© 2012 European Nuclear Society Rue Belliard 65 1040 Brussels, Belgium Phone + 32 2 505 30 54 Fax +32 2 502 39 02 E-mail [email protected] Internet www.euronuclear.org
ISBN 978-92-95064-14-0
These transactions contain all contributions submitted by 7 December 2012.
The content of contributions published in this book reflects solely the opinions of the authors concerned. The European Nuclear Society is not responsible for details published and the accuracy of data presented.
2 of 96
ENC2012-A0026 Development of a thermohydraulic model of the Lazaro Chueca, A. (1); Ammirabile, L. (1); European Sodium Fast Reactor (ESFR) using Martorell, S. (2) the system code TRACE. 1 - JRC-IET, Netherlands 2 - Universidad Politecnica de Valencia, Spain ENC2012-A0028 Generation IV Technology Status including Anderson, G. (1); Lillington, J. (1) recent R & D Activities in ANSWERS 1 - AMEC, United Kingdom ENC2012-A0053 Preliminary Design Assessment of the Molten Merle-Lucotte, E. (1); Allibert, M. (1); Salt Fast Reactor Brovchenko, M. (1); Ghetta, V. (1); Heuer, D. (1); Rubiolo, P. (1); Laureau, A. (1) 1 - LPSC-IN2P3-CNRS / UJF / Grenoble INP, France ENC2012-A0078 Development of materials to withstand the Shepherd, D. (1) extreme, irradiated environments in advanced 1 - National Nuclear Laboratory, United Kingdom nuclear fission reactors ENC2012-A0126 ARCHER:- Material and component challenges Buckthorpe, D. (1) for the Advanced High Temperature Reactor 1 - AMEC, United Kingdom ENC2012-A0258 Pressure Drop Analysis of a Pressure-Tube Type Peiman, W. (1); Saltanov, E. (1); Pioro, I. SuperCritical Water-Cooled Reactor (SCWR) (1); Gabriel, K. (1) 1 - University of Ontario Institute of Technology, Canada ENC2012-A0003 SPES3: THE INTEGRAL FACILITY FOR Ferri, R. (1); Achilli, A. (1); Cattadori, G. SAFETY EXPERIMENTS ON SMALL AND (1); Bianchi, F. (1); Luce, A. (1); Monti, S. MEDIUM SIZED REACTORS (1); Meloni, P. (2); Ricotti, M. E. (3) 1 - SIET S.p.A., Italy 2 - ENEA, Italy 3 - POLITECNICO DI MILANO, Italy ENC2012-A0088 Review of three families of Small Modular Lecomte, M. (1); Beon, J.-Y. (1); Reactors (SMRs): land-based; floating; Poimboeuf, J.-M. (1); Vignon, D. (1) immersed 1 - NucAdvisor, France ENC2012-A0119 PRELIMINARY EVALUATION OF A SEVERE Lo frano, R. (1); Baudanza, V. (1); FLOODING EFFECTS ON AN INNOVATIVE Forasassi, G. (1) SMR. 1 - DIMNP-University of Pisa, Italy ENC2012-A0161 European Design Study on Supercritical Water Schulenberg, T. (1); Starflinger, J. (2); Cooled Reactors Class, A. (1) 1 - Karlsruhe Institute of Technology, Germany 2 - University of Stuttgart, Germany ENC2012-A0202 HEAT-TRANSFER CORRELATIONS FOR Pioro, I. (1); Mokry, S. (1); Gupta, S. (1); SUPERCRITICAL WATER AND CARBON Saltanov, E. (1) DIOXIDE FLOWING IN VERTICAL BARE 1 - University of Ontario Institute of Technology, TUBES Canada ENC2012-A0233 Safety analysis of a sodium-cooled fast reactor Perez-Martin, S. (1); Hering, W. (1); with transmutation capabilities Kruessmann, R. (1); Lemasson, D. (2); Massara, S. (2); Pfrang, W. (1); Ponomarev, A. (1); Schikorr, M. (1); Struwe, D. (1); Verwaerde, D. (2) 1 - Karlsruhe Institute of Technology, Germany 2 - EDF, France
3 of 96
ADVANCED REACTORS
4 of 96
ADVANCED REACTORS I
5 of 96
Development of a thermohydraulic model of the European Sodium Fast Reactor (ESFR) using the system code TRACE.
Aurelio Lazaro1,2, Luca Ammirabile1, S. Martorell2 , G. Verdu2 1 - European Commission (EURATOM), Joint Research Centre Institute for Energy and Transport Westerduinweg 3, 1755 LE, Petten, The Netherlands Telf: +31(0)224 54 5446 Email: [email protected]
2- Departamento de Ingeniería Química y Nuclear. Universidad Politécnica de Valencia Cami de Vera, sn, 46021 Valencia.
Abstract – One of the main goals of the Generation IV International Forum (GIF) nuclear energy systems is to excel in safety and reliability. To pursue such objective, the development of computational tools able to simulate operation conditions that may be critical for the safety of these innovative reactor concepts is essential. As part of the EURATOM contribution to GIF, the FP-7 CP-ESFR project has been launched to study a Sodium Fast Reactor (SFR) design. This paper presents how a thermohydraulic model of the ESFR plant has been developed using the best-estimate system code TRACE. The model simulates the primary, secondary and tertiary system. The primary system includes the reactor core with point kinetic neutronics implemented and it is able to analyse different accident transient scenarios. The work presented in the paper provides the first steps for the development of the transient part of an integrated safety analysis platform with capabilities to perform detailed simulations of the reactor dynamic thermohydraulic-neutronic coupling techniques.
