piFreeze leading to negative head

Hello,<br> piFreeze is leading to gradually drifting values of head and pressure, when a temperature timeseries with oscillations around the freezing temperature is used for temperature boundary condition.<br> This leads to both very high and very low head/pressure values (for instance, head as low as -2e8 m are encountered).<br> Illustration:<br> <img alt="" src="https://i.postimg.cc/DwQmMhzR/Capture-d-cran-de-2023-03-29-19-51-37.png" style="width:1280px;height:665px;"><br> This screenshot shows the gradually increasing head values for a few observations points located below the surface.<br> It also shows the few nodes (selected in yellow on the mesh) that exhibit negative head values.<br> NB : all surface nodes have head BC equal to elevation, which explains why the wrong values appear in the subsurface.<br> <br> Running the same simulation without piFreeze clearly reveals that it is responsible of this.<br> Any idea / suggestions to cope with this bug?<br> <br> The fem file can be downloaded from: https://filesender.renater.fr/?s=download&amp;token=2dec035a-0d0d-42d1-9de9-bea2dd2dffe7 (1.3Gb)

Hi Josue,<br> <br> Without looking the file, my first guess is that the observation points are just directly in the frozen soil. FEFLOW piFreeze reduces significantly the hydraulic conductivity and &quot;close&quot; the porosity in order to mimic the freezing conditions. For example, if you have low K values and together a boundary node at the same location, which could cause outflow, the situation may be an uncontrolled decrease in head.<br> <br> Please get in touch with MIKE Support (mike.de@dhigroup.com) for further discussion.<br> <br> Best regards<br> Carlos

Thank you for your answer Carlos.
I understand (and experience!) that piFreeze can cause diverging values of head. Then, could you suggest a solution to get realistic head values? The idea is obviously not to set (unknown) head limits in the model, but to get computed values everywhere, in order to have a consistent flow.

NB: I contacted the support team, and I am currently waiting for their answer.

Kind regards

Hello, To inform the community: this was due to a bug in the piFreeze. A bugfix has now been released for this plugin (under version v1.001). Regards