• South Africa has two nuclear reactors generating 5% of its electricity.
  • South Africa's first commercial nuclear power reactor began operating in 1984.
  • Government commitment to the future of nuclear energy is strong, with firm plans for further 9600 MWe progressively in the next decade, but financial constraints are severe.
  • Construction of a demonstration pebble bed modular reactor has been cancelled.

Electricity consumption in South Africa has been growing rapidly since 1980 and the country is part of the Southern African Power Pool (SAPP), with extensive interconnections. Total installed generating capacity in the SAPP countries is 54.7 GWe, of which around 80% is South African1, mostly coal-fired, and largely under the control of the state utility Eskom.

Eskom supplies about 95% of South Africa's electricity and approximately 45% of Africa's. Of its total installed net capacity of 40.5 GWe (44.2 GWe gross), coal-fired stations account for 34.3 GWe and nuclear 1.8 GWe2. Early in 2008, demand in South Africa was uncomfortably close to thisa. In 2008, Eskom power stations produced 230.0 billion kWh (TWh) of electricity (out of total South African electricity production of 239.5 TWh), of which the Koeberg nuclear plant generated 12.7 TWh – about 5.3% of total South African generation3.

In 2014 the country produced 253 TWh, this being 232 TWh from coal, 14 TWh from nuclear and 4 TWh from hydro. That year it imported 11 TWh and exported 14 TWh. Consumption was about 4800 kWh per capita.

Over the five years to March 2013, Eskom planned to spend R385 billion (around US$ 50 billion) on new capacity – mainly coal- and gas-fired plants, as well as on returning mothballed coal-fired stations to service. Eskom said the country needs 40 GWe of new generation by 2025, about half of which should be nuclear. In the meantime the country remains heavily dependent on coal, with power plants built near the mines, and the two largest coal-fired plants in the world under construction – 4800 MWe each. Also the country gets 40% of its oil/gasoline needs from coal-to-liquids plants.

In October 2010, the Department of Energy released its draft Integrated Electricity Resource Plan (IRP) for 2010-2030. The IRP outlines the country’s electricity demand, how this demand might be supplied, and what it is likely to cost. Its balanced scenario represents the best trade-off between least-investment cost, climate change mitigation, diversity of supply, localization, and regional development. The IRP requires 52 GWe of new capacity by 2030, assuming 3.4 GWe of demand-side savings. After public consultation the IRP was revised early in 2011 and passed by cabinet in March. According to this scenario, South Africa’s generation mix by 2030 should include: 48% coal; 13.4% nuclear; 6.5% hydro, 14.5% other renewables; and 11% peaking open cycle gas turbine. Although nuclear is included in the energy mix only from 2023, a decision on this "must be finalized as quickly as possible" and a procurement process set up. At least 9.6 GWe new nuclear capacity by 2030 is included in the plan confirmed in mid-2011, significantly less than the 2007 target. In December 2013 the projected 2030 demand was reduced by 6600 MWe to no more than 61.2 GWe.

In the May 2011 budget speech the energy minister reaffirmed that 22% of new generating capacity by 2030 would be nuclear and 14% coal-fired. The budget also provided R586 million ($85 million) for the Nuclear Energy Corporation of South Africa (Necsa) "to continue with its central role as the anchor for nuclear energy research and development and innovation."

The main discussion in the country has been on the need for base-load capacity, rather than specifically nuclear. The IRP remains as policy though the schedule has slipped.

Operating South African power reactors

Reactor Type Net capacity First power Planned closure
Koeberg 1 PWR 930 MWe April 1984 2024
Koeberg 2 PWR 900 MWe July 1985 2025
Total (2)   1830 MWe    

Nuclear industry development in South Africa

South Africa's main coal reserves are concentrated in Mpumalanga in the northeast, while much of the load is on the coast near Cape Town and Durban. Moving either coal or electricity long distance is inefficient, so it was decided in the mid-1970s to build some 1800 MWe of nuclear capacity at Koeberg near Cape Town.

The Koeberg plant was built by Framatome (now Areva) and commissioned in 1984-85. It is owned and operated by Eskom and has twin 900 MWe class (970 & 940 MWe gross) pressurised water reactors (PWRs), the same as those providing most of France's electricity. Stress tests similar to those in the EU were carried out in 2011 with IAEA help. The government plans to extend Koeberg’s operating life from 30 to 40 years, and Eskom solicited tenders for six new steam generators to be installed at Koeberg about 2017-18, aligned with planned maintenance. In August 2014 it awarded the contract to Areva, despite protests from Westinghouse.

The government announced early in 2006 that it was considering building a further conventional nuclear plant, possibly at Koeberg, to boost supplies in the Cape province.

Then early in 2007, the Eskom board approved a plan to double generating capacity to 80 GWe by 2025, including construction of 20 GWe of new nuclear capacity so that nuclear contribution to power would rise from 5% to more than 25% and coal's contribution would fall from 87% to below 70%. The new program would start with up to 4 GWe of PWR capacity to be built from about 2010, with the first unit commissioned in 2016. The environmental assessment process for the so-called 'Nuclear-1' project considering five sites, and selection of technology was to follow in 2008. Areva's EPR and Westinghouse AP1000 were short-listed. Areva headed a consortium of South African engineering group Aveng, the French construction group Bouygues and EDF which submitted a bid to supply two 1600 MWe EPR units. Westinghouse matched this with a bid of three 1134 MWe AP1000 units. The Westinghouse-led consortium included The Shaw Group and the South African engineering firm Murray & Roberts.

Areva and Westinghouse also offered to build the full 20 GWe – with a further ten large EPR units or 17 AP1000 units by 2025. This would have been coupled with wider assistance for the local nuclear industry, in the Westinghouse case including development of the Pebble Bed Modular Reactor (Westinghouse was an investor in the PBMR company and also sponsored the design in the USA – see section on PBMR below).

However, in December 2008, Eskom announced that it would not proceed with either of the bids from Areva and Westinghouse, due to lack of finance, and the government confirmed a delay of several years4.

