Study on the Influencing Factors of CO₂ Storage in Low Porosity-Low Permeability Heterogeneous Saline Aquifer

Abstract

The safety and long-term storage capacity of CO₂ geological storage are necessary factors for project design and engineering development. Evaluating the influencing factors of CO₂ storage and quantitatively analyzing the sensitivity of each parameter have an important guiding role in the design and development of storage projects. In this paper, the Liujiagou Formation in the northeast of the Ordos Basin is taken as an example. Based on the TOUGH/Petrasim simulation tool, the RZ2D geological storage model is established. Seven influencing factors, namely salinity, temperature, horizontal and vertical permeability ratio, pore geometry factor, residual gas saturation, liquid saturation and pore compression coefficient, were compared and analyzed to control the plume migration behavior, interlayer pressure accumulation and storage capacity of low porosity and low permeability heterogeneous reservoirs, and the sensitivity of each parameter to interlayer pressure and storage capacity was quantitatively analyzed. The simulation results show that the uncertain factors affect the safety of CO₂ geological storage to a certain extent by affecting the speed of the residual storage and dissolution storage mechanism. High residual gas saturation and salinity will make CO₂ mostly exist in the form of free state, which will adversely affect the safety and storage capacity of CO₂ saline aquifer storage. High temperature and high vertical permeability ratio will lead to higher interlayer pressure accumulation, which is not conducive to the safety of the storage project but is beneficial to the storage capacity. Temperature, transverse and longitudinal permeability ratio and pore geometry factor control the propagation velocity of plume. The larger these factors are, the faster the plume velocity is. Higher liquid phase saturation is not better; higher liquid phase saturation leads to a large build-up of pressure in the reservoir and can have an adverse effect on the storage volume. The sensitivity analysis of all factors shows that the liquid saturation and temperature have the greatest influence on CO₂ geological storage, and the pore compression coefficient has the least influence. The conclusions of this paper can provide a theoretical reference for the design and development of a CO₂ saline aquifer storage project in a low porosity and low permeability reservoir area.

1. Introduction

Since the 20th century, in order to deal with the increasingly serious environmental problems caused by greenhouse gas emissions, countries are actively exploring ways and means of energy conservation and emission reduction. As a responsible big country, China has committed to peaking carbon emissions by 2030 and achieving carbon neutrality by 2060 [1,2]. Carbon dioxide (CO₂) capture, utilization and storage (CCUS), as a key technology for ‘carbon reduction’, is one of the necessary ways for China to achieve carbon neutrality [3,4]. This technology is mainly divided into two aspects. The front end focuses on the source of CO₂ emissions and realizes the capture and utilization of CO₂ based on engineering innovation and the use of catalysts [5,6,7], and the back end focuses on the closed storage of CO₂. As a direct, effective and large-scale carbon emission reduction method in CCUS, deep saline aquifer sequestration has attracted the attention of a large number of scholars [8,9,10]. However, the injection of a large amount of CO₂ into the deep saline aquifer is bound to change the original geological balance and cause a series of dynamic phenomena in the formation, including large-scale space, time-range pressure conduction, multiphase fluid migration and local formation physical changes. These phenomena will have a significant impact on the safety and stability of the cap and reservoir, thus affecting the effectiveness of the entire storage system. At the engineering scale, the safety of the storage system and the risk and capacity of long-term storage are the most concerning issues for many scholars on CO₂ geological storage [11,12,13]. However, the long-term prediction and risk assessment of CO₂ injection will be affected by many geological system uncertainties. For the storage of saline aquifers, these factors are derived from the uncertainty of the properties and conditions of the deep saline aquifers, including the heterogeneity of the saline aquifers, the uncertainty of the existing data, and the randomness of the stratigraphic characteristics [14,15,16]. Therefore, in order to improve the design of the storage project and ensure the safety and effectiveness of the CO₂ project, it is necessary to evaluate the sensitivity of the influencing factors of CO₂ storage in advance.

Many scholars have made useful discussions on the influencing factors of CO₂ geological storage, including formation physical properties, simulation parameters and well location implementation plan [17,18,19,20,21,22,23,24,25,26].

