Short-term safety risk assessment of CO₂ geological storage projects in deep saline aquifers using the Shenhua CCS Demonstration Project as a case study

Abstract

Although CO₂ geological storage has been recognized as an effective strategy to lower carbon emissions directly, there are no suitable guidelines for safety risk assessment of CO₂ geological storage projects in deep saline aquifers in China and elsewhere. When CO₂ is injected into deep saline aquifers, stratigraphic and structural trapping is the major basic mechanism controlling CO₂ storage capacity and migration in reservoirs. Therefore, a safety risk assessment method is proposed in this paper using perspectives from hydrogeological and environmental geology. The uncertainties and risks consist of CO₂ leakage, ground deformation, and induced earthquakes. Identifying and assessing potential risks are the first and most important step in the process of risk assessment. Based on the identification of risks of CO₂ geological storage projects, we built an elementary risk evaluation index system in an analytic hierarchy process framework. Meanwhile, the possibility of occurrence and damage to the environment and public caused by CO₂ leakage, ground deformation, and induced earthquakes was analyzed in detail, and current risk criteria were also summarized. Furthermore, using the Shenhua CO₂ Capture and Storage Demonstration Project as a case study, we performed a risk identification and evaluation by using qualitative or semi-quantitative methods in sequence, as well as developing the related preliminary risk management measures. This method and case study for short-term safety risk assessment could provide a guideline for site selection, injection design, and monitoring of CO₂ geological storage projects in deep saline aquifers.

Introduction

CO₂ geological storage has been recognized as an effective strategy to lower carbon emissions directly. Twenty-three million tons of CO₂ are stored each year at eight large-scale integrated CO₂ Capture and Storage (CCS) Projects in the world, and this figure is expected to increase to over 36 million tons of CO₂ per year by 2015 (Global CCS Institute 2012).

There are several large-scale successful CO₂ geological storage projects in deep saline aquifers in the world. The first commercial-scale CO₂ geological storage project, the Sleipner Project operated by Statoil, removes CO₂ from natural gas and injects it into a saline formation under the North Sea. The Shenhua CCS Demonstration Project in the Ordos Basin of China sequesters 100,000 tons of CO₂ per year and is the largest coal-based full-chain CCS project in the world (Wu 2013). CO₂ can be injected into deep geological formations such as depleted oil and gas fields, coal beds, and saline aquifers (IPCC 2005). In contrast to CO₂ storage in deep saline aquifers, CO₂-enhanced oil recovery (CO₂-EOR) and coal bed methane (CO₂-ECBM) are CO₂ geological utilization technologies, where part of the CO₂ is recovered during the project process. With the development of CO₂ geological storage projects in deep saline aquifers in China and elsewhere, potential safety risks, monitoring, and assessment has attracted increasing public attention, and has become a key research challenge (IPCC 2005; Bachu 2008). Intergovernmental Panel on Climate Change (IPCC 2005) presented potential environmental hazards and corresponding risks of CO₂ geological storage. Oldenburg (2008) developed a screening and ranking framework for CO₂ geological storage site selection based on health, safety, and environmental (HS&E) risk. Li et al. (2013) performed applied research on the Shenhua CCS Demonstration Project by using an improved framework. There are no laws or regulations in China suitable for risk assessment of CO₂ geological storage projects in deep saline aquifers. However, “Technical Guidelines for Environmental Risk Assessment on Projects” (HJ/T169-2004), developed by the Ministry of Environmental Protection of China, could provide a basic method for evaluation of CO₂ diffusion.

Risk is the probability or threat of damage, injury, liability, loss, or any other negative event that is caused by external or internal vulnerabilities, and that may be avoided through preemptive action. More formally, risk equates to the product of probability that some events will occur and the consequences of those events if they do occur (Stenhouse et al. 2009). Safety risk assessment is the identification and analysis of the dangerous and harmful factors existing in projects by using safety system engineering theory and methods and estimating the possibility and severity of events in order to propose safety precautions. In a broader context, risk assessment is simply an assessment of the safety of a specific event and the possible harmful consequences (Kaplan 1997), as shown in the following Formula (1):

$$R = P \times C,$$

(1)

where R is the evaluated value of the risk, P is the possibility of the potential risk event, and C is the harmful consequence of the risk event.

Therefore, using a hydrogeological and environmental geological approach and using information from comparative nuclear (Benson et al. 2005; Maul et al. 2007) and solid waste disposal projects (Ayomoh et al. 2008; Korucu and Erdagi 2012), a safety risk assessment method for CO₂ geological storage projects in deep saline aquifers has been developed, emphasizing risk identification of safety status and short-term evaluation. Furthermore, using the Shenhua CCS Demonstration Project as a case study, we performed risk identification and evaluation, and present preliminary risk management measures.

