Variance-based determination of dominant model parameters for sand migration in homogeneous gas hydrate-bearing reservoir

Abstract

Sand migration in gas hydrate-bearing reservoir poses a serious problem for a successful long-term gas production. Because gas production is achieved through hydrate dissociation, often driven by depressurization, the process of sand migration involves highly coupled multiphysics behavior. For example, hydrate dissociation causes sediment deformation and may increase the potential of sand migration but hydrate dissociation can also increase permeability, which may lower hydraulic gradient at a given flow rate, leading to reduction in sand migration. Other factors include that sand inflow (or outflow) would cause the increase (or decrease) in pore pressure due to void volume change and thus may halt (or accelerate) hydrate dissociation. An analytical thermo-hydro-mechanical sand migration model to incorporate these interacting features requires a number of parameters and it is important to quantify the significance of each parameter to this complex process of sand migration in gas hydrate-bearing reservoir. This study, therefore, conducts a series of sand migration analyses in field-scale homogeneous gas hydrate-bearing reservoir subjected to depressurization and presents the relative importance of each parameter to the volume of produced sand from a production zone. It is found that the volume of produced sand is mostly dominated by the parameter converting shearing deformation to sand detachment potential and by the parameter increasing critical hydraulic gradient for sand detachment with hydrate saturation.

Mechanistic Understanding of Sand Production in Marine Hydrate Exploitation: Insights from Field Tests and THMC Simulations

“…The immobile sand will be mobilized when the hydraulic gradient exceeds the critical value. The equation for generating eroded solid mass can be expressed as follows d m ssi = prefix− ω 1 m ssi ( ω 3 ε d + ln true( m italicssi m ssi 0 true) ) italicdt , italicif i w > i w cri where ω 1 is the controlling parameter for the rate of detachment, s –1 ; m ssi is the intact solid mass, kg/m 3 ; m ssi 0 is the initial intact solid mass, kg/m 3 ; ε d is the deviator strain; ω 3 is the model parameter accounting for the increase in potential due to shearing deformation; i w the hydraulic gradient of water; i w cri is the critical hydraulic gradient for grain detachment, which is influenced by the hydrate saturation through i w cri = i w cri 0 ( 1 − S H ) ω 2 where ω 2 is the increasing factor of the critical gradient with hydrate.…”

“…3,5,7 For example, Uchida et al first proposed a practical sand production model which described detachment and migration of skeletal particles during hydrate dissociation, and established a thermo-hydro-mechanical sand production coupling model for hydrate reservoir exploitation. 8,9 Their coupling model presented good predictive ability for sand production in Nankai field tests, 3 and thus has become a primary reference for subsequent modeling studies about sand production. et al numerically simulated gas production behavior in the case of sand production during hydrate exploitation, and analyzed the effects of various degrees of sand production on gas production performance.…”

“…The detached particles enter the flowing system and would migrate with fluids. Because sand production does not occur in isolation but rather is closely related to other physical processes such as fluid flow, a few studies have utilized multiphysics coupling modeling methods to predict and analyze sand production processes. ,, For example, Uchida et al first proposed a practical sand production model which described detachment and migration of skeletal particles during hydrate dissociation, and established a thermo-hydro-mechanical sand production coupling model for hydrate reservoir exploitation. , Their coupling model presented good predictive ability for sand production in Nankai field tests, and thus has become a primary reference for subsequent modeling studies about sand production. Adopting the similar coupling model with Uchida et al, Zhu et al numerically simulated gas production behavior in the case of sand production during hydrate exploitation, and analyzed the effects of various degrees of sand production on gas production performance. , However, it has to be acknowledged that numerical simulation studies on sand production during hydrate exploitation are still limited so far.…”

Sand Migration Simulation during Gas Production from Gas Hydrate Reservoir at Kuparuk 7–11–12 site in the Prodhoe Bay Unit, Alaska

“…Uchida et al developed the sand migration model for a hydrate reservoir that incorporates the mobilization of sands, sand flow, and its coupling effect with pressure, temperature, and the sediment responses into a thermo-hydro-chemo-mechanical formulation that solves the governing equations for soil, water, gas, and hydrate. The model was used for history matching, including sand production during the 2013 Nankai offshore gas production test, and for understanding the sand mobilization and migration phenomena in interbedded sediments. , Furthermore, the authors previously studied the variance-based sensitivity of the model parameters and determined the most influential parameters that govern sand migration processes . This has resulted in eliminating three model parameters (out of six) that incorporated sand settling, sand lifting, sand flowing with gas, and effective stress reduction by sand mobilization.…”

“…11,12 Furthermore, the authors previously studied the variance-based sensitivity of the model parameters and determined the most influential parameters that govern sand migration processes. 13 This has resulted in eliminating three model parameters (out of six) that incorporated sand settling, sand lifting, sand flowing with gas, and effective stress reduction by sand mobilization. The simplified formulation is as follows.…”