Real-Time Solution for Down Hole Torque Estimation and Drilling Optimization in High Deviated Wells Using Artificial Intelligence

Abstract

AbstractThis project provides a new realistic solution for the accuracy of down hole torque measurements using the integration of the Artificial intelligence (AI) technology with the downhole challenges being faced while drilling deep and high deviated wells. The new estimates are based on surface measurements which have the major influence on the bit torque (downhole torque) values while drilling. Artificial intelligence technology and its related applications such as; artificial neural network (ANN), support vector machine (SVM) and adaptive neuro fuzzy interference system (ANFIS) will be utilized to predict and estimate accurate wellbore torque which will be applied effectively to prevent real time stuck pipe situation through a friendly user software which will maintain the downhole torque within the SAFE zone by controlling the unified surface drilling variables such as; weight on bit (WOB), Rate of Penetration (ROP) and Flow Rate.This downhole torque model will be validated and verified through a real drilling scenario from a field in north of Africa. The field data includes weight on bit, surface torque, stand-pipe pressure, and rate of penetration were collected from the mentioned well which had experienced a costly stuck pipe situation. However, with the provided model the same encountered scenario will be avoided, due to the optimization of the real time drilling variables and hence, saving the well and evade a costly non-productive time.