Risk identification

Identifying and assessing potential risks are the first major step in the risk assessment process. Risk identification relies on distinguishing various potential factors in sequence which could cause safety risk events and harmful consequences. The contribution of varying trapping mechanisms of CO₂ geological storage in deep saline aquifers at multiple time scales adds complexity to risk assessment. Once CO₂ is injected into geological formations underground, it can be trapped in the pore spaces by four main processes. The first process is stratigraphic and structural trapping. Another process is residual gas trapping, and Solubility trapping refers to CO₂ dissolving into the fluid of the geological formations, such as water. Finally, CO₂ can react with solid materials and become mineralized (Bachu et al. 1994; Baines and Worden 2004; IPCC 2005; Xu et al. 2005; Akbarabadi and Piri 2011; Liu 2012). However, because stratigraphic and structural trapping is the basic major mechanism (Davison et al. 2004) controlling CO₂ storage capacity and migration characteristic in saline aquifers and geological safety, risk identification should focus on the operational and short-term phase of CO₂ geological storage projects. Similarities can be found in radioactive waste disposal where performance assessment calculations must consider the return of radionucleotides to the accessible environment over periods longer than 10,000 years (Maul et al. 2007).

The objects of risk identification include injection facilities, risk materials, and risk categories. The risks of CO₂ geological storage have been analyzed to some extent for surface equipment to underground reservoirs (IPCC 2005; IEA 2008). Generally speaking, the uncertainties and risks are categorized as CO₂ leakage, ground deformation, and induced earthquakes.

The initiating risk events could be caused externally or internally in CO₂ geological storage projects. Risk events caused by earthquakes, active faults, tectonic ground fissures, fractures within caprocks, and abandoned deep wells around the CO₂ geological storage project site are considered external geological factors. However, unscientific injection design, operation violations, and other human-made factors could be considered internal factors.

CO₂ leakage

Supercritical CO₂ is less dense than water and tends to migrate to the top of reservoirs due to buoyancy. It is inevitable that the highly pressurized CO₂ will leak to some extent due to the permeable nature of the porous rocks, causing uncertainty in the storability of a reservoir (Xie and Economides 2009). Abrupt leakage and gradual leakage are two different types of leakage scenarios (IPCC 2005).

Potential leakage pathways

The main potential leakage pathways are usually recognized are human-made, tectonic, leakage trough the caprocks, and hydraulic traps (IPCC 2005; Pruess 2008; Bachu and Celia 2009; Lemieux 2011). CO₂ may leak along anthropogenically created leakage pathways causing abrupt leakage. Injection, monitoring, and abandoned wells are recognized as potential leakage pathways due to poor or unknown conditions of cementing, alteration, abandonment, or plugging (Gasda et al. 2004; Carey et al. 2007).

Tectonic pathways include faults, fractures, ground fissures, and other tectonic leakage pathways created by earthquakes. Active faults are not only CO₂ leakage pathways, but also damage strata continuity, causing CO₂ leakage through the caprock. Meanwhile, tectonic ground fissures especially near active faults may become CO₂ leakage pathways near the surface. Appropriate site selection could avoid earthquakes and volcanoes, which are recognized as major factors for large-scale CO₂ leakage and global climate change. If CO₂ leaks along these tectonic pathways, it is usually difficult to manage.

CO₂ leakage through the caprock occurs when the capillary entry pressure threshold is reached. While the caprock may have the proper hydrogeological conditions for free phase CO₂ containment, fractures and faults can modify its integrity. These features may naturally exist but geomechanical damage can also occur when the yield strength of the caprock is exceeded by the pressure build-up created by the injection of CO₂ (Rutqvist and Tsang 2002; Rutqvist et al. 2007). A fourth potential escape pathway involves dissolved CO₂ that flows up-dip with regional groundwater flow. It is usually not considered because it is thought to take hundreds of thousands of years before the CO₂ reaches the atmosphere and it is expected that most CO₂ will precipitate before reaching the surface (Lemieux 2011).

Safety and environmental impacts of CO₂ leakage

The safety and environmental impacts of geological storage related to the risk of release of stored CO₂ fall into two broad categories (IPCC 2005): local environmental and safety impacts and global effects resulting from the release of stored CO₂ into the atmosphere. The features, processes, and mechanisms of all of the risk events are closely related to the impact and movement of injected CO₂ underground, and the risks of CO₂ leakage out of the storage complex, either to shallower formations or to the atmosphere (European Commission 2011).

Local environmental and safety impacts

Damen et al. (2003) provided a detailed discussion of potential health, safety, and environmental risks and may be the first paper related to potential environmental impacts from geological CO₂ storage. Saripalli et al. (2003) presented perceived risks to humans and other species (e.g., trees and fish) associated with elevated levels of CO₂ in different media. Subsequently, much research has been performed on the environmental and safety effects of CO₂ geological storage, with a focus on CO₂ leakage.

CO₂ leakage to potable water aquifers can contaminate groundwater. CO₂ in a gas phase can dissolve partially or completely within fresh groundwater. CO₂ itself is not a concern to the water quality of an underground drinking water source, but it will change the geochemical conditions in the aquifers (Lemieux 2011). The dissolution of CO₂ in groundwater increases the total concentration of dissolved carbonate which in turn increases acidity and lowers the natural pH of groundwater (Langmuir 1997). Increased acidity can enhance the dissolution of minerals, including those containing hazardous trace elements (Zheng et al. 2009; Apps et al. 2010). Decreased pH could also mobilize hazardous trace elements adsorbed on clays, iron oxyhydroxides or the surface of other rock-forming minerals (Zheng et al. 2009; Apps et al. 2010). The resultant increase in concentration of hazardous trace elements can affect groundwater quality, possibly to the extent that safe drinking water limits are exceeded. However, in highly buffered aquifers, the potential for a pH decrease can be compensated by the dissolution of alkaline minerals like calcite (Bethke 2008). Under supercritical conditions, CO₂ is a highly effective solvent and capable of extracting contaminants from geologic materials such as polycyclic aromatic hydrocarbons. Those toxic compounds could be mobilized and could compromise water quality in nearby aquifers (Stevens et al. 2000).

