HB Thermal Load Balance - face energy flow unexpected behavior

Hi all,

I think I’ve found some unusual behavior in the HB Thermal Load Balance component.

Oberservation
The storage term is unusually large, even when plugging in everything correctly:

What I’ve noticed is that the face energy flow term that is supposed to be plugged into the Thermal Load Balance is unusually large for the outdoor façade:

It suggests that there’s a 2000+ W energy gain into the zone just through my 10,5 m² outdoor façade; roughly 200 W/m² (!). Whereas the ΔT between the indoor surface temperature and zone air temperature doesn’t exceed 2.4 ºC… This seems impossible to me.

Possible explanation
Now, I think I’ve found the reason as to why it reports these large numbers. The outdoor surface conducts heat to both the outside and inside. The outside being much larger because of, I assume, the larger ΔT between outside surface temperature and the outdoor air temperature because of the sun shining on the surface.
HB Read Face Result gets the Surface Average Face Conduction Heat Transfer Energy for each face, which is the average between the conduction to the outdoors and conduction to the indoors. This means that even though most of the energy is “lost” to the outdoors, it’s still counted in the energy balance.
I’ve added a custom simulation output to the E+ simulation: Surface Inside Face Conduction Heat Transfer Energy, which reports much more reasonable values with regards to the energy gain:

For good measure I’ve done the same for the outside to check if indeed that heat transfer is so large, and indeed it is:

When I plug in the Surface Inside Face Conduction Heat Transfer Energy into face_energy_flow, I’m getting a much more sensible Storage component:


The storage jumps at 7AM and 10PM are due to changing internal loads.

Conclusion
In conclusion, I think the face_energy_flow input of the HB Thermal Load Balance component is supposed to just have the Surface Inside Face Conduction Heat Transfer Energy instead of the average, so that the energy lost to the outdoors is not accounted for in the balance nor storage.
But if I’m mistaken I would greatly appreciate to hear the explanation as to why the surface flow / storage is high.
Only thing I’m not sure about is that now only opaque surfaces are considered in face_energy_flow, I couldn’t find a Surface Inside Face Conduction Heat Transfer Energy equivalent for window surfaces. So if anyone knows how to include that if necessary, please let me know!

Thank you!

1 Like

Hey @marctavenier ,

I appreciate the thorough analysis and I agree that your explanations are correct, though I would also consider the current behavior of the load balance correct. It’s all just a matter of where you want to draw the thermodynamic boundary of “the building.” Specifically, should it include the mass of the exterior constructions as part of the building or should it exclude them and only include the mass of interior constructions and the room air?

A long time ago, I made the decision to include the mass of the exterior constructions and I feel that changing this now could create some confusion. But, if enough people say that they prefer your definition of “the building” boundary, I am open to changing it. Your proposed methods are certainly a valid interpretation of the first law of thermodynamics and they bring to light the fact that “the closed system” mentioned in this law is arbitrary and can effectively be whatever you want.

2 Likes

Really interesting discussion.

At our company IES VE is the standard simulation software. That defines external conduction gain as:

External conduction gain: Heat conducted into (or if negative, out of) the room through the internal surfaces of externally exposed elements, including ground floors.

https://help.iesve.com/ve2021/room_variables.htm?ms=AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA%3D%3D&q=RXh0ZXJuYWwgY29uZHVjdGlvbg%3D%3D&st=Mg%3D%3D&sct=MA%3D%3D&mw=MzUw

It appears Design Builder also uses the inside face, which is my go to for E+ comparisons with HB.

https://designbuilder.co.uk/helpv2/Content/Calculation_of_DesignBuilder_Output_from_EnergyPlus_Report_Variables.htm

That pushes me in favour of changing HBs definition, but honestly as my HB experience is still fairly limited don’t put too much weight on this.

I think in the future I would use Marc’s variable request so that my HB results align with other software.

1 Like

That said, this conversation has enlightened me on where external surfaces thermal mass falls in the energy balance - so thanks both!

1 Like

Hi @chris,

Thank you very much for your reply. I understand your reasoning behind it and it makes sense. Though I would argue that given we’re making a thermal balance of a specific zone, and whatever happens outside of that zone is out of scope for the thermal balance of that zone.
This way, the storage remainder term suggests that all that energy is stored in the building’s mass, whereas a significant amount is actually transferred to the outdoors and not stored in the building mass, right?

Given that for this project I would like to exclude the transfer of energy to the outdoors. Is this approach with using the inside face heat transfer correct? Especially regarding my uncertainty about the Surface Inside Face Conduction Heat Transfer Energy equivalent for window surfaces since I don’t know which E+ output I should use for that if at all.

Hey Marc,

On the window conduction Design Builder calculates it from these E+ variables:

Glazing heat gain (Design Builder) = Window Heat Gain [W] - Window Heat Loss [W] - Window Transmitted Solar [W]

Check out the link in my comment above for more details.

1 Like

Hi Charlie,

Thank you, that seems to give valid results :slight_smile:. For reference, here’s the storage remainder:

Alright, you won me over. I realize that what we’re talking about here has no bearing on the monthly load balances that people typically create with Ladybug Tools since any heat that enters/leaves the mass of the exterior envelope within a given day is going to cancel out over a time period as long as a month.

But I can see how your suggestion makes single-day visualizations of the load a lot cleaner. For example, looking at the load profile over the summer design day, this is what we would have had previously when we counted the exterior constructions as part of the building:

… and this is what we get if we discount the mass of the exterior constructions and only look at the heat that makes it to the interior:

This also makes monthly-per-hour visualizations like the following more understandable:

I just merged the change is here and you can get it with the LB Versioner in an hour:

Thanks again for the suggestion.

2 Likes

That’s great Chris, thank you for the fix!