Main controlling factors of fracturing and productivity prediction of low permeability reservoirs in Nanpu Sag based on LIBSVM model
Year: 2025
Type: conference-paper
Abstract: Block Z of Nanpu Sag in Bohai Bay Basin is a low permeability reservoir. The implemented fracturing practice has effectively improved the reservoir plane and interlayer production conditions. However, there is still a lack of understanding in terms of the main controlling factors of fracturing and production capacity evaluation, which restricts the improvement of fracturing technology. To solve this problem, an intelligent production capacity prediction method based on LIBSVM model is proposed, aiming to achieve accurate prediction of fracturing production capacity through machine learning and improve fracturing efficiency. Based on the geological parameters, engineering parameters and full-cycle oil increase data of 120 fractured wells, six key factors closely related to production capacity prediction were screened out through correlation analysis of characteristic parameters: remaining recoverable reserves, permeability, pressure coefficient, sand addition amount, designed half-fracture length and designed fracture height. The LIBSVM algorithm based on radial basis kernel function (RBF) was used to train and test these main controlling factors, and a fracturing production capacity prediction model was constructed. The test results show that the model maintains a prediction accuracy of 90% under different construction pressures and fracturing volumes, verifying its effectiveness and practicality. Compared with neural networks, this method has low computational complexity, can avoid overfitting, and has better generalization ability. The research results not only have important guiding significance for fracturing well selection and layer selection and engineering parameter optimization of low permeability reservoirs in Nanpu Sag, but also show strong practical value in capacity prediction.
Authors: Ying Zhang, Jie Zhang, Xianmei Wang, Zhimin Wu, Xiaoge Li, Lili Qu
Language: English
Subfield: Mechanical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4414126483