Stability of the South African Power Grid ---A data-driven Statistical Physics-based Approach

South Africa (SA) primarily relies on coal-fired power plants for its electricity supply. At least 12% of the population does not have access to power and roughly 10% cannot adequately afford electricity, particularly in rural areas. There is a particular challenge with reliable electricity supply in SA, as currently there is inability to deliver sufficient power according to the country’s demand. This has led to the implementation of rolling blackout  load shedding events across the country. Load shedding has marked deleterious societal effects. In 2021, the citizens and industries of SA were afflicted by a lack of power and periodic load shedding for over 48 days of the year. There are also unplanned outages (known as non-technical losses) for parts of the network.  During electricity outages, people and households typically use Diesel generators (if they can afford them), others simply remain without power. The use of Diesel generators during load shedding periods has severe detrimental effects, in financial, environmental, and health terms. Diesel generators are also frequently used in other African countries if there is no reliable connection to the power grid.
Our project aims to better understand, model, and mitigate the above load shedding situations in SA, working towards sustainable solutions (alternatives to Diesel generators) with no Carbon emissions that can be afforded by all. The overall aim is to model, understand and improve the stability of the African power grid using methods from statistical physics. To model the South African power grid as a whole, we will be using cutting-edge research methods in statistical physics modelling of complex systems, data-driven analysis and machine learning. A central aspect of our work plan will be the analysis of frequency fluctuations in the main grid, the control of microgrids, and the analysis of wind energy statistics, working towards future implementation of zero-emission generators based on wind power, solar panels, and batteries. We will model and analyse the overall demand patterns of electricity consumers in SA in a data-driven way, to finally arrive at practical solutions and concrete mitigation strategies. We aim at solutions that are particularly suited for the poorest in SA. At the same time our approach will contribute to lowering the Carbon footprint of SA in the long-term. The main general objectives of our proposal are as follows:

Model and forecast the stability of the SA power grid. Model the fluctuating electricity demand of individual households in a data-driven statistical-physics inspired way. From a complex system point of view, take up the challenge of modelling a system where demand and supply don’t match.
Model microgrids that use Diesel generators and/or zero-emission generators during load shedding periods. Measure frequency fluctuations in the grid and feed the data into theoretical statistical-physics based models. Develop statistical physics models that capture the essential features of the dynamics.
 Using neural nets, predict wind power fluctuations in SA. Prepare the ground for long-term mitigation strategies and a reliable electricity supply for all (in particular the poorest communities in SA) during load shedding periods and beyond, based on wind power, photovoltaic systems, and batteries.
 Foster new scientific collaborations between SA and the UK, dealing with statistical physics-based modelling of power grids. Work together towards a long-term strategy where power is provided in a reliable way, at the same time reducing the Carbon footprint of SA.