Prediction of H2S solubility in aqueous NaCl solutions by molecular simulation

Abstract

The solubility of hydrogen sulfide in aqueous NaCl solutions has been investigated by calculation of Henry constants through Monte Carlo simulations and using a nonpolarizable force field. The Lorentz-Berthelot combining rules appeared the most appropriate to simulate this system compared to other usual rules. The physical behavior and the trends experimentally observed are qualitatively well reproduced by this purely predictive approach for the binary system H 2 S+H 2 O as well as for the salted system H 2 S+H 2 O+NaCl: as experimentally expected, the simulations indicate a maximum in the Henry constant curve with temperature and an increase of the Henry constant with the salt concentration (salting-out effect). From a quantitative point of view, Henry constants are systematically overestimated, and more specifically for high temperatures. By introducing an empirical temperature-independent binary interaction coefficient between water and hydrogen sulfide, it is possible to obtain a good agreement between experiments and calculated Henry constants for both salt-free and salted systems, over a large temperature range (290 to 590 K) and salinity ranging from 0 to 6 mol/kg. For temperatures less than 400 K, the deviations are smaller and within the experimental uncertainties. At higher temperatures, deviations are higher but only few experimental data are available to confirm this result. Considering statistical and experimental uncertainties, this approach can thus provide a reasonable estimation of H 2 S solubility in NaCl aqueous solution in conditions encountered in geological formations. For a given temperature, the variation of the Henry constant with salt concentration is not exactly the same as the variation experimentally observed, and the salting-out effect is overestimated. This suggests for future works that additional forces such as polarization have to be taken into account for modeling high salt concentration systems.

Predicting the phase behavior of hydrogen in NaCl brines by molecular simulation for geological applications

“…Thus, Widom insertion method could appear in this context less efficient than more advanced techniques to evaluate chemical potential, such as umbrella sampling (Torrie and Valleau, 1977), slow-growth method (Nezbeda and Kolafa, 1991) or thermodynamic integration (Frenkel and Smit, 1996). Nevertheless, a prior study on the solubility of H 2 S in NaCl aqueous solution (Fauve et al, 2017) shown that, with a sufficient number of Monte Carlo steps, the Widom insertion method lead to similar results than thermodynamic integration technique within statistical uncertainties.…”

“…Previous studies have demonstrated the ability of Monte Carlo simulation to predict gas solubility in brines, such as CO 2 (Creton et al, 2018;Jiang et al, 2017;Liu et al, 2013;Tsai et al, 2016;Vorholz et al, 2004), H 2 S (Fauve et al, 2017), SO 2 and other diatomic light gases (Creton et al, 2018). However, to the best of our knowledge, no molecular simulation study dealing with the solubility of hydrogen in electrolyte solutions has been published.…”

Modeling of mixed-solvent electrolyte systems

“…Many molecular simulation studies exhibited this structure-forming phenomenon. As an example, figure 2 shows the radial distribution functions (rdf ) in an aqueous NaCl solution at 1 mol/kg and 298.15K between the pairs Na + /O and Cl -/O (where O is the oxygen atom of the water molecules) obtained from Monte Carlo simulations [100]. Using an integral oof the rdf, this figure also shows the number of water molecules surrounding each ion, also called the coordination numbers (CN).…”

“…Radial distribution function (rdf ) and coordination number (CN)between Na + /O pairs and Cl-/O pairs obtained from Monte Carlo simulations[100] …”

Molecular simulation of equal density temperature in CCS under geological sequestration conditions

“…NaCl is the most common ion in groundwater brine, and different ion models can have a great influence on the density of solution. Fauve et al 40 used OPLS 41 + SPC/E models to simulate the density of NaCl saline solutions with different concentrations at a wide range of temperatures (from 25 to 275°C), and the average deviation between the calculation and the experimental data was 1%. However, the performance of the SPC/E model is not as good as that of the TIP4P/2005 model under reservoir conditions.…”