Modelling of high temperature (600 ℃) air injection.

No module can handle such high temperature in TOUGH3. If you believe no supercritical water will appear, you may try EOS6 in TOUGH4 (https://lbl-1.gitbook.io/tough4-user-manual/process-modeling/eos6)

Dear Xin,

you posted in the TOUGH2 section, so I assume you are using TOUGH2.

No standard TOUGH2 module can go above 350°C, which is the limit of liquid water properties of Region 1 in the IAPWS P-T diagram.

But the correlations for region 2 for the steam phase already implemented in TOUGH2 can go up to 800°C. TOUGH2 implements the IFC-67 correlations, while the IAPWS-IF97 correlations perform a bit better.

So, in principle, if in your simulation Temperatures higher than 350°C occur only  in single gas conditions, actually in steam-like supercritical conditions, then Region 2 correlations needed are already in place in TOUGH2. By modiying the EOS you should be able to go up to 800 bar remaining into the Region 2 boundaries. Modifications should include:

- removing the T<350°C in single gas conditions;

- using IAPWS-IF97 correlations for water and steam properties;

- using IAPWS-2008 for steam viscosity to have reasonable steam viscosity values at supercritical steam conditions;

- implementing checks that intercept conditions not allowed, such as entering into Region 3 (supercritical).

- implementing correlations for the gas enough accurate at high T. 

 I implemented this approach in EOS1H and EOS2H modules, that is the steam-like supercritical versions of EOS1 and EOS2. Something about EOS2H has been published in:

Battistelli A., Finsterle S., Marcolini M., Pan L. (2020). Modeling of coupled wellbore-reservoir flow in steam-like supercritical geothermal systems. Geothermics, 86, 101793. DOI: 10.1016/j.geothermics.2019.101793.

Battistelli A., Finsterle S., Pan L. (2018). Modeling of coupled wellbore-reservoir flow for supercritical geothermal systems in steam-like conditions. TOUGH Symposium 2018.

I also modified TMVOC to run above 350°C when no VOCs are included in the simulation. In this non distributed TMVOC version (under Petrasim)  you can handle air (or a mixture of NCG) and water in Regions 1, 2 (up to 800°C) and 4. Same limitations mentioned above for the water phase and Region 3. It was used to model the heating of an unsaturated zone, initially at atmospheric P and two-phase conditions, using electrical heaters. Below sample plots of T history in one grid element and a map of T distribution at a given simulation time. 

So, the answer is: yes, EOS3 can be modified to handle steam-like supercritical conditions. Modification of EOS3 for full supercritical conditions is feasible, but would require MUCH higher efforts.

But be sure that your modeled system will not go into the supercritical region 3. You want to simulate hot air injection into an aquifer and you estimate total P will be between 10 to 60 bar. Are you sure about that? Injecting air will immediately generate two-phase conditions. The air plume will displace the aqueous phase at a speed higher than that of the heat front, steam will be generated by the evaporation of residual water left behind the two-phase front and this will increase the P because of steam generation. 

Complex processes indeed. Not completely sure that depending on model and injection parameters, conditions for region 3 will not be encountered. The water should evaporate completely at T lower than 350°C in order to be able to run the simulation. Probably the regulation of air injection by limiting the injection air P should help in controlling what happens in the aquifer.

Regards,

Alfredo