This repository contains an example of how to execute a power system scheduling simulation in Sienna using the PowerSimulations.j package, expanding on the data assembled for the PowNet application for the power grid in Cambodia.

This simulation is based on three open-source data and modeling tools used to model power systems with renewable resources such as wind and solar:

1. The RE-Data Explorer can be used to investigate where to site wind and solar plants and to download wind and solar time-series resource files for selected latitude & longitudes where the plants will be located.

   • This repository already contains resource data files in the REDE_resource_data/Input/ folder from the RE-Data Explorer for 5 hypothetical wind and solar plants in Cambodia.

2. The System Advisor Model (SAM) can be used to model the power output (MW) from a wind or solar power plant, using the hourly or subhourly resource data from the RE-Data Explorer as inputs.

   • There are multiple ways for users to access SAM, including but not limited to:

     • A downloadable GUI, which can be used for manual analysis of individual locations, i.e., power plants
     • A Python interface called PySAM, which can be used to programmatically process multiple locations and/or plant configurations
     • The Renewable Energy Potential (reV) tool, which can be used for very large scenario analysis and batch runs

   • In the REDE_resource_data folder, there is a Python script that uses PySAM to process multiple solar and wind time-series resource files downloaded from the RE-Data Explorer for locations in Cambodia. This notebook is based on a PySAM example notebook, available here.

   • This processing has already been complete for the 5 example plants, and the outputs are available in the REDE_resource_data/Output/ folder. However, if users would like to add other wind and solar locations and process those, here are the required steps:

     1. Add wind or solar resource files from the RE-Data Explorer to the REDE_resource_data/Input/ folders

     2. Update the REDE_resource_data/RE_plant_config.csv file, which is a manually generated .csv file to make it easier to load the same plant metadata into both PySAM (to generate the plant-specific power profiles) and Sienna\Data (to attach the hypothetical solar and wind plants to the existing PowerSystem.jl model). It contains the following 4 pointers:

        • type (used by Sienna\Data): either "Solar_PV" or "Wind_WT"
        • node (used by Sienna\Data): Acronym of the node in the PowNet system that the plant should connect to (for example, the closest node based on the latitude/longitude of the plant).
        • resource_file (used by PySAM): Resource file (from RE-Data Explorer or the National Solar Radiation Database (NSRDB))
        • specification_file (used by PySAM): Plant configuration file, exported from the SAM GUI. Example wind and solar plant specification files are already included in the REDE_resource_data/Input folders for the following configurations:
          • Solar: (1) A 100 MWdc fixed (12 degree ~= Cambodia's latitute) tilt system or (2) a 100 MWdc one-axis tracking plant
          • Wind: Models plant with Gamesa G114 2.0 MW turbines, which is selected to emulate the 'T200' representative turbine in the RE-Data Explorer's wind technical potential study (more info Table 12, page 23 here); A review of Cambodia's mean wind speeds in the RE-Data Explorer indicates that the T200 or perhaps T237 representative turbines are a good choice for Cambodia.

     3. Then, run the REDE_timeseries_prep.py script by navigating to the REDE_resource_data folder in a terminal with conda installed, and running:

            conda env create --name pysam_env --file=environment.yml
            conda activate pysam_env
            python REDE_timeseries_prep.py
        

        This will install PySAM and run process the plant data. (Note: There are two ways to install PySAM: pip and conda. If you decide to install PySAM on your own rather than using the environment.yml file, it is recommended to use pip due to delays updating the conda version of PySAM: pip install nrel-pysam.)

3. Finally, Sienna can be used to run simulations of the Cambodia system, including the new renewable resources.

This example includes two pieces of code:

• Data preparation using Sienna's PowerSystems.jl package and construction of a System. This includes pulling in the wind and solar power profiles.
• Example Simulation of a day-ahead unit-commitment scheduling sequence using Sienna's PowerSimulations.j package.

Examples are provided in .jl script format. To autogenerate ipython notebooks, execute the following commands from a Julia REPL:

] activate . # Activates the required environment
include("literate.jl")

From there, users can walk step-by-step through the generated example, PSI-Cambodia.ipynb, using Jupyter Notebooks.