WORLDWIDE DRILLING DEWS I TV Number of wells drilled per 10.000 sq km l -100 100 -3M

Figure 5.27 World oil and gas well distribution and density (courtesy of IHS Energy).

CO2 through the subsurface will be slow. For example, Cawley et al. (2005) studied the effect of uncertainties in parameters such as the flow velocity in the aquifer and capillary entry pressure into caprock in their examination of CO2 storage in the Forties Oilfield in the North Sea. Over the 1000 year time scale examined in their study, Cawley et al. (2005) found that less than 0.2% of the stored CO2 enters into the overlying layers and even in the worse case, the maximum vertical distance moved by any of the CO2 was less than halfway to the seabed. Similarly, Lindeberg and Bergmo (2003) studied the Sleipner field and found that CO2 would not begin to migrate into the North Sea for 100,000 years and that even after a million years, the annual rate of release would be about 10-6 of the stored CO2 per year.

Simulations designed to explore the possible release of stored CO2 to the biosphere by multiple routes, including abandoned wells and other disturbances, have recently become available as a component of more general risk assessment activities (Section 5.7.5). Two studies of the Weyburn site, for example, assessed the probability of release to the biosphere. Walton et al. (2005) used a fully probabilistic model, with a simplified representation of CO2 transport, to compute a probability distribution for the cumulative fraction released to the biosphere. Walton et al. found that after 5000 years, the probability was equal that the cumulative amount released would be larger or smaller than 0.1% (the median release fraction) and found a 95% probability that <1% of the total amount stored would be released. Using a deterministic model of CO2 transport in the subsurface, Zhou et al. (2005) found no release to the biosphere in 5000 years. While using a probabilistic model of transport through abandoned wells, they found a statistical mean release of 0.001% and a maximum release of 0.14% (expressed as the cumulative fraction of stored CO2 released over 5000 years).

In saline formations or oil and gas reservoirs with significant brine content, much of the CO2 will eventually dissolve in the brine (Figure 5.7), be trapped as a residual immobile phase (Figure 5.8) or be immobilized by geochemical reactions. The time scale for dissolution is typically short compared to the time for CO2 to migrate out of the storage formation by other processes (Ennis-King and Paterson, 2003; Lindeberg and Bergmo, 2003; Walton et al., 2005). It is expected that many storage projects could be selected and operated so that a very large fraction of the injected CO2 will dissolve. Once dissolved, CO2 can eventually be transported out of the injection site by basin-scale circulation or upward migration, but the time scales (millions of years) of such transport are typically sufficiently long that they can (arguably) be ignored in assessing the risk of leakage.

As described in Section 5.1, several CO2 storage projects are now in operation and being carefully monitored. While no leakage of stored CO2 out of the storage formations has been observed in any of the current projects, time is too short and overall monitoring too limited, to enable direct empirical conclusions about the long-term performance of geological storage. Rather than providing a direct test of performance, the current projects improve the quality of long-duration performance predictions by testing and sharpening understanding of CO2 transport and trapping mechanisms. Assessing the ability of operational geological storage projects to retain CO2 for long time periods Assessment of the fraction retained for geological storage projects is highly site-specific, depending on (1) the storage system design, including the geological characteristics of the selected storage site; (2) the injection system and related reservoir engineering; and (3) the methods of abandonment, including the performance of well-sealing technologies. If the above information is available, it is possible to estimate the fraction retained by using the models described in Section 5.4.2 and risk assessment methods described in Section 5.7.5. Therefore, it is also possible, in principle, to estimate the expected performance of an ensemble of storage projects that adhere to design guidelines such as site selection, seal integrity, injection depth and well closure technologies. Table 5.5 summarizes disparate lines of evidence on the integrity of CO2 storage systems.

For large-scale operational CO2 storage projects, assuming that sites are well selected, designed, operated and appropriately monitored, the balance of available evidence suggests the following:

• It is very likely the fraction of stored CO2 retained is more than 99% over the first 100 years.

• It is likely the fraction of stored CO2 retained is more than

99% over the first 1000 years.

