Estimation of pressure drop of single-phase flow in horizontal long pipes using artificial neural network

Abstract

Large- pressure drop and drag along the pipe route is one of the problems with fluid transfer lines. For many years, various methods have been employed to reduce the drag in fluid transmission lines. One of the best ways for this purpose is reducing friction coefficients by utilizing drag lowering materials. Experimentally by adding minimal amounts of this material at the ppm scale to the lines and reducing the drag of the flow, fluid can be pumped without the need to change the size of the pipe. In this study, the effect of carboxymethylcellulose biopolymer on the water flow reduction in a 12.7- and 25.4-mm galvanized pipe was investigated. In order to have a comprehensive analysis of process conditions, experiments were carried out with three different levels of concentrations, flow rate and temperature. Also, as a new innovation in this investigation, the outputs of the experimental data were evaluated and analyzed using the Taguchi method and neural network system, and optimized through a genetic algorithm. In this study, the highest rate of drag reduction will be achieved at 39 ° C and at a concentration of 991.6 ppm and flow rate of 1441.1L/h was 59.83% at 12.7-mm diameter.