Geological sequestration is a necessary complement to direct air capture of carbon dioxide (CO2) (see Chapter 5) and bioenergy with carbon capture and sequestration (Chapter 4). After CO2 is captured, it is compressed into a supercritical fluid, then injected down a well into a geologic formation that is deep enough for the CO2 to remain as a supercritical fluid, typically 1 km or more. Compression of the gas to a supercritical fluid allows more CO2 to be sequestered. This is due to the high-density of the fluid (~600 kg/m3) relative to gaseous CO2 and the reduced buoyancy forces in water-filled geological formations, although the system maintains a strong buoyant drive between CO2 and brine (Benson et al., 2005).
Suitable geological formations for storing CO2 comprise a porous and permeable reservoir rock, overlain by an impermeable rock (Figure 7.1). Prospective reservoir rocks include sandstone, limestone, dolomite, or mixtures of these rock types. Because the supercritical CO2 is less dense than the fluids that initially fill the pore spaces in the rocks, it will rise by buoyancy forces through the reservoir rocks until it encounters a low permeability rock, typically called a reservoir seal. Seals are composed of shale, anhydrite, or low permeability carbonate rocks. Once trapped below the seal the CO2 is expected to remain sequestered permanently unless the CO2 encounters a permeable fault or fracture in the seal or a leaky wellbore.
Reservoir rocks are characterized by rock type and by whether the pore spaces in the rocks contain salty water or are filled with oil or gas. Oil and gas reservoirs are further characterized by whether the injected CO2 is also used for increasing oil or gas recovery, a process referred to as CO2 enhanced oil or gas recovery. Coalbeds have also been investigated for CO2 sequestration, but technical challenges have precluded large-scale testing or application (Gale and Freund, 2001; Shi and Durucan, 2005). More recently, hydraulically fractured shale formations have also been suggested as an option for sequestration, with or without enhanced hydrocarbon recovery (Tao and
Sequestration of supercritical CO2 in the pore spaces of sedimentary rocks is the most mature of the options available today for reliable storage. Several comprehensive reviews and research roadmaps lay out the scientific and engineering basis for secure sequestration; discuss site characterization and selection; describe effective monitoring and risk assessment and management approaches; and estimate global sequestration capacity and costs (Bachu, 2015; de Coninck and Benson, 2014; DOE, 2017d, g; Rubin et al., 2015; U.S. Geological Survey, 2013). More importantly, nearly a half-century of experience with CO2 injection for enhanced oil recovery (EOR), and nearly two decades of commercial experience with saline aquifer sequestration are available today (Furre et al., 2017; Hansen et al., 2013; IEA, 2016; Koottungal, 2014; Wilson and Monea, 2004). Finally, a large number of pilot-scale experiments have been carried out to test various aspects of sequestration, including methods for monitoring the fate and transport of CO2 in the surface and leakage (Ajo-Franklin et al., 2013; Greenberg et al., 2017; Hovorka et al., 2006; Ivandic et al., 2015; Jenkins et al., 2012; Mito et al., 2008; Rodosta et al., 2017; Spangler et al., 2010).
As a result of extensive research and experience, scientists now have a good idea of the global and regional distribution of capacity to sequester CO2, integrity, risks, and costs of geologic sequestration. Saline aquifer and CO2-EOR projects are currently sequestering about 31.5 Mt/y of anthropogenic CO2 (GCCSI, 2017). A technical potential of 2,000 Gt CO2 is likely for the coming century, enough to make a substantive contribution to greenhouse gas mitigation strategies (IPCC, 2005). The 2,000 Gt technical potential is large compared to 125 Gt captured and sequestered from fossil fuels and industrial emissions required by 2100 to meet the 2°C target of sequestration (IEA, 2014). This sequestration capacity may not be co-located with all large sources of CO2 emissions, thus requiring either large scale and long-distance CO2 transport by pipelines or ships for some of the emission sources. Carbon capture and sequestration is expected to contribute about 14 percent of the emissions reductions needed to stabilize the climate at 2oC warming (IPCC, 2005). The primary impediments to scaling up to that level concern geology (e.g., site characterization), regulations (e.g., difficult permitting), as well as real and perceived risks (e.g., induced seismicity, leaks). This chapter summarizes what has been learned about geologic sequestration and discusses research needed to ensure secure and reliable sequestration of gigatons of anthropogenic CO2 per year.
BACKGROUND: REQUIREMENTS FOR SECURE AND RELIABLE SEQUESTRATION
Geological Formations Suitable for Sequestration
There are two basic requirements for secure sequestration in sedimentary formations. The first is a thick reservoir, typically a sandstone or carbonate, with sufficient porosity to sequester large volumes of CO2 (~50-100 Mt/project) and sufficient permeability to accommodate injection at commercially meaningful rates (~Mt/year). The second is a seal, typically composed of shale, with high enough capillary entry pressure and low enough permeability to retain the CO2 over geological time periods. Beyond this are site-specific requirements regarding the absence of permeable faults and fractures penetrating the seal, a known and ideally low number of existing wells that could provide leakage pathways, favorable geomechanical conditions to avoid fracturing the reservoir or seal during injection, suitable conditions for monitoring, low likelihood of affecting groundwater, and compatibility with existing land and resource use.
Sequestration security may also be enhanced by secondary trapping mechanisms that act over time to reduce the risk of leakage of CO2 out of the storage reservoir. These mechanisms include solubility trapping (dissolution of CO2 into the brine), residual gas trapping (immobilization by capillary forces in the post-injection period), and
mineralization through geochemical interactions between the CO2, brine, and rock (Emami-Meybodi et al., 2015; Krevor et al., 2015; Talman, 2015; Zhang and DePaolo, 2017). The relative importance of these secondary trapping mechanisms is highly site specific (Figure 7.2) and should be assessed using advanced multiphysics, multiscale numerical simulation models. For example, in a closed structural trap (left panel), solubility trapping is slow and capillary trapping is minimal because the CO2 does not move far. In a hydrodynamic trap, such as a large saline aquifer (middle panel), the CO2 may migrate a substantial distance horizontally. In this situation, solubility trapping is more rapid, capillary trapping can be extensive, and mineralization can be enhanced because the CO2 is exposed to a large volume of water-filled rock.
In some cases, the combination of the secondary processes may be sufficient to entirely mitigate any future risks within decades to centuries after the injection is stopped, whereas in other cases, CO2 will still remain potentially mobile for thousands of years. Engineering design of a sequestration project can accelerate trapping through optimal placement of injection wells, co-injection or sequential injection of water, time-varying injection rates, and potentially other approaches yet to be developed (Cameron and Durlofsky, 2012; Ide et al., 2007; Pawar et al., 2015).
Modeling and Simulation
Robust numerical modeling of CO2 plume migration, pressure buildup, geomechanical effects, and geochemical reactions are required for the design, optimization, and performance confirmation of sequestration projects. Finite difference, finite volume,
or finite element approaches for discretizing the subsurface are used to solve a set of coupled nonlinear partial differential equations that describe the processes occurring in the subsurface (DOE, 2017b). Both proprietary and publicly available codes are available for simulating subsurface processes. Intercomparison studies show that calculations from mature codes using similar spatial and temporal discretization, thermodynamic models for fluid properties, and physical processes agree quite well with each other (Class et al., 2009; Pruess et al., 2004). In practice, however, due to the large spatial domains, long time frames, multiphysics nature of these problems, modelers use a variety of approaches for making these simulations more tractable, including using reduced physics models, upscaling, and nonconverged discretizations (Nordbotten et al., 2013). Intercomparisons that provide the flexiblity to use different approaches for solving even simple problems are found to provide large differences in important model outputs such as plume extent, plume center, plume spread, and CO2 phase distribution, leading to the conclusion that better modeling tools are needed (Nordbotten et al., 2012). Rapid growth in computational power combined with advanced algorithms for solving large sets of coupled nonlinear equations provide the opportunity to make rapid advances in several areas that will support the scale-up to Gt/y sequestration.
