Fig. 9 . (a) multi–level MRI images in Case 4; (b) hydrate saturation in different areas
Fig. 9(a) was the multi–level MRI images of reactor after 158 min in Case 4. The camera on the MRI device remained stationary, and moved the reactor 30 mm upwards after each shot, for the reason the FOV was 30 mm\(\times\) 30 mm, complete reservoir status is presented by this operation. The gas–water injection was stopped and the inlet pressure was constant at 13.0 MPa during the whole shooting process. Significant brightness differences could be found in the hydrate reservoir. By comparing the hydrate saturation in different areas, as shown in Fig. 9(b), the hydrate saturation was higher in farther area alone the gas–water injection direction. The pore seawater in area C was almost consumed up, tiny amount of seawater still remained in the hydrate reservoir and it is too saline to form hydrate, it was hydrate zone. The brighter area (areas A and B) compared with area C was water zone (seawater, hydrate, gas coexisted). The free gas was hindered out of the reactor inlet. Thus, the hydrate reservoir can be divided into three layers. This is very similar to the hydrate reservoir with three structures in the South China Sea: gas hydrate layer; mixing layer consisting of hydrates, free gas and seawater; free gas layer coexists from top to bottom (Qin et al. , 2020). And the experimental phenomenon is consistent with the model for free gas migration presented by Liu et al (Liu and Flemings , 2006).