Model converging in steady state but not in transient

[Francesca Moschini]

Hello there!

I'm experiencing a problem with my model in Modflow NWT. I have converge in the steady state, but when I use it in transient the model does not converge anymore. The .lst file gives me many "Non-convergence in ROUTE_CHAN". I don't know if this is the problem or there are other parameters that I should check.

Any advice about how to solve the problem?

[Ashutosh Singh]

Try selecting initial heads as solution of steady state model and increasing the number of iterations for convergence!

Hpc

Sent from my iPhone

--
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 post to this group, send email to mod...@googlegroups.com.
Visit this group at http://groups.google.com/group/modflow.
For more options, visit https://groups.google.com/d/optout.

[Francesca Moschini]

Thank you for your answer.

I'm using Model Muse and I can't find the option to select the initial heads as solution for the steady state model!

What I'have noticed as well is that during my steady state I don't have any interaction with my SFR package, the only input is the recharge and the only output is from my constant heads. When the transient starts I start having input and output from SFR as well (surface and stream leakage). I have also noticed that for steady and transient I have this message in my listing "SECANT METHOD FAILED TO FIND SOLUTION FOR STREAM SEGMENT " for many of my rivers!

[ashutosh singh]

Start from initial heads mean that initially you can define a stress
period from -1 to 0 (hour or day) and mark it as steady state. From 0
onwards the initial head will be selected from the previous solution
i.e. steady state. Use the same boundary coditions for simulating
initial heads which means that you should keep the SFR in the steady
state model as well.

My advice is that try both steady state and transient state in the
first step. Then move to introducing SFR in steady state and transient
state model.
Try to read more about SFR package and reasons why the solver may not converge.

hpc

[rbwi...@mindspring.com]

If the steady-state model is a separate model from the transient model, you can select the head file generated by the steady-state model in the "Model|MODFLOW Options" dialog box in the "Options" tab. However, normally, you would just make the first stress period a steady-state stress period and subsequent stress periods transient stress periods. You do that in the "Model|MODFLOW Time" dialog box.

-----Original Message-----
>From: Francesca Moschini <fr.mo...@gmail.com>
>Sent: Nov 21, 2014 7:23 AM
>To: mod...@googlegroups.com
>Subject: Re: [MODFLOW] Model converging in steady state but not in transient
>
>Thank you for your answer.
>I'm using Model Muse and I can't find the option to select the initial
>heads as solution for the steady state model!
>What I'have noticed as well is that during my steady state I don't have any
>interaction with my SFR package, the only input is the recharge and the
>only output is from my constant heads. When the transient starts I start
>having input and output from SFR as well (surface and stream leakage). I
>have also noticed that for steady and transient I have this message in my
>listing "SECANT METHOD FAILED TO FIND SOLUTION FOR STREAM SEGMENT " for
>many of my rivers!
>
>Il giorno giovedì 20 novembre 2014 17:07:39 UTC+1, hpc ha scritto:
>>
>> Try selecting initial heads as solution of steady state model and
>> increasing the number of iterations for convergence!
>>
>> Hpc
>>
>> Sent from my iPhone
>>
>> On 20 nov 2014, at 15:26, Francesca Moschini <fr.mo...@gmail.com

>> <javascript:>> wrote:
>>
>> Hello there!
>> I'm experiencing a problem with my model in Modflow NWT. I have converge
>> in the steady state, but when I use it in transient the model does not
>> converge anymore. The .lst file gives me many "Non-convergence in
>> ROUTE_CHAN". I don't know if this is the problem or there are other
>> parameters that I should check.
>> Any advice about how to solve the problem?
>>
>> --
>> 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 <javascript:>.
>> To post to this group, send email to mod...@googlegroups.com <javascript:>

[Francesca Moschini]

Now I see what you mean! My model was already the way you described it. steady state -1,0 and then transient 0-30.
I managed to make my model converge by changing the  Brooks-Corey epsilon. by doing that everything seems to work (for now at least!).
Thank you for your comments Richard and hcp.