Vahid Mojarradi - Department of Petroleum Engineering, Shahid Bahonar University of Kerman, Kerman, Iran Energy and Environmental Research Center (EERC), Shahid Bahonar University of Kerman, Kerman, Iran
Mohammad Ranjbar - Department of Mining Engineering, Shahid Bahonar University of Kerman, Kerman, Iran
Khabat Amiri Hoseini - Institute of iron ore and steel, Kerman, Iran

Horizontal directional drilling (HDD) is widely used in petroleum industry because of its application to reduce costs, drilling in offshores, drilling multilateral wells and several other benefits. In a variety ofconditions it is necessary to predict the torque required for performing the reaming operation (nearsurface), However, because of complicated mechanical and geomechanical parameters which effect rotational torque, presently, there is no appropriate and convenient method to accomplish these tasks. The objective of this study is to improve the prediction of the torque during drilling operations usingFuzzy C means clustering (FCM).Although available information for drilling techniques does provide some means of predicting the torque, it is not sufficient for meeting the present needs. In this study FCM is used to predict the value of torque in horizontal drilling based on operational field data. Number ofclusters which is most effective parameter on FCM performance, is changed from 1 to 84 (number of data) to find best condition for FCM. Results showed that this model could predict rotational torque with correlation coefficient with real data equal to R2= 8759 for Sugeno FIS and R2= 8452 for Mamdani FIS. So this method could predict Rotational torque better than previous numerical and analytical models

Fuzzy C means Clustering, Horizontal Drilling, Rotational Torque