I. INTRODUCTION includes “working horse” designs (pool and loop type, oxide and carbide fuel) that describe cores and systems The Generation IV International Forum (GIF) is an which allow developing and testing different options in a international initiative to develop a new generation of common agreed basis. The design that has been considered nuclear power plants that will excel in safety and reliability in this paper is the pool-type oxide-fuel concept. and will improve in other key-issues as the waste management or the optimization of the fuel usage [1]. This new technology requires specific tools to assess its safety behaviour. To pursue such an objective, the Among the different proposals framed in the GIF, the development and validation of computational tools able to Sodium Cooled Fast Reactor (SFR) has a unique position analyse transients that may affect the plant safety is since related projects have already been developed in essential. several countries for nearly 50 years. A demonstration project within the “European Sustainable Nuclear Industrial In this line, the JRC-IET is developing an integrated Initiative” (ESNII) is planned. The so called ASTRID safety analysis tool [3] with the objective to perform an prototype would be the first Generation IV system based on integrated core and safety analysis of nuclear reactor SFR concept [10]. systems. This platform will assist to fulfil the JRC-IET task to provide independent safety assessment of nuclear In addition to these initiatives, the Collaborative reactors and to contribute in this way to the policy support Project on the European Sodium Fast Reactors, framed in on nuclear safety in the European Union. the 7th Framework Programme, is part of the EURATOM contribution to GIF and merges the efforts of 24 European For its implementation to the ESFR pool-type concept Partners to indentify, organize and implement the R&D it has been pointed as system code the thermo-hydraulic effort needed to develop such a project [2]. This project code TRACE v5.0 [6]. In the following section of this
6 of 96
paper it will be explained how a one-dimensional model of B) and 24 Control and Shutdown Device (CSD) that has been built up and it will be compared with an contains natural boron carbide (19.9% B). equivalent model implemented in the thermo-hydraulic code RELAP5 [5]. A picture of the core layout is shown in Figure 2.
II. THE ONE-DIMENSIONAL MODEL
Two one-dimensional models have been developed in TRACE and RELAP5 for the ESFR pool-type oxide-core plant following the technical specifications fixed in the Working Horse documents .
The plant layout is composed by the pool-type primary system where the heat is generated in the reactor core and transferred via the IHX to the secondary system. Here the heat is extracted from the IHXs and conveyed to the tertiary system where the steam generators produce the steam which drives the turbines, closing the thermodynamic cycle. The secondary system consists of 6 loops, which are thermally linked through the IHX with the pool-type primary and 6 tertiary loops, so that the tertiary system is formed by 36 separate circuits. Fig. 2. Oxide core layout [2].
The pool concept is featured by nearly all the primary The main thermodynamic variables in nominal sodium coolant inside the reactor vessel, enclosing the conditions are listed in Table I. primary pumps and the Intermediate Heat Exchangers (IHX) in addition to the internal structures surrounding the TABLE I core [7]. This configuration is shown in Figure 1. Nominal conditions of the ESFR pool-type plant.
Variable Reactor Power (MWth) 3600 Core inlet temperature (˚C) 395 Core outlet temperature (C) 545 Average core structure temperature (˚C) 470 Average fuel temperature (˚C) 1227
II.A. The primary system
The thermo-hydraulic configuration of the core is modelled by seven different components according to the core power distribution as proposed in the core technical specifications. These seven components correspond to the hot channel, the inner core zone, the outer core zone (two components), the central dummy assembly and control assemblies, the reflector and the by-pass. Fig. 1. ESFR pool-type concept [2]. These seven core regions (Figure 3) are modelled by a The core is composed by an inner and a outer zone PIPE component, and only six (by-pass excluded) are with different Pu mass content. There are 225 inner fuel attached to a heat structure (HTSTR component) to subassemblies with a Pu mass content of 14.5% and 228 simulate the heat transfer to the coolant. These core outer core sub-assemblies with a mass content of 17%. The components are connected both, to the hot (upper) and cold control rod system is composed of 9 Diverse Shutdown (lower) plena of the plant. The fuel thermal conductivity Devices (DSD) that contains enriched carbon carbide (90% was evaluated according to the Phillipponneau model [9].
7 of 96
The radial core expansion (diagrid) feedback is calculated based on the expansion of the core support diagrid. The thermal inertia between the sodium core inlet temperature and the average diagrid temperature is taken into account considering a structure of 5 m diameter and 0.05 m thickness.
Fig. 3. Core channel model [2]. The control rod feedback is calculated taking into account the expansion of the control rod bar and the axial The hot (upper) plenum region is modelled with 3 expansion of the core due to the variation in the sodium PIPE components according to the geometrical core outlet and inlet temperature respectively. configuration of the design. They convey the hot coolant towards the Intermediate Heat Exchanger (IHX), where the II.B. The secondary system heat is transferred to the secondary loop. The secondary system is simplified into two main pipe This IHX is the interface between primary and components that simulate the secondary side of the IHX secondary systems with the secondary sodium flowing and SG and two pipe components that close the circuit. vertically inside the heat exchanger tubes where it is heated by the primary sodium flowing downwards on the shell Their geometrical characteristics have been set up to side. provide the adequate sodium inventory in the secondary system that plays an important role in the transient After the heat is exchanged to the secondary loop, the behaviour of the plant. coolant reaches the cold (lower) plenum region, which is A second BREAK component acts as a pressure modelled by 4 PIPE components taking into account the control to keep the secondary system at atmospheric different cross sections that the coolant encounters pressure. conveying into the core inlet. II.C. The tertiary system A BREAK component acts as a pressure control to keep the primary system at atmospheric pressure. The tertiary has been modelled lumping the 36 steam generators composing the system into one equivalent steam A point kinetic (PK) model has been implemented to generator. A pipe component with its heat structure model the neutronic behaviour of the core. The main connected to the secondary system simulates the tube reactivity feedbacks have been taken into account bundle. A FILL component represents the lower plenum following the recommendations in the core description. The and determines the inlet water mass flow while a BREAK main reactivity feedbacks are calculated as follows: component is the upper plenum where the design pressure is defined. The Doppler feedback is calculated using the BOL coefficient as function of the average fuel temperature II.D. Modifications in the code computed by averaging the fuel temperature over the whole core fuel volume. Both codes RELAP5 and TRACE were modified to be able to deal with liquid Sodium as a coolant. RELAP5 is a The coolant density feedback is calculated based on code designed for light water reactors analysis so the liquid the axial weight of its reactivity coefficient and the relative Sodium thermodynamic properties should be specifically worth of the core channels. implemented. TRACE code has this properties built in by default. Nevertheless, some modifications were done in The fuel axial expansion feedback is calculated as order to be able to work with Sodium and Steam Tables function of average fuel temperature (like for Doppler) (IAPWS-97) [8] simultaneously. These tables are needed to taking into account that at BOL condition, the fuel is obtain accurate calculations of the thermodynamic assumed to expand independently from the cladding. properties of the steam that are reached in the Steam Generators. The clad axial expansion is calculated as function of average cladding temperature taking into account that at In addition, the Ushakov’s Heat Transfer Coefficient BOL condition, the cladding is assumed to expand Correlation was implemented in both codes as it is a better independently from the fuel. approach to simulate the heat transfer to liquid metals than the default built in correlations [4].
