Hydrate Phase Equilibria of Natural Gas Mixture plus Carbon Dioxide in the Presence of Thermodynamic Inhibitors: Experimental Measurements and Modelling

Global natural gas consumption increases by 1.9% per year and it is still forecasted to grow over the next 20 years. Thus, a challenge for petroleum companies is the exploitation of new types of hydrocarbons reservoirs rich in acid gases. 40% of untapped natural or associated gas fields contain high concentrations of CO2 and/or H2S. Despite dealing with different compositions of natural gas, producing these reservoirs involves also low temperatures combined with high pressures working conditions, presence of water and transport times, leading to the possibility of gas hydrate formations in pipelines. To accurately predict hydrates phase boundaries of such systems and to quantify the amount of thermodynamic inhibitor that has to be injected, reliable thermodynamic models are necessary. Moreover, experimental data are also required to test these models and adjust their parameters. However, to our knowledge, existing date are limited for multicomponent systems containing acid gases in the open literature.
In this work, we report experimental measurement of the locus of incipient hydrate-liquid water-vapour curve for systems containing natural gas spiked with 10%mol and 25%mol of CO2 in the presence of 25%wt and 50%wt of thermodynamic inhibitors (methanol, ethanol and monoethylene glycol or MEG). A total of seven mixtures prepared gravimetrically have been studied.
Dissociation point measurements were conducted using the isochoric step-heating method. Measurements were done in a mixed autoclave rig made of Hastelloy with a volume of about 125 mL. The cell is immersed in a temperature controlled liquid bath. To achieve thermodynamic equilibrium and to mix well the fluids, a stirrer with a magnetic motor is used. The working temperature range is from 203 to 323K and up to 70 MPa. A platinium resistance probe measures the temperature. The accuracy of temperature measurements is ±0.1 K. A Quartzdyne pressure transducer is mounted directly on the cell. The accuracy of pressure measurements is about ±5 kPa. Both pressure and temperature monitors are connected to a computer for a direct acquisition.
Different observations can be made from these experiences:
• The presence of CO2 in a natural gas mixture slightly displaces the hydrate dissociation curve (about 1 K less)
• Whatever the concentration, methanol has the highest inhibiting power among the three inhibitors. Indeed, adding 25%wt of methanol decreases the hydrate formation temperature by 12 K, while adding the same amount of ethanol of MEG decreases the temperature by 6 K.
• Ethanol and MEG have a similar inhibiting power when 25%wt are added. However, when increasing the concentration to 50%wt, MEG is a better inhibitor than ethanol.
These data are of a great importance to better understand hydrate phase boundaries of multicomponent mixtures with CO2 and allow also validating thermodynamic models, like the GC-PR-CPA equation of state (EoS) used in this work. The model results from the sum of two terms: a cubic equation of state (the Peng-Robinson EoS) and an associative term to take into account hydrogen bonding. It is thus adapted for systems with water. Moreover, the GC-PR-CPA is a group contribution method, in which parameters are adjusted for different groups. As for the hydrate phase modelling, the GC-PR-CPA has been coupled with Van der Waals and Platteuw (1959) model, as implemented by Parrish and Prausnitz (1972). The model is in good agreement with experimental data.