Proposed South African power reactors

Reactor Type Gross capacity Construction start First power
Thyspunt and/or Duynefontein
(total about 8)
9600 MWe   2024-2030

In the 2011 Draft Integrated Electricity Resource Plan for South Africa – 2010 to 2030 (IRP),5 nuclear prospects were revived, for 9600 MWe, supplying 23% of the electricity. In November 2011 the National Nuclear Energy Executive Coordination Committee (NNEECC) was established as the authority for decision-making, monitoring, and general oversight of the nuclear energy expansion program. An IAEA Integrated Nuclear Infrastructure Review (INIR) was carried out in 2013.

Although the draft IRP included six new 1600 MWe reactors coming online in 18-month intervals from 2023, Eskom has said that it would be looking for lower-cost options than the earlier AP1000 or EPR proposals, and would consider Generation II designs from China (perhaps CPR-1000) or South Korea (perhaps OPR). The capital cost per installed MWe of a CPR-1000 was said to be about half that of an AP1000 or EPR. Before approval, a safety report for the selected specific design must be submitted to the National Nuclear Regulator for evaluation and approval and a nuclear installation licence obtained. Following the Fukushima accident it is likely that a Generation III design will be favoured.

Early in 2011 Areva stepped up its involvement with the Nuclear Energy Corporation of South Africa (Necsa), and early in 2013 Rosatom declared its interest in bidding. Bids were expected to be called early in 2014 so that the contractor/vendor could be on site in 2016, with a view to 2023 operation of the first unit. Initially about 30% local content was expected in the project, rising to 40% later.

In November 2013 Necsa signed a broad agreement with Russia's NIAEP-Atomstroyexport and its subsidiary Nukem Technologies, to develop a strategic partnership including nuclear power plants and waste management, with financial assistance from Russia. It is uncertain just what this means, but Rosatom said that it “offers South Africa to build the entire process chain of NPP construction and operation.” “The strategic partnership implies joint implementation of the national nuclear power development program of South Africa. The key project is construction of new nuclear power plants with the Russian VVER reactors totaling 9.6 GW (up to eight power units) in South Africa. Besides, the parties intend to build a research reactor to the Russian technology, which would lay the basis for joint business in the area of isotope production and sales in the international market.”

“A small-scale analysis says that implementation of a NPP construction project will allow placing 40% up to 60% of equipment orders with the RSA companies, and this will give not less than US $16 billion at the stage of construction only, US $40 billion at the stage of operation; 15,000 up to 30,000 high-skill highly salaries jobs will be created. In the framework of our partnership we will be happy to invite RSA’s companies to the third country markets,” according to Rosatom. Finance was not mentioned then.

In September 2014 Rosatom signed an agreement with South Africa’s energy minister to advance the prospect of building up to 9.6 GWe of nuclear capacity by 2030. The minister said: “This agreement opens up the door for South Africa to access Russian technologies, funding, infrastructure, and provides proper and solid platform for future extensive collaboration." It is expected to involve some $10 billion in local supply chain provision, with localisation of up to 60%. Necsa later said that the new agreement "initiates a preparatory phase for the procurement process for the new nuclear build in South Africa. Similar agreements will be signed with other vendor countries that have expressed an interest in assisting South Africa with the build program...No vendor country has been chosen yet and no technology has been decided. The agreement refers only to what Russia could provide if chosen.” Rusatom Overseas confirmed the likelihood of a Russian government loan, and said that the build-own-operate (BOO) model was preferable. OKB Gidropress and NIAEP-ASE subsequently presented the VVER-TOI design as appropriate, each unit 1255 MWe gross, 1115 MWe net. The reported cost would be $6 billion for the first two units.

In October 2014 a nuclear cooperation agreement with France was signed. The energy minister said, "This paves the way for establishing a nuclear procurement process." Areva welcomed the agreement, and said that it was ready to support the development of new RSA nuclear projects, “notably through its Generation III+ EPR reactor technology."

In November 2014 a similar inter-governmental cooperation agreement was signed with China. The energy ministry said that the agreement "initiates the preparatory phase for a possible utilization of Chinese nuclear technology in South Africa." Three further agreements in December were between Necsa and China National Nuclear Corp (CNNC) to establish a cooperative partnership supporting the country’s nuclear industry, between China’s State Nuclear Power Technology Corp (SNPTC), the Industrial & Commercial Bank of China and South Africa's Standard Bank Group with a view to financing new nuclear plants, and between Necsa and SNPTC for training South African nuclear professional staff. In February 2015 Necsa signed a further skills development and training agreement with SNPTC and China General Nuclear Power Corp (CGN), funded up to 95% by China. CGN has had an office in Johannesburg since 2010.

Agreements with the USA and South Korea are in place, and a further agreement is pending with Japan.

Westinghouse has been active in South Africa's nuclear industry, mainly through support to Koeberg, since the 1990s. In October 2013 Westinghouse signed an agreement with the Sebata Group of engineering companies to prepare for ‘potential construction’ of new nuclear plants in RSA.

In March 2014 it was reported that China’s main nuclear power companies were lining up to bid for a $93 billion contract to build six reactors by 2030. China’s Ministry of Commerce reported that negotiations towards a nuclear cooperation agreement were proceeding. The Energy Minister said that this could involve the joint marketing and supply of nuclear energy products along with infrastructure funding to promote nuclear power developments across the region. SNPTC is focused on the possible supply of CAP1400 reactors. Chinese industry officials in December 2015 expressed confidence in securing the $80 billion order for CAP1400 units, though the first of these in China was not yet under construction.

