Title: Drilling efficiency enhancement in oil and gas domain using machine learning

Authors: Aditi Nautiyal; Amit Kumar Mishra

Addresses: School of Computing, DIT University, Dehradun, India ' School of Computing, DIT University, Dehradun, India

Abstract: Oil and gas are non-renewable natural resources that require a great marvel of engineering, planning, and huge capital investment. Drilling efficiency enhancement means increasing the rate of penetration (ROP) and avoiding downhole complications like stuck pipe, wellbore collapse, etc. Drilling is a complex activity, prediction of ROP and downhole complications depend upon the various drilling parameters, bottom hole assembly designing, drill bit selection, mud type, mud weight selection, well trajectory designing, revolutions per minute (RPM), and formation parameters which exhibit linear or nonlinear relationships with the objective function. This research work is focused on machine learning algorithms like random forest, artificial neural network, etc., for the development of the ROP prediction model. Hyperparameter tunning methods like RandomisedSearchCV() were used to enhance model accuracy, the evolutionary particle swarm optimisation (PSO) algorithm to maximise ROP, and an exhaustive list of parameters deemed necessary for the development of an accurate and generalised ML model. [Received: June 28, 2022; Accepted: September 18, 2022]

Keywords: drilling efficiency enhancement; machine learning; random forest; artificial neural network; ANN; rate of penetration; ROP; downhole complications.

DOI: 10.1504/IJOGCT.2023.129577

International Journal of Oil, Gas and Coal Technology, 2023 Vol.32 No.4, pp.340 - 373

Received: 18 Jun 2022
Accepted: 18 Sep 2022
Published online: 14 Mar 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article