Dear group,
In our current projects we noticed a big difference when modeling solar
gains between the software IES VE and Designbuilder (EnergyPlus). IES VE
underestimate the values of solar gains, by almost 1 to 2 of difference
than EnergyPlus,
So I just try to ask if someone of you are experienced the same
difference and if you know the reasons of that big difference. We test a
"cell-test" and we found also this underestimation of solar gains.
Thanks in advance,
David Garcia Sanchez
David,
One thing to check is the definition of solar gain in the 2 programs. For example, in DesignBuilder EnergyPlus the solar gain reported is the gross solar gain, including any solar heat that is subsequently re-reflected back out of the windows. I don't know the definition of solar gains in IES VE, but if it displays the net solar gain then this might explain the difference, especially in zones with large areas of glazing where re-reflection is more significant.
Regards,
Andy Tindale
David,
I suspect that you are not comparing apples to apples. Intermodel comparisons are generally difficult to complete without a thorough understanding of the inputs and algorithms used in the tools. Its very easy to think you have equivalency in your input models when you do not. You have the added complication in your case that the DesignBuilder interface adds another layer of translation to the problem.
Also note that both simulation engines (IES VE and EnergyPlus) have been through the BESTEST/ASHRAE 140 process as well as many other validation exercises. This should mean that given the same input the output will be similar (within some bound due to different algorithms and implementations).
I suggest that you make a careful comparison of all the inputs and understand how they impact the simulation, then do the same with the output variables that you are comparing.
Regards,
Iain
David,
The attached research paper may help you on this.
Thanks.
That wasn't my question, but the topic is interesting to follow. Thanks
Javed! for the paper.
Vaibhav Jain (Veb)
David,
I would caution you not to be so quick to use the term "underestimate," which implies that you are certain that your test is set up correctly and that know the EnergyPlus calcs to be accurate. Furthermore, given the magnitude of the disparity you've described-which does not show up either in the published ASHRAE 140 tests comparing results for the VE and EnergyPlus or in the SimBuild paper from Cassie Waddell and Shruti Kaserekar that Javed Iqbal attached to his reply-there may be something significantly amiss in your comparison.
You need to ensure that you are consistent with your glazing properties and you may want to be running the "detailed solar calculations" in EnergyPlus and, likewise, linking the simulation to results from the SunCast module within the VE if you want a suitably consistent basis for the comparison. This will facilitate calculation of incident solar radiation on specific surfaces, etc.
While I'm not sure how this works in EnergyPlus, when you run simulations coupled to the SunCast module in the VE, you need to be cognizant of the relative placement of glazing in your "test cell," as low-angle direct-beam radiation striking an east window in the early morning, for example, can pass through the space and out the other side, in keeping with the location and solar transmission of glazing on the fa?ades through which the solar gain is entering and exiting.
On a related note, whereas the paper from Waddell and Kaserekar attributes the calculation of interior inter-surface radiant exchanges to the SunCast module of the VE, this is actually a function of the ApacheSim engine used for all dynamic thermal simulations in the VE. SunCast determines which surface the solar gain will fall upon and what fraction of each surface is shaded from direct-beam radiation at any given time step, as well as how much of those surfaces are shaded from the cooling effects of the night sky; however, regardless of SunCast, the hot interior surface of a piece of glass having absorbed solar gain, for example, will always exchange radiant energy with any interior surfaces of lower temperature during the simulation run. Conversely, the cold interior surface of a window or exterior wall in winter will always act as a sink for radiant energy from warmer interior surfaces.
Finally, if you're interested in digging deeper into the comparison, I was told by researchers at the Technical University at Eindhoven in The Netherlands attending the last SimBuild conference that their current findings (yet to be published) show that the solar calculations in the VE are actually somewhat more accurate than those in other tools, including EnergyPlus and ESPr, particularly when passing through multiple glazed surfaces, cavities, or thermal zones. As it's not yet published and IES was not involved in this research, I'll leave it to TU Eindhoven to elaborate on what they tested, what they found, and whether or not they believe this to be of any significance. They may be willing to share some results in advance of publication.
Regards,
Timothy Moore