5.7.4 Possible local and regional environmental hazards Potential hazards to human health and safety Risks to human health and safety arise (almost) exclusively from elevated CO2 concentrations in ambient air, either in confined outdoor environments, in caves or in buildings. Physiological and toxicological responses to elevated CO2 concentrations are relatively well understood (AI.3.3). At concentrations above about 2%, CO2 has a strong effect on respiratory physiology and at concentrations above 7-10%, it can cause unconsciousness and death. Exposure studies have not revealed any adverse health effect of chronic exposure to concentrations below 1%.

The principal challenge in estimating the risks posed by CO2 that might seep from storage sites lies in estimating the spatial and temporal distribution of CO2 fluxes reaching the shallow subsurface and in predicting ambient CO2 concentration resulting from a given CO2 flux. Concentrations in surface air will be strongly influenced by surface topography and atmospheric conditions. Because CO2 is 50% denser than air, it tends to migrate downwards, flowing along the ground and collecting in shallow depressions, potentially creating much higher concentrations in confined spaces than in open terrain.

Seepage of CO2 is not uncommon in regions influenced by volcanism. Naturally occurring releases of CO2 provide a basis for understanding the transport of CO2 from the vadose zone to the atmosphere, as well as providing empirical data that link CO2 fluxes into the shallow subsurface with CO2 concentrations

table 5.5 Summary of evidence for CO2 retention and release rates.

Kind of evidence

Average annual fraction released

Representative references

CO2 in natural formations

The lifetime of CO2 in natural formations (>10 million yr in some cases) suggests an average release fraction <10-7 yr-1 for CO2 trapped in sedimentary basins. In highly fractured volcanic systems, rate of release can be many orders of magnitude faster.

Stevens et al., 2001a; Baines and Worden, 2001

Oil and gas

The presence of buoyant fluids trapped for geological timescales demonstrates the widespread presence of geological systems (seals and caprock) that are capable of confining gasses with release rates <10-7 yr1.

Bradshaw et al, 2005

Natural gas storage

The cumulative experience of natural gas storage systems exceeds 10,000 facility-years and demonstrates that operational engineered storage systems can contain methane with release rates of 10-4 to 10-6 yr-1.

Lippmann and Benson, 2003; Perry, 2005

Enhanced oil recovery (EOR)

More than 100 MtCO2 has been injected for EOR. Data from the few sites where surface fluxes have been measured suggest that fractional release rates are near zero.

Moritis, 2002; Klusman, 2003

Models of flow through the undisturbed subsurface

Numerical models show that release of CO2 by subsurface flow through undisturbed geological media (excluding wells) may be near zero at appropriately selected storage sites and is very likely <10-6 in the few studies that attempted probabilistic estimates.

Walton et al, 2005; Zhou et al., 2005; Lindeberg and Bergmo, 2003; Cawley et al., 2005

Models of flow through wells

Evidence from a small number of risk assessment studies suggests that average release of CO2 can be 10-5 to 10-7 yr-1 even in existing oil fields with many abandoned wells, such as Weyburn. Simulations with idealized systems with 'open' wells show that release rates can exceed 10-2, though in practice such wells would presumably be closed as soon as CO2 was detected.

Walton et al, 2005; Zhou et al., 2005; Nordbotten et al, 2005b

Current CO2 storage projects

Data from current CO2 storage projects demonstrate that monitoring techniques are able to detect movement of CO2 in the storage reservoirs. Although no release to the surface has been detected, little can be concluded given the short history and few sites.

Wilson and Monea, 2005; Arts et al, 2005; Chadwick, et al., 2005

in the ambient air - and the consequent health and safety risks. Such seeps do not, however, provide a useful basis for estimating the spatial and temporal distribution of CO2 fluxes leaking from a deep storage site, because (in general) the seeps occur in highly fractured volcanic zones, unlike the interiors of stable sedimentary basins, the likely locations for CO2 storage (Section 5.3).

