1. COUPLED HYDRO-MECHANICAL SIMULATIONS
Dr Paul Duplancic

2. Background
Simulation Aided Engineering

4. Why? Hydro-Mechanical Coupling
Because the effect of fluid pressures acting on the solid skeleton of a porous medium plays an important role for many geotechnical applications (groundwater, reservoir examples)
Effective stress concept presence of pwp influences the constitutive response
Because of Similitude / History Matching comparing like-with-like
Mechanical
Displacements (subsidence)
Stresses
Plastic strains (damage)
RER (seismicity)
Hydrological
Pore Pressures
Flux rates
Similitude = good for ‘calibration process‘
Because of the importance of groundwater for geo-technical applications
Traditionally the solution of the groundwater models was separate from mechanical models (due to restrictions in computational capacity)
However, the mechanical problem and the hydrological problem are coupled-problems in reality (mainly to name the change of conductivity with increase of damage and flow in fractured rock along gaps or in narrowed fault zones)
Mine-scale models require minimum-standards of complexity to achieve similitude, hence full coupled solution approaches are desirable but difficult to implement.
Very often information for the hydrological aspects of the problem are limited and have higher uncertainty, hence staggered coupling schemes can be used.
Due to the complexity required for large-scale models to achieve similitude, both mechanical and hydrological problems have often been analysed separately, de-coupled, in the past. But in reality it is a coupled problem

5. How? Hydro-Mechanical Coupling
For 3D, large-scale (mine- or reservoir-scale) models including high complexity in geometry and sequence
Separate solvers for both the mechanical and hydrological analysis
Both solvers communicate with each other, exchanging data at runtime
𝝈 eff = 𝝈 tot + 𝑏 𝛼 𝑝 𝑤 𝟏
𝑘 𝑤 = 𝑘 𝑤0 ⋅𝑓 𝑃𝑆𝑇
pw: wetting phase pressure
PST: plastic strain (rock mass damage quantity)
Mahyari-Selvadurai (1998), Shirazi-Selvadurai (2005) [exponential law, based on equiv. deviatoric (plastic) strain]
technically
conceptually

6. How? Hydro-Mechanical Coupling
rock mass damage distr.
Rock Mass Damage
Hydraulic conductivity
Minor
Sign.
Moderate
None
Very Sign.
𝝈 eff = 𝝈 tot + 𝑏 𝛼 𝑝 𝑤 𝟏
𝑘 𝑤 = 𝑘 𝑤0 ⋅𝑓 𝑃𝑆𝑇
Mahyari-Selvadurai (1998), Shirazi-Selvadurai (2005) [exponential law, based on equiv. deviatoric (plastic) strain]
Chin et al. (2000) [power law, based on volumetric strain via porosity]
Hsiung et al. (2005) [power law, based on stress and plastic failure strains as well as fracture measures]
Mahyari-Selvadurai (1998), Shirazi-Selvadurai (2005)

7. EXAMPLE: model features
Large extent, relatively close to topology surface
Sedimentary domains
Detailed sequence
𝑘 𝑤,Aquifer ≫ 𝑘 𝑤,Aquitard
orthotropic properties

8. SUBSIDENCE MATCH
Subsidence

9. Rock MASS DAMAGE COMPARISON

10. PORE WATER COMPARISON
Match to piezometric measurements of pore water pressures Capture effect of perched water table above goaf

11. RESULTS: PWP AND DAMAGE
damage and PWP eveolution with sequence
 assessment of hydraulic interconnectivity

12. RESULTS: AQUIFER INTERCONNECTION POTENTIAL
heterogenious (here: orthotropic) permeabilities
independent damage-dependency

13. CONCLUSIONS
almost any fluid-flow simulator could be used (establishing coupling via datafile-i/o at runtime)
fluid-flow simulator reports wetting-phase pressures to geomechanical rockmechanics-simulator
rock mass damage is reported back to fluid-flow simulator, which can use updated hydraulic properties
much more realistic behaviour of fluid mechanics and geomechanics forecasting
additional benefits:
obtaining subsidence-plots
establishing damage-induced ‘new‘ flow-channels
capture fault flow and transient flow
robust coupling framework
each simulator can run on machines located in different parts of the world (good experience Australia <-> Europe)