In particular, important physical and chemical processes span spatial scales from nanometers to kilometers, and temporal scales from milliseconds to millennia (Figure 7.3). Consequently, spatial and temporal averaging is required to make the simulations tractable. Volume-averaged properties typically rely on empirical parameterizations that are obtained from laboratory experiments conducted on small samples over short periods of time under conditions which may or may not be representative of actual conditions during the sequestration project. For example, relative permeability is used to parameterize how multiple fluids phases (e.g., CO2 and water) occupy and flow through the rocks. For practical reasons, the flow rate at which these measurements are made is often much higher than the slow flow rates actually experienced during a sequestration project. Computationally intensive and advanced pore-network models and direct solution of the Navier Stokes equations in realistic pore geometries are now being used to bridge these scales, but much remains to be done (Abu-Al-Saud et al., 2017; Raeini et al., 2018). Similar considerations apply to reaction constants for geochemical processes that occur in the reservoir (Zhang and DePaolo, 2017) or parameterization of the onset of convective dissolution in saline formations (Riaz et al., 2006). To reconcile the challenges, multiscale, multiphysics modeling is needed to bridge the gap between scales to accurately parameterize these models. The supercomputers and experimental tools (e.g., U.S. Department of Energy [DOE]
light sources) needed to accomplish this are improving quickly, and rapid progress can be expected with sufficient support.
The lack of a sufficient description of the subsurface geology is also a major challenge for modeling subsurface processes. Information about the subsurface is available from drill cuttings, geophysical well logs, seismic surveys, and a variety of additional techniques. However, a complete high-resolution model of the subsurface is available only at the wellbores themselves. Everything between the wellbores should be inferred indirectly through seismic or other types of geophysical imaging. To deal with the uncertainty associated with the lack of complete information, probabilistic geostatistical methods are used to characterize the subsurface (e.g., Caers and Zhang, 2004; Kitanidis, 1997). Stochastic simulations are used to obtain probabilistic estimates of plume migration, trapping fractions, and sequestration capacity.
It is important to point out that alternative approaches for predicting plume migration have been developed, so-called reduced physics models, that retain many of the critical processes but require less data and are much less computationally intensive (MacMinn et al., 2012; Nordbotten et al., 2005; Szulczewski et al., 2012). While not a substitute for complex models that include a realistic representation of the geological
setting, they provide useful insights and preliminary estimates of how far the CO2 will migrate from the injection well, residual trapping, and solubility trapping. The models have also been used to develop dynamic capacity estimates for several basins in the United States (Szulczewski et al., 2012).
Finally, in practice today, simulation models are used in concert with monitoring data in an iterative fashion, which through the process of history-matching allows calibration of the model. Calibrated models are shown to perform well for geological sequestration projects, at least over decades during the injection period as has been shown from the existing large projects and dozens of pilot projects.
To ensure that sequestration projects are safe and effective, it is necessary to track the location of the plume of sequestered CO2, measure the pressure buildup in and above the storage reservoir, confirm that the injection wells or other wells penetrating the storage formation are not leaking, and look for leakage into groundwater. Requirements or guidelines for monitoring are a key part of government regulations for CO2 sequestration projects (e.g., EPA, 2010). Numerous pilot tests and commercial operations have demonstrated a wide range of monitoring techniques. Seismic imaging is the most commonly used monitoring method for tracking the location of the CO2 plume (Ajo-Franklin et al., 2013; Hovorka et al., 2006; Ivanova et al., 2012; Jenkins et al., 2015; Pevzner et al., 2011; White, 2013). An example of the application of seismic imaging for tracking the location of the CO2 plume is shown in Figure 7.4. The data demonstrate that CO2 is trapped beneath the seal and that several intra-reservoir shale layers act as baffles to trap CO2 deep in the reservoir. The map-view image shows that CO2 is traveling underneath the seal along a north-south axis that coincides with the topography of the seal.
Pressure buildup can be measured in the sequestration reservoir at the injection wells and the monitoring wells, and in aquifers above the storage reservoir. Pressure data from the injection wells is used to ensure that the pressure does not increase to levels that would hydraulically fracture the seal (EPA, 2010). Pressure data from the monitoring wells can be used to assess the extent to which the pressure buildup extends throughout the storage reservoir, which is important for understanding the “area of review”1 and for predicting how multiple sequestration projects in the same reservoir might interact. Vertically distributed pressure sensors can also be used to
1 The region surrounding the geologic sequestration project where underground sources of drinking water may be endangered by the injection activity. The area of review is delineated using computational
track migration of the plume (Strandli and Benson, 2013; Strandli et al., 2014). Pressure changes measured above the storage reservoir are a highly sensitive indicator of leakage into overlying aquifers (Kim and Hosseini, 2014; Meckel et al., 2013).
Geomechanical responses associated with the pressure buildup caused by CO2 injection have been successfully monitored using interferometric synthetic aperture radar (InSAR) satellite images to detect land surface deformation. This technique was first used at the In Salah Field in Algeria, where cm-scale uplift was detected and used to infer the presence of a fault in the seal (Vasco et al., 2010). Application of this technique is best suited to regions where the land surface cover does not change much over the year.
Over the past decade, evidence has linked seismic events to underground injection of oilfield brine disposal and, to a lesser extent, hydraulic fracturing of shales for oil and gas recovery (Ellsworth, 2013; Langenbruch and Zoback, 2016; Walsh and Zoback, 2015). Most of these events are not felt at the surface but can be detected either with surface arrays of geophones or with buried geophones. Seismic events induced by pressure increases due to CO2 injection need to be monitored to provide assurance that they do not present a hazard to structures, people, or the integrity of the reservoir seal. Table 7.1 summarizes information about microseismic events associated with geological sequestration and EOR projects. It is important to note that none of the
modeling that accounts for the physical and chemical properties of all phases of the injected CO2 stream and displaced fluids, and is based on available site characterization, monitoring, and operational data (EPA, 2013).
events has been felt at the surface and require sensitive instrumentation to detect and locate them. As shown in the table, most microseismic events are smaller than M2 and are measured most reliably with borehole arrays of geophones. Multiple geophones are needed to locate the origin of the microseismic event.
Monitoring methods are also available for detecting leakage of CO2 to the surface. Many of these techniques are borrowed from ecosystem sciences that study carbon cycling in terrestrial environments such as eddy-covariance towers (Lewicki et al., 2010), flux accumulation chambers, cavity ring-down spectrometers for measuring 13C and 12C isotopes from mobile or stationary platforms (Krevor et al., 2010), open-path lidar detection systems, soil-gas sampling (Fessenden et al., 2010) and hyperspectral imaging of vegetative stress from airborne and stationary platforms (Male et al., 2010; Rouse et al., 2010). In addition, infrared detectors are widely available for detecting and measuring CO2 in air and are commonly used for ensuring safe operations around CO2 sources. The ZERT experiment in Montana showed that leaks of 100 kg/day distributed over an area of 500 m2 could be detected with a variety of methods described above (Spangler et al., 2010).
