Production Capacity Prediction of Tight Sandstone in Z-area Based on ARO-BP Algorithm

Abstract

Tight sandstone reservoirs make capacity prediction a challenging task due to their strong non-homogeneity, complex microporous
structure, and poor correlation between porosity and permeability. The purpose of this paper is to discuss the methods and techniques for
predicting the capacity of tight sandstone reservoirs and to provide a scientific basis for petroleum exploration and development. By comprehensively analyzing geological, logging and production data, this paper proposes a capacity prediction model based on the artificial rabbit
optimized inverse neural network algorithm, to which the logging parameters are preprocessed and input to predict the capacity, and verifies
its effectiveness and accuracy. It not only realizes the qualitative evaluation of single-well production capacity of production wells, but also
provides the support of production capacity prediction for new drilling wells to be put into production at a later stage, which has a positive
role in promoting the production capacity construction and deployment of ultra-low-permeability reservoirs.