You are here


The troubled recent history of nuclear power in South Africa

Nuclear Monitor Issue: 
Greenpeace South Africa

Six days after the nuclear catastrophe at Fukushima in Japan in March 2011, South Africa’s Minister of Energy Dipuo Peters declared her country’s intention to add 9,600 MW of nuclear electricity - or six new nuclear reactors. On September 15 she said she had signed off on a proposal for new nuclear power plants and said it would be presented to cabinet soon. Peters said she expects the cabinet to decide on the plan by the end of this year and the bid process to start early in 2012. The last attempt to build a nuclear plant, led by state-owned power utility Eskom, was scratched on funding woes.

Speaking at the second regional conference on energy and nuclear power in Africa in Cape Town on May 30 this year, Ms Peters went even further, trumpeting the development of a nuclear-export market to the rest of Africa, supported by both the International Atomic Energy Agency, and the African Union.

South Africa spent 13 years pursuing the Pebble Bed Modular Reactor, wasting billions of rands in the process (R9-billion was spent on research and development and another R22-billion would have been needed to complete a demonstration model) as investors across the world shied away from having anything to do with it. Eventually the state cancelled the project and wrote off the monies it had spent. The government and its wholly owned power utility Eskom remain hell bent on securing what it believes will be a cheap and sustainable nuclear solution for its energy supply crisis.

By 2006, South Africa was beginning to run short of power generation capacity. It was clear that the PBMR would not be available to order for a long time. Eskom began to talk about ordering ‘conventional’ nuclear power plants. First in line were the EPR supplied by the French company, Areva and the AP1000 supplied by the Japanese owned company, Westinghouse. Eskom’s implication was that such designs were well proven. In fact, at that point, only one order had been placed for an EPR and none for the AP1000. By 2011, there were four orders for EPRs, two for China, one for France and one for Finland and four for AP1000s, all for China. None of these orders were in service by 2011 and the two EPR orders for France and Finland were seriously over budget and late.

In 2006, the South African government forecast that a new unit could be on-line between 2010 and 2012. By mid-2007, Eskom was targeting construction of 20,000 MW of new nuclear capacity by 2025, although completion of the first unit had slipped to 2014. It expected an overnight construction cost of US$2,500/kW. (Overnight cost is the cost of a construction project if no interest was incurred during construction, as if the project was completed "overnight.")

In January 2008, Eskom received two bids in reply to its call for tenders from November of the previous year for 3,200-3,400 MW of new nuclear capacity in the near term and up to 20,000 MW by 2025. One bid was from Areva for two EPRs (plus 10 more for the long-term) and the other from Westinghouse for the three AP1000s (plus 17 more in the long term).

It was later reported that the bids were for around US$6,000/kW (overnight) – more than double the expected price. It was therefore no surprise when Eskom abandoned the tender in December 2008 on the grounds that the magnitude of the investment was too much for it to handle. This was despite the willingness of Coface, the French government’s loan guarantee body, to offer export credit guarantees and despite Areva’s claims that it could have arranged 85% of the financing.

Eskom in crisis
Three weeks into January 2008, Eskom had hit a brick wall. It could no longer meet all the country’s electricity demands without melting the national grid. Eskom turned to the bulk users, and appealed to them to ration their demand. Even so, for some months the country faced a series of electricity outages (euphemistically called “load shedding”). Not only was this a blow to businesses, agriculture, schools, hospitals and households, but it coincided with global recession.

Eskom had also run out of money and its credit ratings were reduced. Eskom could no longer afford to invest in new infrastructure, without massive extra income. It would take three years before it could make new orders, and until then the board was saying no to new investments. The biggest blow to the nuclear industry was the decision to scrap the tender process for Nuclear-1, the first of a number of new large-scale reactors. The government had to inform vendors Areva and (Toshiba-owned) Westinghouse that their bids would not be considered for the meantime. The policy was not being suspended, but the orders were temporarily shelved.

Newly appointed CEO Brian Dames tried to rebuild Eskom’s reputation and finances. A big hurdle was the steady loss in Eskom’s credit ratings. Eskom hoped to raise electricity tariffs substantially, despite this being opposed by the trade union movement and other sections of civil society. The National Energy Regulator reduced Eskom’s application for 35% increases for three years to 25%, amounting to a doubling of tariffs over the same period, hitting poor and middle-class households, who objected strongly to the sweetheart commercial deals which Eskom had made in the past with smelters and other large users to be charged minimal tariffs.

