Influencing Factors of the Carbon Sequestration Coefficient in Saline Aquifers——Based on Multiphase Flow Displacement Experiments

Chenglong Zhang¹Wentao Zhao²Tieya Jing²Junbin Zhao³Jian Zhang²Mingyi Wei⁴Juan Zhou²Lei Fu^1*

• ¹China Geological Survey, Beijing, China
• ²China Huaneng Clean Energy Research Institute, Beijing, China
• ³Shanxi Institute of Geological Survey CO., LTD., Taiyuan, China
• ⁴National Energy Group Inner Mongolia Shanghai Temple Power Generation Co., Ltd, Ordos, China

Geological CO2 storage refers to the process of injecting CO2-primarily captured from large industrial emission sourcesinto suitable geological formations (e.g., deep saline aquifers and depleted oil and gas reservoirs) through engineering techniques to achieve its long-term isolation from the atmosphere (Metz et al., 2005;Huang, 2021;Guo et al., 2014). It is estimated that by 2060, the contribution of CCUS technology to China's carbon neutrality goal could exceed 500 million tons (Ding, 2021;Li et al., 2006). Among all geological formations suitable for CO2 storage, saline aquifers account for 95.6% of the total storage potential, far exceeding that of other formation types. Owing to their enormous potential and wide distribution, carbon sequestration in saline aquifers is regarded as a critical supporting technology for achieving carbon peak and carbon neutrality targets.Referring to the calculation method for geological CO2 storage in saline aquifers proposed by the U.S. Department of Energy based on volumetric theory, the formula for estimating the CO2 storage potential of saline aquifers is as follows (Goodman et al., 2011;USDOE, 2006;Li et al., 2015):E h A P 2 CO      ρ φwhere P (kg) is geological potential, A (m²) is reservoir area, h (m) is thickness, ϕ (%) is porosity, and ρ CO2 (kg/m³) is CO2 density. E, the carbon sequestration coefficient, is defined as the ratio of the volume of supercritical CO2 to the total pore space of the formation. And it is a dimensionless geological coefficient.As indicated by the formula, the accuracy of CO2 geological storage potential calculations depends largely on the pore volume of reservoir rocks and the effective storage coefficient. While the precision of reservoir rock pore volume can be improved via advanced geological exploration techniques, the current selection of effective storage coefficients lacks a theoretical basis. Thus, obtaining a more accurate effective storage coefficient is crucial for enhancing the precision of CO2 geological storage potential evaluations.Most existing studies on effective storage coefficients for geological CO2 sequestration have been conducted under idealized experimental conditions or via numerical models (Dai et al., 2022;Diao et al., 2023;Fu et al., 2022;Lei et al., 2022;Wei et al., 2013Wei et al., , 2021Wei et al., , 2022;;Yang et al., 2019;Zhang et al., 2005;Zhang et al., 2019), failing to reflect the high complexity of real-world on-site scenarios. Currently, the mainstream storage coefficient used in calculating CO2 storage potential in saline aquifers is 0.02, which involves significant uncertainties (Goodman et al., 2011;Li et al., 2015;Diao et al., 2017;Guo et al., 2014). In practice, saline aquifers exhibit substantial variations in geological characteristics across regions, such as porosity, permeability, formation pressure, temperature, and formation water chemistry. These factors are interrelated, and traditional experimental methods struggle to systematically analyze the role of each factor, making it difficult to determine a unified and accurate storage coefficient.To investigate the carbon sequestration coefficient in saline aquifers, this study adopted an orthogonal experimental design, using artificial cores with significantly different porosities as experimental materials. A multiphase flow displacement NMR analysis system was employed to conduct 9 sets of core multiphase flow displacement experiments via orthogonal design, with porosity, confining pressure, and pressure difference as variables (Lang et al., 2021). The maximum residual CO2 saturation of artificial cores under different conditions was obtained, thereby accurately identifying the relative importance and interactions of each factor on the storage coefficient. This effectively addresses the limitations of existing studies in comprehensively analyzing multiple factors. A multiple linear regression equation for maximum residual CO2 saturation was established, which can provide a reference for determining carbon storage coefficients in saline aquifers in future research. The