relation: https://khub.utp.edu.my/scholars/19905/
title: Explicit Data-Based Model for Predicting Oil-Based Mud Viscosity at Downhole Conditions
creator: Alkouh, A.
creator: Elraies, K.
creator: Agwu, O.E.
creator: Alatefi, S.
creator: Azim, R.A.
description: One of the functions of drilling mud is to transport cuttings out of a drilled well. To fulfill this purpose, the mud must possess optimal viscous characteristics. The main goal of this study is to develop a model for estimating the plastic viscosity (PV) of oil-based muds (OBMs) at downhole conditions, especially when high-temperature, high-pressure wells are drilled. Existing methods used by previous researchers include the following: the multiplicative factor method, the relative dial readings approach, and the nonlinear regression technique. These methods do not result in flexible, generalizable, accurate, or robust models. Furthermore, existing literature mostly contains models that are only applicable under normal temperature and pressure conditions (surface conditions). To overcome these challenges, this study employed the artificial neural network (ANN) technique to predict the PV of OBM under downhole conditions. The data used to develop the model consists of 88 OBM laboratory PV measurements obtained from literature. The findings indicate that the performance of the developed model using the statistical metrics of means square error (MSE), root-mean-square error (RMSE), and correlation coefficient (R) was 0.0185, 0.136, and 0.967 respectively. To test for the generalizability of the developed model, a new data set consisting of 56 data points was used. In this regard, the model had an R value of 0.80, an MSE of 0.95, and an RMSE of 0.975. In order to ascertain the parametric importance of the inputs, a connection weight algorithm was utilized. In this regard, the downhole pressure had a higher influence on the PV (64.5) than that of downhole temperature (35.5). To make the models simple to incorporate in software applications, they were explicitly presented. In terms of memory requirements and processing speed for software applications, the models had a memory footprint of 48 bytes and required 12 floating-point operations to give an output. This model is valid for 20 °C � temperature �315 °C and 0 psi � pressure �40,000 psi. The characteristics of the models proposed in this work for which originality is claimed include the models� computational cost evaluation, their explicitness, and a suggestion for using them in the field. Upon utilizing the proposed models, time-consuming laboratory measurements of PV would no longer be necessary, and real-time results could be provided in the field. © 2024 The Authors. Published by American Chemical Society.
publisher: American Chemical Society
date: 2024
type: Article
type: PeerReviewed
identifier:   Alkouh, A. and Elraies, K. and Agwu, O.E. and Alatefi, S. and Azim, R.A.  (2024) Explicit Data-Based Model for Predicting Oil-Based Mud Viscosity at Downhole Conditions.  ACS Omega, 9 (6).  pp. 6684-6695.  ISSN 24701343     
relation: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85184823065&doi=10.1021%2facsomega.3c07815&partnerID=40&md5=58176a3b5d5eec8d9be308a21cf0c002
relation: 10.1021/acsomega.3c07815
identifier: 10.1021/acsomega.3c07815