Groundwater monitoring can also be used to detect leaking CO2. Field experiments have employed several approaches for monitoring CO2 directly, or indirectly in
TABLE 7.1 Microseismic Events Measured at Sites with CO2 Injection for Sequestration or EOR (from).
|Aneth (USA)||CO2 EOR||Borehole microseismic||Magnitudes: M 1.2 to M0.8 Frequency: 3800 events over 1 year. Two fault-like clusters|
|Cogdell (USA)||CO2 EOR||Regional network||One M4.4 event and 18 M3+ events over a 6 year period. No major seismicity at nearby, similar operations|
|Weyburn (Canada)||CO2 EOR||2000-3 Mtpa||Borehole microseismic||Magnitudes: M3 to M1. Frequency: 100 events over 7 years Diffuse locations|
|Decatur (USA)||CO2 disposal||2011 2014 1 Mtpa||Borehole microseismic and surface array||Magnitudes: M2 to M1 Frequency: 10,123 events over 1.8 years Multiple fault-like clusters|
|In Selah (Algeria)||CO2 disposal||2004-1 Mtpa||Shallow borehole microseismic||Magnitudes: M to M1.7 Frequency: 10,000 events over 1 year Indications of fracture stimulation|
|QUEST (Canada)||CO2 disposal||2015-1 Mtpa||Borehole microseismic array||<100 microseismic events from a localized source region in the basement|
SOURCE: DOE, 2017c.
groundwater from reaction products (Anderson et al., 2017; Hovorka et al., 2006; Jenkins et al., 2012; Romanak et al., 2012; Yang et al., 2013). New sampling methods have been developed to get pressurized samples that are representative of the subsurface conditions (Freifeld et al., 2005). Tracers have also been used to track movement of water and/or dissolved CO2, including perflurocarbons and fluorescein (Kharaka et al., 2009; Ringrose et al., 2009; Würdemann et al., 2010). Because of the high-cost and labor-intensive nature of fluid sampling and analysis, geochemical analyses have been used primarily as research tools.
Expectations and Requirements for Reliable Sequestration in Sedimentary Rocks
Experience with CO2 sequestration in sedimentary rocks has been consistent with expert opinion first synthesized in the Intergovernmental Panel on Climate Change Special Report on Carbon Dioxide Capture and Storage (IPCC, 2005, p. 12):
With appropriate site selection informed by available subsurface information, a monitoring program to detect problems, a regulatory system, and the appropriate use of remediation methods to stop or control CO2 releases if they arise, the local health, safety and environment risks of geological sequestration would be comparable to risks of current activities such as natural gas storage, EOR, and deep underground disposal of acid gas.
As summarized in Table 7.2, underpinning this statement are a number requirements that should be met to assure secure and reliable sequestration (Benson et al., 2005, 2012; de Coninck and Benson, 2014).
EXPERIENCE WITH SEQUESTRATION IN DEEP SEDIMENTARY ROCKS
Injection of CO2 into sedimentary rocks began in the 1970s, largely for enhanced oil recovery. However, the Sleipner Saline Aquifer Storage Project set the stage for sequestering anthropogenic CO2 for greenhouse gas mitigation. The Sleipner Project has sequestered approximately 1 million tonnes of CO2 per year since 1996 in an off-shore formation at a depth of 800-1,000 m below sea level (Furre et al., 2017). The project has demonstrated that commercial-scale quantities of CO2 could be securely sequestered in a permeable sandstone formation beneath a low permeability shale seal. It has also provided a wealth of experience and data to advance our understanding about migration of CO2 in the subsurface and the use of seismic imaging to track migration of the plume.
Including the Sleipner Project, there have been five commercial-scale sequestration projects in saline aquifers (Table 7.3). Four are operating successfully today,
TABLE 7.2 Requirements for Secure and Reliable Sequestration of Supercritical CO2 in Sedimentary Rocks
|Fundamental storage and leakage mechanisms||Solid scientific understanding and predictive ability for multiphase flow, interfacial processes, rock-water-CO2 reactions, secondary trapping mechanisms, pressure buildup, and sealing processes. Multiphysics, multiscale modeling of coupled processes on all timescales up to millennia.|
|Site characterization, site selection, and risk assessment.||Geological, hydrogeological, geomechanical, and geochemical characterization. Assessment of sequestration capacity, sealing potential, and environmental risks to groundwater, natural resources, and people.|
|Storage engineering||Design of injection operations, pressure management, plume containment, and acceleration of secondary trapping.|
|Safe operations||Application of best practices and conduct of operations to minimize risks of worker injury, uncontrolled releases of CO2, and fugitive emissions.|
|Monitoring||Monitoring to track migration of the plume, leakage into groundwater, and leakage to the atmosphere; assure well integrity, control pressure buildup, avoid ecosystem impacts, and ensure public safety.|
|Contingency planning and remediation||Actions planned and executed in the event of unintended releases of CO2 to the atmosphere or leakage into groundwater, and to control public hazards.|
|Regulatory oversight||Effective government oversight of projects to ensure due diligence and accountability in all aspects of a CO2 sequestration project.|
|Financial responsibility||Shared risk pool with financial set-asides to ensure the adequacy of financial resources in the event that remediation is needed after projects are shut down.|
|Public engagement and support||Effective communication, consultation, and support from the communities living in the vicinity of CO2 sequestration projects.|
TABLE 7.3 Large-scale Projects to Sequester CO2 in Saline Aquifers
|Sleipner||Offshore Norway||1996-present||1 Mt/y|
|In Salah||Algeria||2004-2010||0.7 Mt/y|
|Snohvit||Offshore Norway||2008-present||1 Mt/y|
|Decatur||Illinois, United States||2011-2014||0.3 Mt/y|
|Quest||Alberta, Canada||2015-present||1.2 Mt/y|
|Gorgon||Western Australia||?||3.4 Mt/y|
SOURCE: GCCSI, 2017.
sequestering a total of 4.2 Mt/y. The Gorgon Project in Northwest Australia will be the largest saline aquifer sequestration project and is expected to sequester from 3-4 Mt/y and to begin operations shortly (Flett et al., 2009). Injectivity has been inadequate at two of the projects: In Salah and Snohvit (Eiken et al., 2011). The In Salah Project was suspended because of the large pressure buildup during CO2 injection and the associated occurrence of unexpected geomechanical deformation (Eiken et al., 2011; Rutqvist et al., 2010; Vasco et al., 2010). Poor injectivity at the Snohvit field was remedied by injecting into a different interval.
In addition, 125 CO2-EOR projects in the United States are injecting Mt/y of CO2 into depleting oil reservoirs (EPA, 2016a). About 21 Mt/y of the CO2 is captured from anthropogenic sources (EPA, 2016a). allowing for its efficient extraction from the reservoir. Some fraction of the injected CO2 is produced back with the CO2, but is immediately separated, recompressed, and injected back underground. Unless the injected CO2 is intentionally removed from the reservoir after the EOR project is completed, nearly all of the CO2 injected into the reservoir over the lifetime of the project will remain underground. In the United States alone, 100 billion barrels of oil are technically recoverable with CO2-EOR (Kuuskraa, 2013), not including residual oil zone production. At the industry-standard ratio of about 3 barrels of oil per ton of CO2, this corresponds to about 30 Gt of CO2 that could be sequestered in depleted oil reservoirs (IEA, 2015a; Kuuskraa, 2013). Using advanced CO2-EOR, which is designed to co-optimize CO2 sequestration and EOR, the ratio of CO2 injected to oil produced could be tripled or more, leading to a sequestration potential of greater than 90 Gt CO2.
