Sophisticated ROP Prediction Technologies Based on Neural Network Delivers Accurate Drill Time Results

Abstract

Abstract Accurately predicting the rate of penetration (ROP) is a prominent factor affecting economic and engineering decisions during well planning. However, ROP prediction based on simple algorithms applied to offset wells has historically yielded mixed results. To improve ROP predicting capabilities, the provider is applying an artificial neural network (ANN) to analyze offset drilling data. The system has significantly improved the ability to accurately predict drilling performance, despite expected changes in lithology, hole size, bit type and mud properties. The flexibility of the software package allows engineers to analyze a wide range of information and deliver high-quality ROP predictions based on previous experience and data from offset wells. The process includes detail scrutiny of the offset and training of the ANN system until the neural network is validated. The simulation is then run to solve expected ROP, using any changed drilling conditions as input. In a recent case, an operator required a well that would have to be drilled deeper than offsets and with different borehole sizes. By applying the neural network capabilities, engineers were able to deliver analytical ROP predictions for the planned well, including quality result for different lithologies with a wide range of rock strength values. This paper will focus on the new ANN technologies being utilized for predicting drilling performance. It will also include a case study that documents its successful application.