Prediction of the hydrate forming condition of natural gases in the presence of mono-ethylene glycol

An inherent problem with natural gas production or transmission is the formation of gas hydrates, which can lead to safety hazards to production and transportation systems and to substantial economic risks. Therefore, an understanding of the inception of hydrate formation is necessary to overcome hydrate issues. The aim of the first step of this study is to develop a methodology for predicting hydrate forming condition of sweet natural gases in the presence of mono-ethylene glycol. In this work, a novel simple-to-use correlation, which is easier than current available models involving huge number of parameters, requiring more complicated and longer computations. This generalised approach enables the estimation of hydrate formation pressure of sweet natural gases for pressure ranging 1200 to 40,000 kPa and temperatures between 260 K and 298 K as well as molecular weights in the range of 16 to 29. New proposed correlation shows consistently accurate results across proposed pressure, temperature and molecular weight ranges. This consistency could not be matched by either of the widely accepted existing correlations within the investigated range. For all conditions, new correlation showed average absolute deviation to be less than 0.2% and provided much better results than the widely accepted existing correlations. In the next step, novel empirical correlations are developed to predict the required mono-ethylene glycol (MEG) weight percent in the rich solution of monoethylene glycol and the flow-rate for desired depression of the gas hydrate formation temperature. These correlations are generated for a natural gas with relative density of 0.6 at pressures of 3, 5, 7, and 9 MPa, which are applicable to wet gas temperatures of 20, 30, 40, and 50 °C. In order to extend the application of these correlations to wide ranges of natural gas mixtures with specific gravities of up to 0.8, two generalized correction factors are also provided. The accuracy of this simple method is compared with the simulation results obtained by commercial software which showed good agreement