This study analyzed the partial effect of carbon dioxide hydrate in reaction kettle experiments. The particle and bubble characteristics of the crystal nucleus during carbon dioxide hydrate decomposition were observed under the microscope. The results showed that in the temperature range of 0.5°C–3.5°C, the pressure range of 3 MPa–5.5 MPa, phase characteristics in the reaction kettle changed in a complex fashion during carbon dioxide hydrate formation. During hydrate decomposition, numerous carbon dioxide bubbles were produced, mainly by precipitation at high temperatures or in the hydrate cage structure. The hydrate crystal nucleus initially exhibited fluidity in the reaction. However, as the reaction progressed, the hydrate crystal nucleus migrated upward under the influence of gravity and carbon dioxide diffused into the aqueous phase. Next, the hydrate was formed and accumulated, finally forming a solid carbon dioxide hydrate layer.
As a greenhouse gas, carbon dioxide (CO2) can be buried through injection into geological formations. Currently, several oilfields worldwide have adopted carbon dioxide displacement for improving oil recovery. However, CO2 hydrate is produced in the wellbore during injection. Hydrate is a clathrate envelope, water molecules form a lattice structure in hydrates, the gas molecules fill the holes between the lattice (
Toschev et al. [
Sloan et al. [
Some studies have focused on the hydrate crystal nucleus and nucleation [
CO2 (purity 99.99%, provided by Daqing Xuelong Qigong Co., Ltd.) and distilled water (obtained by secondary distillation in the laboratory).
The CO2 hydrate formation experiment was conducted using a visual reaction kettle. A schematic diagram of the experimental device is shown in
Freon in the storage tank (
The experimental device was provided by the China Nantong Huaxing Petroleum Instrument Factory. The height of the reactor was 0.5 m and the temperature error in the reactor during multiple tests was ±0.1°C. The temperature rise rate in the reactor was approximately 3–5 min, whereas the temperature decrease rate was approximately 5–10 min. However, since the temperature control system adopts a constant-temperature water bath cycle, the water bath temperature can be set in advance at the beginning of the experiment. After the temperature was stabilized, the cycle was turned on. This method can allow the temperature in the reactor to reach the target temperature rapidly at the beginning of the experiment (approximately 1 min). Since the reactant is carbon dioxide, we chose a nickel-chromium alloy as the material for the entire reactor because it has good corrosion resistance and thermal stability. The formation state of hydrate in the pipeline can be observed through an observation window at the upper part of the reactor. Since the hydrate formation reaction requires low temperature and high pressure, sapphire glass was selected as the material for the observation window, with a diameter of 50 mm. During the formation of hydrate, the temperature and pressure in the reactor will change; therefore, we placed temperature and pressure sensors into the reactor and measured the change in temperature and pressure by applying an external electrical signal (
Considering that hydrate is gradually accumulated during the freezing process of well bore, the borehole wall maintains a low temperature during the freezing process, and there is a cooling time at the beginning of the experiment, it is unreasonable to characterize the induction time of hydrate formation by monitoring the temperature and pressure. In addition, only the hydrate accumulated on the shaft wall will have an effect on the freeze plugging in the well bore. Therefore, the induction time of hydrate formation was measured through observation window. It was believed that the timing started upon the formation of hydrate in the observation window; this time was the initial time of hydrate formation induction time. If the hydrate is completely filled in the observation window, it is considered that freeze plugging has occurred, and at this moment, the reaction is terminated.
First, distilled water was added into the reaction kettle until the liquid level reached halfway on the observation window. Next, the cylinder was opened to allow CO2 into the rector. The temperature and pressure monitoring system was then switched on to observe temperature and pressure changes in the reactor. The reaction temperature was set using the temperature control system and the reaction pressure was set using the gas cylinder. When the temperature in the storage tank reached the experimental temperature, coolant circulation was initiated and the experiment began. Hydrate formation in the reaction kettle was observed. The time at which hydrate started forming in the reactor was set as the reaction start time, and the time at which hydrate filled the reactor was used as the reaction end time. The characteristics of hydrate formation and changes in temperature and pressure in the reaction kettle were observed. The experimental Steps (1)–(5) were repeated several times to confirm the accuracy of the experimental results.