The environmental impact assessment (EIA) process initiated in 2006 confirmed the selection of three possible sites for the next nuclear power units: Thyspunt, Bantamsklip, and Duynefontein, the last of which is very near the existing Koeberg nuclear plant. All are in the Cape region and were subject to further assessment. A draft environmental impact report (EIR) was published in March 2010 recommending the Thyspunt site in Eastern Cape province near Oyster Bay, Jeffrey’s Bay and a few kilometres west of Cape St Francis. Bantamsklip is east of Cape Town near Gansbai. A final EIR was to be submitted to the Department of Environmental Affairs early in 2011. In March 2016 Eskom submitted site licence applications to NNR for both Thyspunt and Duynefontein to construct and operate "multiple nuclear installations (power reactors) and associated auxiliary nuclear installations.”

The Renewable Energy Plan launched in October 2011 called for 17,800 MWe of "green" energy by 2030, and invited tenders for 1450 MWe of solar PV, 200 MWe of CSP, 1850 MWe of wind power, and various smaller contributions, total US$ 11 billion. In December 2011 the energy minister said that some $50 billion would be spent on nuclear capacity to 2030. The National Development Plan then cast doubt on nuclear power's financial viability, but in November 2012 the cabinet endorsed a "phased decision-making approach for implementation of the nuclear programme", along with the "designation of Eskom as the owner-operator as per the Nuclear Energy Policy of 2008".

The president’s annual state-of-the-nation address in February 2015 reaffirmed the 9.6 GWe target with the first unit on line in 2023 and said that bids would be sought from the USA, China, France, Russia and South Korea. In May, the energy minister said that the procurement process for the new nuclear power plant would begin by September, and she expected that a strategic partner would be selected by March 2016. Early in June Eskom ceded control of the new build programme to the Department of Energy.

Following cabinet approval in December 2015, the Department of Energy prepared to issue its request for proposals for 9600 MWe of nuclear power capacity. Five reactor vendors will be invited to make proposals: Rosatom, SNPTC, KEPCO, EDF/Areva and Westinghouse. Proposals should specify reactor design, the degree of localization, financing, and price. Funding will be decided following responses to this, and will be in line with the 2011 Integrated Resource Plan for Energy. Necsa stressed in September 2016 that the field was wide open and that an initial contract might be for up to three PWR units, about one-third of the total, with an operating reference plant in the country of origin. It did not want a BOO arrangement or a turnkey contract, but favoured a build-own-transfer model such as in the UAE.


Over 1993 to 2010, Eskom (in collaboration with others since 1999b) was developing the Pebble Bed Modular Reactor (PBMR). It is a high-temperature gas-cooled reactor (HTR) design, for both electricity generation (through a steam turbine or direct cycle) and process heat applications. From 1999 to 2009, the South African government, Eskom, Westinghouse, and the Industrial Development Corporation of South Africa invested R9.244 billion (about US$ 1.3 billion) in the projectc. The original concept was for a 400 MWt direct (Brayton) cycle unit, but later a 200 MWt (80 MWe) steam cycle version was proposed.

In September 2010, the Minister of Public Enterprises announced that the Government would stop investing in the PBMR9. The Minister gave the following reasons for this decision:

  • No customer for the PBMR had been secured.
  • In addition to the R9.244 billion ($1.3 billion) already invested in the project over the previous decade, a further R30 billion ($4.2 billion) or more was needed.
  • The project has consistently missed deadlines.
  • The opportunity to participate in the USA’s Next Generation Nuclear Plant (NGNP) program had been lost (see below).
  • Any new nuclear build program in South Africa would use Generation II or III technology. (The PBMR is considered a Generation IV technology.)
  • Government spending had to be reprioritized in the light of the economic downturn.

The project had received a certain amount of interest, but not enough to secure the financing it required. The domestic need in South Africa is for larger units.

Further information is in the Appendix below.

Eskom is managing the PBMR assets. The main PBMR test facilities (fuel development laboratory and helium test facility) are in care and maintenance. The government is concerned to protect the intellectual property involved.

An August 2013 application for federal US funds from National Project Management Corporation (NPMC) in the USA is for an HTR of 165 MWe, apparently the earlier direct-cycle version of the PBMR, emphasising its ‘deep burn’ attributes in destroying actinides and achieving high burn-up at high temperatures.

Uranium mining in South Africa

Uranium production in South Africa has generally been a by-product of gold or copper mining.

In 1951, a company was formed to exploit the uranium-rich slurries from gold mining and in 1967 this function was taken over by Nuclear Fuels Corporation of South Africa (Nufcor), which in 1998 became a subsidiary of AngloGold Ltd, now Anglo Gold Ashanti. Over nearly 30 years to 1980 it had produced some 100,00 tonnes of uranium oxide from varied feed, at a peak rate of almost 6000 t/yr in 1960. The plant is 60 km west of Johannesburg, adjacent to West Rand mines, in Gauteng province. It produces over 600 tonnes U3O8 per year from uranium slurries such as ammonium diuranate (yellowcake) trucked in from various gold mines and the Palabora copper mine. In May 2009 AngloGold announced plans to construct a new uranium recovery plant at its Kopanang mine to lift production to 900 t/yr from 2012.

The road tankers in Nufcor's fleet have two separate stainless steel compartments, each with capacity for about 25 tonnes of ammonium diuranate (ADU) as an aqueous slurry, containing the equivalent of about 7.8 t of U3O8 (earlier: 13 t and 3.5 t respectively).

There are about 400 tailings dams and dumps arising from gold mining in the Witwatersrand area of Gauteng province, and much of the available uranium today is in these. There are further tailings near Klerksdorp, close to the Vaal River. There is some radionuclide and heavy metal pollution arising from some of these and acid mine drainage. Many of the tailings dams and dumps are being re-treated to recover gold and sometimes uranium.

Uranium Production - tonnes U

  2011 2012 2013 2014 2015
Ezulwini-Cooke 34 0 0 69 47
Vaal River 548 465 531 504 346
Total 582 465 531 573 393

Much of the productive and prospective ground for uranium as gold by-product is in the Witwatersrand Basin, an area about 330 km x 150 km south and southwest of Johannesburg. Klerksdorp, Welkom, Carltonville, Parys and Evander are towns also on its fringes, associated with gold mines.

