Prediction of drilling fluid lost-circulation zone based on deep learning

Lost circulation has become a crucial technical problem that restricts the quality and efficiency improvement of the drilling operation in deep oil and gas wells. The lost-circulation zone prediction has always been a hot and difficult research topic on the prevention and control of lost circulation. This study applied machine learning and statistical methods to deeply mine 105 groups and 29 features of loss data from typical loss block M. After removing 10 sets of noise data, the methods of mean removal, range scaling and normalization were used to pre-treat the 95 sets of the loss data. The multi-factor analysis of variance (ANOVA) and random forest algorithm were adopted to determine the 13 main factors affecting the lost circulation. The three typical deep learning neural network models were improved, the parameters in the models were adjusted, the neural network models with different structures were compared according to the PR curves, and the best model structure was built. The pre-treated loss data in 95 sets with 13 features were divided into the training set and test set by a ratio of 4:1. The model performance was evaluated using F1 score, accuracy, and recall rate. The trained model was successfully applied to the G block with severe leakage. The results show that the capsule network model is better than the BP neural network model and the convolutional neural network model. It stabilizes at 300 training rounds, with a prediction accuracy of 94.73%. The improved model can be applied to lost-circulation control in the field and provide guidance on leakage prevention and plugging operations.