PorousFlow: EOS for Water, Salt, Gas with precipitation and dissolution at elevated temperatures

[batodon]

Hello, this issue is related to ticket #23280. It turns out that the optimal EOS for my simulation should have salt precipitation and dissolution and should also account for porosity and permeability variation. TOUGH2’s EWASG (Battistelli,1997), instead of EOS4, seems the most appropriate. I, however, struggle to get the EWASG equivalent in PF. The closest I could get are PorousFlowWaterNCG and PorousFlowBrineCO2. The latter seems more suitable, so I tried simulating TOUGH2’s EWASG example with it, but my simulation did not converge. It is probably because the PorousFlowBrineCO2 does not include the effect of dissolution and precipitation of halite. Could you please recommend a PF fluid state module that can do this? Or do I need to make any code modifications to PorousFlowBrineCO2? I attached my input file for the EWASG example that I tried. Thanks.

[Mesh]
[gen]
  type = GeneratedMeshGenerator
  dim = 2
  xmin = 5
  xmax = 10000 #10000
  nx = 100        #100    
  bias_x = 1.01
  ymin = -500
  ymax = 0.0
  ny = 1
  []
  coord_type = RZ
[]


[FluidProperties]
  [co2]
    type = CO2FluidProperties
  []
  [brine]
    type = BrineFluidProperties
  []
  [true_water]
    type = Water97FluidProperties
  []
  [tabulated_water]
    type = TabulatedBicubicFluidProperties
    fp = true_water
    temperature_max = 900
    temperature_min = 250
    pressure_max = 1E8
    interpolated_properties = 'density viscosity enthalpy internal_energy'
    fluid_property_file = water97_tabulated_11.csv
  []
  [true_co2]
    type = CO2FluidProperties
  []
  [tabulated_co2]
    type = TabulatedBicubicFluidProperties
    fp = true_co2
    temperature_max = 900
    temperature_min = 250
    pressure_max = 1E8
    interpolated_properties = 'density viscosity enthalpy internal_energy'
    fluid_property_file = co2_tabulated_11.csv
  []
[]


[UserObjects]
  [dictator]
    type = PorousFlowDictator
    porous_flow_vars = 'pgas zi temperature'
    number_fluid_phases = 2
    number_fluid_components = 3
  []
  [pc]
    type = PorousFlowCapillaryPressureConst
    pc = 0
  []
  [fs]
    type = PorousFlowBrineCO2
    brine_fp = brine
   co2_fp = co2
#    type = PorousFlowWaterNCG
#    water_fp = true_water
#    gas_fp =  true_co2           
    capillary_pressure = pc
  []
[]

[GlobalParams]
  PorousFlowDictator = dictator
  gravity = '0.8  0.8  0.8'
  temperature_unit = Celsius
[]

[Variables]
  [pgas]
    initial_condition = 6e6
  []
  [temperature]
    initial_condition = 275.55
  []
  [zi]
  []
[]

[ICs]
  [zi]
    type = PorousFlowFluidStateIC
    variable = zi
    saturation = 0.45
    gas_porepressure = pgas
    temperature = temperature
    fluid_state = fs
  []
[]


[AuxVariables]
  [Xnacl]
    initial_condition = 0.3
  []
 [sgas]
   order = CONSTANT
   family = MONOMIAL
 []
  [sbrine]
   order = CONSTANT
   family = MONOMIAL
  []
[]


[AuxKernels]
  [sgas]
    type = PorousFlowPropertyAux
    property = saturation
    phase = 1
    variable = sgas
  []
  [sbrine]
    type = PorousFlowPropertyAux
    property = saturation
    phase = 0
    variable = sbrine
  []
[]

[Kernels]
  [mass0]
    type = PorousFlowMassTimeDerivative
    variable = pgas
    fluid_component = 0
  []
  [adv0]
    type = PorousFlowAdvectiveFlux
    variable = pgas
    fluid_component = 0
  []
  [mass1]
    type = PorousFlowMassTimeDerivative
    variable = zi
    fluid_component = 1
  []
  [adv1]
    type = PorousFlowAdvectiveFlux
    variable = zi
    fluid_component = 1
  []
  [energy]
    type = PorousFlowEnergyTimeDerivative
    variable = temperature
  []
  [heat_adv]
    type = PorousFlowHeatAdvection
    variable = temperature
  []  
  [conduction]
    type = PorousFlowHeatConduction
    variable = temperature
  []
[]


