About the Project

Agora Energiewende aims to model the Kazakh power system, incorporating a substantial increase in variable generation, such as solar and wind, surpassing the current official mid-term policy goal of 15% of all renewable energy sources (RES) in generation by 2030. This endeavor has received support from Open Energy Transition on the modeling side.

Used for 2024 study, "Modernising Kazakhstan’s coal-dependent power sector through renewables - Challenges, solutions and scenarios up to 2030 and beyond"

Development status: Active and Stable

PyPSA-KZ-Data

Extra data for the Kazakhstan model that will be used as input for PyPSA-Earth. Repo design oriented on: https://github.com/pypsa-meets-earth/pypsa-zm-data

Data

Contains openly available data for Kazakhstan

Demand data

Monthly electricity demand data with monthly aggregation provided by Kazakhstan operator of the electric energy and power market

kz_demand_validation.csv (demand_valid_clean.R contains for details on how data were extracted and a simple vizualization; needs to work any R terminal + installation of two libraries)

Model Execution

Setting up the general repositories

The provided workflow builds on PyPSA-Earth. Therefore, first, the PyPSA-Earth repository must be forked and the fork should then be cloned. A fork can be created by navigating to the PyPSA-Earth website. By clicking on the fork-symbol in the upper right corner, a fork is created and linked to the specific user. While making a fork, unclick Copy the main branch only option to fork all branches and tags of pypsa-earth repository.

Next, we also need to fork the pypsa-kz-data repository. A fork can be created by navigating to the pypsa-kz-data website and clicking the fork symbol in the upper right corner.

In order to clone both forks to the correct locations on a local machine, the following commands can be used using the local machines shell:

git clone https://github.com/<user-name>/pypsa-earth

<user-name> must be replaced with the personal github-username.

After that, one must change to the freshly created pypsa-earth repository.

cd pypsa-earth/

Switch to the stable v0.4.0 version of pypsa-earth that is compatible with pypsa-kz-data repository:

git checkout tags/v0.4.0

Repeat the cloning, this time for the pypsa-kz-data repository.

git clone https://github.com/<user-name>/pypsa-kz-data

Again, <user-name> must be replaced with the personal github-username.

In order to install the pypsa-earth environment, instructions are provided in the pypsa-earth documentation, see Install dependencies and Python dependencies. After installing the environment, activate it using

conda activate pypsa-earth

Modeling adaptations for KZ study

To adapt the overall workflow for kz, only two further changes are necessary.

Firstly, open the Snakefile (in pypsa-earth/) and navigate to line 1071-1072, which should read

os.system("snakemake -j all solve_all_networks --rerun-incomplete")
os.system("snakemake -j1 make_statistics --force")

and replace these two lines with

os.system(f"snakemake -j1 networks/{wildcards.scenario_name}/base.nc")
os.system("cp pypsa-kz-data/data/custom_powerplants.csv data/custom_powerplants.csv")
os.system("snakemake -j1 solve_everything --rerun-incomplete")

Secondly, copy the default configuration file to the pypsa-earth folder using:

cp pypsa-kz-data/config.kz_default.yaml config.default.yaml

In case you already have a custom config file, make sure to replace it as well, using

cp pypsa-kz-data/config.kz_default.yaml config.yaml

Note! Run two aforementioned commands in pypsa-earth directory. You are now all set to run all scenarios!

Running KZ scenarios

To prepare running all scenarios, execute

snakemake -j1 prepare_kz_scenarios

Optionally, to save time for future runs, you can now set enable: retrieve_databundle: True in the config.yaml to False. If you already have build all cutouts for 2011, 2013 and 2018, you can also set enable: build_cutout: True to False. Finally, to run all scenarios, execute

snakemake -j1 run_all_scenarios

After all scenarios have executed successfully, all results are generated and locally saved in pypsa-earth/results/<scenario_folder>/networks/.

Potential errors

• A rule is killed. In this case, open the Snakefile in pypsa-earth or open kz.smk in pypsa-kz-data (depending on the rule which is killed), navigate to the rule that is being killed in the workflow and increase the memory assignment (for example, add a 0 at the end).

• The workflow runs into an error during the build_powerplants rule. In this case, try to repeat step 1. of the workflow using the command

cp pypsa-kz-data/data/custom_powerplants.csv data/custom_powerplants.csv

• Unusual error arising from either Snakemake or the Snakefile and proving to be challenging to comprehend: Inspect all indentation. Ensure there is no tab spacing; employ only spaces, i.e., . It is probable that the indentations before

os.system(f"snakemake -j1 networks/{wildcards.scenario_name}/base.nc")
os.system("cp pypsa-kz-data/data/custom_powerplants.csv data/custom_powerplants.csv")
os.system("snakemake -j1 solve_everything --rerun-incomplete")

are tabs instead of four spaces.

• Missing data/ folder or some relevant subfolders. This should normally be executed automatically when executing the rule prepare_kz_scenarios, however might be missing due to incorrect execution. The databundle can be also retrieved manually via:

snakemake -j 1 retrieve_databundle_light

• The rule retrieve_databundle_light always executes with an error. To avoid this, try setting enable: build_cutout: False to True.

• Hydrobasins data might not be accessible in some regions (e.g. in Kazakhstan). It is recommended to retrieve the databundle using VPN.

Comes in handy

After all cutouts were generated (i.e. the three files asia-<year>-era5.nc exist in the folder pypsa-earth/cutouts/, where <year> is 2011, 2013, and 2018, navigate to pypsa-earth/pypsa-kz-data, open the default config file, navigate to line 36, which should read build_cutout: True, and set it to build_cutout: false. This will save you a lot of time when (re-)runnig scenarios. But remember to set it back to true in case one of the cutouts was deleted!

Acknowledgement

Code development and testing: Open Energy Transition

Model assumptions: Agora Energiewende