Since it is difficult for the nonpolar gas molecules and polar water molecules to form stable crystal state, we placed a part of the solid crystal hydrate in the initial structure as crystal nucleus. The energy of the initial structure is minimized by running a short NVT simulation of approximately 400 ps for the relaxation of solid–liquid interface. After relaxation, the entire system was simulated using NPT ensemble, and the trajectory file was obtained. Velocity Verlet algorithm calculated the motion equation of the atoms, and the time step is 1 fs. Water molecules can adopt rigid models, such as SPC or TIP4P, to constrain their bond angle length. Methane is described by the EPM2 model. The potential energy parameters of the interactions between different types of atoms were obtained by the standard Lorentz–Berthelot mixing rule. The temperature and pressure were controlled at 275 K and 10 MPa, respectively. The truncation radius of the short-range interaction was 12 Å, and the long-range Coulomb force adopted the PPPM algorithm.
In this study, the induction time of carbon dioxide hydrate formation was studied based on the hydrate freezing and plugging process in the well bore. So in this study, the time at which hydrate was observed through the observation window was set as the start time, while the time at which hydrate in the reaction kettle complete generated was set as the end time. The hydrate formation process observed at 2.5°C is shown in
As shown in
The experimental pressure was set at 4 MPa. Initial temperatures of 0.5°C, 1.5°C, 2.5°C, and 3.5°C were tested, and temperature changes in the reaction kettle were measured over time, as shown in
As shown in the temperature change curve (
Experiments were conducted at 2°C and initial pressures of 3 MPa, 3.5 MPa, 5 MPa and 5.5 MPa, and the pressure varied over time, as shown in
The pressure initially dropped when the hydrate started to form. An earlier drop point indicated earlier hydrate formation, showing that hydrate formation was easier. At the same temperature, a higher pressure resulted in easier hydrate formation. When the initial pressure was 3 MPa, the pressure in the reactor initially increased. This was attributed to the initial reaction being relatively violent, which increased the reactor temperature, making dissolved gaseous CO2 more likely to escape from the water under low pressure conditions, which caused the reactor pressure to increase.
As shown in
The induction times for CO2 hydrate formation at different temperatures and pressures are shown in
Number | Temperature(°C) | Pressure |
Inductiontime (min) | Number | Temperature(°C) | Pressure |
Inductiontime (min) |
---|---|---|---|---|---|---|---|
1 | 0.5 | 4 | 27 | 5 | 2 | 3 | 40 |
2 | 1.5 | 20 | 6 | 3.5 | 30 | ||
3 | 2.5 | 21 | 7 | 5 | 18.5 | ||
4 | 3.5 | 23 | 8 | 5.5 | 16 |
Under low temperature and high pressure conditions, the induction time for hydrate formation was relatively short. This was attributed to a lower set temperature causing a greater degree of subcooling and greater driving force, resulting in a more complete reaction. Under normal circumstances, the induction time for hydrate formation in the reactor was about 15–40 min. The formation states of CO2 hydrate at different temperatures are compared, as shown in
The liquid carbon dioxide layer is very thin at 0.5°C and 4 MPa; however, this does not suggest that the amount of carbon dioxide is less. In addition to the liquid carbon dioxide, the water phase will still dissolve a large amount of carbon dioxide, and this carbon dioxide will be added to the reactants. Therefore, the temperature, pressure, and phase characteristics in the reactor will change in a complex manner during the formation of CO2 hydrate.
To observe the CO2 hydrate decomposition process and microscopic bubble characteristics inside the reactor, a vertical hydrate reactor was placed horizontally and the microscopic characteristics of CO2 hydrate decomposition were observed via microscope, as shown in
Hydrate was rapidly decomposed from the outside to inside. During the decomposition process, the apparent interface between the hydrate and liquid mixture was observed, accompanied by numerous CO2 bubbles. This was mainly due to the increased reactor temperature during the decomposition process, which caused decreased CO2 solubility, the precipitation of many CO2 bubbles, and the production of CO2 in the hydrate cage structure during decomposition, resulting in the formation of many bubbles. The CO2 hydrate decomposition process was observed using a microscope, as shown in
As the CO2 hydrate decomposed, the apparent interface between the hydrate and liquid mixture was observed. Furthermore, the carbon dioxide hydrate phase contained many air bubbles. The bubble distribution reflexes the hydrate decomposition. The air bubble was produced at a relatively high temperature, and the bubbles increase the speed of hydrate decomposition. Our analysis shows that the formation of bubbles disperses carbon dioxide hydrate into numerous small hydrate nucleation particles. Meanwhile, a large number of bubbles speed up the hydrate decomposition process.
The characteristics of varying temperature and pressure during CO2 hydrate decomposition at different initial temperatures were measured, as shown in
The pressure and temperature stabilization time in the reactor at different generation temperatures was always about 30 min. This indicated that there was no apparent difference in the hydrate decomposition rate under different experimental conditions, which might be due to the good heating effect of the experimental apparatus in its horizontal placement, resulting in no apparent changes in temperature and pressure.
