Posted Tue, 20 Mar 2018 11:05:33 GMT by Roman Laube
Hello everyone,

i am trying to develop a rather simple FEFLOW model in order to simulate the temperatur plume of a shallow geothermal well. Even without any wells, my model is not behaving as I did expect. So far my setup contains of a rectangle with 40 slices with 0.5 m layers (Elevation in total is 20 m). I set the hydraulic head on the west side of the model to 10m and on the nodes of the east side of the model to 5 m hydraulic head to simulate groundwater flow. Fully saturated, steady fluid flow, conductivity in every direction is 86 m/d. Free surface is set on the first slice, all the other are dependent.

When I simulate the model, I observe the hydraulic head with a 3D line from the east to the west-side of the model and expect a linear behaviour. Unfortunately the line is curved. Is there an explanation?

(I attached a picture with an example-file with the same problem just with different parameters:hydraulic head-west:20m hydraulic head-east:10m)

Posted Tue, 20 Mar 2018 11:52:03 GMT by Björn Kaiser
I would need more details about the model to find an explanation. I suggest to contact the FEFLOW Support:
Posted Tue, 20 Mar 2018 13:24:59 GMT by Roman Laube
I will go with the example file and state you pretty much all the parameters that I set and attach the file.
Polygon Dimensions: 100 x 100
Slice1: Elevation is 20 m
Slice 41: Elevation is 0 m
all slices in between are 0.5 m thick

Meshing: Triangle
Proposed Elements: 1000
Meshing Density 1.0

All elements have the same conductivity: K_xx,K_yy,K_zz = 86m/d

All nodes on the western side of the model on all layers: Hydraulic-head BC= 20 m
All nodes on the eastern side of the model on all layers: Hydraulic-head BC= 10 m

Problem Settings:
Simulate Flow via = Standard (saturated) groundwater flow equation
State: Fluid Flow = Steady
Free Surface: Unconfined Aquifer
Slice 1-Status=Free
All other slices= Dependent
Top of model domain= unconstrained head
Bottom of model domain = unconstrained head

Any other informations required? Am i right with the assumption that the line/curve of the hydraulic head in an unconfined aquifer is supposed to be linear? Many thanks!

(The file is attached)
Posted Tue, 20 Mar 2018 15:19:40 GMT by Björn Kaiser
I suggest to contact the FEFLOW Support:
Posted Wed, 21 Mar 2018 07:36:27 GMT by Peter Schätzl
As the transmissivity changes with water level, the hydraulic-head curve in an unconfined aquifer cannot be expected to be linear. Where the water level is higher, transmissivity is higher and the gradient will be lower and vice versa. If you switch to confined, you'll see a linear curve.
Posted Wed, 21 Mar 2018 10:02:40 GMT by Roman Laube
Thank you very much!

You must be signed in to post in this forum.