Natural seeps are widely distributed in tectonically active regions of the world (Morner and Etiope, 2002). In central Italy, for example, CO2 is emitted from vents, surface degassing and diffuse emission from CO2-rich groundwater. Fluxes from vents range from less than 100 to more than 430 tCO2 day-1, which have shown to be lethal to animal and plants. At Poggio dell'Ulivo, for example, a flux of 200 tCO2 day-1 is emitted from diffuse soil degassing. At least ten people have died from CO2 releases in the region of Lazio over the last 20 years.

Natural and engineered analogues show that it is possible, though improbable, that slow releases from CO2 storage reservoirs will pose a threat to humans. Sudden, catastrophic releases of natural accumulations of CO2 have occurred, associated with volcanism or subsurface mining activities. Thus, they are of limited relevance to understanding risks arising from CO2 stored in sedimentary basins. However, mining or drilling in areas with CO2 storage sites may pose a long-term risk after site abandonment if institutional knowledge and precautions are not in place to avoid accidentally penetrating a storage formation. Hazards to groundwater from CO2 leakage and brine displacement Increases in dissolved CO2 concentration that might occur as CO2 migrates from a storage reservoir to the surface will alter groundwater chemistry, potentially affecting shallow groundwater used for potable water and industrial and agricultural needs. Dissolved CO2 forms carbonic acid, altering the pH of the solution and potentially causing indirect effects, including mobilization of (toxic) metals, sulphate or chloride; and possibly giving the water an odd odour, colour or taste. In the worst case, contamination might reach dangerous levels, excluding the use of groundwater for drinking or irrigation.

Wang and Jaffe (2004) used a chemical transport model to investigate the effect of releasing CO2 from a point source at 100 m depth into a shallow water formation that contained a high concentration of mineralized lead (galena). They found that in weakly buffered formations, the escaping CO2 could mobilize sufficient dissolved lead to pose a health hazard over a radius of a few hundred metres from the CO2 source. This analysis represents an extreme upper bound to the risk of metal leaching, since few natural formations have mineral composition so susceptible to the effects of CO2-mediated leaching and one of the expressed requirements of a storage site is to avoid compromising other potential resources, such as mineral deposits.

The injection of CO2 or any other fluid deep underground necessarily causes changes in pore-fluid pressures and in the geomechanical stress fields that reach far beyond the volume occupied by the injected fluid. Brines displaced from deep formations by injected CO2 can potentially migrate or leak through fractures or defective wells to shallow aquifers and contaminate shallower drinking water formations by increasing their salinity. In the worst case, infiltration of saline water into groundwater or into the shallow subsurface could impact wildlife habitat, restrict or eliminate agricultural use of land and pollute surface waters.

As is the case for induced seismicity, the experience with injection of different fluids provides an empirical basis for assessing the likelihood that groundwater contamination will occur by brine displacement. As discussed in Section 5.5 and shown in Figure 5.22, the current site-specific injection rates of fluids into the deep subsurface are roughly comparable to the rates at which CO2 would be injected if geological storage were adopted for storage of CO2 from large-scale power plants. Contamination of groundwater by brines displaced from injection wells is rare and it is therefore expected that contamination arising from large-scale CO2 storage activities would also be rare. Density differences between CO2 and other fluids with which we have extensive experience do not compromise this conclusion, because brine displacement is driven primarily by the pressure/hydraulic head differential of the injected CO2, not by buoyancy forces. Hazards to terrestrial and marine ecosystems Stored CO2 and any accompanying substances, may affect the flora and fauna with which it comes into contact. Impacts might be expected on microbes in the deep subsurface and on plants and animals in shallower soils and at the surface. The remainder of this discussion focuses only on the hazards where exposures to CO2 do occur. As discussed in Section 5.7.3, the probability of leakage is low. Nevertheless, it is important to understand the hazards should exposures occur.

