IMPROVEMENTS IN CLATHRATE MODELLING: I. THE H2O-CO2 SYSTEM WITH VARIOUS SALTS

Abstract

The formation of clathrates in fluid inclusions during microthermometric measurements is typical for most natural fluid systems which include a mixture of H2O, gases, and electrolytes. A general model is proposed which gives a complete description of the CO2 clathrate stability field between 253–293 K and 0–200 MPa, and which can be applied to NaCl, KCl, and CaCl2 bearing systems. The basic concept of the model is the equality of the chemical potential of H2O in coexisting phases, after classical clathrate modelling. None of the original clathrate models had used a complete set of the most accurate values for the many parameters involved. The lack of well-defined standard conditions and of a thorough error analysis resulted in inaccurate estimation of clathrate stability conditions. According to our modifications which include the use of the most accurate parameters available, the semi-empirical model for the binary H2O-CO2 system is improved by the estimation of numerically optimised Kihara parameters σ = 365.9 pm and ɛ/k = 174.44 K at low pressures, and σ = 363.92 pm and e/k = 174.46 K at high pressures. Including the error indications of individual parameters involved in clathrate modelling, a range of 365.08–366.52 pm and 171.3–177.8 K allows a 2% accuracy in the modelled CO2 clathrate formation pressure at selected temperatures below Q2 conditions. A combination of the osmotic coefficient for binary salt-H2O systems and Henry's constant for gas-H2O systems is sufficiently accurate to estimate the activity of H2O in aqueous solutions and the stability conditions of clathrate in electrolyte-bearing systems. The available data on salt-bearing systems is inconsistent, but our improved clathrate stability model is able to reproduce average values. The proposed modifications in clathrate modelling can be used to perform more accurate estimations of bulk density and composition of individual fluid inclusions from clathrate melting temperatures. Our model is included in several computer programs which can be applied to fluid inclusion studies.