Dai et al. studied seven factors affecting the CO₂ storage capacity of the Chang 6 reservoir in the Yanchang Formation of the Ordos Basin. The results show that the sensitivity of the factors affecting the CO₂ capture capacity is in the order of formation stress, temperature, residual gas saturation, horizontal permeability and porosity. The sensitivity of the factors affecting the CO₂ dissolution and capture capacity is formation stress, residual gas saturation, temperature, horizontal permeability and porosity [17]. Taking the Liaohe Basin in China as an example, Mkemai and Gong explored the influence of changes in reservoir water saturation and horizontal and vertical permeability ratios on the reservoir performance of CO₂ storage. The results show that the higher the water saturation of the reservoir, the lower the amount of CO₂ contained in the reservoir. The greater the ratio of Kv/Kh, the more favorable the interaction with formation brine, thus promoting the dissolution of CO₂ [18]. Sarkarfarshi et al. conducted a sensitivity analysis of the influence of different geological parameters on the uncertainty of predicting the evolution of CO₂ plume based on the Nisku aquifer stratigraphic data in Alberta, Canada. The analysis results show that the formation porosity, residual brine saturation and capillary entry pressure are the main parameters affecting the uncertainty of plume evolution [19]. Based on the Qianjiang Depression in the Jianghan Basin, Li et al. discussed the influence of formation water salinity on the process of CO₂ storage. The results show that with the increase in the formation water salinity, the dissolution of CO₂ is significantly reduced, and the storage safety and continuous injection capacity of CO₂ are also reduced [20]. Moreover, using the Petrasim/TOUGHREACT-ECO2N module, Zhang et al. studied the effects of five geological parameters, i.e., formation brine salinity, formation brine composition, formation pressure, formation temperature and permeability, on the CO₂ sequestration in saline aquifers. The results show that the dissolution mass of CO₂ decreases with the increase in pressure and temperature, while the change in gaseous CO₂ saturation is the opposite. The lower the formation pressure and temperature, the more the permeability of the salt water layer is more conducive to CO₂ sequestration [21]. Yoshida et al. used a two-dimensional uniform model to study the influence of the uncertainty of the reservoir model on CO₂ sequestration and analyzed the results based on individual parameter sensitivity studies and Monte Carlo simulation. He believes that the measurement of the end point of CO₂ relative permeability is very important for accurate reservoir modeling [22]. Yang et al. studied the effects of reservoir size, well pattern, injection rate, reservoir heterogeneity, anisotropy ratio and permeability sequence on CO₂ storage capacity and efficiency by using the composition simulator Computer Modeling Group-General Equation of State Model. The simulation results show that simultaneous water intake during CO₂ injection greatly improves the storage capacity and efficiency of CO₂. A certain degree of heterogeneity or anisotropy is conducive to the storage of CO₂. High injection rate is beneficial to reservoir capacity but reduces reservoir efficiency and CO₂ breakthrough time, thus limiting the total amount of CO₂ injection [23]. Zheng et al. used TOUGH2/ECO2N to establish a 2D model to study the effects of seven parameters, including horizontal permeability, porosity, residual liquid saturation, pore distribution index, compressibility, reciprocal of inlet pressure and salinity, on the pressure at the injection well, the total amount of gas-phase CO₂ and the diffusion distance of the CO₂ gas plume. The results show that the sensitivity of porosity is the highest when the pressure at the injection well is the response variable. When the total amount of CO₂ and the diffusion distance of CO₂ plume are taken as the response variables, the sensitivity of horizontal permeability is the highest [24]. In addition, the difference of injection direction and injection location of wells will also affect the storage of carbon dioxide in saline aquifers [25,26,27].

In summary, although the predecessors have done a lot of research on the influencing factors of CO₂ geological storage, it has effectively promoted the progress of CCS engineering. However, most scholars focus on the influence of individual factors on the storage capacity of CO₂ storage. Only a few scholars have comprehensively analyzed the influencing factors of CO₂ geological storage, but they have neglected the influence of storage factors on reservoir pressure accumulation. Moreover, the uncertainty of the influencing factors of most studies comes from the hypothesis, not from the actual properties and matching the physical property model of the field. In addition, little attention has been paid to the sensitivity of factors affecting CO₂ geological storage in low-porosity and low-permeability heterogeneous reservoirs.

Controlled by the unique energy industry structure, the carbon emission intensity of Yulin City in 2020 is four times the national average. The total energy consumption is large, and the carbon emission is large. ‘Carbon emission reduction’ is urgent. Based on this, this study takes the Liujiagou Formation in the Yulin area of the northeastern Ordos Basin as an example to study the effects of salinity (w_B), pore compressibility (C), permeability ratio (Kxyz), temperature (T), pore size distribution parameter (λ), residual gas saturation (Slr) and liquid saturation (Sls) on CO₂ storage in low-porosity and low-permeability saline aquifers, including plume migration behavior, interlayer pressure accumulation and CO₂ storage capacity. The purpose of this study is to compare the control of different uncertain parameters on CO₂ storage in deep saline aquifers through sensitivity analysis, so as to quantitatively evaluate the influence of the above parameters on the long-term prediction and risk assessment of CO₂. It provides a theoretical basis for the influencing factors of CO₂ storage in low-porosity and low-permeability saline aquifers and provides some guidance for the design and implementation of CO₂ storage projects in Yulin and other low-porosity and low-permeability saline aquifers in the future.

2. Regional Geology

The study area in this paper is the Liujiagou Formation in the Yulin area, northeastern Ordos Basin. The Ordos Basin is located in the western part of the North China Block (Figure 1) and is a typical multi-cycle craton basin [28]. In terms of tectonics, the Ordos Basin is divided into six secondary tectonic units [29], namely the Yimeng uplift, the Weibei uplift, the Jinxi flexural fold belt, the Yishan slope, the Tianhuan depression and the western margin thrust belt. The Yulin area is located in the northeastern part of the Yishan slope in the Ordos Basin. The Yishan slope is the main structure of the basin. The internal structural form has the characteristics of low amplitude [30]. The overall performance is high in the northeast and low in the southwest. Tilted structures and local fault structures are less developed [31].

The development of the Liujiagou Formation in the Yulin area is deeply influenced by the structure of the Yishan slope, and the deposition is relatively sTab. Based on 228 stratigraphic depth wells in the study area, the Liujiagou Formation is slowly inclined from east to west, the burial depth range is 1023 m~2591 m and the sedimentary thickness range is 122 m~488 m (Figure 2).

During the sedimentary period of the Early-Middle Triassic Liujiagou Formation, affected by the gradual closure of the Paleo-Tethys Ocean, the North China Plate was in an arid zone, and the climate was very dry and hot [33]. During this period, the northeastern Ordos Basin mainly developed a shallow river-delta environment [34]. According to the drilling data of the Gaojiabao exploration area in Shenmu County, Yulin area, the Liujiagou Formation in the study area is mainly interbedded with sandstone and mudstone. The sandstone is mainly argillaceous sandstone, siltstone and fine sandstone. On the whole, the grain size is fine, mainly gray white and light flesh red. In order to determine the physical conditions of the formation, the physical properties of the regional formation are inverted based on the exploration well in the study area. The porosity is 1~10.18%, and the permeability is 0.11 mD~2.3 mD. It is a low porosity and low permeability formation, and the physical properties of sand and mudstone are not obvious. The bottom of the Liujiagou Formation is the dominant reservoir, and the mudstone, siltstone and clay in the middle and upper parts are interbedded, which is a favorable cap rock (Figure 3), and the overlying Heshanggou Formation is a cap rock widely developed in the region [35,36], which can effectively trap CO₂.

