Question-and-Answer Resource for the Building Energy Modeling Community
 | 1 | initial version |
I have an update after some more tests. I think I have narrowed down the issue to the fact that the Surface Window Net Heat Transfer Energy output does not include a full balance of energy transfer across the windows of the Zone and there is a significant amount of energy transfer across the window to the outside of the zone that is excluded in this output. First, I can say pretty confidently that it's not a fluke of my particular model since I get these large remainders using a wide range of different climate files and they persist whether the building is conditioned or unconditioned. As long as the model is very glassy, I cannot get the balance to sum to zero and there is consistently a net positive load balance.
Insufficiencies in the Surface Window Net Heat Transfer Energy output would explain why, when there are no windows in the Zone and I simply sum the Surface Inside Face Conduction Heat Transfer Energy output, I get something close to zero (since there are no values for the Surface Window Net Heat Transfer Energy in this case). However, when I do the inverse of this situation and I set the floors and roofs to be adiabatic and ensure all of the walls have 100% glazing, summing all of the Surface Window Net Heat Transfer Energy outputs does not give me zero. In fact, I get monthly totals of kWh that are even worse than the 90% glazed case that I looked at earlier:
This is what the load balance looks like for this case:
Where the storage term is, again, this remainder that is trying to account for the fact that Surface Window Net Heat Transfer Energy does not give us the full heat transfer across the window.
Given the way that this Window Net Transfer output is described in the E+ docs, I'm inclined to think that there is either a bug in EnergyPlus or the docs are wrong and one of these terms is not accounted for in the calculation of Surface Window Net Heat Transfer Energy:
I should also clarify that there is no frame and divider in any of these tests that I am running and the window is not shaded. So Just the first 4 terms should describe the full full load balance of my most recent test.
It's also possible that there is still some term other than those accounted for in Surface Window Net Heat Transfer Energy, which I have just been unable to think of and could correct the sum of everything back to zero. I have been racking my brains the last few days and I have been unable to come up with what I am missing.
If I don't hear a response here, I'll assume that everyone else is as stumped as I am and I'll open an issue on the EnergyPlus Github to either fix the E+ docs or fix what may be a bug related to this output.