The government then guaranteed Eskom’s massive investment in two giant coal-fired power stations. Medupi, the first of the two to be built, will be funded by the World Bank despite the enormous carbon emissions the 4,800 MW plant will produce. The loan of US$3,75 billion, was strongly opposed by local NGOs, and even caused countries like the Netherlands, Britain, the US, Norway and Italy to abstain from voting at the bank’s decision making committee.

To help Eskom get funding for its future nuclear power stations, companies like Areva have said they will help to intercede with the French government to release development finance. The potential Chinese bidders for Nuclear-1 (China Guangdong Nuclear Power Group) have linked up with the Standard Bank of South Africa, 20% owned by a Chinese bank (Industrial and Commercial Bank of China), in order to assist Eskom to purchase future reactors.

As a result, Eskom’s financial woes are less of an obstacle to re-launching the bids for Nuclear-1.

2010 onwards
The South African government seemed to assume that cheap reactors can be found, if only they could be identified. This led it to look at a design offered by Korea, which had won four orders for the Unityed Arab Emirates (UAE) with a bid worth about US$4,000/kW (overnight costs), well below the levels offered by Areva and Westinghouse, but 60% above the level assumed by the South African government in 2006.

Despite the precariousness of the Korean option, the South African government has had discussions with the Korean government about the supply of such reactors.

The other design being considered by South Africa is the one that makes up the majority of Chinese orders. China dominates the world market for nuclear power plants accounting for 25 out of 38 of the reactors on which construction has started since January 2008. Of the 25, 19 are supplied by Chinese companies and this CPR-1000 design is based on the design China imported from France in the 1980s. This is the same design as is already installed at Koeberg. Some updating will have taken place, for example taking advantage of better IT equipment, but it is clear that it is fundamentally a 40 year old design. The South African government has also been talking to the Chinese government about importing such reactors.

However, a number of assumptions seem to underlie this attempt:
• That the reactors would be much cheaper than more modern designs, partly because they are older and partly because they would be manufactured in China;
• That China has the spare component manufacturing capacity to export plants; and,
• That the NNR would be comfortable licensing a design that fell well short of the requirements of Western regulators, for example on protection against impact by aircraft.

Eskom seems remote from this process and it is not clear whether it supports the idea of importing older technology. As with its reservations with the PBMR, Eskom could be uncomfortable raising any concerns about South African government policy.

The lessons from the Fukushima disaster in March 2011 have yet to be fully identified, but there does seem to be a strong probability that older designs will be seen, worldwide, not just in the West, as inadequate for new orders. In particular, designs with a greater level of ‘passive’ safety – ones that in an emergency situation do not require the operation of engineered safety systems to bring them to a safe condition – will be required. Even the French EPR does not incorporate strong passive safety features and the Chinese and Korean designs certainly do not have passive safety.

The new call for nuclear tenders
The call for tenders expected for 2012 is based on the Integrated Resource Plan 2010. The rationale for the integrated resource planning process is that it should identify the lowest cost way to meet electricity demand by considering all resources including energy efficiency measures. The plan includes 9,600 MW of new nuclear capacity to be completed between 2023 and 2030. Whether this nuclear capacity really represents the least cost way of meeting demand depends on the accuracy of the assumptions made on the cost.

The IRP 2010 bases its assumptions on a report commissioned from the US Electric Power Research Institute (EPRI, 2010), a US research organisation funded primarily by US electric utilities. Nuclear power costs are dominated by the costs associated with the construction of the plants, the overnight cost of construction and the cost of borrowing, which is related to the discount rate. For the construction cost, the EPRI report gives an overnight cost of R28,375/kW for an Areva EPR and R33,235/kW for a Westinghouse AP1000. If we assume an exchange rate of US$1=R6.75, this equates to about US$4,200/kW and US$4,900/kW. It is hard to understand why the South African government should assume costs that are only 70-80% of the prices bid two years earlier. There is certainly no evidence that estimated nuclear costs have gone down since then.

The discount rate of 8% adopted by the South African government also appears too low. For example, the UK government assumed a discount rate of 10% in 2008 when it assessed the economics of nuclear power. The discount rate is effectively a tool to allocate the limited quantity of capital available as profitably as possible. It should ensure that only projects that achieve the given rate of return on capital – the discount rate – are pursued. If nuclear power is assessed using too low a discount rate, it is likely that relatively unprofitable projects will be pursued at the expense of more profitable projects. The use of too low a discount rate is particularly serious because one of the key reasons the previous tender failed appears to have been because affordable finance was not available. Cape Times reported that Rob Adam, CEO of Necsa, has said:

‘The country’s nuclear programme had been canned in 2008 because “we couldn’t get a bank to lend the money for long enough. Commercial banks’ time frames are too short. So now the vendor must come with a bank or financial institution”, and South Africa would repay this over time.’

