ECO2N / ECO2M: handling of wettability in a dry zone
Hi folks,
When exposing the initially brine-filled grid cells near injection point to CO2 injection, a dry zone is established. At a certain condition of flow and rock properties water saturation becoming zero in some grid cells despite having assigned a positive values for irreducible- or residual water saturation (when entering relative permeability endpoints). I wonder what is the rationale behind it? Is it because we think evaporation should also include residual wet phase saturation? The expectation is that Sg+Sw+Ss=1 and Sw=Sw_mob+Sw_immob. Further, we may expect that solid phase replaces first mobile water, then immobile, if it supposed to do so. If that is the case, the solid saturation should get as high as Sw_mobile before any Swr is removed? But in that case, the grid cell might already had lost so much of its porosity that the permeability has already reached zero. If the fraction of lost porosity to reach a “zero-point-permeability” given by input as 0.7 (for instance) and Swr=0.1, what happens when 70% of initial pore space is filled with solid salt, would it be 30% trapped high saline water left there as capillary front has moved far away? What about if the fraction of lost porosity is 0.95 instead of 0.7, would it be 0.05 irreducible water saturation left there?
Appreciate your feedback
Godarzi
6 replies
-
Hi Godarzi.
When dry CO2, or any other dry gas, is injected in a medium containing water, water is evaporated according to local thermodynamic equilibrium eventually promoting the complete evaporation of the aqueous phase and the precipitation of dissolved solids. This process has been observed at the field during natural gas injection in saline aquifers, as well when producing natural gas from gas reservoirs with high temperature and high saline connate brine.
The process has been replicated at the laboratory and reproduced by numerical simulation when injecting dry CO2-rich mixtures or any dry sour- acid-gas mixture.
TOUGH2/3 assumes instantaneous thermodynamic equilibrium among all the phases present. Thus, as soon as the dry gas enters in contact with the brine, water is evaporated in order to have the same water fugacity (or partial pressure, depending on the EOS used) in both phases.
Looking at the mass balance equations, you can see that there is no distinction between mobile and immobile aqueous phase. Thus, evaporation drives just a reduction of aqueous phase saturation and an increment of NaCl concentration in the bulk of aqueous phase. Similarly, if NaCl reaches the solubility limit for local conditions, the solid halites precipitates reducing the porosity and permeability.
The models used to account for porosity-permeability relationships are a big question mark, as the deposition of solid halite on the surface of rock matrix alters the pore size distribution. As you said, permeability can be reduced to nearly zero when there is still a large fraction of porous volume available for fluid flow, this because it is enough that the pore throaths are clogged to avoid the fluid flow (see Verma & Pruess models and the large literature on the argument published in recent years).
To improve the simulation of the evaporation process, the inclusion of vapor pressure lowering (VPL) could be considered (implemented in EOS4, EWASG and TMGAS). The reduction of aqueous phase saturation produces an increase of capillary pressure which affects the VPL, reducing the vapor pressure of the aqueous phase. The evaporation is slower and liquid water can persist at P lower than the local vapor saturation P. But finally, the aqueous phase is evaporated even with VPL on.
Regards,
Alfredo
-
Hi,
I am simulating CO2 injection in heterogeneous model on TOUGH3 (ECO2N module). I am facing the same water evaporation problem (Xco2 = 1 or 0.999) or Solid saturation (ss = 1 or 0.9999). I turned on the VPL by changing MOP2(4) = 1 (MOMOP 4 =1, in petrasim), [I got this information from output file generated by TOUGH3, ECO2N module] the screenshots attached. Now I am getting problem to run the simualtion, Now its falling after run of 10-20 seconds frequently. Please is check is it right way to turn on VPL for ECO2N for TOUGH3. Thanks for your help.
-
@Prince Kumar
Not sure I have understood which is the issue you have. Having a complete dry-out of the formation around the injection well? With SW going to zero? SS goin to 1 ?
Usully dry-out occurs because undersaturated CO2 is injected and the residual brine left behind the CO2 front is progressively evaporated by the dry CO2.
The injection can stop if a too high accumulation of solid NaCl occurs, driving permeability to negligible values. This may occur before SS reach 1, depending on the porosity-permeability relationship you are using.
High SS values may be obtained when injecting CO2 at low rates in a formation with high Pcap, so that the capillary gradient overcomes the injection P gradient and the mobile brine flows by capillarity towards the wellbore, to finally evaporates in the same grid element.
Including VPL (vapor pressure lowering) may reduce this process as it reduces the water vapor pressure when the saturation of the water phase is reduced and capillary P is increased. VPL is simulated using Kelvin's eq. in EOS4 and EWASG, but I think is not used in the original ECO2N (Pruess, 2005).
You mention MOMOP option in TOUGH3.
First of all, it seems you gave 1 under column 40, block MOMOP.
From TOUGH3 user's guide, it seems only the frst 27 columns are used. So your input has no effect.
MOMOP describes additional options as shown in Table 7.
Format(27I1)
MOP2(I), I = 1,27MOMOP(4) option is not a true VPL according to Kelvin's eq., but just a reduction (probably linear with respect to SL) of vapor pressure below SL=0.02. It is not available in all EOS modules. You should check if the option is available in ECO2N under TOUGH3.
4 Vapor pressure reduction
0 No vapor pressure reduction at low liquid saturation
>0 Reduces vapor pressure for Sl < 0.02 to prevent liquid disappearance by evaporation (only certain EOS modules)
I think that if you get a too high halite accumulation (if this is your issue), you should try to understand which are the specifications of your problem that give this result, not just try to suppress brine evaporation. Check your input values about rock permeability, capillary pressure, permeability-porosity relationship, content of H20 in injected CO2, injection T, etc.
Regards,
Alfredo
-
Thank You so much for the detailed explanation.
Yes, You are right, I insert wrong MOMOP, and best of my knowledge I did not find any discussion on VPL for ECO2N module in TOUGH3 manual. Still I am searching for that.
Mainly, the simulations are failing for these conditions:
1. Near the injection well, cells have Xco2 = 1 or 0.999
2. Near the injection well, cells have solid saturation = 1 or 0.999
3. Some cells have condition 1 and adjacent cells have condition 2
I also injected water (10^-5*mass flow of CO2) but it doesn't worked. The simulation failed on the same time and same element.
As you suggested, I am trying permeability-porosity relationship and using model (Verma and Pruess (1988)). If you recommend any other model or any other change, I would be thankful.
-
Having XCO2 going to 1 it's OK, as you are injecting CO2 (with a tiny amount of water, which is a good practice to avoid just one component in any of the grid blocks).
If SS goes to 1, of course, you are completely clogging the porous medium. I guess you have already used one of the permeability vs porosity models available in ECO2N. The fact that SS =goes to 1 suggests you have chosen one of the model without a critical porosity. But I guess that even using a model giving k=0 at SS lower than 1, this might not solve the issue of brine flowing towards the well with persistent vaporization in the same element.
Have a look to the Pcap curve and check if unrealistically high values are computed when SL is going to 0.
About the "VPL" option controlled by MOMOP, you may initialize two elements at the same two-phase conditions (water-steam) and at different gas saturations, such as SG=0.5 and SG=0.01. If the initial conditions computed by ECO2N are the same, then it means the "VPL" is not implemented in ECO2N. This if you do not have access to the source code.
The option should not be implemente, if it is not mentioned in the user's guide.
Finally, in case there is no ways to avoid an excessive halite precipitation, you may try to first inject a cushion of pure water, followed by the injection of CO2. Pure water will display the brine during both water injection and CO2 injection (at least until all the water is left in the residual aqueous phase). This might help avoiding the clogging of porous medium. The volume to be injected is likely very case specific.
Regards,
Alfredo
-
-
glad to get your feedback- as always, and clarification on this.
It makes sense to think that evaporation does not distinguish between mobile or immobile water, as long as we think of it as a prevailing equilibration mechanism driven by partial component pressure in different phases. But I can't find any documentation to see how it is handled when permeability / porosity relation intervenes. This is to see whether it is any synchronization or reconciliation among these 3 parameters in a grid block: por/perm relation, material balance and equilibration.
Thank you
Godarzi
