A Statistics-Based Correlation for the Assessment of Water-Based Mud Performance: Yield Point Prediction

A suitable design of drilling fluids is essential to improve the drilling efficiency and decrease drilling time/cost. Yield point is one of the important rheological properties of drilling fluids used to delineate the performance of a drilling fluid. Although laboratory measurements of the YP are possible, they are time-consuming and costly. Therefore, predicting YP at different conditions and additive concentrations is quite advantageous. This work used response surface methodology (RSM) to build an accurate correlation that predicts YP based on pressure, temperature, and concentrations of nanosilica, nanoclay, and bentonite. The correlation was statistically analyzed by validating it through the F-statistics test, ANOVA, fit statistics, Pareto chart, and diagnostics plots to increase its accuracy and degree of confidence. The YP correlation was verified by using unseen data to check its efficiency. The influences between the predictor and target variables were shown in the contour and 3D surface response plots. The correlation results fairly agree with the measured data with a Pearson's correlation coefficient of 0.9947. The results show that the temperature and concentrations of bentonite and nanosilica highly affect the YP in water-based drilling fluids.