Sequestration Capacity and Footprint in Sedimentary Basins
Global sequestration capacity estimates for sedimentary formations range from several thousand to more than 25,000 Gt CO2 (Benson et al., 2005, 2012). Due to the enormity of these estimates, more salient issues are the geographical distribution and co-location of potential sequestration sites with large emission sources as well as the potential rate at which CO2 can be sequestered. If CO2 is captured from concentrated sources, having nearby reservoirs to sequester the CO2 is desirable to avoid the complications and costs of long-distance transport. Thus, investable storage reserves may be a fraction of these overall assessment, and factors such as a transport distance may have implications for which locations are prioritized for development. Figure 7.5 shows the location of sedimentary basins that are highly prospective (dark gray) and prospective (medium gray) for CO2 sequestration. One of the major benefits of direct air capture is that CO2 can be captured near the best sequestration sites to take full advantage of all the available capacity, thus avoiding limitations associated with co-location with large emission sources.
In the United States, both DOE and the U.S. Geological Survey (USGS) have made estimates of the CO2 sequestration resource in sedimentary basins. USGS used a geology-based probabilistic assessment methodology to obtain a mean estimate of approximately 3,000 Gt of subsurface CO2 sequestration capacity that is technically accessible below on-shore areas and state waters (U.S. Geological Survey, 2013). This amount is more than 500 times the 2011 annual U.S. energy-related CO2 emissions of 5.5 Gt. DOE provides a range of 2,600-22,000 Gt CO2 based on methods similar to those used by USGS (DOE, 2015b).
Although much work has been done to produce a globally harmonized assessment of the capacity available for geological sequestration, many uncertainties remain about how much of the pore space identified in these assessments will actually be usable (Bachu, 2015). Pressure buildup and associated risks is one of the major constraints on the sequestration capacity (Birkholzer et al., 2009). Some have argued for this reason that the capacity will be on the low end of this range (Ehlig-Economides and Economides, 2010; Zoback and Gorelick, 2012). The extent to which pressure buildup limits sequestration capacity is captured by the concept of a dynamic capacity that is constrained by the maximum rate of injection that will avoid excessive pressure buildup in a geological formation. Dynamic capacity is site and context specific (e.g., Are others also using the formation for sequestration?) and will depend on whether active pressure management (e.g., brine extraction to offset the pressure buildup caused by CO2 injection) is
implemented (Buscheck et al., 2012). Only when many more commercial-scale projects are implemented will we be able to refine these estimates and gain confidence in the extent to which these sequestration resources can be realized.
The footprint per ton of CO2 for a sequestration project is highly site specific, depending on the architecture of the storage formation and seal, petrophysical properties of the rocks, pressure and temperature of the storage formation, and extent of secondary trapping. For example, footprint estimates for Sleipner are about 1-3 t/m2 (Furre et al., 2017). For thinner storage reservoirs, it could be significantly less. A reasonable range for a wide variety of sequestration sites is about 0.5 to 5.0 t/m2.
Making the decision about whether a particular prospective site is suitable for an actual project requires matching site attributes of the sequestration formation to proposed emissions sources. That is, a sequestration formation with a static capacity of 5 Gt, for example, offers limited guidance in a carbon capture and sequestration (CCS) investment decision. Investors in CCS need confidence that a sequestration reservoir can accommodate, for example, 20 Mt/y for 25 years with an acceptable level of risk around permanence, cost, and license to operate. A range of technical and nontechnical risks and uncertainties serve to limit estimates of investable storage reserves (as compared to static assessments) including the following:
- Containment: Validation of seal and absence of faults;
- Unit Characterization Cost: Access, exploration, and appraisal capability and costs;
- Unit Storage Cost: Mainly well cost and count, which depends on initial injectivity and decline rates; and measuring, monitoring, and verification (MMV)
- End-of-Life Closure and Abandonment: Technical standards, ongoing MMV requirements, and relinquishment of liabilities.
The risks are basin, play, and jurisdictionally independent, and so methods and recommendations for practice may vary depending on the setting.
De-risking and proving-up storage reserves requires traditional exploration, appraisal, and field development planning activities, which are the domain of oil and gas operators and service companies. Such activities include drilling and seismic and extended well tests and require significant time (several years) and costs in excess of $100M to complete a single project. Close cooperation and engagement of the oil and gas industry and service providers will provide valuable guidance and advice for research in this area. That guidance should extend to developing a view on the efficacy of current estimates of geo-sequestration resources/reserves.
COST OF SEQUESTRATION IN DEEP SEDIMENTARY BASINS
Costs for CO2 sequestration in sedimentary rocks have been estimated based on experience with the projects described above, together with model estimates from scenario-based assessments. Estimates range from $1 to $18/tCO2 (2013 dollars; Table 7.4). The most recent estimates from DOE narrow the range to $7 to $13/tCO2 in the United States. The wide range reflects the highly site-specific nature of geologic sequestration projects. Primary variables include the depth of the formation, number of injection wells required, existing land uses, and ease of deploying monitoring programs. Appendix F provides additional information about costs of compression, drilling and completing injection wells, and pipeline transportation.
The cost estimates in Table 7.4 include well drilling, injection, monitoring, maintenance, reporting, land acquisition and permits, and other incidental costs. They do not include costs associated with remediation activities, which may be required in the case of well leakage, groundwater contamination, or management of the risks of induced seismicity with active pressure management (Brunner and Neele, 2017; Kuuskraa, 2009; Zahasky and Benson, 2016). Proper design and operations should avoid these complications; therefore, costs associated with remediation are not included here.
TABLE 7.4 Compilation of All-In Costs for Sequestration in Deep Sedimentary Basins
|Study||Low Estimate (2013$/tCO2)||High Estimate (2013$/tCO2)|
NOTE: These estimates do not include the cost of compression (Appendix F).
SOURCE: Rubin et al., 2015.
REGULATIONS, BEST PRACTICES, AND STANDARDS
In the United States, each year more than 2.5 Gt of brines are injected into deep underground formations for disposal (EPA, 2018). This experience underpins a substantial body of work on technical, administrative, and regulatory approaches for achieving secure and reliable sequestration of supercritical CO2 in sedimentary basins. The United States has developed regulations for CO2 sequestration in saline aquifers, covering issues such as siting, well construction, monitoring, and risk management, particularly with regard to the presence of active or abandoned wells in the so-called area of review (EPA, 2010). Injection into oilfields is regulated under a different set of requirements (EPA, 2018). These regulations are promulgated through the U.S. Environmental Protection Agency’s (EPA’s) Underground Injection Control Program, which is intended to protect freshwater resources.
International guidelines for tracking and reporting greenhouse gas emissions have been developed for inventory accounting of CO2 capture and sequestration projects (IPCC, 2006). In addition, the International Organization for Standards has developed standards for activities related to CO2 capture, transportation, and geologic sequestration, including the design, construction, and operation of the project; monitoring and verification; environmental planning and management; and risk management (ISO/TC 265, 2011). Finally, DOE has developed best practice manuals. covering site screening, selection, and characterization (DOE, 2017d); public monitoring, verification, and accounting (DOE, 2017g); outreach and education (DOE, 2017b); operations (DOE, 2017f); risk management and simulation (DOE, 2017e); and storage formation classification (DOE, 2010). These regulatory approaches and practices are likely to evolve as experience grows with the implementation of new commercial projects.
