Estimation of Pressure Loss of Herschel-Bulkley Drilling Fluids During Horizontal Annulus Using Artificial Neural Network

Accurate estimation of the pressure losses for non-Newtonian drilling fluids inside annulus is quite important to determine pump rates and select mud pump systems during drilling operations. Therefore, in this study, pressure losses of Herschel-Bulkley drilling fluids in concentric and eccentric annulus are predicted using simple, reliable, and cost-effective artificial neural network (ANN) method. The average relative error was less than 5% with correlation coefficient (R) of 0.999 for the prediction of pressure loss (P) taking the ratio of pipe diameter to casing diameter (D-i/D-o), eccentricity of annulus (E), and properties of the non-Newtonian liquid, that is, flow behavior index (n), consistency index (K), yield stress ((y)), and liquid flow rate (Q) as inputs to an ANN for Herschel-Bulkley fluids. Experimental data from the literature were used to train the ANN for predicting pressure losses in eccentric annuli.