Jump to content
GMS, SMS, and WMS User Forum

SAMG and GMG Solver Error - Dewatering System?

Dirk Herding

Recommended Posts

Hello everyone,

I‘m trying to model a regional scale model with 12 layers and ~5 mil. cells. Because I do not work with different active areas for every layer, within the outcrop area plenty thin cells occur. They can fall dry. The eastern model boundary is built by an extensive constant head boundary in layer1, the western outcrops receive low recharge.

The model is divided into a fictive upper system (layer1) and a lower system (layer 2 to 12). Layer 2 is a widespread aquitard which separates the lower system from the upper system. Layer 1 represents the upper aquifer system (simplification to allow downward leakage related to extensive pumping in the lower system).

The steady state calibration of the lower system is tedious but relatively normal. The calibration of the upper system is nearly impossible. Only when using relative low conductivities the iteration process starts. Otherwise I get the following errors:

GMG Solver:

“GMG Assembly Error in Subroutine GMG1AP”

SAMG Solver

*** ERROR: Allocation failed for <ifg,ipatt,nflag_thread> in (pwint2_aa)

*** ERROR: zero diagonal element in (myILUfct)!

*** ERROR: (M)ILU(0) decomposition failed: smoothing (ilux_prepare)!

>>>>> SAMG returned with error code 230 <<<<<


*** ERROR: SAMG re-start requires an initial call with iswit=4!

*** ERROR: Allocation failed in (cgstab)

>>>>> SAMG returned with error code 70 <<<<<


I do not really understand the problem (I know the different damping methods). The sensitivity is extreme. Only slight changes in hydraulic conductivity (e.g. 1.0e-6 m/s to 1.3e-6 m/s) cause the mentioned errors. When I get a nearly adequate solution and turn on wells, drains (every stress which removes water)…the same errors.

Maybe, too much cells fall dry within one outer iteration and/or isolated cells causes high residual changes between iterations. Why so, I have widespread artesian pressure and enough space to reduce the heads until the system is running empty.

Any suggestions?! Is the model to big ~5mil. cells and arrays to small?

Best Regards,


Link to comment
Share on other sites

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.


  • Create New...