Although much has been done to develop the legal and regulatory framework for CCS (Dixon et al., 2015), several challenges and unresolved issues remain regarding the regulatory and legal framework for CO2 sequestration in deep geological formations, including the following:
Financial responsibility for long term liability: CO2 sequestered in deep geological formations is expected to persist in a supercritical phase for hundreds to thousands of years or longer. Although the risks of leakage or other environmental harm are expected to decrease over time because of the secondary trapping mechanisms and pressure decreases in the post-project period, the possibility nevertheless remains that some CO2 could leak out of the reservoir. After the project has been shut down, who is responsible for monitoring and remediation, for how long, and with what mechanism? Although proposed as solutions to this issues, bonds, shared risk pools, and insurance have not matured because of the early stage of this technology (e.g. Gerard and Wilson, 2009). This issue is less salient in parts of the world where the subsurface is “owned” by the government, which will assume responsibility after assurances have been provided that the CO2 is expected to remain trapped. Lack of clarity on this issue is one of the largest barriers to scale-up of CO2 sequestration in deep geological formations (Davies et al., 2013).
Pore space ownership: Who can grant the right to sequester CO2 in subsurface pore space depends on national and sub-national laws defining mineral rights, water rights, surface rights, and other beneficial land uses. Clarification of rights where conflict might exist is needed to expedite scale-up of CO2 sequestration. The large footprints of full-scale CO2 sequestration projects, which can potentially extend over 100 km2 or more, present additional complications. Aggregating access to the subsurface from 10s to hundreds of landowners can be timing consuming and expensive. When landowners desire to use their pore space for sequestration, or if people do not support the project, legal measures used by the oil and gas industry, such as unitization (management of adjacent properties under a common operating regime) may be necessary, and the rules for doing so vary by state.
Regulatory impediments: As expected when new regulatory requirements are promulgated, some requirements are too challenging or too expensive for operators. Examples include overly prescriptive monitoring programs that do not provide the operator with the flexibility to tailor the monitoring program to site-specific attributes; rules for transitioning from a CO2-EOR project to a sequestration project; and state-based requirements that conflict with national requirements.
Over time, none of these issues is insurmountable, as evidenced by practices in more mature mineral and resource extraction activities. However, maintaining a
science-based approach to assessing risks, managing risks, and updating regulations will be essential to support Gt-scale deployment of this technology.
Scaling up global CO2 sequestration in deep sedimentary formations to 5-10 Gt/y CO2e is an enormous task that requires research to ensure its secure and reliable implementation. To put this into perspective, 5-10 Gt/y sequestration in deep geological formations would require more than a 100-fold scale-up from current sequestration operations and would assume the scale of global oil production, which is a $2 trillion/y industry. The enormous sequestration capacity of geological formations combined with the permanent nature of geological storage warrant a significant investment in research and development (R&D). The more than 100 years of oil and gas operations created a sufficient foundation of knowledge to continue expansion of geological sequestration projects in oil and gas reservoirs and saline aquifers. Scale-up would occur gradually with learning-by-doing as a key component of capacity building and knowledge generation. However, if this technology is to expand to the Gt/y CO2 scale and beyond, much more intensive use of our sequestration resources will be required, which depends on better information to assess risks, select sites, and provide assurances of their safety and effectiveness. The following 10-year research agenda is designed to develop the scientific and engineering knowledge needed to assess the conditions under which Gt/y CO2 sequestration in deep geological formations would be possible, as well as best practices for engaging local communities and the general public on geological sequestration.
The key research needs are summarized below, and their costs are listed in Table 7.5. Some of these needs arise from experiences with existing sequestration projects or analogous operations, such as large-scale water injection or natural gas storage. Others arise from our limited ability to predict the outcome of complex coupled multiphysics, multiscale processes on timescales of centuries to millennia. All of these needs are critical to scale sequestration to the Gt/y level.
A large number of new projects is expected to be deployed under the 45Q rule over the coming decade (45Q is a new tax incentive providing $35/t for CO2-EOR and $50/t for saline aquifer sequestration2). This provides an enormous opportunity for the R&D community to partner with industry to gain real-world experience in Mt-scale sequestration. In fact, many of the research needs can only be met through partnership with
2 See http://uscode.house.gov/view.xhtml?req=(title:26%20section:45Q%20edition:prelim) (accessed January 28, 2019).
industry, which facilitate the leveraging of the large investments in site characterization and infrastructure that will take place and the two-way transfer of knowledge in the context of real-world projects.
These research needs, although largely framed in the context of the United States, relate to valuable fundamental knowledge that can be transferred to regions around the world through international collaborations.
Many of these research topics have been supported during the past decade, to one degree or another, by DOE, the Department of the Interior (DOI), EPA, and the National Science Foundation (NSF). Significant progress has been made to support sequestration of emissions resulting from the use of fossil fuel sources in electricity generation and industrial processes. This research agenda is designed either to supplement existing programs or to place a stronger emphasis on certain topics than currently exists. The needs arise when considering prospects for multi-Gt-CO2/y negative emissions over many decades, and perhaps a century or more. Negative emissions with sequestration in deep geological formations presents unique challenges because (1) CO2 may be sequestered in regions lacking large emission sources, that is, regions that have not been the focus of current investigations and (2) the use of deep geological formations on a large scale for the purpose of negative emissions may double or triple (or more) the total amount of geological sequestration. This massive scale-up in geological sequestration will require more intense resource utilization and development of new sequestration resources that may rely on advanced reservoir engineering practices that remain in their infancy, such as accelerated secondary trapping mechanisms and reservoir pressure management. These activities will support net emissions as well as sequestration from fossil fuel sources.
1. Quantifying and managing the risks of induced seismicity. During the past 5 years an unprecedented number of induced seismic events have occurred in regions with historically low rates of seismicity, largely in the midwestern United States. Notably, Oklahoma experienced a rapid increase in the number and magnitude of earthquakes (Walsh and Zoback, 2015). These events have been attributed primarily to the disposal of oilfield brines into saline aquifers (Keranen et al., 2014), which increases the pore pressure and sometimes causes critically stressed faults to slip. The injection of brine into aquifers directly above fractured basement rocks has resulted in the most and largest earthquakes. A small number of induced earthquakes have been attributed to hydraulic fracturing (Ellsworth, 2013). Although most events are very small, some are large enough to be a nuisance, and at worst, capable of property damage and harm to people. Some argue that sequestration in deep saline formations poses similar risk because of the widespread pressure buildup associated with injection
of large volumes of CO2 at high rates (Zoback and Gorelick, 2012). Information and knowledge are not sufficient to distinguish low-risk from high-risk sites, particularly for sequestration in saline aquifer directly over basement rock. Susceptibility to induced seismic events ranges over two orders of magnitude for slip along basement faults, and current knowledge gaps prevent prediction of those areas susceptible to induced seismic events. In addition, while many sites have experienced a large number of micro-seismic (not felt at the surface) events because of CO2 injection, whether these indicate the potential for larger events is uncertain. Research is needed to understand better the mechanisms, risks, and consequences of pressurization-induced slip and associated seismicity along pre-existing fractures in basement rocks. In addition, field testing techniques are needed to quantify risks in advance of sequestration site selection. For example, pressure transients measured during fluid injection are routinely used to measure the permeability of formations. Similar tests could be used to address susceptibility to induced seismic events. This information can be used to develop a methodology for selecting sites with low risk and/or to manage injection rates to limit the potential to induce seismic events.
