TL;DR
This study uses NMR and mercury injection to map how residual gas distributes within sandstone pores under different water saturation and flow conditions. They found that while initial water saturation plays a role, the capillary number (representing the balance between viscous and capillary forces) significantly dictates where trapped gas ends up, particularly in larger pores.
What Problem Does This Paper Address?
Understanding how gas gets trapped in reservoirs, especially after waterflooding, is crucial for maximizing oil and gas recovery. This "residual gas" can be difficult to dislodge. While previous studies have looked at factors like initial water saturation (Swi) and capillary number (Ca) separately, their combined effect on the microscopic distribution of this trapped gas within different pore sizes (micropores, mesopores, macropores) has remained ambiguous. This paper aims to fill that gap by providing a detailed, pore-scale picture of residual gas distribution under varying experimental conditions.
Methods
The researchers employed a multi-pronged approach:
- High-Pressure Displacement Experiments: They conducted both spontaneous and forced imbibition (waterflooding) experiments on sandstone core samples.
- Nuclear Magnetic Resonance (NMR): This was the primary tool for visualizing and quantifying the gas saturation and its distribution in real-time. Specifically, they used the NMR transverse relaxation time (T2) spectrum. The T2 spectrum is sensitive to pore size, allowing them to differentiate between gas in micropores (< 1 µm), mesopores (1-10 µm), and macropores (> 10 µm).
- Mercury Intrusion (MI): This technique was used to determine the pore size distribution of the rock samples independently.
- Calibration: The NMR T2 spectrum was calibrated against the mercury intrusion data to establish a reliable conversion coefficient between T2 relaxation time and pore radius. This allowed for quantitative analysis of residual gas distribution across different pore sizes.
The experiments systematically varied the capillary number (Ca) across four orders of magnitude (logCa from -8.68 to -5.68) and initial water saturation (Swi) at three levels (0%, 39.34%, and 62.98%).
Datasets
Not discussed in the paper. The study used sandstone core samples from a gas reservoir.
Key Insights and Ideas
- Capillary Number is King (for distribution): While initial water saturation influences the amount of trapped gas, the capillary number (Ca) has a more significant impact on where that gas is distributed within the pore network. This highlights the importance of flow rate and viscous forces in determining the final resting place of trapped gas.
- Larger Pores are Gas Havens: In spontaneous imbibition, over 90% of the residual gas consistently ended up in pores larger than 1 µm (meso- and macropores), regardless of initial water saturation. This suggests that larger pores act as preferential sites for gas trapping.
- Flow Channels Evolve with Ca: The nature of the flow channels formed by water during imbibition changes with the capillary number. At lower Ca, flow is dominated by smaller pores (micro- and mesopores). As Ca increases, flow increasingly utilizes larger pores (meso- and macropores), which in turn influences where the residual gas is left behind.
- Swi's Role is More Subtle: While higher initial water saturation generally leads to increased capillary trapping efficiency, its impact on the distribution of residual gas across pore sizes is less pronounced than that of Ca, especially in spontaneous imbibition.
Key Results
- Spontaneous Imbibition: Above 90% of residual gas was found in pores > 1 µm across all tested Swi levels. Capillary trapping efficiency increased with increasing Swi.
- Forced Imbibition:
- At low Ca (logCa = -8.68, -7.68), flow channels were primarily micro- and mesopores.
- At high Ca (logCa = -5.68), flow channels shifted to meso- and macropores.
- With increasing Ca (from -6.68 to -5.68) at low Swi (0% and 39.34%), residual gas in macropores decreased significantly, while it slightly increased in micropores.
- The impact of Ca on residual gas distribution was greater than Swi for low to intermediate Swi. For the highest Swi (62.98%), both Ca and Swi had a less pronounced effect.
Limitations and Future Work
- Rock Stress Sensitivity and Gas Slippage: The authors acknowledge that while they set confining pressure at 2 MPa and pore pressure below 0.1 MPa, and used high permeability/porosity samples, the influences of rock stress sensitivity and gas slippage effects on the experimental results are considered negligible. However, in other reservoir conditions, these factors might play a more significant role.
- Hysteresis and Capillary End Effects: The study mentions that hysteresis and capillary end effects can cause nonuniform initial conditions but were neglected in this work. Future research could explore their specific contributions.
- Critical Ca: The critical capillary number, beyond which significant changes in trapping occur, was not definitively observed within the experimental range. Further investigation into this threshold could be valuable.
- Generalizability: The study focused on sandstone samples. Future work could explore these effects in other reservoir rock types (e.g., carbonates, shales) to assess the broader applicability of the findings.
- Real-world Application: The authors suggest that understanding these mechanisms can help optimize gas production rates to improve recovery efficiency in gas reservoirs, particularly by reducing residual gas in larger pores. Further work could focus on translating these microscopic findings into macroscopic reservoir management strategies.
Summary updated