3. Research Methods and Schemes

3.1. Simulation Method

The research tool of this paper is TOUGH2/Petrasim-ECO2N module [37,38,39,40]. PetraSim is an interactive preprocessor and postprocessor for TOUGH series code, which can quickly develop models compatible with TOUGH series simulators and view the results of TOUGH series simulators [39]. TOUGH2 is a widely used numerical simulation tool, which is mainly used to simulate the non-isothermal flow of multi-component and multi-phase fluids in porous media and fractured media [38]. ECO2N is a module of TOUGH2, which is designed for the geological storage of carbon dioxide in saline aquifers under certain temperature, pressure and salinity conditions. It can fully describe the thermodynamic and thermophysical properties of the mixture of H₂O, NaCl and CO₂ and model the flow process isothermally or non-isothermally. The phase conditions characterized can include single-phase (containing H₂O or rich in CO₂) and two-phase mixtures. During the simulation process, the fluid phase may appear or disappear, and the solid salt may precipitate or dissolve [37].

The basic principle of TOUGH numerical simulation is based on Darcy’s law of multiphase flow and the conservation of fluid mass energy. The general mathematical equation can be expressed as follows:

$${\displaystyle \frac{d}{dt}}\underset{V_n}{{\displaystyle \int }}{A}^{k}dv=\underset{{\Gamma }_n}{{\displaystyle \int }}{F}^k\cdot nd\Gamma +\underset{V_n}{{\displaystyle \int }}{q}^{k}dv$$

In the formula, $V_n$ is the volume of the control unit; ${\Gamma }_n$ is the surface area of the unit; $k$ is the component labeling superscript, $k$ = l, 2, 3, …, Nc; Nc is the component fraction; $k$ = Nc + 1 represents the energy conservation equation; $A^k$ is the mass term of component $k$ in the unit; when $k$ = Nc + 1, $A^k$ is the energy term in the unit body; $F^k$ is the mass exchange term; $q^k$ k is the source-sink term in a unit; $n$ is the unit normal vector in the unit; and $t$ is time.

Among them, the mass conservation equation is as follows:

$$A^k=\varphi {\displaystyle \sum }_{\beta =1}^{NPH}{S}_{\beta }{\rho }_{\beta }{X}_{\beta }^k$$

where $\beta$ is the phase sign; ($\beta$ = l, 2, … NPH, gas or liquid phase, etc.); $\varphi$ is the porosity; ${\rho }_{\beta }$ is the density of the phase; $S_{\beta }$ the saturation of the phase; and $X_{\beta }^k$ is the mass content of the components in the phase.

The energy conservation equation is

$$A^{Nk+1}={\displaystyle \sum }_{\beta }\left(\varphi {\rho }_{\beta }{S}_{\beta }{U}_{\beta }\right)+\left(1-\varphi \right){\rho }_s{U}_s$$

where ${\rho }_s$ is the density of the solid rock; $U_s$ is the thermal fusion of the solid rock; and $U_{\beta }$ is the internal energy of each phase in the fluid.

3.2. Design and Establishment of the Geological Model

A RZ2D (two-dimensional radial) reservoir model was created in TOUGH/Petrasim (Figure 4). The reservoir depth range is 1368~1439 m, the thickness (Z direction) is generalized to 50 m and the X direction is 3 km. Based on the change in permeability and porosity, the reservoir is divided into 9 layers (

Table 1. The stratigraphic physical characteristics of the fine division of reservoirs in the Triassic Liujiagou Formation under the Gao 2 borehole.

Reservoir	Stratification	Depth Interval [m]	Thickness [m]	Porosity (%)	Permeability [mD]
LJG3	LJG34	1368–1374	6	5.6	0.17
	LJG33	1374–1379	5	6	0.22
	LJG32	1379–1387	8	5.13	0.11
	LJG31	1387–1392	5	8.99	1.34
LJG2	LJG22	1405–1410	5	10.18	2.3
	LJG21	1410–1418	8	5.83	0.2
LJG1	LJG13	1426–1429	3	8.17	0.88
	LJG12	1429–1434	5	8.56	1.1
	LJG11	1434–1439	5	9.67	1.84

). In the Z direction, every 1 m is divided into 1 layer, for a total of 50 layers. The X direction is divided by equal ratio, and the scale factor is 1.11, which is divided into 100 layers. The whole reservoir model is 5000 grids, and the fixed grid is 100. Considering the poor physical properties of the reservoir in the study area, the constant flow rate injection will cause the injection point pressure to continue to accumulate to the fracture pressure of the mudstone caprock of the Liujiagou Formation (about 40 MPa) [41]. In order to meet the engineering safety requirements and facilitate the operation of the model, according to the international research experience [42], a constant pressure of 1.5 times the hydrostatic pressure of the reservoir is used to inject CO₂, and the injection well runs through the entire reservoir. The injection point is located at the bottom of the left boundary of the model, the wellbore is 3 m and the injection time lasts for 30 years. A monitoring well is set up at a distance of 100 m from the injection well. The monitoring point is located at the top grid of each layer to quantitatively characterize the influence of the change of each uncertain factor. At the same time, in order to facilitate the convergence of the model calculation, this paper only gives heterogeneity in the vertical direction of the formation and assumes that it is homogeneous in the horizontal direction.

