A method and device for intelligently predicting natural gas productivity of a well subzone combination test

By dividing the reservoir into smaller layers and constructing machine learning models and deep learning algorithms, and integrating reservoir parameters and inter-layer interference coefficients, the problem of multi-layer reservoir parameter integration and inter-layer interference in traditional natural gas production capacity prediction is solved, and high-precision natural gas combined well test production capacity prediction is achieved.

CN120952263BActive Publication Date: 2026-06-26RICHFIT INFORMATION TECH +1
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
RICHFIT INFORMATION TECH
Filing Date
2025-09-25
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

Traditional natural gas production capacity forecasting methods cannot effectively integrate multi-layer reservoir parameters and inter-layer interference factors, resulting in low production capacity forecasting accuracy in scenarios such as tight gas and low-permeability multi-layer combined testing, which makes it difficult to support scientific mining decisions.

Method used

By dividing the target gas well reservoir into sub-layers, a machine learning model for predicting the production capacity of different types of sub-layers is constructed to determine the inter-layer interference coefficient. Then, a deep learning algorithm is used to integrate reservoir sub-layer parameters, single-layer production capacity and interference coefficient to establish a combined logging sub-layer production capacity prediction model.

Benefits of technology

It improves the accuracy and efficiency of natural gas well production prediction, solves the prediction distortion problem caused by ignoring the heterogeneity of small layers and inter-layer interference factors in traditional methods, and achieves high-precision production prediction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120952263B_ABST
    Figure CN120952263B_ABST
Patent Text Reader

Abstract

The application discloses a kind of intelligent prediction methods and devices of natural gas productivity of well logging small layer combination test, wherein the method comprises: according to the multiple well logging curve data of the target gas well obtained, the multiple target small layers of the target gas well reservoir are divided, the corresponding productivity prediction machine learning model is respectively constructed, single layer daily gas production and single layer daily water production are predicted;For multiple target small layers, determine the interference coefficient representing the interlayer interference of multiple target small layers;According to the interference coefficient, the single layer daily gas production and single layer daily water production of each target small layer, the well logging small layer combination test productivity prediction result after deducting interference loss is obtained;And / or, using deep learning algorithm, establish well logging small layer combination test productivity prediction model, output corresponding target small layer combination test combination test productivity prediction result.The application is used to improve the efficiency and accuracy of natural gas combination test well productivity prediction.
Need to check novelty before this filing date? Find Prior Art