8 of 96
In the Figure 4, it is shown the mass flow in both, II.E. Steady state results primary and secondary systems, all along the transient. As they are imposed to trigger the transient their behaviour is The steady-state nominal power conditions for ESFR exactly the same between the two codes. (pool-type) plant as computed by TRACE and RELAP5 are compared in Table II. The difference in temperatures is within 5 °C.
There is in general a good agreement with the plant technical specifications. It can be noticed that a high primary sodium mass flow is needed to provide the design ΔT of 150 ºC across the core with 3600 MW power.
The different mass flows are due to the fact that the codes use different (lower) specific heat for sodium than in the technical specifications. This results in a higher mass flow needed to compensate for a lower specific heat.
TABLE II
TRACE and RELAP5 nominal conditions for ESFR plant Fig. 4 . Primary and secondary massflows.
PARAMETER ESFR TRACE RELAP5 Figure 5 shows the comparison between the two codes P. Mass flow (kg/s) 19535 20692 20860 of the core, IHX and SGs power. S. Mass flow (kg/s) 15330 16444 16907 Core Inlet (˚C) 395 391 395 Both codes present similar results in the evolution of Core Outlet (˚C) 545 545 545 the power. It is worth to note that in the simulation and for IHX Inlet (˚C) 340 335 332 both codes a constant gap size model was adopted that IHX Outlet (˚C) 525 522 517 might overestimate the reactor power compared to a SG outlet Temp. (˚C) 490 493 487 variable gap size. SG pressure (bar) 185 185 185 SG mass flow (kg/s) 1650 1650 1650 The oscillations that appear in the SG power are due to the change of the flow regime and of the heat transfer in the II.F. Transient results SGs following the change in the mass flow during the transient. In order to evaluate the neutronics and thermal- hydraulic behaviour of the different system codes involved in the project, a reference transient was fixed. The transient duration is 700 seconds and it consists in the application of three simultaneous perturbations in the flowrate of the three coolant systems. These perturbations are listed in Table III.
TABLE III Transient definition Perturbations Description 1 Reduction of the primary flowrate to 40 % 2 Reduction of the secondary flowrate to 40 % 3 Reduction of the feedwater in SG to 50%
At nominal steady-state power conditions these perturbations are applied triggering a transient with strong Fig. 5. Core, IHX and SG powers. influence on the behaviour of the main plant parameters such as coolant temperatures, reactivity and consequently, In Figure 6 it is shown the evolution of the coolant the power generated. temperatures in the three systems along the transient. These temperatures are; the sodium temperatures in the inlet and
9 of 96
outlet of the core, the sodium temperatures in the inlet and considered in the Point Kinetic calculation proposed , outlet of the secondary side of the IHX and the water namely, the Doppler effect, the coolant density effect, the temperatures in the inlet and outlet of the SGs. It can be expansion of fuel, cladding and diagrid, and the control rod highlighted the strong increase of the core outlet partial-insertion due to its relative movement with the temperature due to the abrupt reduction of the primary structure. massflow during the firsts seconds of the transient. After these initial seconds, the temperature goes down smoothly due to the decrease of the core power shown in Figure 4.
Fig. 8. Total Reactivity
The Figure 9 shows the evolution of all these reactivity Fig. 6. Coolant temperatures. effects separately. The positive reactivity that can be observed in Figure 5 during the very first seconds (6.5s) is Figure 7 shows the evolution of the temperatures of caused by the coolant density effect. Due to the remarkable fuel and cladding. Namely, the peak and mean temperature increase of the temperature of the coolant through the core of the fuel and the peak cladding temperature of the hot in the first seconds, as it was shown in Figure 5, its density assembly. These temperatures are much related with the decreases, decreasing its neutronic absorption rate that gap size model considered. As mentioned before, RELAP5 causes the positive reactivity. and TRACE have both a constant gap size model. Nevertheless, after these initial seconds the reactivity turns negative due to the influence of the Doppler effect caused by the initial increase of the fuel temperature and, mostly, the negative reactivity caused by the neutronic absorption due to the partial insertion of the control rods.
Fig. 7. Fuel and Cladding temperatures.
In Figure 8, the total reactivity feedback that commands the reactor core power is shown. It should be highlighted that the transient is unprotected, which means that no scram is activated during this transient. So, in this picture shows the spontaneous evolution of the core Fig. 9. Reactivity Feedbacks. reactivity as the sum of the effects of different phenomena
10 of 96
IV. CONCLUSIONS 2. FIORINI, G.L., VASILE A., European Commission - 7th Framework Programme, “The Collaborative The results of both RELAP5 and TRACE code Project on European Sodium Fast Reactor (CP- calculations for the key parameters of the benchmark ESFR)”. Nuclear Engineering and Design (2011). transient are shown in figures 3 to 8.The overall analysis indicates a good agreement between the results provided by 3. AMMIRABILE, L., TSIGE-TAMIRAT, H. the two codes. This is due to the strong similarity in the “Development of the European Safety Analysis model chosen in both TRACE and RELAP5 to represent Platform (ESAP) for Integrated Core and Safety the ESFR plant. This is particularly comforting when Analysis of Nuclear Reactor Systems”. Proc. Of the considering the evolution of the reactivity feedbacks in the International Congress on Advances in Nuclear Power core. (ICAPP), 2011, Nice, France.
