Too many timesteps

[hayleycas]

I am trying to simulate the injection of contaminated water into an aquifer at an injection rate of 25L/s over a 16 month period using x2t. The aquifer has a flow rate of 0.001 m/day and the domain consists of cells 100cm x 100cm x 2m with a total of 100 cells. When I run the model the number of timesteps is arond 10 billion, which would take forever to run. I have tweaked the parameters as much as I possibly can but do not seem to be able to reduce the number of timesteps.

 

I would really appreciate someone having a look at my script and telling me if I have an error in the script or whether the scenario is the problem.

 

I have attached the script to this email.

 

Cheers

Scenario 4 v1d.x2t

[Brian Farrell]

Hello,

 

According to your script, your entire domain is actually 100cm x 100cm x 2m (The length and width commands are for the entire domain, not individual nodal blocks). Assuming you want your individual cells to be 100cm x 100cm x 2 m, this comes to a volume of 2,000,000 cm3 per cell. With 5% porosity, the volume of water per cell is 100,000 cm3. Since you are pumping water into the domain at a rate of 25 L/s (25,000 cm3/s), the entire fluid volume of the node with the well is replaced every four seconds. Is this realistic?

 

I think the pumping rate you specify might be too high (at least relative to the size of your domain), which creates a very large hydraulic potential gradient and hence large fluid velocities. To remain stable, time steps must be kept very small, so that fluid does not travel farther than the length of a single nodal block in one time step. You should look into the von Neumann and (especially here) the Courant stability criteria, and whether all of your parameters are correct. The large pumping rate and the small domain work together to force very small time steps, and because the simulation lasts a relatively long time, you need many time steps.

 

Hope this helps,

 

Brian Farrell

Aqueous Solutions LLC