Impacts of elevated CO₂ concentrations in the shallow subsurface could include lethal effects on plants and subsoil animals. High fluxes in conjunction with stable atmospheric conditions could lead to local high CO₂ concentrations in the air which could harm animals or people. Pressure build-up caused by CO₂ injection could trigger small seismic events (IPCC 2005).

Global environmental impacts

CO₂ geological storage also has global environmental impacts, in that successful storage will reduce emissions from fossil fuel use and increase its potential as a greenhouse gas emission reduction option. In contrast, high-CO₂ release rates from storage sites would reduce the effectiveness of CO₂ geological storage projects.

Ground deformation

Essentially, the process of CO₂ injection exerts continuous pressure into reservoir rocks underground, increasing pore-fluid pressure, as well as the stress and volume of reservoir rocks. Because of a lower density and smaller coefficient of viscosity than saline water, a supercritical CO₂ plume will migrate driven by injection pressure and pore water buoyancy. The CO₂ plume with associated saline buoyancy and volume expansion force of reservoir rocks would transfer the overlying caprocks and cause deformation in the vertical direction. If the accumulated deformation is large enough, the shallow surface will uplift around the injection well, resulting in ground deformation.

The In Salah Gas project in Algeria involves injection of about 4,000 tons of CO₂ per day into the Krechba Carboniferous sandstone (a 20 m thick, methane producing reservoir) at a depth of 1,800 m near the Krechba gas field. The monitoring of injected CO₂, borehole surveys and geochemical, geophysical as well as geomechanical investigations are still underway. Because of a relatively deep reservoir, a relatively stiff overburden, and with the volume of CO₂ being injected fairly small compared to the overburden, the initial view of the In Salah Project was that no significant ground deformations would occur. However, InSAR data from the first few years of injection show a surface uplift on the order of 5–10 mm per year above active CO₂ injection wells and the uplift pattern extends several km laterally (Vasco et al. 2008a, b). The observed uplift can be explained by pressure changes and associated vertical expansion within the 20-m thick injection zone and the approximately 100-m thick zone of shaly sands immediately above the injection zone. Meanwhile, permeability at injection wells is strongly heterogeneous, affected by the degree of fracturing and perhaps by intersecting faults. Although the ground deformation in In Salah Gas project is a special case because of its sequestration characteristics, it still provides information valuable to all CCS projects.

Induced earthquakes

Deep-well injection of waste fluids may induce earthquakes with moderate local magnitudes (M _L), as in the 1967 Denver earthquakes (M _L of 5.3; Healy et al. 1968; Wyss and Molnar 1972) and the 1986–1987 Ohio earthquakes (M _L of 4.9; Ahmad and Smith 1988) in the United States. Seismicity induced by fluid injection likely results from increased pore-fluid pressure in the hypocentral region of the seismic event (Talebi et al. 1998). Microseismic data analysis together with interpretation of injection data at the In Salah CO₂ storage site provides a valuable tool for improved understanding of the subsurface injection and storage processes. More than 1,500 microseismic events have been detected semi-automatically between August 2009 and May 2012 and the occurrence of the events correlates clearly with increased injection rates and well-head pressures. Most likely the fracture pressure has been exceeded temporarily, resulting in a sudden increase of macroseisms (Oyea et al. 2013).

Once CO₂ injection-induced earthquakes occur near CO₂ geological storage project sites during implementation, it may damage infrastructure, including equipment for CO₂ injection, such as wellbores and monitoring devices. At the same time, the earthquakes may cause more public opposition to CO₂ geological storage projects. Large-scale CO₂ storage in deep saline aquifers may induce macroseisms. Therefore, scientific injection design is important to the success of projects.

Risk evaluation

Risk evaluation index system

The analytic hierarchy process (AHP) is an effective tool for dealing with complex decision making, and may aid the decision maker to set priorities and make the best decision (Saaty 1980; Saaty and Kearn 1985). It has been widely applied in risk or safety assessment because of its comprehensiveness and easy operability (Shi et al. 2009).

By reducing complex decisions to a series of pairwise comparisons, and then synthesizing the results, the AHP helps to capture both subjective and objective aspects of a decision, while at the same time reducing bias in the decision making process. The AHP considers a set of evaluation criteria, and a set of alternative options among which the best decision is to be made. It generates a weight for each evaluation criterion according to the decision maker’s pairwise comparisons of the criteria. Higher weights indicate the higher importance of the corresponding criterion. For a fixed criterion, the AHP assigns a score to each option according to the decision maker’s pairwise comparisons of the options based on that criterion. Higher scores indicate a better performance of the option with respect to the considered criterion. Finally, the AHP combines the criteria weights and the option scores, thus determining a global score for each option, and a consequent ranking. The global score for a given option is a weighted sum of the scores obtained with respect to all of the criteria.