MesoMR12-060H-I multiphase flow displacement NMR system (Fig. 1) was used, featuring: Relaxation spectrum testing: Analyzes porosity, pore size distribution, saturation, and layered water content. NMR measures nuclear magnetic resonance of magnetic nuclei in external magnetic fields, providing rapid, non-destructive analysis of porous media properties (e.g., porosity, permeability, water saturation). In CO2 sequestration research, it is used to track the migration of wetting-phase fluids and characterize reservoir properties. Artificial sandstone cores (supplied by Yilai Bo Technology Co., Ltd.) with porosities (5%, 10%, 15%) were used, and their porosity accuracy was pre-tested (Tab.1). A three-factor, three-level orthogonal design (L9(3 3 )) was adopted, with variables including porosity (5-15%), confining pressure (8-10 MPa), and pressure difference (0.3-0.9 MPa), while temperature was maintained constant at 33 ℃ . Details of the nine experimental groups are provided in Table 2. Cores were fully saturated with water using a vacuum-pressure saturation device (Fig. 2) to ensure uniform wetting. This device is primarily composed of a vacuum system, a liquid storage tank, a sample saturation chamber, a manual pump, a pressure gauge, and valves. When the liquid level in the storage tank ceases to drop, the valve is closed, and the hand pump is used to pressurize to 25 MPa, with this pressure maintained for 24 hours. At this stage, the core is considered to be fully water-saturated. The core sample to be tested, which was fully saturated with water, was prepared prior to conducting the NMR core displacement experiment, following these steps:(1) System self-inspection: The NMR measurement system was powered on according to specifications, a calibration sample was placed into the core holder, and a system self-inspection was performed.(2) Sample preparation and loading: The core sample to be tested was retrieved, a heat-shrinkable tube was applied, and a heat gun was used to shrink the tube tightly around the core. The prepared core sample was loaded into the core holder.(3) Pipeline connection: The core holder was positioned within the NMR coil, and pipelines for the confining pressure unit, gas injection unit, back pressure unit, and low-temperature circulation unit (which maintains the NMR system temperature at 25 ℃ to eliminate temperature-induced interference with NMR signals) were connected.(4) Pressurization: After purging the confining pressure fluid (fluorinated fluid FC770), the fluid was heated to the required experimental temperature using the heating unit. The industrial control computer was used to synchronously apply confining pressure and back pressure in the following sequence: 3-5-7-9-12 MPa for confining pressure and 1-3-5-7-8 MPa for back pressure.Meanwhile, CO2 from the gas cylinder was injected into the intermediate container and pressurized to the required experimental pressure via a constant-rate, constant-pressure pump.(5) Initial signal acquisition: T2 curves and images of the sample in its initial saturated state were collected using the Carr-Purcell-Meiboom-Gill (CPMG) sequences. Additionally, the industrial control computer was used to set the sampling sequence and time intervals for the dynamic displacement process.(6) Water injection: Since high temperatures from the heat gun during sample preparation might lead to partial water loss in the core, the core was re-saturated with water at the experimentally designed back pressure for 12 hours prior to the displacement experiment. This ensured 100% water saturation in the core before gas injection.(7) CO2 injection: Once water saturation was complete, CO2 gas (pressurized to the required experimental level) was injected into the core holder-with the confining pressure maintained 3-5MPa higher than the inlet pressure throughout the experimentto initiate the displacement process.(8) Data collection: Relevant NMR signals were acquired at predefined time points during displacement using the set sampling sequence, and the dynamic displacement process was monitored in real time.