parameter constraints year_of_retrofit

[lamingblu]

Issue Description

I'm working with TEASER 1.3.0 to create individualized retrofit scenarios for real building data. My Python script automates TEASER model creation from Excel data and applies various retrofit standards (EH40/EG40, Custom City Standard and Passivhaus) with custom U-values.

Current Implementation

Excel integration for real building data (20 buildings)
Mapping from building types to TEASER parameters
Custom U-value configurations
Retrofit application
Export to OpenModelica for energy consumption analysis

retrofit_configurations = { 'EG40_Standard': { 'base_parameters': { 'year_of_retrofit': 2025, # <-- Potential issue? 'type_of_retrofit': 'adv_retrofit', 'window_type': 'Kunststofffenster, Waermeschutzverglasung, dreifach',#aus JSON 'material': 'EPS_032' }, 'custom_u_values': { 'exterior_wall': 0.20, 'roof': 0.14, 'windows': 0.95 # ... } } }

Problem

To my understanding of example e7 and the project.py, year_of_retrofit represents the intended (future) year of retrofit and needs to be set. However my retrofit year_of_retrofit configurations currently use 2025+. As a result division by zero errors occur.

Questions

Is year_of_retrofit limited to specific year ranges and how do I retrofit a specific scenario in the future.

Should I use a historical year (e.g., 2016) for current standards like EH40/EG40 and individual u-values, window_type etc.? In a previous approach I just set them to 2014 and 2016 respectively but that's not necessarily wanted.

Maybe there is something obvious that I don't see. Thanks for any idea into the right direction.

[PGorzalka]

Division by zero often happens when somewhere there is no layer configuration available, and this could be the case for future-planned retrofits... You might have to check which data classes are loaded and if there is a specification available for the year of retrofit that you set. If no, you might be able to import your own data for it with the new possibility of using custom data classes. Didn't try that out though.

[lamingblu]

@PGorzalka, thank you for confirming the issue behind the ZeroDivisionError. Only by setting year_of_retrofit to 2015 and creating custom U-Values, modelica models were created and worked. I just tested year by year before coming to that solution.

• Is 2015 still the newest "old" year that internal files of TEASER refer to in a stable way?
• Following the results of discussion Italian tabula archetypes integration #735 and to my understanding, the scope of using "my own data" implies creating three new JSON files for defining the thermal properties of a retrofitted building: teaser/data/input/inputdata/MaterialTemplates.json, teaser/data/input/inputdata/TypeElements_KFW.json and teaser/data/input/inputdata/TypeElements_TABULA_DE.json right?

[PGorzalka]

@lamingblu to be honest, I never tried, I have only been working with the internal files and Python debugging to get to where you need to add what. But I had a quick check and found that all TABULA DE default buildings indeed do have both retrofit and adv_retrofit options. So what I suggested before should not solve the issue (sorry).

If you don't know enough of Python to dig into the TEASER code and find where the issue comes up (i. e., where the element is not filled with layer data), could you please specify the full code of how you apply the retrofit? And maybe find out if the issue comes up for all of your 20 buildings or just for some? I have another idea what the problem could be, but I don't want to make another wrong guess before testing ;)