3.3. Initial and Boundary Conditions

Because the reservoir thickness is not large, the model can be set to isothermal. Considering that there are no deep formation geophysical data in the study area, the initial default condition refers to the measured value of the middle part of the Liujiagou Formation reservoir in the SH-CCS project in the adjacent area [41,43]. The temperature is 52.4 °C, the linear interpolation is performed at 2.7 °C/100 m and the pressure is 15.26 MPa. The average formation pressure coefficient of 0.88 is used for the linear calculation of each depth. (Note: The reason why the temperature and pressure of the SH-CCS Liujiagou Formation in the adjacent area are directly cited here is that the reservoir depth of the SH-CCS Liujiagou Formation is also located in the middle layer of the Liujiagou Formation reservoir in the study area (Figure 2). The use of this temperature and pressure has better indicative significance.) The formation is assumed to be a saturated saline aquifer, and the salinity is taken as an average of 30 g/L according to water sample analysis. The original CO₂ saturation of the formation is 0. The ratio of horizontal permeability to vertical permeability of the formation rock is set to 10, the liquid residual saturation Slr is 0.30, the gas residual saturation Sgr is 0.05, the gas and liquid saturation Sgs and Sls values are 1 or 0.99 and the rock density of the Liujiagou Formation is 2400 kg·m⁻³ [41,43,44]. The thermal properties of all rocks are the thermal conductivity of salt-saturated rock, which is 2.51W/(m*k). The specific heat is 920J/(kg*k) [45], the rock pore compression ratio is 4.5 × 10⁻¹⁰ [37] and the default parameters of other strata are shown in

Table 2. Default hydrological parameters of strata.

Fm	T °C	P MPa	wB %	CWET W/(m*k)	SPHT J/(kg*k)	Slr	Sgr	Sgs	Sls	C Pa−1	P0 kpa	ρ kg.m−3	Kxyz
LJG	52.4	15.26	3	2.51	920	0.30	0.05	1	1&0.99	4.5 × 10−10	11.2	2400	1:10

Note: w_B is the salt content; CWET is the thermal conductivity of rock; SPHT is the special enthalpy of rock particles; Slr and Sgr are the saturation of residual water and residual gas, respectively; Sgs and Sls are the maximum saturation of gas phase and liquid phase, respectively; C is the compression coefficient; P0 is the pressure coefficient; ρ is the formation density; Kxyz is the ratio of horizontal and vertical permeability.

. In addition, the gas-water relative permeability and capillary pressure in the formation are simulated using the van Genuchten model. The reason for using this model is that the van Genuchten relative permeability model is widely used in CO₂ geological storage simulation [18,46,47,48,49], and the van Genuchten capillary pressure model is one of the most suiTab capillary pressure functions for CO₂ injection simulation [47]. For the boundary conditions, the radial boundary of the reservoir model is a constant boundary to characterize the connection with the open saltwater layer. Due to the barrier of caprock at the top and bottom of the model, the zero flow boundary is used to characterize the sealing of caprock. Before the injection of CO₂, the whole model achieves pressure gradient balance through 1000 years of initialization to achieve a steady state.

3.4. Simulation Scheme Design

It is generally believed that mineralization provides only a small contribution to the storage of CO₂ in the saline aquifer within a hundred years [50,51]. Therefore, this study does not consider the impact of mineralization plume migration on the safety and storage capacity of the saline aquifer storage project. The purpose of this paper is to modify the assignment of uncertain parameters in the reservoir and analyze the influence of various factors on CO₂ geological storage, including CO₂ diffusion range, interlayer pressure accumulation and geological storage capacity. The interval value of the uncertainty of the physical properties of the strata is taken from the previous research experiments on the physical properties of the Liujiagou Formation [41,43,44]. Among them, the pore compression coefficient C is 1.10~6.8(× 10⁻¹⁰), the pore geometry factor λ is 0.35~0.95, the gas residual saturation Sgr is 0.01~0.25 and the liquid saturation Sls is 0.89~0.99. For salinity and the ratio of horizontal and vertical permeability, it is reduced and amplified based on the standard value. According to the default value of the parameters and the uncertainty interval, the seven influencing factors are divided into default value, scaling value and larger value, as shown in

Table 3. Uncertain parameter simulation scheme.

	wB/%	T/°C	Kxyz	λ	Sgr/%	C		Sls/%
Scaling	0.01	41.10	1:30	0.35	0.01	1.10	Scaling 1	0.95
Default	0.03	52.4	1:10	0.457	0.05	4.50	Scaling 2	0.98
Larger	0.1	76.6	1:1	0.55	0.25	6.80	Default	1

.

4. Results and Discussions

4.1. The Influence of Various Uncertain Factors on the Storage of Saltwater Layer

4.1.1. Salinity (w_B/%)

The effects of different salinity values (0.01, 0.03, 0.1) on the storage of the saltwater layer were compared. It can be seen from Figure 5 that after 30 years of injection, the change in salinity will not affect the diffusion range of the plume, and the most affected by the salinity change is dissolved CO₂. With the increase in salinity, the dissolved CO₂ saturation decreases significantly. When the salinity increases to 0.03, the CO₂ saturation decreases by 10%. When the salinity is 0.1, the CO₂ decreases sharply to 70% of the low salinity (0.01) scenario. This is because the increase in salinity can significantly reduce the solubility of CO₂ in water [52,53], In addition, the increase in salinity also reduces the molar volume difference between fresh saltwater and CO₂-dissolving saltwater [52], reduces the effectiveness of mixed convection, and reduces the contact area between saltwater and CO₂, resulting in a decrease in dissolution efficiency.