Cooke, with Ezulwini – Sibanye

First Uranium Corp of Canada, built a US$ 55 million uranium processing plant at Ezulwini gold-uranium mine in the West Rand, 40 km southwest of Johannesburg, which has 3200 tU in measured and indicated resources and 85,000 tU inferred resources. The main part of the plant, part of a $280 million recommissioning project, was completed and the first uranium produced in May 2009. Calcining is off-site, by Nufcor. A seven-year ramp-up of underground production from the Middle Elsburg reef was planned, but the uranium plant was placed on care and maintenance in February 2012. FY 2012 production (to end March) was 34 tU. The mine earlier produced over 6000 tU from 1961 to 2001. In August 2012 Gold One bought the mine and plant for $70 million, and it became Cooke 4.

Gold One planned to recommission the Ezulwini plant by March 2013, to treat both gold and uranium separately. The mine would form part of the company’s Cooke underground operations, but in August 2013 Sibanye Gold Ltd (spun off from Goldfields in February 2013) bought the whole Cooke operation, with all of the underground and surface plant, and including the dumps. Sibanye issued Gold One with about 150 million shares, some 17% of the company.

Nearby, Rand Uranium was spun off from Harmony Gold Mining Co and in joint venture with Pamodzi Resource Fund was reopening part of the Randfontein mine (also in Gauteng, 40 km west of Johannesburg). The mine, at the western end of the main Witwatersrand gold orebody, produced uranium in the 1980s, though its Cooke section had only been mined for gold. Rand identified JORC-compliant resources of some 41,000 tU both in tailings and underground. This included probable reserves of 15,200 tU in the Cooke tailings, which comprised the chief asset of the new company. Production at 1000 tU/yr was envisaged, but Randfontein Surface Operations (RSO) became part of the Cooke underground operations of Sibanye by about 2013.

The shallow Cooke 1, 2, & 3 mines, Cooke 4 including Ezulwini, and Randfontein Surface Operations comprise the Cooke operation of Sibanye, which describes it as a short- to medium-term asset, with life of mine estimated to extend to 2023. Cooke A1 proved & probable reserves were 1180 tU at 0.028%U in total A1 resources of 9420 tU at 0.045%U, with 16,950 tU as B1 inferred resources at the end of 2015. The Ezulwini plant enables recovery of uranium as by-product of about 4 tonnes of gold production annually. Cooke 4 is a 1634 m deep mine adjacent to Ezulwini plant, and a separate part of the operation.

Underground material from Cooke 1, 2 and 3 is processed at the Doornkop plant, operated by Harmony Gold Mining Company Ltd, on a toll treatment basis for recovery of gold. Run-of-mine ore from Cooke 4 is treated at the Ezulwini gold-uranium plant near Cooke 4 Shaft. Ore from the uranium section at Cooke 3 is hoisted separately and trucked to the Ezulwini gold-uranium plant for treatment.

In May 2014 Sibanye sent its first consignment, of 10 tonnes ammonium diuranate, from Ezulwini/Cooke to Nufcor for calcining to U3O8. In 2014, 69 tonnes of by-product uranium from Cooke was stockpiled. Production in 2015 was 47 tU, also stockpiled. The company earlier expected to ramp up to 230 tU/yr by-product by 2016.

West Rand Tailings Retreatment Project (Driefontein and Kloof)

The measured and indicated resources/probable reserves at Sibanye’e West Rand Tailings Retreatment Project (Driefontein, Kloof and Cooke historical tailings) in Gauteng province were 38,190 tU at 0.0054%U at the end of 2015. A feasibility study is under way in 2016.

(Driefontein has six operating shaft systems and three metallurgical plants, and operates at depths of 700 m to 3400 m below surface. Kloof is an intermediate to ultra-deep level mine with operating depths of between 1300 m and 3500 m below surface. Both are major gold producers.)

In July 2016 Sibanye Gold said that over 17 months the deep Cooke 4 operation had failed to meet production and cost targets, and losses had accumulated. Its continued operation was in question.


Sibanye Gold has its main operation at Beatrix, 20 km south of Welkom, in the Free State province 240 km southwest of Johannesburg. The company describes it as a low-cost and high-productivity operation with life of mine to about 2029. Sibanye announced uranium resources (SAMREC code) at December 2013 for Beatrix West Section – previously Oryx – on the Beisa reef of 9900 tU at 0.079%U. In February 2016 the company announced a maiden A1 probable reserve for Beisa of 4490 tU at 0.06%U, in A1 measured and indicated resources of 10,390 tU at 0.09%U. Beisa North has inferred resources of 13,630 tU at 0.09%U. In early 2016 the company completed a feasibility study on the project. The Beisa mine had been closed in 1984 by the predecessor of Goldfields. The company said it might construct a uranium plant at Beatrix or ship all the uranium north to be processed at the Cooke 4 Ezulwini plant in West Rand.

Buffelsfontein, Vaal River – Anglo Gold Ashanti Surface Operations

First Uranium had been building a larger $260 million uranium processing plant at the Buffelsfontein gold mine in the North West province, 160 km from Johannesburg and within the Klerksdorp gold field of the Witwatersrand basin. It is about 10 km east of Klerksdorp, and has some 21,000 tU as proven and probable reserves in old mine tailings, some near Stilfontein – its Mine Waste Solutions (MWS) project. The MWS tailings dams cover an area about 14 km across. Uranium production was expected to be 600 tU/yr over 16 years at full capacity, ramping up to 350 tU/yr from late 2010. The provincial government in July 2009 approved construction of a large new tailings storage facility for MWS, and a new plant was due to be commissioned in 2014. AngloGold Ashanti acquired the operation for $335 million in mid-2012.

Anglo Gold Ashanti produces uranium as by-product from its Surface Operations division which extracts gold from marginal ore dumps and tailings storage facilities on surface at various Vaal River and West Wits operations. Surface Operations includes MWS which operates independently. The company quoted by-product resources (mostly indicated) of 130,560 tonnes U3O8 including by-product reserves (mostly probable) of 53,700 t U3O8 at the end of 2015. Some 60% of the resources were in MWS or other surface tailings.

