Introduction

History

Historically, commercial tools have been extensively utilised by both governments and utilities to develop capacity expansion plans. One of the challenges of utilising such commercial tools to support transparent energy policy is the closed nature of the models, which are essentially a ‘black-box’. For this reason, initiatives such as openmod have been developed with the goal of creating open energy modelling platforms that are based on open-source software as well as open data.

PyPSA (Python for Power System Analysis) is a diverse toolbox of modules that can be assembled to produce a power / energy system modelling framework to meet specific requirements. As described by Brown et al PyPSA includes a toolbox of modules for multi-horizon investment planning, unit commitment of conventional generators, time-varying renewable generators, storage units, all combinations of direct and alternating current electricity networks, and the coupling of electricity to other energy sectors, such as gas, heating and transport.

An initial PyPSA-RSA model of the South African power system called PyPSA-RSA was developed by Hörsch and Calitz who studied a future renewable energy based South African electricity network for a single reference year (in other words, it was not an intertemporally optimised power system development model – showing an optimised transition over time – it only provided a snapshot of an optimised high renewables system at a particular point in time – 2040, albeit with reasonable spatial (network) resolution. The initial PyPSA-RSA model is described and evaluated in the paper PyPSA-RSA: Investment and operation co-optimization of integrating wind and solar in South Africa at high spatial and temporal detail.

PyPSA has also been utilised to build an extensive interconnected power system model of the European power system PyPSA-Eur. This model has been further extended to include full sector coupling of power, transportation and heat demand PyPSA-Eur-Sec. Like the PyPSA-RSA model, the PyPSA-Eur model framework has been designed to handle high spatial resolution (detailed network specification, i.e. grid constraints), but it does not perform multi-year optimisation and instead analyses a single year snapshot. In the African context the project PyPSA-Meets-Earth is currently ongoing and aims to develop an interconnected, high-level, African power system model on the same basis as PyPSA-Eur. Multi-horizon planning functionality, based on the assumption of “perfect foresight”, was only recently added to PyPSA in 2021, and currently PyPSA-Eur, PyPSA-RSA, and PyPSA-meets-Earth do not yet incorporate this capability. The PyPSA-Eur-Sec model does include myopic expansion planning capability that is based on separating the planning horizon into several shorter periods that are solved as separate optimisations.

Currently, there is no existing PyPSA model for South Africa that can be used to address the multi-year, least cost optimised, transition path to a decarbonised power sector over the coming decades. Therefore, an updated and expanded PyPSA-RSA model is required that is comparable with the existing, mostly proprietary, modelling approaches that are used in the country.

In 2022 Meridian Economics leveraged the experience from PyPSA-Eur, PyPSA-RSA and PyPSA meets Earth to significantly shorten the development time frame for the new model. Meridian Economics has undergone extensive work in terms of input datasets, model development, and validation against existing results from commercial tools to establish PyPSA-RSA credibility to ensure that the outcomes can be trusted to support policy decisions. The updated PyPSA-RSA model has the following capabilities:

  • Updated input data: Includes latest assumptions around technology performance and learning rates.

  • Multi-horizon capacity expansion planning: Myopic and/or perfect foresight expansion capabilities which can solve in five-year time steps up to 2050.

  • Validation test case : Benchmarked case studies using existing energy planning results in the public domain such as those developed by the CSIR.

  • Solar and wind resource assessment : Utilises the Atlite module and Global Wind Atlas to generate renewable potentials for South Africa’s Renewable Energy Development Zones.

  • Model file : Input excel spreadsheet for easier model input.

Folder structure

The content in this package is organized in folders as described below; for more details, please see the documentation.

  • data: Includes input data that is not produced by any snakemake rule.

  • scripts: Includes all the Python scripts executed by the snakemake rules.

  • resources: Stores intermediate results of the workflow which can be picked up again by subsequent rules.

  • networks: Stores intermediate, unsolved stages of the PyPSA network that describes the energy system model.

  • results: Stores the solved PyPSA network data, summary files and plots.

  • benchmarks: Stores snakemake benchmarks.

  • logs: Stores log files about solving, including the solver output, console output and the output of a memory logger.

  • envs: Stores the conda environment files to successfully run the workflow.

Note that the folders: resources, networks, results, benchmarks, and logs are created by the snakemake workflow and are included in .gitignore.

License

PyPSA-RSA work is released under multiple licenses:

  • All original source code is licensed as free software under MIT License.

  • The documentation is licensed under CC-BY-4.0.

  • Configuration files are mostly licensed under CC0-1.0.

  • Data files can be found here and are licensed under different licenses as noted below.

Licenses of the data used in PyPSA-RSA:

  • BY: Attribute Source

  • NC: Non-Commercial Use Only

  • SA: Share Alike