It can be seen intuitively from the monitoring points that the change in salinity affects the pressure accumulation and CO₂ saturation at the same time (Figure 6a). With the increase in salinity, the pressure accumulation value in the layer increases. This phenomenon may be due to the feedback of salting-out effect. During the process of CO₂ displacing saline water, CO₂ will push saline water to the opposite area of the injection to produce salt precipitation [54]. The precipitation of salt will encroach on the surrounding pore space, resulting in the reduction of effective porosity and the accumulation of local pressure. This indicates that the influence of interlayer pressure accumulation on engineering safety must be considered when CO₂ is stored in saline aquifers with high salinity. It can be seen that at the monitoring point, the change in salinity does control the solubility of CO₂, and the dissolved CO₂ saturation between reservoirs decreases greatly with the increase in salinity, which also confirms the conclusion shown in the above cloud map (Figure 6b). The increase in salinity will restrict the effectiveness of the dissolution and storage mechanism. The difference is that the data of the monitoring points show that the change in salinity also affects the saturation of gaseous CO₂ (Figure 6c). Although the degree of influence is small, it can be seen that when the salinity increases from 0.03 to 0.1, the saturation of gaseous CO₂ still decreases to a certain extent, which indicates that higher salinity will greatly reduce the effectiveness of residual storage and dissolution storage. In addition, the change in salinity does not seem to affect the migration velocity of the plume (Figure 6). The time of the pressure plume, dissolved CO₂ and gaseous CO₂ reaching the monitoring point does not change with the change in salinity.

4.1.2. Temperature (°C)

The distribution of gas saturation and dissolved concentration after 30 years of CO₂ injection in different reservoir temperatures (41.10 °C, 52.4 °C, 76.6 °C) was compared. It can be seen from Figure 7 that the saturation of dissolved CO₂ gradually decreases with the increase in temperature. When the temperature increases from 41.1 °C to 52.4 °C, the saturation of dissolved CO₂ decreases by 10%. When the temperature increases to 76.6 °C, the saturation of dissolved CO₂ decreases to 90% of the default value. However, it has a positive effect on the swept area of gaseous CO₂ and the dissolved CO₂ plume. The increase in temperature increases the horizontal and vertical migration of the plume. Therefore, the influence of temperature on the distribution of CO₂ gas saturation and dissolved concentration is bidirectional. This is because the change in temperature affects the solubility and viscosity of CO₂. As the temperature increases, the solubility of CO₂ gradually decreases [53], and more CO₂ is stored in the form of gas. In addition, the higher the temperature, the smaller the viscosity of CO₂ and the stronger its fluidity, which undoubtedly increases the contact area between gaseous CO₂ and fresh saline water. Therefore, although the increase in temperature reduces the saturation of dissolved CO₂, it also increases the plume migration distance and the area of dissolution and storage. However, due to the influence of reservoir heterogeneity, the increase of vertical migration is less.

In addition, according to the data of monitoring points, the change in temperature also affects the pressure accumulation between reservoirs and the velocity of plume propagation (Figure 8). As the temperature increases, the pressure accumulation between the reservoirs during the injection will increase, but at the end of the injection, the interlayer pressure values of each temperature scene will accumulate to a similar value (Figure 8a). This shows that in the scenario of temperature change, the influence of the increase in interlayer pressure accumulation value caused by the increase in temperature during injection on reservoir integrity should be considered. In addition, the monitoring points clearly show that the propagation speed of the CO₂ plume will increase with the increase in temperature. At the same time, the data also confirm the conclusion of the above cloud map, that is, the dissolved CO₂ saturation decreases greatly with the increase of temperature, and the gaseous CO₂ saturation increases with the increase in temperature (Figure 8b,c). This indicates that the increase in temperature may weaken the dissolution and storage mechanism, resulting in a decrease in the effectiveness of the dissolution and storage mechanism. A large amount of CO₂ exists in the form of free state, which will pose a hidden danger to the safety of CO₂ storage engineering.

4.1.3. Horizontal and Vertical Permeability Ratio (Kxyz)

Permeability ratio is an important parameter to describe reservoir heterogeneity, and it is also one of the factors affecting CO₂ migration, distribution and storage capacity. Therefore, it is particularly important to study the influence of different permeability ratios on CO₂ geological storage. The effects of different permeability ratios (1:30, 1:10, 1:1) on CO₂ storage in low-porosity and low-permeability heterogeneous reservoirs were compared. It can be seen from Figure 9 that the difference in permeability ratios mainly restricts the horizontal and vertical migration of the gaseous CO₂ plume, thereby controlling the sweep area of dissolved CO₂. When the permeability ratio is 1:30, the gaseous CO₂ plume mainly migrates horizontally, which is subject to the lower longitudinal permeability. At this time, the area of CO₂ plume is limited, mainly distributed in the lower part of the reservoir, and the swept area of dissolved CO₂ is small. With the gradual increase in the permeability ratio, the lateral migration of CO₂ decreases slightly. Due to the high permeability in the longitudinal direction, the vertical migration is greatly improved under the action of pressure and density difference, and the range of overall CO₂ migration and diffusion becomes larger. Affected by reservoir heterogeneity, under high permeability ratio, the gaseous CO₂ plume shows a tongue-like advantage in areas with good physical properties, and the swept area increases. At this time, the diffusion range of dissolved CO₂ also becomes larger.

Based on the monitoring points, the change in the ratio of horizontal and vertical permeability also affects the pressure accumulation between reservoirs and the velocity of plume propagation. It can be clearly seen that the pressure accumulation between reservoirs increases significantly with the increase in the permeability ratio (Figure 10a). This shows that in other scenarios with the same conditions, the formation with a high permeability ratio may accumulate higher pressure values between layers. Therefore, when selecting the storage site, the formation with a certain difference in horizontal and vertical permeability may be better, or the injection pressure is controlled in the formation with a high ratio of horizontal and vertical permeability to weaken the pressure accumulation between the layers. In addition, the plume propagation speed increases rapidly with the increase in the permeability ratio (Figure 10b,c). The change in the permeability ratio also restricts the saturation of CO₂. High permeability ratio corresponds to high gaseous CO₂ saturation, but the change in permeability ratio cannot change the saturation of dissolved CO₂ (Figure 10b,c). This indicates that the permeability ratio may affect the residual gas storage mechanism by controlling the migration of the gaseous CO₂ plume, which in turn affects the effective time of the dissolution storage mechanism.