Therefore, based on the risk identification of CO₂ geological storage projects and using information about CO₂ leakage, ground deformation, and induced earthquakes, we analyzed all of the risk factors in order to build an elementary risk evaluation index system using the AHP framework (Table 1). The risk events for CO₂ geological storage projects in deep saline aquifers include CO₂ leakage, ground deformation, and induced earthquakes. The risk factors of risk events are injection implementation, wellbore integrity and devices, geological safety conditions, and mineral resource exploitation. However, our simplified evaluation used risk evaluation indexes for ground deformation and induced earthquakes only, because damage to the public and the environment for these two factors is more immediate than damage from CO₂ leakage.

Table 1 Elementary risk evaluation index system

Risk evaluation method

Occurrence probability of risk events

Although there are several demonstration or commercial CO₂ storage projects in deep saline aquifers, research on environmental impacts and safety risks is limited. We could only perform a risk evaluation by using qualitative or semi-quantitative methods. As is shown in Table 2 (Li et al. 2006), we transformed the qualitative descriptions of occurrence probabilities of risk events to quantitative values.

Table 2 Qualitative descriptions of occurrence probabilities and corresponding quantitative values

Damage degree classification of consequence of risk events

The risk evaluation set is the collection of possible consequences caused by risk factors. We can classify the degree of damage as a consequence of risk events and transform qualitative descriptions to quantitative values, by using relationships shown in Table 3 (Pruess 2008).

Table 3 Damage degree classification of consequence of risk events and corresponding quantitative values

Risk calculation

Based on the transformed quantitative values of weights, occurrence probabilities, and damage degrees of risk factors, we could evaluate risk by using Formulas (1) and (2). Therefore, we could analyze potential risk events and maximum risk in order to make emergency precautions and avoid risk event occurrence.

$$R = \sum\limits_{i = 1}^{n} {R_{i} A_{i} } \quad (i = 1,2,3, \ldots ,n),$$

(2)

where R is total evaluated risk, R _i is the evaluated risk of factor i, n is the number of risk factors, A is the weight of risk factors, and A _i is the weight of risk factor i.

Risk criteria

Acceptable risk criteria for individuals

Acceptable risk criteria for individuals indicate the risk level of people around the CO₂ geological storage project, with maximum and minimum values usually provided. The China Academy of Safety and Technology (2006) proposed acceptable risk criteria for individuals suitable for China based on study of the rules and regulations published by authoritative governments abroad (Table 4).

Table 4 Object types affected by risk events and acceptable risk criteria for individuals

Individuals affected by risk events could be divided into three types (Table 4) based on sensitivity and population density. Values of tolerable and negligible risk per year for typical objects are recommended. The evaluated risk of CO₂ geological storage project in highly sensitive and high-population density areas should not be larger than 0.5 × 10⁻⁵ per year, but could be neglected if lower than 1 × 10⁻⁷ per year. Similarly, the largest tolerable risk per year in moderately sensitive and medium population density areas is 1 × 10⁻⁵, and the largest negligible risk per year is 1 × 10⁻⁶. The largest tolerable and negligible risk per year is 5 × 10⁻⁵ and 1 × 10⁻⁵, respectively, in minimally sensitive and low-population density areas.

Acceptable risk criteria for the public

Acceptable risk criterion for the public is used to reduce accidental risks to the public. As shown in Fig. 1 (China Academy of Safety and Technology 2006), two lines represent the maximum and minimum risks, respectively. The F–N graph provides acceptable, unacceptable, and ALARP areas of risk criteria for the public in the Netherlands, Great Britain, and Hong Kong.

Fig. 1

Acceptable risk criteria for the public in the Netherlands, Great Britain, and Hong Kong

Other risk criteria

The purpose of CO₂ geological storage projects is to reduce CO₂ emissions and project the earth’s environment, so it is necessarily to develop risk criteria associated with CO₂ leakage, ground deformation, and induced earthquakes, which could provide more information for risk assessment in CCS projects.

Acceptable and limiting CO₂ leakage rates

While the goal of geological CO₂ storage is to store injected CO₂ underground permanently, there are likely to be projects where some CO₂ leakage occurs. Acceptable rates of leakage in the past have been expressed as a percentage of the total volume injected and typically range from 0.01 % per year (1 % over 100 years) to 0.001 % per year (1 % over 1,000 years) (Bowden and Rigg 2005; Shuler and Tang 2005). For a CO₂ storage site which annually releases 0.001 of the amount stored, effectiveness is around 60 % after 1,000 years. This rate of release would be equivalent to a fraction retained of 90 % over 100 years or 60 % over 500 years. It is likely that, in practice, geological and mineral storage would have lower rates of release than this and hence higher effectiveness—for example, a release rate of 0.01 % per year would be equivalent to a fraction retained of 99 % over 100 years or 95 % over 500 years (IPCC 2005).

The performance assessment carried out during Phase 1 of the IEA WeyburnCO₂ Monitoring and Storage Project focused on the capabilities of the reservoir to contain the injected CO₂ (IEA GHG 2008). In terms of the expected evolution of the storage system (Base Case) and possible CO₂ migration via natural leakage pathways, the geology would keep the CO₂ underground for at least 5,000 years. In terms of human-made pathways (abandoned wells), a small amount of leakage was predicted over 5,000 years, with the mean value for maximum leakage rate (for several hundred wells) of 4 × 10⁻⁴ kg/day, with 95 % of the simulations yielding a value <1.6 × 10⁻³ kg/day (Zhou et al. 2005).

