Using Different Artificial Neural Network Models to Predict Rate of Penetration for Oil and Gas Wells

Rate of penetration (ROP) is an important parameter that determines the success and cost of a drilling operation. In this paper, the research approach is based on different artificial neural network (ANN) models to predict ROP for oil and gas wells in Nam Con Son basin. First, the study will collect and evaluate drilling parameters as input data for the model. Second, using the data after evaluation, train the network and then create an optimal network capable of predicting ROP with high accuracy. The evaluation and ranking of input parameters obtained from the drilling process shows that the parameters that greatly affect the ROP prediction of the model are weight on bit (WOB), mud weight (MW) etc. The construction of local and global ANN models with different input data also leads to many different results, in which the global model is evaluated higher based on the mean square error. (MSE) and the correlation coefficient R squared (R²). The ROP prediction results obtained from different ANN models are also compared with traditional models such as Bingham model, Bourgoyne & Young’s model. This comparison has shown that the predictive power of ANN models is higher than that of traditional models. These results have shown the competitiveness of the ANN models and its high applicability to drilling operations.