In the last three decades, microbes dubbed 'extremophiles', living in environments where life was previously considered impossible, have been identified in many underground habitats. These microorganisms have limited nutrient supply and exhibit very low metabolic rates (D'Hondt et al., 2002). Recent studies have described populations in deep saline formations (Haveman and Pedersen, 2001), oil and gas reservoirs (Orphan et al., 2000) and sediments up to 850 m below the sea floor (Parkes et al., 2000). The mass of subsurface microbes may well exceed the mass of biota on the Earth's surface (Whitman et al., 2001). The working assumption may be that unless there are conditions preventing it, microbes can be found everywhere at the depths being considered for CO2 storage and consequently CO2 storage sites may generally contain microbes that could be affected by injected CO2.

The effect of CO2 on subsurface microbial populations is not well studied. A low-pH, high-CO2 environment may favour some species and harm others. In strongly reducing environments, the injection of CO2 may stimulate microbial communities that would reduce the CO2 to CH4; while in other reservoirs, CO2 injection could cause a short-term stimulation of Fe(III)-reducing communities (Onstott, 2005). From an operational perspective, creation of biofilms may reduce the effective permeability of the formation.

Should CO2 leak from the storage formation and find its way to the surface, it will enter a much more biologically active area. While elevated CO2 concentrations in ambient air can accelerate plant growth, such fertilization will generally be overwhelmed by the detrimental effects of elevated CO2 in soils, because CO2 fluxes large enough to significantly increase concentrations in the free air will typically be associated with much higher CO2 concentrations in soils. The effects of elevated CO2 concentrations would be mediated by several factors: the type and density of vegetation; the exposure to other environmental stresses; the prevailing environmental conditions like wind speed and rainfall; the presence of low-lying areas; and the density of nearby animal populations.

The main characteristic of long-term elevated CO2 zones at the surface is the lack of vegetation. New CO2 releases into vegetated areas cause noticeable die-off. In those areas where significant impacts to vegetation have occurred, CO2 makes up about 20-95% of the soil gas, whereas normal soil gas usually contains about 0.2-4% CO2. Carbon dioxide concentrations above 5% may be dangerous for vegetation and as concentration approach 20%, CO2 becomes phytotoxic. Carbon dioxide can cause death of plants through 'root anoxia', together with low oxygen concentration (Leone et al., 1977; Flower et al., 1981).

One example of plant die-off occurred at Mammoth Mountain, California, USA, where a resurgence of volcanic activity resulted in high CO2 fluxes. In 1989, a series of small earthquakes occurred near Mammoth Mountain. A year later, 4 ha of pine trees were discovered to be losing their needles and by 1997, the area of dead and dying trees had expanded to 40 ha (Farrar et al., 1999). Soil CO2 levels above 10-20% inhibit root development and decrease water and nutrient uptake; soil oil-gas testing at Mammoth Mountain in 1994 discovered soil gas readings of up to 95% CO2 by volume. Total CO2 flux in the affected areas averaged about 530 t day-1 in 1996. Measurements in 2001 showed soil CO2 levels of 15-90%, with flux rates at the largest affected area (Horseshoe Lake) averaging 90-100 tCO2 day-1 (Gerlach et al., 1999; Rogie et al., 2001). A study of the impact of elevated CO2 on soils found there was a lower pH and higher moisture content in summer. Wells in the high CO2 area showed higher levels of silicon, aluminum, magnesium and iron, consistent with enhanced weathering of the soils. Tree-ring data show that CO2 releases have occurred prior to 1990 (Cook et al., 2001). Data from airborne remote sensing are now being used to map tree health and measure anomalous CO2 levels, which may help determine how CO2 affects forest ecosystems (Martini and Silver, 2002).

There is no evidence of any terrestrial impact from current CO2 storage projects. Likewise, there is no evidence from EOR projects that indicate impacts to vegetation such as those described above. However, no systematic studies have occurred to look for terrestrial impacts from current EOR projects.

