MATSHELA KOKO
I write to you, Minister Gwede Mantashe because, according to the Electricity Regulation Act 4 of 2006, the Minister of Energy is the electricity supplier in the last resort. My submission is that unless you act differently, purposefully and with haste, South Africans can expect under-frequency blackouts to continue up to 2035 and beyond.
The energy plan announced by President Cyril Ramaphosa on July 25, 2022, together with Eskom’s Strategy 2035, is a threat to the security and quality of power supply up to 2035 and beyond. This plan will not stop the current electricity blackouts in the short to medium term. To put it bluntly, the last-resort electricity supplier is failing in his constitutional mandate.
The only short-term solution other than resorting to electricity blackouts is a coordinated government effort to improve Eskom’s energy availability to at least 72%. The government is on record stating that Eskom is too big to fail. Eskom will have failed if it cannot restore the energy availability to at least 72% in the next 18 months.
The long-term solution is a power system with adequate rotational inertia and inertial power responses to support the future low-inertia power grid. This context must recognize the need for a massive increase in private investment in electricity generation.
From 1998 to 2003, before Alec Erwin was appointed the minister of public enterprises, the country focused on introducing new independent players into the generation market and restructuring the electricity distribution industry (Chettiar et al., 2009).
The diminishing electricity reserve margin problem was not considered urgent or immediate.
The “strategies employed to achieve these two objectives did not materialize due to inappropriate legislative and regulatory frameworks and stakeholder alignment challenges.”
The net effect was the electricity blackouts amounting to 642 GWh in 2007–2008.
There were 1 528 GWh of electricity blackouts in 2014–2015. The blackouts were unrelated to the government inaction from 1996–2003; they were related to the catastrophic failure of the coal silo at the Majuba Power Station in November 2014.
An electricity generation capacity loss of 3 000 MW occurred at the station due to the silo’s collapse. In 2018–2021, the under-frequency electricity blackouts amounted to 5 863 GWh; as in 2014–2015, what happened in 1996–2003 was neither the contributory nor the root cause.
The actual cause of the under-frequency electricity blackouts in 2018–2021 was the poor energy availability factor relative to the IRP 2019 plan. Connecting more intermittent generation capacity to the grid is a poor technical response to the problem due to a baseload fleet’s poor performance.
Speaking during an energy transition breakaway at the South African Investment Conference of 2022, Eskom CEO Andre De Ruyter was quoted by the Engineering News online magazine as saying: “The utility estimated that some 68 000 MW of mostly new variable renewable-energy capacity would have to be added to cater for the 22 000 MW of coal [sic] that the utility would be retiring by about 2035.”
In your capacity as the electricity supplier of the last resort, you should be mindful that converter-interfaced generators help the grid, but they are not the grid. It would be best if you built a reliable power grid now and in the future. The last resort’s electricity supplier cannot outsource this obligation to others, including the private sector.
Historically, “we would forecast load and deploy generation.” This is how the second and third industrial revolutions were built. Constructing the next integrated resource plan on the basis that in the distant future, “we would forecast generation and deploy load” is negligent.
This will negatively impact the fourth industrial revolution’s pace and scale. The fourth industrial revolution requires reliable, clean, and affordable electricity.
Meridian Economics (2022) reports that, “rather than increasing system risk as many observers expect, the analysis based on the empirical data unequivocally shows that adding variable renewable generators to the existing distressed South African power system will result in a disproportionate reduction in load shedding, and an increase in system reliability” (my emphasis).
Scholars with more extraordinary expertise in power system engineering would disagree with Meridian Economics. The power system’s total inertia gradually decreases with increasing variable renewable generators.
System inertia helps counter frequency variations from the nominal value (50Hz) in response to a system disturbance from an initially balanced system before the primary response is activated. The primary response is activated within 10 seconds to arrest a frequency excursion outside the deadband and is sustained for at least 10 minutes (Tshwagong et al., 2021).
According to Milano et al. (2018), the bigger the total inertia, the higher the system’s kinetic energy. Thus, the higher the system’s kinetic energy, the more the system can compensate for power imbalances from system disturbance or load variations.
Adding 68 000MW of mainly new variable renewable energy to replace the 22 000MW of coal power stations that Eskom would be retiring by 2035 will produce a low inertia system.
A grid with low inertia is intrinsically less secure than one with high inertia. Lower levels of inertia generate a steeper rate of RoCoF slopes and low-frequency nadir.
RoCoF is the rate of change of system frequency, and frequency nadir is the minimum value of the frequency reached during the transient period. A steeper rate of RoCoF slopes and a low-frequency nadir are harmful in the event of a sudden load change.
RoCoF was of minor relevance for generation systems primarily based on synchronous generators. This has since changed due to the low system inertia caused by the disposal of synchronous generation.
The detailed studies from EirGrid and ENTSO-E cited in Milano et al. (2018) recommend relaxing the limits on RoCoF and frequency nadir as well as remedial RoCoF control actions, such as emulating virtual inertia. Control actions, such as emulated inertia, sometimes called synthetic or virtual inertia, are promising but still under development, and there are no indications of the costs.
I am persuaded by the literature that there are significant technical challenges to be addressed in low inertia systems before South Africa can go with the proposal by Meridian Economics. I hope that you are persuaded too.
It is thus an intriguing thought experiment for Meridian Economics to submit, “that adding variable renewable generators to the existing distressed South African power system will result in a disproportionate reduction in load shedding and an increase in system reliability.”
Based on the arguments stated above, it is evident that the electricity plan for the country cannot be based on such an experiment. The Minister is the electricity supplier of last resort. It would help if you acted differently, purposefully, and with haste in consultation with the National Energy Regulator of South Africa to keep the lights on until 2035 and beyond.
This will require an immediate commitment to the next pump storage scheme in Steelpoort (Tubatse), and an acceleration of the 2000 MW nuclear build program. It would be ideal for the two projects to be connected to the grid by 2035. These are no-regret moves. Unfortunately, the window of opportunity for new gas to power plants is out.
As I conclude, I submit that the problem is not with renewable energy technologies or batteries. The problem is that people are not planning for their use or how renewable energy technologies might be most helpful (Angwin, 2020).
Koko is the managing director of Matshela Energy
IOL