Uranium is produced at Vaal River by processing the reef material from Moab Khotsong, Great Noligwa and Kopanang, all next to to the Buffelsfontein mine. The reef ore is milled at the Noligwa Gold Plant and processed at the South Uranium Plant for uranium oxide extraction by the reverse leach process. Ammonium diuranate (‘yellowcake’) is the final product of the South Uranium Plant and is transported to Nufcor (located in Gauteng) where the material is calcined and packed for shipment to the converters.

Total production in 2013 was 532 tU, partly as by-product from the Great Noligwa mine. In 2014, 504 tU was produced.

Both Shiva and MWS operations are in the Klerksdorp area southwest of Johannesburg, and in 2008 First Uranium announced plans to build an acid plant using pyrite from MWS and 30 MWe of power generating capacity to service the two operations.

Shiva/Dominion Reefs

In 2006, Uranium Onei obtained its mining right for the Dominion Reefs project at Haartebeesfontein, 20 km east of Klerksdorp in the North West province,160 km southwest of Johannesburg. This had been the northern part of the adjacent Buffelsfontein Gold Mine operation. Production commenced early in 2007 and was planned to increase to 1730 t/yr U3O8 by 2011. Production cost was earlier expected to be US$ 14.50/lb U3O8 from the conglomerate reefs to 500 metres depth, but evidently increased well beyond this. The first sales contract for 680 tonnes was announced in November 2006. Production in 2007 was 78 tonnes and that for 2008 was 86 tonnes U3O8, reflecting slower and more difficult underground development than anticipated. A small amount of uranium was purchased from Australia in 2008 to meet sales commitments.

Dominion, including the Rietkuil section, had indicated resources of 51,000 tonnes U3O8 at 0.063% and inferred resources of 62,800 tonnes U3O8 at 0.036%. Within these, reserves however are only 14,240 tonnes at 0.077% and with US$ 46.50/lb production cost. The mine was closed in October 2008 due to a labour dispute coupled with power shortages and increased project costs in the context of lower uranium spot prices. The mine was then put on care and maintenance.

Uranium One sold it in April 2010 for $37 million to Shiva Uranium, a 74% subsidiary of Oakbay Resources & Energy Ltd, which is 85% Indian-owned. Shiva resumed uranium production early in 2011, but since then only gold has been produced while uranium workings are developed with a view to substantial production. The onsite uranium plant is the only one in South Africa using pressure leaching, which achieves uranium recovery of up to 92%, significantly more than other leaching methods.


Australian-based Peninsula Energy has reported a JORC-compliant resource of 21,930 tU at its Karoo project straddling the East and West Cape provinces. This includes indicated resource of 8440 tU grading 0.089%U in sandstone. Drilling continues to convert historical resource information from 1970s to JORC-compliant. Some of the resources in the Ryst Kuil channel have molybdenum by-product. Peninsula holds a 74% interest in the project, the remaining 26% is held by black economic development partners.

Uranium and molybdenum mineralisation is hosted in fluvial channel sandstone deposits chiefly in the western and central parts of the Main Karoo basin. The occurrences are epigenetic, tabular and sandstone-hosted, and the thickest sandstone bodies tend to contain the highest proportion of mineralisation. The company’s further exploration target is up to 110,000 tU.

The Ryst Kuil part of the Karoo project was held by Uramin Inc, which was then taken over by Areva to become Areva Resources Southern Africa. The deposit had been discovered by Esso in the 1970s. Some 16,000 tU resources were estimated on historic basis at 0.1% grade. Areva suspended the project at the end of 2011, and in April 2014 it was acquired by Peninsula Energy to form part of the Karoo project. The Ryst Kuil channel is the focus of ongoing exploration.


The Namakwa Henkries uranium project in the Northern Cape province is being explored by Namakwa Uranium, which is now owned 74% by Aarvark Uranium Ltd and 26% by the company's black economic development partner, Gilstra Exploration. Anglo American did a feasibility study on the project in 1979. Xtract Resources investigated the prospect in 2014, but did not proceed with acquisition.

Fuel cycle, historical R&D

Eskom procures conversion, enrichment and fuel fabrication services on world markets. Nearly half of its enrichment is from Tenex, in Russia. However, historically South Africa has sought self-sufficiency in its fuel cycle, and this is again becoming a priority. In December 2012, the Draft Mineral and Petroleum Resources Development Amendment Bill was approved. This aims to regulate uranium production and provides for the implementation of an approved beneficiation strategy through which strategic minerals can be processed domestically.

The South African nuclear industry dates back to the mid-1940s, when the predecessor organisation to the Atomic Energy Corporation (AEC) was formedj. In 1959, the government approved the creation of a domestic nuclear industry and planning began the next year on building a research reactor, in cooperation with the US Atoms for Peace program. The Pelindaba site near Pretoria was established in 1961, and the 20 MWt Safari-1 reactor there went critical in 1965k. In 1970, the Uranium Enrichment Corporation (UCOR) was established as South Africa commenced an extensive nuclear fuel cycle program, as well as the development of a nuclear weapons capability. In 1985, UCOR was incorporated into the AEC, which was restructured to become the South African Nuclear Energy Corporation (Necsa) as a state-owned public company in 1999.

A 1200 tU/yr conversion plant was established and ran in the 1980-90s.

Enrichment was undertaken at Valindaba (also referred to as Pelindaba East) adjacent to the Pelindaba site by the unique Helikon aerodynamic vortex tube process developed in South Africa, based on a German design. Construction of the Y-Plant pilot uranium enrichment plant commenced in 1971 and was completed in 1975 by UCOR. At this time, the USA stopped exporting highly enriched uranium (HEU) fuel for the Safari-1 reactor in protest against the construction of Y-Plant and South Africa's nuclear weapons program. Due to technical problems, Y-Plant only started producing 45%-enriched uranium in 1979 and in 1981 the first fuel assemblies for Safari-1 from Valindaba were fabricated. The Y-Plant produced about 990 kg of HEU with average enrichment of 68% until operations ceased in 1990, 336 kg of average 45% enrichment being assigned to Safari-1 for fuel or irradiation targets. The Y-Plant has been dismantled under International Atomic Energy Agency (IAEA) supervision. Much of the high-enriched uranium is stored.

