Energy Demand too low TABULA Germany

[pnguyen-dh]

Describe the problem
In the newest version, CEA already implements the data for Germany from TABULA. So I use this database for my simulation, I would expect the energy demand (specific heating demand) should be roughly the same as in TABULA, or maybe about 10-20% different. But after the simulation the building's energy demand are too low, I expect 150 kWh/(m2a) for single family house 1969-1978 (source), but get only about 15-65 kWh/(m2a). I would say that the simulation demand were too good to be true.

To Reproduce
If you need the input/output data, I can zip it and send it. I used the standard SFH_F, SFH_NR

Expected behaviour
The energy demand should be then higher? Or I actually might miss something.

Screenshots
Energy End-use Intensity for District

Thank you for your support!

[ddceruti]

Hi! Thanks for your feedback and using the database! Can you provide an input file to reproduce (maybe 1 or 2 synthetic buildings)? I would look into it and see if it depends on the TABULA values within the database or some other database parameter.

[pnguyen-dh]

inputs.zip

Hi, here is the input. I also tested the same buildings with the database from CH, and the energy demands are much higher. I also tried to look into the DE database but couldn't find anything special. Feel free to tell me if you need anything else.

[ShiZhongming]

Hi @pnguyen-dh

Thanks for raising this issue. It is very much appreciated as the DE Database is new and needs time to stabilise.

I have also used a random single-family house neighbourhood near Berlin.

For buildings using SFH_F, the heating demand I got was ranging from ~90 kWh/m2 - ~120 kWh/m2, which is rather similar to the reference value of TABULA at 139.5. Please note that in the graph, the heating energy demand calculated is based on a floor area equaling 0.85 X gross floor area, while energy end-use intensity equals this heating energy demand divided by 100% of the gross floor area. This could be one of the reasons that lead to the low normalised results in the graph.

For buildings using SFH_F_NR, the heating demand I got was much lower than that of TABULA, which is as what you had described.

We will need a bit more investigation over this. I am also tagging @ddceruti here. Please share your opinion on this.

Meanwhile, we would like to mention that CEA has been a tool designed for district-scale building energy simulation using simplified building geometries. We would recommend applying CEA for comparing multiple design scenarios or finding patterns when adjusting input parameters.

[pnguyen-dh]

HI @ShiZhongming,

Thanks for your reply. I also tried to change the Hs_ag = 1, and recalculated everything. The number for SFH_F_NR and SFH_F_AR were still too good. I would suggest that, there is some problem with the refurbishment settings, I can see that it was even better than passivhaus.

I'm working on a small project now for district labeling, so I also need to compare the average heating demand in a district, and how good is a building before and after refurbishment, for example. The energy potentials calculation is kind of on point for me. Anyway, thanks for your supporting!

[ddceruti]

Hi,
I checked the U-Values of the whole envelope again and I could not find any discrepancies with the reported TABULA values.
I found a mistake in the use_type warm water per person: 11.6 ldp instead of the 40 in CH database and 36 % of 121 ldp = 43.6 by the BDEW (https://www.bdew.de/service/daten-und-grafiken/trinkwasserverwendung-im-haushalt/). I will update this and check the database again more throughly next week.
Thanks again @pnguyen-dh for your feedback and @ShiZhongming for your reply.