CO₂ concentration

Rice (2013) suggested that CO₂ concentrations in the range of 0.5–1.5 % are well tolerated by healthy humans, and “prolonged exposure to CO₂ concentrations of 1 % may significantly affect health on the general population.” (A limit of 0.35 % CO₂ in indoor air is recommended by Health Canada).

CO₂ is not considered as a pollutant in the current legal system of China; however, the “Occupational exposure limits for hazardous agents in the work place—Chemical hazardous agents” (GBZ 2-2002) enacted by the National Health and Family Planning Commission of China states that CO₂ concentration exposure should not be larger than 18,000 mg/m³ for a short time. “Indoor air quality standard” (GB/T18883-2002) also enacted by the National Health and Family Planning Commission of China recommends the indoor average CO₂ concentration of 0.1 %.

Case study

Risk identification and engineering analysis

The Shenhua CCS Demonstration Project is the first CCS project in the Ordos Basin of China, and is the largest coal-based full-chain project of CO₂ capture and geological storage in deep saline aquifers in the world. There is a single injection wellbore called the Zhongshenzhu 1# wellbore and two monitoring wells called the Zhongshenjian 1# and Zhongshenjian 2#. A total of 167 kilo-tons of CO₂ was injected into the deep saline aquifers of Mesozoic sandstone and Paleozoic Majiagou carbonate rocks from September 2011 to November 2013 (Wang 2013).

Shenhua CCS Demonstration Project is located in the northeast of Yimeng uplift of Ordos Basin. 3D seismic exploration (Fig. 2) and drilling data (Fig. 3) show that the formations are sloping from northeast to southwest slowly at the dip, around 1° as a gentle southward monoclinal structure (Wu 2013; Guo et al. 2014; Wei et al. 2014). And the stratigraphy sequences from bottom to top are as follows: Lower Ordovician Majiagou Formation (O₁ m); Middle Carboniferous Benxi Formation (C₂ b) and Upper Carboniferous Taiyuan Formation (C₃ t); Lower Permian Shanxi Formation (P₁ s), Lower Permian Shihezi Formation (P₁ sh), Upper Permian Shiqianfeng Formation (P₂ sh); Lower Triassic Liujiagou Formation (T₁ l) and Heshanggou Formation (T₁ h), Middle Triassic Zhifang Formation (T₂ z), Upper Triassic Yanchang Formation (T₃ y); Lower Jurassic Yan’an Formation (J₁ y), Middle Jurassic Zhiluo Formation (J₂ z) and An’ding Formation (J₂ a); Lower Cretaceous Dongsheng-Luohandong Formation (K ^d-l₁ ); Quaternary (Q).

Fig. 2

Typical profile of 3D vertical seismic profiling (VSP) seismic exploration

Fig. 3

Reservoir-seal assemblages of the Shenhua CCS Demonstration Project

Using data from geological surveys, 3D seismic exploration, drillings, and numerical modeling, risk identification and engineering analysis were performed as follows.

Potential leakage pathways and damage

Injection operation

To date, the Shenhua CCS Demonstration Project runs well, with monitoring showing no CO₂ leaks or related environmental hazards (Li et al. 2013). As the first deep saline aquifer demonstration project, the Shenhua CCS project operates under scientific management regulations, and has made emergency precautions in order to avoid potential risk event occurrence and damage.

Wellbore integrity and devices

The injection and monitoring wellbores were constructed strictly in accordance with related rules, and monitoring results suggested that devices run well. CO₂ leakage along this kind of leakage pathways is a low possibility.

Geological safety

Historical earthquakes: The peak ground acceleration of the tectonic area in which the Shenhua CCS Demonstration Project is located is 0.05 g, and the basic seismic intensity is VI. Only a few small earthquakes occurred near the injection site with no earthquake disasters on record, and the crust is stable.

Fracture structures and activity: Data from a ground geological survey and 3D seismic exploration show no active faults near the project site. Although there are several structural fractures developing shallowly, the fault throws are small and no faults are found near the Zhongshenzhu 1# well. There are no obvious fractures in the core samples from the injection well.

Sealing capacity: We identified three main regional caprocks in the Shenhua CCS Demonstration Project, including the Benxi Formation mudstone, the upper part of Shihezi Formation and lower part of Shiqianfeng Formation mudstone, and the Heshanggou and Zhifang Formation mudstone, with large thickness, good continuity, and sealing ability (Fig. 3). Furthermore, if CO₂ breaks through the middle part of the Zhifang Formation, which is the most important regional mudstone seal as mentioned above, there are two reservoir-seal assemblages below 800-m depth which could provide buffering—the upper part of Zhifang Formation sandstones—top of Zhifang Formation mudstones and the bottom of the Yanchang Formation sandstones and mudstones above. The monitoring data of the Zhongshenjian 2# wellbore, the bottom of which is above all the reservoirs, show no CO₂ breakthrough of the Heshanggou and Zhifang Formation mudstone. Compared with the caprocks of CO₂ gas fields in Lishui Sag of the East China Sea Shelf Basin, where the influence thickness in the vertical direction is <10 cm because of corrosion of supercritical CO₂ on carbonate minerals, the major caprocks of Shenhua CCS Demonstration Project should provide effective sealing for injected CO₂.

However, although the 3D seismic exploration (Fig. 2) and drillings (Fig. 3) show that the mudstone continuity in the major caprocks is sound (Wu 2013; Li et al. 2013; Guo et al. 2014), the fluvial depositional environment of Mesozoic strata and heterogeneity could not ensure that there is no possibility of CO₂ breakthrough of caprocks.

Tectonic ground fissures: The topographic change activities are relatively small in general, and there are no tectonic ground fissures near the injection well.

Hydrodynamic condition: The reservoirs of the Shenhua CCS Demonstration Project involve two aquifer systems, which are Cambrian-Ordovician carbonate saline aquifers and Carboniferous-Jurassic sandstone saline aquifers. Both of them mainly accept surface water and shallow groundwater recharge flowing along the dip direction from the east to the center of the Ordos Basin. Therefore, the hydrodynamic condition could play a prominent part in blocking the CO₂ plume from migrating to the east, which is suitable for CO₂ geological storage. Liu et al. (2013) used the ECO₂N module of TOUGH2 to simulate flow and pressure configurations in response to small-scale CO₂ injection into multilayer saline aquifers, and the results showed that the lateral distance reached by the CO₂ plume was limited to within a radius of 200 m from the injection point, and its vertical movement was restricted by the low permeability siltstone and mudstone layers. Generally speaking, the distance of CO₂ migration in the lateral direction driven by hydrodynamic force and injection pressure is limited. The CO₂ plume should not reach faults below 1,000 m underground around the Zhongshenzhu 1# well.

Geological hazards: landslide, mudflow, and rockfall were not found in Shenhua CCS Demonstration Project site by ground geological survey, suggesting the site is suitable for engineering construction.

Mineral resources exploitation

There was no abandoned well identified by the geological survey data within 100 m. The Shenhua CCS Demonstration Project is located in the coal pillar of the Yan’an Formation at a depth of 400 m. About 10 km away from the Shenhua CCS Demonstration Project in the east, the nearest ground subsidence area induced by coal mining is obvious. Therefore, together with coal mining, subsidence may threaten the safety of injection and monitoring wells, which could cause CO₂ leakage.

Ground deformation

Numerical stimulation

Li (2012) established a coupling analysis model for the Shenhua CCS Demonstration Project considering hydraulic field and pressure field, and performed a numerical simulation based on the Biot equation (Biot 1941). The numerical simulation result is shown in Fig. 4, we could conclude that the ground deformation in the Shenhua CCS Demonstration Project site was small and difficult to discern in the first 10 years. Although ground deformation is more and more obvious along with continuous CO₂ injection from the first to tenth years, the maximum deformation displacement in the vertical direction is <5 × 10⁻⁹ mm. Ground deformation displacement near the injection well is affected by project construction, but CO₂ injection into deep saline aquifers subjected the ground to uplift overall. However, surface subsidence occurs 3,000–5,000 m away from the injection well as well, but is not obvious.

Fig. 4

Numerical simulation results for ground deformation based on hydraulic and pressure fields

Monitoring using D-InSAR technology

As one of the key members in the Shenhua CCS group, China geological survey center for hydrogeology and environmental geology (CHEGS) has carried out ground deformation monitoring four times using D-InSAR technology from 2010 to 2011. There are ten obvious ground deformation areas, with the nearest one being ~10 km away from the injection well and there is no obvious ground deformation in the Shenhua CCS Demonstration Project site (Fig. 5). The ten obvious ground deformation areas are mainly caused by coal mining which has been confirmed by geological survey.

Fig. 5

Ground deformation monitoring using D-InSAR technology

Induced earthquakes

Based on geological analysis, many researchers (Wu 2013; Guo et al. 2014) evaluated the CO₂ geological storage potential by numerical simulation in Shenhua CCS Demonstration Project. And all the results suggested that the reservoirs in the Shenhua CCS Demonstration Project site could handle 10 × 10⁴ t/a of CO₂ geological storage per single injection well for 3 years. After 167 kilo-tons of CO₂ injected, Liujiagou sandstone reservoir becomes the most effective reservoir with the largest capacity without hydraulic fracturing. It has stored over 90 % of the CO₂ injected, which reflects the superiority of sandstone saline aquifers for CO₂ geological storage (Wang 2013). Therefore, there is still much potential space for CO₂ geological storage, without any felt earthquakes before.

Meanwhile, because of discontinuous captured CO₂ production from Shenhua coal to liquid and chemical factory and the safety of reservoirs, the injection rates are kept low. Although monitoring for small seismic events has not been conducted before, the rational CO₂ injection rate and pressure could ensure that induced earthquakes will not be likely to occur.

Risk evaluation

Risk calculation

Weight evaluation

Comprehensively considering the geological survey, injection, monitoring, and numerical simulation results of the Shenhua CCS Demonstration Project in risk identification mentioned above, we improved the risk evaluation index system outlined in Table 1, as illustrated in Table 5. The hierarchy model was classified into four layers, i.e., goal layer, level 1 criterion layer, level 2 criterion layer, and level 3 criterion layer. The goal is to evaluate the risk of Shenhua CCS Demonstration Project. At level 1, there exist three evaluation criteria which are risk events of CO₂ leakage, ground deformation, and induced earthquakes. However, ground deformation and induced earthquakes were simplified to evaluate the corresponding risk. At level 2, CO₂ leakage in turn consists of injection implementation, wellbore integrity and devices, geological safety conditions, and mineral resources exploitation. At level 3, injection implementation consists of overlarge injection design and irregular operation. Similarly, we can classify the risk events, risk factors, and sub risk factors in different levels.

Table 5 Index system for risk evaluation of Shenhua CCS Demonstration Project

Taking into account the contribution of each risk factor to risk size of each corresponding risk event, and the contribution of each event to total risk, we calculated the weights of risk factors using AHP in the following four steps (Saaty 1980, 1985; Spires 1991; Abdullah et al. 2013).

• Step 1 Construct the hierarchical structure and obtain normalized matrix.

  First, the criteria are compared with respect to the goal. (a _ij )n × n matrix, denoted as A, is created using the pairwise comparisons with the elements a _ij indicating the value of criterion i relative to criterion j, as shown in the following Formula (3).

  $$A = \left[ {\begin{array}{*{20}l} {a_{11} } \hfill & {a_{12} } \hfill & {a_{13} } \hfill & \cdots \hfill & {a_{1n} } \hfill \\ {a_{21} } \hfill & {a_{22} } \hfill & {a_{23} } \hfill & \cdots \hfill & {a_{2n} } \hfill \\ \cdots \hfill & \cdots \hfill & \cdots \hfill & \cdots \hfill & \cdots \hfill \\ {a_{n1} } \hfill & {a_{n2} } \hfill & {a_{n3} } \hfill & \cdots \hfill & {a_{nn} } \hfill \\ \end{array} } \right]_{n \times n}$$

  (3)

  The values a _ij are obtained by the a _ii  = 1, a _ij  = 1/a _ji , where a _ij  > 0, for all i. Therefore, if a number is assigned to element i when compared to element j, then j has the reciprocal value when compared with i. Second, its entries are normalized by dividing them by their sum. This is repeated for all columns to obtain the normalized matrix A (A _norm) as follows.

  $$A_{\text{norm}} = \left[ {\begin{array}{*{20}l} {a_{11} /a_{1}^{{\prime }} } \hfill & {a_{12} /a_{2}^{{\prime }} } \hfill & {a_{13} /a_{3}^{{\prime }} } \hfill & \cdots \hfill & {a_{1n} /a_{n}^{{\prime }} } \hfill \\ {a_{21} /a_{1}^{{\prime }} } \hfill & {a_{22} /a_{2}^{{\prime }} } \hfill & {a_{23} /a_{3}^{{\prime }} } \hfill & \cdots \hfill & {a_{2n} /a_{n}^{{\prime }} } \hfill \\ \cdots \hfill & \cdots \hfill & \cdots \hfill & \cdots \hfill & \cdots \hfill \\ {a_{n1} /a_{1}^{{\prime }} } \hfill & {a_{n2} /a_{2}^{{\prime }} } \hfill & {a_{n3} /a_{3}^{{\prime }} } \hfill & \cdots \hfill & {a_{nn} /a_{n}^{{\prime }} } \hfill \\ \end{array} } \right]_{n \times n},$$

  (4)

  where a _ij in the above matrix is the pairwise comparisons of ith row relative to ith column, $a_{n}^{{\prime }}$ is the sum of the pairwise comparisons in the ith column.

• Step 2 Find the criteria weight and geometric means of $\sqrt[n]{{\mu_{i} }}$ , respectively.

  $$\sum\limits_{i = 1}^{n} {\mu_{i} } = \sum\limits_{i = 1}^{n} {\sum\limits_{j = 1}^{n} {a_{ij} } }$$

  (5)

• Step 3 Find the eigenvector by normalized the pairwise comparisons, and calculate the weights (W) in different level layers.

  $$W_{i} = \frac{{\sqrt[n]{{\mu_{i} }}}}{{\sum\nolimits_{i = 1}^{n} {\mu_{i} } }}$$

  (6)

• Step 4 Check the consistency ratio (CR), the comparison matrix will be considered to be consistent if there exist CR < 1.

  Calculate the maximal latent root λ _max.

  $$\lambda_{\hbox{max} } = \sum\limits_{i = 1}^{n} {\frac{{(AW)_{i} }}{{nW_{i} }}}$$

  (7)

  Calculate the coincidence indicators (CI)

  $${\text{CI}} = \frac{{\lambda_{\hbox{max} } - n}}{n - 1}$$

  (8)

  Check the CR. The CR is consistent when its value is <0.1.

  $${\text{CR}} = {\text{CI}}/{\text{CR}}$$

  (9)

  When RI is the random index and depends on the number of element being compared, n and takes on the values are shown in Table 6.

  Table 6 The value of RI

Risk calculation

Based on Tables 2 and 3, we evaluated the occurrence probabilities and degree of damage of all risk factors to give corresponding quantitative values.

