%0 Journal Article
%T Numerical modeling of three-phase flow through a Venturi meter using the LSSVM algorithm
%J Journal of Computational & Applied Research in Mechanical Engineering (JCARME)
%I Shahid Rajaee Teacher Training University (SRTTU)
%Z 2228-7922
%A Khayat, Omid
%A Afarideh, Hossein
%D 2020
%\ 09/01/2020
%V 10
%N 1
%P 153-170
%! Numerical modeling of three-phase flow through a Venturi meter using the LSSVM algorithm
%K Measurement
%K Three phase flow
%K Venturi meter
%K Computational Fluid Dynamics
%K LSSVM algorithm
%R 10.22061/jcarme.2019.3637.1423
%X One of the challenging problems in the Oil & Gas industry is accurate and reliable multiphase flow rate measurement in a three-phase flow. Application of methods with minimized uncertainty is required in the industry. Previous developed correlations for two-phase flow are complex and not capable of three-phase flow. Hence phase behavior identification in different conditions to designing and modeling of three-phase flow is important. Numerous laboratory and theoretical studies have been done to describe the Venturi multiphase flow meter in both horizontal and vertical flow. However, it is not possible to select the measurement devices for all similar conditions. In this study a new venturi model was developed that implemented in Simulink/Matlab for predicting mass flow rate of gas, water and oil. This models is simple and semilinear. Several classified configurations of three phase flow were simulated using Computational Fluid Dynamics (CFD) analysis to get hydrodynamics parameters of the flows to use as inputs of the model. The obtained data, used as test and train data in Least squares support vector machine (LSSVM) algorithm. The pressure drop, mass flow rate of gas, oil and water have been calculated with LSSVM method. Two tuning parameters of LSSVM, namely γ and σ^2, obtained as 1150954 and 0.4384, 53.9199 and 0.18163, 8.8714 and 0.14424, and 10039130.2214 and 0.74742 for pressure drop, mass flow rate of oil, gas mass flow rate, water mass flow rate, respectively. Developed models was found to have an average relative error of 5.81%, 6.31% and 2.58% for gas, oil and water respectively.
%U https://jcarme.sru.ac.ir/article_1046_d58c6904c3ca2d75f5abe6347a30263e.pdf