Characteristics of Crude Oil Mobilization and Displacement Mechanisms in Low-Permeability Conglomeratic Sandstone Reservoirs under Different Development Modes: An NMR-Based Study

Technical Summary: Characteristics of Crude Oil Mobilization and Displacement Mechanisms in Low-Permeability Conglomeratic Sandstone Reservoirs under Different Development Modes

Problem Statement
Ultra-low-permeability sandy conglomerate reservoirs, particularly those in the Xinjiang Oilfield (e.g., Mahu Oilfield), represent a critical resource for reserve growth in China's continental basins. However, their development is severely constrained by complex pore-throat structures, strong heterogeneity, and the coexistence of gravel-framework support with sandy-muddy infilling. These geological characteristics result in low porosity, high threshold pressure gradients, and poor sweep efficiency for injected fluids. While water flooding is the standard development method due to its maturity and cost-effectiveness, it frequently suffers from preferential breakthrough through large channels and the inability to mobilize oil trapped in micropores and small pores. Conversely, while CO₂ flooding offers mechanisms like viscosity reduction and hydrocarbon extraction, its performance is also highly sensitive to heterogeneity and prone to gas channeling. A systematic understanding of how different displacement methods (spontaneous imbibition, water flooding, CO₂ flooding, and their combination) selectively mobilize oil across specific pore-size classes (micropores, small pores, mesopores, and macropores) remains insufficient, particularly regarding the complementary effects of these methods at the pore scale.

Methodology
This study employed a low-field Nuclear Magnetic Resonance (NMR) approach integrated with core displacement experiments to investigate oil mobilization mechanisms.

• Sample Selection: 28 core samples were collected from the Kexia and Baikouquan formations in the Xinjiang Oilfield, covering ultra-low, low, and medium-low permeability intervals. Fourteen representative samples were selected for displacement experiments.
• Experimental Design: Four distinct experimental modes were conducted:
  1. Spontaneous Imbibition: Using heavy water (D₂O) at 60°C for 12 days to observe capillary-driven mobilization.
  2. Water Flooding: Constant-rate injection of simulated formation water to evaluate pressure-driven displacement.
  3. Supercritical CO₂ Flooding: Injection of supercritical CO₂ above the minimum miscibility pressure (MMP) to assess extraction and swelling effects.
  4. Water Flooding Followed by CO₂ Flooding: A sequential process to evaluate the potential of CO₂ to recover bypassed oil after water flooding.
• Measurement and Analysis: NMR T₂ spectra and T₁–T₂ maps were acquired before and after each displacement stage. The pore system was quantitatively classified into micropores (T₂ < 1 ms), small pores (1 ≤ T₂ < 10 ms), mesopores (10 ≤ T₂ < 100 ms), and macropores (T₂ ≥ 100 ms). Changes in signal amplitudes were used to calculate oil mobilization efficiency and residual oil distribution for each pore class.

Key Results

• Spontaneous Imbibition: This mechanism primarily mobilized oil from micropores and small pores driven by capillary forces. In ultra-low permeability cores, strong capillary forces effectively displaced adsorbed oil from tight matrices. However, in higher-permeability cores, water preferentially occupied large pores, leaving micropores largely untouched due to insufficient capillary pressure gradients. Recovery factors ranged from 31.2% to 40.8%, showing no simple monotonic relationship with permeability; pore structure heterogeneity was a more dominant control than permeability alone.
• Water Flooding: This method exhibited strong pore-size selectivity, preferentially displacing oil from mesopores and macropores. Due to high capillary resistance and threshold pressure gradients, injected water failed to effectively penetrate micropores and small pores. Consequently, final recovery factors were limited to 31.2–46.7%, with significant residual oil remaining in smaller pore systems. Water breakthrough occurred rapidly, often accompanied by viscous fingering and channeling.
• Supercritical CO₂ Flooding: CO₂ flooding demonstrated superior microscopic displacement efficiency, extending the lower pore-size limit of oil mobilization. Through molecular diffusion, hydrocarbon extraction, oil swelling, and viscosity reduction, CO₂ effectively mobilized oil from micropores and small pores that were inaccessible to water. Final recovery factors reached 61.6–79.4%. While gas channeling was observed in higher-permeability cores, the overall recovery was significantly higher than water flooding.
• Water Flooding Followed by CO₂ Flooding: Subsequent CO₂ injection after water flooding provided substantial incremental recovery (an additional 30–38 percentage points). This combined approach effectively targeted residual oil in mesopores and small pores that were bypassed during the water flood. However, micropores remained the principal storage space for residual oil even after the combined process.

Key Contributions

• Pore-Scale Quantification: The study quantitatively delineated the selective mobilization patterns of different displacement methods, identifying that water flooding targets larger pores while CO₂ flooding targets finer pores.
• Mechanism Elucidation: It clarified that the inefficiency of water flooding in ultra-low permeability reservoirs is due to high capillary resistance and threshold pressure gradients, whereas CO₂ overcomes these barriers through diffusion and physicochemical interactions.
• Complementarity: The research demonstrated that water flooding and CO₂ flooding are complementary at the pore scale: water provides effective pressure-driven displacement in larger channels, while CO₂ enhances microscopic mobilization in finer pores.

Significance and Claims
The authors state that these findings provide an experimental basis for selecting and sequencing enhanced oil recovery (EOR) methods in heterogeneous ultra-low-permeability sandy conglomerate reservoirs. The study suggests that a transition from water flooding to CO₂ flooding can effectively recover bypassed oil, particularly from mesopores and small pores. However, the authors maintain a modest scope, noting that these results are based on plug-scale experiments. They caution that while a complementary effect is observed at the core scale, establishing a universal reservoir-scale synergistic response requires further consideration of heterogeneity, injectivity, gas channeling risks, and specific operating conditions. The work advocates for a reservoir evaluation framework centered on pore structure and pore-throat size distribution rather than permeability alone to better predict oil mobilization potential.