Modeling Rate of Penetration for Deviated Wells Using Artificial Neural Network

Abstract

The advanced technology has made directional drilling widely used to enhance the production of mature fields. The rate of penetration (ROP) contributes strongly towards the cost of drilling operations, where achieving higher ROP leads to substantial cost saving. The main objective of this study is to develop a model that predicts the ROP for deviated wells using artificial neural networks (ANNs). The model was developed based on the most critical variables affecting ROP using ANNs. In addition to the azimuth and inclination of the well trajectory, the controllable drilling parameters, unconfined compressive strength (UCS), pore pressure, and in-situ stresses of the studied area were included as inputs. 1D Mechanical earth modeling (1D-MEM) data, geophysical logs, daily drilling reports, and mud logs (master logs) of deviated wells drilled in Zubair field located in Southern Iraq were used to develop the ANN model. The results displayed that the ANN’s outputs are close to the measured field data. The correlation coefficient (R) and average absolute percentage error (AAPE) were over 0.91 and 8.3%, respectively, for the training dataset. For testing data, the developed model achieved a reasonable correlation coefficient (R) of 0.89 and average absolute percentage error (AAPE) of 9.6%. Unlike previous studies, this paper investigates the effect of well trajectory’s (azimuth and inclination) and their influence on the ROP for deviated wells. The major advantage of the present study is calculating approximately the drilling time of the deviated well and eventually reducing the drilling cost for future neighboring wells.