The Vapour-Vapour Interface Observation and Appraisement of a Gas-Condensate/Supercritical CO2 System

Injecting supercritical CO2 into gas reservoir is a novel trial to improve condensate gas recovery and decrease the hydrocarbon liquid dropout. A good understanding of the effect of supercritical CO2 on the phase behavior properties of these hydrocarbons is essential for accurately forecasting the displacing performance and storing process of the reservoirs with numerical simulators. This paper presents novel phase behavior experimental procedures and phase equilibrium evaluation methodology for gas-condensate phase system mixed with supercritical CO2 over a wide range of temperatures and pressures. A unique phase behavior phenomena was also reported. The mass transfer between two vapour phases was also measured. In order to interpret and identify the interface property between condensate gas and supercritical CO2, a multiphase thermodynamic VLV equilibrium model was established. Finally, taken YKL condensate gas in Northwest China as an example, the region where the conditions in terms of pressure, temperature and CO2 concentration can yield VLV equilibrium was found. The calculation results of multiphase thermodynamic model for condensate-CO2 system in this paper are close to the experimental data and can truthfully reflect the phase behavior of interface between CO2 and condensate gas. The research results indicate that it is the existence of the interface between CO2 and condensate gas that makes CO2 possible be an attractive option to successfully displace condensate gas and decrease CO2 emissions.

CO 2 Enhanced Gas Recovery (EGR) has been studied in laboratory settings and in field-scale simulations for both tight and conventional gas reservoirs, but application in the field remains limited to a handful of cases [13][14][15][16] .
Excessive mixing is a key risk associated with CO 2 EGR as it would result in undesirable contamination of the natural gas asset and/or early breakthrough of the injected CO 2 at the production wells. This risk of excessive mixing of CO 2 and natural gas within the reservoir has limited the practice of EGR. At present, rare literature reported the mix phenomena between CO 2 and natural gas under reservoir conditions and clearly revealed under what kind of conditions natural gas is not easy to be mixed by CO 2 . Therefore, finding suitable gas injection conditions is the key to the promotion of EGR in suitable gas fields.
Although, CO 2 flooding has been discussed in dry gas reservoirs, still, it has complex challenges in condensate reservoirs. One challenge is how to enhance condensate recovery; Another challenge is how to enhance condensate gas recovery. Previous research related to enhancing recovery of condensate gas field by injecting CO 2 has until now focused on boosting condensate recovery. For this reason, the studied fluid of most reservoirs has been high-condensate content. Plenty of studies have been conducted to discuss the physical and chemical processes when CO 2 contact with condensate liquid. These experimental data revealed the possibility of improving condensate by CO 2 flooding and can be utilized to validate the numerical studies in modeling tools. For example, JJ Chaback et al. 17 , Ahmed et al. 18 , E. Shtepani 19 , and RM Ramharack 20 observed the density and composition of CO 2 in the process of retrograde evaporation of condensate oil through PVT experiment, and approved that CO 2 can inhibit condensate precipitation and move condensate . Their studies indicated that the dew point drops, the liquid-volume percent (liquid saturation) decreases, the compressibility factor falls and the two-phase envelope diagram shrinks with the increase of CO 2 composition. All these trends showed the positive impact of CO 2 have on liquid recovery. Millán A E et al. 21 and Al-Abri et al. 22,23 quantitatively studied the condensate recovery and relative permeability with long core experiment and numerical simulation. Research showed that injecting SC-CO 2 can decrease capillary instability, improve mobility ratio, delay the breakthrough time of gas injection and improve the condensate sweep efficiency. Heron GM et al. 24 and Taheri A et al. 25 carried out experimental study of injecting CO 2 , N 2 and dry gas into fractured gas condensate reservoir. They came to conclusions opposed to each other.
However, there has been little investigation on how CO 2 mix with condensate gas and whether CO 2 can improve the recovery of condensate gas, despite these numerous studies on mitigating condensate precipitation and enhancing oil recovery by CO 2 flooding. In fact, condensate gas reservoirs produce both condensate gas and condensate, and both have high profits. In order to elucidate the possibility of CO 2 EGR in condensate gas reserovirs, a series of experiments and model were used to analyze mixing procedure when CO 2 contact with condensate gas. The focus of this research is to demonstrate under what kind of conditions in terms of pressure, temperature and CO 2 concentration, CO 2 and condensate gas are not easily mixed into one phase.

Experiment
Experiment materials. The condensate gas from Well YK1, located at north of China is composed of 87.368% methane, 3.570% ethane and 0.718% propane etc. CO 2 has purity up to 99.95%. The physical properties of condensate gas are listed in Table 2. The experimental conditions are that in the YKL gas reservoir (at 43.5 MPa and 132.18 °C).

