A Modified Framework to Describe Stress-Strain Behavior and Volumetric Response of Hydrate-Bearing Sand

Abstract

Gas hydrate-bearing sands (GHBS) contain a large volume of methane in the form of hydrate, which makes them an attractive source of energy. Hydrates exert a strong influence on the mechanical properties of sands, where increasing hydrate saturation (Sh) of the pore space leads to an increase in peak strength, post-peak strain softening, and dilation. Methane recovery from hydrates requires dissociation of the hydrate, which involves coupled and complex processes involving methane gas generation, increases in fluid pressure and reduction in effective stresses, along with changes in the mechanical behavior of GHBS, all of which may present potential geo-hazards and engineering challenges. Field-scale hydrate production tests have been carried out to evaluate the potential for methane recovery, however, these have sometimes ended abruptly due to technical failures. For this reason, numerical models are required to evaluate the long-term feasibility of hydrate production and reduce the risk of failure. Thus, the development of an appropriate constitutive geomechanical model is necessary in order to conduct realistic numerical analyses to assess the long-term response of the GHBS reservoir during hydrate production. Early attempts at modeling the geomechanical behavior of hydrate-bearing soils modified the Mohr-Coulomb (MC) model by incorporating a relationship between cohesion (𝑐) and hydrate Sh of the pore space. As the MC model did not capture the overall stress-strain, including volumetric, response of hydrate-bearing sands, recent models have considered Rowe's stress-dilatancy theory. In this model, the increase in strength is related to dilation of hydrate-bearing sand and assumes the impact of hydrate in the pore space to be kinematic in nature (function of soil friction) and ignores any 𝑐 component of the hydrate. Although modifications to Rowe’s theory have included a 𝑐 component, both soil friction (kinematic) and 𝑐 are considered to be constants, although recent laboratory studies suggest that 𝑐 is mobilized and subsequently lost during shearing. In this study, a stress-dilatancy model is developed to better describe stress-strain response of GHBS, including 𝑐 effects. The model includes extensions to the well-known equations for Rowe’s theory to incorporate functions that consider the unique hydrate characteristics, as well as soil density and 𝜎3. The developed model better represents the geomechanical stress dilatancy behavior applicable to GHBS that can easily be implemented in standard elastoplastic models for use in numerical simulations to assess the impact of hydrate, and its dissociation, on the long-term response of a GHBS reservoir.