Abstract: Engineering simulators used for steady-state multiphase pipe flows arecommonly utilized to predict pressure drop. Such simulators are typically basedon either empirical correlations or first-principles mechanistic models. Thesimulators allow evaluating the pressure drop in multiphase pipe flow withacceptable accuracy. However, the only shortcoming of these correlations andmechanistic models is their applicability. In order to extend the applicabilityand the accuracy of the existing accessible methods, a method of pressure dropcalculation in the pipeline is proposed. The method is based on wellsegmentation and calculation of the pressure gradient in each segment usingthree surrogate models based on Machine Learning algorithms trained on arepresentative lab data set from the open literature. The first model predictsthe value of a liquid holdup in the segment, the second one determines the flowpattern, and the third one is used to estimate the pressure gradient. To buildthese models, several ML algorithms are trained such as Random Forest, GradientBoosting Decision Trees, Support Vector Machine, and Artificial Neural Network,and their predictive abilities are cross-compared. The proposed method forpressure gradient calculation yields $R^2 = 0.95$ by using the GradientBoosting algorithm as compared with $R^2 = 0.92$ in case of Mukherjee and Brillcorrelation and $R^2 = 0.91$ when a combination of Ansari and Xiao mechanisticmodels is utilized. The method for pressure drop prediction is also validatedon three real field cases. Validation indicates that the proposed model yieldsthe following coefficients of determination: $R^2 = 0.806, 0.815$ and 0.99 ascompared with the highest values obtained by commonly used techniques: $R^2 =0.82$ (Beggs and Brill correlation), $R^2 = 0.823$ (Mukherjee and Brillcorrelation) and $R^2 = 0.98$ (Beggs and Brill correlation).