On the neighbouring Pelindaba site, construction on a semi-commercial enrichment plant commenced in the late 1970s. This Z-Plant began commissioning in 1984, with full production in 1988. It had a capacity of 300,000 SWU/yr and supplied 3.25%-enriched uranium for the Koeberg plant. (Originally fuel for Koeberg was imported, but at the height of sanctionsl the AEC was asked to set up and operate conversion, enrichment and fuel manufacturing services.) Z-Plant was uneconomic and closed in 1995. It has been fully demolished.

Both centrifuge and molecular laser isotope processes were also being explored. Construction of the prototype module for the Molecular Laser Isotope Separation (MLIS) project was carried out in the Y-Plant building. The MLIS program started in 1983 and was joined by Cogema of France in a 50:50 funding arrangement in 1995. In 1997 the program was cancelled due to technological difficulties and AEC budget cuts.

Some fuel fabrication facilities operated from 1962. The BEVA fuel fabrication plant with 100 t/yr capacity operated in 1980-90s and supplied 330 PWR fuel assemblies for the Koeberg reactors.

A pebble fuel plant at Pelindaba was planned. Meanwhile, in December 2008, PBMR’s pilot fuel plant manufactured 9.6% enriched fuel particles, which were shipped to the USA for testing at the Idaho National Laboratory. In August 2009, PBMR (Pty) shipped 16 graphite spheres (containing 9.6%-enriched fuel particles) to Russia for irradiation tests to demonstrate the fuel’s integrity under reactor conditions. The irradiation tests, conducted by the Institute of Nuclear Materials in Zarechny near Ekaterinburg, were the final step in the development of the fuel for the PBMR demonstration unit.

A 2007 draft nuclear energy policy outlined an ambitious program to develop all aspects of the nuclear fuel cycle, including a return to conversion, enrichment, fuel fabrication and also reprocessing of used fuel as strategic priorities related to energy security. A new 5.0 to 10.0 million SWU/yr centrifuge enrichment plant built in partnership with Areva, Urenco or Tenex was envisaged, the larger version allowing scope for exports. These ideas seem to have faded but with 9600 MW of new capacity being built a significant level of local content in fuel cycle services is anticipated.

Initial feasibility studies on the re-establishment of nuclear fuel cycle programs were completed in 2011. Necsa’s pre-feasibility study for enrichment capacity, in collaboration with potential partners, is proceeding and the study is to be reviewed every two years. South Africa’s proposed new power plants are expected to need about 465 tonnes of enriched uranium annually by 2030. Necsa proposes to establish fuel fabrication capacity for PWRs ‘to ensure eventual security of fuel supply for South Africa’. In March 2013 Westinghouse signed a cooperation agreement with Necsa on development of local facilities for fuel assembly components.

In 2012 the vision was for 1800 tU/yr conversion plant,1.3 million SWU/yr enrichment plant and 200 tU/yr fuel fabrication plant, all established at one fuel cycle site from 2016. The conversion capacity might involve re-commissioning the old plant, or an international joint venture to commission in 2026. Enrichment would be by centrifuge, possibly with international partner, to commission 2026-257. Fuel fabrication was to be in partnership with the new NPP vendor, commissioning in 2023-25.

Research & development today

Necsa was established from AEC as a public company under the 1999 Nuclear Energy Act, and is wholly owned by the State. Its main functions are to undertake and promote research and development in the field of nuclear energy and radiation sciences and technology, and to process source material, special nuclear material and restricted material.

Necsa operates the 20 MWt Safari-1 reactor at its Pelindaba nuclear research centre. Safari-1 is the main supplier of medical radioisotopes in Africa and can supply up to 25% of the world's molybdenum-99 needs. By 2009 the reactor was converted from using HEU to low enriched uranium (LEU) fuel11, and conversion of the targets used for radioisotope production from HEU to LEU was achieved in 2010.12 Following this, Necsa and its subsidiary NTP Radioisotopes (Pty) Ltd in October 2010 were awarded a $25 million contract by the US Department of Energy's (DoE's) National Nuclear Security Administration (NNSA) to supply Mo-99. The commercial-scale production of the medical isotope from low-enriched uranium will be in collaboration with the Australian Nuclear Science and Technology Organization (ANSTO), whose 20 MWt Opal reactor also uses LEU fuel and targets for Mo-99 production.

With the end of operating life of Safari in sight, proposals have been for Dedicated Isotope Production Reactor (DIPR) and, with more probability, a new research reactor which includes isotope production among other roles.

In mid 2013 a nuclear cooperation agreement was signed with the EU, to support research, with PBMR and medical isotopes mentioned. This will be implemented by Euratom.

Klydon Corporation, which emerged from the AEC, has been developing its Aerodynamic Separation Process (ASP) employing so-called stationary-wall centrifuges with UF6 injected tangentially. It is based on Helikon but, pending regulatory authorisation from Necsa, has not yet been tested on UF6 – only light isotopes such as siliconmwhich it is evidently most suited to. Klydon Element 92 Division is focused on uranium prospects, while its Stable Isotopes Division is concerned with silicon-28, zirconium-90 and medical isotopes.

A private company, Steenkampskraal Thorium Limited (STL), is designing the TH-100, a 35 MWe (100 MWth) pebble bed HTR reactor.

Radioactive waste management

The 2008 National Radioactive Waste Disposal Institute Act provides for the establishment of a National Radioactive Waste Disposal Institute (NRWDI) to be responsible for radioactive waste disposal in South Africa. Establishment of this was announced in March 2014.

