Björn Kaiser

If you use FEFLOW 7.0 or a later version please use [b]Edges[/b] rather than Join Edges.

0.1

Sure, sorry for not writing earlier. Are you using a 2D or 3D model? Did you change settings for the anisotropy in the Problem Settings?

Yes, the origin for 2D vertical models is quite important, because the head is based on a level over datum. In contrast, 3D models usually have implicitly this level in the input data. If you refer to sea level (0 m), I suggest to keep the origin at 0 m without any transformation or shifting. You simply need to assure that the y-coordinates (gravity axis) in the 2D Supermesh refer to elevations above sea level.

At the moment, FEFLOW 7.1 does not support Visual Studio Community 2017. I suggest to use 2015.

If you are interested in Spline Models you need to define both relative conductivity VS. saturation and capillary pressure head VS. saturation. Otherwise, you can use one of the empirical models as also suggested by Coastal.

Unfortunately, it is not possible to say a priori which parameters are most sensitive with regard to the distribution of the temperature, because there are too many dependencies. As a simple example, in geothermal applications different heat driver may co-exist within the system. If advection is the dominant driver, I would expect a smaller contribution of changes in thermal properties in the computed thermal field. I think a sensitivity study you carry out is quite helpful to identify the major candidates affecting the thermal field.

Hi Anner, yes, you are right. The budget you observe as distributed source/sink are attributed to buoyant flow. If you want to go more into the details, you may contact the FEFLOW Support.

Adaptive Mesh Refinement (AMR) does not exist in recent FEFLOW versions. Indeed, long time ago there was the capability to use AMR. As far as I know, this capability reached rather an experimental stage only.

At the moment FEFLOW does not provide a direct import from Petrel. If you want to import the K-values as you described, I suggest to use an ASCII based dat-file. If you work with partially- or fully-unstructured meshes, I suggest to use centroides of the elements. Centroids of the elements allows you either to regionalize your K-values by means of a 3D interpolation method or by use a Point in Element Algorithm. Here is an example how the file format may look like.

      Id                  X                  Y                  Z                  M  Element            CENTER_X            CENTER_Y            CENTER_Z    LAYER              CONDX
      1  3410643.5567414304  5817349.5977908531      44.2379718868        0.0000000000        1  3410643.5567414304  5817349.5977908531      44.2379718868        1        0.0003802296
      2  3410645.2541029272  5817299.4178183805      43.9502428466        0.0000000000        2  3410645.2541029272  5817299.4178183805      43.9502428466        1        0.0003888353
      3  3410605.6775149289  5817255.8355963556      43.7894600469        0.0000000000        3  3410605.6775149289  5817255.8355963556      43.7894600469        1        0.0003985416

If you do not work with partially- or fully-unstructured meshes, you may regionalize your data by using layered based 2D interpolations. Here is an example how the file format may look like:

      Id  Element            CENTER_X            CENTER_Y            CENTER_Z    LAYER              CONDX
      1        1  3410643.5567414304  5817349.5977908531      44.2379718868        1        0.0003802296
      2        2  3410645.2541029272  5817299.4178183805      43.9502428466        1        0.0003888353
      3        3  3410605.6775149289  5817255.8355963556      43.7894600469        1        0.0003985416

Both file-format examples has been derived by a simple export of the K-values from FEFLOW. It is always helpful to check the export data to get an understanding of the import data.

I also woule like to draw your attention about importing geological models as 3D Supermeshes directly.

FEFLOW has the capability to import triangulated surfaces as a 3D Supermesh. The triangulated surfaces must represent the outer borders of the model domain and optionally internal surfaces which separates different units (e.g. physical rock properties).

The following file formats are supported:
[list]
[li]TetGen Surface Mesh Files (*.smesh)[/li]
[li]Triangle Mesh 3D files (*.mesh)[/li]
[li]Autodesk 3DS files (*.3ds)[/li]
[li]VTK Files (*.vtu)[/li]
[li]Leapfrog Mesh Filesn  (*.ts)[/li]
[li]GOCAD (*.ts)[/li]
[li]GeoModeller (*.gmod)[/li]
[li]Object File Format Mesh Files (*.off)[/li]
[li]Object File Format Mesh Files (*.obj)[/li]
[li]STL Mesh Files (*.stl)[/li]
[/list]