The one-dimensional model has been of great 4. Mikityuk K. “Heat transfer to liquid metal: Review of importance to: data and correlations for tube bundles”. Nuclear Enginerring and Design 239 (2009) 680-687. Prove that RELAP5 and TRACE are thermohydraulic codes able to simulate transient 5. RELAP5/MOD3 Code Manual Vol2. NUREG/CR- behaviours in a SFR technology plant with minor 5335-Vol II. Office for Nuclear Regulatory Research. modifications in its source code. Washington. The results of both codes have an excellent agreement in general terms. 6. NRC, 2007. TRACE v5.0 Theory and User’s manual. The simulations done with the model can be used as Office for Nuclear Regulatory Research. Washington. a fair approximation of the value of safety parameters and grace times in the analyses of the 7. Dan Gabriel Caccuci (ed), Handbook of Nuclear behaviour of the plant in transients that may Engineering, Chapter 21, Springer (2010). compromise the plant safety. It has been a first step in the development of a three- 8. Wagner, W., Kruse, A. “The industrial standard dimensional model able to analysis more complex IAPWS-97 for the thermodynamic properties and transients that will lead to the implementation of a supplementary equations for other applications”. full-scope platform with capabilities to perform a Properties of Water and Steam. Springer. Heidelberg. wide range of dynamic analysis in SFR technology Germany. plants. 9. Y. Phillipponneau. “Thermal conductivity of (u, NOMENCLATURE Pu)O2-x mixed oxide fuel.”. Journal of nuclear materials. Vol. 188, p.p. 194-197, 1992. ESAP - European Safety Analysis Platform. CP-ESFR - Competitive Project on the European Fast 10. SNETP (Sustainable Nuclear Energy Technology Reactor. Platform), 2009. Strategic Research Agenda. NRC - Nuclear Regulatory Commission. http://www.snetp.eu/www/snetp/images/stories/Docs- IHX - Intermediate Heat Exchanger. AboutSNETP/sra2009.pdf SG - Steam Generator. BOL - Beginning of life. ACS - Above Core Structure. DSD - Diverse Shutdown Devices. CSD - Control and Shutdown Device.
REFERENCES
1. Gen IV Roadmap, 2002 - US DOE Nuclear Energy Research Advisory Committee and the Generation IV International Forum, A technology Roadmap for Generation IV Nuclear Energy Systems, GIF002-00, December 2002.
11 of 96
GENERATION IV TECHNOLOGY STATUS INCLUDING RECENT R&D ACTIVITIES IN ANSWERS
J. N. Lillington, P.N. Smith, T.D. Newton, R Thetford AMEC Technical Services Kimmeridge House, Dorset Technology Park Winfrith
G. S. Anderson AMEC Technical Services Thomson House, Birchwood Park Warrington
ABSTRACT
One objective of the paper is to review the current status of three of the Gen IV designs, which include high temperature gas cooled, liquid sodium and molten salt systems and their stage on the journey towards commercial exploitation. The first two of these designs represent some degree of evolution from plants that are or have been in operation, in prototype or have received some demonstration of principle. However all will also require additional R&D before they can be realised commercially to their full Gen IV potential. Collectively the designs could meet all or most of the demand criteria outlined in a companion paper presented to this conference. A second objective of the paper is to include a commentary on some of the key R&D requirements for these various designs, with particular reference to those pertinent to the reactor, fuel and fuel cycle.
The ANSWERS codes developed by AMEC have radiation transport methodologies for evaluating fuel response, core design and structural component radiation dose. The codes encompass both state-of-the-art deterministic and Monte Carlo methods. They apply for normal core operation and for fault studies and are validated for a wide range of current generation and advanced reactors. The third objective of this paper is to include a discussion of ANSWERS codes developments that are pertinent to the Gen IV designs.
1 Introduction
This paper addresses the Gen IV designs that may be expected to meet various energy and nuclear market demands looking forward towards 2050. In addition to the need for high- efficiency electricity generation and co-generation, these include the aspiration for a more sustainable fuel cycle, improved management of high-level waste and spent fuel, possibly including recycling [1]. Other demands are for the management of stockpiled plutonium from earlier weapons programmes and the consideration of a thorium fuel cycle.
The Very High Temperature gas-cooled Reactor (VHTR) offers best prospects for high- efficiency electricity generation, including high-temperature process heat and cogeneration. There is much experience of liquid Sodium cooled Fast Reactors (SFR) for electricity generation. The Gas cooled Fast Reactor (GFR) offers improved sustainability through combining the high-temperature process heat capability of the VHTR with the capability of a fast reactor to operate in a self-sustainable closed fuel cycle. The helium coolant of GFR is transparent and single-phase with excellent nuclear and chemical stability. These coolant properties permit simple in-service inspection and repair, and the lack of phase change and favourable neutronics practically eliminates energetic core disruptive accidents. All fast reactors have the capability to reduce both the volumes and radiotoxicity of wastes through recycling and transmutation of minor actinides. The flexibility of the fast reactor concept permits the management of plutonium stockpiles through their ability to be net burners, breeders or “iso-generators” of plutonium. However, since the ANSWERS R & D
12 of 96
developments for VHTR and SFR apply also to GFR, this system per se will not be addressed in detail in this paper. Finally the Molten Salt Reactor (MSR) is the least studied of the Gen IV candidate designs but offers the best conversion ratios among the Gen IV designs, the potential for thorium utilisation and other fuel management options.
ANSWERS has contributed to the UK’s Advanced Reactor R&D Programme since the 1980s [2]. This programme initially had a strong fast reactor focus. It provided key inputs in the areas of core neutronics, nuclear data and fuel modelling and behaviour through the CAPRA programme on plutonium burning and the CAPRA/CADRA programme on plutonium management and minor actinide burning. This programme evolved to cover work relevant to all three of the systems of interest to the UK (VHTR, SFR and GFR). The ANSWERS team contributed to EC programmes of relevance to Gen IV reactors, such as CONFIRM, FUTURE, RAPHAEL, AFTRA and PuMA, and participated in OECD/NEA international benchmarks through the OECD working groups IRPhE, JEFF and ICSBEP as well as IAEA Co-ordinated Research Programmes and international benchmarks.
