Question about Modeling Gas Dispersion

[Caleb Taylor]

Hello everyone,

I am working on a project using CONTAM to model the dispersion of gas species throughout a 3-story building (~90 zones) that does not normally exchange air with the ambient atmosphere (unique application). I connect the “exterior” zones to ambient with flow elements set to have a negligible airflow. This has yielded steady-state airflow results consistent with the expected values. I defined the initial atmosphere (composed of N2, O2, Ar, and CO2) and then used a burst model to release ~10 lbs of a heavy gas contaminant into the building’s atmosphere. I defined the molar mass, diffusion coefficient, effective density, and default concentration for all gas species. Ultimately, I want to determine the steady-state concentration of the contaminant gas in each zone and then ventilate with the ambient atmosphere to restore the buildings air.

When I perform transient contaminant simulations the results converge to the same concentration in each zone, regardless of zone elevation. This is inconsistent with the expected result of the heavier gas having a higher concentration on lower elevation zones. When reviewing the program documentation I have been struggling to determine if this is an intended result, based on the well-mixed assumption. I am currently experimenting with 1-D zones to correct this, but haven’t had much success. Is it possible to simulate gas dispersion with CONTAM and have buoyancy effects enabled?

Thank you,

Caleb

[William Dols]

Caleb,

CONTAM is generally a "well-mixed" model, so zones are represented as having a single concentration within each zone/node.

It does not provide for the separation or settling of a gas species within different regions of air of a given zone based on species properties.

There are several aspects of CONTAM that you may be misunderstanding and would take more than a single email to clarify.

Please start with a simple model and explore the behavior before working with such a complex, 90-zone system.

You might also consider working with a different model altogether, depending on your requirements. 

Feel free to reach out to me directly via email if you wish to discuss further.

- Stuart