Necsa had been operating the national repository for low- and intermediate-level wastes at Vaalputs in the Northern Cape province. This was commissioned in 1986 for wastes from Koeberg and is financed by fees paid by Eskom. From about 2008 this Vaalputs facility became the National Radioactive Waste Disposal facility, and continued to be managed by Necsa until 2014 when NRWDI was set up. Some low- and intermediate-level waste from hospitals, industry and Necsa itself is disposed of at Necsa's Pelindaba site.

Used fuel is stored at Koeberg. In May 2015 Holtec won a contract from Eskom to supply HI-STAR 100 dual-purpose metal casks for transport and storage of used fuel by September 2018.

In August 2008, the nuclear safety director of the Minerals and Energy department announced that Eskom would seek commercial arrangements to reprocess its used fuel overseas and utilize the resulting mixed oxide (MOX) fuel.

Regulation and safety

In 1948, the Atomic Energy Act created the Atomic Energy Board, which later became the Atomic Energy Corporation (AEC). In 1963, the Nuclear Installations Act provided for licensing and in 1982 the Nuclear Energy Act made the AEC responsible for all nuclear matters including enrichment. An amendment to it created the autonomous Council for Nuclear Safety, responsible for licensing.

The Nuclear Energy Act of 1999 gives responsibility to the Minister of Minerals & Energy for nuclear power generation, management of radioactive wastes and the country's international commitments. The South African Nuclear Energy Corporation (Necsa) is a state corporation established from the AEC under the Act, and is responsible for most nuclear energy matters including wastes and safeguards, but not power generation.

The National Nuclear Regulator Act of 1999 sets up the National Nuclear Regulator (NNR) – previously the Council for Nuclear Safety – covering the full fuel cycle from mining to waste disposal, and in particular the siting, design, construction, operation and decommissioning of nuclear installations. The NNR is being strengthened in preparation for an expanded role with new nuclear power plants. In November 2015 NNR established a relationship with China’s National Nuclear Safety Administration (NNSA) to collaborate on several fronts and benchmark its regulatory practices.

The Department of Minerals and Energy (DME) has overall responsibility for nuclear energy and administers the above Acts. However, Eskom is under the Department of Public Enterprises.

The Department of Environmental Affairs is responsible for environmental assessment of projects, and has a cooperative agreement with the National Nuclear Regulator for nuclear projects.

An IAEA Integrated Nuclear Infrastructure Review (INIR) mission was undertaken early in 2013 to evaluate the status of the country’s nuclear infrastructure development. It was the first such mission to a country with established nuclear power, and was valuable.


Having been a foundation member of the IAEA in 1957, South Africa is the only country to develop nuclear weapons and voluntarily give them up. It embarked on a nuclear weapons programme around 1970 and had a nuclear device ready by the end of the decade. The weapons programme was terminated by President F. W. de Klerk in 1990 and, in 1991, the country signed the Nuclear Non-Proliferation Treaty (NPT). In 1993, de Klerk announced that six nuclear weapons and a seventh uncompleted one had been dismantled. In 1995, the International Atomic Energy Agency (IAEA) was able to declare that it was satisfied all materials were accounted for and the weapons programme had been terminated and dismantled.

In 1996, South Africa signed the African Nuclear Weapon Free Zone Treaty – also called the Pelindaba Treaty. In 2002, the country signed the Additional Protocol in relation to its safeguards agreements with the IAEA. South Africa is member of the Nuclear Suppliers' Group, and the country’s dual-use capabilities are regulated by the South African Council for the Non-Proliferation of Weapons of Mass Destruction (NPC).

Further Information


a. In January 2008, Eskom was forced to curtail power exports as well as introduce load shedding. The reserve margin of the electricity system was around 5% in January 2008 but by January 2009 the reserve margin had recovered to about 14%. This was due to economic slowdown, and hence lower electricity demand, as well as the recovery of coal-related problems experienced by the company in early 2008. [Back]

b. Eskom holds all the shares in the PBMR company, PBMR (Pty) Ltd., but several investment partners have provided financing for the feasibility stage of the project. In June 2000, the UK's British Nuclear Fuels Limited (BNFL) took a 22.5% stake in the venture. Soon after, US utility PECO (later Exelon, following the merger with Commonwealth Edison) took a 12.5% stake. The South African government-owned Industrial Development Corporation (IDC) took 25%, leaving Eskom with 40%, of which 10% was reserved (but never taken up) for an Economic Empowerment Entity. Exelon withdrew from the project in April 2002. Also, around the same time, BNFL reduced its stake to 15%, and IDC reduced its to 13%. In 2006, BNFL's 15% stake was transferred to its Westinghouse subsidiary, which was later sold to Toshiba.

Under an investors' agreement made in 2005, BNFL/Westinghouse had a 15% stake, IDC 14%, the South African government 30%, leaving Eskom with 41%. These shares were expected to move to 4% Westinghouse, 15% IDC, 30% South African government and 5% Eskom by 2012, with 46% being held by another investor. However, in August 2006, this agreement lapsed and a new agreement could not be reached. (Had a new investors' agreement been reached, Westinghouse would have had rights to 5% of the company, the South African Industrial Development Corporation to 5%, and 81% for the government, leaving Eskom with 9%.)