2 High Temperature Reactors
Current Status
The VHTR design is based on a helium-cooled, graphite-moderated thermal neutron flux reactor with target outlet temperatures as high as 1000 0C. It offers the possibility of high- efficiency electricity generation as well as high-temperature process applications and hydrogen production.
HTR designs were first mooted in the 60s and there were prototype designs including Dragon at Winfrith, AVR and THTR in Germany and Peach Bottom and Fort Saint Vrain in the US. There are two fundamental designs in HTR technology; the prismatic core design and the pebble-bed design. The prismatic design was adopted in Japan (HTTR), in France (ANTARES) and in the NGNP programme in the US. For pebble-bed designs, much work was carried out in the South African Pebble-Bed Modular Reactor (PBMR) programme but this was halted in 2010. Only the HTR-10 pebble bed reactor (10MWt) in China is operational although the larger HTR-PM (2*250MWt units) reactor is also under construction.
R&D Requirements
The main areas for R&D on HTRs are centred on fuel and fuel cycle, the robustness of component materials, the thermal conversion cycle, BOP and the process heat and hydrogen production technologies for other applications.
The basic HTR (TRISO) fuel is based on UO2 particles coated with four layers: a buffer of porous carbon to accommodate fission products, a dense inner layer of pyrolytic carbon, a strong and impervious ceramic layer of silicon carbide and a dense outer layer of pyrolytic carbon. The design objectives are good structural integrity and fission product retention for high burn-up at high temperature. An alternative fuel kernel of UCO with a zirconium carbide layer (in place of silicon carbide) has also been proposed. Thorium and plutonium-based fuel could be utilised within a closed fuel cycle, but at present fuel recycling is not envisaged. R&D is required to confirm these design requirements. HTR coated particle fuel can be utilised for Pu and Minor Actinide (MA) management e.g. in a GFR.
ANSWERS Developments
ANSWERS codes have been developed to model advanced HTR fuel; in particular, reactor physics modelling capabilities for the PBMR using both Monte Carlo and deterministic methods [3], [4].
13 of 96
The ANSWERS Monte Carlo code MONK can model randomly-placed particles within a pebble and the pebbles within the core using algorithms developed to give the required core packing fraction without modelling partial pebbles or opening artificial streaming paths in the core that would result from simply stretching a close-packed hexagonal array. Each pebble in the core is ascribed an individual burn up to allow a full modelling of the whole core.
The ANSWERS deterministic WIMS modular code package allows a triple heterogeneity modelling of resonance self shielding. Sub-group methods are used to treat the heterogeneity of the fuel particles, the particles within the fuel pebbles and the interaction between pebbles of different burn up or composition. The WIMS code allows the modelling of the fuel flow through the core for both once-through and multi-pass cycles. In addition, the temperature distribution in the core can be calculated in pseudo-steady-state conditions and the resulting temperatures fed back to the modules performing the resonance self-shielding.
Computational Fluid Dynamics (CFD) methods have been developed for both normal and accident conditions in PBMR. Developments extended to modelling the fuel pebbles, coolant, and reactor structure. Coupled CFD and ANSWERS WIMS code capabilities have also been developed for HTR, particularly for PBMR and to a lesser extent for a prismatic HTR in one of the EC research programmes [5]. For GFR, the methodology was based on ERANOS.
Fuel performance modelling capability exists for PBMR, using the STRESS3 and STAPLE codes. STRESS3 models the behaviour of UO2 fuel particles in a high-temperature reactor, up to the failure of the coatings. It includes simple models of fission gas release (via the Booth model) and thermo-mechanical behaviour. STAPLE is a statistical wrapper that performs a Monte Carlo series of STRESS3 calculations to estimate particle failure probabilities. The TRAFIC code (discussed later) has detailed mechanistic modelling of fission gas, temperatures, stresses and strains in a fuel pin. Presently TRAFIC is limited to a cylindrical geometry but the models could be applied in spherical geometry and VHTR fuel.
3 Fast Reactors
Current Status
The Gen IV SFR is characterised by liquid sodium coolant at relatively low pressure but with high core power density. With a closed fuel cycle it would offer good fuel sustainability and allow management of waste, plutonium and other actinides. There are three reference system designs: a large loop-type, a medium pool-type with MOX fuel and an aqueous processing fuel cycle, and a smaller modular design incorporating U-Pu-MA metal alloy fuel with pyrometallurgical processing.
The SFR concept has been established in various countries for many decades but for the primary purpose of electricity generation. Large-scale prototypes have successfully operated in the UK (DFR, PFR), US (EBR I, Enrico Fermi, EBR II, FFTF), France ( Rapsodie, Phénix, Superphénix), Russia (BR5, BOR 60, BN 350, BN600), Japan (Joyo, Monju) and Germany (KNK II). There are fast reactors currently under construction in Russia (BN800). In France, the ASTRID programme aims to deploy a Gen IV SFR demonstrator by the early 2020s.
R&D Requirements
For the development of a Gen IV system, the main R&D requirements concern the development of advanced fuels and MA-bearing fuels, systems integration, component design and materials and safety issues associated with sodium chemical and neutronic reactivity. Possible fuels include mixed oxide, nitrides and carbides and, particularly for small reactors, metal alloys. The R&D requirements on the fuel are largely focussed on the performance of oxide fuel including cladding, oxide strengthened steel and metal alloy fuel and ferritic-martensitic stainless steel cladding. Fuels containing MAs and fission products
14 of 96
tend to have a low thermal conductivity; research work is examining the consequences of embedding MA fuel particles in an inert matrix that has higher conductivity. Fuel performance codes are being developed in parallel with experimental work. Another area for R&D concerns the various fuel cycle options associated with the above processes, the recycling of highly radioactive fuel and avoidance of the separation of pure plutonium.
