Question-and-Answer Resource for the Building Energy Modeling Community
Get started with the Help page
Ask Your Question

VAV System Not Converging

asked 2014-12-15 12:50:06 -0600

jdunn gravatar image

updated 2015-07-27 11:02:46 -0600

Simulation Community -

I'm getting the following error when running my central VAV template system when running for only January:

* Warning * SimHVAC: Maximum iterations (20) exceeded for all HVAC loops, at RUN PERIOD 1, 01/04 10:00 - 11:00 * ~~~ * The solution for one or more of the Air Loop HVAC systems did not appear to converge

I'm modeling for LEED, and I'm getting this warning as ONLY when I hardsize the supply fan maximum flow rate in the hvactemplate:system:vav object according to the mechanical schedules for the project.

The designed flow rates and capacities are about 2x the autosized rates, not sure if that would cause a problem. Here's a table that shows the fan and cooling coil sizes for both the autosizing baseline and the proposed case.

image description

When using the output:diagnostics object to display extra warnings, it says that this particular air hander does not converge for temperature, energy, or mass flow rate. I've doubled the Maximum Hvac Iterations field in the ConvergenceLimits object to 50, but to no avail. Not sure if I need to adjust the temperature tolerance values in the building object.

Any idea why this might be happening? Also Is this something that LEED is going to have an issue with? This causes a lot of warnings given it happens on during all times in the run period.

Here is a link to download the full error file -

edit retag flag offensive close merge delete

2 Answers

Sort by » oldest newest most voted

answered 2014-12-16 12:01:42 -0600

Archmage gravatar image

The usual reason that a larger size air handler will tend to show more of the dreaded SimHVAC: Maximum Iterations warnings is because with the larger system the terminal units spend more time running at the minimum stop on the VAV damper. A common situation for getting these warnings is during times of low load when loads are just starting to pick up and the dampers begin to ramp up from the minimum stop. If you can diagnose that these errors are occuring during such times, by studying detailed frequency time series results, then I believe it is safe to ignore them. What usually happens is that when the damper varies, both the flow and supply air temperature are changing at the same time and the air system solver doesn't settle into one solution. The supply deck temperature changes because of economizer action as the ratio of outdoor to return air changes with variations in overall flow. This has been a long standing problem with no easy solution.

edit flag offensive delete link more

answered 2014-12-16 04:55:41 -0600

updated 2014-12-16 04:59:04 -0600

LEED asks you how many warnings you have, and you'll have to provide a supplemental narrative describing the warning messages and why they do not adversely affect the results of the energy model.

Without looking at your idf file, I'm not sure what's causing this specific error. Nevertheless, a few things to mention:

  • I wouldn't necessarily be worried if it doesn't converge for only one timestep.
  • I would try running the simulation with a smaller time step, say 6 per hour instead of 1. If your model is complex and you're afraid of computing time, try for just January
  • It seems that while the supply inlet temperatures are indeed oscillating, they are not wild. Here's one extract:

    * ~~~ * Node named VAV2-37 SUPPLY INLET shows oscillating temperatures across iterations with a repeated value of 15.419107

    * ~~~ * Node named VAV2-37 SUPPLY INLET temperature [C] iteration history trace (most recent first): 15.419107,15.416305,15.426937,15.419107,15.416305,15.426937,15.419107,15.416305,15.426937,15.419107,

The recent history shows that it stays very close.

That being said, did you mean to have some apparently in cooling mode and others in heating mode? (You have 24 VAVs delivering air below 13°C/55°F, 21 VAVs delivering over 24/75°F, and 6 VAVs around 16°F/61°F)

  • Something useful is to request an output at "Detailed" (= HVAC variable timesteps) reporting frequency of the supply inlet temperatures, and then plot it to see what's going on (I suggest using SQLite output and using the ResultsViewer from Openstudio)

  • Check what convergence tolerances you're using for your VAVs (I think default is 0.001°C)

  • If after all of the above you're not satisfied, try the helpdesk.

edit flag offensive delete link more


Thanks for the information and reply!

Some notes:

  • This happens for multiple timesteps when running for the entire year, about 8.
  • The cooling during the wintertime is due to the large occupancy of the classrooms, and the relatively mild winters of New Orleans.
  • I tried running the simulation on a 6-timestep, but the converge errors still happen.
  • I'll definitely report and check out the supply inlet temperatures, and see if there is anything funky.
  • Given that temperatures are just oscillating, as you mentioned, I might be able to make the case that the warning is innocuous.
jdunn gravatar image jdunn  ( 2014-12-16 18:14:28 -0600 )edit

Especially for 8 warnings only. If it was repeating thousands of times it'd be harder to explain, though doable... If you can prove that nothing delirious is happening (no real change in energy use), you'll be just fine for LEED. Be proactive and send them a narrative explaining that on first submittal. Check your convergence tolerances for the VAVs too. (Do a search in notepad++ for example, you'll see convergences are used on different levels. But what matters is probably the one attached the the VAV object itself). Try increasing it a little (0.005°C? 0.01°C?) to see what happens.

Julien Marrec gravatar image Julien Marrec  ( 2014-12-17 05:30:12 -0600 )edit

Just so we have a trace here, Jean made some good points on E+ support: "Why are the proposed capacities double those of the baseline? This would make me question the baseline autosizing, i.e. how your loads are timed with the schedules for day types "SummerDesignDay" and "WinterDesignDay" and if you are indeed setting the minimum ODA rates equal in both models."

Julien Marrec gravatar image Julien Marrec  ( 2014-12-17 05:31:36 -0600 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Training Workshops

Question Tools



Asked: 2014-12-15 12:50:06 -0600

Seen: 697 times

Last updated: Dec 16 '14