Although PBMR (Pty) Ltd continued to list its investors as the South African government, IDC, Westinghouse and Eskom, its funding following the completion of the feasibility stage in 2004 was principally from the South African government (through its Department of Public Enterprises). In March 2010, the government drastically cut funding for the PBMR, then in September 2010 it announced that all funding was to be cut. [Back]

c. Of the R9.244 billion (about US$ 1.3 billion) invested in the PBMR project, the South African government contributed 80.3%, Eskom 8.8%, Westinghouse 4.9%, the Industrial Development Corporation of South Africa 4.9%, and Exelon 1.1%. Figures given by Barbara Hogan, Minister of Public Enterprises, to the National Assembly on 16 September 2010 (see Reference 9 below). [Back]

d. Developed from the 200 MWt Siemens/Interatom HTR-Modul reactor design, the initial PBMR was a 268 MWt (110 MWe) design. In order to lower the capital costs of the plant, relatively minor changes led to a 302 MWt design. However, as a result of issues arising during more detailed analysis, in 2002 it was decided that a complete review of the design was to be carried out. This resulted in the 400 MWt (165 MWe) version with a fixed central reflector in the core (the 268 MWt design has a dynamic central reflector column of graphite spheres). In addition, the power conversion unit was changed from three-shaft vertical to single-shaft horizontal turbine-compressor configuration. Reactor outlet temperature is 900ºC.6 [Back]

e. In 2009, the PBMR company announced it had decided to focus on a 200 MWt (80 MWe) design for the PBMR rather than the 400 MWt version (see Note d above). The 200 MWt version uses a conventional Rankine cycle to deliver super-heated steam (750ºC) through a steam generator for electricity generation and process heat applications. [Back]

f. Construction of the 400 MWt demonstration plant was originally envisaged to commence in April 2007 but, partly due to delays in licensing, was put back to 2009. Later, following the decision in 2009 to focus on the 200 MWt PBMR design, the construction schedule was delayed indefinitely. A total of R 1 billion was spent on equipment, including R 268 million of reactor vessel parts made by Spain's Equipos Nucleares.

In 2003, the South African Department of Environmental Affairs and Tourism (DEAT) issued positive Record of Decisions on the environmental impact assessment (EIA) studies for the PBMR demonstration module and pilot fuel plant. However, these decisions were set aside by the Cape High Court following appeals from anti-nuclear group Earthlife Africa. This ruling, along with design changes to the PBMR – the Brayton cycle turbine design was simplified from 3-shaft vertical to single shaft horizontal configuration and the reactor capacity increased from 302 MWt to 400 MWt (see Note d above) – led to the decision to enter into a new EIA process for the demonstration PBMR. This process remains unfinished and is not likely to be completed in light of the change to a 200 MWt version of the PBMR - and the subsequent cancellation of funding in 2010. In January 2007, the Minister of Environmental Affairs and Tourism upheld the positive Record of Decision for the pilot fuel plant EIA.

Several contracts for the 400 MWt design had been awarded. In April 2005, PBMR (Pty) awarded a US$ 20 million contract to Uhde, a local subsidiary of Germany's Thyssenkrupp Engineering, to build a plant at Pelindaba near Pretoria to manufacture the fuel pebbles for the planned demonstration PBMR. The fuel plant was expected to be completed by 2010 but was delayed by regulatory issues. In August 2008, a contract was awarded to the joint venture company Murray & Roberts SNC-Lavalin Nuclear (Pty) Ltd for the provision of engineering, procurement, project and construction management (EPCM) services for the demonstration PBMR plant, then envisaged to be at Koeberg. [Back]

g. In September 2010, Barbara Hogan, South African Minister of Public Enterprises, told the National Assembly that one reason for the withdrawal of government support for the PBMR was: "The opportunity afforded to PBMR to participate in the USA’s Next Generation Nuclear Plant (NGNP) program as part of the Westinghouse consortium was lost in May this year when Westinghouse withdrew from the program" (see Reference 9 below). In fact, Westinghouse did not withdraw from the NGNP project, but did withdraw from the conceptual design phase (sometimes referred to as Phase 1). Following Hogan's September announcement, Westinghouse released a statement which it says was made in May 2010 by Kate Jackson, Senior Vice President of Research and Technology and Chief Technology Officer of Westinghouse: "In early March, the team of Westinghouse, PBMR (Pty) and Shaw Group was selected by the Department of Energy (DOE) to negotiate an award under a funding opportunity announcement for Phase 1 of the DOE's NGNP project. Since that time, the team has not been able to reach agreement on a way forward, and therefore, will not participate together in this project phase." [Back]

h. The small operating HTR-10 research reactor at China's Tsinghua University is the basis of the 250 MWt (105 MWe) HTR-PM reactor (one 210 MWe module consisting of twin reactor units driving a single steam turbine), which also derives from the earlier German development. [Back]

i. See the Uranium One website (www.uranium1.com) [Back]

j. In 1944, the USA and UK requested forecasts from South Africa on its potential to supply mineable uranium. This led to the formation of the Uranium Committee in 1945, and, in 1948, the Atomic Energy Board (AEB) was formally established to oversee uranium production and trade. In 1959, research, development and utilisation of nuclear technology was added to AEB's remit. In 1970, the Uranium Enrichment Corporation (UCOR) was established, initiating an extensive fuel cycle program. In 1982, the AEB was re-established as the Nuclear Development Corporation of South Africa (NUCOR) under a new controlling body – the Atomic Energy Corporation of South Africa (AEC). In 1985, UCOR was incorporated into the AEC. The South African Nuclear Energy Corporation (NECSA) was formed out of the AEC in 1999. [Back]

k. The Safari-1 (South African Fundamental Atomic Research Installation) reactor initially operated at 6.75 MW and was upgraded to 20 MW in 1968. The pool-type reactor is an Oak Ridge National Laboratory (ORNL) design fuelled by highly enriched uranium (HEU). A program to convert to low enriched uranium (LEU) fuel commenced in 2006. [Back]

l. South Africa's policy of apartheid – which ended with the 27 April 1994 general election – attracted extensive international sanctions. However, it was not until the late 1970s/early 1980s that international pressure intensified, culminating in 1985-1991 with trade sanctions by the USA, British Commonwealth and Europe, as well as disinvestment campaigns in many countries. [Back]


m. Extrapolating from test results, ASP is expected to have an enrichment factor in each unit of 1.10 (cf 1.03 in Helikon) with about 1000 kWh/SWU. Development of it is aiming for 1.15 enrichment factor and less than 500 kWh/SWU (compared with about 10,000 kWh/SWU in the Z-plant). However, to achieve gas speeds sufficient for enrichment, heavy elements such as uranium need to be greatly diluted with hydrogen, and the process appear uneconomic for uranium