Zhengshan Qin

Dry-out and salt precipitation (DSP) near the wellbore is a frequently observed phenomenon in underground gas storage during injection–withdrawal cycles. It causes pore-space contraction and localized pore–throat clogging, thereby reducing reservoir porosity and permeability. However, existing clogging relations often fail to explicitly represent how clogging material occupies and evolves within the pore space, and they also struggle to capture the strongly nonlinear permeability response in a unified manner. To address these limitations, we conducted long-term gas-flooding experiments and measured key petrophysical properties and clogging-state variables before and after DSP. Mercury intrusion porosimetry (MIP) combined with fractal analysis was further used to characterize the evolution of salt-precipitation–induced microscale heterogeneity. Building on two classical conceptual frameworks—capillary bundles (CB) and equal-diameter sphere packings (EDSP)—we propose a clogging state-variable–driven strategy to construct porosity–permeability relations for occupancy-type clogging. By incorporating dual-fractal theory, we further establish a generalized porosity–permeability relation that simultaneously accounts for cumulative pore cross-sectional-area scaling (described by Ds) and volumetric space filling of the connected pore network (captured by the pore-volume fractal dimension Dv), together with a parameter-reduced form for data-limited applications. The CB and EDSP relations emerge as assumption-dependent, geometry-specific special cases of the proposed dual-fractal framework. Experiments show that DSP decreases both porosity and permeability, with permeability being markedly more sensitive. Permeability exhibits pronounced nonlinear decay with increasing solid-salt saturation and clogging factor. The MIP-derived Dv exhibits physically interpretable associations with petrophysical properties and clogging-state variables, whereas the corresponding Ds, inferred from Dv through the stated additive approximation, provides a model-consistent representation of pore-surface scaling rather than independent experimental evidence. Taken together, these results indicate that salt clogging is manifested not only by pore-volume reduction but also by changes in the effective scaling characteristics and flow-relevant heterogeneity of the connected pore system. Evaluations across relations and published DSP and other occupancy-type clogging datasets confirm that the generalized dual-fractal relation provides the best overall in-sample fit among the four proposed forms, while the CB/EDSP and simplified-fractal forms provide comparable descriptions of the dominant trend and remain useful lower-complexity alternatives in data-limited settings. This work provides a mechanism-informed and implementable framework for interpreting and predicting DSP-induced permeability impairment, and it supports embedding clogging relations into numerical simulators for performance assessment in occupancy-dominated clogging scenarios.

Based on dual-fractal theory, this paper proposes a novel methodology to transform nuclear magnetic resonance (NMR) measurements into relative permeability model, thereby realized the challenge of accurately quantifying multiphase fluid flow in dual-fractal porous media. The novel relative permeability model integrates a multitude of parameters associated with the microscopic pore architecture of porous media. It can be verified through case studies that the newly constructed relative permeability model demonstrates excellent applicability. Furthermore, we have discovered a correlation between the segmentation characteristics of the relationship between the pore radius (r) and transverse relaxation time (T2) in low-permeability sandstone reservoirs and the number of peaks in the NMR curve. Specifically, the single-peak type can usually be divided into two segments, while the multi-peak type can generally be divided into three segments. The comprehensive fractal dimension (Dc) derived from the weighting method effectively captures the holistic core heterogeneity. The characteristics of small diameter pores are particularly sensitive to variations in the pore tortuosity fractal dimension (Dτ), the pore structure fractal dimension (Df), Dc, and ε (represents the pore size ratios). Given that the wetting phase preferentially occupies small pores, its relative permeability is significantly affected by these parameter changes. Conversely, the non-wetting phase mainly flows through large pores and is thus less influenced by changes in these parameters.