Finally, we performed the risk calculation by using Formula (10) improved from Formula (1) and Formula (2), and the detailed risk values are shown in Table 7.

$$\begin{aligned} R &= R_{a} \cdot W_{a} + R_{b} \cdot W_{b} + R_{c} \cdot W_{c} = \sum\limits_{i = 1}^{4} {R_{{a_{i} }} \cdot W_{{a_{i} }} } + P_{b} \cdot C_{b} \cdot W_{b} + P_{c} \cdot C_{c} \cdot W_{c}, \hfill \\ R_{{a_{ij} }} &= \sum\limits_{j = 1}^{m} {P_{{a_{ij} }} \cdot C_{{a_{ij} }} \cdot W_{{a_{ij} }} } \quad (j = 1, \ldots ,m;\;m\;{\text{is}}\;{\text{the}}\;{\text{number}}\;{\text{of}}\;{\text{sub}}\;{\text{factors}}) \hfill \\ \end{aligned}$$

(10)

where R is the total evaluated risk, R _a is the evaluated risk of CO₂ leakage, R _b is the evaluated risk of ground deformation, R _c is the evaluated risk of induced earthquakes, $R_{{a_{i} }}$ is the risk of factor i, $R_{{a_{ij} }}$ is the risk of sub factor ij, P _b is the occurrence probability of ground deformation, P _c is the occurrence probability of induced earthquakes, P _ij is the occurrence probability of sub factor ij, W _a is the weight of CO₂ leakage, W _b is the weight of ground deformation, W _c is the weight of induced earthquakes, $W_{{a_{i} }}$ is the weight of factor i, $W_{{a_{ij} }}$ is the weight of sub factor ij, C _b is the harmful consequence of ground deformation, C _c is the harmful consequence of induced earthquakes, and $C_{{a_{ij} }}$ is the harmful consequence of sub factor ij.

Table 7 Risk calculation result of the Shenhua CCS Demonstration Project

Risk analysis

Inferred by the analysis on the quality of the reservoir, geological safety, and public and environmental conditions, the Shenhua CCS Demonstration Project is suitable for CO₂ geological storage. Based on the object types affected by risk events and the acceptable risk criteria for individuals (Table 4) and acceptable risk criteria (Fig. 1), we can conclude that the total evaluated risk of the Shenhua CCS Demonstration Project is 6.47E−8, which is within the allowable range for the public and could be negligible for individuals.

As shown in Table 7, the possible maximum risk event is CO₂ leakage, and the most likely risk factor is mineral resource exploitation, as ground deformation or ground fissures caused by coal mining may damage the integrity of injection and monitoring wellbores and devices. In addition, the risk of CO₂ breakthrough of the caprocks, causing leakage, should be given more attention by all CO₂ geological storage projects as well.

Risk management

Improve the quality of injection

Injection pressures, potentials, and rates that are too large may cause risk, even in abandoned reservoirs. More numerical simulation and injection tests are needed to better guide CO₂ injection. Meanwhile, workers should implement injection methods strictly according to current scientific information and management regulations in order to avoid human-made risk events.

Improving monitoring

Because CO₂ migration characteristics are affected by geological conditions, monitoring of the CO₂ plume in the subsurface is important to the success of the project, providing key data for effectiveness, safety, and continuity of the sequestration project. Monitoring should be improved to identify risk and possible CO₂ leakage pathways. Tracking the distribution of trapped CO₂ in the fluid, dissolved and solid phases are needed for plume confirmation, leakage detection, and regulatory oversight. Existing monitoring methods include well testing and pressure monitoring, use of chemical tracers, chemical sampling, surface and borehole seismic analysis, electromagnetic, and other geotechnical instruments (Benson and Myer 2002; Klara et al. 2003; Sato 2006; Saito et al. 2007; Loizzo et al. 2011; Wiese et al. 2013; Eshiet and Sheng 2014). The spatial and temporal resolution of current methods is unlikely to be sufficient for confirmation of performance and leakage detection. Remote sensing requires high-resolution mapping techniques, such as Interferometric Synthetic Aperture Radar (InSAR) (Pritchard et al. 2014; Ramirez and Foxall 2014), for tracking migration of sequestered CO₂ and its by-products as well as deformation and microseismicity monitoring (Rutqvist et al. 2009).

Taking precautions

The purpose of CO₂ geological storage is to reduce CO₂ emissions and slow climate change. If large risk events occur, remedy methods should be performed to lower or avoid damage. Precautions should be taken in the early stage of CCS project implementation, and should be accepted by experts and the public.

While there is limited experience with geological storage, closely related oil industrial experience, nuclear (Benson et al. 2005; Maul et al. 2007) and solid waste disposal research (Ayomoh et al. 2008; Korucu and Erdagi 2012) could serve as a basis for appropriate risk management, including remediation. The effectiveness of the available risk management methods still needs to be demonstrated for use with CO₂ storage. If leakage occurs at a storage site, remediation to stop the leakage could involve standard well-repair techniques or the interception and extraction of the CO₂ before it would leak into a shallow groundwater aquifer (IPCC 2005).

Conclusions

As an environmental project to reduce greenhouse gas emissions, safety risk assessment of CO₂ geological storage projects in deep saline aquifers should consider risk events including CO₂ leakage, ground deformation, and induced earthquakes. In this paper, a safety risk assessment method, based on hydrogeology and environmental geology information, is proposed, emphasizing the risk identification of safety status and short-term evaluation. Although the method is only based on qualitative or semi-quantitative information, it can be used to preliminarily assess whether the selected site for a project is suitable. Using the Shenhua CCS Demonstration Project as a case study, risk identification of potential leakage pathways could provide a site selection framework. However, more numerical simulations and experiments need to be used to improve occurrence probabilities of risk events. Additionally, damage to the environment and the public caused by risk events should be further studied in comparison to natural CO₂ emissions, such as Mammoth Mountain in California and Nyos Lake in Cameroon. Furthermore, the criteria of risk events are vitally important for CCS projects in regard to CO₂ emissions, and risk assessment methods should be in accordance with any laws and regulations.