(6a) Phase Behaviour of Isobutane + CO2 and Isobutane + H2 at Temperatures between 190 and 400 K and at Pressures up to 20 MPa

Mixtures containing isobutane, carbon dioxide (CO₂), and/or hydrogen (H₂) are found in various industrial processes, from refrigeration to alkylation, and in pipelines. Understanding how these mixtures behave, and being able to model this behaviour, is essential for these processes. Isobutane and CO₂ are both considered green refrigerants, with low ozone depletion factors and low global warming potentials, which are increasingly relevant as the world moves towards net zero. The use of pure isobutane or CO₂ in low to moderate temperature refrigeration cycles has become well established (1). Mixtures of these two components have been investigated, with promising balances between flammability, performance, and operating pressure, though issues have been encountered due to limited literature data (2), sparking new interest in vapour liquid equilibrium (VLE) measurements. There is also interest, especially, in new experimental VLE data on hydrogen mixtures, whose behaviour is poorly studied in juxtaposition to its importance during the ongoing energy transition. Isobutane + H₂ illustrates this well, as there has only been one prior investigation into its behaviour or properties (3).

As such, an investigation has been carried out into the phase behaviour of binary mixtures of isobutane with carbon dioxide or hydrogen, extending the range of data available in the literature for both mixtures. The data produced has been used to optimise model parameters for cubic and multiparameter equations of state (EoS). Compressed liquid density measurements have also been carried out for a couple of compositions, primarily to lessen the issues commonly encountered when attempting to predict liquid densities with thermodynamic models.

VLE Method

The VLE measurements were performed using a previously described apparatus (4), employing the static-analytical method (5). A high pressure VLE cell, with a maximum working pressure of 20 MPa, contains the mixture under investigation and is immersed in a thermostatic bath capable of operating at temperatures from 183 to 473 K. Samples of the vapour and liquid may be withdrawn through two ROLSI electromagnetic sampling valves, with capillaries that pass down into the upper and lower regions of the VLE cell. These samples are carried to a gas chromatograph (GC), via a heated and passivated line, for analysis using a thermal conductivity detector. A fluid handling system allows fluids to pass to and from the VLE cell. Some modifications have been made, most notably the addition of an ISCO syringe pump connected to the cell allowing for liquid handling and greater degrees of automation.

GC calibration was performed through the use of gas sampling loops and the absolute area method (4). The uncertainties associated with this method have been discussed previously (6). The absolute area method has been compared to a more complicated analysis which uses an exponentially modified gaussian distribution to describe the detector signal as a direct function of concentration.

Density Method

Density measurements were carried out using a vibrating tube densimeter (VTD) with an operating temperature range of 283 to 473 K and maximum working pressure of 70 MPa. The apparatus has been described previously (7), consisting mainly of the densimeter itself and an ISCO syringe pump, able to reach pressures up to 69.5 MPa. The ISCO may be filled with a homogeneous mixture above the bubble pressure using a variable volume cell (VVC) (8). The mixtures were prepared gravimetrically inside the VVC using a high accuracy balance. Calibrations were performed using nitrogen and water, whose densities can be modelled with low uncertainties.

Modelling

Modelling has been carried out primarily using variants of the volume-translated Peng-Robinson (PR-vdW) EoS with the Soave alpha function (9), using the van der Waals mixing rule and the modified Wong-Sandler mixing rule of Orbey and Sandler coupled with the NRTL free energy model (PR-mWS-NRTL) (10). The binary interaction parameters, expressed as a function of temperature, were regressed from the new experimental data and selected high-quality literature data. The more complex PR-mWS-NRTL model slightly outperformed the classic PR-vdW model, however both were capable of adequately reproducing the data. Some mixture parameters for the multiparameter EoS used by REFPROP were also regressed, as these were previously absent.

Isobutane + CO₂

VLE measurements of the binary system isobutane + CO₂ were carried out at three isotherms (240, 280, and 310 K) at pressures from the lower limit of the sampling system (~0.5 MPa) to the mixture critical pressure. Literature data was available for the 280 and 310 K isotherms (11), allowing for comparisons. There is strong agreement between the VLE data produced in this work and the existing data, for both phases, providing useful validation for the other measurements. The 240 K isotherm is 10 K below any previous studies and aids the modelling of the mixture for low temperature refrigeration cycles.

The mixture was found to display Scott-van Konynenburg (12) type I behaviour, with a continuous critical curve between the critical points of the pure components. This was to be expected, as the pure component critical points are relatively close. Expressed in terms of the reduced temperature (T_r) and reduced pressure (p_r) of isobutane, the critical point of CO₂ is located at T_r = 0.75 and p_r = 2.03.

Isobutane + H₂

VLE measurements of the binary system isobutane + H₂ were carried out at nine isotherms (190, 240, 280, 311, 339, 363, 375, 390, and 400 K) at pressures up to the maximum working pressure of 20 MPa. Three of the isotherms were carried out at, or close to, temperatures examined in the one prior study (3), finding good agreement for the liquid compositions but less so for the vapour compositions. This is not unusual, given the nearly eight decade gap between investigations and the limitations of the older apparatus (13). At low temperatures, the liquid remains rich in isobutane and the vapour consists almost entirely of H₂. At higher temperatures the difference in composition between the two phases narrows, and, when approaching the critical temperature of pure isobutane, the mixture critical pressure enters the operating range of the apparatus, which has allowed for data on the critical curve to be produced.

The mixture was found to display Scott-van Konynenburg type III behaviour, with a discontinuous critical curve emerging from the critical point of the less volatile component, typical of mixtures of significantly different molecules (12). It is expected that the upper critical end point will not manifest before the formation of solid phases occurs, as the melting line for isobutane lies between the critical points of the pure components. Expressed in the same terms as above, the critical point of H₂ is located at T_r = 0.08 and p_r = 0.36.

References

1. Bhatkar V.W., Kriplani V.M., Awari G.K. (2013) Alternative refrigerants in vapour compression refrigeration cycle for sustainable environment: a review of recent research. Int. J. Environ. Sci. Technol., 10, 871-880

2. Sarkar J., Bhattacharyya S. (2009) Assessment of blends of CO2 with butane and isobutane as working fluids for heat pump applications. Int. J. Therm. Sci., 48, 7, 1460-1465

3. Dean M.R., Tooke J.W. (1946) Vapor-Liquid Equilibria in Three Hydrogen-Paraffin Systems. Ind. Eng. Chem., 38, 4, 389-393

4. Fandiño O., Trusler J.P.M., Vega-Maza D. (2015) Phase behaviour of (CO2 + H2) and (CO2 + N2) at temperatures between (218.15 and 303.15) K at pressures up to 15 MPa. Int. J. Greenh. Gas Control, 36, 78-92

5. Dohrn R., Peper S., Fonseca J.M.S. (2010) High-pressure fluid-phase equilibria: Experimental methods and systems investigated (2000–2004). Fluid Phase Equilib., 288, 1-2, 1-54

6. Al-Ghafri S.Z., Forte E., Maitland G.C., Rodriguez-Henríquez J.J., Trusler J.P.M. (2015) Experimental and modeling study of the phase behavior of (methane + CO2 + water) mixtures. J. Phys. Chem., B 118, 14461–14478

7. Tay W.J., Trusler J.P.M. (2018) Density, sound speed and derived thermophysical properties of n-nonane at temperatures between (283.15 and 473.15) K and at pressures up to 390 MPa. J. Chem. Thermodyn., 124, 107-122

8. Tay W.J. (2017) Thermophysical Properties of Hydrocarbons with Dissolved Carbon Dioxide. PhD thesis. Imperial College London

9. Péneloux A., Rauzy E., Fréze R. (1982) A consistent correction for Redlich-Kwong-Soave volumes. Fluid Phase Equilib., 8, 1, 7-23

10. Orbey H., Sandler S.I. (1998) Modeling Vapour-Liquid Equilibria: Cubic Equations of State and Their Mixing Rules. Cambridge Series in Chemical Engineering

11. Nagata Y., Mizutani K., Miyamoto H. (2011) The precise measurement of the (vapour + liquid) equilibrium properties for (CO 2 + isobutane) binary mixtures. J. Chem. Thermodyn., 43, 3, 244-247

12. Rowlinson J.S., Swinton F.L. (1982) Liquids and Liquid Mixtures. 3rd ed. Butterworths Monographs in Chemistry, 191-229

13. Katz D.L., Hachmuth K.H. (1937) Vaporization Equilibrium Constants in a Crude Oil–Natural Gas System. Ind. Eng. Chem., 29, 9, 1072-1077