There are a number of issues impacting the choice of ADS neutronic parameters, in particular the ADS reactivity, keff. The degree of sub-criticality (keff) must be a balance between safety and acceptable economics. Here keff represents the sum of the initial reactivity and all other possible effects, e. g. burn-up reactivity swing including Np or Pa effects, power and void reactivity, etc.

ADS can be used for minimising the sources of long-term radiotoxicity, e. g. reactor fuel inventories, fuel wastes from reprocessing, and long-lived radioactive fission products (Slessarev, 1997). According to Salvatores et al. (1995), the latter two of these sources are the most important in terms of the accumulation of radiotoxicity.

For example, consider the neutronic potential of a representative ADS within a uranium fuel cycle complex (Slessarev, 1997) in the following system. A slightly sub-critical lead — cooled fast breeder reactor with nitride fuel and proton beam source with a keff of 0.98 would exhibit a neutron surplus of about 0.4 neutrons per fission (zero breeding gain in the fuel) plus 0.05 neutrons/fission due to spallation in the lead target. The lead is used as a liquid and target. This gives a total neutron surplus of 0.45 neutrons/fission, sufficient to burnout all dangerous fission products and/or reproduce new fuels for further nuclear power utilisation.

In this system, there is no need for control rods; it is a dual circuit, and a relatively inert coolant from the point of view of safety, e. g. fire hazard. The neutronics are sub-critical plus a stabilised reactivity increment. This system provides an apparently good balance

with regard to economics, reduction of fuel waste potential and safety for the uranium fuel cycle.

The thorium fuel cycle has a much lower waste toxicity level for both thermal and fast reactors than does the uranium fuel cycle. This is because of its smaller production of trans-plutonium (Carminati et al., 1994; Rubbia et al., 1995) and, therefore lower minor actinide concentrations (at least for about 1000 years before some build up of long-term toxic U, U, Pa). From a neutronic perspective, however, every fission of Th produces fewer neutrons than does 238U. There are other disadvantages in relation to achieving sub-criticality at economic cost and a protactinium effect, which implies a low keff value. Thus for the thorium cycle, it is necessary to have a compromise between the economics, sub-criticality level and safety margin. This is difficult because a low keff can only be achieved at more expense; reduced cost would be at the expense of higher keff and less safety margin.

The SVBR-75 reactor module is designed by EDB Gidropress and SSC RFIPPE for steam production to replace VVER-440 reactors that are being decommissioned (SSC RF-IPPE, EDB, 1996; Stepanov et al., 1998) (Figure 14.3). Specifically it has been designed for application in the Novovoronezh power plant facilities as units 2, 3 and 4 are decommissioned. The concept is flexible and can be applied for combined generation of heat and electricity. The SVBR-75 concept exhibits the important features of lead — bismuth coolant systems (Gromov et al., 1996).

14.6.1 ANSTREM

The ANGSTREM project (Stepanov et al., 1998) is based on the concept of a modular, transportable nuclear power and heating station, utilising fast reactor technology with lead-bismuth eutectic cooling. The main design organisation is EDB ‘Gidropress’ together with IPPE, Obninsk providing scientific consultancy. The ANGSTREM technology is envisaged for a number of applications including electricity generation, heat supply, freshwater and possibly hydrogen production.

Separate effects tests were carried out for the AP600 design to demonstrate the feasibility of using a passive core cooling system to mitigate all design basis accidents. There were also confirmatory tests to verify the performance of the various system components. These included: passive residual heat exchanger, automatic depressurisation, passive core cooling system check valve and core make-up tank tests.

In addition to separate effects tests, there were also passive core cooling system tests to demonstrate the overall system performance for both pressurised and de-pressurised conditions. The test facility for this programme was the Oregon State University APEX facility, and the programme was carried out within a Westinghouse/USDOE collaboration.

There were a number of thermal-hydraulic facilities commissioned and operated during the 1980s and 90s in support of the needs of currently operating plant. Many of these facilities have been dismantled but others remain either in standby or in operation to service the needs of evolutionary water reactors. Facilities include PKL, SPES for PWR, PIPER-ONE for BWR, PACTEL and PMK for VVER and PANDA for BWR (Addabbo et al., 2001).

The SPES facility (Bacchiani et al., 1994) at the SIET facilities in Piacenza, Italy was modified to include a passive core cooling system and used for high-pressure system loop thermal -hydraulic tests in support of AP600. All the safety systems were simulated and a series of tests addressed LOCA, steam generator tube rupture (SGTR) and SLB thermal — hydraulic issues. PKL is currently in use to simulate boron mixing effects, in connection with a present day reactor transient issue involving boron dilution during reflux condensation in a LOCA.

Although configured for VVER geometry, PACTEL tests (Kervinen et al., 1990), have been carried out to simulate passive injection during a LOCA, which is of relevance to the AP600 safety system function.

15.10.1 Containment Tests

Many of the confirmatory tests for AP600 were in justifying the passive containment cooling system. Separate-effects tests to characterise the decay heat removal character­istics of the containment design were carried out. These tests included the investigation of heat removal from wetted steel plates simulating the containment surface. Also containment external cooling air flow path pressure drop tests were carried out to characterise the frictional losses. Steam condensation tests on surfaces at different angles were performed to simulate condensation inside the containment in the presence on non­condensable gases.

Composite containments, including a steel inner liner and an outer concrete shell, have been considered to meet potential European requirements for licensing. The outer concrete shell provides greater strength to mitigate the consequences of some severe accidents. Experiments to establish passive containment cooling for such containments were carried out in the PASCO facility at FZK, Germany (Erbacher et al., 1995).

Passive systems are a feature of a number of advanced evolutionary LWRs, both for primary coolant system heat removal and for containment cooling. Tests are in progress in the PANDA facility in Switzerland in the EC TEMPEST programme (Wichers et al., to be published), to resolve outstanding issues of the effects of light gases for confirming the long-term LOCA response of the passive containment cooling systems for SWR100 and ESBWR.

Advanced gas reactors (AGRs) were designed to overcome some of the inherent limitations of the Magnox design. The main problem with the Magnox design was the low power density, pressure and operating temperatures.

The first prototype AGR was built at Windscale in 1962. The commercial AGRs that were subsequently built were twin 620-660 MW plants. Seven stations were built; these entered commercial operation in the late 1970s and 1980s. The first industrial plant was at Hinkley B commissioned in 1976. These plants ran into difficulties during their construction and design phases due to problems that were both industrial and technical. In all, three different industrial groups were commissioned with different design approaches. The Dungeness B loop is shown in Figure 1.6.

The AGR uses carbon dioxide as a coolant, like the Magnox plants, but in order to achieve higher coolant pressures (~ 40 bars) and temperatures (outlet temperatures ~ 650°C), a new fuel design was required. The fuel became uranium dioxide pellets, inside stainless steel tubes.

AGR fuel had to be enriched to about 2.3% uranium-235 in order to overcome the significant neutron absorption of the stainless steel fuel cans. With this enrichment, it was

possible to achieve a 3-fold increase in volumetric power density with an average fuel rating of 4-fold increase compared with the best Magnox stations.

The more onerous pressure and temperature operating conditions created difficulties for the designers associated with vibration, chemistry (corrosion) and concrete insulation problems.

In the AGR, the coolant gas is circulated from the core to steam generators. These are mounted inside the pre-stressed concrete pressure vessel. These steam generators comprised 4 or 8 steam raising units. Good efficiencies are achieved as high as 40%. The steam generators provide steam at around 170 bars and 560°C, conditions that are comparable with those in an efficient fossil fuel plant.

A problem of concern for the AGR designers was attack of the graphite moderator by the carbon dioxide gas, which could oxidise the graphite and reduce its strength. This was overcome via controlled coolant chemistry with an appropriate level of water vapour content together with a small concentration of methane. This was however a delicate balance, because too much methane could result in carbon deposition on the fuel elements and consequent degradation of heat transfer.

AGRs can be refuelled on load and the fuel can remain in the core for long periods, up to 5 years. They have high fuel efficiency, up to about 40%; they have a more efficient use of fuel compared with LWRs. The AGR has a number of inherent safety features; e. g. the graphite has a large thermal capacity in the event of a primary circuit rupture.

A disadvantage of AGRs has been the limited investment of international vendors to support their technology. This, coupled with the lack of standardisation, has led to higher capital costs. It has not competed successfully outside of the UK in comparison with the PWR and BWR.

1.4.2 High Temperature Reactors

There is clearly a strong incentive to maximise the thermodynamic efficiency of nuclear power plants and one way of achieving this is to increase the temperature of the coolant. From the early days of nuclear power there has been considerable interest in helium cooled high temperature reactors (HTRs).

A 20 MW prototype, the Dragon reactor, was built and operated at Winfrith between 1964 and 1975. The plant was operated as part of an international OECD co-operative programme. Although, there were plans for a follow-on programme to Dragon, these were not pursued.

Another 13 MW prototype, the AVR, was built in 1966 at Julich in Germany based on the ‘pebble bed’ design. In this design, the fuel consists of particles of thorium or uranium dioxide fuel surrounded by carbon. These particles are a fraction of millimetre in diameter and are bonded into balls. Following AVR, a 295 MW plant was built at Schmehausen in Germany in the early 1980s and this achieved power in 1985.

In this design, the core is filled with approximately 675,000 spherical graphite fuel particles. The helium coolant is pressurised to about 40 atm and exits the core at 750°C. Heat is transferred to water and steam, circulating in stainless steel tubes within the helium. Steam passes to the steam generator at 530°C and 181 atm.

Another model, taken forward in the US was the prismatic core design. In this design, the fuel particles are formed into cylindrical rods and placed in hexagonal graphite blocks with coolant channels. An initial 40 MW prototype designed by the General Atomic Company was built at Peach Bottom in the US. This operated from 1966 to 1974. It was followed by a 330 MW prototype at Fort Saint Vrain, which came onto the grid in 1976. Here, 10,000 fuel particles are fixed in a graphite matrix with 210 fuel channels and 108 helium channels. The helium is at 48 atm, there are 1482 fuel blocks. Somewhat higher coolant outlet temperatures were achievable with this design.

The helium-cooled reactors have a number of attractions in principle. Helium is a preferred coolant to carbon dioxide in the presence of graphite since it is inert and therefore does not oxidise graphite — a problem at higher temperatures in carbon dioxide- cooled reactors.

Another attraction of the helium-cooled reactor designs discussed above was that they could be used to produce fissile material from less useful uranium fertile material. Uranium-238 is converted to plutonium-239 and uranium-233 to thorium-232. There was, therefore, the possibility of achieving very high burn-up with targets up to 100,000 MW days tonne-1.

Difficulties were encountered in the early days of these reactors and new orders were not placed following these prototypes. However, there has been a recent revival of interest in HTRs in recent years, e. g. the ESKOM project in South Africa.

At the time of writing, there is no commercial power plant of this type in operation. However, this type of reactor is one of the designs under consideration in the US Generation IV programme. These designs are discussed in detail in subsequent chapters.

Human factor issues affect most aspects of plant design, operation and maintenance. The subject has received increasing attention over recent years from both regulators and utilities. Operating experience has shown that plant personnel and the systems within which they operate, play a very important contribution to safety.

There have been recent efforts in human factors engineering, ergonomics, and biomechanics to improve understanding and safety operation (Ramsey, 1998). These have also included human/machine interfaces and the development of special purpose systems. These must manage various data inputs relating to information gathering and output to maximise human and machine performance. Techniques for human error rate prediction (THERP) have been established. These compile human error rates for various industrial tasks. Comparisons with other industries indicate that nuclear facilities generally meet very high comparative degrees of safety. Studies of human attitudes and physical limitations are given in USNRC (1992). An account of workers responses to events and their compliance with control measures is given in SOER92-1 (1992).

Human factors’ issues have been recognised by the UK regulator via specific safety assessment principles addressing human factors issues (Dixon, 1998).

One method of improving safety is to identify factors that impact on performance of a particular job. Utilities have developed a number of techniques to help them to analyse particular tasks. Factors that impact on performance include design of interfaces, the procedures in place and staffing levels and training. Good practice guidelines that have been recommended include the adoption of user friendly operating instructions and peer review of proposed changes, etc.

The HSE in the UK has also recognised the importance of broad-ranging organisational factors including safety management systems and safety culture (Dixon, 1998) (Table 3.5). The establishment of good safety culture within organisations is important to ensure the implementation of safety principles at all levels within the plant. The HSE, in common with most other safety authorities, believes that the licensee should own its safety cases. This is particularly important today in many countries including the UK, where nuclear industries are undergoing rapid change and where there is increasing use of contractors.

BNFL/Magnox Generation commissioned a study to establish the relationships, if any, between employee safety awareness and safety performance (Spooner and Vassie, 1999). This study identified five factors — training/experience, safety initiatives, communication, organisation and personnel.

The participants in the study felt that safety awareness was developed partly via common sense and partly via specific skill training. Personal experience, particularly of an unsafe event, was not surprisingly, highly influential on an individual’s safety awareness, but in addition learning from a colleague’s experience was also influential. Training/ experience were considered to be key influences in employee safety awareness.

Safety initiatives such as ‘near miss’ reporting were regarded as effective in promoting safety awareness. It found that performance-related bonus schemes, including the achiev­ement of specific safety targets did not significantly influence safety-related behaviour. This would appear to contrast the situation in certain organisations in North America.

The study appeared to show that passive forms of communication, e. g. notices, had less impact than verbal communication; e. g. team briefings were found to be more effective. The employment of active communications’ systems, e. g. PA and VDU systems, was also considered to be more effective.

It was concluded that improved communication of learning events was useful and that the organisation should be in place to do this. How the organisation responded in resolving

Table 3.5. Human factors issues

Impinge on plant design, operation and maintenance Task analysis — procedures, training, interface design, staffing levels Safety culture at all levels — organisation, plant management, staff Licensee ownership of safety case — cf. use of contractors

safety issues was also considered but the participants did not feel that this contributed significantly to safety awareness. However, further work was required.

Personal relationships, including interactions within a group and responsibility for others were considered to influence safety awareness significantly.

Although the study cited was specifically for the BNFL/Magnox Generating Group, it was felt that the conclusions have a wider application to other similar organisations.

Safety management and safety culture have been reviewed in a recent IAEA international conference on safety culture in nuclear installations held in Rio de Janeiro in December 2002 (IAEA International Conference, 2002; IAEA/NSR/2002, 2003). This conference confirmed that safety culture is now regarded internationally as an important element of nuclear safety. The IAEA Nuclear Safety Standards Committee endorsed a proposal in 2002 to develop safety standards specifically addressing safety management and culture. Two particular issues identified at this conference were that although safety culture is now embraced by top management, there is still a need to broaden appreciation through to the shop floor. It was also noted that safety culture is being embraced more enthusiastically in countries with a developing industry, than in those which had long — established nuclear programmes.

There are differences in national approaches in respect of once-through vs. a reprocessing and recycling policy. These approaches are linked with national policy on the management of natural resources, view on the relative radiological risk, domestic energy resources, security of supply and the relative economics, see for example, Bertel and Wilmer (2003).

From a sustainable energy perspective, recycling offers the option of better utilisation of resource and reduced radioactive waste. A MOX fuel cycle offers plutonium burning and reduction of radiotoxicity of spent fuel. In terms of public risk, an NEA study (OECD/NEA, 2000) concluded that the differences in public exposures between the fuel cycles were not significant.

The position adopted on the second and third issues depends on the country’s requirement for autonomy.

The economics depend on the expected prices of uranium and fuel cycle costs and specific national conditions. The current position favours the once-through option, even with a significant growth in nuclear energy production.

Through the evolution of the 1970-1980s the approach was generally to build bigger and more sophisticated reactors (Mourogov et al., 1999; Anand, 1999). This approach was perceived to suffer from several disadvantages. The larger reactors were not suitable for developing countries with smaller grids. Also the increasing sophistication was not commensurate with reducing capital cost.

As a consequence, smaller and simpler designs were put forward, perhaps the best known was AP600 incorporating passive decay heat removal systems. This system has now been extended to larger scale AP1000, see above, but the approach was first introduced and verified on the lower rated AP600 design.

Another ‘approach’ to provide a flexible capability is to consider modular units, which can be designed, manufactured and assembled using production line processes and standardised procedures (Hatcher, 1999). The 100 MWe gas reactor pebble bed modular reactor (PBMR) is an example of this approach, introduced earlier in Chapter 2 and discussed in more detail later.

7.3. INNOVATIVE DESIGNS

A wide range of advanced reactor types has been considered over recent years but many of these would require substantial investment and development. A set of the most promising reactor types has been put forward by the Generation IV International Forum (GIF) Member Countries. Design requirements for these systems are considered later in the book.

India has a vibrant nuclear power programme with currently 14 units in operation, 9 under construction, and more new reactors planned (Table 9.3). There are 5 research reactors in operation (World Nuclear Association, 2003). Currently, nuclear power supplies less than 4% of the country’s electricity requirement. There is a target to reach 10% in 2005. Capacity factors are now much improved compared with a few years ago, reaching 85% in 2001-2002.

The Tarapur plants are increased capacity plants based on domestic technology and are expected to begin operation in 2004-2005. The other PHWRs will follow later; the Rawatbhata units are scheduled to be in operation by 2007. The design for future PHWRs, the first of these are likely to be Kakrapar 3 & 4, has now been raised to 680 MWe.

Two large VVER-1000s are being built by Russia. The first unit is forecast to be commissioned in 2007.

The construction of a 500 MWe fast breeder reactor is in progress at Kalpakkam. This is contributing to the government’s objective to utilise India’s abundant thorium resource as a nuclear fuel. The intention is for this reactor to be operating in 2010. This reactor will be fuelled by uranium-plutonium-carbide fuel. The plutonium resource would come from currently existing PHWRs.

The intention is to develop an advanced heavy water reactor (AHWR) thorium cycle based of the following route. Existing PHWRs will burn natural uranium to produce plutonium. Fast-breeder reactors of the type above will then burn plutonium and breed U-233 from thorium. AHWRs will then burn the U-233 with thorium. The first AHWR is due to start construction in 2004.

There are plans to build a mix of reactor types to meet India’s requirements. The forward plan is to have a 300 MWe AHWR together with a mix of 500 MWe FBRs, 680 MWe PHWRs and 1000 MWe LWRs by 2020.

India’s civil nuclear strategy is to achieve complete independence in the fuel cycle. The country has a fuel fabrication facility at Hyderabad for PHWR and BWR. There are also spent fuel and reprocessing plants at Trombay, Tarapur and Kalpakkam. There is a waste immobilisation plant and storage facility at Tarapur.

Research is in progress in setting up a deep geological repository for high-level wastes.

The Indira Gandhi Centre for Atomic Research at Kalpakkam is working on fast reactor technology development. The Bhabha Atomic Research Centre near Mumbai is working on thorium-based systems. In particular, the Centre is working on the AHWR. In addition, India is also developing accelerator-driven systems for driving sub-critical reactors.

The PBMR design has been put forward by the South African Utility, ESKOM in partnership with an international consortium. It also meets Generation IV design objectives in that it includes passive safety features to meet public acceptance criteria and offers competitive economics. The units are relatively small at 110-120 MWe with good economic and safety characteristics. The PBMR is also flexible in that it can be built virtually anywhere. It operates with a direct Brayton thermo-dynamic cycle, with target efficiency of around 45%. In principle, it can also use a thorium fuel cycle as well as a traditional uranium cycle. The design is modular in order to enable an operating utility to match the size of his station to the demand. The present capital cost is estimated at about $1000US per kWe, the construction period is estimated to be very short at around 2 years.

The PBMR offers a potential complementary service to the energy market in terms of present plant capabilities as both an electrical and non-electrical energy generator. It is of medium size, comparable with current-sized gas plants. It could offer a capability for the co-generation of heat or even dedicated nuclear heating applications, as expanded below.

The PBMR design is based on the HTR-MODULE design previously licensed in Germany for commercial operation. Present activities are aimed at the engineering design, independent safety reviews by participating countries in the ESKOM project and in making provisions for the licensing process.

The HTGRs have desirable features from various safety perspectives. The cores have a large thermal inertia, low power density and a strongly negative Doppler reactivity coefficient. As for most reactor types, the transients can be categorised into two broad categories, reactivity-initiated events and loss of flow events, either with or without depressurisation. For an un-scrammed core heat-up, the maximum core temperatures are reached within 3 days but fuel temperatures do not exceed above 1600°C ensuring that fuel particle integrity is preserved.

One concern with HTGRs is that air could ingress the core resulting in oxidisation of the graphite. This would require a multiple failure scenario of ruptures in the pressure vessel and surrounding concrete. However as noted above, even if such events occurred, there would still be several days to breach the opening of the reactor vessel.

A considerable advantage of gas systems described above is that they are free from the usual problems associated with loss of cooling in LWR systems. Thus there are no phenomena of concern such as ‘Departure from Nucleate Boiling’ loss of heat transfer or ‘Pellet Clad Interaction’ failure.

The reactor has diverse and redundant safety systems. For example, the reactor can be shutdown by three independent control systems. Each system is sufficient in itself to achieve this requirement.

In summary, in addition to electricity generation, HTGRs are being proposed as candidate plants for process heat applications that require high-temperature conditions. These include hydrogen and methanol production in a steam reformer, a process that requires high-temperature heating of steam and methane. Steam could be produced and then utilised for processes such as coal densification and steam injection for the recovery of hydrocarbons. These plants would also be suitable for de-salination processes, which require low-temperature heat. There may be potential to take waste heat from the pre­cooler that would otherwise be wasted. The ways of operating HTGRs in these multi­generation modes would add significantly to the thermal efficiencies that would be achievable with the plant.

If the accelerator beam is switched off, the external spallation source will turn off and the reactor will go sub-critical with the power at decay heat levels. In a critical reactor, shutdown is achieved via the mechanical insertion of control rods. In both cases there is a delay from the trip signals for shutting off the beam or for activating the control rod release, which might be of the order of 0.5 s. However, overall, the time to switch off the current to the accelerator would be much faster than the control insertion time in a critical reactor, e. g. 1.5-3 s in a PWR (a little faster for a fast reactor with a smaller core).

A number of different beam shut-off systems are being considered. Diverse trip signals are necessary that result in beam shut-off. Since shut-off is important in cooling failure accidents, the current could be coupled to that driving the coolant pumps on the various cooling loops or on the feedwater pumps. Other passive means involve dropping the spallation target. This could be achieved by supporting with a low melting point metallic structure, which would melt in the event of sufficient temperature increase. Another could be via a magnetic structure, which would drop the target once the Curie temperature is reached. Other methods include deflecting the proton beam or, as in the ADS Rubbia design, by interrupting the beam by the rising lead level in the event of a cooling failure.