Real-Time Yield Point Prediction for Water-Based Drilling Mud using Particle Swarm Optimised Neural Network

Abstract

Yield point (YP) is an essential rheological property of drilling mud that measures the fluid's resistance to initial flow, giving insight into the attractive forces among the solids in the mud. YP influences the ability of the mud to lift well cuttings from the annulus to the surface, impacting the overall drilling efficiency. Despite its significance, YP is typically measured only once or twice a day using complex rheometers, which can be time-consuming and costly. To address these limitations, this study investigates the application of artificial intelligence (AI), specifically particle swarm optimisation-back propagation neural network (PSO-BPNN), for real-time prediction of the yield point from Marsh funnel experimental data. Unlike the rheometers, Marsh funnel experiment is conducted every 10 to 15 minutes using less complex field instruments. It was identified from the study that PSO-BPNN outperformed BPNN in the estimation of YP in terms of correlation coefficient (R) and mean square error (MSE). During testing PSO-BPNN attained 0.923 and 1.129 as R and MSE score, respectively, while BPNN had 0.604 and 1.733 for the R and MSE score, respectively. The findings suggest that PSO-BPNN offers a more reliable and efficient approach to predicting drilling fluid yield point from Marsh funnel experimentation.