Experimental apparatus.
To observe and calibrate how CO 2 mix with condensate gas, we set up a non-equilibrium experiment with high-pressure and high-temperature laboratory. A schematic of the experimental setup is presented in Fig. 1 the injected sample was monitored by a cathetometer. Incandescent lamp, which was filled with hydrogen and N 2 , was placed at the opposite side of the sapphire cell to observe the change of the transmittance of the system. Fluid phase change in PVT cell was recorded via the computer video acquisition camera.
Before the experiments, the sapphire cell was dismounted from the apparatus, washed with distilled water and dried, and then installed back into the apparatus. Subsequently, the cell was evacuated to ensure the absence of air. The desired temperature was set through the temperature chamber. After that, the right amount of the sample that was equilibrated without condensate was prepared to be injected into the sapphire cell at desired conditions of pressure and temperature.
Non-equilibrium phase measurement of CO 2 mixed with condensate gas. Usually, it is difficult to survey the mixing process of CO 2 and condensate gas in experiments through visual observation, because CO 2 and condensate gas are both transparent. Inspired from the critical opalescence of CO 2 , critical opalescence is used to distinguish CO 2 from condensate gas in these experiments. Critical opalescence is a striking light scattering phenomenon, which was elegantly explained by Einstein. In the critical region the light scattering is so large that the substance appears milky white in reflected light and brownish dark in transmitted light. The phenomenon arises from the large fluctuations in the critical region of the substance 26 . CO 2 critical phenomenon around the critical temperature and pressure was reported in many literatures [27][28][29] . The critical temperature is 31 °C and critical pressure is 7.53 Mpa. However, the pressure and temperature of reservoir conditions are much higher than the critical pressure and critical temperature of CO 2 . No reports were mentioned about opalescence or density  fluctuation of CO 2 existed in high temperature and high pressure, which is far away from critical pressure and temperature.
The non-equilibrium phase measurements are aimed at finding the existence conditions of opalescence of CO 2 over a wide range of temperatures and pressure, and characterizing the mix characteristic of CO 2 and condensate gas at specific pressures and temperatures.
In the experiment, starting from a high pressure, it is gradually depressurized in the sapphire cell to observe physical changes through the glass window into the cell. Three different tests conducted: (1) Pressure and temperature search to find the opalescence of CO 2 existence conditions; (2) Characterization of CO 2 mixing with condensate gas; (3) Diffusion test of the system of condensate gas and SC-CO 2 with high temperature.
Pressure and temperature search to find the opalescence of CO 2 existence conditions. When the temperature was kept constant, the pressure search method was used to determine the CO 2 opalescence existence conditions. The first step was to pump CO 2 into sapphire cell at reservoir temperature (132.18 °C) and pressure (43.5 MPa). Then, pressure reduced from 43.5 MPa to 18 MPa, and CO 2 status in sapphire cell was recorded every 3 minutes. To confirm the accuracy of the onset and end of the pressure of appearance and disappearance of opalescence CO 2 , three repetitive tests by increasing and decreasing the pressure around the appearance and disappearance of opalescence CO 2 were preformed. Similar to the pressure search method, the temperature search method was used.
Characterization of CO 2 mixing with condensate gas. Based on the pressure and temperature range at which CO 2 displayed opalescence phase behavior, the pressure and temperature used to test the mixing of CO 2 and condensate gas were selected. Then at specific temperature and pressure, injecting condensate gas from the top of sapphire cell steadily, took photos of the SC-CO 2 -condensate gas every five seconds from the visualization window.
Diffusion test of the system of condensate gas and SC-CO 2 with high temperature. One group of temperature and pressure where CO 2 appear opalescence phenomena were chosen. After the system standing for 30 minute, gas was discharge from sapphire cell with no pressure fluctuation. And their composition of different parts was measured with cathetometer.

Results and Discussion
Experiment Results. Pressure and temperature search to find the opalescence of CO 2 existence conditions.
Obvious phase changes of CO 2 was recorded when pressure drops from 25 MPa to18MPa at 132.18 °C (Fig. 2). Figure 2(1) and (2) shows that under the temperature and pressure of the gas reservoir (132.18 °C 43.5Mpa), the initial state of CO 2 behaves like fluid with high density and stable physical properties. It is a kind of "liquid-like" supercritical fluid with the same light transmittance in the whole system. When the pressure drops to 25 MPa, CO 2 opalescence appears in reservoir temperature for the first time (Fig. 2(3)). When the pressure drops continually, CO 2 continues presenting opalescence phenomena in sapphire cell until 18Mpa (from Fig. 2(3) to (9)). When the pressure reaches to 18 MPa, opalescence phenomenon disappears and light transmittance increases significantly (Fig. 2(10)).
