Microscopic pore-throat grading evaluation in a tight oil reservoir using machine learning: a case study of the Fuyu oil layer in Bayanchagan area, Songliao Basin central depression

Tight oil reservoirs are key research targets of petroleum geological exploration. Due to their strong heterogeneity, reservoir classification is fuzzy and efficiency is low, which affects reservoir evaluation accuracy. This study aims to improve the accuracy of the evaluation of microscopic pore throats of tight reservoir. Tight oil reservoir is classified on the basis of microscopic pore-throat grading with machine learning technology including Box-Cox transformations, Grey relational analysis, Q-cluster analysis, and discriminant analysis, in combination with traditional measures of reservoir quality in the Fuyu oil layer, Bayanchagan area, Songliao Basin. The results show that Class I reservoirs are high-quality, medium compaction-strong dissolution-weak cementation reservoirs, with large pore-throat radii, concentrated from 0.16 μm to 1.00 μm. Class II reservoirs have medium potential, with strong compaction-medium dissolution-medium cementation. Pore throats have relatively uniform distribution with low connectivity, and radii mainly between 0.016 μm to 0.16 μm. The Class III reservoir has the lowest potential and strong heterogeneity, with strong compaction, weak dissolution, and strong cementation. Small pore-throat radii range from 0.002 μm to 0.016 μm. We infer that this method can effectively evaluate tight oil reservoirs and provide a basis for the prediction of favorable areas for tight oil exploration and development.