0

Query on Early Steady State in Methane Discharge Model at 1% Hydraulic Gradient

Dear TOUGH2 Group,

I am modeling methane discharge from a 1m^3, source producing methane at a rate of 1.16E−7 kg/(s⋅m^2). The objective is to observe accumulation and migration through a homogeneous hillslope under three hydraulic gradients: 0%, 1%, and 10%.

For both 0% and 10% gradients, the simulation runs successfully for the entire 10-year duration. However, for the 1% gradient, the model reaches steady state after only 0.64 years. This early stabilization prevents me from observing methane migration to the atmosphere.

I have attempted to extend the simulation time by adjusting:

  • Time step control parameters (e.g., DELTEN)
  • End time (TIMAX)
  • Solution weighting and convergence controls

Despite these efforts, the simulation consistently stabilizes prematurely at the 1% gradient.

Questions:

  1. Could there be a specific physical or numerical reason why the intermediate gradient reaches steady state so quickly, while the two end-member gradients do not?
  2. Are there additional adjustments or solver options that might help prolong the transient phase to allow for gas migration to the atmosphere?

Any insights or suggestions would be greatly appreciated!

5 replies

null
    • kenny
    • 9 days ago
    • Reported - view

    please post the output log file here. There are many reasons that cause the simulation to stop. If it stops (reaches steady) for continually converging on one single time step, you may try using a large initial time step size.

      • Jordon_Massey
      • 8 days ago
      • Reported - view

      @kenny Thank you for your response! I have attached the output file from my simulation for your review.

      Based on previous forum discussions, I attempted adjusting the time step and convergence criteria to delay reaching steady state. However, these changes did not yield the desired results.

      OUTPUT.7z
      • kenny
      • 8 days ago
      • Reported - view

       You model has reached the maximum number of time steps of 9999. If you want to run more time steps than 9999, you  may input MCYC=-20. 

      By the way, it seems your model has a wrong initial conditions which lead to wrong phase transition (You can find the printing out pressure is negative value or NaN at the beginning of time stepping). With such a small time step size,  the steady state can never be reached.

      • Jordon_Massey
      • 5 days ago
      • Reported - view

       Was able to activate -20 = MCYC! So Thank you!

      My thermodynamic skills are admittedly a bit rusty, so I spent much of the weekend delving into the controls of phase transitions in TOUGH to better understand your additional comments. Thank you for pointing me in the right direction—Henry’s law stood out as particularly relevant in this context.

      I do have a couple of follow-up questions to clarify some new thoughts that arose as I worked through your feedback.

      For the initial conditions in this model, I started by running a simulation with no gas source. I created an empty aquifer and allowed fluid to flow at the boundaries until the system reached steady state. I then used the conditions from that steady-state run to define the initial conditions for the current model.

      In the current setup, water is still allowed to flow in and out of the boundaries, but I added a gas source. This may explain why I observe the evolution of gas and liquid phases at the beginning of the simulation run.

      If this is indeed the cause, should I be concerned about the implications?

      Thank you again for your guidance—I greatly appreciate your help!

    • kenny
    • yesterday
    • Reported - view

    I do not think any problem for the set up of initial conditions.  Adding gas sources should not bring negative pressure or NaN value. I would suggest your do numerical experiments by running  the model with no/tiny gas source. If the problem is still there, it may indicate that the steady state solution is not correct. 

Login to reply

Content aside

  • yesterdayLast active
  • 5Replies
  • 40Views
  • 2 Following