It appears the South African government did not learn from the previous tender when it assumed far too low a construction cost and proceeded with a call for tenders that had to be abandoned because the prices bid could not be financed. The government also seems heavily involved with the process, with ministers and sometimes the president conducting negotiations and signing agreements with governments of potential suppliers. These efforts have been particularly intense with France with whom an undertaking to explore an intergovernmental agreement on spent-fuel management, co-operation between the countries’ nuclear safety authorities, and implementation of the agreement on nuclear R&D between the Necsa and its French counterpart have been agreed.

Sources: This article (except the lead) is reprinted from a new Greenpeace South Africa report, called 'The true costs of nuclear power in South Africa'. It is available at:
Contact: Greenpeace South Africa, 10A and 10B Clamart House, Clamart Road, Richmond, Johannesburg, South Africa

S-Africa: Eskom: record loss; PBMR "indefinitely postponed"

Nuclear Monitor Issue: 
WISE Amsterdam

Eskom, South Africa's state-owned utility, has reported a record annual loss and has warned of a funding gap for an expansion program needed to prevent a repeat of the blackouts the country experienced in 2008. The company, which supplies about 95% of South Africa's electricity and more than 60% of Africa's, reported a loss of 9.7 billion rand (US$ 1.25 billion) for the year that ended 31 March. In the previous year, Eskom made a loss of 210 million rand (US$ 27 million).

The utility foresees a funding shortage of some 80 billion rand (US$ 10 billion) for its expansion program aimed at reducing the risk of power shortages. In January 2008, as domestic supply reached its limit, South Africa suffered crippling blackouts and electricity exports to neighbouring Botswana and Zimbabwe were stopped. This led to a wider grid failure affecting Zambia.

In August 2009, Bobby Godsell, chair of the utility, noted, "We need to mobilize greater equity resources to fund the build program. The government has already provided 60 billion rand (US$ 8 billion) in a loan with equity characteristics. Government revenues are likely to be severely constrained in the near future. We need to find other sources of expansion funding, perhaps in the form of a development bond that will enable South Africans to invest in the expansion of our country's energy system."

"The capital costs of our build program have escalated considerably," Godsell added.

"Prior to the recent global economic crisis, construction costs were escalating worldwide and across all industries. The global recession has created new market circumstances."

And the nuclear program?

In early 2007, Eskom's board approved a plan to boost electricity output to 80 GWe by 2025. This included the construction of 20 GWe of new nuclear capacity, which would see the contribution of nuclear energy grow to 25% from the present 5%. The plan for the nuclear new-build program would kick-start with up to 4 GWe of pressurized water reactor (PWR) capacity, to be constructed from about 2010 with commissioning in 2016. Five sites in the Cape Province were under consideration, although the most likely initial site (Nuclear-1) would be that of Koeberg, the site of South Africa's only existing nuclear power plant. The Nuclear-1 project was established after the very ambitious scenario for development and construction of the Pebble Bed Modular Reactor (PBMR) failed to meet even the most modest time schedule.

Having already made "considerable progress" in the process to procure a PWR, Eskoms board of directors decided in December 2008 not to proceed with the project due to ‘the magnitude of the investment’; the companies own financial constraints and the global economic situation. The investment was increasingly impossible to justify, with a plunging rand, global lines of credit frozen, and a new government with potentially different priorities.

On September 11, addressing the World Nuclear Association Annual Symposium in London, UK, Jaco Kriek, CEO of the PBMR company, said that South Africa's pebble bed modular reactor (PBMR) Demonstration Power Plant (DPP) project has been indefinitely postponed due to financing constraints. He said the PBMR company has had to adopt a new business model "to reduce the funding obligations on the South African government."

Sources: World Nuclear News, 28 August 2009 / Nuclear Monitor 681, 16 December 2008: ‘Eskom cancels PWRs; major blow to nuclear expansion’ / World Nuclear News, 11 September 2009
Contact:  CANE, Coalition Against Nuclear Energy South-Africa, Tel: +27-72 628 5131, Email: