A computational tool for EoS parameter estimation : evaluation of new methodologies and application in a natural gas processing simulation

Abstract

Brazilian pre-salt reservoirs represent the discovery of large amounts of light oil, but at high pressures and containing high levels of CO2, being a challenge in its thermodynamic modelling by conventional equations of state (EoS). In order to perform more accurate predictions, it is necessary to execute the procedure of parameter estimation. Therefore, in this work a computational tool was developed that reliably carries out this parameter estimation with interactive and user-friendly interface, allowing the creation of charts containing parametric and objective function evaluations in each point of the optimization. New methodologies for parameter estimation were systematized utilizing this tool: the first one by applying simultaneously vapour-liquid and liquid-liquid equilibria and the other handling water content in dew point condition. Utilizing the EoS Cubic-Plus-Association (CPA), these methodologies were properly validated with literature data. New CPA parameters for water, which take into account the experimental value of its hydrogen bonds energy, and for its mixture with CO2 were obtained, whose average absolute deviation in the water content at pressures higher than 200 bar fell from 23.7% to 3.3% compared to previous publications. Finally, this study has been applied into a natural gas process unit, modelled with the parameters found here.