Similar phase changes of CO 2 was recorded when temperature drops from 132 °C to 8 °C at 30 MPa (Fig. 3).
Distribution characterization of CO 2 mixed with condensate gas. The pressures to test the mixing of CO 2 and condensate gas were selected based on above pressure search experiments. 25Mpa was chosen as experiment pressure for mixing CO 2 and condensate gas. Keep the temperature of this system in 132.18 °C. Injecting condensate  gas from the top of sapphire cell containing CO 2 steadily, photographs of SC-CO 2 -condensate gas from the visualization window were recorded very five seconds. The unequilirium experiment of CO 2 and condensate gas snapshots revealed that obvious phase behavior change between CO 2 and condensate gas at 25Mpa and 132.18 °C (Fig. 4). First the pure CO 2 in the PVT cell, with weak background light source, was chosen as the basic condition before injecting condensate gas ( Fig. 4(1)). At this point, CO 2 stays in opalescence state with high density. After starting to inject condensate gas from the top of PVT cell, condensate gas looks like bright light appears at the top of CO 2 , which shows injected condensate gas has good transmittance. There is no obvious difference of the phase and physical characteristic between condensate gas and supercritical CO 2 (Fig. 4(2 and 3)). After keeping injecting condensate gas from the top of PVT cell, there is convection diffusion and mass transfer between condensate gas and CO 2 . As condensate gas was injected from the top continually, the layer between two fluids with dark color appears. Similar to the phenomena near critical region reported in literature 27 , even temperature and pressure far away from critical point, these domains are dynamic and fluctuate strongly when the layer is between two fluids; we capture them immobilized during the transfer process (Fig. 4(4 and 5)). After keeping injecting condensate gas from the top of PVT cell, the layer between CO 2 and condensate gas becomes clear (Fig. 4(6~10)). Continuing injecting condensate gas from the top of sapphire cell, the layer between CO 2 and condensate gas shrinks and then contracts to a bright yellow opaque. There exists transparent uniform condensate gas on the top and transparent supercritical CO 2 phase at the bottom (Fig. 4(11 and 12)).
Similar phenomena were also recorded at 22 MPa and 20 MPa. The temperatures to test the mixing of CO 2 and condensate gas were selected based on above temperature search experiments. Similar phenomena about mixing of CO 2 and condensate gas were also recorded at 100 °C, 50 °C and 30 °C.
The experimental data with the onset and end pressure of layer appearance with different temperature, pressure and CO 2 mole percentage were listed in Tables 3 and 4. When temperature drops from 132 °C to 30 °C, there is more possibility of the layer existed between CO 2 and condensate gas. From 132.18 °C to 30 °C, there is a gradual declined reduction in the end pressure of layer, reaching a figure of 7.5 Mpa.   Diffusion test of the system of condensate gas and SC-CO 2 with high temperature. To perform a detailed assessment of the spatial distribution of CO 2 and condensate gas and discuss interface property, the diffusion test was carried out with the composition analysis of 10 equal division subdomains based on pressure search experiment. Fig. 5 and Table 5 shows the composition of different subdomains in Sapphire cell containing condensate gas and CO 2 when pressure is 25 MPa and temperature is 132.18 °C, after mix experiment was stood still for 30 minutes.
The gas was discharged and tested with cathetometer. Composition shows that domain 1-3 is primarily composed of condensate gas. Domain 4 to 6 see a dramatic rise of CO 2 content rise and changed from 3.519 to 86%; a considerable decrease of the C 2~C6 content occurs from 5.39% to 1.95%; as for the domain 6 to 9, the content of     compositions remain still. Thus, compared with the composition of different domains, the interface between condensate gas and CO 2 locate at domain 4-5, which means that the composition of interface may be a compound of CO 2 and condensate gas. Based on the definition 30 , the layer between condensate gas and CO 2 is also called interface. Aforementioned conclusion needs further investigation.