[Materials]
  [temperature]
    type = PorousFlowTemperature
    temperature = temperature
  []
  [brineCo2Properties]
    type = PorousFlowFluidState
    gas_porepressure = pgas
    z = zi
    temperature = temperature
    xnacl = Xnacl
    capillary_pressure = pc
    fluid_state = fs
  []
  [porosity]
    type = PorousFlowPorosityConst
    porosity = 0.05
  []
  [permeability]
    type = PorousFlowPermeabilityConst
    permeability = '50e-15 0 0   0 50e-15 0   0 0 50e-15'
  []
  [relperm0]
    type = PorousFlowRelativePermeabilityCorey
    n = 4
    s_res = 0.30
    sum_s_res = 0.35
    phase = 0
  []
  [relperm1]
    type = PorousFlowRelativePermeabilityCorey
    n = 4
    s_res = 0.30
    sum_s_res = 0.35
    phase = 1
  []
  [rock_heat]
    type = PorousFlowMatrixInternalEnergy
    specific_heat_capacity = 1000
    density = 2600
  [] 
  [rock_thermal_conductivity]
    type = PorousFlowThermalConductivityIdeal
    dry_thermal_conductivity = '2 0 0  0 2 0  0 0 2'
  []
[]


[DiracKernels]
  [fluid_produce]   
    type = PorousFlowSquarePulsePointSource
    point = '5 -250 0'
    mass_flux = -1.3e-4   # -65  #-0.033   # 
    variable = pgas
  []
[]

[VectorPostprocessors]
  [vars]
    type = NodalValueSampler
    sort_by = x
    variable = 'pgas zi Xnacl'
    execute_on = 'timestep_end'
    outputs = spatial
  []
  [auxvars]
    type = ElementValueSampler
    sort_by = x
    variable = 'sbrine sgas'
    execute_on = 'timestep_end'
    outputs = spatial
  []
[]

[Postprocessors]
  [Pgas]
    type = PointValue
   point =  '5 0 0'
    variable = pgas
  []
  [sgas]
    type = PointValue
    point =  '5 0 0'
    variable = sgas
  []
  [sbrine]
    type = PointValue
    point =  '5 0 0'
    variable = sbrine
  []
  [Xnacl]
    type = PointValue
    point =  '5 0 0'
    variable = Xnacl
  []
[]


[Preconditioning]
  [smp]
    type = SMP
    full = true
    petsc_options_iname = '-ksp_type -pc_type -sub_pc_type -sub_pc_factor_shift_type'
    petsc_options_value = 'gmres bjacobi lu NONZERO'
  []
[]


[Executioner]
  type = Transient
  solve_type = NEWTON
  end_time = 2e6
  nl_max_its = 25
  l_max_its = 100
  dtmax = 1e5
  nl_abs_tol = 1e-20
  [TimeStepper]
    type = IterationAdaptiveDT
    dt = 100
  []
[]

[Outputs]
  exodus = true
  sync_times = '1e4 1e5 2e6'
  [time]
    type = CSV
  []
  [spatial]
    type = CSV
    sync_only = true
  []
[]

[cpgr]

Yeah, WaterNCG is probably closest but doesn't do brine so no salt. BrineCO2 is functionally equivalent to TOUGHS ECO2N, so not quite what you are after.

I think extending WaterNCG to be BrineNCG is probably what you want, but it would take some coding. I'm happy to chat about this and work out how to do it.

[batodon]

Hello @cpgr, I’m not getting any luck with the ewasg example, so I’m approaching my simulation differently based on the publication by Grimm et al. https://doi.org/10.1029/2020WR027213. Specifically, I’m updating my permeability and porosity models to reflect the salt precipitation when Xnacl reaches (and exceeds) the solubility limit instead of introducing a new (solid) phase.

Following this approach, the porosity and permeability should evolve as a function of the fracture aperture during halite precipitation. However, I’m unsure about the routine used (in the publication) to code the precipitation rate (R) and the fracture aperture (b) evolution. Since salt precipitates when the concentration (or the mass fraction in the aqueous phase, i.e., Xnacl) reaches (and exceeds) the solubility limit, I’m guessing that R and b are (coded as) material property in the BrineCO2 fluid property user-Object. After that, they were exported as aux-variables, with the fracture aperture eventually used to update the porosity, permeability and capillary pressure. Is this correct, or I’m I missing something? I searched for the repo corresponding to these publications but found nothing useful. I’ll be deeply grateful if you could share any repo regarding this problem that could help me. Thank you.

[cpgr]

I can't look at this today, but will do it first thing next week. I want this to work so will allocate some time next week to helping getting this going

[batodon]

Thanks for the feedback, @cpgr, It’s very refreshing! I’ll keep working on it too. Hopefully, it works out.

[batodon]

Hello, @cpgr. Please, any updates?

[robo-boluo]

Hello, @batodon and @cpgr,

I am also curious about the evolution of porosity and permeability due to salt precipitation and would like to learn of any updates on this topic.