Eskom sees a nuclear future in the pebble bed

30 November 1998

Is there a reactor type for which public acceptance, and the costs of safety in the achievement of that acceptance, would allow a renewal of a nuclear power programme? In South Africa, Eskom successfully operates the two unit Koeberg PWR station, but it does not see LWRs as a solution for the present. Rather, it is putting its technical and financial resources behind a modular pebble bed high temperature reactor project which it sees as the best approach to take.

The case for nuclear power in South Africa at load centres some distance from the coal fields is still strong enough for the country’s national utility, Eskom, to consider its possibilities. Following an assessment in 1993 the utility once again discounted conventional light water reactors. However, this time further investigations were carried out which identified the two main problems: the increased costs of nuclear power and public acceptance. Both of these were perceived to be the result of the safety issues related to potential accidents, so that the only way to achieve competitive costs with nuclear power would be to remove the potential, however remote, for accidents with significant off-site consequences.

The small high temperature reactor (HTR), using coated particle fuel, appeared to be the only type which met these requirements. The design was investigated in much greater detail and this work has progressed to the stage of a concept design for a 100 MWe Pebble Bed Modular Reactor (PBMR). The level of detail in the concept design was sufficient to ensure that the safety arguments could be shown to be valid and that realistic costing could be made.

The key initial requirements identified for undertaking a PBMR programme were an adequate technology level in local industry with a backer of sufficient size, a non-prescriptive licensing regime and a cost structure for power generation which excluded the present technology. All these requirements could be met in South Africa, with Eskom well placed both in terms of its size and its legal position.

The key design requirement is that no physical process, however unlikely, could cause a significant radiation-induced hazard outside the site boundary. This can be achieved in the PBMR principally by demonstrating that the integrated heat loss from the reactor vessel exceeds the decay heat production in a post-accident condition and that the peak temperature reached in the core during a transient must be below the demonstrated fuel degradation point, and far below the temperature at which the physical structure is affected. Heat removal from the vessel is achieved by passive means. These features preclude a “core-melt” scenario.


The PBMR project was launched as part of Eskom’s Integrated Electricity Planning Supply Side Working Group activities. The initial work in 1993 reviewed potential nuclear options and identified the possibility of a small, inherently safe reactor based upon German designed high temperature reactors. The funding was initially small but grew as the potential economic benefits were increasingly confirmed by the work being done. Throughout the project the target set were to be competitive on cost with large Eskom mine-mouth coal fire power stations without limitations on siting.

Technological features of the PBMR project were based on the experience gained in a number of projects, leading to a high degree of confidence in the design. Different experience could be gleaned from each project, but in all cases it was fundamentally based on extensive research of materials, components, fuel, core and overall plant technology. Specific examples include the neutronic design of the 265 MWth PBMR reference proposal based on the HTR-MODUL (200 MWth) and HTR-100 (250 MWth) designs of pebble bed reactors by Siemens/Interatom and BBC/HRB, respectively. The concepts are based on the well-proven technology and the operational histories of the AVR experimental reactor in Jülich and the THTR demonstration plant in Uentrop-Schmehausen (a 300 MWe demonstration pebble bed reactor plant with steam turbine that operated for 5 years). These were projects costing billions of US dollars and were followed by the licensing of the HTR-MODUL reactor design in 1987 for commercial operation in Germany, showing that the key technologies have been mastered.

Concept design and costing studies showed that the technology being adopted for the base line design had all been demonstrated adequately to avoid fundamental technical risks. This was supported by technology contracts with the original commercial developers in Germany (Siemens and ABB) through their subsidiary (HTR GmbH) as well as the related research institute (KFA in Jülich). Support for other key technology areas included detailed involvement of expert companies in several countries, such as GEC Alsthom on the helium submerged generator, UCAR Carbon on the reactor graphite, NFI of Japan on the fuel manufacture and Mitsubishi Heavy Industry on the main reactor vessel, both from a technical and costing point of view.

There have been over ten independent reviews of the technical and commercial aspects, both those funded by the project and those requested by potential joint venture partners. Only two specific concerns were raised: the back end fuel cycle costs (by the IAEA expert on nuclear costing) and the performance of the recuperator (in the review from GEA, and to a lesser extent, by NREC).

As far as the back end fuel cycle costs are concerned, the figures for the PBMR are based on the agreed rationale for Koeberg. The current cost of disposal is less than 1% of the fuel cost and therefore even a large (factor of 10) increase would not significantly increase the levelised cost of power. In the case of the recuperator the concern is principally over the compactness of the design and this has been addressed by doubling the available volume for the recuperator. This did not have a significant impact on the overall costs.

During the period of these reviews, the available core power was increased by changes to the fuelling regime from 226 to 265 MW while increasing the available margins (lowering the peak fuel temperature during loss of cooling events). At the same time the reflector structure was made substantially simpler.

The reviews have included a market survey for this class of plant covering 19 countries which indicated a substantially larger overseas market than used for the economic evaluations. This is because of the strong cost advantage of this design over other options where very cheap coal is not available.


The performance figures are listed in the following table. The dynamic performance of the design has been validated by use of an engineering simulator.

The overall circuit layout is shown diagrammatically above.

The process cycle used is a standard Brayton cycle with closed circuit water-cooled intercooler and precooler. Separate turbo compressors and power turbine with adjustable stator blades are used. This separation simplified the design and qualification process and the adjustable stator blades are used for short-term control. Magnetic type bearings avoid any contaminants in the helium circuit and reduce maintenance. The reactor systems are placed inside a reactor pressure vessel (RPV) and the turbomachinery and heat exchangers in a Power Conversation Unit (PCU).


Substantial progress has been made in preparing the PBMR project.

• Milestones

The PBMR programme has already achieved the following milestones:

• Application for nuclear licence of the design with the South African regulator, the CNS (Council for Nuclear Safety).

• Initiation of environmental impact assessment process to allow selection of the first site.

• Establishment of single programme team (of over 40 full time staff) outside the utility head office.

• Finalisation of concept design.

• Prequalification of key suppliers.

• Negotiations with potential joint venture partners.

• Initiation of tender process for detail design of long lead items.

This programme is intended, when combined with public and stakeholder consultations, to enable the decision on the potential construction of the first reactor to be taken by the end of 1999. This would include full consideration of the development, construction and operating costs, design parameters and the site location.

• Licensing

The South African licensing system is based on a fundamentally probabilistic basis, with the requirement to meet international standards. This has been the case since the beginning of the South African nuclear power programme in the early 1970s. The process requires that any nuclear activity (including design process) shall be covered by the Council for Nuclear Safety.

In light of this, and the need to establish the design rules for the PBMR, Eskom has formally applied for a license for the PBMR design. The application has started a programme aimed at achieving the initial “licensability” statement on the PBMR by the third quarter of 1999. This will include the safety criteria that the plant must meet, the general and specific design criteria, the event list, the classification systems, and a review of the design basis.

These activities are underway with the involvement of international consultants (both to support the CNS and Eskom). The initial Technical Description has been issued to the CNS and the draft SAR is due to be completed by the end of the year.

• Environmental impact assessment

Under South African legislation there is a requirement for a full EIA for any new power plant. Accordingly, Eskom has started the process and has called for potential suppliers to submit their capabilities. A formal inquiry to the qualified suppliers will be issued shortly. As part of the EIA process there will be a number of coastal sites considered, as well as the overall societal impact as to the value of the PBMR project (to allow for the “no go” option).

• Engineering

The concept design is now largely complete, with the basic design process underway. Those areas of the plant which are not covered in significant detail are those which are standard “off the shelf” equipment and do not contribute significantly to the cost (e.g. the air compressor system), but even in these cases a performance specification has been generated.

The load-following capability has been a very specific emphasis in the work to date. An engineering simulator, based on a G2 platform, has been developed to allow non-real time dynamic analysis. Two other system analysis tools are being used in the cycle development, FLOWNET and a MATHCAD-based model. While these tools are sufficient to handle the concept and basic design phases they are not seen to be appropriate for the detailed design phase and a new engineering simulator is under development due to be in service by the middle of 1999.

Another key area has been maintenance analysis. There has been extensive work on the maintainability of the design, and all components are classified by their life expectancy and difficulty of repair. This leads to some components having very easy access (eg the bypass and interrupt valves) which can be maintained without breaching the helium circuit, and some which can be changed, but only in the same manner as changing steam generators on a PWR (eg the recuperator). This analysis has allowed the maintenance cycles and removal routes to be established (along with, for example, their impact on building design and crane requirements).

Test rigs are under construction in particular areas. These are required to demonstrate very specific features which would be valuable to include in the design and are separate from the inquiry documents now being issued for the detail design and manufacture of long lead-time components (the manufacture option in these contracts would be exercised only once the commitment to the first unit has been made).

• Fuel manufacture

Fuel is a key element of the PBMR programme, and the quantities required exceed any previous HTGR project. Small scale HTGR fuel-making facilities exist in various parts of the world (in Japan and China, for example), but the PBMR requires a new fuel manufacturing facility. The present intention is to construct it at the South African Atomic Energy Corporation (AEC) site, in the complex that built the fuel for Koeberg. This project has identified the layout for such a line (initially to manufacture 1.4 million spheres a year) and the equipment specifications.

