Implementing artificial neural networks and support vector machines to predict lost circulation

Abstract

Lost circulation is one of the major challenges encountered during drilling operations. The events related to the lost circulation can be responsible for losses of hundreds of millions of dollars each year. This paper presents a study on the application of artificial neural networks (ANNs) and support vector machine (SVM) to develop a robust system that can be used to predict the lost circulation occurrence. In the first step, field dataset, including drilling operation parameters, formation type, and lithology of the rock, as well as the drilling fluid characteristics, were collected from 385 wells drilled in southern Iraq from different fields. Then, the user-controlled parameters for ANNs (e.g., training function, number of hidden layers, transferring function, and number of neurons in each hidden layer) and SVMs (e.g., regularization factor, the type of kernel function, and its specific parameters) were optimized using the most common conventional performance criteria. Finally, the best-proposed models were examined using a few examples of real lost circulation cases from the field. The results of the analysis have revealed that both ANNs and SVM approaches can be of great use, with the SVM results being more promising. The application of the machine learning methods could assist drilling engineers in modifying drilling parameters to minimize the likelihood of lost circulation.