by Mark Swilling

Urgent steps must be taken this year to bring the cheapest and most reliable energy on to the grid within the next three years. Coal, nuclear and substantial gas infrastructure cannot be brought online within that time frame. Only renewables (wind and solar) can be built that quickly. That is an undeniable fact, no matter what your biases may be. Furthermore, renewables are also the cheapest option and prices are expected to continue to drop.

When South Africans discuss their problems, two words dominate the discussion: lockdown and load shedding. Economists disagree on most issues, but there’s one thing they all agree on: without a solution to load shedding, no economic strategy is going to work, no matter how fiscally conservative or ambitious. Investors and ratings agencies also agree, as do our politicians, that as long as load shedding persists, a post-Covid-19 recovery is a pipe dream.

But when it comes to solutions to load shedding, there is no sense of urgency. Yes, we now hear Eskom CEO André de Ruyter saying repeatedly and with ever-greater urgency that we need to face the reality that unless we get new generation capacity on to the grid within the next few years, we face increasingly serious load shedding in the coming years. But who is really listening? Seems like the president might be, but his ministers are at sixes and sevens.

It’s like 1998 all over again. In 1998, a white paper on energy was adopted that said a new generation needed to be built urgently, but nothing was done for a decade until then-president Thabo Mbeki apologised to the nation for ignoring Eskom. Two years later, the deal with the World Bank to build Medupi and Kusile in a reckless rush was signed. The sense of déjà vu is overwhelming – not least because we don’t have 10 years this time around. We have at most a year before things start going horribly wrong – maybe more like six months.

It may be time for radical pragmatism: those who resist “divisionalisation” and want to retain Eskom as a fully vertically integrated publicly owned company, may need to accept that this will (unintentionally) reinforce those forces with vested interests in the status quo; and those who want unbundling and privatisation might need to accept that divisionalisation under the auspices of (publicly owned) Eskom Holdings is the more realistic option.

According to a CSIR report and Eskom COO Jan Oberholzer (in a recent radio interview), we face load shedding for at least the next 18 months. Indeed, even if we make the big decisions this year (which is unlikely), 2022 looks to be the most challenging year ever. Even the government’s Integrated Resource Plan (IRP) expects a bad year in 2022, thereafter improving and close toresolved by 2025 if we follow the IRP script, which we are not doing. The IRP stipulates that 1GW PV and 1.6GW wind should have come onstream in 2022, which is not going to happen.

However, even if we did follow the IRP script, it assumes coal plants’ reliability will improve, i.e. that the energy availability factor (EAF – % fleet available at any given time) increases from 67% in 2019, to roughly 76%. In reality, the EAF will probably continue to decrease because the coal plants are old and creaky, and therefore increasingly unreliable (according to various public statements by the Eskom CEO and COO).

The coal plant EAF has been steadily dropping for 20 years, and – according to the CSIR report –dropped from 80% to 67% in just the past three years!

As everyone agrees, urgent steps must be taken this year to bring the cheapest and most reliable energy on to the grid within the next three years. Coal, nuclear and substantial gas infrastructure cannot be brought online within that time frame. Only renewables (wind and solar) can be built that quickly. That is an undeniable fact, no matter what your biases may be. Furthermore, renewables are also the cheapest option and prices are expected to continue to drop.

Although prices for renewables contracted via bid windows 1-3 were clearly high, to constantly refer to this as a reason to deny the reality of rock-bottom prices today no longer makes sense. Further, to think gas can be brought online and at scale in this time frame, at lower prices, is delusional.

Fortunately, the Department of Mineral Resources and Energy finally published in September 2020 its latest so-called Section 34 Determination to bring 6.8GW of renewables, 3GW of gas and 1.5GW of coal on to the grid. Unfortunately, in the gazetted statement, Eskom is specified as the only off-taker, despite references to the contrary by the department’s representatives in various presentations.

If non-Eskom off-takers were allowed, new opportunities would have been created: it would have meant other more financially viable entities can be off-takers, but this would mean Eskom agreeing to  “wheeling” across its grid (for a fee, of course, because it costs Eskom money to maintain the grid). This is an important scenario because Eskom is not trusted by the financial sector as an off-taker due to the dire state of its balance sheet.

The Request for Proposals for the renewables will be published in December 2020, after which bids must be submitted. If the Independent Power Producers office processes these speedily (which is unlikely), new capacity could come online by the end of 2021/22, thus offsetting what could be our worst year ever at exactly the moment we should be emerging from the “pancession”.

The 1.5GW of new so-called clean coal envisaged by the Department of Mineral Resources and Energy, if it can be funded (which is highly unlikely), could take another decade to come fully online (if the Medupi process is anything to go by).

Solar and wind can already produce electricity 40% cheaper than coal, so this is also the best route to minimise the upward pressure on electricity prices. The best estimates of the levelised cost of energy per kWh are wind at 66c, solar at 60c, coal at R1.31 (30% more for “clean coal”) and nuclear at R1.76.

The low price of renewables is important because for years now, Nersa has denied Eskom the right to implement cost-reflective tariffs that cover the “weighted average cost of capital” (i.e. what it costs to fund the assets owned by Eskom) – those who think Eskom wants higher prices to pay for inefficiencies and corruption are wrong. This is a major cause of Eskom’s ballooning debt, now an eye-watering R480-billion.

Following Eskom’s recent court victories, tariffs will now increase to make up for the backlog. Bringing cheaper energy on to the grid helps ameliorate this upward pressure to the benefit of many businesses and households. “Clean coal” will do the opposite.

If we don’t urgently act to bring the renewables online, we can (as per the CSIR document) expect load shedding by 2022 that is 2-3 times worse than what we’re seeing today, and permanent load shedding by 2025 (with as much as 20% of electricity at any one time cut off).

There is, however, a solution. But it will mean creating policy certainty for the first time and bringing to an end the political bickering about renewables. All this could be done at no cost to the state. The easiest way to avoid plunging South Africa into an energy-driven economic crisis that exacerbates the current “pancession” (and also potentially triggers a sovereign debt crisis) is to ensure that a handful of key decision-makers make the right decisions within the next three months. Millions will benefit.

To argue that the energy transition means less jobs is totally wrong. A just transition is a real, tangible, achievable reality that can be unlocked relatively easily.

For the sake of the argument, let’s assume a stretch target of 25GW of new, reliable, fast-deployment generation over the next two to five years – that means solar, wind, storage and gas backup, because no other technology mix can be brought online this fast. This notional stretch target of 25GW is derived from the Meridian/CSIR report that envisages a ramp-up to a build rate of 5GW pa for the next 20 years, hence 25GW for the first five is not unreasonable.

Building out this low-cost generation infrastructure could become the most significant post-Covid-19 economic recovery boost.

It will trigger a huge jobs opportunity in the ongoing construction of more and more solar and wind sites (including self-generation), and in the upstream manufacturing and assembly of components (e.g. wind towers, panels, steel production).

On average, 10,000 jobs are created per GW of renewable energy (and this excludes the upstream jobs that will be created), plus 200 permanent operational jobs per TWh. But once you start building renewable energy, you cannot stop, because what you build this year (say 5GW), must get rebuilt in 20 years. To put this in perspective, there are 56,000 jobs in coal mines that supply Eskom – and they reduce over a long 20-year period as coal mines that service Eskom close (with most of these workers reaching retirement age during this time).

To argue that the energy transition means less jobs is totally wrong. A just transition is a real, tangible, achievable reality that can be unlocked relatively easily.

As far as the manufacturing plants and jobs are concerned, they could be concentrated in (the recently promulgated) Renewable Energy Development Zones to be located near coal mines coming offline where grid capacity is the most substantial (even though the wind and solar resource might not be optimal).

Not only that, with South Africa’s ultracompetitive solar and wind resource, if we can tap into that, we can unlock globally competitive industries in South Africa, and start exporting goods produced on the back of this clean electricity around the world. We will be the preferred supplier compared with those still using dirty fuels to manufacture.

One example: fuels such as ammonia, which is what ships use, can be produced with “green hydrogen”. Green hydrogen is made from running renewable electricity through water using mature electrolysis technologies – a future that Sasol is seriously considering.

Some of the 292 billion litres of water that Eskom used to run the coal-fired power stations in 2019 could be repurposed after the power stations close. According to Tobias Bischof-Niemz, one litre can make 9kg of hydrogen using mature well-known electrolysis technologies. This means 292 billion litres can make 33 billion kg of hydrogen.

The energy content of 33 billion kg of hydrogen is 1,287TWh of energy, or 4,633PJ of energy per annum. South Africa’s coal mines produce 6,100PJ of energy in the form of coal (which uses up two-thirds of this energy to make the final energy that comes out of a coal-fired power station).

The potential is enormous, which is why there are major new green hydrogen projects popping up around the world.

To unlock the potential of a renewables-led industrialisation programme, the following will need to happen:

Customer generation: 5GW

According to the CSIR, the fastest route to new generation, and thus some relief, is to open the floodgates to let customers, mostly businesses, build onsite generation. Right now, anything above 1MW needs a generation licence, which everyone knows is such a hassle to get it is not worth trying.

If this cap was lifted to 10MW, it could bring as much as 5GW on to the system before 2022 (CSIR report). The Department of Mineral Resources and Energy claims it is working towards enabling this in some way, but it seems largely by making the Nersa process more efficient (which is unlikely to be achieved).

Lift cap on renewables generation: 300MW

Existing solar PV and wind farms are licensed to generate a fixed quantity of energy to put on to the grid. They can, however, produce more – between 150MW (according to the Independent Power Producer office) and 300MW according to some experts in the wind and solar environment. This should be a quick win. The department claims there is a process under way to address this.

REIPPP: 7-10GW ‘shovel ready’; can deliver 5-10GW per year (from 2022 forward)

Developers report that they have 7-10GW of “shovel-ready” projects that could come online in the next one to two years… so by mid- to end-2022.

Key adjustments, however, will be necessary to enable this. They include:

  1. An updated IRP in 2021 that includes a commitment to a pipeline of new, bigger auctions under Section 34 Determinations, for 5+ GW of renewables per year, starting in 2021 (on top of the September 2020 Section 34 Determination) and the politically motivated cap on solar PV and wind should be lifted;
  2. Renewable Energy Development Zones: ensure these are established in locations with existing infrastructure that make it easier/faster to build renewables, including (a) grid transmission access, and (b) completed environmental impact assessments. In the short term, these might not be in locations where the best wind and solar resources are available, but where there is existing grid capacity (mainly in Mpumalanga). Longer term, we should build transmission capacity to sites where the best solar and wind resource is available, including establishing Renewable Energy Development Zones in these locations. This is not expensive, but the lead times are very long because they include obtaining servitudes and such over long distances;
  3. Establishment of the publicly owned Independent Transmission, System & Market Operator (ITSMO) as buyer as envisaged by the Eskom Roadmap – for practical purposes, this entity would be owned by Eskom Holdings, with its own board. This would provide a fully financially stable entity to act as the buyer from both Independent Power Producers and Eskom generation (if a decision is taken that enables Eskom to competitively bid for and build renewable energy plants). This would reduce “counterparty risk”, allowing Independent Power Producer developers to apply a lower cost of capital, which means cheaper electricity for South Africans. If Eskom as currently structured is the only off-taker, funders will once again require sovereign guarantees (which may not be forthcoming, given the current fiscal crisis, thus putting the entire programme at risk);
  4. Allowing non-Eskom off-takers (like municipalities and even corporates or community organisations) to buy energy from renewables generators (i.e. bilateral power-purchase agreements) would offset the requirement for sovereign guarantees. But Eskom would need to agree to “wheeling” of its grid (which happens already on a large scale), for a fee; and
  5. Establishment of a “war room” involving the Independent Power Producer office, Nersa, Eskom and the relevant departments: the purpose would be to fast-track approvals across multiple government departments, including the redesign of certain procurement processes, including the frequency and size of each bid window.

Bringing Eskom into the game

The IRP and determinations currently exclude Eskom from the renewables programme, other than as off-taker, although the IRP did acknowledge that a strategy should be developed for Eskom’s participation in new generation (see p.17). This reflects a widespread view within the renewables sector that Eskom’s job is simply to step out of the way. The Department of Mineral Resources and Energy is reconsidering this. New energy generation regulations have been drafted that envisage a role for Eskom and local authorities in renewables generation. This opens up the possibility for renewables to be procured via the REIPPPP mechanism (Section 34 Determinations, Nersa concurrences, Independent Power Producer office RfPs) and via the Eskom-owned ITSMO.

Imagine a total of 5GW of renewables per annum procured via both the ITSMO and the REIPPPP (especially if non-Eskom off-takers were allowed) in accordance with their respective capacities. Depending on how Eskom divisionalises, it could establish a “green genco” owned by Eskom Holdings that could in theory also bid competitively for projects, thus ensuring it achieves the same efficiencies that the private sector can achieve. Ring-fencing the ITSMO from the “green genco” will, therefore, be necessary to avoid conflicts of interest.

Many of Eskom’s existing plants may already be generating electricity that is more expensive than electricity from new solar and wind plants, especially where the EAF is particularly low. One opportunity that Eskom’s newly established Just Energy Transition office may want to consider, is to use the already existing licensed grid connection that these coal plants have to build solar, wind and storage at the same locations to provide low-cost and reliable electricity into the grid. The obstacle might be technology-specific licences, but maybe there is a way around that. If there is, new renewables could be brought on to the grid in this way without a Section 34 determination from the Department of Mineral Resources and Energy. Here’s how:

Experts advise me that within the current legislation and regulations, there are no impediments to Eskom projects which build new generation without a Section 34 determination, where it is aligned with the Integrated Energy Plan (IEP) and shows compliance with the IRP. Section 34 provides for the minister responsible for energy to determine the need for new generation capacity, which would otherwise not have been possible. Section 34 determinations do not, therefore, seek to circumscribe Eskom’s capacity to create new generation capacity in terms of the National Energy Act and Electricity Regulation Act.

To deliver these assets, Eskom could partner with the private sector (giving preference to black-owned South African companies, municipalities and metros) to leverage their capital and project delivery capacity. For this purpose, Eskom could use the treasury regulations (Section 16) that provide an operational framework for structuring public-private partnerships or joint shareholding of special-purpose vehicles. Although not applicable to Schedule 2 public entities such as Eskom, this does provide a framework to work from, with flexibility for entities such as Eskom. Private sector players could compete on both price and percentage of free carry for Eskom in the projects.

Battery storage: 5GW (four-hour duration) by 2023

Similar to solar and wind, costs of storage – utility-scale batteries – have dropped dramatically: 85% reduction since 2010. To enable renewables to provide power on to the grid when it’s needed, there are two key elements: (1) deploying both solar and wind across many geographical locations which often produce at different times; and (2) storage. For the volumes of renewables we are discussing, and following the proportions in the Meridian report, for 25GW of renewable energy the amount of storage we might want on the system may be as high as 5GW (i.e. 5GW batteries that run for four-plus hours).

Storage should be incentivised in all the procurement documentation, no matter via the REIPPPP or Eskom procurement processes (if these occur).

Building on government’s Section 34 Determination of 10 September

On 10 September 2020, Minister of Mineral Resources and Energy Gwede Mantashe welcomed the concurrence by Nersa to a Section 34 Determination that enables the department to undertake procurement of 11.8GW of power, in line with IRP 2019.

Relative to the solutions proposed in this article, this represents a very good first step. However, these are the challenges that will need to be faced to achieve the 25GW stretch target:

  • Insufficient scale: this determination will enable a good portion of the “shovel-ready” renewables projects to be developed. That will help considerably, but relative to the 25GW gap, this is insufficient to resolve load shedding by 2025;
  • Expensive coal: Bid window 4 saw solar PV come in at c. 62 cents/kWh. This provides electricity generation at >40% lower cost compared with a new coal plant. Further, solar and wind technology prices have continued to drop in recent years, meaning the savings by building renewables instead of coal would be even larger today. For reference, the world record for lowest-cost solar, previously held by Abu Dhabi, was recently claimed by Portugal with a solar auction delivering power at 1.31 US cents/kWh (c. R0.22/kWh). South Africa has much better solar resources (i.e. more sun) than Portugal; and
  • Dirty coal: beyond the cost argument and the as yet commercial unviability of clean coal technologies (which only cut out at most a third of emissions and cost 30% more), there is also the issue of emissions from coal. South Africa is seeking to attract financial support to address financing issues with Eskom and the power sector generally. This support is premised on evolving South Africa’s power system to modern and low-carbon energy sources, i.e. renewables. New construction of coal would compromise the potential large-scale investment referred to by President Cyril Ramaphosa at the UN Climate Summit in September 2019, when he suggested a transaction that traded voluntary accelerated decarbonisation for $11-billion of concessional climate finance.

In short, the critical next steps should be:

  1. Customer generation: lift the cap on self-generation from 1MW to 10 MW;
  2. Enable existing renewables plants to maximise output: lift all restrictions on output;
  3. REIPPP: while the September determination is a good start, it is not ambitious enough and needs to ensure that the off-taker ecosystem is diversified to include non-Eskom off-takers that may not require sovereign guarantees;
  4. New generation regulations: speed up the promulgation of these regulations to diversify the role that Eskom, public entities, local authorities and communities can play in renewables generation;
  5. Renewables-led industrialisation: To encourage developers to build up local supply chains, and invest in other local operations which will also enable them to bring their costs down further, the government should signal an expectation of a continued cadence of Section 34 Determinations and procurement by Eskom over the next five to 10 years targeting, for example, 40GW of solar and wind by 2030, with an annual build target of 5+ GW of new capacity added per year from 2022 onwards;
  6. Implement the adjustments for enabling the delivery of the lowest-cost energy possible and the most jobs, plus rapid decarbonisation: these include establishing Renewable Energy Development Zones in the right places; incentives for Eskom to deliver transmission connections in a timely manner; establishing the ITSMO; enabling purchase power agreements with non-Eskom off-takers; streamlining the Independent Power Producer procurement process;
  7. Enabling Eskom to use sites where coal-fired power stations are shutting down to establish large-scale renewables generators: this is not impacted by Section 34, and could possibly allow Eskom to use existing licences and existing grid capacity; and
  8. Storage: the 0.5GW of storage (specified in the September 2020 determination) is a good start, but insufficient. Progress towards 5GW of storage will be required, which could possibly be slightly reduced if gas backup proves to be as viable and more affordable.

If there is one issue that could finally finish off the ANC as a leading political force, it would be the way Eskom is handled by the government. Eskom still has the leadership and capacity to do what is necessary. But only if the government, at no additional financial cost to the state, creates the appropriate policy framework – and this, at sonic speed.

This article was first published on on 30 September 2020.