pvlib python is a community-supported open source tool that provides a set of functions and classes for simulating the performance of photovoltaic energy systems. pvlib python aims to provide reference implementations of models relevant to solar energy, including for example algorithms for solar position, clear sky irradiance, irradiance transposition, DC power, and DC-to-AC power conversion. pvlib python is an important component of a growing ecosystem of open source tools for solar energy [@Holmgren2018].
pvlib python is developed on GitHub by contributors from academia, national laboratories, and private industry. pvlib python is released with a BSD 3-clause license allowing permissive use with attribution. pvlib python is extensively tested for functional and algorithm consistency. Continuous integration services check each pull request on multiple platforms and Python versions. The pvlib python API is thoroughly documented and detailed tutorials are provided for many features. The documentation includes help for installation and guidelines for contributions. The documentation is hosted at readthedocs.org as of this writing. A Google group and StackOverflow tag provide venues for user discussion and help.
The pvlib python API was designed to serve the various needs of the many
subfields of solar power research and engineering. It is implemented in
three layers: core functions, the Location and PVSystem classes,
and the ModelChain class. The core API consists of a collection of
functions that implement algorithms. These algorithms are typically
implementations of models described in peer-reviewed publications. The
functions provide maximum user flexibility, however many of the function
arguments require an unwieldy number of parameters. The next API level
contains the Location and PVSystem classes. These abstractions
provide simple methods that wrap the core function API layer. The method
API simplification is achieved by separating the data that represents
the object (object attributes) from the data that the object methods
operate on (method arguments). For example, a Location is
represented by a latitude, longitude, elevation, timezone, and name,
which are Location object attributes. Then a Location object
method operates on a datetime to get the corresponding solar
position. The methods combine these data sources when calling the
function layer, then return the results to the user. The final level of
API is the ModelChain class, designed to simplify and standardize
the process of stitching together the many modeling steps necessary to
convert a time series of weather data to AC solar power generation,
given a PV system and a location.
pvlib python was ported from the PVLib MATLAB toolbox in 2014 [@Stein2012, @Andrews2014]. Efforts to make the project more pythonic were undertaken in 2015 [@Holmgren2015]. Additional features continue to be added, see, for example [@Stein2016, @Holmgren2016] and the documentation's "What's New" section.
pvlib python has been used in numerous studies, for example, of solar power forecasting [@Gagne2017, @Holmgren2017], development of solar irradiance models [@Polo2016], and estimation of photovoltaic energy potential [@Louwen2017]. Mikofski et. al. used pvlib python to study the accuracy of clear sky models with different aerosol optical depth and precipitable water data sources [@Mikofski2017] and to determine the effects of spectral mismatch on different PV devices [@Mikofski2016]. pvlib python is a foundational piece of an award, "An Open Source Evaluation Framework for Solar Forecasting," made under the Department of Energy Solar Forecasting 2 program [@DOESF2].
Plans for pvlib python development includes the implementation of new and existing models, addition of functionality to assist with input/output, and improvements to API consistency.
The source code for each pvlib python version is archived with Zenodo [@pvlibZenodo].
The authors acknowledge and thank the code, documentation, and discussion contributors to the project.
WH acknowledges support from the Department of Energy's Energy Efficiency and Renewable Energy Postdoctoral Fellowship Program (2014-2016), Tucson Electric Power, Arizona Public Service, and Public Service Company of New Mexico (2016-2018), and University of Arizona Institute for Energy Solutions (2017-2018).
CH acknowledges support from the U.S. Department of Energy's Solar Energy Technology Office.
WH and CH acknowledge support from the Department of Energy Solar Forecasting 2 program.
MM acknowledges support from SunPower Corporation (2016-2017).
Sandia National Laboratories is a multi-mission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-NA-0003525.