well condition: distributing flux according to transmissivity

Hi everyone!

I have a feflow model

->multi-layered
->3D heterogeneous K field
->pumping wells

I'm using the 4th kind boundary condition for the wells, with well bore condition (specific option setting).

I initially thought that I would enter a single value as a global pumping rate and that the flow rate would be distributed to the various layers on which the wells have been place with a flow rate proportional to the transmissivity of each layer of the model.

I realized that I was wrong and entered the pumping for each layer as:

(total pumping rate / n slices = slice pumping rate).

This way, all of the slices have the same pumping rate... even tough it makes no sense in a heterogeneous K field setting...

My question is : ''Do I have to calculate transmissivities by hand and enter the 4th conditions on each slice as a transmissivity related proportion of the whole pumping rate?''

Or is there some built-in function?

Any help or suggestions appreciated!!!

Thanks!
Thomas

Thomas,

You do not have to calculate the rates for each slice per hand. Using the well bore condition, all the well nodes are internally connected by a 1-dimensional element - like a tube. Thus putting the extraction/injection to just one node (usually the bottom node) and a 0 value to the others the extraction/injection will be distributed to all the slices. The distribution is a result from the simulation, it is not done in advance. Besides the transmissivities of the layers also the head distribution has an influence: If the head is generally higher around the well in one slice, the inflow into the well will be higher, too.

You have probably used the Budget Analyzer to track the flow into the well. This doesn't give you the results you are looking for  in case of a multi-layered well. You'll always see the total inflow at the node where you have put the extraction rate, which is due to the fact that the Budget Analyzer only shows boundary fluxes - and the boundary flux indeed occurs at that very node. You cannot use the Budget Analyzer to track internal water fluxes, and the flow within the 1D well element is an internal flow.

The latest version FEFLOW 5.3 contains a new option to calculate the distribution well extraction for each layer as an option in the Fluid Flux Analyzer, which is the tool to calculate internal fluxes. Up to FEFLOW 5.2 there's no such functionality, but if your model doesn't show extreme vertical head gradients you can assume that the extraction is approximately weighted by transmissivity.

Best regards,
Peter

sorry!
think I forgot to say thank you for your reply... I've read it a while ago now!
thomas

Dear Peter

is it possible that multi - layerd wells do not work as 1D - connected elements for abstract water from the model, but work also as 1D - elements in all ways. That means, is it possible, that water could flow vertical through an aquitard layer along the multi - layerd well - boundary condition from a higher aquifer to a lower aquifer ?

My results seem to show a vertical flow along the multi layered well boundary condition through aquitard layers.

do you have any ideas ??

best regards christian.

Yes, the multi-layered well acts like a 1-dimensional discrete element with a very high conductivity. If the well abstraction/injection is low, it can happen that the flow from one layer/slice to another one exceeds the abstraction. In this case, I would recommend to turn off the multi-layer well option for the whole model and insert discrete elements manually with realistic properties of the well pipe. This will avoid a vast overestimation of these flows between the layers.

wow....nice post.....
[right][url=http://salestrackingsoftware.org][color=#F8F8F8][u]sales tracking software[/u][/color][/url][/right]