Confined/unconfined best practise

[David Ström]

Hi everyone,

TLDR; in what situations do you apply convertible or confined and when is it most important. Often convertible (unconfined) cause instabilitet. 

In many cases when I set up groundwater models I end up with non-convergence issues if the model is simulated with convertible layers.

When the layers are set to confined, this is rarely an issue. I suppose there is alot of ocillating water levels in the uppermost layers with the convertible alternative.

So, what are your experiences and suggestions when it comes to selecting between these options? And in what situations do you find it more or less important to define them "correctly" (whatever correctly is in a given situation).

To provide a recent example; I'm building a transient model over two years. Following the first steady state timestep there are 3 transient stress period with incremental length to smooth the solution.

The largest part of the model domain has a top layer with clay followed by silt, other parts of the domain can be either sand/gravel, till or bedrock.

I'm running the set up in modflow 6 which gives me the option to assign confined or convertible with the IDOMAIN setting for individual cells.

Conceptually i would expect the areas covered by clay to be confined and the other parts convertible. Switching from fully confined to this gives me convergence problems. 

The top three layers are soil and bedrock, after which an additional 6 layers are bedrock. Would it be more or less correct that my bedrock layers are confined or convertible?

Lastly; I am aware that some solutions have been suggested in various sources. One being to have a smoother horizontal transition between conductivity-zones.

Can this be handled in another way? Seems tedious to add zones between different k-zones with interpolated k-values or something similar.

Best regards

David

[C. P. Kumar]

Dear David,

This group was created by me and now there are few moderators of this group.

This is in reply to your another message (not circulated to the group). There is no limit on who can send messages to the group. Please understand that group messages are moderated. Once a message is posted to the group then one of the moderators of the group need to approve the message (to ensure relevance of the message) before it is circulated to the members. That is why there may be some time lag between posting of a message (starting a new thread or replying to an existing thread) by a member and actually being circulated to all the group members after moderation by one of the moderators.

C. P. Kumar

Former Scientist 'G'

NIH, Roorkee

[David Ström]

Thank you for clearing this out. I was not aware of this. Disregard my double posts, it will not be repeated 

--
You received this message because you are subscribed to a topic in the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/modflow/JUqPaYkX4RM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to modflow+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/1ed66c5c-f560-441f-887d-28d9a0238062n%40googlegroups.com.

[Jakab Andras - Gmail]

David,

MODFLOW considers a cell as confined if the simulated head is above the top of the cell, otherwise the cell will be treated as unconfined provided that it is convertible. Setting a cell as confined just for the sake of making the solution convergent is not highly recommended. In the field, however, the confined status depends on a more complex set of conditions, and a lower permeability unit locally overlying a higher permebility aquifer domain will not necessarily turn that into a confined one. Depending on the modeling objectives, I would leave all cells convertible, and avoid the explicit representation of the heterogeneities within the layers that are straddled by the water table, especially if they don't play a major role in the flow conditions. Use a simplified conceptualization with a lower number of layers in that vertical region. Also, make sure you are using the Newton-Raphson method as it will result in a more stable solution. Alternatively, you may use a package or an engine that simulates unsatirated flow in the vadose zone.

--
You received this message because you are subscribed to the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to modflow+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/9988995f-2f2d-4e5c-a453-a4deaba88b1fn%40googlegroups.com.

[David Ström]

Jakab,

Thank you for an interesting explanation. I can't see any obvious flaws causing the convergence-troubles I'm having. A simplified conceptualisation may suffice in many cases, however. There is a quite sturdy knowledge of the geological units in play here and I would not want to generalize the stratography too much. I will have to ponder on which parameters are best suited for modification to solve whatever is causing this behavior. 
My personal experience is that these issues are more common when dealing with layers of clay or other confining units on top of till and bedrock. I often struggle with the boundary between the confining clay unit and the surrounding till or other material where it moves up through the stratography.

Sincerely
David

[Randall Hanson]

Hey David,

You may want to try MF-OWHM as it has much better diagnostics for convergence issues. You can also set solver properties within NWT solver different for each stress period.

Here’s the USGS download link for MF-OWHM

Cheers 

Randy Hanson

One-Water Hydrologic

http://www.One-WaterHydrologic.com/

Sent from my iPhone

On Dec 14, 2022, at 7:53 AM, David Ström <david.s...@gmail.com> wrote:

Jakab,

To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/4724b986-7106-4d4d-9cd1-1969ed21d187n%40googlegroups.com.

[David Ström]

I will do so. Thank you for your suggestion. It can be tricky at times so simulate bortom low-permeable grout in excavations when there are many different excavations and depths.

The LGR option in ohwm couls be great for this with local vertical discretization att different levels.

Regards 

David 

You received this message because you are subscribed to a topic in the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/modflow/JUqPaYkX4RM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to modflow+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/B60A3171-7698-441F-90EF-049E8C3A362D%40gmail.com.

[Randall Hanson]

Hey David 

Also please know that OWHM offers transient HFB barriers that can be both lateral or vertical barriers. These have been used to simulate temporary barriers for long-wall coal mining in Germany so this may be useful for your possible applications too!!

Cheers 

Randy Hanson

One-Water Hydrologic

Sent from my iPhone

On Dec 15, 2022, at 9:31 PM, David Ström <david.s...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/CAC_Nd8S-ZN08Vfv9kNixLUpVMyHOtP6eijLNhy4WmafBgJEeeQ%40mail.gmail.com.

<OWHM_Icon.png>