ANSWERS Developments
The TRAFIC code [6] includes the major underlying mechanisms and models for fast reactor fuel performance in normal and accident conditions. Stresses and displacements of fuel and cladding depend on elastic deformation, creep, plastic yield, solid swelling, friction between fuel and cladding and thermal expansion. Temperatures are determined by accounting for heat flows associated with heat generation by fission, conduction in fuel and clad, conduction across the fuel-clad gap and heat transfer to the coolant. TRAFIC calculates the distribution of fission gas and helium between the fuel, the plenum and any inert matrix. Models of grain growth, clad corrosion, solid-state thermo-migration and chemical equilibrium and kinetics describe the physical and chemical changes in the fuel as irradiation proceeds.
The code has been developed and validated via fuel-related research programmes. Most important are the experimental test programmes such as CABRI and PFR/TREAT, but data have also been taken from other experimental programmes such as routine post-irradiation examination, out-of-pile annealing studies of fission gas, void swelling under irradiation in neutron and electron beam facilities ([ISIS, Diamond]. Theory and modelling work (e.g. atomistic modelling of metals and UO2) have also contributed. Research has investigated other issues relevant for SFRs (and for other advanced reactors), such as the radiochemistry of reprocessing actinides and vitrifying high-level waste, and geological disposal of wastes.
4 Molten Salt Reactors
Current Status
MSRs have some unique features among the other Gen IV designs, enabling the optimum use of fuel resources, high burn-up of the fuel and reduced production of high level waste. There is considerable fuel cycle flexibility. They can be used for breeding in both thermal and fast neutron spectra modes. They can be used for actinide burning in a fast spectrum. The unique feature is that fuel is dissolved in a liquid salt coolant, typically a mixture of lithium / beryllium fluoride. Lithium / sodium / potassium and sodium / zirconium fluorides are other possible salts. The MSR is amenable to U-Pu and thorium fuel cycles; the latter offers particular attractions for breeding.
In the 1950s and 1960s, two thermal spectrum reactors were built and operated at ORNL, US (the Aircraft Reactor Experiment and the Molten Salt Reactor Experiment (MSRE)) and there were also significant programmes in Russia starting in the 1970s. Early work of molten salt fast reactors was done in the UK. No fast spectrum reactors have been built though there has also been early work on conceptual fast spectrum designs in the US (ORNL, ANL) and in Switzerland (Institute of Energy Research). Recent work has been mainly in China that is launching a programme to develop a Thorium Molten Salt Reactor (TMSR), also referred to as the Liquid Fluoride Thorium Reactor (LFTR).
R&D Requirements
The R&D requirements are concerned with molten fuel and salt technology, the fuel cycle, the chemistry and other properties of liquid salt, and safety. There is one fewer layer of defence in depth than for the other fuel designs in that there is no cladding. R&D is focussed on the fuel and coolant performance in the reactor, including gaseous fission product extraction and fuel reprocessing.
15 of 96
ANSWERS Developments
The ANSWERS codes have been developed to allow modelling of general geometries and material compositions. Recent developments include optimization of the physics modelling for thorium fuel applications. This has included modelling of a Th232/U233 thermal breeder core in the Shippingport facility [7] using the deterministic WIMS code and the Monte Carlo MONK code [8]. The study of multiplication factor and depletion included comparison of the thorium data in the latest nuclear data evaluations (including ENDF/B-VII, JEFF3.1 and JEF2.2), comparison of continuous energy and broad group representations of nuclear data, and comparison of deterministic and Monte Carlo methods.
The CRITEX code was developed jointly with CEA for criticality excursion analysis of fissile solutions. The code has been extensively validated against the French CRAC and SILENE fissile solution experiments [9] and the Japanese TRACY experiments [10]. With suitable modifications the code could be used for transient analysis of accident conditions in MSRs.
5 Conclusions
The SFR concept has received the most attention and has the most operational experience, followed by the HTR. The MSR concept is the least mature. Global investment in the Gen IV evolutions (VHTR, SFR and MSR) of all these designs is confined to only a few countries.
The VHTR, SFR and MSR designs all exhibit some combination of higher temperature and a need for better material corrosion resistance. They require substantial R&D in order to reach Gen IV design goals. This is particularly so for the fuel and fuel cycle.
The ANSWERS codes capabilities are being extended for modelling fuel and core design for the above reactor concepts. This paper gives some examples; the work continues.
6 References
[1] Anderson, G S, Lillington J N , Market Demands and Nuclear Power in the United Kingdom in the Early Years of the Twenty First Century, ENC 2012, 9-12 December 2012. [2] Smith, P N, Lillington J N, Middlemas C, Radiation Transport Modelling and the ANSWERS Codes Suite, Nuclear Future, Vol 7, Issue 2, March/ April 2011. [3] Newton, T D, The Development of Modern Design and Reference Core Neutronics Methods for the PBMR, ENC Conference, Lille, France, October 2002. [4] Franklin, B M and Newton, T D, PROTEUS PBMR Modelling Using MONK, ICNC Third International Conference on Natural Computation, St Petersburg, Russia, May 2007. [5] Stainsby, R, Sunderland, R E, UK Activities in support of Gas Reactor Development, OECD/DOE/CEA PBMR Coupled Neutronics/ Thermal Hydraulics Transient Benchmark. [6] Mignanelli, M A, Thetford, R, Chemical Modelling with the TRAFIC code, Transactions of the American Nuclear Society; (United States); Journal Volume: 66; ANS/ ENS international meeting on fifty years of controlled nuclear chain reaction: past, present, and future, Chicago, IL (United States), 15-20 Nov 1992. [7] Atherton R, “Water Cooled Breeder Program Summary Report (LWBR Development Program)”, WAPD-TM-1600, October 1987. [8] Harrington C, “Reactor Physics Modelling of the Shippingport Light Water Breeder Reactor”, MPhil Dissertation, Cambridge University, August 2012. [9] Mather D J, Bickley A M and Prescott A, “CRITEX: A Code to calculate the Fission Release Arising from Transient Criticality in Fissile Solutions”, UKAEA Report, R 2385, 1994. [10] Miyoshi Y, et al, “Inter-Code Comparison Exercise for Criticality Excursion Analysis, Benchmarks Phase - 1: Pulse Mode Experiments with Uranyl Nitrate Solution in the TRACY and SILENE Facilities”, OECD/NEA/NSC Report 2008.
