Rate of Penetration Prediction using Mathematical Drilling Models (A Case Study of Niger Delta Formation)

Danasabe I. T., Tanko A. and Akindele U. M.

Department of Mineral and Petroleum Engineering, College of Engineering, Kaduna Polytechnic, Kaduna

Keywords: Drilling Efficiency; Rate of Penetration; Burgoyne and Young Model; Exploratory Well; Maurer's model; Bingham's model; Maurer's model.

Abstract

The rate of penetration (ROP) prediction is one of the most important factors in improving drilling efficiency and minimizing drilling cost. Mathematical drilling models provide a technique to predict and control drilling process. The prediction of rate of penetration using well data from an exploratory well in Niger Delta has been conducted. Three models were assessed with various degrees of precision. The predictive accuracy of the models was confirmed by comparison with actual ROP data and calculation of Mean Absolute Percentage Errors. The B & Y's model presented the best matches with the actual ROP data in comparison to the other two models used in the study, with a Mean Absolute Percentage Error of 25%. The Bingham's model, followed closely with good matches with the actual ROP data with a Mean Absolute Percentage Error of 35%. Conversely, Maurer's model showed a great degree of discrepancy with the actual data by over-predicting the ROP across the total well depth, thereby yielding a Mean Absolute Percentage Error of 3188%. The outcome shows that B & Yis the best model for ROP prediction in the Niger Delta sedimentary basin going by its low Mean Absolute Percentage Error while Maurer's model is the least accurate. The model can forecast ROP as a function of several drilling parameters such as WOB, RPM, Mud weight, SSP, Torque, flow rate etc. with a sufficient precision.