gas recovery operation. At the end of the field life, the CO₂ molecule is trapped in a depleted oil field. At some point, it will dissolve in the formation brine and ultimately, depending on the availability of Fe and Mg cations, even react to form carbonate minerals. This thought experiment demonstrates that the five storage options summarized above are not necessarily fully separate and that transitions amongst them are likely to occur.