During CO2 hydrate decomposition, owing to bubbles escaping, the hydrate decomposed into hydrate particles (hydrate crystal nucleus), which must be generated at the moment of formation. The light transmittance of CO2 hydrate from the beginning of the reaction to complete formation was observed and compared to analyze the CO2 hydrate particle phenomenon, as shown in
CO2 hydrate was not fully formed at the reaction start, with only scattered particles floating at the CO2 and water interface. The light transmittance of the reactor was strong at this time, while the light transmittance of CO2 hydrate at reaction completion was poor. This phenomenon was attributed to the CO2 hydrate crystal nuclei not initially being closely connected after formation, with the single crystal nucleus being relatively independent (as shown in
The particle effect of the hydrate crystal nucleus not only affected the light permeability of the hydrate in the reactor, but also influenced its formation characteristics. CO2 hydrates at the two-phase interface of CO2 and water initially and after a period of the reaction were compared, as shown in
Owing to the particle effect, the hydrate crystal nucleus had a flowing property at the initial stage of the reaction, existing as a hydrate slurry. Furthermore, as the hydrate density was lower than the densities of liquid CO2 and water, and the hydrate crystal nucleus was relatively independent and had a certain fluidity, the hydrate crystal nucleus was transported upward. CO2 also dissolved and diffused into the water phase, causing further hydrate formation, accumulation, and, finally, formation of a solid CO2 hydrate layer.
At the initial stage, when the pressure and temperature conditions met the hydrate formation conditions, the hydrate crystal nucleus started to form (
This particle effect was due to disordered hydrogen bonds at the junction of hydrate crystal nuclei during their gradual growth (
To investigate the particle effects of carbon dioxide hydrate, molecular dynamics simulation of hydrate formation was conducted. Several researchers have studied the molecular dynamics simulation of hydrate formation [
As shown in
To verify our results, we used Material Studio software to model and interleave two groups of hydrate crystal nuclei. We believe that the contact location of the two groups of crystal nuclei is the chaotic region of molecules, and the molecules on both sides are the boundary regions, as shown in
Since hydrogen bonds between the water molecules break rapidly during the decomposition of hydrate crystal cells, it is impossible to accurately observe the motion characteristics of molecules at different positions during the decomposition of hydrate. Therefore, by investigating the changes in the root mean square displacement and velocity correlation function with time during hydrate decomposition, we analyzed the molecular motion law during hydrate decomposition. The molecular simulation ensemble was selected as the NPT ensemble, decomposition temperature was 293 K, pressure was 0.04 GPa, simulation time was 10 ps, and time step length was 0.1 fs. The changes in root mean square (RMS) displacement and velocity correlation functions of molecules at different positions during the hydrate decomposition were simulated, as shown in
During the hydrate decomposition, the RMS displacement of molecules changes linearly, as shown in
(1) During CO2 hydrate formation, the reaction kettle temperature, pressure, and phase behavior underwent relatively complex changes. Under the conditions tested, the induction time of hydrate generation was around 15–40 min. However, the induction time was extended by the liquid CO2 layer hindering contact between gaseous CO2 and water.
(2) A large number of CO2 bubbles were produced during hydrate decomposition. During the decomposition process, CO2 bubbles were mainly formed by precipitation due to increased temperature in the reactor and decreased solubility of CO2 and the CO2 bubbles in the hydrate cage structure.
(3) When the reaction started, only floating scattered particles were present at the interface between CO2 and water. At this time, the reactor showed strong light transmittance and flow performance. As the reaction progressed, the crystal nucleus of the hydrate moved upward under the influence of gravity, CO2 was dissolved and diffused into the water phase, and hydrates were further formed and accumulated, finally forming a solid CO2 hydrate layer. After reaction completion, CO2 hydrate showed poor light transmittance.
(4) The particle effect during the hydrate decomposition can be attributed to the hydrogen bond disorder between the hydrate crystal nucleus during the formation of the hydrate crystal nucleus.
This research was conducted at the Key Laboratory of Enhanced Oil & Gas Recovery of Ministry of Education at Northeast Petroleum University (Daqing, China). The authors gratefully acknowledge the support of the National Natural Science Foundation of China (No. 51574089) and Heilongjiang Provincial Department of Education (TSTAU-R2018018) and the Innovative scientific research project for Postgraduates of Northeast Petroleum University (YJCX2016-013NEPU).