Jump to content
RockWare Support Forum

Alfredo

Members
  • Posts

    7
  • Joined

  • Last visited

Posts posted by Alfredo

  1. ECO2N V.1.0 (Pruess, 2005, ECO2N: A TOUGH2 Fluid Property Module for Mixtures of Water, NaCl, and CO2) was developed primarily to simulate the geologic CO2 disposal in sedimentary environments at temperatures below 110°C.

    The EOS uses a slightly modified version of the phase equilibria developed by Spycher and Pruess (2005) for  H2O-NaCl-CO2 mixtures. The approach followed for the phase equilibria, which gives the mutual solubilities of CO2 in the aqueous phase and of H2O in the CO2-rich non-aqueous phase, cannot be used at higher temperatures.

    Thus, simply extending the temperature range of CO2TAB file is not enough to perform reliable simulations of CO2 injection in EGS.

     

    ECO2N V.2 (Pan et al., 2015, ECO2N V2.0: A TOUGH2 Fluid Property Module for Mixtures of Water, NaCl, and CO2) has been developed for these applications.

    Alfredo

  2. The primary variable Xsm is used in the EWASG EOS module. Depending on its value, it can be (see the TOUGH2 V.2.0 user's guide (Pruess et al., 1999) pag. 55-56):

    - XS,  the mass fraction of NaCl in the aqueous phase (0-1);

    - SS+10, the solid salt saturation (10-11).

     

    With the same meaning Xsm is also used in the old ECO2 and in ECO2N.

     

    Starting with Petrasim 2017, in EWASG XS can be the mass fraction of NaCl in either the aqueous or gas phase, as the very small solubility of salt in the water vapor is accounted for.  

    In practice: XS = kg NaCl over kg solution, that is kg of NaCl+water+NCG.

     

    Alfredo

     

  3. TOUGH2 (Pruess et al., 1999) does not simulate the mechanical dispersion. It simulates just the molecular diffusion.

    Full hydrodynamic dispersion is simulated by TOUGH2 in 2D cartesian geometries by coupling the T2DM module (Oldenburg C.M e Pruess K. (1993). A two-dimensional dispersion module for the TOUGH2 simulator. Earth Sci. Div., Lawrence Berkeley National Laboratory report LBL-32505, Berkeley, CA.)

    Full hydrodynamic dispersion can be simulated in 3D non structured grids by a special version of TOUGH2 called T2R3D (Wu Y-S, Pruess K. (2000). Numerical simulation of non-isothermal multiphase tracer transport in heterogeneous fractured porous media. Adv. in Water Resour., 699-723.)

     

    For a cartesian grid, for which the IFD method is equivalent to the finite difference method, the numerical dispersion induced by the space discretization  is equivalent to the mechanical dispersion for which the longitudinal dispersion coefficient is half of grid spacing (Bear, 1979). Emulating the mechanical dispersion with grid spacing can work only for very simple cartesian grid geometries.

     

    Alfredo

  4. TOUGH2 (Pruess et al., 1999) does not distinguish between 'total' and 'effective' porosity. A look to the way the mass balance equations are written shows that 'total' porosity is used.

    I was working with an in-house version of TOUGH2 to simulate the waterflooding (with a polymer) in oil reservoirs, and we had the need to account for a fraction of the pore volume which is not directy displaced by the injected fluid. Actually, we wanted to simulate that a fraction of the water saturation (the connate water saturation) was not displaced by the polymer solution.

    We tested two different approaches:

     

    1. use a dead pore volume approach, following what is done in ECLIPSE by Schlumberger.  The accumulation term for the polymer was modified as shown below.

     

    image.png.b1d4aedb4d58fbe02e3d79cffc490b37.png, where Sdpv is the dead pore volume water saturation.

     

    2. use a MINC approach with 2 nested media. The internal one was emulating the fraction of pore volume not directly accessible to the injected fluid,  actually that containing the connate water. The other was representing the effective porosity. The approach was inspired to the work by Falta (2000) Numerical modeling of kinetic interphase mass transfer during air sparging. Water Resources Research, 36, 12, p. 3391-3400.

     

    By a proper choice of model parameters, the two approaches were equivalent as far as the transport of polymer is concerned.

    Of course the MINC approach makes directly inaccessible the 'connate' water to all the mass components simulated.

     

    The first approach requires the modification of source code. The second one can be used with the standard TOUGH2 version, but requires the doubling of the element number.

     

    Alfredo

     

     

  5. What TOUGH2 does not like is to solve  the mass balance of a component which is not present at all in a grid block.

    For this reason initializing a non zero mass fraction of water 2 allows to proceed with the simulation. Actually, initial concentrations much lower than 0.01 can be succesfully used.

    Again, in case of injection (sources), if pure water is injected for long times the mass fraction of water 2 can be reduced to 0. All sources should not inject just water, but also a tiny fraction of water 2, adding an additional source. 

     

    This problem is in principle present in all the EOS with more than just one component.

    Alfredo

×
×
  • Create New...