Natural CO2 seepage in volcanic regions, therefore, provides examples of possible impacts from leaky CO2 storage, although

(as mentioned in Section 5.2.3) seeps in volcanic provinces provide a poor analogue to seepage that would occur from CO2 storage sites in sedimentary basins. As described above, CO2 seepage can pose substantial hazards. In the Alban Hills, south of Rome (Italy), for example, 29 cows and 8 sheep were asphyxiated in several separate incidents between September 1999 and October 2001 (Carapezza et al, 2003). The measured CO2 flux was about 60 t day-1 of 98% CO2 and up to 2% H2S, creating hazardous levels of each gas in localized areas, particularly in low-wind conditions. The high CO2 and H2S fluxes resulted from a combination of magmatic activity and faulting.

Human activities have caused detrimental releases of CO2 from the deep subsurface. In the late 1990s, vegetation died off above an approximately 3-km deep geothermal field being exploited for a 62 MW power plant, in Dixie Valley, Nevada, USA (Bergfeld et al., 2001). A maximum flux of 570 gCO2 m-2 day-1 was measured, as compared to a background level of 7 gCO2 m-2 day-1. By 1999, CO2 flow in the measured area ceased and vegetation began to return.

The relevance of these natural analogues to leakage from CO2 storage varies. For examples presented here, the fluxes and therefore the risks, are much higher than might be expected from a CO2 storage facility: the annual flow of CO2 at the Mammoth Mountain site is roughly equal to a release rate on the order of 0.2% yr-1 from a storage site containing 100 MtCO2. This corresponds to a fraction retained of 13.5% over 1000 years and, thus, is not representative of a typical storage site.

Seepage from offshore geological storage sites may pose a hazard to benthic environments and organisms as the CO2 moves from deep geological structures through benthic sediments to the ocean. While leaking CO2 might be hazardous to the benthic environment, the seabed and overlying seawater can also provide a barrier, reducing the escape of seeping CO2 to the atmosphere. These hazards are distinctly different from the environmental effects of the dissolved CO2 on aquatic life in the water column, which are discussed in Chapter 6. No studies specifically address the environmental effects of seepage from sub-seabed geological storage sites. Induced seismicity

Underground injection of CO2 or other fluids into porous rock at pressures substantially higher than formation pressures can induce fracturing and movement along faults (see Section 5.5.4 and Healy et al., 1968; Gibbs et al., 1973; Raleigh et al., 1976; Sminchak et al., 2002; Streit et al., 2005; Wo et al., 2005). Induced fracturing and fault activation may pose two kinds of risks. First, brittle failure and associated microseismicity induced by overpressuring can create or enhance fracture permeability, thus providing pathways for unwanted CO2 migration (Streit and Hillis, 2003). Second, fault activation can, in principle, induce earthquakes large enough to cause damage (e.g., Healy et al., 1968).

Fluid injection into boreholes can induce microseismic activity, as for example at the Rangely Oil Field in Colorado, USA (Gibbs et al., 1973; Raleigh et al., 1976), in test sites such as the drillholes of the German continental deep drilling programme (Shapiro et al., 1997; Zoback and Haijes, 1997) or the Cold Lake Oil Field, Alberta, Canada (Talebi et al., 1998). Deep-well injection of waste fluids may induce earthquakes with moderate local magnitudes (ML), as suggested for the 1967 Denver earthquakes (ML of 5.3; Healy et al., 1968; Wyss and Molnar, 1972) and the 1986-1987 Ohio earthquakes (ML of 4.9; Ahmad and Smith, 1988) in the United States. Seismicity induced by fluid injection is usually assumed to result from increased pore-fluid pressure in the hypocentral region of the seismic event (e.g., Healy et al., 1968; Talebi et al., 1998).

Readily applicable methods exist to assess and control induced fracturing or fault activation (see Section 5.5.3). Several geomechanical methods have been identified for assessing the stability of faults and estimating maximum sustainable pore-fluid pressures for CO2 storage (Streit and Hillis, 2003). Such methods, which require the determination of in situ stresses, fault geometries and relevant rock strengths, are based on brittle failure criteria and have been applied to several study sites for potential CO2 storage (Rigg et al., 2001; Gibson-Poole et al., 2002).

