As Eskom consults a range of experts to assess potential solutions for resolving the load-shedding crisis, independent energy analyst Clyde Mallinson has a proposal that he believes will end rotational power cuts in two years without creating stranded assets or increasing tariffs.
The proposed solution is based on detailed modelling of the performance of the current system, which a new Council for Scientific and Industrial Research (CSIR) report shows to be relying increasingly on rotational cuts of between 1 GW and 4 GW to balance supply and demand.
In fact, the CSIR’s latest statistical analysis shows that 2021 was the country’s worst-ever for load-shedding, with an upper limit of 2 521 GWh shed, affecting 13% (1 169 hours) of the year.
For the period to the end of May, there has already been 1 023 GWh of load-shedding, affecting 19% (677 hours) of the year to date and placing 2022 on a path to setting a new load-shedding record.
To reverse this economically destructive trend in the shortest possible time, Mallinson is proposing that South Africa, through its large metropolitan councils and district municipalities, as well as through energy intensive clients, invests in a 6 GW/24 000 MWh battery storage fleet over a 24-month period, while accelerating a large-scale solar and wind programme in parallel.
Such front-loading of battery storage would, Mallinson’s model shows, halt load-shedding even in the absence of additional wind and solar, which would then have more construction time to ramp up to the scale required to free up space for coal-fleet maintenance and to cater for progressive coal decommissioning.
Such an approach would also eliminate the need for gas-to-power capacity, including power ships, which run a growing stranded-asset risk as the economics of a solar-wind-storage (SWS) system continue to improve and as the world intensifies it efforts to exit fossil fuels.
Mallinson’s modelling shows that, notwithstanding its unreliability, the coal fleet is typically able to deliver 22 GW on most days, before ramping down after the evening peak.
An analysis of a week in May this year shows that there was potential for the coal fleet to have delivered an additional 78 GWh of electricity if it was not being ramped up and down, but rather run more consistently.
“During that week, there was more surplus energy under a steady state coal curve then what was required to be shed during load-shedding.
“So, the question then, is why don't we run the coal at that steady-state speed?”, Mallinson asked during an interview with Engineering News.
The answer is that there is no way currently to absorb that coal surplus other than by recharging the upper dams at the country’s pumped-hydro schemes.
Eskom is, however, only able to fill up or recharge the pumped storage at a maximum capacity of 2 700 MW, and normally does this during the night.
On most days, Eskom uses the pumped storage in generation mode, and is not able to easily switch quickly between charge and generation mode. The pumped storage units typically generate at less than maximum capacity, as full capacity is held back to maintain a safe reserve margin in case of further coal unit breakdowns or a depletion of diesel stocks.
“It’s a big storage facility, yes, but it has a capacity constraint that can be best described as a large water bottle with a very thin neck.”
The solution, using Mallinson’s analogy, would be to create many smaller storage bottles with thick necks, which in the electricity industry would imply distributed short-duration battery storage facilities with a capacity large enough to absorb the surplus, as well as handle the short duration peaks.
“We will then fill that short-duration storage with the surplus that we can generate if we run the coal fleet at a steady state and we will release the energy from that additional storage at peak times and eliminate load-shedding.”
Should the coal fleet not be in a position to supply 22 000 GW, Mallinson believes there is even potential to deploy the battery fleet such that it is able to tap the 1 000 MW of instantaneous and regulating reserves – two sub-categories within a bigger reserve margin that Eskom is compelled by regulation to sustain and which, at times, are being maintained through load-shedding.
Such reserves need to be able to enter the grid within two seconds, which is possible with batteries, which can stop charging within milliseconds.
“In other words, if the coal surplus were not available, the instantaneous and regulating reserve of 1 000 MW could provide 24 000 MWh, which is precisely what a 6 GW battery fleet, with four hours duration, would require to deliver.”
“Charging battery energy storage systems with reserve margin allows the system operator to still classify the charging capacity as reserve margin, as it can be reversed within milliseconds.”
For this reason, Mallinson believes an emergency investment programme for a 6 GW/24 000 MWh battery fleet should be prioritised for delivery over the coming 24 months to address load-shedding.
Such a programme would be unprecedented in scale, but Tesla’s 90-day delivery of a 100 MW/129 MWh battery system in South Australia in 2017 is indicative of what is possible under emergency conditions.
Such conditions could exclude onerous requirements for local content, for instance, as that would stymie the roll-out. However, it would make South Africa a leading destination for utility-scale stationery storage, which could spur manufacturing investment once companies understand the scale of the country’s ongoing storage opportunity.
The front-end loading of the battery investment would also not, in Mallinson’s view, disrupt the larger deployment of an SWS-based electricity system, which will require short-duration storage to complement the variable renewable energy plants.
In his view, such a system would be built at a ratio of 8 GW of solar photovoltaic (PV), 5 GW wind and 3 GW storage. By 2030, such a system in South Africa would require 45.5 GW of solar PV, 24 GW of wind and 17 GW of storage, which together with residual coal, diesel and embedded generation could consistently meet 289 TWh of yearly demand.
At prevailing battery storage costs, Mallinson estimates the investment would cost R100-billion, which could be spread across the eight metropolitan councils, which are all moving to find solutions to end load-shedding, some of the larger district municipalities, and large energy intensive users.
From an economic perspective, the payback would be rapid, given that 24 hours of stage two load-shedding is conservatively estimated to cost about R1-billion. “So, in 100 days of avoided load-shedding, at a macroeconomic level, we would have paid it off.”
However, Mallinson believes the commercial proposition to be equally attractive, as the batteries would be “trickle charged” at off-peak tariffs and injected at peak-time tariffs.
“Therefore, the investment does not need to lead to an increase in a resident’s electricity bill. In fact, the tariffs could fall over time as the council sources cheaper forms of electricity from wind and solar.”
To add even further citizen momentum, Mallinson believes there is an opportunity to allow residents to invest directly into the batteries, receiving annuity income dividends over and above the benefits of having staunched load-shedding.
“If we could find 50 000 people or businesses in Johannesburg, for instance, to each stump up R100 000 from their access bonds, we would have the R5-billion required to invest in a 300 MW battery that could end load-shedding across the city.”