4.1.4. Pore Geometry Factor (λ)

Pore geometry index is the main parameter in the relative permeability model and capillary pressure model, and it is one of the influencing factors in CO₂ geological storage. The uncertainty of pore geometric parameters will cause deviations in the prediction results of CO₂ storage [46]. Based on the measured data, this paper compares the effects of different pore geometric indexes (0.35, 0.457, 0.55) on the simulation results.

Figure 11 shows that the diffusion area of the CO₂ plume increases with the increase in the pore geometry index. The specific performance is that the lateral migration distance increases with the increase, and the longitudinal migration distance increases with the increase and decreases slightly, but the degree of increase in the lateral migration distance is greater than the degree of decrease in the longitudinal migration distance. Therefore, the diffusion area of the CO₂ plume increases with the increase in the pore geometry index. At the same time, the saturation of gaseous CO₂ is also restricted by the pore geometry index, and the average saturation of gaseous CO₂ will increase.

In addition, according to the monitoring points, the uncertainty of the pore geometry index also affects the pressure accumulation and plume propagation speed in the layer. With the increase in the pore geometric index, the pressure accumulation value in the layer also increases (early-middle injection) and then decreases with the increase in the pore geometric index (Figure 12a). The plume propagation velocity increases with the increase in the pore geometry index (Figure 12b,c). In addition, the pore geometry index also affects the gaseous CO₂ saturation, which shows that it increases with the increase in the pore geometry index, and the dissolved saturation is not affected, which is consistent with the conclusion of the above cloud map. In addition, compared with the results of horizontal and vertical permeability ratios, the control of the pore geometry index on gaseous CO₂ saturation seems to be greater than the control of plume migration velocity.

4.1.5. Residual Gas Saturation (Sgr/%)

Residual gas saturation is one of the parameters of formation physical properties, and it is also one of the important parameters in the relative permeability model and capillary pressure model. The effects of different residual gas saturations (0.01, 0.05, 0.25) on CO₂ geological storage were compared. It can be seen from Figure 13 that the result of increasing residual gas saturation is to fix more gas during gas plume migration. The gas CO₂ saturation gradually increases, and the residual gas storage mechanism becomes more effective. Although the saturation of dissolved CO₂ is not changed, the migration distance of the plume is reduced, thus reducing the effectiveness of the solubility storage mechanism.

The data of the monitoring points show that the pressure accumulation in the high residual gas scenario during the injection period is much smaller than that in the low residual gas scenario (Figure 14a), which may be attributed to the lower migration velocity of the CO₂ plume (Figure 14c). At the end of the injection, the pressure accumulation value of the high residual gas scenario will be greater than that of the low residual gas scenario, which may be the feedback of a large amount of residual storage (Figure 14c), resulting in an increase in pore pressure. With the increase in residual gas saturation, the migration speed of CO₂ plume decreases, and the time of gaseous CO₂ and dissolved CO₂ reaching the monitoring point lags behind with the increase in residual gas saturation (Figure 14b,c). The final saturation of dissolved CO₂ is not restricted by residual gas saturation, while the saturation of gaseous CO₂ increases significantly with the increase in residual gas saturation. Therefore, although the increase in residual gas saturation has an adverse effect on the CO₂ dissolution and storage mechanism, it may greatly enhance the effectiveness of the residual storage mechanism.

4.1.6. Liquid-Phase Saturation Degree (Sls/%)

In previous simulation studies, most scholars generally assumed that the formation is saturated with water. In fact, the saturated saline water layer is difficult to achieve under actual conditions. Therefore, little is known about the effect of different total liquid saturation on CO₂ geological storage in low-porosity and low-permeability heterogeneous reservoirs. This paper compares the distribution of plumes with different liquid saturations (0.95, 0.98, 1) after 30 years of CO₂ injection. It can be seen from Figure 15 and Figure 16 that the increase in liquid saturation inhibits the migration of gaseous CO₂ plume and does not promote the dissolution of CO₂. The saturation of gaseous CO₂ decreases with the increase in liquid saturation, while the saturation of dissolved CO₂ has almost no change (Figure 16b,c). This indicates that for CO₂ geological storage, the higher liquid saturation is not the better. The reason for the above results may be that in the saturated state of the saline water layer, due to the liquid phase occupying the pore space, the rate of CO₂ displacing the saline water is slowed down, and the pore space cannot accept a certain amount of residual gas when the liquid phase is refluxed. At the same time, the convection between the saline water dissolving CO₂ and the fresh saline water is weakened, resulting in a decrease in the saturation of gaseous CO₂ and almost no change in the saturation of the dissolved state.

The change in liquid saturation will also affect the accumulation of interlayer pressure (Figure 16). Before the middle stage, the interlayer pressure increases with the decrease in liquid saturation. After the middle stage, the interlayer pressure increases with the increase of liquid saturation. In addition, the change in liquid saturation will not affect the migration speed of the plume.

4.1.7. Pore Compressibility (Pa-1)

The pore compression coefficient is the compression coefficient of pore volume to pore pressure [55]. Pore space is an important parameter to determine the reserves of CO₂ geological storage. The difference in the pore compression coefficient will inevitably affect the effect of CO₂ geological storage. Previous studies have pointed out that the error caused by the uncertainty of the pore compression coefficient should not be ignored in geological modeling [56]. Based on this, the effects of different pore compression coefficients (1.10, 4.50, 6.80) on the storage of CO₂ saline aquifers were compared by referring to field experimental data. The results show that under the same order of magnitude (e-10), different pore compression coefficients have little restriction on the effect of CO₂ geological storage. The change in pore compressibility almost does not affect the distance of CO₂ plume migration, nor does it affect the distribution range of gas and liquid CO₂ saturation (Figure 17).