The monitoring of microseismic events, especially in the vicinity of injection wells, can indicate whether pore fluid pressures have locally exceeded the strength of faults, fractures or intact rock. Acoustic transducers that record microseismic events in monitoring wells of CO2 storage sites can be used to provide real-time control to keep injection pressures below the levels that induce seismicity. Together with the modelling techniques mentioned above, monitoring can reduce the chance of damage to top seals and fault seals (at CO2 storage sites) caused by injection-related pore-pressure increases.

Fault activation is primarily dependent on the extent and magnitude of the pore-fluid-pressure perturbations. It is therefore determined more by the quantity and rate than by the kind of fluid injected. Estimates of the risk of inducing significant earthquakes may therefore be based on the diverse and extensive experience with deep-well injection of various aqueous and gaseous streams for disposal and storage. Perhaps the most pertinent experience is the injection of CO2 for EOR; about 30 MtCO2 yr-1 is now injected for EOR worldwide and the cumulative total injected exceeds 0.5 GtCO2, yet there have been no significant seismic effects attributed to CO2-EOR. In addition to CO2, injected fluids include brines associated with oil and gas production (>2 Gt yr-1); Floridan aquifer wastewater (>0.5 Gt yr-1); hazardous wastes (>30 Mt yr-1); and natural gas (>100 Mt yr-1) (Wilson et al., 2003).

While few of these cases may precisely mirror the conditions under which CO2 would be injected for storage (the peak pressures in CO2-EOR may, for example, be lower than would be used in formation storage), these quantities compare to or exceed, plausible flows of CO2 into storage. For example, in some cases such as the Rangely Oil Field, USA, current reservoir pressures even exceed the original formation pressure (Raleigh et al., 1976). Thus, they provide a substantial body of empirical data upon which to assess the likelihood of induced seismicity resulting from fluid injection. The fact that only a few individual seismic events associated with deep-well injection have been recorded suggests that the risks are low. Perhaps more importantly, these experiences demonstrate that the regulatory limits imposed on injection pressures are sufficient to avoid significant injection-induced seismicity. Designing CO2 storage projects to operate within these parameters should be possible. Nevertheless, because formation pressures in CO2 storage formations may exceed those found in CO2-EOR projects, more experience with industrial-scale CO2 storage projects will be needed to fully assess risks of microseismicity. Implications of gas impurity

Under some circumstances, H2S, SO2, NO2 and other trace gases may be stored along with CO2 (Bryant and Lake, 2005; Knauss et al., 2005) and this may affect the level of risk. For example, H2S is considerably more toxic than CO2 and well blow-outs containing H2S may present higher risks than well blow-outs from storage sites that contain only CO2. Similarly, dissolution of SO2 in groundwater creates a far stronger acid than does dissolution of CO2; hence, the mobilization of metals in groundwater and soils may be higher, leading to greater risk of exposure to hazardous levels of trace metals. While there has not been a systematic and comprehensive assessment of how these additional constituents would affect the risks associated with CO2 storage, it is worth noting that at Weyburn, one of the most carefully monitored CO2 injection projects and one for which a considerable effort has been devoted to risk assessment, the injected gas contains approximately 2% H2S (Wilson and Monea, 2005). To date, most risk assessment studies have assumed that only CO2 is stored; therefore, insufficient information is available to assess the risks associated with gas impurities at the present time.

5.7.5 Risk assessment methodology

Risk assessment aims to identify and quantify potential risks caused by the subsurface injection of CO2, where risk denotes a combination (often the product) of the probability of an event happening and the consequences of the event. Risk assessment should be an integral element of risk-management activities, spanning site selection, site characterization, storage system design, monitoring and, if necessary, remediation.

The operation of a CO2 storage facility will necessarily involve risks arising from the operation of surface facilities such as pipelines, compressors and wellheads. The assessment of such risks is routine practice in the oil and gas industry and available assessment methods like hazard and operability and quantitative risk assessment are directly applicable. Assessment of such risks can be made with considerable confidence, because estimates of failure probabilities and the consequences of failure can be based directly on experience. Techniques used for assessment of operational risks will not, in general, be readily applicable to assessment of risks arising from long-term storage of CO2 underground. However, they are applicable to the operating phase of a storage project. The remainder of this subsection addresses the long-term risks.