(9) Residual weight measurement: Throughout the displacement process, CPMG sequence tests were performed at regular intervals to obtain a series of core T2 curves, one-dimensional water distribution maps, and MRI images after varying displacement durations. As displacement progressed, reduced water content gradually weakened the hydrogen proton signal intensity. When the signal intensity approached the background level, MRI images were simultaneously observed:if fluid distribution boundaries became indistinct (i.e., imaging quality was impaired), it indicated that pore water had been displaced to irreducible saturation. Alternatively, the experiment could be terminated when changes in the core's T2 curve were negligible and the curves nearly overlapped.The core sample was removed, and its residual weight was measured. All other experimental groups followed the same procedure. The CPMG sequence parameters were: waiting time (TW) = 5000 ms, echo time (TE) = 0.2 ms, 16 accumulations, with an acquisition time of 3 minutes (Tab.3). Using the Laplace equation:2 c γ COSθ P R  (Where Pc is capillary entry pressure, γ is interfacial tension, θ is contact angle, and R is pore radius), smaller pores (nanopores/micropores) require higher Pc, explaining why residual water concentrates in these regions.To conduct an in-depth analysis of displacement processes across different pore size groups, the final CO2 displacement efficiency for each pore size group in each experimental set was calculated using the initial T2 curve from the displacement experiment and the T2 curve at the final state upon displacement completion. The formula for calculating CO2 displacement efficiency is as follows.1 2 1 A A R A  Where: R is the CO2 displacement efficiency; is the total nuclear magnetic signal within the range of nanopores, micropores, mesopores, or macropores in the initial T2 curve; is the total nuclear magnetic signal within the range of nanopores, micropores, mesopores, or macropores in the T2 curve at the final state after displacement.Water can only migrate into nanopores if the pressure is sufficient to overcome the capillary entry pressure, which is typically high due to the small pore radii. CO2 displacement efficiency (Fig. 6) shows that mesopores and macropores have displacement efficiencies of >60% and >80%, respectively, while nanopores and micropores have <50%. Negative efficiency in some nanopore calculations (Groups 5, 8, 9) reflects water influx from larger pores. The distribution of residual CO2 (Fig. 7) shows that mesopores hold ≥40% of residual CO2, due to their higher proportion in the pore structure and the limited space in nanopores/micropores occupied by residual water. Maximum residual CO2 saturation (SR) and residual water saturation (SW) were calculated using:1 2 1 0 100% R m m S m m     , 1 W R S S  where m0, m1, m2 are dry, saturated, and residual core weights, respectively. Results for nine groups are listed in Tab.4. The volume coefficient is associated with the area and formation thickness, while the heterogeneity of the formation complicates its calculation. In practical engineering, the volume coefficient is typically derived via three-dimensional geological modeling and numerical simulation methods. By integrating the maximum residual CO₂ saturation equation developed in this study, the geological CO₂ storage coefficient for actual engineering sites can be calculated, thereby supporting the evaluation of CO₂ geological storage potential in saline aquifers. To investigate the factors influencing the effective storage coefficient for CO2 sequestration in saline aquifers and accurately assess their storage potential, 9 sets of multiphase flow core displacement experiments based on orthogonal design were conducted, with porosity, confining pressure, and pressure difference as variables. The experimental materials used were artificial cores with significantly different porosities. The key findings are as follows:(1) 80% of the variation in "maximum residual CO2 saturation" is influenced by "porosity," "confining pressure," and "pressure difference." Among these 3 factors, porosity has the most significant impact on maximum residual CO2 saturation.(2) Qualitative analysis of water migration in cores during displacement via NMR T2 curves revealed that water migration during displacement is closely related to pore structure: water in mesopores and macropores is preferentially displaced, while water in nanopores and micropores is more resistant to displacement.(3) NMR was used to analyze the maximum residual CO2 saturation of artificial cores under different conditions, and a multiple linear regression equation relating maximum residual CO2 saturation to "porosity," "confining pressure," and "pressure difference" was established. By incorporating the volume coefficient derived from numerical simulations, the geological CO2 storage coefficient for actual engineering sites can be estimated. This serves as a reference for subsequent calculations of geological CO2 storage potential in saline aquifers.