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

Calibrate E+ model with AFN based on measured temperature

asked 2020-04-20 16:53:07 -0500

rafael.alonso gravatar image

updated 2020-04-23 12:09:28 -0500

Hi all,

I´m looking for the best tool to calibrate an E+ thermal model based on measured hourly Air Temperature. I´m exploring the following variables: infiltration through an Airflow Network, Internal Heat Gains (lights, people) and, maybe, HVACIdealAirLoads system size.

There are a few posts around the forum, but most of them are several years old. Only one option seems current, the PAT tool from OpenStudio, but it only runs with an osm model and not idf (I need the idf to include the Airflow Network, not yet implemented in OS).

Does anyone know a tool to perform the calibration? Ideally, something that runs the idf, compares the hourly temperature for a given period to the measured data and then adjusts the model before running again.

Thanks in advance! Rafael

UPDATE We are discussing a potential solution in the comments of this other thread. I´ll share it back here if it succeeds.

UPDATE 2 As per the post above, the only solution so far is to write an E+ or OS measure to use in PAT. It´d be of great help if anyone knows an E+ measure related to AFN to use it as a base.

edit retag flag offensive close merge delete


Note that as of OpenStudio 2.5.0, AirflowNetwork has some support in the SDK. Quoting the release notes: "Preliminary support for the EnergyPlus AirflowNetwork feature is now provided in OpenStudio. Envelope leakage and core elements of distribution networks are supported. Changes are expected to support ongoing enhancements of the AirflowNetwork model (e.g. multiple air loops)."

shorowit gravatar image shorowit  ( 2020-04-20 18:54:23 -0500 )edit

Hi @shorowit, thanks for the tip. I´m not very familiar with the relation between OS SDK, OS and E+. Does that mean that I can now edit the AFN through an OS measure? Or how would you use that capability in the SDK?

rafael.alonso gravatar image rafael.alonso  ( 2020-04-21 03:57:43 -0500 )edit

That is correct, you could edit the AFN through an OS measure.

shorowit gravatar image shorowit  ( 2020-04-21 08:26:22 -0500 )edit

Ok, thanks! And do you know if such a measure already exists? I´ve never made a measure before and I couldn´t find any AFN related measures in BCL, maybe there are other websites with E+, OS measures?

rafael.alonso gravatar image rafael.alonso  ( 2020-04-21 08:48:02 -0500 )edit

1 Answer

Sort by » oldest newest most voted

answered 2020-04-24 04:53:42 -0500

aubr gravatar image

Hello, if you want to calibrate such a model by modifying the IDF, I would suggest using GenOpt and trying to minimize a discrepancy function of indoor temperature (e.g. mean squared error). The drawback is that you would have to write this function. But the solution would be quite flexible.

edit flag offensive delete link more


Thanks, I´d take a look into GenOpt. Do you know if it works with E+ 9.3?

rafael.alonso gravatar image rafael.alonso  ( 2020-04-28 11:58:10 -0500 )edit

Your Answer

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

Add Answer


Question Tools



Asked: 2020-04-20 16:53:07 -0500

Seen: 526 times

Last updated: Apr 24 '20