I’ve been playing around with the Heating demand simulation, as I will be testing different proposals for an urban plan for energy-releated properties. My test consist of few single-family houses and two larger buildings. The value that I am looking for is the kWh/m2a for heating, to test which building forms have better performance potential.
The first value next to a building is the annual heating demand, the second is the heating demand / brutto area. The construction are set to a passive standard, my problem is the two larger buildings. How come their energy demand is just as high as of a typical single-house? That leads then to a crazy low energy demand / area value.
Because this concerns energy modelling in an urban scale, I figured modelling the HBzones quick & dirty- meaning 1 HBzone = 1 floor of a building. Could that be an issue?
In the 2nd picture as I was looking for causes, I found a zone that seem to have suspicously low heating demand - although I am aware of the probable heat transfer from the lower floors But the numbers are quite low for the lower floor too.
Maybe asking the very basic questiontoo helps: The heating values coming from the Energy+ simulation component are purely the kWh’s needed to keep the zone warm for the set period, right? It is not automatically converting the value into (E)/m2 or somehow taking into account the area of the HBzones, or?
(I would love to share my GH definition, but the forum says new users can’t upload attachments?)
Yes, something looks wrong with the way this is being mapped to the buildings, if we’re assuming all other energy parameters are equivalent. I think something is wrong in the post-processing of the results and zone area. Are you running all these models at the same time? As a quick gut check, I would isolate one of your large buildings, and run the simulation on just those zones, to see if you’re still getting the same kWh. I expect your larger buildings to be more efficient (per unit area), but not this much more efficient.
Because this concerns energy modelling in an urban scale, I figured modelling the HBzones quick & dirty- meaning 1 HBzone = 1 floor of a building. Could that be an issue?
Your energy demand will likely be a lower then the actual building because you’re essentially averaging out extreme conditions by not splitting your floors into thermal zones – but this shouldn’t effect the kWh/m2 error you’re having.
Maybe asking the very basic questiontoo helps: The heating values coming from the Energy+ simulation component are purely the kWh’s needed to keep the zone warm for the set period, right? It is not automatically converting the value into (E)/m2 or somehow taking into account the area of the HBzones, or?
Yes that’s right. You can use the Normalize Data by Floor Area to divide your energy by area. If you aren’t using this component (doesn’t look like it based on your screenshots) I think that’s what’s causing the problem. I wouldn’t trust the order of the energy values that come out of the GH components to necessarily sync with the order of the zones.
thank you for you reply. I checked multiple times everything that has to do with the orders of zones, everything is fine. I also tried the Normalize Data by Floor Area component, it gives the same values.
Here I run the simulation just for the 2 types of buildings, same extreme result. You can see the energy demands for each zone (yellow txt) floor area (green txt), name (black txt) + on the GH definition you can see it’s the numbers straight from the e+ component, so no shuffling of zones there. (These values are the same if I run simulation only with solely with the large building.)
As another test I tried the same simulation with no changes in Construction - taking default Honeybee constructions for MidRiseApartment zone program, I guess. Then my end energy values seem quite logical: Tiny house - 58.88 kWh/m2a and the Apt. house with 40.8 kWh/m2a.
The constructions are layers that I took from real passive house examples like triple glazed windows with U=0,8 W/m2-K, 40cm insulation walls. So my guess is these values are simply too optistimic for a larger apartment house? Maybe they don’t calculate well will such large Zone as a whole floor and I should try splitting the mass at least 1 HBzone / apartment.
In that case, I think it might be the lack of thermal zone splitting. In retrospect, I realize that not splitting thermal zones actually seems to create differences in the modeling assumptions between the two shapes. Specifically, your small buildings have a small enough footprint that the difference between its EUI and a more realistic thermally split zone won’t be that large (relatively). But for the larger building, there’s a much larger difference between its EUI and a more realistic thermally split zone. So you’re actually modeling very different interior conditions.
From that perspective, splitting the large building into thermal zones makes the assumptions about interior conditions closer to the small buildings, which should give us a better apples to apples comparison.
Other thoughts/questions if that doesn’t work:
Is the WWR in both shapes the same?
I also suggest running a heat balance with a representative small and large building to get an idea of where the differences lie.
There’s some interesting nonlinear differences occurring as your shape changes. The perimeter to area ratio gets smaller as you stretch out a rectangle. The geometry of your peaked roof is also nonlinearly changing your volume if you change the short axis.Watch out, since changing multiple parameters that have nonlinear impacts on the energy thermodynamics can get confusing when trying to evaluate the model.
so I kind of ignored this problem for quite a few weeks now, but as it turns out in my project, I really should solve this issue.
SaeranVasanthakumar
Saeran, as you suggested I tried to compare a bit more “apples to apples” now, where a larger 3 floor residential building is modelled both in my previous “urban scale” principle of 1 HB zone = 1 Floor, but also from a bunch of “modules” roughly the size of a smaller Tiny houses that I have to compare them to. Still, energy demand of more than 10x times difference:
I applied Passive house construction to all buildings, and after consulting with energy experts on this project, they are expecting the Tiny houses to have maximum 2x times heating demand over a regular-sized passive building.
Replies to your comments:
• Yes, the WWR is set to 0.2 to all
• I’m not sure what you meant, if you meant checking the “thermal Load Balance” then the values are likeso:
Actually, I think splitting zones for the larger buildings, (and possibly removing the peaked roof) has solved your problem. Your original large buildings were 3.9 kWh/m2 and now are 11.57 kWh/m2. It’s no longer a factor of 10 difference between the large building and your smaller buildings (which have a 16 - 35 kWh/m2 demand in your original image). The 3.9 kWh/m2 was what seemed so unrealistic to me in the original post, and zone splitting has addressed the issue.
Those small buildings with a ~100 kWh/m2 are not a realistic basis for comparison because they have double (or more) of the amount of exterior envelope of an equivalent thermal zone in the larger building. It is reasonable for the heating loads to be extremely high in such a space.
The larger buildings are still going to perform better then your medium-sized buildings because even though their subdivided zones will more realistically capture perimeter (extreme) spatial conditions, they have less exterior envelope exposure. The fact that the larger 1 floor = 1 zone is up to 9.73 here is a reflection of the lack of peaked roof and lower aspect ratio from the 3.9 buildings, and is actually consistent with our geometry/energy assumptions.
If modeled correctly, we should expect zones, with the same boundary conditions, geometry and size to have the same heating demand across the buildings. Once that is correctly modeled, the ratio of corner zones to non-corner zones is likely what’s driving the difference in heating between your various building sizes. You can check this by checking the heating demand per zone.
You can start digging into various other metrics to identify where any discrepancies is occurring (zone metrics, zone surface metrics, the energy balance component), and there may be further errors contributing to some swings (e.g. I’m still unsure why your original image has two midsize buildings that look the same have difference of 2x heating demand, you may also want to look at perimeter/core zone splits for larger areas) - but based on the results you’re showing - zone splitting was the problem for your larger buildings.