Based on the monitoring points, it can be seen that the change in the pore compression coefficient will have a certain impact on pressure accumulation (Figure 18a). It is shown that in the early stage of injection, the higher the pore compression coefficient is, the lower the pressure value is, and then the pressure value tends to be uniform. The effect on CO₂ diffusion rate and saturation is negligible (Figure 18b,c).

4.2. Analysis of Influencing Factors of CO₂ Storage in Saline Aquifer

4.2.1. Change of Pressure Field in Reservoir

According to the accumulation value of interlayer pressure after 30 years of CO₂ injection simulated by the model (

Table 4. The influence of different factors on the accumulation of interlayer pressure.

	Salt	Kxyz	T	λ	Sgr		Sls
Scaling case/Pa	1.741 × 107	1.698 × 107	1.741 × 107	1.754 × 107	1.741 × 107	Scaling1 case/Pa	1.737 × 107
Default case/Pa	1.746 × 107	1.746 × 107	1.746 × 107	1.746 × 107	1.746 × 107	Scaling2 case/Pa	1.742 × 107
Larger case/Pa	1.765 × 107	1.775 × 107	1.743 × 107	1.737 × 107	1.757 × 107	Default case/Pa	1.746 × 107

), it can be seen that the interlayer pressure value gradually increases with the increase in salinity, horizontal and vertical permeability ratio, temperature, residual gas saturation and liquid saturation. As the pore geometric factor increases, the pressure value gradually decreases. The temperature factor is special. The pressure value first increases with the increase in temperature (41.1–52.4 °C) and then decreases with the increase in temperature (52.4–76.6 °C).

4.2.2. Changes in Capture Amount Under Different Capture Mechanisms

After 30 years of CO₂ injection in low-porosity and low-permeability heterogeneous reservoirs, the total CO₂ gas storage and total dissolved storage are shown in

Table 5. The influence of different factors on gaseous CO₂ storage.

	Salt	C	Kxyz	T	λ	Sgr		Sls
Scaling reserve/kg	1.42 × 108	1.41 × 108	1.28 × 108	1.27 × 108	1.08 × 108	1.43 × 108	Scaling1 reserve/kg	1.53 × 108
Default reserve/kg	1.41 × 108	1.41 × 108	1.41 × 108	1.41 × 108	1.41 × 108	1.41 × 108	Scaling2 reserve/kg	1.46 × 108
Larger reserve/kg	1.38 × 108	1.42 × 108	1.67 × 108	1.46 × 108	1.61 × 108	1.11 × 108	Default reserve/kg	1.41 × 108

and

Table 6. The influence of different factors on dissolved CO₂ storage.

	Salt	C	Kxyz	T	λ	Sgr		Sls
Scaling reserve/kg	3.70 × 107	3.44 × 107	3.11 × 107	3.08 × 107	3.13 × 107	3.95 × 107	Scaling1 reserve/kg	3.46 × 107
Default reserve/kg	3.45 × 107	3.45 × 107	3.45 × 107	3.45 × 107	3.45 × 107	3.45 × 107	Scaling2 reserve/kg	3.46 × 107
Larger reserve/kg	2.65 × 107	3.46 × 107	3.87 × 107	4.23 × 107	3.51 × 107	1.64 × 107	Default reserve/kg	3.45 × 107

. As shown in Table 5 and Table 6, with the increase in salinity, residual gas saturation and liquid saturation, the amount of CO₂ gas storage and CO₂ dissolution storage decreased. With the gradual increase in the pore compression coefficient, horizontal and vertical permeability ratio, temperature and pore size geometric factor, the amount of CO₂ gas storage and CO₂ dissolution storage increased.

4.2.3. Sensitivity Analysis

Previous studies have pointed out that the degree of influence of a parameter on the model can be expressed by the sensitivity of the model to the parameter [57,58,59]. Sensitivity refers to the degree of influence of one factor’s change on other factors. In order to quantify the sensitivity of storage capacity to physical parameters, based on the research of Yuqun Xue [57], the sensitivity of reservoir physical parameters is calculated by the following formula.

$${\beta }_{i, k}={\displaystyle \frac{\partial H_{\dot{i}}}{\partial {\alpha }_k}}\approx {\displaystyle \frac{H_i\left({\alpha }_k+\Delta {\alpha }_k\right)-H_i\left({\alpha }_k\right)}{\Delta {\alpha }_k}}$$

In the formula, the subscript $i$ is the time observation point; $H_i$ is the CO₂ storage capacity MCO₂(kg) calculated at the $i$th time observation point when the $k$th reservoir physical parameter is ${\alpha }_k$; ${\alpha }_k$ is the value of the $k$th reservoir physical parameters; $\Delta {\alpha }_k$ is the increment of ${\alpha }_k$; ${\beta }_{i, k}$ is the sensitivity of the variable $H_i$ to the $k$th reservoir physical parameters at the observation point of the $i$th time.

Because of the different dimensions of reservoir physical parameters, it is impossible to compare their sensitivity, so the above formula is transformed into a dimensionless form, which is

$${\beta }_{\dot{i}, k}={\displaystyle \frac{\partial H_i}{\partial {\alpha }_k}}∕{\displaystyle \frac{H_i}{{\alpha }_k}}$$

Because the sensitivity parameter ${\beta }_{i, k}$ changes greatly with time, here the $H_i$ of the $k$th reservoir physical property parameter at the observation point of the $i$ observation time is calculated by taking the average value $H$ of the $k$th reservoir physical property parameter in the whole time period.

