excavations with feflow

I want to set up a model of a local or regional groundwater flow system for calibration. Afterwards i want to insert one or more excavations in this model to determine the flow rate that will enter this excavation and has to be pumped out. As well, i want to to determine the impact that the excavation has on the groundwater flow, especially the change of the groundwater table.

I don't know how to do this with feflow????

The simplest option is to set seepage face boundary conditions (head bc with head = elevation of the bottom of the excavation plus constraint condition max flux = 0) at the bottom of the excavation.

How about increasing K for the part of the model domain that gets excavated? That's the second change we normally make, but is that even required/suggested?

Thanks, Chris

If the excavation is completely dewatered, I'd not change K. A high K is useful if there is an excavation which fills with water (like a gravel pit), and you'd like to simulate the water flow within the pit a little bit more realistically. For underground construction, like tunnels etc. I'd decrease K as otherwise water might infiltrate at the boundary conditions on top of the tunnel, flowing out at the bottom. This usually does not change the flow field, but it influences the budget on different parts of the structures.

The above posts very been very helpful, as I also wish to model an excavation which will both
(a) be empty
(b) fill with water

However, I was wondering, can these both be done with the same model mesh?  My understanding of the seepage face boundary condition is that it would be applied at the edge of my model mesh, hence suggesting there should be a hole in the mesh to represent the excavation.  On the other hand, when the excavation is full of water then there shouldn't be a hole in the mesh.  Or is it possible to apply the seepage face internally inside the mesh?

Any insight would be most appreciated.

Seepage face boundary conditions can also be applied around an excavated section within the mesh. Then you would simply not look at what happens inside the excavation.

[pre][/pre]The above is interesting. I want to gets estimates of inflows into an excavation. To do so, I have first assigned a seepage face to the bottom and slopes of the excavation in my model. I find that the excavation does not dewater all the way to the base. However, I then increase K for the part of the model domain that gets excavated (as mentioned by criga above), and this shows an improved result, i.e. a pit that is nearly dry to the base.

So, my queries are as follows:
1) Is it ok to use an increased K in this case? If yes, would you increase the K (a factor of 10 maybe?) for each layer separately, or would you simply assign a single high value (1E-5 m/s?) of K to all relevant layers?
2) To estimate the inflows into the excavation, I believe I can use the fluid flux analyzer and work out the flow rates into the excavation for each layer?

Any tips on the above would be much appreciated.

Cheers,
Dev

Hi Dev,

please consider that the seepage level will not reach the base of your excavation. From well hydraulics it is known that emptying a well will not lead to seepace face at the base. In respect to head reality does not follow Dupuit assumptions (however in respect to outflow ol' Jules-Étiennes  model is pretty reliable!). There will be always a certain seepage face above the base, The height of which depends on geometrical parameters like the ratio of the equivalent radius of the pit to the distance of your boundary and the ratio of the equivalent radius of the pit to the aquifer thickness. We did some parameter studies for fully penetrating wells and developed dimensionless charts for wells. similar results can be expected in excavation pits.

Regards

Héctor