- Hundreds of pages of "briefing notes" reveal bidders in the emergency power programme warned the government rules would favour fossil fuels and leave renewables out in the cold.
- Nonsensical rules, they charged, would also result in higher electricity prices for consumers.
- As they predicted, the cheapest price was R1.46/kWh, more than double the 62c/kWh that Eskom secured in 2018 for solar- and wind-powered IPPs.
When the Department of Mineral Resources and Energy launched its urgent tender last August to mitigate loadshedding, it had one claimed goal: get cost-effective power to Eskom as quickly as possible.
But when hopeful bidders looked at the rules of the Risk Mitigation Independent Power Producer Procurement Programme (RMI4P), it seemed to many there was another agenda at play: channel the lion's share of the multibillion-rand, 20-year contracts to the budding gas industry.
"Is it really the department's intention to ensure that renewables [and] batteries ... are excluded from participating in this tender?" one exasperated bidder would ask officials in the months that followed.
As part of the tender process, the department distributed questions, answers and clarifications in the form of "briefing notes" to all prospective bidders.
The most common answer they received? The RMI4P is "technology agnostic... There is no exclusion of any qualifying technology."
AmaBhungane's investigation suggests this was a half-truth at best. Instead, the department, using requirements from Eskom as client and adding rules of its own, pushed gas ahead of renewables - and the Turkish-led Karpowership consortium ahead of everyone else.
It said in a written response this allegation was "baseless and misplaced" and the tender was "fair, transparent and competitive" and not designed with any specific supplier in mind. As evidence of the "level playing field", it pointed to "the hybrid nature of the selected projects" - many combine renewables, batteries and fossil fuel - and the fact that 28 bids were received.
The department added the same boilerplate answer it had given many frustrated bidders: "The [tender] design was technology agnostic to meet system requirements as identified by [Eskom]."
In detailed answers, Eskom defended the requirements it had conveyed to the department, but made clear some of the department's rules were not based on its requirements.
A 20-year emergency
The RMI4P was designed to provide an urgent solution to Eskom's chronic power crisis. (Hence "risk mitigation" makes up the first part of the tender's awkward name.)
Eskom predicts it will have a shortage of up to 6 000MW for the next five years. Currently, that shortfall is being plugged, if at all, by peaking plants that run on diesel. But Eskom would desperately like to find a cheaper solution.
Enter the RMI4P, which aims to secure up to 2 000MW of power from independent power producers.
When the department issued a request for information to sound out the market in December 2019, it asked potential bidders to provide different tariff estimates "in the event that the tenure of the [contract] is for three years; five years, 10 years, 15 years or 20 years".
But by the time the tender was issued in August 2020, the department had decided it was only interested in procuring for 20 years. This was "non-negotiable", it told bidders.
This should have been the first sign the department had objectives beyond solving Eskom's immediate crisis.
The 20-year tenure of the power purchase agreements sits uneasily with the fundamental design features of the RMI4P: the specific hours during which bidders will need to provide power. The tender calls for projects that can provide 100% of their capacity between 05:00 and 21:30 on any given day.
While this fitted the current loadshedding risk profile, there was no reason to believe this would be a logical requirement for 20 years, one bidder pointed out in a briefing note.
It might well be, the bidder said, that 10 years from now enough renewables would have been built to make the daytime capacity sufficient with the risks to the system actually occurring after 21:30.
In other words, the risks that need mitigation will inevitably change before the end of this "risk mitigation" contract ends.
Eskom said in written responses to amaBhungane that although it asked for some criteria to be included in the RMI4P tender, it "did not specify that the [contracts] have to be for 20 years".
Longer contracts are riskier for Eskom, especially when electricity is generated by burning fossil fuel. While renewables offer largely predictable tariffs, fuel-based projects shift the risk to Eskom.
This is because fuel is treated as a "pass through" cost, meaning Eskom is potentially exposed to both fluctuating fuel prices and the rand-dollar exchange rate: if the price of fuel goes up, Eskom (and the consumer) pays more.
"Any exposure to commodity and exchange rate fluctuations does result in risk and opportunity. Eskom will employ mechanisms such as hedging, if available, to mitigate against the risk," it said.
"Eskom ... acknowledges that there is risk in these programmes that ultimately are borne by consumers... We are concerned that consumers are having to bear all these risks through price increases."
Another feature of the RMI4P amplifies this risk enormously. The power purchase agreements will guarantee a minimum off-take by Eskom of 70% of all the power the bidder could provide - whether Eskom needs it or not. For a programme meant to provide power on demand to prevent loadshedding, this is, arguably, illogically generous.
"It is not the worst thing to have powerships for three to five years - at a low factor of 15%. Anyone can be happy with that. But not with a guaranteed 70% for 20 years. That's too much gas," one bidder told amaBhungane.
And there is a provision for the contracts to last even longer. The RMI4P allows the department to add up to 10 years onto the contract, if a "change in law" forces the company selling the power to incur extra capital or operational expenses.
