Simulating CO2 injection using EOS2

Hi Saul,

This is a frequently asked question, so I try to explain it in some detail:

(1) It is essential to realize the difference between injection (positive rates) and production (negative rates).

(2) For injection, one has to specify the rate of a component (not a phase) with a given enthalpy. Variable TYPE indicates which component is injected. Whether the injected component is in the gas, liquid (or oil) phase depends on the thermodynamic conditions of the injection element; it may change during the injection.

(3) For production, one specifies a total mass flow rate. One cannot select which component or phase is produced. What comes out of the well is a mixture of phases and components (whatever is in the reservoir!). The only option for TYPE is therefore MASS. This is physically plausible: There is no membrane or filter in the production well that allows, for example, to select that only water molecules but no CO2 molecules enter the well, or that only the gas phase but not the liquid phase is produced (once can "install" such filters, e.g., a phase separator, by choosing appropriate relative permeability functions for the production element). The phase ratio is determined by the mobility ratio, and the mass ratio is a function of the phase ratio and the thermodynamic conditions in the production element (check also MOP(9) for additional details).

(4) So, if you want to change the "feed composition" for injection, you have to split the total rate (in example rfp it is 3.25 kg/s) into two GENER entries, one for the water component and one for the CO2 component. Have both sources inject into the same element, but use different SOURCE names and choose the right TYPE.

(5) People often want to make sure that the injection happens in a specific phase, which cannot be perfectly controlled, as discussed in Point (2) above. For example, if an aqueous solution containing dissolved CO2 is to be injected, make sure that the injection rate ratio Q_CO2/Q_H2O is smaller than the solubility limit - otherwise a gas phase may evolve (unless the injection pressure increases sufficiently to prevent degassing). Similarly, if you may want to inject a gas consisting of air and water vapor, make sure the ratio Q_H2O/Q_AIR is smaller than the maximum water mass fraction in the gas phase (determined by the saturated vapor pressure for the given temperature) to prevent condensation (i.e., a liquid phase) to be injected.

(5) I've just realized that example rfp is misleading in this regard: It has MASS as the TYPE for injection instead of a proper component name. If MASS is chosen, TOUGH2 automatically assigns the first component, i.e., TYPE=COM1 (which is the same as TYPE=WATE). The GENER block in rfp should look like the following for clarity:

GENER----1----*----2----*----3----*----4----*----5----*----6----*----7----*----8
 AA 1INJ 1                         WATE       3.75     5.0E5
 KA 1PRO 1                         MASS      -3.75

Hope this is useful.

Stefan