Risk assessment methodologies are diverse; new methodologies arise in response to new classes of problems. Because analysis of the risks posed by geological storage of CO2 is a new field, no well-established methodology for assessing such risks exists. Methods dealing with the long-term risks posed by the transport of materials through the subsurface have been developed in the area of hazardous and nuclear waste management (Hodgkinson and Sumerling, 1990; North, 1999). These techniques provide a useful basis for assessing the risks of CO2 storage. Their applicability may be limited, however, because the focus of these techniques has been on assessing the low-volume disposal of hazardous materials, whereas the geological storage of CO2 is high-volume disposal of a material that involves comparatively mild hazards.

Several substantial efforts are under way to assess the risks posed by particular storage sites (Gale, 2003). These risk assessment activities cover a wide range of reservoirs, use a diversity of methods and consider a very wide class of risks. The description of a representative selection of these risk assessment efforts is summarized in Table 5.6.

The development of a comprehensive catalogue of the risks and of the mechanisms that underlie them, provides a good foundation for systematic risk assessment. Many of the ongoing risk assessment efforts are now cooperating to identify, classify and screen all factors that may influence the safety of storage facilities, by using the features, events and processes (FEP) methodology. In this context, features includes a list of parameters, such as storage reservoir permeability, caprock thickness and number of injection wells. Events includes processes such as seismic events, well blow-outs and penetration of the storage site by new wells. Processes refers to the physical and chemical processes, such as multiphase flow, chemical reactions and geomechanical stress changes that influence storage capacity and security. FEP databases tie information on individual FEPs to relevant literature and allow classification with respect to likelihood, spatial scale, time scale and so on. However, there are alternative approaches.

Most risk assessments involve the use of scenarios that describe possible future states of the storage facility and events that result in leakage of CO2 or other risks. Each scenario may be considered as an assemblage of selected FEPs. Some risk assessments define a reference scenario that represents the most probable evolution of the system. Variant scenarios are then constructed with alternative FEPs. Various methods are used to structure and rationalize the process of scenario definition in an attempt to reduce the role of subjective judgements in determining the outcomes.

Scenarios are the starting points for selecting and developing mathematical-physical models (Section5.4.2). Suchperformance assessment models may include representations of all relevant components including the stored CO2, the reservoir, the seal, the overburden, the soil and the atmosphere. Many of the fluid-transport models used for risk assessment are derived from (or identical to) well-established models used in the oil and gas or groundwater management industries (Section 5.4.2). The detail or resolution of various components may vary greatly. Some

Table 5.6 Representative selection of risk assessment models and efforts.

Project title

Description and status


New model, CQUESTRA, developed to enable probabilistic risk assessment. A simple box model is used with explicit representation of transport between boxes caused by failure of wells.

Weyburn/Monitor Scientific

Scenario-based modelling that uses an industry standard reservoir simulation tool (Eclipse3000) based on a realistic model of known reservoir conditions. Initial treatment of wells involves assigning a uniform permeability.

NGCAS/ECL technology

Probabilistic risk assessment using fault tree and FEP (features, events and processes) database. Initial study focused on the Forties oil and gas field located offshore in the North Sea. Concluded that flow through caprock transport by advection in formation waters not important, work on assessing leakage due to well failures ongoing.

SAMARCADS (safety aspects of CO2 storage)

Methods and tools for HSE risk assessment applied to two storage systems an onshore gas storage facility and an offshore formation.

Was this article helpful?

0 0
Guide to Alternative Fuels

Guide to Alternative Fuels

Your Alternative Fuel Solution for Saving Money, Reducing Oil Dependency, and Helping the Planet. Ethanol is an alternative to gasoline. The use of ethanol has been demonstrated to reduce greenhouse emissions slightly as compared to gasoline. Through this ebook, you are going to learn what you will need to know why choosing an alternative fuel may benefit you and your future.

Get My Free Ebook

Post a comment