power-model

These files are for running the power model (OSeMOSYS).

1. Get set up first

The recommended way to get these files on your computer is to clone using Git:

git clone https://github.com/asia-pacific-energy-research-centre/power-model.git

Make sure you run this command in the folder you want.

You need to create a conda environment. To install, move to your working directory and copy and paste the following in your command line:

conda env create --prefix ./power_model_env --file ./config/environment.yml

conda activate ./power_model_env

2. To run the model

Make sure you did Step 1 (only need to do it once).

1. Make sure you are in the working directory: In the command prompt you can check with pwd

2. Edit the data input sheet. Configure your model run using the START tab. Leave Config file and Solver as data_config_all_calculated.yaml and coin_mip if you are unsure. They should be what you use 99% of the time anyway.

3. Run the model by typing in python ./src/main.py <input_data_sheet_file>" a. where <input_data_sheet_file> is the name of the input data sheet file in ./data (e.g., data.xlsx) b. For example, python ./src/main.py data.xlsxis the command to run the model with the data in data.xlsx c. You can also add> output.txt 2>&1to save the complete output to a text file. For example,python ./src/main.py data.xlsx > output.txt 2>&1. This will print a lot of stuff, so the file could become quite large. If you want to just see the most important outputs, the tmp/ECONOMY/SCENARIO/process_log_ECONOMY_SCENARIO_FILEDATEID.txt` is useful.

4. The model will run. You can take a look at the input and intermediate data in the tmp/ECONOMY/SCENARIO/ folder. There will be results in the results/ECONOMY/SCENARIO folder and visualisations in the visualisations/ECONOMY_SCENARIO folder.

3. Debugging model runs

When the script runs, the following temporary files are saved in ./tmp/ECONOMY NAME/SCENARIO:

• combined_data_ECONOMYNAME_SCENARIO.xlsx
• datafile_from_python_ECONOMYNAME_SCENARIO.txt
• model_ECONOMYNAME_SCENARIO.txt
• process_log_ECONOMY_SCENARIO_FILEDATEID.txt
• specs_FILEDATEID.txt
• ECONOMYNAME_SCENARIO_config.yaml The above files are created before OSeMOSYS runs. If you notice that OSeMOSYS gives you an error, check these files.

If the model solves successfully, a bunch of CSV files will be written to the same tmp folder. These are then combined and saved in the results folder as Excel files.

If there is an error message saying the model is infeasible, check your model data. You can also double check the process_log_{economy}_{scenario}.txt file for outputs from the solving process. If the model is infeasible, the results files will not be written and you will get a "file not found" error message or something. This is your clue that the model did not solve. You always want to see a message in the solver output saying "OPTIMAL LP SOLUTION FOUND".

You may find issues with the 'datafile_from_python_ECONOMYNAME_SCENARIO.txt' file not being created and this causing the solver to not work. One possible reason that is hard to spot is that installing/uninstalling otoole might fix it. this is because when otoole is installed it seems to record the location of itself in its code. So if you install it in one location, then move it, it will still be looking for the old location. To fix this, uninstall otoole, then reinstall it in the new location. I have put an issue on Otooles github about this so maybe it will be fixed.

4. Adding results to config yml files

To add results (e.g., capacity factor) you need to edit the following files:

• osemosys.txt
• XYZ_config.yaml

The osemosys_fast.txt file is where the calculations occur. Following the pattern from the other results. The results_config.yml file tells the script to include that result and add it to the combined results Excel file.

5. Using results

Saved in the results folder will be a few different files. The ones with name ~ tall_...xlsx, will be a combination of all the results in one csv.

Also, if you have set the variable SAVE_RESULTS_VIS_AND_INPUTS to True in the main.py file, then you will also have a folder in results/ with the FILEDATEID and some general settings in it's name (i.e. /2023-04-11-134916_19_THA_Reference_coin_mip/) which will contain all the resutls, input, intermediate and visualisation files. This can be useful for debugging and testing.

Extras:

Creating visualisation of RES

You can create a visualisaton of the RES as stated within the config/config.yml files. The script to run this will be outputted at the end of each model run, but you will need to run it in command line yourself.

Avoiding errors when running the system:

I have included a few checks to make sure things in the input data are as they should be, however it is hard to cover for all of the possible ones, and a bit of a waste of time to cover for ones which will eventually be caught by otoole/cbc/osemosys. So if you find some error you dont expect take a look below:

Validation:

https://otoole.readthedocs.io/en/latest/functionality.html#otoole-validate - to do when they develop its capacity a bit more. It would help to prevent possible bugs from changes to the config and input data.