14.1. INTRODUCTION/OBJECTIVES

The development of nuclear power has been primarily concerned with electricity generation. However, there is increasing interest in utilising nuclear power for other purposes. Some of these have already been described in the two preceding Chapters 12 and 13, including systems for the destruction of plutonium, the conversion of minor actinides in waste and for the production of hydrogen. This chapter covers more generally, further applications of nuclear plant for other than electricity generation, e. g. reactor systems for district heating, desalination and other process plant.

Nuclear energy can provide an alternative to carbon fuels as a useful heat source. This was realised early in the history of nuclear power development. Nuclear reactors have already been utilised in many of the nuclear operating countries for supplying energy for district heating, seawater desalination and other industrial processes. Much of this energy has been produced from power reactors operating in co-generation mode with electricity production together with one of the heat applications above.

IAEA (IAEA-TECDOC-1056, 1998) is acting as a forum to facilitate interest in nuclear heat applications. It has co-ordinated reviews of progress in the technology, including operating experience, technological developments and experience in the above applications. There are now over 60 reactors supplying heat in district heating, desalination and other industrial processes together with over 500 reactor-years of operational experience. The technical or safety-related issues in regard to nuclear heat applications have been considered in the international community. There are few additional issues compared with electricity generation applications.

Of the overall world energy consumption, about one third is used for electricity generation. Of the remainder, heat utilised by residential and industrial consumers represents a major share, the majority of this heat produced by burning fossil fuels, coal, gas, oil and wood. The next significant energy consumer is transport. Nuclear energy supplies about 6% of the world energy requirement and about 17% of the electrical supply. Although only about 1% of the heat produced by nuclear reactors is used for heat applications there are some signs of growing interest (Csik and Kupitz, 1997). Significant experience in co-generation of electricity and heat has been gained in Russia, Europe, North America and Japan, dedicated heat producing plants are now also receiving attention, e. g. Russia and China (IAEA-TECDOC-1056, 1998).

Historically, there has been more interest in district and process heat applications than in desalination. However, with the obvious requirements for freshwater in the developing

countries, there is increasing interest in desalination applications in the IAEA Member States (IAEA, 1998). There is likely to be an increasing need for freshwater in much of the developing world over the next few decades (Wangnick, 1995).

High-temperature applications are again mentioned briefly in this chapter to complete the survey. There are some additional applications (other than hydrogen production) under consideration in some countries.

High fast neutron fluence in RPV internals can change the ductility and fracture resistance of the material. Cracking has been detected in some RPV internal components such as the core shrouds and top guides of BWRs and this has resulted in the need for more data on the irradiated material properties. A concern has also been expressed as to whether high neutron doses could cause void swelling and, therefore embrittlement induced by the voids. These phenomena could clearly impact on the life of a plant. In order to address these issues, the EC PRIS project has been set up to examine the properties of irradiated stainless steels for predicting the lifetime of such nuclear power plant components (Nordgren et al., 2001).

The project involves the procurement of representative top parts of BWR control rods blades of type AISI 304L and type AISI 316L stainless steel with fast neutron fluences in the range 2 X 1021 -5 X 1021 ncm~2. These specimens are being examined, mechanical properties are being determined and the microstructure is being characterised.

A thimble tube of type AISI 316 stainless steel from the Swedish Ringhals 2 plant that has been irradiated for 23 years to between 0 and 70 dpa is also being examined. Tensile and hardness properties, fracture properties and radiation-induced micro-structural and micro-chemical changes will be determined. Fracture properties will be determined using previously established pin-loading fracture toughness test techniques (Grigoroev et al., 1995, 1997).

The properties of both the BWR — and PWR-irradiated materials will be compared with non-irradiated archive materials.

Stress corrosion cracking in PWR and BWR shroud internals are also under study in the EC INTERWELD project (Youtsos et al., 2001). The objective of this project is to define better the radiation-induced material changes in the heat-affected regions of austenitic stainless steels.

Test welds of stainless steel type 304 and type 347 are being produced with weld residual stresses, microstructure and properties that are representative of core shroud applications. These are being irradiated to two neutron fluence levels in the HFR at Petten, the low level at 0.3 dpa and the high level in the range 0.8-1.2dpa. These levels are representative of LWR internal irradiations. The results will be compared with an in­service weld from the BR3 reactor. This weld has been irradiated from 1962 to 1987 in a coolant of temperature 260-300°C with maximum dose irradiation of 2.4 X 1020 n cm-2.

The weld residual stresses of the irradiated materials are being measured by neutron diffraction and the corrosion characteristics of the material will be determined by further tests. Mechanical properties are being determined for both the test specimens and the in-service material. The microstructure and microchemistry properties are being obtained by optical, EPMA and other techniques.

Computational dynamics (CFD) codes provide solutions to modelling more general thermal — hydraulics situations, which are not modelled adequately by the system codes. Examples of such codes are the CFX code (CFX 4.3, 1999) developed for general fluid flow applications; in particular, it can be applied for reactor safety analysis. Another example is the FEAT code including coupled thermal-hydraulic and structural modelling capabilities developed by British Energy.

CFD codes are used to model flows where 3D effects and/or turbulent mixing phenomena are important. They are also useful in modelling complex geometries with arbitrary boundary shapes and internal structures. They are used for detailed phenomenological modelling to gain understanding but also in supplying mixing models for benchmarking system codes. For LWR applications, they are used in transient analysis of boron dilution events, thermal mixing in overcooling transients, and cold water mixing in steam line breaks. They are also used for modelling pools in advanced reactor passive systems where thermal mixing processes are often important in modelling heat transfer mechanisms.

CFD codes are being validated for reactor safety applications in a number of different European research projects. The codes include CFX-5 and FLUENT for modelling flow mixing and flow distribution in the primary circuit (FLOMIX-R) (Weiss et al., to be published). CFX-5, CODE-SATURNE and TRIO-U (ECORA) (Scheuerer et al., to be published) are being validated for a range of applications including primary loop flow mixing, pressurised thermal shock (PTS) flow modelling and 3D containment analysis.

To date, CFD codes applications in reactor safety are largely concerned with single­phase applications. The ASTAR project (Paillere et al., to be published) has looked to extend the modelling limitations of the systems codes such as CATHARE, ATHLET, TRAC, RELAP5, etc. For example, a multi-dimensional model FLUBOX was coupled to ATHLET within this project. CFD codes are now being developed for multi-scale (termed CFMD (Yadigaroglu, to be published)) applications and these are being examined at the research level.

Fluid flow modelling in gas reactors where only single-phase flows are present, is a much more straightforward proposition than the modelling of two-phase flows in LWRs. CFD codes have been applied to gas reactor flow modelling in normal operation and accident conditions. They are particularly amenable for modelling such flows and they have also been coupled with neutronics codes to provide power variation feedbacks.

There have been substantial analytical methods developments for modelling sodium cooled LMFBRs (IAEA-TECDOC-1083, 1999). Codes have been developed with the support of extensive experimental facilities in Europe and the US. Codes have been produced for modelling decay heat removal under various accident conditions. The requirements have been to model forced and natural circulation in various components under steady-state and transient conditions. There has been particular attention paid to the development of multi-dimensional codes for modelling disturbed turbulent liquid metal flows. Much of this experience will be relevant to future liquid metal systems.

Clearly decisions on granting life extension will rest with the regulator of the country in question. He will need to have established procedures in place and if not already available, these will take time to develop. There are now precedents in a number of countries on regulators considering or having already granted applications. For example, in the UK, the NII have already extended the lifetime of some of the Magnox reactors beyond 40 years. Several US utilities have applied for lifetime extension. The issue is now under consideration in Canada, Japan and European countries, including Russia (for some designs).

2.8.3 Political Factors

Undoubtedly, the climate of acceptability of nuclear power has changed during the lifetime of plants in many countries. Many plants commenced operation when the view that ‘nuclear power would be too cheap to meter’ was being expounded and attitudes towards nuclear power were very positive. Now, decades later there may be a moratorium in certain countries on extending plant operation beyond the original operating life. Nevertheless, there are a number of countries that are not opposed to the continued operation of nuclear power plant and the decision will then become one of economics and safety compliance.

Various performance indicators have been defined for measuring the success of a plant in terms of its availability to produce energy safely and economically. The maintenance of good availability depends primarily on the following (IAEA-TECDOC-1098, 1999): control of outage activities, reduction of unplanned outage, reduction of plant transients, improvement of thermal efficiency, good housekeeping of the facilities, minimising plant ageing and optimising staff utilisation.

Different agencies have put forward different performance indicators (IAEA-TECDOC — 1098, 1999; WANO, 2002) but they have much in common. Important indicators include, e. g. the energy availability factor (EAF) and the unit capacity factor (UCF) or cumulative EAF. These are included among the IAEA and its power reactor information system (PRIS) measures and WANO measures. The EAF is defined to be the ratio of

the actual energy generation (net) in a given period, as a percentage of the maximum energy that could have been produced by continuous operation. Figure 4.1 shows how the worldwide UCF (WANO) has steadily improved over the last decade. Precise definitions of the WANO measures are given in Table 3.1.

Unavailability factors are also considered by both agencies and others. The unavailability factor is usually broken down into planned (PUF) and unplanned energy unavailability factors (UUF) (IAEA-TECDOC-1098, 1999). In IAEA-TECDOC-1098 (1999), energy losses are considered to be planned if scheduled 4 weeks in advance. Planned energy losses include planned outages for refuelling, maintenance, testing, etc. under management control. Unplanned outages include not only unplanned outages requiring similar activities, but also for losses beyond the control of management. Figure 4.2 shows how the unplanned capability loss factor of WANO has steadily reduced over the last decade. Figure 4.3 shows a similar reduction in unplanned automatic scrams.

These improvements are attributed to improvement in plant maintenance management and through taking advantage of the benefit of experience.

Across the range of reactor types, the PWR, BWR and AGR units have kept a broadly constant level of performance in recent years. For VVER & RBMK units, there was a decrease in energy availability in the early 1990s, but there has been recovery more recently. The initial decrease was due to the implementation of back-fitting programmes and increase in other maintenance activities during this period.

In general, the IAEA Study (IAEA-TECDOC-1084, 1999) found that the variations and uncertainties found in the data (costs levelised to 1997) for decommissioning after long­term storage (Table 6.7) had similarities with the data for decommissioning immediately (Table 6.8). In particular, this correspondence related to variations from country to country and also from specific case to case. Also not surprisingly, decommissioning costs were sensitive to national labour resource estimates.

6.11.1.1 PWRs. Estimates were provided for reactors from Belgium, Germany, Japan, Korea, Netherlands and the US for units in the range 500-1300 MWe. It was found that costs ranged between 200 and 700$US per kWe, and for the capacity range considered there were no economies of scale. It was found that in most cases the total over-night costs for decommissioning with long-term storage were higher than for immediate decommissioning, considered later. However, it is recognised that the net present value would normally be on the contrary, dependent on the decommissioning schedule and the assumed discount rate.

6.11.1.2 BWRs. Estimates were provided and shown for Finland, Germany, Italy, Japan, Netherlands and the US for reactors in the range 160-1300 MWe. Decommission­ing costs were found to be in the range 150-600$US per kWe (a small BWR-60 plant in

the Netherlands was also analysed; it was found that scaling effects did exist for the smaller capacity plant). In addition, the relative costs between decommissioning after long-term storage and immediate decommissioning were similar to those for PWRs. For example, long-term storage undiscounted costs were again higher than immediate decommissioning but as noted earlier, the situation would be different if discounted costs are taken into account.

6.11.1.3 VVERs. Data were available from Bulgaria, Czech Republic, Germany, Russia and Slovakia. Reactors considered were the VVER 440 MWe plants 440/230 and 440/213. The costs were found to differ widely from 120 to 130$US per kWe in the Russian Federation up to 1400$US per kWe in Germany. Much of this difference is reflected in labour rate costs. As for PWRs and BWRs, decommissioning with long-term storage is more expensive than immediate dismantling (undiscounted costs). The differences were relatively small for Slovakia.

6.11.1.4 HWRs. Decommissioning costs from three available assessments of Canadian units were estimated. It was found that there was a substantial difference between units of similar capacity, largely reflecting the situation that the cost estimates were made at different times. For some Canadian plants, decommissioning with long-term storage was found to be cheaper than decommissioning with immediate dismantling even in undiscounted costs.

6.11.1.5 LWGRs (RBMK). Estimates were available from the Ukraine and Russia. Due to the large amount of graphite in the core, decommissioning with long-term storage is a more feasible option for LWGRs than decommissioning with immediate dismantling. Most assessments for long-term storage are again higher (undiscounted) than those for immediate dismantling.

6.11.1.6 GCR and AGRs. Gas-cooled reactor data were supplied from the UK and for an old reactor in Spain, results are shown for the range 200-660 MWe. Due to technical design reasons, decommissioning with long-term storage is preferable to immediate dismantling. This is because there are some operations that can be carried out manually that are not possible for PWRs and BWRs. The costs for GCRs are higher than for other reactor types 1000-3000$US per kWe for the above capacity range. Part of the reason is not only due to the smaller size of GCR units but also there are larger volumes of radioactive waste that need to be processed. There are also increased man-power requirements.

6.11.1 Immediate Dismantling

Immediate dismantling decommissioning costs are summarised in Section 6.11.2, for the major reactor types of interest (IAEA-TECDOC-1084, 1999).

6.11.2.1 PWRs. Data were collected from a number of countries including Belgium, France, Korea, Netherlands, Sweden, UK and the US. A range of plants was considered covering the range 500-1400 MWe. The costs spanned between 150 and 700$US per kWe, reflecting large deviations in the key decommissioning parameters across the countries considered. These related particularly to differences in labour requirements, on the amount of decommissioning wastes and the duration of decommissioning activities. In general, it was found that the effect on reactor scale was small compared with differences between countries and differences between the estimates for the same reactor made at different times.

6.11.2.2 BWRs. Estimates were considered for Finland, Germany, Japan, Nether­lands, Sweden and the US, covering reactor units in the range 470-1300 MWe. Decommissioning costs were found to be in the range 170-650$US per kWe, i. e. similar to those for the PWRs (the effects of scale, however, were more visible than for BWRs), but again these were small compared with cross country variations of estimates with time.

6.11.2.3 VVERs. VVER plants have certain design differences from PWRs which impact on decommissioning costs, e. g. there is a high share of common systems and components in twin units.

In IAEA-TECDOC-1084 (1999), costs were presented for Bulgaria, Finland, Germany, Russia and for Slovakia for largely 440 MWe units of the 230 and 213 specification.

In general, the costs for VVERs were similar to those for PWRs and BWRs, except in Germany and Russia. Costs were higher in Germany and lower in Russia. These differences were not quantified but differences in labour rates and also in labour requirements were contributing factors.

6.11.2.4 HWRs. Data were available from Canada, Korea and India covering plants in the range 200-1300 MWe. In general, costs for HWRs are of the same order as for PWRs, BWRs and VVERs, but the cost variation from case to case appears less. However, the sample of plants considered was smaller. The costs for Korea were higher than Canada and India.

6.11.2.5 LWGRs (RBMK). Estimates were considered from Lithuania (1500 MWe Ignalina NPP) and Russia (1000 MWe plants). In general, assessments for LWGRs are lower than for other types, almost certainly reflecting low labour rates in these countries. The estimates were higher in Russia than in Lithuania; however, in the Russian data the costs of handling irradiated graphite were not included. In the Lithuanian data, these costs were taken into account.

Practically all the costs above were derived on the assumption of planned decommissioning. There may be cases when decommissioning is required urgently. This might be due to economic, safety, political or social reasons. In such cases, additional financial losses may be incurred (IAEA-TECDOC-1084, 1999).

Future plants may be of a number of designs since there is no general agreement on what features should be included in future designs. Further, there is no universal agreement on the design basis and how improvements in safety can be quantified. Further there are worldwide differences in licensing positions and engineering design standards across the world.

There has been progress in Germany and France towards developing harmonised approaches to design and licensing, e. g. within the EPR initiative.

In the US, the design of the AP600 has been certified by the USNRC as meeting accepted standards. This provides a demonstration to a potential regulator that the design has been certified against a particular standard. Generic design requirements can be derived from IAEA standards. These provide a norm for vendors to demonstrate how their particular designs meet these requirements.

Future designs will also have to satisfy URs and the evidence to date is that there are different vendors proposing a wide range of designs in the market. Different regulating bodies may be sympathetic to different designs and indeed different safety solutions for the same design. A harmonisation of design requirements and safety solutions (if they can be agreed by regulators) would clearly be desirable for vendors who could then seek design certification that would be acceptable in a number of different countries. For the same reasons, it would be an advantage for utilities.

Since there are likely to be operator applications for quite different designs in the future, there would clearly be a benefit in a more ‘technical neutral’ approach to licensing if it could be acceptable to the regulator, i. e. the licensing process would become less design specific than it is today.

The purpose of the reactor coolant system is to maintain cooling during normal operation and also during transients. There must be sufficient water inventory and safety water injection systems to ensure that water reaches the core. Heat is then transferred by circulation (forced or natural circulation) to the ultimate heat sink (Fil et al., 1999).

In many advanced plants, water to replenish any reduced water inventory in the primary circuit is stored entirely inside the containment. This provides additional protection against external events and other types of accident, e. g. containment bypass. Other features that are included to ensure protection of the primary circuit inventory include:

— pressurizer relief to a water storage tank;

— heat rejection to a water storage tank via heat exchangers;

— water storage tank joined with the containment sump;

— water storage tank, located high above the core for gravity driven injection, and;

— core make-up tanks (CMTs) at full circuit pressure to provide high pressure injection.

High-pressure passive injection systems are not present in currently operating reactors. The CMTs provide this function for AP1000 and AP600, (written AP1000/600 in this section). If the initiation set points are reached, valves open and cold water from the CMT flows into the reactor coolant system. If the CMT water level falls too low, then stepwise depressurisation of the reactor coolant system is initiated to ensure that medium and low pressure systems initiate.

Passive injection from accumulators is available in advanced passive designs, as it is in present generation plant. Modern designs have been optimised to increase system reliability and to broaden the pressure window of operation. Examples of such plants are AP1000/600, Mitsubishi APWR and Indian HWR designs. In addition, the Russian W-392 and W-407 designs adopt this principle. The Mitsubishi APWR accommodates an advanced accumulator system which eliminates the need for low pressure injection.

Passive low-pressure injection from the water storage tank is placed at high elevation across the core. Discharge can only take place when the reactor system pressure is at the last stage of depressurisation. Examples of such designs are AP1000/600 and the VVER-640/W-407 designs.

In advanced systems, sufficient heat transfer is attained provided there is sufficient water to cool the core. It is ensured by natural circulation from the heat source (core) to the heat sink, e. g. water storage tank in the AP1000/600 designs or the SGs in the Russian VVER-1000/W-392 design. These paths can exist in single — or two-phase water/steam modes. Different designs can make use of a range of different natural circulation paths.

Under accident conditions, heat is transferred to water tanks inside or outside the containment. Heat is then transferred to the surrounding atmosphere either via the containment shell or via a special heat exchanger, discussed below.

Passive feedwater systems have been considered in connection with the CANDU reactor design. There is an elevated tank above the boilers. Valves are opened to depressurise the boilers and allow flow by gravity.

The MSR is part of the Generation IV programme (The US Generation IV Implementation Strategy, 2003). In this design, the fuel is a liquid mixture of sodium, zirconium and uranium fluorides. The system is low pressure, with the coolant outlet temperature around 700°C. The power for the reference plant is 1000 MWe. It is a flexible system for actinide destruction. The economics are less favourable because of a large number of support systems for the maintenance of fuel and coolant. The system will require significant advances in chemistry plant design before it can realise a more mature design. The MS system will largely be for electricity production and plutonium and minor actinide destruction. Some example reactor types are given in Table 12.7.

KAERI in Korea are developing an integrated desalination system coupled with the system integrated modular advanced reactor (SMART) 330 MWt integral PWR system (Chang and Kim, 1998) (Figure 14.2). SMART is an evolutionary PWR incorporating passive safety features, simplified systems, cost-effective component fabrication and a load-follow operational capability. It is designed to be a co-generation system, which aims to produce 40,000 m3 per day of potable water, the remaining energy to be converted to electricity. The MSF and the RO options are both under review and investigation in Korea. The fundamentals of the SMART reactor design are shown in Seo (1997). Details on the licensing programme are given in Kim and Chang (1997).