16 of 96 PRELIMINARY DESIGN ASSESSMENT OF THE MOLTEN SALT FAST REACTOR E. MERLE-LUCOTTE, D. HEUER, M. ALLIBERT, M. BROVCHENKO, V. GHETTA, A. LAUREAU, P. RUBIOLO LPSC-IN2P3-CNRS / UJF / Grenoble INP, 53 rue des Martyrs, F-38026 Grenoble Cedex, France
ABSTRACT In the frame of developing future energy resources and reducing nuclear wastes, the molten salt reactor concept offers very good potential. Molten salt reactors are liquid fuel reactors so that they are flexible in operation but very different in the design and safety approach compared to solid-fuelled reactors. This paper will address design issues of the MSFR, detailing some technological choices for the system components (fuel salt composition and distribution, core geometry, fuel heat exchangers…).
1. Introduction Starting from the Oak-Ridge Molten Salt Breeder Reactor prototype, parametric studies were performed, focusing on breeding capabilities, reprocessing requirements, safety issues and more recently on nuclear power plant design, resulting in an innovative breeder concept: the Molten Salt Fast Reactor or MSFR [1-6]. The MSFR, with a fast neutron spectrum and operated in the Thorium fuel cycle, may be started with 233U, enriched U, and/or TRU elements as initial fissile load. This concept has been recognized as a long term alternative to solid-fuelled fast neutron systems with a unique potential (large negative temperature and void coefficients, lower fissile inventory, no initial criticality reserve, simplified fuel cycle, wastes reduction…) and is thus one of the reference reactors of the Generation IV International Forum [7].
Fig.1: Schematic conceptual MSFR design The reference MSFR is a 3000 MWth reactor with a total fuel salt volume of 18 m3, operated at a mean fuel temperature of 750°C. Figure 1 sketches the general component outlines for such a MSFR. The core consists of a circulating fluoride salt loaded with the fuel (note the absence of solid matter in core). The fuel salt considered in the simulations is a binary fluoride salt with 77.5% of lithium fluoride; the other 22.5% are a mix of heavy nuclei fluorides. This proportion, set throughout the reactor evolution, leads to a fast neutron spectrum. The total fuel salt volume is distributed half in the core and half in the external fuel circuit This MSFR system thus combines the generic assets of fast neutron reactors (extended resource utilization, waste minimization) with those associated to a liquid-fuelled reactor. In preliminary designs developed in relation to calculations, the core of the MSFR is a single compact cylinder (2.25m high x 2.25m diameter) where the nuclear reactions occur within the liquid fluoride salt acting both as fuel and as coolant. The external core structures and the fuel heat exchangers are protected by thick reflectors made of nickel-based alloys, which have been designed to absorb more than 99% of the escaping neutron flux. These reflectors
17 of 96 are themselves surrounded by a 20cm thick layer of B4C, which provides protection from the remaining neutrons. The radial reflector includes a fertile blanket (50 cm thick - red area in Fig. 1) to increase the breeding ratio. This blanket is filled with a fertile salt of LiF-ThF4 with 232 initially 22.5mole % ThF4. The return circulation of the salt (from the top to the bottom) is divided into 16 groups of pumps and heat exchangers located around the core [8]. The neutronic reflectors, made of NiCrW-based alloy, constitute the lower and upper walls of the core. The lower reflector is connected to a draining system: in case of a planned shut down or incident/accident leading to a temperature increase in the core, the fuel configuration may be changed passively by gravitational evacuation of the fuel salt in tanks located under the reactor where a passive cooling will be achieved. Conceptual design activities are currently underway so as to increase the confidence that MSFR systems can satisfy the goals of Generation-IV reactors in terms of sustainability (Th breeder), non proliferation (integrated fuel cycle, multi-recycling of actinides), resource savings (closed Th/U fuel cycle, no uranium enrichment), safety (no reactivity reserve, strongly negative feedback coefficient) and waste management (actinide burner). Two of these studies related to the fuel salt are detailed in this paper. The selection of the liquid fuel composition is presented in the second section, based on neutronics, materials and chemical considerations. The third section presents the global method developed to assess the design of the heat exchangers while taking into account the requirements of the entire fuel circuit, since the fuel salt is also used as the coolant in such reactors. One of the main constraints on the design of the fuel circuit of the MSFR is indeed the ability to evacuate the heat generated while restraining the fuel salt volume mobilized out of the core for this task.
2. Which liquid fuel? The use of a liquid fuel has significant potential benefits: • The homogeneity of the fuel allows uniform combustion, thereby avoiding loading plans • Fuel management involves only fluid transfers • Reprocessing and fuel preparation require no change of state • In an emergency, the fuel can be transferred quickly by gravitational flow to vessels designed to evacuate the residual power passively • Fuel reprocessing can be done online or in batch mode on discrete samples and therefore without requiring reactor shutdown. Liquid-fuelled reactors are « homogeneous reactors » that have intrinsic safety properties thanks to the fuel's expansion coefficient that induces large negative thermal and void feedback coefficients. Because of this, such reactors can be controlled without control or command rods. The choice of liquid fuel is guided by operational considerations, but also the need to meet the GEN IV recommandations. The main criteria are: • A melting temperature not too high and a sufficiently high boiling point • Low vapour pressure • Good thermal and hydraulic properties • Good stability under irradiation • Sufficient solubility of fissile and fertile elements • Avoid the production of unmanageable radioisotopes • A high neutron transparency • An identified fuel reprocessing method. Taking into account all these constraints, the choice is reduced to two possible types of liquid: a fuoride or a chloride salt. We thus compare in this paper the characteristics of a MSFR operated in the Thorium fuel cycle while using fluoride or chloride salts. Two fuel salts have been considered in our reactor simulations: LiF-(HN)F4 with 22.5% heavy nuclei (HN), and NaCl-(HN)Cl4 with 28% heavy nuclei. Both salt compositions correspond to eutectic points of their respective phase diagram, with a melting temperature respectively of 565°C (fluoride
18 of 96 salt) and 375°C (chloride salt). More precisely, we have considered a fluoride salt enriched in 7Li (99.999 % of 7Li and 0.001% of 6Li) and a chloride salt enriched in 37Cl (99% of 37Cl and 1% of 35Cl), at the feasibility limit in both cases. Considering only the chemical characteristics of the salts is not discriminatory, each salt having its own drawbacks / advantages. For example, the reference technique to extract U and Pu is fluorination, without a corresponding process for a chloride mixture. The actinide solubility is higher in chloride than in fluoride salts, which could be a limiting factor to start a MSFR with actinide elements. On the other hand, the boiling temperature of chloride salts is 300°C lower than that of the fluoride salts (respectively around 1400°C and 1700°C), which would have to be taken into account in safety studies for transient analyses. We now compare chloride and fluoride salts according to neutronics considerations.
