Natural and industrial analogues for leakage of CO₂ from storage reservoirs: identification of features, events, and processes and lessons learned

Abstract

Instances of gas leakage from naturally occurring CO₂ reservoirs and natural gas storage sites serve as analogues for the potential release of CO₂ from geologic storage sites. This paper summarizes and compares the features, events, and processes that can be identified from these analogues, which include both naturally occurring releases and those associated with industrial processes. The following conclusions are drawn: (1) carbon dioxide can accumulate beneath, and be released from, primary and secondary shallower reservoirs with capping units located at a wide range of depths; (2) many natural releases of CO₂ are correlated with a specific event that triggered the release; (3) unsealed fault and fracture zones may act as conduits for CO₂ flow from depth to the surface; (4) improperly constructed or abandoned wells can rapidly release large quantities of CO₂; (5) the types of CO₂ release at the surface vary widely between and within different leakage sites; (6) the hazard to human health was small in most cases, possibly because of implementation of post-leakage public education and monitoring programs; (7) while changes in groundwater chemistry were related to CO₂ leakage, waters often remained potable. Lessons learned for risk assessment associated with geologic carbon sequestration are discussed.

Introduction

The injection and storage of anthropogenic CO₂ in deep geologic formations is a potentially feasible strategy by which to reduce CO₂ emissions and atmospheric concentrations (e.g., International Energy Agency 1997; Reichle et al. 1999). While the purpose of geologic carbon storage is to trap CO₂ underground, CO₂ could migrate away from the storage site into the shallow subsurface and atmosphere if permeable pathways (such as well bores or faults) are present. While limited CO₂ leakage does not negate the net reduction of CO₂ emissions to the atmosphere, adverse health, safety, and environmental risks associated with elevated CO₂ concentrations must be evaluated, particularly if the release at the surface occurs quickly and/or is spatially focused. Cases of CO₂ and CH₄ leakage from reservoirs to the near-surface environment as a result of natural and industrial processes serve as analogues for the potential release of CO₂ from geologic storage sites (e.g., Allis et al. 2001; Stevens et al. 2001a, b; Benson et al. 2002; Beaubien et al. 2004; Shipton et al. 2004; NASCENT 2005). Analysis of these analogues thus provides important insight into the key characteristics of CO₂ leakage; the resulting impact of the leakage on human health and safety, ecology, surface water, and groundwater; and quality, and the effectiveness of remedial measures applied. Lessons can then be learned from natural and industrial analogues for risk assessment associated with geologic CO₂ injection and storage.

The features, events, and processes (FEPs) relevant to the geologic disposal of radioactive waste have been compiled and used in systems analysis to assess performance and safety (e.g., NEA/OECD 2000). Based on this work, Savage et al. (2004) developed a framework for compiling a database of generic FEPs for the evaluation of CO₂ storage sites. However, FEPs associated with geologic sequestration of CO₂, in particular potential CO₂ leakage from storage sites, have not been identified from actual cases of leakage from natural CO₂ reservoirs. The purpose of this paper is to summarize and compare the FEPs of CO₂ leakage examples from natural geologic reservoirs and CH₄ from natural gas storage sites. To this end, a broad (although not exhaustive) set of natural and industrial analogues for CO₂ leakage is summarized (Appendices A and B; online supplementary material). In addition, several examples of CH₄ leakage associated with industrial processes are described. Based on this summary, the causes and consequences of CO₂ leakage resulting from natural and industrial processes are described, with particular emphasis on (a) the geologic model for CO₂ accumulation in the reservoir, (b) events leading to the leakage of the CO₂ from the reservoir, (c) pathways for CO₂ migration to the surface, (d) the magnitude and consequences of the release, and (e) remedial strategies applied. Implications for geologic carbon storage and the related risk assessment work are then discussed.

Overview of natural and industrial analogues

Leakage of CO₂ has occurred naturally from geologic reservoirs in numerous volcanic, geothermal, and sedimentary basin settings worldwide. These systems serve as natural analogues for the potential leakage of CO₂ from geologic carbon sequestration sites and provide information on both the short and long-term causes and effects of CO₂ leakage. It should be noted, however, that in some environments the rates of CO₂ filling of natural geologic reservoirs may not be comparable to (i.e., may be lower than) CO₂ injection rates at storage sites. In these cases, natural analogues may not provide information on processes related to relatively rapid injection rates, such as pressure-induced geomechanical damage.

Leakage of CO₂ from geologic reservoirs resulting from industrial processes has occurred relatively infrequently. A limited number of examples of CO₂ and CH₄ leakage associated with exploration for or exploitation of geothermal, CO₂, natural gas, and water resources have been reported in the literature. Nonetheless, these relatively rare leakage events serve as analogues for the potential release of CO₂ from sequestration sites caused by human-related practices. Because most reported events of industrial CO₂ leakage are associated with well construction and injection and withdrawal practices, and few events have been reported associated with abandoned wells, industrial analogues usually provide information on short-term causes and effects of CO₂ leakage. Although the densities and solubilities of CH₄ and CO₂ are different, they have similar viscosities and behave as buoyant fluids. Furthermore, CO₂ injection rates into storage reservoirs will likely be comparable to CH₄ injection rates into natural gas storage sites. Thus, examples of CH₄ leakage from natural gas storage sites can serve as industrial analogues for CO₂ leakage.

