The invisible power plants helping to end load-shedding

South African households and companies are increasingly ‘pooling’ their renewable energy generation to optimise their output, reducing the demand for electricity and ensuring a stable supply of electricity to the grid.

South Africa has experienced a boom in renewable generation over the past few years as the severity of load-shedding has increased. 

According to Eskom’s latest Weekly System Status Report, South African households have installed over 5,500 MW of rooftop solar capacity compared to just over 3,000 MW a year ago. 

This has been coupled with significant investments from private companies to reduce their reliance on Eskom’s electricity, cut their carbon emissions, and save money on electricity tariffs. 

Private companies have registered over 7,000 MW of renewable energy generation with Nersa, with RMB estimating that companies will add over 6,000 MW to the grid by the end of next year and contribute nearly 20,000 MW by 2030.

However, there are some problems with renewable energy generation that make it inefficient and unstable at the utility-scale. 

In particular, renewable energy is unreliable and unstable as the sun does not always shine, and the wind does not always blow. This makes it difficult for Eskom to manage its grid and ensure its stability. 

In Gauteng alone,  the variable output of rooftop solar can cause a sudden increase in electricity demand of 2,000 MW, depending on weather conditions.  

There are over 1,000 MW of rooftop solar panels in Gauteng. During overcast conditions, this 1,000 MW has to be drawn from Eskom’s grid and power supplied from other sources. 

Head of power for RMB’s infrastructure sector solutions, Daniel Zinman, told 702 that AI is increasingly helping companies and Eskom improve the stability of this power supply and optimise the output of renewable energy systems. 

Zinman said AI enables companies to create ‘virtual power plants’ that combine various energy sources, from wind to solar to battery storage, into a single plant to get the optimal output from these sources. 

This can be applied at the utility scale and on a residential level, helping estates go completely off the grid and optimise their savings on electricity tariffs. 

As it works right now, households and businesses have rooftop solar, inverters, and batteries to supplement their own electricity needs and mitigate against load-shedding. 

“It could work much more efficiently and optimally if a whole series of houses or companies were used in a combined fashion to optimise generation and consumption,” Zinman explained. 

Effectively, the generation and storage capacity of the houses and companies would be pooled to provide central control over a microgrid to smooth over supply and demand spikes. 

At the utility scale, technology would be used to create the most efficient energy mix between various renewable sources and battery storage. 

This would result in large renewable generation facilities providing more stable and reliable power to Eskom’s grid, making it easier for the utility to manage. 

Zinman explained that AI could be used to smooth over the peaks and troughs of renewable energy supply to enable it to more effectively meet peak demand in the mornings and evenings.