TY - JOUR
TI - Machine learning-based evaluation of acidizing effectiveness and optimization of acidizing parameters for carbonate gas reservoir horizontal wells
AU - Ji Shi-Wen
AU - Zheng You-Cheng
AU - Zhang Yan
AU - Lu Xiao-Feng
AU - Zhu Si-Rong
AU - Ye Kai
JN - Thermal Science
PY - 2025
VL - 29
IS - 2
SP - 1043
EP - 1048
PT - Article
AB - The Dengsi formation in the Gaoshiti-Moxi block of the Sichuan Basin is  characterized as having low porosity and low permeability. Typically, the  development is carried out using horizontal wells and segmented acid  fracturing techniques. In this study, based on a data-driven approach,  geological, engineering, and well testing data were collected from 22  horizontal wells in the study area. Then, a high precision acid fracturing  productivity model was established using Gaussian process regression. This  model exhibited a high level of prediction accuracy, with an average  relative error of only 8.77% for the test dataset. Furthermore, leveraging  the established productivity model and employing a particle swarm  optimization algorithm, research was conducted to optimize acid fracturing  parameters and predict well productivity. The practical application of this  approach in one well yielded favorable results, which hold promise for  providing guidance on segmented acid fracturing design in the study area.