Together, natural and industrial analogues for CO₂ leakage provide important information about the key FEPs that are associated with leakage, as well as the health, safety, and environmental consequences of leakage and mitigation efforts applied. Appendices A and B (online supplementary material) describe these aspects of natural and industrial analogues, respectively, for CO₂ leakage. Appendix A begins with a more detailed summary for Mammoth Mountain (USA) due to the large amount of data available for this site, and follows with more generalized descriptions of CO₂ leakage in other volcanic, geothermal, and sedimentary basin systems. While the Appendices (online supplementary material) do not represent an exhaustive list of natural and industrial cases of CO₂ leakage that have occurred worldwide, a suite of well-documented cases representing a range of geologic settings is included. In addition, several examples of CH₄ leakage related to industrial processes are given. Focus is placed on CO₂ leakage cases presented in the peer-reviewed literature. Also, emphasis is placed on natural rather than industrial analogues because (1) they provide information on both the short- and long-term causes and effects of CO₂ leakage and (2) a relatively large number of examples of natural CO₂ leakage events are reported in the literature.

To compare the key characteristics of CO₂ or CH₄ leakage associated with natural and industrial processes detailed in Appendices A and B (online supplementary material), leakage cases are classified according to the key features of the CO₂ or CH₄ accumulation, the events leading to the leakage of CO₂ or CH₄ from the reservoir, and the processes by which the CO₂ or CH₄ was released at the surface (Tables 1, 2). In Tables 1 and 2, columns 1 through 3 describe these key features, including site location, the source of the CO₂ or CH₄ in the accumulation, and the geologic model for CO₂ or CH₄ accumulation, for example, the reservoir, reservoir depth (if known), and capping rocks. Column 4 in Tables 1 and 2 describes the event triggering the leakage of CO₂ or CH₄ from the reservoir. Columns 5 and 6 give the processes by which the CO₂ or CH₄ was emitted at the surface, including the pathway(s) for leakage and the style of surface emission. References for each of the leakage analogues are given in Appendices A and B (online supplementary material).

Table 1 Summary of FEPs of natural leakage of CO₂ (see Appendix A for references)

Table 2 Summary of FEPs of industrial leakage of CO₂ or CH₄ (see Appendix B for references)

Common FEPs

As shown in Tables 1 and 2, several key similarities exist between the FEPs of different natural and industrial cases of CO₂ or CH₄ leakage. First, the sources of CO₂ in natural accumulations are most commonly thermal decomposition of carbonate-rich sedimentary rocks and/or degassing of magma bodies at depth (analogues A1–A5, A7–A13, B1–B5). Second, CO₂ from these sources often accumulates in highly fractured and/or porous rocks (e.g., sandstones, limestones) under low-permeability cap rocks (analogues A1–A6, A12, A13, B1–B5). The cap rocks may be low-permeability rock units (e.g., shale and siltstone) or a zone of hydrothermal alteration.

In the case of natural CO₂ leakage, once the CO₂ leaks from the storage reservoir, fault and/or fracture zones are the primary pathways for CO₂ migration to the surface (analogues A1–A9, A12, A13). These high-permeability zones may be pre-existing, or be created/enhanced due to seismic activity associated with, for example, fluid migration and pore-fluid pressurization. In the case of CO₂ or CH₄ leakage associated with industrial activity, the event triggering the release is commonly a well blowout, related to injection/withdrawal practices or a defect in a well (analogues B1–B8). Also, the pathway for CO₂ or CH₄ migration to the surface is usually the well bore and/or fractures that have formed around the well bore.

Differing FEPs

There are several key differences between the FEPs of the various examples of CO₂ or CH₄ leakage (Tables 1, 2). The depth of the CO₂ or CH₄ source and the reservoir(s) in which the CO₂ accumulates or is injected varies widely from <1 km (e.g., analogues B3–B7) to multiple km (e.g., analogues A1, B1, B8). At an individual site, CO₂ may accumulate in a single reservoir (e.g., analogues A1, A4–A6, B4), or within multiple vertically stacked and/or horizontally compartmentalized reservoirs (analogues A12, A13, B2, B3, B5). Cap rocks and/or low-permeability fault zones can serve to separate multiple CO₂ reservoirs at a given site.

Some examples of natural CO₂ leakage have been correlated with specific triggering events, such as seismic activity or magmatic fluid injection (analogues A1, A3, A7, A10, A11), while other events have not been correlated with such events. However, the lack of correlation in the latter cases may be due to the absence of observations or data collection at the time of the leakage event. Where a trigger event was identified, it was commonly an event that caused geomechanical damage to cap rocks sealing the CO₂ reservoir.

