Re: [EnergyPlus_Support] Energy plus model calibration

Dear Heba,

Assuming the zone temperature sensor is accurate AND representative if the zone average temperature (that's what E+ assumes), then it can be used as an indicator of good calibration.
I think that it is not likely that you have defined the building materials and infiltration sufficiently well that you can expect E+ to match measured data for zone temperature. That would be highly unusual for a real, occupied building.
For that reason, "calibration" of non-conditioned zones is not a normal activity since the goal of "calibration" is to match actual and predicted energy use. Perhaps you are referring to a research project?
Actual / real installations almost never have the same air flow, water flow, motor power or COP as indicated on manufacturer data; they must be verified by measurement. Normally major energy components should be measured and perhaps a sampling of the small components.
Measurement limits the unknown variables such as infiltration to a smaller group so that calibration does not need to evaluate as many possibilities of variation. There are normally many possible combinations of variables which result in "calibration" that is within acceptable limits for indicators like RMSE. Your job is to find the most likely solution based on available information. Hopefully the solution can be tested against future performance in order to refine it.
Consider calibration as a challenge to your understanding of the building and its energy systems and also to your creativity in finding the best calibration solution.

On Tue, Oct 11, 2016 at 5:54 AM, hassan_kamel77@xxxxxxxxxxx [EnergyPlus_Support] <EnergyPlus_Support@xxxxxxxxxxxxxxx> wrote:

Â

Thanks Jim for your attached file, it helped me too...
I'm sure that was a great presentation (Y).
I'm calibrating a model with DesignBuilder. and have a question !!

If it is hybrid apartment, I calibrated the conditioned and unconditioned zones separately,

For the unconditioned zone, I compared the real measured temperature with the output attempts from DesignBuilder, after changing in the uncertainty variables, and made them as less as possible.
then performed the stats. tests like MBE, RMSE, and correlation coefficient, to choose the near-to-actual curve from the measured temperature.

For the conditioned zone, I made the same actually, to compare between temperatures.

My questions are:
- is that right?
- what's the difference between the calibration of a conditioned and non-conditioned zones?
- If I have the actual energy use of the air-condition of that zone, should I compare it with the cooling loads of the modeled one?

Thanks for concern, and if you could send me also some reading materials or URLs, I appreciate.

Heba

James V Dirkes II, PE, BEMP, LEED AP
CEO/President
The Building Performance Team Inc.
1631 Acacia Dr, GR, Mi 49504

Direct: 616.450.8653
jim@xxxxxxxxxxxxxxxxxxxxxxxxxxx

Website l Â LinkedInÂ | May 2016 ASHRAE Journal article on energy modeling

?an open mind is not an end in itself but rather is like an open mouth, designed to bite down on something solid and nourishing.