open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear All
I am trying to couple Telemac2d with WAQTEL. My water flow simulation with Tel2d had 7000000 time steps and took 13 hours on HPC. Now I have coupled with WAQTEL and the runtime is more than 3 days(Parallel Processors=576) . I have the limitation of 3 days in parallel mode. Can any one help me to decrease the runtime?
If I increase the time interval and decrease the number of time steps the currant number highly increases and I will receive out range outputs.
Regards

Hello Arman,

You can split your computation into several runs lasting less than 3 days CPU time (see keywords COMPUTATION CONTINUED = YES + PREVIOUS COMPUTATION FILE in the TELEMAC-2D manuals).

You can also think of the extent of your computational domain. Do you need it so large or can it be reduced? You can also use a coarser grid, in particular where you do not need such refinement.
These ideas may be means to save elements and reduce your CPU time.

Looking at your steering files, I have a few suggestions:
- if the maritime boundaries are far from the insteresting area you study, you can use classical conditions (OPTION FOR LIQUID BOUNDARIES = 1) + only prescribe water depth (code 5 4 4 5 in the boundary conditions file),
- do you need to write every variable in the results + as often as every hour?
- if O2 or BIOMASS module are enough for WAQTEL to model water quality, as they solve less advection-diffusion equations for tracers, it may be tried,
- you do not seem to have dry areas in your model as you use TIDAL FLATS = NO and use the default values for the advection scheme for tracers (= 1 = characteristics). Do not forget that characteristics are not conservative and you may face issues. As you do not have dry areas, prior to run a 1 year simulation, I would suggest you to try LIPS schemes for tracers as they may be efficient (and conservative), in particular with wet elements: try SCHEME FOR ADVECTION OF TRACERS = 5;...;5 (as many as the number of tracers) and SCHEME OPTION FOR ADVECTION OF TRACERS = 4.

Hope this helps,

Chi-Tuan

Dear Chi Tuan
I used TDAL FLATS=YES
Then I received negative depths. I think it is better to split to several runs.
Regards

Hello Arman,

If you do not want to have negative depths, you should use option TREATMENT OF NEGATIVE DEPTHS = 2 or 3 (which are no default values).
In that case, you also have to use the combo of options:
- MASS-LUMPING ON H = 1.,
- CONTINUITY CORRECTION = YES,
- SUPG OPTION = ...;0;... (i.e. no SUPG upwinding on depth).

Please read the TELEMAC-2D user manual, in particular subsection for tidal flats (7.5).

Hope this helps,

Chi-Tuan

Dear Chi Tuan
Thanks for your help. Ill go through it to see what happen.
Kind Regards

Dear Chi Tuan
I was looking in euro example on svn. The water process=3 is selected in that example while it should be =5 as it is mentioned in waqtel manual. I used =5 and I receive some errors of eutro steering file. what do you think about that? Which one I should use and if I want to use eutro with water process=5 then is there any convenient example for steering file?
Regards

Hello Arman,

What do you mean by "I was looking in euro example on svn"? What release are you looking at?
Since release v8p0, the numbering of WAQTEL processes has changed and now, to use EUTRO module, you have to set WATER QUALITY PROCESS = 5 (or a multiple of 5). 3 was the old choice for EUTRO but it is not good now.

As told in many locations on this forum, please use the latest release available, now v8p3r0, see the announcement:
opentelemac.org/index.php/latest-news-de...mac-v8p3r0-available
You have to switch from svn to git now (see the installation instructions).

What do you mean by " I used =5 and I receive some errors of eutro steering file." Can you give these errors please if you want any help? Do not forget to read the forum rules available here:
opentelemac.org/index.php/kunena/help
in particular #7

So please prior to other posts, read the WAQTEL user manual, WAQTEL validation manual + example of the latest release available, i.e. v8p3r0:
wiki.opentelemac.org/doku.php?id=documentation_latest

Hope this helps,

Chi-Tuan

Dear Chi Tuan
The other issue with waqtel is the prescribed tracers values order. In the mannual it is mentioned by boundary order and in the site it is mentioned by tracer order as you can see in the attachments. On the other hand we input 8 tracers when we use eutro by telemac and we receive 5 output tracers. I dont know how I can enter the amount of each tracer in each boundary. Is there any example in this case?
Regards
Arman

Hello again Arman,

What do you mean by "site"?
If it is the webpage:
wiki.opentelemac.org/doku.php?id=user_manual_telemac-2d
which comes from
wiki.opentelemac.org/doku.php?id=user_manuals
which comes from
wiki.opentelemac.org/doku.php
it is written that it is deprecated (and may have to be deleted).
Indeed, there are corrections of mistakes in the manuals and you should read the latest releases, which are available with the latest release of the TELEMAC-MASCARET system, i.e. v8p3r0 as told in my previous post.
A hint is that the last update of this webpages was 7 years ago...

Anyway, the answer of your questions (in particular your last one) is in the user manuals of v8p3r0 (in the TELEMAC-2D one).

Hope this helps,

Chi-Tuan

Dear Chi Tuan
Sorry for the little information about my model.
I am using Telemac2d V8P2 couple with WAQTEL. I tried to go through gitlab as you mentioned and used urtro steering (WAQTEL example) file with WATER QUALITY PROCESS = 5. I have attached my steering file and uetro_steer. Unfortunately I receive the error of 'exceeding maximum iterations'.
The error is as attachment. I think something is wrong with my eutro steering file. Can you please let me know if there is a convenient example or guide me what should I do.
Thanks and Regards
Arman