Earthquakes have been induced at some CO2 sequestration sites, notably the In Salah Project in Algeria and the Decatur Project in Illinois. These events have been small (< ~M 1) and not felt at the surface or caused any damage. Other sequestration projects have not induced any detected seismic events, even when injecting into basal aquifers above a crystalline basement rock. The potential for harmful earthquakes and the extent to which cautionary exclusion of some locations will affect estimates of sequestration capacity need to be evaluated. Specific research directions include the following:
- Understanding why some CO2 sequestration sites experience induced earthquakes and other do not;
- Determining the potential for pre-qualifying sites to avoid induced events using the best available models and data;
- Developing short-term tests to assess the risks of induced earthquakes before committing to a project;
- Understanding the implications of induced seismicity on sequestration capacity estimates and injectivity rates;
- Developing mitigation approaches for minimizing risks, such as brine extraction, injection pressure management, or requiring both a top and bottom seal; and
- Understanding the potential for fault slip to increase leakage from the sequestration reservoir.
TABLE 7.5 Costs and Components for a Geologic Sequestration Research Agenda
|Recommended Research||Estimated Research Budget ($M/y)||Time Frame (years)||Justification|
|Basic Research/Development||Reducing seismic risk||50||10||The proposed budget would allow for 3 experiments in different U.S. regions, each at a cost of about $15M/y for 10 years. The region-specific projects would be supported by $5M/y of model development, laboratory studies, and analysis of new and existing data sets. This research would improve understanding of and reduce the risks of induced seismicity at geological sequestration sites, develop methods for assessing and mitigating risks of seismicity, improve capacity estimates by screening sites that are high risk for induced seismicity, and help quantify the risk of leakage from fault slippage.|
|Improving secondary trapping prediction and methods to accelerate secondary trapping||25||10||This research program would support a 10-y multi-investigator team to perform a large-scale experiment designed to quantify the effectiveness of natural and accelerated trapping for immobilizing CO2 in the post-injection period. The experiment would require a combination of field experiments, multiscale laboratory experiments, numerical modeling, and monitoring. The goals of this coordinated program would be to improve understanding of the coupled, multiscale, multiphysics processes governing secondary trapping, reliably predict and verify their effectiveness, and develop and demonstrate methods to accelerate secondary trapping of CO2.|
|Improving simulation models for performance prediction and confirmation.||10||10||This program would support 2-3 teams of researchers to develop improved simulation models for predicting the fate and transport of CO2 in the subsurface, particularly with regard to the effects of geological heterogeneity, secondary trapping mechanisms, geochemical reactions, geomechanical responses to CO2 injection, and the coupling between them over thousands of years. Simulation models will be built at a hierarchy of relevant scales, from the nano-scale to the basin-scale. Robust approaches for translating between these scales will be developed to ensure that reservoir and larger scale models incorporate accurately the range of relevant physical processes influencing plume migration during injection operations and the post-injection period.|
|Recommended Research||Estimated Research Budget ($M/y)||Time Frame (years)||Justification|
|Development/Demonstration||Increasing the efficiency and accuracy of site characterization and selection||45||10||Partner with industry to develop and test innovative approaches for characterizing greenfield sites, which usually require about $100M to assess the suitability of a site. The program could be carried out by expanding the CarbonSAFE program to include 2 sites with sequestration quantities of 200+ Mt CO2 to assist states and commercial entities in qualifying sites for large-scale deployment (4 projects over a 10-y period). The data collected from this program should be made publicly available through NETL’s EDX platform, USGS, and university data archives, and federated data volumes such as Nat-Carb. This program would develop the knowledge needed to develop and demonstrate the following: efficient and effective methods for characterizing geological sequestration sites over the large footprint of a commercial-scale CO2 sequestration project (~100 km2), methods for identifying and characterizing faults in seals and basement rocks; and methods of characterizing geological heterogeneity and associated trapping of CO2. $5 M/y of the proposed budget would be used to support academic, laboratory, and industrial research developing innovative approaches that could be tested in the above-mentioned field programs.|
|Improving monitoring and lowering costs for monitoring and verification.||50||10||Many new sequestration projects are likely to be developed over the next 10 years as a result of the 45Q rule. These projects provide an ideal opportunity to partner with industry to develop, test, and deploy the next generation of integrated monitoring systems for commercial projects. The proposed research program would provide for 4-6 projects at a cost of $5-10M/y. The collaborative projects would develop and demonstrate approaches to optimize integrated monitoring programs that reduce costs while increasing quality and access to real-time information about the status of stored CO2. In addition to the field experiments, $10 M/y would be used to support fundamental research to develop and test new approaches to quantify mass balances, measure CO2 saturations, and quantify leakage.|
|Developing reservoir engineering approaches for co-optimizing CO2-EOR and sequestration||50||10||Develop and demonstrate reservoir management practices to co-optimize CO2-EOR and CO2 sequestration to achieve negative emissions during oilfield operations. Quantify the extent of negative emissions that can be achieved by co-optimization. Two field-scale experiments in partnership with industry are proposed, each with a budget of $20 M/y for 10 years. $10 M/y will support academic, national laboratory, and industry research to develop new approaches for co-optimization.|
|Deployment||Assessing and managing risk in compromised sequestration systems.||20||10||Improve understanding of the impact of leakage on groundwater systems and the vadose zone. Quantify the extent to which these interactions attenuate CO2 migration and mitigate risks of leakage to the atmosphere.|
|Social sciences research to improve public engagement effectiveness with local communities and the general public||1||10||Establish best practice for community engagement, rules of practice, and regulation guidelines. Provide educational materials for increasing awareness of the need, opportunity, risks, and benefits of geological sequestration for negative emissions.|
2. Increasing the effectiveness of site characterization and selection methods. Site characterization and selection is arguably the single-most important factor for secure and reliable CO2 sequestration in sedimentary rocks, but it poses challenges beyond what is required for oil and gas exploration and production. Research needs associated with the scale-up of commercial sequestration projects include the following:
- Developing and demonstrating efficient and effective methods for characterizing the reservoir and seal. The area requiring characterization and assessment for a 50-100 Mt scale sequestration site can be about 100 km2, which is large compared to even large oil and gas fields. This work should be performed in collaboration with industry, as long as all data will be made available to the public.
- Identifying small faults in the seal that may provide leakage pathways if they are permeable. Such faults are difficult to resolve from seismic imaging. Experience from In Salah and other reservoirs has shown that small faults, often called sub-seismic, are not revealed by conventional analysis. In this case, unexpectedly rapid transport through the reservoir and unusual geomechanical uplift highlighted the presence of fractures. Better methods for detecting faults with small offsets are needed. Improved methods for seismic data acquisition and processing together with field experiments in collaboration with industry are also needed.
- Imaging faults in igneous and metamorphic basement rocks that may create risks for induced seismicity. Such faults are rarely imaged because seismic data sets are optimized to provide the highest possible resolution in overlying resource-bearing sedimentary formations. Moreover, base fault detection is challenged by the near vertical nature of the faults together with the lack of stratigraphic markers that support detection of faults in sedimentary rocks. Reanalysis of existing data sets to identify basement faults, together with acquisition of new data sets that are optimized for basement fault detection, would help to develop the methodology for basement fault detection. The work could be done in collaboration with industry because a large amount of useful data already exists but is currently inaccessible.
