DEM simulations of methane hydrate exploitation by thermal recovery and depressurization methods

• Published in: Computers and Geotechnics
• Main Authors: Jiang, M, Fu, C, Cui, Liang, Shen, Z, Zhu, F
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2016
• Subjects: Methane hydrate
• Online Access: http://epubs.surrey.ac.uk/811016/1/DEM%20simulations%20of%20methane%20hydrate%20exploitation.pdf; http://epubs.surrey.ac.uk/811016/7/SRI_deposit_agreement.pdf; https://doi.org/10.1016/j.compgeo.2016.05.011

Methane hydrate (MH, also called fiery ice) exists in forms of pore filling, cementing and load-bearing skeleton in the methane hydrate bearing sediment (MHBS) and affects its mechanical behavior greatly. To study the changes of macro-scale and micro-scale mechanical behaviors of MHBS during exploitation by thermal recovery and depressurization methods, a novel 2D thermo-hydro-mechanical bonded contact model was proposed and implemented into a platform of distinct element method (DEM), PFC2D. MHBS samples were first biaxially compressed to different deviator stress levels to model different in-situ stress conditions. With the deviator stress maintained at constant, the temperature was then raised to simulate the thermal recovery process or the pore water pressure (i.e. confining pressure for MH bond) was decreased to simulate the depressurization process. DEM simulation results showed that: during exploitation, the axial strain increased with the increase of temperature (in the thermal recovery method) or decrease of pore water pressure (in the depressurization method); sample collapsed during MH dissociation if the deviator stress applied was larger than the compression strength of a pure host sand sample; sample experienced volume contraction but its void ratio was slightly larger than the pure host sand sample at the same axial strain throughout the test. By comparison with the laboratory test results, the new model was validated to be capable of reproducing the exploitation process by thermal recovery and depressurization methods. In addition, some micro-scale parameters, such as contact distribution, bond distribution, and averaged pure rotation rate, were also analyzed to investigate their relationships with the macroscopic responses.