Modelling Filtration Loss Rate of Nanoparticles Based Mud at Downhole Conditions Using Neural Network

Student: PRINCE USEN WILLIE (Project, 2025)
Department of Petroleum Engineering
University of Uyo, Akwa Ibom State

Abstract

ABSTRACT Optimizing mud filtrate invasion is crucial in oil and gas drilling to minimize formation damage. Recent studies indicate that nanoparticles (NPs) present promising outcomes in mitigating filtrate loss when incorporated as additives in drilling fluid (mud). Modeling the impact of NPs accelerates the selection of the optimal type, size, concentration, etc., streamlining the process to align with specific drilling conditions. In this research, an artificial neural network (ANN) model was formulated to forecast the filtrate invasion of nano-based mud across a broad spectrum of pressures and temperatures, reaching up to 500 psi and 300°F, respectively. The model was constructed using a dataset comprising 1,003 data points. The dataset was partitioned into 70% for training and 30% for validation. The model's performance was assessed by calculating statistical parameters. The developed ANN model demonstrated efficiency in predicting filtrate invasion across different pressures and temperatures, with a mean square error (MSE) of 0 00065492 and a coefficient of determination (R2) exceeding 0.9774 for the entire dataset. This ANN model spans a wide range of pressures, temperatures, and various types, concentrations, and sizes of nanoparticles (NPs), confirming its usability and comprehensiveness.

Keywords
Nanoparticle Drilling Fluid Mud filtrate Artificial Neural Network
Filters
Institutions