Finally, the style of natural CO₂ leakage at the surface varies widely between different sites, as well as within individual sites. Surface releases occur in the form of diffuse gas emission over large land areas, focused vent emissions, eruptive emissions, degassing through surface water bodies, and/or release with spring discharge (analogues A1–A6, A8, A9, A10–A12). In rare cases, the CO₂ release may have been a self-enhancing or eruptive process (analogues A7, A10). In the case of CO₂ or CH₄ leakage associated with well failures, the gas may be emitted at the surface in a focused form, free flowing or geysering from the well and/or diffusely through soils, water pools, or fractures around the well site (analogues B1–B8).

Additional considerations

The magnitude and consequences of CO₂ or CH₄ leakage events, as well as the type and success of strategies that were implemented to monitor and/or remediate the leakage, are important additional considerations that should be taken into account in risk assessment associated with geologic carbon sequestration. Table 3 and Appendices A and B (online supplementary material) detail these aspects for the natural and industrial leakage analogues, showing that the magnitude of the surface CO₂ or CH₄ release varies widely between different cases and does not necessarily depend on the style of the release. For example, the magnitude of diffuse CO₂ emissions from soils varies greatly between different, as well as within, individual sites. At sites where groundwater chemistry was monitored, chemical changes were sometimes observed related to CO₂ leakage (e.g., analogues A1, A3, A11, A12) and resulting groundwater acidification and interaction with host rocks along flow paths. However, groundwaters remained potable in most cases.

Table 3 Additional aspects of CO₂ or CH₄ leakage associated with natural (N) and industrial (I) processes (see Appendices A and B for references)

In many of the leakage examples, CO₂ was monitored in the near-surface environment within and around CO₂ leakage areas, often on a regular basis (analogues A1–A6, A8–A12, B2, B4). Monitoring strategies include measurements of soil CO₂ flux using accumulation chamber or eddy covariance methods and soil, atmospheric, or vent gas CO₂ concentration using gas analyser or chromatography techniques. Controlled degassing of CO₂-rich lakes has also been carried out to mitigate CO₂ buildup (analogues A10). In many cases, public education programs were implemented to advise people visiting or living near the CO₂ release areas of potential health, safety, and environmental hazards (e.g., analogues A1–A3, A6, A10). Zoning bylaws have also been established in some cases to control development near high CO₂ emission areas (e.g., analogues A3, A6). The hazard to human health was small in most examples of surface CO₂ releases; this could be attributed in part to public education and CO₂ monitoring programs.

Conclusions and lessons learned

Leakage of CO₂ has occurred naturally from geologic reservoirs in numerous volcanic, geothermal, and sedimentary basin settings. In addition, CO₂ and CH₄ have been released from industrial reservoirs as a result of well defects and/or injection/withdrawal processes. These systems serve as natural and industrial analogues for the potential release of CO₂ from geologic storage reservoirs and provide important information about the key FEPs associated with releases, as well as the health, safety, and environmental consequences of releases and monitoring and mitigation efforts that can be applied. While the detailed characteristics of CO₂ or CH₄ leakage are unique to each case, five general FEPs were identified based on analysis of a range of natural and industrial analogues for CO₂ and CH₄ leakage, from which lessons can be learned and should be applied to risk assessment associated with geologic carbon storage:

• 1. Carbon dioxide can both accumulate beneath, and be released from, primary and secondary reservoirs with capping units located at a wide range of depths. Both primary and secondary reservoir entrapments for CO₂ should therefore be properly characterized at potential geologic carbon sequestration sites.

• 2. Many natural releases of CO₂ have been correlated with a specific event that has triggered the release, such as magmatic or seismic activity. The potential for processes that could cause geomechanical damage to sealing cap rocks and trigger the release of CO₂ from a storage reservoir should be evaluated.

• 3. Unsealed fault and fracture zones can act as fast and direct conduits for CO₂ flow from depth to the surface. Risk assessment should therefore emphasize determining the potential for and nature of CO₂ migration along these structures.

• 4. Wells that are improperly constructed or abandoned, and become structurally unsound over time, have the potential to rapidly release large quantities of CO₂ to the atmosphere. One focus of risk assessment should therefore be an evaluation of the potential for both active and abandoned wells at storage sites to transport CO₂ to the surface, particularly in depleted oil or gas reservoir systems, where wells are abundant.

• 5. The style of CO₂ release at the surface varies widely between and within different leakage sites. In rare circumstances, the release of CO₂ can be a self-enhancing and/or eruptive process; this possibility should be assessed in the case of CO₂ leakage from storage reservoirs.

Furthermore, analysis of natural and industrial analogues has demonstrated two important points related to human health and safety and groundwater quality. First, the hazard to human health was small in most examples of CO₂ leakage. This could result from implementing public education and CO₂ monitoring programs; these “remedial” programs should therefore be employed to minimize potential health, safety, and environmental effects associated with CO₂ leakage. Second, while changes in groundwater chemistry can be related to CO₂ leakage caused by acidification and interaction with host rocks along flow paths, waters remained potable in many cases. Groundwaters should be monitored for changes in chemistry that could result from CO₂ leakage from storage sites.