Machine Learning Technique Improves Resistivity Logging Data in Oil and Gas Industry

Strategic Insights -

Researchers at Shandong University have introduced a machine learning technique based on "training while drilling" to enhance the integrity and accuracy of resistivity logging data in the oil and gas industry. This approach integrates a Long Short-Term Memory (LSTM) neural network into embedded devices, allowing for real-time downhole data completion and improving the efficiency and immediacy of data processing. The method, which utilizes downtime for model updating, has demonstrated notable improvements in predictive accuracy, outperforming other deep learning algorithms.

Key Takeaways:

  • The research proposes a machine learning technique based on "training while drilling" to improve resistivity logging data accuracy in the oil and gas industry.
  • The technique integrates an LSTM neural network into embedded devices for real-time downhole data completion and enhances the efficiency and immediacy of data processing.
  • The method utilizes downtime for model updating and has demonstrated notable improvements in predictive accuracy.
  • The LSTM model outperformed other deep learning algorithms, including Fully Connected Neural Networks (FCNN), in terms of prediction accuracy and stability.
  • The dataset used in the study was based on field data from the Shengli Oilfield in Dongying, spanning depths of 1140 to 1690 m.
  • The LSTM model achieved a mean squared error (MSE) of 0.0610, significantly lower than the MSE of 0.0961 observed for the FCNN and other traditional methods.
  • The research has been peer-reviewed and has been published in the journal Earth Science Informatics.
  • The study was financially supported by Sinopec and Key Technology Research on MatriNavi II Rotary Steerable System.

Statistics:

  • 1140-1690 m: depths of the field data used in the study
  • 0.0610: MSE achieved by the LSTM model
  • 0.0961: MSE observed for the FCNN and other traditional methods
  • 95.1%: improvement in predictive accuracy compared to the initially trained model

Sources:

  • NewsRx. Reports from Shandong University Provide New Insights into Information and Data Processing (A Method for Training While Drilling To Predict Electromagnetic Wave Logging Curves Based On Long Short-term Memory Neural Networks). Information Technology Newsweekly. September 2, 2025; p 668.
  • A Method for Training While Drilling To Predict Electromagnetic Wave Logging Curves Based On Long Short-term Memory Neural Networks. Earth Science Informatics, 2025;18(3).