- Obtaining reliable information on geological heterogeneity in the reservoir, which has significant effects on secondary trapping mechanisms and may be an important part of the security case for sequestration.
- Developing publicly available data sets that can be used by the research community and private developers interested in sequestration projects.
- Expanding on programs such as CarbonSAFE3 to incorporate highly prospective sequestration sites where bioenergy carbon capture and sequestration (BECCS) and direct air capture are the likely sources of CO2.
3. Improving monitoring and lowering costs for monitoring and verification. Monitoring is critical for determining whether a project is performing as designed. Over the past 20 years, impressive progress has been made to adapt and demonstrate component technologies to track plume migration and detect leakage, land surface deformation, pressure buildup, rock-water-CO2 reactions, and other factors. Important gaps remain, however, in several areas that will be essential to large-scale deployment of sequestration in geological formations:
- Mathematical approaches are needed for co-inversion of multiple data sets coupled to performance prediction models to provide more information than is available from one data set alone. For example, it is very difficult to measure the CO2 saturation distribution in a plume with seismic data alone, but co-inversion of seismic data with electrical resistance tomography data using a 3-D reservoir model may provide detailed information about saturation distributions.
- Current monitoring methods can identify the location of CO2 in the storage reservoir and detect leakage. However, they could be improved with more quantitative information about CO2 saturations and mass balances. This information can calibrate and validate the simulation models that are used for a variety of purposes, including resource optimization, forecasts of plume migration during the post-injection period, and regulatory compliance.
- Strategies and technologies are needed for an adaptive monitoring program that is site-specific and responsive to the changing needs and conditions of the sequestration projects. Fit-for-purpose approaches with sufficient flexibility to respond to the evolving understanding of project’s uncertainties and risks are needed.
- Today, seismic monitoring is the primary method used to track the location of sequestered CO2. Surveys are labor intensive, time consuming, and expensive and therefore conducted at multiyear intervals. Alternative methods to deploy
3 The Carbon Storage Assurance Facility Enterprise (CarbonSAFE) Initiative projects focus on development of geologic storage sites for the storage of 50+ million metric tons (MMT) of CO2 from industrial sources. CarbonSAFE projects will improve understanding of project screening, site selection, characterization, and baseline monitoring, verification, accounting (MVA), and assessment procedures, as well as the information necessary to submit appropriate permits and design injection and monitoring strategies for commercial-scale projects. These efforts will contribute to the development of 50+ MMT storage sites in anticipation of injection by 2026.
- seismic imaging that allows continuous acquisition and real-time analysis in combination with subsurface pressure data can enable rapid leak detection and response.
- Advanced methods for locating and characterizing leaks, should they occur, are needed to guide remediation and compliance efforts.
4. Improving confidence in secondary trapping mechanisms and accelerating their trapping speed. Secondary trapping mechanisms can be viewed as an insurance policy for geological sequestration of CO2. “Secondary” does not imply that these mechanisms are of secondary importance, only that they serve backup if the wells are leaky or the seal does not perform as expected. The benefits of secondary trapping mechanisms, such as solubility trapping, residual gas trapping, and mineral trapping, compensate for flaws in either the seal or in the leakage pathways created by wells penetrating the seal above the sequestration reservoir (see “Geological Requirements for Secure and Reliable Sequestration” above). For many reservoirs, the risks for leakage are low, and therefore secondary trapping mechanisms are not important. For others, secondary trapping may be the factor that ensure the reservoir’s suitability for sequestration. When secondary trapping mechanisms are an important part of sequestrtion security, it will be necessary to improve understanding of how they work and how to accelerate them. Research is needed to improve understanding of the coupled, multiscale, multiphysics processes governing secondary trapping, and to reliably predict and verify their effectiveness. For example, it is not yet possible to accurately simulate convection-driven dissolution of nearly all of the injected CO2 over the 1,000s of years during the post-injection period. Some research simulators are starting to tackle the problem, but commercial simulators have neither sufficiently high spatial resolution nor the advanced numerics needed to simulate this reliably (Riaz et al., 2006). Similarly, the Land trapping model was developed for application in the context of waterflooding oil reservoirs (Land, 1968). Because CO2 is soluble in water, Ostwald Ripening has the potential to redistribute the residually trapped CO2, potentially leading to remobilization of the gas (de Chalendar et al., 2018). This phenomenon is not captured in the Land trapping model or any other trapping model available in commercial simulators today. Similarly, no models can accurately simulate the coupling between solubility trapping and residual gas trapping, changes in wettability caused by mineral precipitation or dissolution and consequent redistribution of the CO2, and the millennial-scale evolution of a plume of CO2 subject to these processes.
In addition, a little-explored area of research is the opportunity to accelerate secondary trapping of CO2. This acceleration would limit the footprint of the CO2 plume, shorten the period over which leaks could occur, reduce the amount of CO2 that would leak in the case of leaking wells and poor-quality seals, and reduce the need
for risk management activities, such as monitoring and maintenance of contingency plans.
This research activity would require a combination of theory, simulation, and laboratory and field experiments.
5. Developing reservoir engineering approaches for co-optimizing CO2-EOR and sequestration. Sequestering CO2 in oil and gas reservoirs has the potential to be a carbon neutral or carbon negative activity, while simultaneously producing hydrocarbons that can be used for applications that are difficult to decarbonize (e.g., air transport). Moreover, oil and gas reservoirs have both a seal that demonstrably retains buoyant gases over long timescales and a reservoir that has been characterized extensively. Current CO2-EOR projects are designed to maximize profits by minimizing the amount of CO2 injected into the oil reservoir for every barrel of oil produced. In a world that values CO2 sequestration, the economic drivers will change to co-optimize revenues from oil production and CO2 sequestration. For such projects to become carbon negative, they must significantly increase the ratio of CO2 injected per barrel of oil recovered. Research is needed to develop reservoir engineering methods to co-optimize CO2 sequestration and enhanced oil recovery because current approaches for CO2-EOR are not likely to efficiently sequester more CO2 simply by increasing the amount of CO2 injection. Increasing the amount of CO2 injection using today’s reservoir engineering approaches for CO2-EOR will only result in more recycling of CO2, driving up costs and decreasing efficiency. Alternative approaches such as gravity stable injection scheme using horizontal injection wells could minimize CO2 recycle and increase oil recovery while increasing the efficiency and amount of sequestration. In addition, conventional CO2-EOR has been optimized for miscible recovery (oil and CO2 become a single phase, thus making it easier to produce the oil), which is only possible for deep, light crude oils. Many oil reservoirs are not suitable for miscible recovery because of the oil composition, temperature, or pressure of the reservoir. but still have significant sequestration capacity. Methods for co-optimization of oils that are not miscible with CO2 are needed. In addition, methods for co-optimizing oil recovery from residual oil and transition zones (i.e., areas with oil saturations that are too low for conventional oil recovery operations such as primary production or water flooding) (Koperna et al., 2006). Opportunities for sequestration and co-optimization also exist in (a) near off-shore formations, which have important infrastructure considerations, (b) stacked formations where one location has several different zones in which CO2 can be sequestered, and (c) by sequestering CO2 either in flanks of the reservoir or below main pay zones for pressure support.
