Fig. 11 . MRI images and inlet pressure and hydrate saturation changes in Cases 6 and 7
Fig. 11 shows MRI images in FOV and the characteristics of hydrate saturation during the whole flow process. In the stage 1 (\(S_{1}\)), the inlet pressure of both cases remained stable around 4.0 MPa. However, the inlet pressure is slightly higher than the back pressure (about 4.2 MPa), because the contact of seawater and hydrate causes the capillary force in the sediments. The MRI images of \(S_{1}\) brightened, indicating the hydrate dissociated during the gas–water flow process with the flow rate of 1–2 ml/min. Because the large amounts of seawater flow enhanced the mass transfer between hydrate phase and seawater phase, the driving force is the chemical potential difference caused by low methane concentration in seawater (Sun et al. , 2020a; Yang et al. , 2019). Meanwhile, the random distribution of hydrate in porous media was the reason that the seawater–hydrate interface in Case 6 and Case 7 was different (Zhang et al. , 2019). The hydrate dissociation areas in Fig. 11(a) could be considered the chimney in the hydrate reservoir (Torres et al. , 2004), the reservoir had no sealing effect and showed well permeability characteristics.
In the stage 2, the gas–water flow in hydrate–bearing sediments with chimney due to hydrate dissociation in stage 1, it was obvious that the reactor inlet pressure was increasing, shown in Fig. 11(b), and the corresponding MRI images went dimmed, hydrate saturation increased. As the seawater flow rate decreased and methane gas flow rater increased, the methane dissolved more fully in seawater. Thus, the hydrate formed inside the hydrate thermodynamically stable area, the gas–water flow process in \(S_{2}\) was the formation process of hydrate–containing sealing layer. In Case 6, no more seawater flowed out of the reactor after 246 min, hydrate–containing sealing layer had sealing effect at this moment. The inlet pressure fluctuated in 274~279 min, and seawater flowed into the FOV. Under the driving of pressure difference, the hydrate–containing sealing layer was partially destroyed and seawater migrated in the reservoir, but the seawater did not break through the sealing layer. The hydrate continuously formed after 279 min, the pore volume is reduced, thus the inlet pressure increased faster. In Case 6, the turning point appeared at 202 min, the pressure difference of in and out is 2.34 MPa in Case 5 and 2.71 MPa in Case 6, respectively. They were all close to the mean value of pressure different (2.47 MPa), the pressure difference of in and out could be a marker to judge the formation of hydrate–containing sealing layer. In hydrate–bearing sediments, the hydrate–containing sealing layer can be formed when hydrate reforms in hydrate reservoir, even though hydrate has dissociated before.