Interface property analysis. Interface property at 132.18 °C. The onset pressure of layer/interface appearance: In the experiment, when the temperature is 132.18 °C, the onset pressure of interface appearance is 25 MPa when 50 mol% of CO 2 mixes with of 50 mol% condensate gas. The PR equation of state, which was fitted to experiment data, was used for two-phase flashing calculation under different experimental conditions. The calculation results are shown in Table 6. From the results, it can be seen that when the pressure is 25 MPa, the system with 50 mol% of condensate gas and 50 mol% of CO 2 begin be condensed, which corresponds to the onset pressure of the interface. Therefore, it is believed that there may be an interface when condensate is condensed in CO 2 -condensate gas system. The end pressure of layer/interface: However, with the decrease of pressure, the condensate content in the interface rise and the interfacial tension between CO 2 and condensate grows, and phase separation occurs. When the condensate in the interface is precipitated continuously, the interface disappears gradually due to the influence of gravity differentiation.
In order to prove above idea, numerical simulation of slimtube in CO 2 drive condensate under different pressure was carried out to analyze the interfacial tension between CO 2 content and condensate.
Analysis of interfacial tension in rich CO 2 condensate. The CO 2 mole percentage and interfacial tension were plotted for driving simulation of condensate. Dimensionless distance 0.375 is the CO 2 displacement front of in the displacement process of condensate, shown in Fig. 6. At the displacement front of CO 2 , there is low interfacial tension between CO 2 and condensate, which is about 0.6 dyn/cm, when the pressure is 25 MPa. With the pressure decreasing, the interfacial tension between CO 2 and condensate increases gradually (Fig. 7). Therefore, when the temperature is 132.18 °C and the pressure is 25 MPa, CO 2 can mix with condensate and forms interface. But with the decrease of pressure, condensate in the interface is separated, and the interface begins to be unstable.  Density analysis. The density of interface was calculated based on molar percentage of the condensate and CO 2 . At 132 °C, the density of supercritical CO 2 , condensate gas and interface were plotted in the Fig. 8. When the pressure is 24.97 MPa, there is a stable density difference among three phases. With the pressure decreasing, the density of interface is closer to the density of supercritical CO 2 . When the pressure is below 10 MPa, the densities of interface and condensate gas are exactly the same. Thus, in order to form a stable interface, it is considered that there is a certain density difference between the three phases.
Modeling Results. Obtaining this experiment conditions including pressure, temperature and CO 2 concentration which can yield interface phenomena is a very labor intensive and time consuming process, especially when the components differ greatly from different gas fields. A possible solution to this problem is to develop a new approach to describe experimental data that would allow construction phase transition interface out of   a limited data set. Such approach will make it possible to obtain not only reliable information on properties of mixtures, but also data on composition. Based on interface property analysis, novel mathematical model to model the interface phenomena during CO 2 mixing with condensate gas was established, in order to find the region where the conditions in terms of pressure, temperature and CO 2 concentration can yield VLV equilibrium containing interface, which is presented in Appendices.
Reliability of VLV calculation. The accuracy of the calculations was evaluated by the absolute average relative deviation (AARD) defined as follows, As shown in Fig. 9, experimental pressures at which interface appear and disappear with different CO 2 mole percentage at 30 °C, 50 °C, 100 °C and 132 °C were in well agreement with that calculated by the VLV model with AARD = 5.41%, indicating that the VLV model in present work was reliable for interface envelop calculation.
PTX phase diagram of interface. The PTX phase diagram of interface was drawn in Fig. 10. Figure 10 shows that the interface envelops looked like J column or "stomach". At the temperatures above CO 2 critical temperature (31 °C), interface occurs when CO 2 contact with condensate gas (see Fig. 11). With the temperature rising, the up pressure and low pressure envelope of the interface increase gradually. Figure 12 illustrates that when CO 2 , the content of which is between 10% and 82%, mix with condensate gas, the system shows interface characteristics in a certain pressure range; when the content of CO 2 is between 40 and 50 mol%, a wide pressure range of interface appears in the system.

Conclusion
The following conclusions can be drawn from the current study: • Measures show opalescence or density fluctuation of CO 2 in high temperature and high pressure. Opalescence phenomena of CO 2 in high temperature and high pressure was used to observe the mix process of CO 2 and condensate gas, which shows there is interface when CO 2 mix with condensate gas at high (reservoir) temperature. • Thermodynamic characteristic analyses indicate the onset pressure of interface appearance between CO 2 and condensate gas is around dew pressure in the CO 2 -condensate system. With the pressure dropping, the condensate in the interface is precipitated continuously. The end pressure of interface appearance is related to gravity differentiation because of increase of the interfacial tension and phase separation between CO 2 and condensate. • The multiphase thermodynamic model for condensate-CO 2 system in this paper can truthfully reflect the interface behavior during CO 2 injecting condensate gas over a wide range of temperatures and pressures, and calculation results of which are close to the experimental data. • Our study provides the possibility to predict the enhanced condensate gas recovery with CO 2 injection with novel trial.