If stricter environmental laws are imposed on fuel-based projects, a 20-year contract could become a 30-year contract, "so as to put the seller in the same overall economic position it would have been but for such change in law".
Clauses like this, combined with the fuel and currency risk, should have given renewable projects an edge. But other aspects of the tender would almost guarantee gas the most significant seat at the table.
First, there is the dispatchability rule, based on Eskom's requirements, which demands that the full contracted amount of power be "dispatchable" within minutes of Eskom calling for it. The bidder should then be capable of getting turned off just as quickly when Eskom says stop.
And this service must be available for 16.5 hours per day between 17:00 and 21.30.
"These projects will be available to generate energy output when called upon [to avoid loadshedding] and provide ancillary services to reduce the significant cost of using diesel that is currently the only option available to [Eskom]," the department said.
This takes 100% renewable projects out of the running completely; their only option is to add batteries or fuel so that power can be turned on at the flick of a switch, come rain or shine.
If the RMI4P's only job was to replace Eskom's diesel-fired peaking plants with a cheaper, somewhat greener option, then fair enough. But the design of the RMI4P effectively means the only projects that qualify are ones that can both act as peaking plants and deliver baseload power for the next 20 years.
And few projects can do both.
One way in which many renewable projects were effectively disqualified or discouraged from bidding was the mandatory "reliability run". This involved projects having to demonstrate they can provide 100% of their installed capacity for the full 16.5 hours per day for 15 consecutive days. This is a nearly impossible feat for a battery-renewable combination.
As a result, many bids looked like Frankenstein monsters, with fossil fuel-burning components bolted on to meet the dispatchability rule.
"Our modelling shows that if [Eskom] is able [reasonably] to accommodate a very small relaxation in technical requirements then a [solar and battery] solution can provide a tariff below R1 and with savings to the taxpayer of R10 billion/50MW project over 20 years," one bidder told the department.
Energy analysts, such as University of Cape Town professor Anton Eberhard, argued it was far more cost-effective to balance the grid as a whole, rather than set absolute dispatchability requirements for each individual generator.
Carefully managed, renewables can provide the bulk of the power needed at a cheaper rate. The most recent solar and wind projects contracted by Eskom supply power is at around 62c/kWh. The cheapest project on the RMI4Ps list of preferred bidders came in at R1.46/kWh.
The strict dispatchability rule leads to the second problem: oversizing.
As another potential bidder argued, in order for renewable projects to deliver their contracted megawatts throughout the day, they would need to be "oversized". In other words, for a solar farm to deliver the contracted 50MW when the sun is less than optimal, you may need to build a 100MW plant coupled with a massive battery.
"After massive modelling efforts, it is clear that the reliability run [to test dispatchability] effectively precludes renewables plus batteries [because of the requirement for oversizing and therefore tariff increase].
"There seems to be no clear basis for this... A small relaxation in the availability requirements [to, say, 90%] ... will provide for significantly lower tariffs without compromising the system in any way," the bidder argued in an October email to the department.
Its response was to remind the bidder the RMI4P was "technology agnostic", provided it was "dispatchable, flexible and reliable generation".
The disappointed supplier did not submit a bid.
One solution to the "oversizing" problem would have been to let the projects sell any excess power to private buyers.
While the department did not ban this outright, it clearly dissuaded bidders, saying in a briefing note suppliers could not "sell or deliver ... or make available any capacity to any third party" without written consent from Eskom and it.
"This will be dealt with on a case by case basis... Neither [Eskom] nor the [department] are inclined to consider such requests at the moment."
These kinds of responses led a riled bidder to ask: "Given the significant cost and effort of preparing a bid, and to determine whether it is worthwhile to prepare bids on this proposed technical solution ... can the department confirm it understands the technical limitations of a battery renewable-based solution, and can it unequivocally confirm that it will not, a priori, dismiss any battery renewable-based bids as not being dispatchable and non-compliant?"
No, no and no
There were other rules too that hamstrung renewable bidders: generation units and batteries had to be "co-located", there would be no "wheeling" by transporting power through the Eskom grid, and batteries could not charge off the grid when there was spare capacity.
At the bidder's conference in September, one potential bidder pointed out that this was at odds with Eskom's own use of the enormous Ingula pump-storage plant: "None of Eskom's large energy storage systems are co-located with Eskom's generation assets, and yet these storage assets constitute the most reliable and dispatchable assets held by Eskom - as long as there is surplus energy available in the system to top up this storage."
Eskom defended imposing these rules on the tender, even if they frustrated or eliminated the renewable sector. This was, it told us, to prevent "arbitrage" where a company could buy power cheaply from the grid and sell it back to Eskom at a higher price.
"Future batteries may be treated differently," Eskom said. "Although Eskom did not specifically ask for co-location ... the rule for not taking power from the grid was required. Apart from the arbitrage issue, the complexity in the reconciliation required and the wheeling arrangements to be put in place would make this difficult to administrate."