Discussions are now being held with vendors for the equipment. For the actual fuel technology, Eskom has involved a number of suppliers and potential suppliers in various ways (such as partnership agreement, commercial contract, or in some cases negotiations are still underway). In the meantime the SA AEC, under an Eskom contract, has started laboratory scale work and has started the manufacture of fuel kernels (with a target of 5 kg by the end of the year) and is setting up a small coater which should be commissioned by the end of the year. This work is to support the external technology that is being obtained.


The PBMR project has been analysed from three perspectives to gauge its value to the country, the utility and the investor.

• National benefits

Eskom established a base case scenario to allow analysis of potential national benefits. The project was subjected to an input-output analysis, using the base case model, for construction of the units only (without the development and fuel projects). The model assumed 10 units a year for local construction and 20 for export. The South African content of the local plant is 81% and 50% for export plant. These values are based on the plant equipment breakdown and a realistic assessment of current local manufacturing capability.

The figure of 20 export units a year was based upon an assessment of the world market carried out for Eskom. This would equate to a 2% share of the overall world power market or (over the 20 years considered) approximately 14% of the replacement market for current nuclear plants. The figure for the local market was based on the long-term Eskom growth trend of 3.53% (1980-1993), or 1500 MWe/y on a 410 GWe base. This equates to the long term medium to high growth assumptions of Eskom. In both local and export cases, the impact is on a linear basis. In other words, the impact of one unit a year for the local market is 10% of the impact of 10 units a year. To retain the local content however, there must be an adequate annual production level, at least 5-10, to maintain the economies of scale.

The analysis showed that when the project had matured (approximately 10 years) the effect on the South African economy from the local and export market was:

• Utility benefits

The PBMR studies were started to meet a future need for distributed generation at a competitive cost to Eskom’s current coal generation. A combination of advantages of the PBMR over any other identified options are:

• Distributed generation.

• Short construction period.

• Small unit size.

• Excellent load following.

• Low environmental impact.

• Competitive economics .

The key impacts of the PBMR option on Eskom’s expansion planning would be allowing for the construction of multi-unit power stations near to coastal load centres, limiting the need for extensive transmission system strengthening. It would also reduce the uncertainty, risks and therefore the costs associated with the current long-term planning requirements and reduce Eskom’s exposure to negative environmental claims, such as carbon dioxide emissions and use of highveld water resources. It would result in the improvement of quality of supply at remote locations (eg the Eastern Cape) without the need for new line compensation equipment. The PBMR would also provide an economic mitigation strategy for greenhouse gas reductions.

The current IEP analysis indicates that the series PBMR would be constructed in parallel with a new 4000 MWe coal-fired station, and these cost assessments are currently being made without full costs of including integration of transmission to remote load centres (such as Eastern and Western Cape). These costs can be significant. Their inclusion would increase the competitive advantage of the PBMR.

It is of note that the development cost of the PBMR (R365 million in 1998) is some 2.5% of the capital cost of a 4000 MWe coal fired station, or the interest on a three-month total project delay! • Investor benefits

This project was started solely on the basis of the utility requirements. It is, however, clear that the economic advantages of the PBMR need not be limited to the South African grid. Unlike Eskom’s other options, coal and hydro, the PBMR costs are virtually independent of location. The base-load costs of about 1.43¢(US)/kWh† (including 6% profit on construction and 20% on fuel) are extremely low compared with overseas costs (average cost in China is 3¢/kWh, and in Japan 9¢/kWh). Therefore there should be an extensive export potential, particularly as the safety standards being applied to the PBMR design are stricter than those applied to other modern western nuclear stations.

Following the work in 1997 an analysis of the project investment returns was undertaken, assuming it captured some 2% of the world market for new power plant. This analysis assumed that an owner’s generation cost of 1.6¢/kWh would be attractive and resulted in a base case where the internal rate of return on invested equity and loan capital over a 25 year period is a real 25.5% after tax. No sales to Eskom are included in this analysis beyond the first module and no external gearing is assumed. Analysis showed that the result was not highly sensitive to the startup cost (100% over expenditure of development, fuel plant and first unit brought the IRR to 18.6%) but it was to the construction period gearing after the first unit.

On this basis, the project can be seen to be a viable and attractive investment opportunity.

There are several other competitive advantages specific to South Africa that can be exploited:

• Adequate technology level in the local industry.

• Limited (compared to USA/EU/Japan) anti-nuclear movement.

• Large enough utility to provide backing for the project.

• Non-prescriptive nuclear licensing regime.

• Cost structure for power generation that imposes a strong cost cap.

