Jump to content
GMS, SMS, and WMS User Forum

# how to design an open excavation in modflow model

## Recommended Posts

Hi,

i have to model how much water has to be pumped out of some excavation which will be build for the purposes of National Library.

I designed 3d model of geology with different layers and i also have constructed a TIN with the geometry of this excavation.

No my question is, how would i "subtract" this excavation out of a 3D model, so that model would show an excavation. The purpose is

because, this excavation will be secured with impermable ( no flow) diaphragm wall and the aqufer beneth is confined.

Any ideas would be greatly appreciated!

thank you, matevz

#### Share this post

##### Share on other sites

There are a few ways to go about it. Only one is easy, but I would bet that your problem isn't that simple. Easy - if the excavation has flat sidewalls and a constant depth, determine which layer the bottom is in, use a polygon, and set the layer top elevation to be that value. The layer above would be inactive (no flow).

You could try converting your TIN to scatter points, interpolating to a 2-D grid which was created from your 3-D grid, and then working with 2-D data sets (e.g., calculating the minimum of Layer 1 top elevation and the TIN elevation in the data calculator).

You could also copy existing elevations out to Excel and attempt to change them there before copying them back into GMS.

Those last 2 ideas become a bit more complex if the excavation crosses over multiple layers.

Just a few ideas.

#### Share this post

##### Share on other sites

Hi,

thank for your help.

Although i have another question/problem.

I have a multilayer geology model od the area with known water level (s).

The problem (for me) is that i have water level in 2nd layer (unconfined aquifer) than there are two leyer of imermeable sediments and lower there is the main aquifer which is also unconfined. How can i deisgn two water tables in GMS?

thank you, m

#### Share this post

##### Share on other sites

If you're trying to get it to show up in a cross-section view, from what I know, GMS will only draw a water table on the upper-most head.

#### Share this post

##### Share on other sites

hi,

no not to show in the cross section. I'd just like to use the both water tables in the computation. How can i "tell" modflow that there are two water tables?

thank you, m

#### Share this post

##### Share on other sites

Well, I'm sure others have dealt with this before (I have not), but knowing that each layer has its own resulting head, I could imagine that, if the model is set up with two layers of very low conductivity material between the higher conductivity material layers, you will see the difference in head in each layer. I don't know if you will run into convergence problems or not, though.

#### Share this post

##### Share on other sites

hi,

this is what i dont comprehend quite well. how can every layer have its own head? let say there are two layer one on top of another one with k = 10e-3 other 10e-5. the lower one is saturated fully and upper one half. than there is actualy only one (in real life - nature) water table, is it not? But in GMS you get contours in both layers. WHY? hope u understood. thanks anyway, m

#### Share this post

##### Share on other sites

Well, this is how you get an upward or downward gradient. If the lower layer has a higher head (due to more pressure from flow coming from other parts of the aquifer and tighter soils above) than a layer above it, there is an upward gradient. Similarly, a higher head in an upper layer (say due to increased recharge) will drive water down to the lower layer.

With the situation you described in your last post, yes, you have one water table, but the driving forces in the aquifer are shown via different heads in each layer.

Think of the following scenario: An unconfined layer with a head around 100'; a tight clay layer confining the layer below it. That lower layer has a head of 102'. If you poke a hole through the confining layer with a piezometer, you will measure a water level of 102'.

Going back to your first example, if not enough water is getting to that second layer to keep it saturated (perhaps due to well withdrawl in the area), but the confining layer above it is very tight, you now have two very different water elevations in the same area.

#### Share this post

##### Share on other sites

Hmm....

I still do not quite get the display in GMS after the computation.

Attached are the results of my first run of the model that i desgned. First 5 pictures (from left to right, up to down) are the layer of sediments. I understacnd that the first three are dry. I do not quite understand the display that layer 4 and 5 represent (first and second pic from left to right in second row). I guess that flooded cells are colored blue, dry cells are colored red?

Bur the real problem is in cross sections. What does the contours there represent? I dont understand that?

I really appreciate you help,

thanks, m

Edited by matevz

#### Share this post

##### Share on other sites

You are correct - red is dry, blue is flooded. So what has happened is that the head in the blue cells is actually higher than the cell's top elevation and wants to be in the cell above it, but while MODFLOW was calculating a solution, the cell above dried out in an iteration. Once a cell dries out, it won't turn on again, unless you use cell rewetting, but that has its own issues to deal with. A good way to get those cells to be wet is to have the starting heads as close as possible to the desired solution, but that is no guarantee. Depending on what you are doing in the model, it may not matter, but that is for you to determine. Also, if you know that all of those cells will stay dry, it is often better to just make them inactive.

To your cross sections: The picture is not large enough for me to see the contour labels, but I'm assuming that the head is higher as you go up. This means there is a downward gradient. How's about trying this: To help you visualize what is happening, you should go to your plan view of layer 4 (the upper-most layer with wet cells. Select a cell, got to MODPATH and Generate Particles at Selected Cells. Start with one particle on the surface. GMS will automatically generate a pathline showing you the way the particle travels (perpendicular to flow lines). It looks to me like you have more vertical flow than horizontal flow in your model, so you should see that happening.

#### Share this post

##### Share on other sites
You are correct - red is dry, blue is flooded. So what has happened is that the head in the blue cells is actually higher than the cell's top elevation and wants to be in the cell above it, but while MODFLOW was calculating a solution, the cell above dried out in an iteration. Once a cell dries out, it won't turn on again, unless you use cell rewetting, but that has its own issues to deal with. A good way to get those cells to be wet is to have the starting heads as close as possible to the desired solution, but that is no guarantee. Depending on what you are doing in the model, it may not matter, but that is for you to determine. Also, if you know that all of those cells will stay dry, it is often better to just make them inactive.

To your cross sections: The picture is not large enough for me to see the contour labels, but I'm assuming that the head is higher as you go up. This means there is a downward gradient. How's about trying this: To help you visualize what is happening, you should go to your plan view of layer 4 (the upper-most layer with wet cells. Select a cell, got to MODPATH and Generate Particles at Selected Cells. Start with one particle on the surface. GMS will automatically generate a pathline showing you the way the particle travels (perpendicular to flow lines). It looks to me like you have more vertical flow than horizontal flow in your model, so you should see that happening.

Another thing that can help with cells going dry is to make the acceleration parameter of your MODFLOW solver small and increase the number of iterations.

#### Share this post

##### Share on other sites

Hi,

thanks for your reply. What is an accelaration parameter of an MODFLOW solver?

m

#### Share this post

##### Share on other sites

Hi,

how would i draw a diaphragm wall in GMS in the excavation. Attached is the picture of a cross section of this problem. The diaphragm does not completely penetrate layer 5, but if i draw an arc wtih line being the end (bottom) of diaphragm, and use the featrure "Barrier", GMS draws the barrier through the whole 5th layer. Does anyone has an idea?

thank you, m

Edited by matevz

## Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×

×

• #### Activity

• Leaderboard
×
• Create New...