Neural network-based reservoir micropore structure evaluation method and device

Abstract

Embodiments of the invention provide a neural network-based reservoir micropore structure evaluation method and device. The method comprises steps of: acquiring rock core pore structure data obtainedbased on a rock core high pressure mercury experiment, and the type of a rock core pore structure corresponding to the rock core pore structure data; aligning rock core depth and well measurement depth according to a density curve in rock core density and well measurement data, obtaining well measurement data corresponding to the rock core, and obtaining pre-processed well measurement data; training an initial neural network model with pre-processed well measurement data as input and the type of the rock core pore structure as output, and obtaining a trained neural network model; and inputtingtarget well measurement data into the trained neural network model when the reservoir microstructure of the target well needs to evaluate, and obtaining pore type corresponding to the target well measurement data. The reservoir micropore structure can be evaluated rapidly and accurately.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of oil and gas field development, and in particular to a neural network-based evaluation method and device for reservoir microscopic pore structure. Background technique [0002] Oil and gas are stored and flow in the pores of rocks, so the shape, size, development and connectivity of rock pores, that is, the pore structure of rocks, can directly affect the storage capacity and productivity of oil and gas, especially the microscopic Pore ​​structure is the basis for studying the physical properties and storage properties of rocks, and is of great significance to the exploration and development of oil and gas resources. [0003] At present, the methods for studying the microstructure of reservoirs are mainly divided into two categories. The first category is core analysis, which mainly includes capillary force curve method, cast thin section method and scanning electron microscope me...

AI Technical Summary

Problems solved by technology

[0004] However, the core analysis method is generally manual identification, and the analysis speed is slow; the logging data field evaluation method mainly reflects the macroscopic pore structure of the rock, and the identification accuracy is low

Images