This arguably might have made sense for a short-term emergency tender, but not for a 20-year programme, given that the ability to administer and audit such inputs and outputs will be key to Eskom's stated task of establishing an independent grid operator.
Another rule required contractors to show they have a track record of delivering similar sized projects that were at least two years old, a clause that one bidder described as "exceptionally onerous and excluding".
The contractor noted although "the [battery] market has grown a lot over recent years ... there were few [over 20MW battery] projects being constructed in 2017".
"In practical terms: it excludes almost all contractors that have a presence in the South African market and all projects proposing [solar and battery]," the contractor complained.
Taken together, these rules set by Eskom and the department hamstrung most renewable projects.
As one prospective bidder put it: "As it currently stands, it will be technically challenging for wind and/or [solar] to participate in the tender... [B]y allowing [renewable energy] to participate, a lower [average cost] will emerge with lower tariffs for the country: i.e.: value for money."
And as another bidder put it, projects that combined renewables and batteries would either be "disqualified from the start ... or will price themselves out of the market due to excessive over-sizing...
"Is it really the department's intention to ensure that renewables plus batteries [without ANY fossil fuel inputs and associated carbon emissions] are excluded from participating in this tender?"
This bidder, it turns out, would go on to be selected as one of the preferred bidders. But only after adding a diesel-burning component to its bid.
Give us gas
Beyond having a distinct anti-renewable bent, the RMI4P was very much geared towards liquified natural gas (LNG) - which Karpowership proposes to use - rather than other fuels.
One example: If you were bidding with a reciprocating engine or gas turbine you would need to ensure they could burn two types of fuel, one of which had to be LNG "without the need for major conversions or refurbishments in the future". If you were planning to burn LNG, this requirement would fall away.
The absurdity of this clause is that LNG is not widely available in South Africa. Although significant gas fields has been discovered offshore, it will be years before its product is commercially available.
Yet, when one bidder asked if the department would accept diesel during "an initial phase until natural gas will be available", it said "no".
"There is no exclusion of any qualifying technology, however, one of the objectives of the [RMI4P] was to reduce the overreliance on diesel as a result the use of diesel will not be allowed in the programme even in the 'initial phase'," it wrote in an October briefing note.
A month later it had to backtrack: "We hereby correct the previous reply... Diesel is not specifically excluded."
The about-turn is a meaningful one. On the understanding that diesel was not allowed, a project that needed a fossil fuel stopgap would likely have relied on liquid petroleum gas (LPG), which is locally available but far more expensive. When diesel suddenly becomes an option it completely changes the economics of the project.
Currently, of the eight projects selected as preferred bidders, three will use diesel to supplement renewables and batteries, while one will use LPG.
But that could still change. Included in the proposed contract is a clause which allows the department to engineer a switch to LNG: "If natural gas or liquefied natural gas [LNG] becomes available or is reasonably likely to become available for utilisation in the project, the department shall have the right to propose a variation to ... the [contract] by serving a written notice on the seller indicating which natural gas or LNG sources the department requires the seller to consider for utilisation as a fuel source for the project."
Whether that would allow the department to cherry-pick a specific LNG supplier is not clear. However, in an interview with Chris Yelland, department officials laid out their vision of how gas-fired powerships would open the door for a domestic LNG market.
"Once we have [Karpowership's infrastructure] at the ports, we will have the opportunity to bring in LNG, and we can then start an engagement on how we can utilise these facilities, and the gas, in other sectors of the economy.
"So, I think this is a good start, with infrastructure that can be used as a stepping stone for the country," Maduna Ngobeni, the chief operating officer of the department's independent power producer office, told Yelland.
The department's deputy director-general, Jacob Mbele, agreed: "Yes, we currently don't have gas in South Africa. But there are explorations that are happening, and there are findings. It is likely that, in the future, the gas for these powerships will come from local fields."
The integrated resource plan, the country's energy road map, envisages adding 3 000MW of gas to the grid by 2030. Any gas-powered projects signed under the RMI4P will be added on top of this because the RMI4P falls under the plan's "other" category, which was envisaged to be taken up by projects like biomass and co-generation.
While it is widely understood that this forms part of a formal plan to develop our gas industry, the department's choices raise the suspicion there is also an informal plan to develop a new generation of politically connected gas tycoons.
Last month, DNG Energy, a rival gas bidder, filed an explosive case against the department, Mineral Resources and Energy Minister Gwede Mantashe, and all the companies selected as preferred bidders.
Chief executive Aldworth Mbalati said in court papers he believed shadowy political figures used "undue influence" to ensure that two-thirds of the contract was awarded to Karpowership SA.
As a gas bidder, he did not take issue with how the renewable industry was systematically excluded. Instead, he focused on how the RMI4P was seemingly designed and then manipulated to ensure that one bidder - Karpowership - came out on top.