• Utility having good public image and credibility.

The competitive edge of the PBMR over potential suppliers of similar technology will be maintained due to the initial time lead (which must be protected) combined with an ongoing active product development and enhancement programme funded from revenue. This is included in the current financial mode as 4% of total turnover.

Safety advantages of the Pebble Bed Modular Reactor

To achieve a “catastrophe-free” design, irrespective of probability, a study into the relative risk between the PBMR and other nuclear activities was established (Audit of the Eskom pebble bed modular reactor, by John H Gittus). It should be noted that the risk from other industrial activities substantially exceeds that of the highest nuclear risk by a substantial margin. Under these analyses a PBMR could, for all sensible purposes, be considered to be a normal industrial plant, in terms of siting and emergency planning. The following is a Risk Index table which provides a comparison of various risks. Risk Index table* 37 Public risk due to accidents, including a “Chernobyl” at a PWRa. 35 Occupational risk due to normal operation at a PWR. 6.1 Public risk due to normal operation of reprocessing plant. 1.8 Public risk due to normal operation of a PWRb. 1.4 Public risk due to normal operation of an Advanced PWRc. 0.4 Public risk due to accidents, including a “Chernobyl”, at an Advanced PWR. 0.2 Occupational risk due to normal operation of reprocessing plant. <0.1 Public risk due to the most severe accidents at a PBMR. a The public risk is for members of the public living within 1000 km of the installation. The PWR chosen is one of those at Tricastin, as it is typical of modern reactors. b The public risk due to normal operation is the environmental risk. c The Advanced PWR chosen is Sizewell B in the UK. It incorporates additional safety features compared with the earlier designs of PWRs. * An index of 1 is equivalent to a dose of 0.01 man-Sv/TWh

PBMR compared to GT-MHR

There are three main differences between the main system design of the PBMR (Pebble Bed Modular Reactor) and the GT-MHR (Gas-Turbine-Modular High temperature Reactor) which is being developed by the international group of GA/MINATOM/Framatome/Fuji Electric. Fuel design The GT-MHR uses block fuel as against the spherical fuel elements of the PBRM. The main advantage of the block fuel is that it has structural strength and fully defined geometry (as with LWR fuel). It is therefore possible to place control rods inside the fuel region, in-core instrumentation etc and to share the loads on the overall core structure. The disadvantages are that the refuelling can only be done with the reactor off line, the excess reactivity to have an extended fuel cycle must be managed and the actual coated particles are in “compacts” with high packing densities, and therefore greater prospect of interaction between the particles. The main advantage of the spherical fuel elements is that the fuel movement takes place while the reactor is on load. This therefore removes the need for “refuelling” outages, limits the interest cost on fuel in the core and minimises the excess reactivity which has a significant impact on certain reactivity events. The other advantage is that the coated particles are dispersed throughout the fuel zone and therefore the problem of particle interaction is very much less. Turbine layout The GT-MHR uses a single shaft, on which there are the power turbine and the LP and HP compressors. The PBMR has three separate shafts, one for the power turbine (driving the generator) and the two turbo-compressor units (each with a turbine driving a compressor). The main advantages of the GT-MHR layout is that the potential overspeed of the shaft on load rejection is less, due to the load of the compressors. In the case of the PMBR there is no load on the shaft when the electrical load is removed suddenly. This leads to the need to limit the free speed of the PBMR with full gas velocity and no load, to design a generator with substantial overspeed capacity (40+%) and to install an interrupt valve system which can stop the gas flow (in the cold part of the circuit). The advantage of the multi-shaft system is to allow the individual shafts speeds to be optimised independently, to have dynamic control (including adjustable stator blades) and to limit the size of the individual machines and therefore reduce the problems of shaft dynamics. The multiple shaft also allows for easier maintenance when the different stages have different operating lives (mainly due to high temperature creep). In the case of the PMBR with a 50 Hz (vs 60 Hz for GT-MHR) frequency and a very high load following requirement, the multi-shaft solution was seen to be better. Gas ducts The GT-MHR has an annular gas duct with the cold (~470°C) in the outer section and the hot (~850°C) gas in the centre. The PBMR has an outer manifold which is kept at the HP compressor outlet temperature (~110°C) and inside this the separate hot (~900°C) and cold (~540°C) gas ducts. The PBMR arrangement leads to a larger diameter outside duct but does not subject the high pressure boundary to the temperature of the “cold” gas. This high temperature on the GT-MHR requires the use of a more advanced steel than on the PBMR while still placing an upper limit on the cold gas temperature. The higher return temperature on the PBMR should allow higher cycle efficiencies if all other component efficiencies are the same.

Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.