Modeling of stability conditions of natural gas clathrate hydrates using least squares support vector machine approach

Abstract

This paper concerns a novel computer-based approach namely Least Square version of Support Vector Machine (LSSVM) algorithm to estimate/represent hydrate forming/dissociating conditions of various gases in the presence of pure water or aqueous solutions of salt(s) and/or alcohol(s). To this end, several models have been presented to predict the hydrate dissociating temperature (HDT) of C1, C2, C3, i-C4, CO2, H2S, N2, and natural gas mixtures in pure water or additive containing solutions. For modeling purpose, an extensive database comprising more than 3900 experimental data have been gathered from the literate from 1940 to 2013. The collected databank covers wide range of experimental conditions at solid/liquid/vapor or solid/ice/vapor equilibrium of distinct hydrate systems. All the proposed models reproduce the targets with R2 of greater than 0.97. The predictions of the developed LSSVM models for hydrate systems of C1, C2, C3, i-C4, CO2, H2S, N2, and natural gas mixtures are in good agreement with corresponding experimental data with the average absolute relative deviation percent (%AARD) equal to %0.22, %0.33, %0.26, %0.07, %0.47, %0.56, %0.08, and %0.34, respectively.