Sand production in the reservoir is a key issue restricting the safe exploitation of natural gas hydrate, and reservoir permeability is a key parameter affecting the production capacity of hydrate exploitation. The effects of particle size distribution of reservoir, hydrate saturation, depressurization amplitude and pore size of sand prevention layer on reservoir sand production and permeability during hydrate depressurization exploitation were investigated by laboratory model tests. The correlation between the reservoir sand production and change in reservoir permeability was analyzed. The results show that: (1) Factors influencing reservoir sand production: the lower the clay content, the larger the depressurization amplitude, and the greater the pore size of sand prevention layer were, the larger the amount of reservoir sand production. With the same pore size of sand prevention layer, the amount of reservoir sand production has approximately a linear relationship with the average flow rate of water. (2) Factors influencing reservoir permeability: the reservoir permeability generally increased with time, and finally gradually tended to a stable value. The greater the hydrate saturation, and the greater the clay content were, the smaller the initial value of permeability. The greater the depressurization amplitude, and the larger the pore size of sand prevention layer were, the greater the ratio of final value to initial value of permeability. (3) The correlation between the sand production and variation in permeability: with the same hydrate saturation, the more the amount of reservoir sand production was, the relatively greater the increase amount of reservoir permeability. With the same hydrate saturation and particle size distribution of reservoir, the more the amount of reservoir sand production was, the relatively greater the increase rate of reservoir permeability.

The leakage and erosion of shield tunnel segment will change the pore pressure distribution of surrounding soil, arise deformation of surrounding soil, and arise variation of internal force of segment, which will affect the safe operation of tunnel. The effects of sand particle size, pressure head, overburden pressure and relative density of sand on water leakage, sand leakage and ground surface settlement were investigated by a self-developed test device for simulating leakage and erosion of shield tunnel segment. The test results show that: 1) The larger the average particle size of sand was, the greater the coefficient of permeability, the greater the flow rate, the greater the water leakage with the same water crossing section, the smaller the sand leakage and surface settlement of specimen. When the ratio of average particle size of sand to diameter of leakage hole was small, sand was continue to leak out, and the cumulative sand leakage increased linearly. When the ratio of average particle size of sand to diameter of leakage hole was greater than 0.15, the sand particles could form the stable soil arch around the leakage hole, and the average particle size of sand forming the soil arch was much smaller than the diameter of leakage hole. 2) The greater the pressure head on the top of specimen was, the greater the hydraulic gradient, the greater the flow rate and water leakage, the greater the seepage force applied on sand particles, the greater the sand leakage and surface settlement of specimen. 3) The greater the overburden pressure was, the relatively smaller the coefficient of permeability, the smaller the flow rate and water leakage, the greater the friction resistance when the sand particles moved, the harder the sand particles to move, so the smaller the sand leakage and surface settlement of specimen. 4) The relatively looser the specimen was, the relatively larger the void ratio and coefficient of permeability, the relatively larger the flow rate and water leakage, the relatively easier the sand particles to move due to the weak dilatancy, the relatively larger the sand leakage and surface settlement of specimen.