[TRNSYS-users] Convergence problems

David BRADLEY d.bradley at tess-inc.com
Mon Dec 5 07:23:01 PST 2022


Julien,

   Convergence problems are often the result of not having enough 
thermal capacitance in a loop. For instance if you have a massless solar 
collector, a pump and a constant effectiveness heat exchanger  all in a 
loop and have neglected the pipes then there are no components in the 
loop that have thermal capacitance and the loop can change temperatures 
very very fast. You don't necessarily need to go through the whole 
system and add in every cm of piping but look at the warnings to find 
specifically what components are not converging and then look at those 
loops to see if you can add piping or a buffer tank to the loop to make 
it numerically more stable.

   You can also drop to a 1 minute time step as part of your solution.

   If you do go the route of using a time delay on inputs I would 
suggest that you delay the values that your controllers are watching. In 
my example above, for instance, you might delay the outlet temperature 
of the collector that the controller is watching. I would not advise 
delaying the temperature of the collector that is being passed to the 
next component in the loop.

kind regards,

  David


On 12/5/2022 7:23 AM, Julien Gambade via TRNSYS-users wrote:
> Dear Trnsys users,
>
> I have some problem with convergence issues. I am working with a 
> simulation with a duration of several months and a timestep of 5 min.
>
> For several time steps, some inputs do not converge and after a 
> certain time, the amount of warnings exceeds the number of warnings 
> before error.
>
> Two options allowed me to end the simulation : increase the number of 
> warnings allowed and use Type "delayed inputs" (last option results in 
> a change in my outputs).
>
> Do you think the first solution is fine or should I go to the second 
> solution ? Or, does anyone know an other solution I could try ?
>
> Best regards,
> -- 
> *Julien GAMBADE*
> *Doctorant*
> Institut de Recherche Dupuy de Lôme (UMR CNRS 6027)
> Centre de recherche Christian Huygens
> Rue de Saint-Maudé - 56100 Lorient - France
> julien.gambade at univ-ubs.fr
> www.irdl.fr <http://www.irdl.fr>
>
> _______________________________________________
> TRNSYS-users mailing list
> TRNSYS-users at lists.onebuilding.org
> http://lists.onebuilding.org/listinfo.cgi/trnsys-users-onebuilding.org

-- 
***************************
David BRADLEY
Principal
Thermal Energy System Specialists, LLC
3 North Pinckney Street - suite 202
Madison, WI  53703 USA

P:+1.608.274.2577
d.bradley at tess-inc.com

http://www.tess-inc.com
http://www.trnsys.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/trnsys-users-onebuilding.org/attachments/20221205/4494af2b/attachment.htm>


More information about the TRNSYS-users mailing list