Warming up days - open studio/energy plus

[Aimee]

Hi,

I’m trying to run an energy simulation in a cold climate and the default number of warm up days (25) is not enough. I am using the simulation control component to increase this, (see the old thread below) but when I go into open studio it still ends at 25.

If anyone has any ideas for how to get around this that would be great.

Many thanks,

Aimee

The old thread with the component and file I used:

Warming up period

Hi, I am using EnergyPlus on HoneyBee. I get question about analysis period. Can I make analysis period more than one year or repeat simulation to make correct initial setting? It needs warming up period when I run thermal simulation to get correct output. If we run simulation without warming up, results in former period is incorrect because initial thermal situation is determined by program.

[chris]

Hi @Aimee ,

Welcome to the forum! It seems the reason why it does not work is because the Building object in OpenStudio SDK does not have the maximum number of warmup days exposed on it:
https://openstudio-sdk-documentation.s3.amazonaws.com/cpp/OpenStudio-2.8.0-doc/model/html/classopenstudio_1_1model_1_1_building.html

You can see that it is on the equivalent EnergyPlus object here: https://bigladdersoftware.com/epx/docs/9-0/input-output-reference/group-simulation-parameters.html#field-maximum-number-of-warmup-days

We are currently working with the OpenStudio development team and I will ask about exposing this property the next time that we talk. In the meantime, I’m sorry that I can’t offer anything better than the following options:

1. Use the direct E+ exporter in Honeybee (and sacrifice detailed HVAC systems)
2. Use an EnergyPlus measure to automatically edit this property after export from OpenStudio
3. Manually edit the number of warmup days in the IDF and use the ReRun IDF component to run it through E+ and bring the results back into Grasshopper

Also, I don’t see a warning about convergence in your screenshot. Is there something that I am missing?

[Aimee]

Hi Chris,

Thank you, I’ve managed to update the warm up days but unfortunately it does not converge even after 500 days! So I think I have a much larger error elsewhere. I will go back through the rest of the script and try to identify it

Yes you’re right the screenshot doesn’t show a warning. The script I was using wasn’t all my work so I screenshotted a sample to show the component I was having trouble with.

Thanks again,

Aimee

[chris]

@Aimee ,

I should also mention that, sometimes, the lack of convergence isn’t necessarily erroneous or incorrect if the type of things going on in the building are intentionally dynamic. For example, I know some of simulations that I have run with natural ventilation don’t converge but this is because the windows keep opening and closing in response to the zone temperature. In this case, I tend to just note the warning but proceed with analyzing the results.

[minggangyin]

@Aimee

Thank you, I’ve managed to update the warm up days but unfortunately it does not converge even after 500 days! So I think I have a much larger error elsewhere. I will go back through the rest of the script and try to identify it

Hi,@Aimee
I suggest you upload your grasshooper definition for further discussion.

[chris]

I have a correction to make to my previous comment. The WarmUp Days are exposed in OpenStudio SDK. They are just on the SimulationControl object rather than the Building object, which I have to say makes a lot of sense.

So I will add the ability to set the number of Warmup days on the OpenStudio component now.

[chris]

This has been fixed:

fix(simulationcontrol): Adding Support for WarmUp Days in OpenStudio Component

You can now set the WarmUp day limits on the OpenStudio component.

[MichaelDonn]

I have v 1.9.25 of LBT.

I cannot see how to vary the warmup days as an input to the Simulation Control component. And I cannot find where else it may have gone in the system.

[chris]

Hey @MichaelDonn ,

We did not expose warmup days on any LBT component because the E+ warning you are getting, which is driving you to ask for this feature, is misleading.

If your simulation is not converging after the 25 warmup days that we use in Honeybee, then it is almost certainly not going to converge after more warmup days. Maybe if we used a default of 5 warmup days (or something like that), the E+ warning would have actual meaning to us.

But I can confidently tell you that the reason why your simulation is not converging after 25 warmup days does not have to do with too few days. Probably, there are a set of unstable controls in the model (like maybe some natural ventilation controls, which are causing the windows to open and close in rapid succession). Unstable controls like this can prevent the simulation from ever reaching convergence and you may need to tweak them.

I have gotten this question enough that maybe I should just overwrite the warning that E+ gives in these cases. In any event, if you want help identifying which controls are causing the erratic behavior, @MichaelDonn, please open a separate issue.