dwind 0.4

Distributed Wind Generation Model

Distributed Wind Generation Model (dWind)

Please note that at this time the model can only be run on NatlLabRockies's Kestrel HPC system. Though a savvy user could recreate our data in their own computing environment and update the internal pointers in the example configuration at examples/larimer_county_btm_baseline_2025.toml and examples/model_config.toml.

Installing dwind

1. Install Anaconda or Miniconda (recommended) if not already installed.

2. Clone the repository

   git clone https://github.com/NatlLabRockies/dwind.git
   

3. Navigate to the dwind repository.

   cd /path/to/dwind
   

4. Create your dwind environment using our recommended settings and all required dependencies.

   conda env create -f environment.yml
   

Running

Configuring

dWind relies on 2 configuration files: 1) a system-wise setting that can be shared among a team, and 2) a run-specific configuration file. Both will be described below.

Primary model configuration

The primary model configuration should look exactly like (or be compatible with) examples/model_config.toml to ensure varying fields are read correctly throughout the model.

Internally, dWind is able to convert the following data to adhere to internal usage:

• Any field with "DIR" is converted to a Python pathlib.Path object for robust file handling
• SQL credentials and constructor strings are automatically formed in the [sql] table for easier construction of generic connection strings. Specifically the {USER} and {PASSWORD} fields get replaced with their corresponding setting in the same table.

Configuration, the primary class handling this data allows for dot notation and dictionary-style attribute calling at all levels of nesting. This means, config.pysam.outputs.btm and config.pysam.outputs["btm"] are equivalent. This makes for more intuitive dynamic attribute fetching when updating the code for varying cases.

Run configuration

The run-specific configuration should look like examples/larimer_county_btm_baseline_2025.toml, which controls all the dynamic model settings, HPC configurations, and a pointer to the primary model configuration described above.

Run the model

dwind has a robust CLI interface allowing for the usage of python path/to/dwind/dwind/main.py or by directly callingdwind. For more details on using the CLI, use the --help flag, or visit our CLI documentation page https://natlabrockies.github.io/dwind/cli.html

To run the model, it is recommended to use the following workflow from your analysis folder.

1. Start a new screen session on Kestrel.

   screen -S <analysis-name>
   

2. Load your conda environment with dwind installed.

   module load conda
   conda activate <env_name>
   

3. Navigate to your analysis folder if your relative data locations in your run configuration are relative to the analysis folder.

   cd /path/to/analysis/location
   

4. Run the model.

   dwind run config examples/larimer_county_btm_baseline_2025.toml
   

5. Disconnect your screen Ctrl + a + d and wait for the analysis to complete and view your results.