Deep Learning and Time-Series Analysis for the Early Detection of Lost Circulation Incidents During Drilling Operations

Abstract

Drilling operations consist of breaking the rock to deepen a wellbore for oil or gas extraction. A drilling fluid, circulating from the surface through the drill pipe and from the annulus to the surface, is used to remove rock cuttings and maintain hydrostatic pressure. Drilling fluid lost circulation incidents (LCIs) are major sources of non-productive time (NPT) in drilling operations. These incidents occur due to preexisting natural fractures (vugs, caverns, etc.) and/or drilling-induced hydraulic fractures. The initiation of an LCI could lead to other hazardous drilling phenomena, such as formation influx or kick/blowout, stuck pipe incidents, among others. LCIs are typically monitored at the rig site by observing drilling fluid levels in the fluid tanks. This manual process incurs missing the occurrence or late detection of LCIs. Machine learning (ML) and deep learning (DL) classification algorithms are powerful in processing time-series data and achieving early detection of such temporal phenomena. In this study, we performed a large-scale analysis of the surface drilling and rheology data obtained from historical wells with LCIs. This analysis includes primary and secondary preprocessing steps including, aggressive sampling, feature engineering, and window normalization to derive generalizable DL models for real-time operations. Focal loss was utilized to account for data class imbalance and train robust and generalizable models. The results obtained from different ML/DL algorithms showed that one-dimensional convolutional neural network models resulted in the best performance with state-of-the-art precision, recall, and F1 scores of 87.34%, 73.40%, and 79.77%, respectively, on unseen test drilling data.

Anomaly Detection Using Explainable Random Forest for the Prediction of Undesirable Events in Oil Wells

“…Additionally, the offered system could alert the crew 9 hours and 7 hours before incidents took place in Well-1 and Well-2, respectively. Furthermore, Aljubran et al [14] developed a DL-based model for early predicting fluid lost circulation incidents (LCIs) in drilling operations. e dataset utilized in their study was based on an analysis of historical drilling data derived from standard drilling rig equipment and apparatus.…”

Forecasting the abnormal events at well drilling with machine learning

“…As was mentioned in Section 3.3, the general model quality was estimated using the ROC curve and ROC AUC metric. To benchmark model quality with other models, we trained Convolution Neural Network (CNN), similar to the approaches [5,31], which uses t segment values as features, and random classifier, which predicts accident probabilities for each time step in segment t randomly. CNN model has tree convolution units, where each convolution layer is followed by the batch normalization layer, dropout unit (p=0.1), and two fully connected linear layers with 128 and 32 input neurons, respectively.…”

“…A similar approach used mud telemetry data as input data and neural network as a classifier as described in papers [5,31], which allows forecasting sticking and loss of circulation accidents while drilling. A similar solution can also be applied for benchmarking our approach.…”

The application of deep learning algorithms to classify subsurface drilling lost circulation severity in large oil field datasets

“…In that study, the models were fed with the real-time data of a 16.7-min drilling period (recorded every 0.02 s) with high accuracy. Ljubran et al [48] also investigated the role of Random Forest (RF), Artificial Neural Network (ANN) and DL including LSTM and CNN models to predict lost circulation. That work revealed the superiority of CNN over other algorithms.…”