Answers are grounded in this paper and cite their sources.
A new method of quantifying modal decomposition and grouping pore-size distributions is presented. This approach enables the identification of sample groups with distinct fluid displacement phenomena. A lognormal Gaussian decomposition model was used to fit the observed multimodal pore size distributions of rock samples obtained from 2D SEM images. The model exhibited high fitting accuracy, with coefficients of determination (R2) approaching 1 and a minimum value of 0.81 for cumulative distribution function (CDF) and probability distribution function (PDF) data respectively. This confirms the robustness of the fit. Notably, key Gaussian fitting parameters, such as weighting factor (W) and mean values (M) of the decomposed distributions, exhibited strong correlations with core permeability across two well locations. This validation underscores the effectiveness of this approach in classifying fluid displacement using the Gaussian parameters. Cluster analysis classified rock samples into three distinct groups based on pore structural features derived from Gaussian parameters and permeability, a key petrophysical property. W1 represents Gaussian weighting factor of mode-1 subpopulations, W23 reflects well-connected Gaussian W of mode-2 and mode-3 subpopulations, while M2, M3, and M23 define the mean pore sizes of mode-2 and mode-3 subpopulations, characterising dominant peaks in the pore volume distribution. Group 1 (Low fluid flow) exhibited high capillary forces and low permeability, limiting fluid mobility and making movement challenging for monitoring fluid flow. Group 2 (Transitional flow behaviour) facilitated moderate, predictable fluid flow with uniform flow paths, reducing fingering effects and making it ideal for water flooding, underground gas storage, and monitoring. Group 3 (High fluid flow) displayed complex flow behaviour with preferential pathways causing uneven displacement fronts, pronounced fingering due to heterogeneity, and high permeability. This methodology enhances reservoir characterisation by revealing pore size distribution heterogeneity. This provides valuable insights for optimising hydrocarbon recovery, carbon storage, and groundwater remediation, ensuring more effective management of fluid displacement dynamics. • Introduced Lognormal Gaussian Decomposition Model for fluid dynamics in porous rock. • Novel multimodal Gaussian model enhances pore system analysis from SEM images. • Cluster analysis groups pore systems into distinct categories for fluid flow classification. • Gaussian decomposition improves permeability estimation across reservoir formations.
No takes yet. Share an insight, caveat, or question.
Agunwoke et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: