A nuclear hybrid energy system would operate dynamically. In this sense, the system could be considered ‘load-following,’ in that the electricity produced would match the grid demand; however, the system is perhaps better described as ‘load-dynamic.’ Rather than modify the system production to match grid demand by varying the reactor power level (which would describe a load-following reactor), the system would be designed and controlled to maintain reactor power level while shifting output from the grid to the alternate output stream — such that the reactor itself is ‘load-dynamic.’ Hence, a nominally baseload nuclear power plant would be used flexibly within the system to produce electricity to mitigate either variable renewable input to the integrated system or variable electricity demand from the grid. Increased efficiency can thus be achieved by matching electrical output to demand while utilizing excess generation capacity for other purposes when it is available. The envisioned system operation offers a new approach to system optimization from a financial perspective, resulting from the opportunity to produce revenue from a variety of product streams while avoiding inefficiencies of underutilized capital. This approach recognizes and takes advantage of the benefits of operating a nuclear power plant at its nameplate capacity in a steady state fashion to reduce wear on nuclear system components. It allows the nuclear plant to operate at a high capacity factor to recoup the relatively high cost of capital compared to fossil fuel plants, while taking advantage of the relatively low recurring cost of uranium fuel sources.

At the present time, the focus of this ART research area is on conducting basic R&D on new materials to enable the ultimate deployment of innovative A-SMR designs. It is aimed at addressing key long-term design needs for application of advanced materials. High-temperature design methodology is an enabling reactor technology for most A-SMR concepts. Removal of unnecessary conservatism in design methodology will lead to more flexibility in the construction and operation of A-SMRs. Gaining mechanistic understanding of long-term deformation and degradation mechanisms such as creep, creep-fatigue, and thermal aging will provide guidance on the extrapolation of accelerated time-at-temperature design data for a 60-year design life, and beyond, with higher confidence [4]. Equally important in terms of realizing the complete development of these advanced materials is the code qualification and acceptance of these candidate materials.

DOE-NE’s Office of Advanced Reactor Technologies is responsible for directing the ART R&D program for the development of advanced reactor concepts that focused on larger size advanced concepts under its former Advanced Reactor Concepts R&D program, which included the high-temperature gas-cooled reactor in support of DOE’s Next Generation Nuclear Plant (NGNP) program and sodium — cooled fast reactors (SFRs). The objective is to develop innovative technologies that resolve key feasibility and performance challenges and that have the potential to provide significant safety, economic improvements, and lower fabrication, construction, and operations costs. As for any new class of reactors, materials are one of the key technology areas that must be understood and demonstrated in terms of performance. Given that advanced concepts for small reactors would likely use the same sets of coolants — helium, liquid metals, and fluoride salts — these materials R&D projects are applicable for both large and small advanced reactors.

The current scope of R&D in this research area is categorized into three areas:

• high-temperature design methodology;

• materials development for A-SMRs; and

• materials property data to support R&D and code case development

Current materials research areas are listed in Table 14.2 including a brief summary of the scope for each research area.

Figure 14.2 presents the test facilities as just described in Table 14.2 for generating experimental data for Alloy 617. The two servo-hydraulic test machines shown are coupled electronically to allow for applying equal amounts of elongations at all times where the sum of the loads is equal to the preselected total load at all times. This two-bar thermal ratcheting test enables development and verification of material models for inelastic analysis and verification of elastic-perfectly plastic strain limits code case. This work is being conducted at DOE’s Oak Ridge National Laboratory.

The function of the severe accident mitigation system (SAMS) is to prevent the egress of molten corium resulting from a severe accident out of the containment. This egress of corium can be avoided due to the design characteristics of the reactor cavity and containment together with the operation of the safety systems. A small air gap under the RPV is filled with water from the containment spray system (CSS) for a severe accident. The in-vessel cooling prevents an egress of the corium out of the RPV. In addition, the water in the in-containment refuelling water storage tank (IRWST) provides an external cooling of the RPV and prevents an egress of the corium out of the RPV. Passive autocatalytic recombiners (PAR) are provided in the containment to remove the explosive hydrogen generated during a severe accident.

15.4.8.3 Safety assessments

The safety analysis of the SMART design is conducted in which methods of both deterministic and probabilistic analyses are applied. It is demonstrated that the SMART is designed to be capable of meeting the prescribed limits acceptance criteria.

Many fuel operating limits (performance and safety) for LWRs (including iPWRs) are directly related to the maximum linear power density of the fuel. As a first approximation, for a given fuel rod geometry, the peak fuel temperature, the surface heat flux, the decay heat generation rate, and the stored thermal energy are proportional to the fuel rod linear power density. As such, it is a key role of the nuclear designer to maximize total power output, while minimizing power peaking (and hence the linear power density), both for the fuel assembly as a whole in the core, and for the individual fuel rods.

The attempt by the nuclear designer to reduce the maximum rod power in the core does not negate the need for a full thermo-mechanical and thermal hydraulic analyses for the reactor, but it does aid the overall design process for the fuel and the core. The key considerations for the nuclear designer are that the power distribution of the core must be such that:

• the fuel will not exceed specified peak linear heat rates under normal operating conditions

— a value that is determined as part of the safety analysis for the core;

• the fuel does not exceed the design basis set for departure from nucleate boiling (DNB)

— DNB is the point at which the heat transfer from a fuel rod rapidly decreases due to the

insulating effect of a steam blanket that forms on the rod surface when the temperature continues to increase — this value is calculated by thermal hydraulics analysis;

• under normal and abnormal operating conditions, the maximum linear heat rate will not cause the fuel to melt — a value provided to the nuclear designer;

• the fuel rod power and burnup are consistent with the assumptions and analysis in the fuel rod thermo-mechanical performance analysis — a subsequent analysis, once the nuclear design is completed and can be iterative if subsequent violations in design criteria for the fuel performance are found.

To demonstrate these requirements, the nuclear designer has also to complete the analysis under a range of extreme power shapes and variations, not just under nominal reactor operating conditions. The extreme shapes used reflect experience at operating reactors (in the first instance based on large PWRs), but the shapes are chosen to be deliberately conservative, e. g., axial power skewed by control rod insertion/withdrawal, and load follow conditions.

For the fuel assembly power in a PWR or in a specific iPWR, there is no design limit or constraint. (It is often said that fuel rods, not fuel assemblies fail!). But it is an important design optimization consideration in order to ensure that those assemblies with continually higher powers do not lead to limitations in the fuel rod performance and its associated integrity (e. g., violations in rod internal pressure due to fission gas release), or that assembly burnups do not vary too significantly within the batch such that the fuel assemblies are used inefficiently (see Section 4.3.3 for details).

Temperature is one of the basic measurements required in a nuclear reactor. Typical PWR temperature measurements have been accomplished using either RTDs or thermocouples. Thermocouples have been used in core temperature measurements where the temperature can be extremely hot. RTDs are typically used in the hot and cold legs of a PWR for reactor coolant temperature. Qualified versions of thermocouples and RTDs have been developed and there is much longevity data on these devices.

Another temperature measurement device, not currently used in nuclear power reactors, but with some nuclear potential, is the Johnson noise thermometer (JNT). This device is based on the thermal fluctuation of a conductor. It uses the mean square thermal noise voltage generated by a resistor or capacitor to establish temperature. The application of this methodology is highly accurate for temperature measurements but the voltage measured is in the range of micro-volts and spread over a wide bandwidth. Although the use of amplifiers and digital filters enhances the signal strength, there are many practical hurdles that must be solved for an industrial use. One of the key advantages for the JNT is the elimination of typical RTD drift. A JNT device is self-calibrating and it has the ability to maintain its accuracy indefinitely without the need for human intervention for periodic calibration. There are currently no commercial models available for purchase; however, with the ability to maintain its calibration indefinitely, the JNT may be a good new solution for iPWRs.7

It is likely that iPWRs will continue to use thermocouple and RTD methods to measure core and coolant temperatures, but the traditional mountings and instrument size may have to be re-engineered for the unique and smaller geometries involved. In these conditions, alternate temperature measurement devices might be considered.

Optical fiber technology offers another new approach for temperature sensors. This method has not been used in traditional PWR applications and represents a new option for safety system temperature measurements. Over the last decade, Luna Innovations has introduced several new fiber optic sensor technology options for making temperature measurements (Figure 6.3). Notably, Luna is able to use

a single, inexpensive, and commercially available optical fiber to make distributed measurements; characterizing thousands of temperature or strain measurements made at points along the length of the fiber in a fraction of a second. One demonstration of this technology was conducted during a short-term test in a nuclear research reactor. Temperatures were calculated from Rayleigh backscatter signals, without the use of fiber Bragg gratings, every 1 cm along a variety of commercially available single mode C-band telecommunication optical fibers. Luna’s distributed sensing technology can provide high spatial resolution temperature measurements in a compact format, without suffering from electromagnetic interference, and along fiber lengths up to 30 m for selected radiation environments.8

Sporian Microsystems Inc. offers another temperature sensing option in the MEMS (micro-electro-mechanical sensors) technology field that has a high range of temperature sensitivity (up to 1300 °C) with a high degree of accuracy (Figure 6.4). The device is composed of polymer-derived ceramic material, is extremely small, and has been tested in research reactors. Sporian is currently designing these sensors under a Department of Energy (DOE) small business contract.9

The nature of HSIs for advanced reactors can be better understood if they are characterised in terms of the ‘operational domains’ where they are used and also the architecture of HSIs themselves. We can define ‘operational domain’ as the physical, structural, logical or functional characteristics that distinguish different areas in the plant where work is performed and where humans interact with technology.

There are nine distinct work domains in new NPPs where HSIs will play an important role. Some of these are dedicated and enclosed areas; other areas inside or outside the plant have variable boundaries within which functions are performed:

• Control room — This is an enclosed area, often in close proximity to the reactor and turbine building.

• LCSs throughout the plant, typically consisting of one or more small control panels.

• Materials and waste fuel handling. Forklifts, cranes and similar tools are typically found in these domains.

• Refuelling operations, using specialised equipment to handle radioactive materials.

• Maintenance inside and outside the plant, using a range of conventional and specialised tools.

• Outage Control Centre, characterised by many desktop computers, large displays, printers, planning boards and communication equipment.

• Fuel processing installations, characterised by specialised equipment to handle hazardous materials, such as robotic manipulators.

• TSC. This centre is typically somewhere on site and like the outage control centre would have large displays, but also limited HSIs that provide access to some of the displays found in the control room.

• EOF. This facility is located at a more remote location outside the plant perimeter and would also have access to data from the control room.

Most of these domains have a greater or lesser degree of interdependence, overlap or redundancy, as shown in Figure 7.2. This diagram indicates that the control room dominates in terms of range and number of HSIs applied in that environment. The arrows indicate the potential links or relationships between domains. Six related operational domains, some of which may functionally overlap with each other or with the main control room, are the LCSs throughout the plant, HSIs used for materials handling, refuelling operations, fuel and waste handling, and HSIs used in maintenance and outage management. Fuel processing plants could feature strongly

in future at plants using fast breeder reactors and fuel reprocessing. The interfaces between the control room and the related, or interdependent, domains consist primarily of status displays and communication devices. These interfaces enable the operating crew to maintain situation awareness of all activities throughout the plant and during all conditions.

Two other domains interface with the control room only during upset or emergency conditions: the TSC and the EOF. The most important domains are described briefly below.

The requirements and regulations related to the off-site emergency planning are different in different countries, but the overall considerations and impact are similar. The impact of an EPZ extending beyond the NPP site boundary has both social and economic consequences. Socially, it projects the image of nuclear power being unlike other industries with the potential of impacting population outside the site boundary. (This is true of course for many other industries, but it is not necessarily recognized in public perception.) Economically, a large EPZ has associated large cost, resulting from the need for redundant evacuation routes, control of population density, etc. Large NPPs, frequently built in multiples, may accommodate this penalty and offset it by their large power output.

SMRs, on the other hand, are ideally suited for a range of power outputs, for co-generation (including for example district heating), for placement closer to the end-user (thus within higher population areas), and for more diffuse siting. At the same time, their enhanced safety characteristics promote the possibility of licensing without the requirement for off-site emergency response, in other words, collapsing the EPZ to the site boundary.

Indeed, SMRs provide the technical basis for implementing such a change in EPZ:

• They have a smaller source term than large reactors, at least on a per-reactor basis. While a common-cause multiple unit failure cannot be excluded, probabilistically it is not very likely.

• iPWR SMR tend to have smaller CDF and LERF — typical CDF values are on the order of 10-7, significantly less than in the current Generation-II plants.

• Smaller source term combined with smaller CDF/LERF results in significantly reduced probability of consequences.

Thus, from the purely technical viewpoint, it should be justifiable to reduce the size of the EPZ for SMRs, and, with adequate safety characteristics, this reduction would result in collapsing the EPZ to the site boundary (i. e. no off-site emergency planning is needed).

In practice, the situation is more complex. The current US regulation, for example, institutes a uniform 10 mile EPZ size without regard to the plant specifics. Thus, a new regulation for SMRs is needed where the EPZ size would be based on the risk — based or risk-informed approach; in the US, the latter is more likely to be acceptable. This issue has been raised several times in the past. The reader is encouraged to consult Carelli et al. (2008) for an example of a proposed technical approach of EPZ evaluation.

The recent strong interest in SMRs has led the NRC to explore the possibility of a generic resolution to emergency planning requirement issues for SMRs. The staff proposed a new approach for SMRs in SECY-11-0152 dated October 28, 2011 (NRC, 2011). Evaluating, accepting and implementing the intrinsic capability of SMRs to limit their EPZ to the site boundary is one of the top-level targets pursued by the SMR developers.

Economic competitiveness of a power generation technology depends on the ability to provide electricity with affordable LCOE, and/or of repaying the investors by means of adequate cash inflows, granting a minimum acceptable capital remuneration compared to the level of risk and to the PBT duration. Given the relevance of the capital costs in the nuclear electricity generation cost (i. e. given the ‘capital intensity’ of the nuclear investment), capital cost, including overnight construction and financial costs, has a relevant impact on the key economic performance indicators.

With very few SMR projects under construction and no actual data on overnight actual costs, cost estimation of SMRs is usually performed on a top-down basis, as recommended in Section 10.1.2, starting from available information on large, advanced pressurized water reactor (PWR) units, as a starting reference cost. (The only SMR reactors under construction in 2014 were CAREM in Argentina, HTR-PM in China and the twin barge-mounted KLT-40S in Russia (Akademik Lomonosov), planned to be located near Vilyuchinsk. Construction was started in 2007 and, owing to some economic-financial problems, the plant is now expected to be completed in 2016 (http://www. world-nuclear. org/info/inf33.html)). Carelli et al. (2010) present a parametric methodology to compute the overnight construction cost of SMRs, based on the application of dimensionless coefficients, related to the most important differential economic features between SMRs and LRs: e. g. expected learning effect, degree of modularization, co-siting economies and simplified design. Many of these factors are dependent on the number of units built on the same site and on the plant output size.

The above-mentioned design-related economies, learning effects on costs, plant modularization, and co-siting economies account for an expected reduction of multiple plant construction cost. Based on these factors, it is estimated that the SMR economic paradigm might bring unit construction cost in line with expected costs of Generation III+ (GENIII+) large PWRs, thus overcoming the loss of economy of scale. Figure 10.6 provides a qualitative sketch of SMR economic features’ recovering the loss of economy of scale on unit construction cost, as far as multiple SMRs are considered an alternative investment opportunity to LR power stations, with the same overall power at site-level.

Actual information on LRs under construction (in western countries) gives evidence of relevant time-schedule and cost overruns. It must be highlighted that this comparison applies on SMR versus large NPP expected costs. This means that capital cost overruns, which seem to systematically affect actual costs of large NPP projects, are not considered. When actual costs of construction are considered, it is expected that, as stated in Section 10.2.2, SMRs might have better control on construction schedule and costs, and higher probability to meet capital budgeting. The main assumption is that, the simpler the design, the easier the procurement, manufacturing and assembling process and the project management.

In any case, a possible cost overrun on a single SMR unit would necessarily have a lower incidence on total investment than in an LR, due to the lower cost of a single SMR.

If a time/cost schedule mismatch affected the initial SMR unit(s), the simple fact of fractioning a nuclear power station into multiple smaller units makes unlikely that such a mismatch could be repeated on all of the units, on account of learning and improved practices in the supply process, construction work and project management.

Thus, as far as actual construction costs are considered, including possible cost overruns and financial interests escalation, SMRs might improve their cost competitiveness against LRs, as compared to the mere theoretical expectations.

Economy of scale has been the key driver of the nuclear industry over the past. The evolution of nuclear power technology is characterized by a constant trend in the output size increase. The US utilities converged to 1000-1400 MWe sized plants, French NPPs were scaled from 950 to 1550 MWe in the 1971-1999 period, up to the recent 1600 MWe European Pressurized Reactor (EPR). As a capital-intensive industry, nuclear power generation technology pursued the economy of scale law, to decrease the incidence of fixed costs over a higher output base.

In principle, SMRs are heavily penalized by the loss of economy of scale: applying a typical scale exponential law (usually with coefficient in the range 0.6-0.7), a stand-alone 335 MWe SMR may bear 70% cost increase on a unit base (7/kWe) over a 1340 MWe LR (Carelli et al., 2007a). SMR units with smaller size would bear a greater penalty (up to 350%); this should be recovered by other means, in order to uphold cost competitiveness.

Nevertheless, the evidence of construction cost escalation of GW-scale reactors triggers considerations about the applicability of the economy of scale law on NPPs (Grubler, 2010): an increase in the plant scale apparently produces an increase in the intrinsic complexity, which challenges the project management and other activities in the plant design, construction and assembly. This translates in construction schedule delays and dramatic cost overruns. (For example Olkiluoto, Flamanville, South Texas Project, Vogtle and, more recently, Hinkley Point 3, Olkiluoto and Flamanville are under construction, while US-based projects are in the early site preparation phase; therefore the cost overrun have different nature in the two cases.) Projected cost and the lead time of the new projects under construction in Europe or under construction in US, have all been dramatically revised upwards, with a rate of increase per year of delay in the plant commissioning in excess of 20% (Table 10.6). This cost

1000

"D c ro

CD ГО <D

500

escalation coherent with a historical trend of construction cost increase over time (Figure 10.7).

A detailed analysis of the French NPP fleet (all PWRs) shows that construction costs and schedule have increased over time with the size of the plants (Figures 10.8 and 10.9). The French PWR program exhibited substantial real cost escalation, in spite of a unique institutional setting allowing centralized decision making, regulatory stability and dedicated efforts for standardized reactor designs. This evidence challenges the applicability of a learning economy on NPP construction, as far as ‘traditional’ NPP are considered, without introducing the concepts of design simplification and modularization, discussed in the following Sections 10.5.1 and 10.5.2.

CP0 ACPI ДСР2 □ P4 P’4 • N4

□ 500 MW GCR A 950 MW PWR

1350 MW PWR P4

□ 1350 PWR P’4

• 1550 MW PWR N4 О Super-Phenix О 1650 MW EPR

1965 1970 1975 1980 1985 1990 1995 2000 2005

Year of grid connection

Figure 10.9 Construction time of French reactors (construction start to first grid-connection, in months). Data on 1650 MWe EPR reactor in Flamanville is a projection submitted by the French authorities to the IAEA. Source: IAEA PRIS Data Base (2009) http://www. iaea. org/ pris/home. espx. GCR = graphite gas reactors (Grubler, 2010).

The production line is based around a continuous flow of product through a factory and has been successfully demonstrated for food products, telecoms products, cars and aircraft. However, for these products the market volume of sales is assured before the factory capital investment is made. With a product that is the physical size of a small reactor vessel the investment in workstations and tooling is significant. Equally for a small reactor, manufactured in a factory at a specific volume of units per year in a repetitive manner, it can be conceived that there is an economic case to support a level of investment in jigs and tooling. This appetite for facility investment is tempered by the realism of the market demand. With an emerging customer base there is no basis to justify the capital investment on facilities at the outset — compromising the appeal of a pure production line implementation. There are three potential solutions to this;

1. Investment based on rational industrial business risk — ‘Step of faith’.

2. State involvement — state ownership of the same business risk.

3. Alternative implementation methods.

The step of faith is a little lighthearted; however, it makes the point that all business investment contains an element of risk. The risk with lower levels of customer orders prior to establishing a small reactor facility could restrict the net level of investment. This restriction could in turn limit the ability of the factory to deliver product at mature volumes in an effective and cost-efficient manner in the longer term.

The second option offered is state involvement. The options for state involvement will also vary from region to region across the globe. The US Department of Energy (DOE) has recognised this with their four-phase approach to the acceleration of small reactor deployment. Their third phase recognises this ‘step of faith’ transition from early technology adopters to full-scale production. This state stimulus can come in many forms ranging from a public-private partnership to a commitment to power purchase.

The third option is an alternative manufacturing flowline assembly. Flowline is an interim option between a full-scale production line and individual batch manufacturing, and incorporates the best features of both approaches. With batch production manufacture the unit is produced in a series of stages, each stage building on the preceding stage. Running repeat operations in batches does recover some of the operating efficiency. The point with batch production is that the item remains stationary throughout and the workstation is reconfigured around it for each operation. The time to set up and re-configure between each of these operations leads to delays. For assured processes differences from set-up to set-up cannot be tolerated. Batch manufacture gives greater opportunity for batch to batch tolerance variation, which is a bad thing. However, it also creates a window for product customisation. In our circumstance this could be the difference between a generator protection system that operates at 60 Hz instead of 50 Hz.

Multiple SMR concepts are currently being studied and are in various stages of development. These concepts range from advanced LWR concepts, specifically integral pressurized water reactors (iPWRs), to advanced concepts such as the high-temperature gas-cooled reactor (HTGR), fluoride salt-cooled high-temperature reactor (FHR) and liquid-metal-cooled reactor (LMR) concepts, all of which were identified (in a conceptual sense) as advanced reactor concepts warranting further study by the Generation IV International Forum (GIF) at a meeting held in 2000 [6]. The fundamental neutronic and thermal hydraulic differences that exist between a water-cooled, thermal spectrum reactor, cooled using either natural convection or forced convection, thermal spectrum reactors using alternate coolants (HTGR, FHR), and fast spectrum reactor (LMR) concepts could result in significant differences in technically feasible HES architectures and their associated system operating parameters. The operating temperature for each of these concepts is the primary driver in the available options for coupling to different process applications. Key features for each of these reactor types are summarized in Table 13.1.

In defining operating strategies for an integrated hybrid system, which by definition

should be designed to accommodate variability, it should be noted that the preferred approach is to operate the reactor or reactors at the nameplate power level over long periods of time (i. e. between standard maintenance and refueling cycles) to minimize stress on the system components and to maximize core life. Cycling from approximately 65 to 100 percent is achievable for much of the core life of LWRs, but this cycling should not be the standard approach to system operation. To maintain robust reactor performance at a steady power level, the anticipated cycling of power demand and the subsequent need to switch between production of electricity and direct application of thermal energy should be buffered from the reactor. Buffering might be accomplished by use of energy storage systems and possibly be aided by selection of a reactor design that would have a protracted response time to changes on the secondary side of the system (e. g. one that is least coupled to the balance of plant). Significant analysis must be performed to assess the overall operational performance characteristics of reactor concepts that could fit within the NHES model, including system response times (e. g. as it is affected by reactivity feedback coefficients, thermal capacity of the coolant, requirement for an intermediate heat exchanger, etc.), reactor coupling with the balance of plant, possible load leveling schemes, etc.