STATISTICAL EVALUATION OF VOID FRACTION CORRELATIONS FOR THE NUMERICAL MODELING OF TWO-PHASE FLOW IN GEOTHERMAL WELLS

Predicting flowing pressure and temperature profiles in geothermal wells is a fundamental task to study the in flow production mechanisms. The gas void fraction is one of the most important parameters required for the better prediction of production profiles. Eight empirical correlations (Bonnecaze, Dix model, Duns-Ros, Krilov, Hasan, Rouhani, Homogeneous Model and Orkiszewski) for the estimation of gas void fractions and to model their implications on the two-phase flow inside geothermal wells were evaluated. These correlations were assessed through the two-phase flow modeling (using the wellbore simulators GEOPOZO and GEOWELLS) in four producer wells from Mexican geothermal fields: Los Azufres, Mich. (Az-18), Los Humeros, Pue. (H-1), and Cerro Prieto, B.C. (M-90 and M-201). The simulated pressure and temperature profiles were statistically compared with actual field data. An acceptable agreement (< 10%) between the simulated profiles, obtained with the Dix model correlation, and measured data was obtained. These results enabled the modeling of two-phase flow inside geothermal wells to be reliably performed, which constitute an advantage due to the limited number of available correlations to calculate the gas void fraction in geothermal wells.