PySoda is an irradiance-based synthetic Solar Data generation tool to generate realistic sub-minute solar photovoltaic (PV) power time series. Soda emulates the weather pattern for a certain geographical location using 30-min averaged irradiance and cloud type information from the National Solar Radiation Database (NSRDB)

Use Git to install pysoda in your current python environment

git clone https://github.com/Ignacio-Losada/SoDa.git
cd SoDa
pip3 install -r requirements.txt

Use Git to install a conda environment for pysoda

git clone https://github.com/Ignacio-Losada/SoDa.git
cd SoDa
conda env create -f environment.yml

Once PySoda is installed, solar time series can be generated as follows.

First, you'll need to create an object with the coordinates of interest

import soda
lat = 33.9533
lon = -117.3962

site = soda.SolarSite(lat,lon)

Then, obtain the closest NSRDB point to the specified coordinates and retrieve the neccesary irradiance values. We recommend retrieving the 30-min average NSRDB irradiance data to obtain the best results

year = "2015"
leap_year = False
interval = "30"
utc = False
df = site.get_nsrdb_data(year,leap_year,interval,utc)

You'll also need to specify the solar panel configuration and obtain the 30-min averaged solar time series

clearsky = False
capacity = 1
DC_AC_ratio = 1.2
tilt = 33
azimuth = 180
inv_eff = 96
losses = 15
array_type = 0

pwr = site.generate_solar_power_from_nsrdb(clearsky,capacity,DC_AC_ratio,tilt,azimuth,inv_eff,losses,array_type)

Finally, you can generate stochastic solar time series for a given date at different time resolutions, e.g. 5 seconds

date = "2015-01-10"
resolution = "5S"
solar_data = site.generate_high_resolution_power_data(resolution, date)

This function will return a pandas dataframe with the solar generation. And we compare our results

If you find SoDa useful in your work, we kindly request that you cite the following publication:

@inproceedings{,
  author = {Ignacio Losada Carreno and Raksha Ramakrishna and Anna Scaglione and Daniel Arnold and Ciaran Roberts and Sy-Toan Ngo and Sean Peisert and David Pinney},
  title = {SoDa: An Irradiance-Based Synthetic Solar DataGeneration Tool},
  booktitle = {2020 IEEE International Conference on Smart Grid Communications (SmartGridComm)},
  year = {2020},
  month = {November},
  pages = {1-6},
  doi = {}
}