6. Assessing and managing risk in compromised sequestration systems. The potential impacts of CO2 leakage, particularly into freshwater aquifers, must be better understood. Long before CO2 is leaked into the atmosphere, it migrates upward from the sequestration reservoir toward the land surface. Along the way it will interact with the geological system (e.g., rocks, groundwater, and microbiota) and manmade materials (e.g., casing and cement). These interactions may attenuate leakage into the atmosphere, which is beneficial from a climate-change perspective, but may have negative consequences. In particular, leakage of CO2 into aquifers reduces the pH of the fluids and changes the geochemical equilibrium. Under certain conditions at some sites, hazardous elements such as arsenic may be mobilized (Zheng et al., 2009). The risks associated with leakage into the subsurface are highly site specific and localized, and although the likely consequences are improbable and negligible, some risk of substantial impacts remains. Consequences will depend on the amount of leakage, composition of the leaked gases (for oil or gas field sequestration, some hydrocarbons may be carried along with the CO2), and characteristics of the subsurface hydrogeological setting. Research is needed to quantify the likely impacts of leakage for deep, intermediate depth, and shallow aquifers, as well as the vadose zone.
7. Improving simulation models for performance prediction and confirmation. Simulating the multiscale, multiphysics, coupled processes that influence the fate and transport of supercritical CO2 injected into sedimentary rocks remains a grand challenge that underpins critical aspects of geological sequestration of CO2. Site selection, storage engineering, risk assessment, and project performance confirmation all rely heavily on the veracity of simulation models. Particularly challenging issues for today’s generation of simulation models that are unique to geological sequestration stem from three factors, namely: the very large footprint of a typical sequestration project (100 km2); the thermodynamic properties of CO2 which result in a complex coupling between gravitational, buoyant, and viscous forces, along with dissolution in brine and reaction with the rock; and the permanent nature of sequestration, which entails understanding of the behavior of geologically sequestered CO2 on timescales of 1,000s of year and longer.
These challenges necessitate dealing with geological heterogeneity over very large spatial domains and finding effective ways to volume average rock properties; simulating processes over timescales of centuries to millennia; quantifying uncertainty; and incorporating coupled processes, including diffusive and convective transport, geomechanical deformation and associated risks of induced seismicity, and the kinetics of rock-water-CO2 reactions. Probabilistic treatments of the subsurface geology are needed to quantify likely outcome and confidence levels. Importantly, reliable methods for incorporating what is learned about secondary trapping processes are needed
to predict post-injection performance and support regulatory decisions about important issues such as how long monitoring will be required.
8. Improving community engagement and informing the general public about the need, opportunity, risks, and benefits of CO2 sequestration in deep geological formations. Several large-scale CCS demonstration projects have been delayed, abandoned, or relocated globally as a result of public opposition or regulatory action. Any research agenda should include a focus on establishing best practices for community engagement, rules of practice, and regulatory guidelines. The general public and policymakers should be educated about the risks and benefits of geological sequestration.
Implementation, Cost, and Management of the Research Agenda
Implementing the research agenda outlined above requires data, community coordination, and funding. Data are the lifeblood of research, but much relevant data are poorly accessible, either because they are scattered in laboratories around the world or because they are proprietary. In particular, the oil and gas industry has collected decades of data on CO2 injection, sequestration, oil recovery, brine recovery, and other processes. However, only a small fraction of these data is available, and only with restrictions, such as not publishing raw data. Likewise, expertise in laboratory analysis, computational modeling, and monitoring, as well as practical experience with sequestration projects resides in many countries. Yet, it will take a collective effort to understand and develop effective co-optimization techniques for storing CO2 in sedimentary formations. A virtual data repository would facilitate the necessary data sharing and collaboration.
This budget reflects a substantial increase of the DOE 2017 budget for CO2 sequestration in deep geological formations (DOE, 2018), as well as the additional and unique needs associated with sequestration for the purposes of negative emissions. This research supports ramping up implementation of CO2 sequestration in deep sedimentary formations to the 100Mt and then Gt scale over the coming decade. Much of the research agenda requires field experimentation and testing. In most cases, new wells should be drilled and completed at costs of about $5M per well, research infrastructure such as roads and power lines must be built, and CO2 must be purchased at a cost typically in the order of $100/t. Close partnership with existing or planned industrial projects can help reduce these costs. However, past experience has shown that research activities do not always align with industry schedules and priorities.
DOE, NSF, EPA, and DOI have distinct and important roles to play in pursuing a comprehensive research agenda. In addition, capacity building to support regulatory and land-use decisions would benefit from a coordinated and rigorous multiagency research program. The potential agency contributions to the research needs and data repository are identified in Table 7.6 and summarized below.
Department of Energy: The Office of Science and the Office of Fossil Energy have supported fundamental and applied research on geological sequestration for the past two decades. New research and development needs in DOE’s purview include research on trapping mechanisms; multi-scale, multi-physics modeling of the fate and transport of CO2 in the subsurface; and development of storage engineering approaches that optimize and accelerate trapping, co-optimize hydrocarbon production, reduce the costs of real-time monitoring, assess the risks of induced seismicity, and expedite site characterization and selection.
National Science Foundation: NSF plays an important role in engaging and leveraging university research on Earth processes that are relevant to sequestration. Cutting-edge advances in hydrology, geochemistry, geophysics, biogeochemistry, and social science all have bearing on the efficacy and acceptance of geological sequestration.
TABLE 7.6 U.S. Federal Agency Responsibilities for the Research Needs and Data Repository
|Characterization and site selection||X||X||X|
|Co-optimization of EOR/sequestration||X|
|Environmental impacts and risk assessment||X||X|
In addition, NSF can support translational research, such as bringing the latest innovations to sequestration science and engineering, and can build capacity for industry and regulatory agencies to rapidly scale this technology.
Environmental Protection Agency: EPA regulates the site selection and compliance with operational, monitoring, and reporting requirements, and establishes greenhouse gas inventory accounting requirements for geological sequestration projects. EPA is also concerned with risk and could work with DOE and DOI to support the development of reliable approaches to assess, minimize, and monitor the risk of groundwater contamination at sequestration sites. The agency could also work with DOE to develop monitoring-modeling workflows to reduce the ongoing costs of meeting regulatory requirements.
Department of Interior: USGS conducts research to improve understanding of induced seismicity. In addition, both USGS and Bureau of Land Management (BLM) play important roles in the scale-up of geological sequestration. USGS has assessed geological storage resources in U.S. sedimentary basins (Blondes et al., 2013) and is well placed to identify prospective regions for geological sequestration of CO2. Continuous updates of USGS estimates on prospective sequestration and resource size using information from real-world experience would support scale-up to Gt/y and would guide site characterization and site selection. Federal lands will be needed to achieve Gt-scale sequestration, BLM could undertake a study of their sequestration potential.
Tens of millions of tons of CO2 are injected into subsurface pore space annually, mainly to facilitate enhanced oil recovery. For this reason, it is often taken for granted that the oil industry and its partners can easily provide unlimited CO2 sequestration when and if it becomes economically viable. However, there is an enormous difference between megaton-scale EOR projects in oil fields and sequestration of billions of tons of CO2 per year in deep saline aquifers. Implementing the research agenda outlined above will yield information essential for storing enough CO2 to make a substantial contribution to greenhouse gas mitigation. It will also help project operators to avoid secondary impacts that could harm people and generate opposition to geologic sequestration. Some of the research is aimed at reducing the potential for groundwater contamination above CO2 reservoirs, even those with otherwise imperceptible leaks. Other research is aimed at reducing the number and size of induced earthquakes or even avoiding them through careful site selection. The proposed program of early-stage laboratory research combined with site-specific research will lay the foundation for scale-up of CO2 sequestration in deep geological formations to the Gt level.
This page intentionally left blank.