In addition, due to the particularity of the horizontal and vertical permeability ratio parameters, it cannot directly use the above formula. In fact, the change in the ratio of horizontal and vertical permeability is only the vertical permeability, so the change in vertical permeability can be used to refer to the change in the ratio of horizontal and vertical permeability. Here, assuming that the unit permeability in the horizontal direction of the $g$ layer in a formation is $x$, the horizontal permeability of the g layer is ${\mathsf{\mu }}_{g}x$. For the average permeability $\overline{X}$ of the whole formation, it is

$$\overline{X}=\frac{{\mathsf{\mu }}_{1}x+{\mathsf{\mu }}_{2}x+{\mathsf{\mu }}_{3}x+\cdots +{\mathsf{\mu }}_{g}x}{g}$$

In the formula, $g$ denotes the number of layers in the reservoir, $x$ denotes the unit permeability in the horizontal direction, ${\mathsf{\mu }}_g$ denotes the value of the horizontal permeability of the $g$th reservoir and $\overline{X}$ denotes the average permeability of the whole reservoir.

According to the assumed horizontal and vertical permeability ratio (1:30, 1:10, 1:1), the average vertical permeability values Y in the formation are 1/30 $\overline{X}$, 1/10 $\overline{X}$, $\overline{X}$, respectively. At this time, the horizontal and vertical permeability ratio parameters without numerical meaning can be transformed into the value Y (1/30$\overline{X}$, 1/10$\overline{X},\overline{X}$) of the vertical permeability with numerical meaning and substituted into the above formula to obtain the sensitivity.

It can be seen from Figure 19 that the order of influencing pressure accumulation in the layer from high to low according to the standard sensitivity is liquid saturation, horizontal and vertical permeability ratio, pore geometry factor, temperature, salinity and residual gas saturation (0.126 > 0.04 > 0.021 > 0.014 > 0.004 = 0.004). Liquid saturation, horizontal and vertical permeability ratio, salinity and residual gas saturation are positively correlated with pressure accumulation, while pore geometry factor is negatively correlated with pressure accumulation. The change in temperature is not a single linear relationship with pressure accumulation. When the temperature is between 41.1 °C and 52.4 °C, it has a positive correlation with pressure accumulation. When the temperature increases from 52.4 °C to 76.6 °C, it has a negative correlation with pressure accumulation. In addition, the ratio of horizontal and vertical permeability is also a special influencing factor. When the ratio of horizontal and vertical permeability increases from 1:30 to 1:10, the sensitivity of horizontal and vertical permeability to pressure accumulation is enhanced. When the ratio of horizontal and vertical permeability increases from 1:10 to 1:1, the sensitivity to pressure accumulation is weakened.

The sensitivity of the CO₂ gas storage capacity to default values from high to low is liquid saturation, pore geometric factor, temperature, horizontal and vertical permeability ratio, residual gas saturation, salinity and pore compression coefficient (9.8 > 5.527 > 2.54 > 0.764 > 0.098 > 0.058 > 0.02) (Figure 20). Pore geometry factor, temperature, horizontal and vertical permeability ratio and pore compression coefficient are positively correlated with gas storage capacity, while liquid saturation, residual gas saturation and salinity are negatively correlated with gas storage capacity. The pore geometry factor and temperature show a special positive correlation. As the pore geometry factor gradually increases, its sensitivity to gas storage gradually decreases. At medium-low temperatures (41.1 °C~52.4 °C), its sensitivity to gaseous storage capacity gradually increases, and at high temperatures, the sensitivity decreases.

The sensitivity of CO₂ dissolved storage capacity from high to low is temperature, residual gas saturation, horizontal and vertical permeability ratio, liquid saturation, pore geometry factor, salinity and pore compression coefficient (2.74 > 1 > 0.817 > 0.8 > 0.68 > 0.6 > 0.021) (Figure 21). Temperature, horizontal and vertical permeability ratio, pore geometry factor and pore compression coefficient are positively correlated with dissolved storage capacity, and residual gas saturation and salinity are positively correlated with dissolved storage capacity. Although the pore geometric factor shows a positive correlation, its sensitivity to dissolved storage decreases with the increase in its value. When the ratio of transverse and longitudinal permeability is high (1:1), its sensitivity to dissolved storage is also weakened. More special is the liquid phase saturation. When the liquid phase saturation is 0.95, the sensitivity to the dissolved storage capacity is positive. As the liquid phase saturation increases, the sensitivity to the dissolved storage capacity shows a negative correlation.

5. Conclusions

In this paper, the Liujiagou Formation in the Yulin area of the Ordos Basin is taken as an example to study the effects of seven uncertain factors, namely salinity (w_B), temperature (T), horizontal and vertical permeability ratio (Kxyz), pore geometry factor (λ), residual gas saturation (Sgr), liquid saturation (Sls) and pore compression coefficient (C), on the geological storage of CO₂ in low porosity and low permeability heterogeneous reservoirs. The main results are as follows:

Uncertain factors affect the safety of CO₂ geological storage to a certain extent by affecting the speed of the residual storage and dissolution storage mechanism. The increase in residual gas saturation is not conducive to the diffusion of the plume, and the diffusion range of the plume increases with the increase in temperature, longitudinal permeability ratio and pore geometric factor.

Higher residual gas saturation and salinity will adversely affect the safety and storage capacity of CO₂ saline aquifer storage. High temperature and high vertical permeability ratio will lead to higher interlayer pressure accumulation, which is not conducive to the safety of the storage project but is beneficial to the storage capacity.

The decrease in salinity and temperature increases the saturation of dissolved CO₂. Gaseous CO₂ saturation increases with the increase in temperature, permeability ratio, pore geometry factor and residual gas saturation and decreases with the increase in liquid saturation. The larger the temperature, the ratio of horizontal and vertical permeability and the pore geometric factor, the faster the plume velocity. The greater the residual gas saturation, the slower the plume velocity.

The sensitivity analysis of all factors shows that the liquid saturation and temperature have the greatest influence on CO₂ geological storage, and the pore compression coefficient has the least influence.