2.1 Breeding capabilities and irradiation damages The neutronic characteristics of a MSFR, based on the Thorium fuel cycle, and using a chloride / fluoride fuel salt, are listed in Tab 1. These results have been obtained via full numerical simulations of each system.
Parameter Fluoride Salt Chloride Salt Th Thorium capture cross-section σ C in core (barn) 0.61 0.315 Thorium amount in core (kg) 42 340 47 160 Thorium capture rate in core (mole/day) 11.03 8.48 Th Thorium capture cross-section σ C in blanket (barn) 0.91 0.48 Thorium amount in the blanket (kg) 25 930 36 400 Thorium capture rate in the blanket (mole/day) 1.37 2.86 233U initial inventory (kg) 5720 6867 Neutrons per fission ν in core 2.50 2.51 233 233U U capture cross-section σ C in core (barn) 0.495 0.273 233 233U U fission cross-section σ f in core (barn) 4.17 2.76 Capture/fission ratio α (spectrum-dependent) 0.119 0.099 Total breeding ratio 1.126 1.040 Tab. 1: Neutronic characteristics of the MSFR
Fig 2. Neutron Spectra for a chloride (purple curve) and a fluoride (green curve) salt in a MSFR
19 of 96 The salt density is equal to 4.1 for the fluoride salt and to 3.2 for the chloride salt. The chloride salt is consequently more transparent to neutrons, and, with an identical reprocessing (typically some ten litres of fuel salt per day), breeding is obtained only with a larger chloride salt volume of around 40 m3 in the core and 20 m3 in the fertile blanket, instead of respectively around 20 m3 and 8 m3 for a fluoride salt. This transparency to neutrons is probably due not only to the lower density of the chloride salt, but also to the absence of inelastic scattering on Cl as compared to F. The neutron spectrum is thus faster in the chloride salt (cf. Fig. 2) and neutrons have a smaller probability of interaction. The mean capture cross-section on Thorium is thus equal to 0.61 barn in the fluoride salt and to 0.315 barn only in the chloride salt. As the amounts of Thorium in each case are quite similar, the capture rate on Thorium in the fluoride salt is more important (11.03 mole/day) than that in the chloride salt (8.48 mole/day). We can easily evaluate the ratio of the breeding rates in the two systems, assuming that all the fissions occur on 233U and that the 233Pa does not capture. 233 τν U f == The reactivity is then equal to: k All 1 τ Abs Y with τ X the reaction rate X on the nucleus Y and ν the number of neutrons produced per fission. The breeding ratio is then equal to: 233 τ Th τν Th σ U R = c = c , α being the ratio c . 233U + τα All 233U ()1+ τα f ()1 Abs σ f Noting that the total absorption rate is identical in both cases since the total powers are equal, we deduce the ratio of the breeding rates in MSFRs based on fluoride and chloride salts: R ν ()1+α ()τ Th Cl = Cl F c Cl Th = 0.935 RF ν F ()1+α Cl ()τ c F 04.1 This ratio calculated with a full simulation of both systems is equal to = 924.0 126.1 As a conclusion, the breeding ratio of a MSFR operated with a chloride salt is clearly degraded as compared to that of a MSFR operated with a fluoride salt, in spite of the fact that the chloride salt volume considered is twice that of the fluoride salt.
Finally, the radiation damages in neutron-irradiated materials, dependent on many factors like the irradiation dose and the neutron spectrum, and expressed in dpa (displacements per atom), is directly impacted by the choice of the salt. Our calculations show that, in the most irradiated area corresponding to the first two centimetres of the central area (radius 20 cm and thickness 2 cm) of the axial reflector, the damages are 4 times higher (30 dpa/year against 7.5 dpa/year) for the chloride salt compared to the fluoride salt. This is due to the chloride salt neutron spectrum which is faster than that of the fluoride salt (see Fig. 2).
2.2 Production of problematic elements The presence of Cl in the salt leads to the production of 36Cl, whose radioactive period is 301 000 years. This element is very mobile, it is thus impossible to confine it over such large periods. 36Cl is produced through two production modes: 35Cl(n,γ)36Cl and 37Cl(n,2n)36Cl. The first mode is far more probable, requiring a salt enriched in 37Cl. Although we have chosen a significant enrichment of 99%, the first mode is still dominant with a production of 36Cl of around 10 moles per year (360 grams/year, with a total production of 373 grams/year). This production can be compared with the production of Tritium in the case of the Lithium fluoride salts, which amounts to 55 moles/year (166 grams/year). However the Tritium, also mobile, has a radioactive period of 12 years only, being thus truly less problematic. This 36Cl production represents one of the major drawbacks of chloride salts. Finally the presence of Cl in the salt also leads to the production of Sulphur, mainly through the reactions 37Cl(n,α)34P(β-[12.34s])34S and 35Cl(n,α)32P(β-[14.262 days])32S. Even with the
20 of 96 large enrichment of 99% in 37Cl, these reactions produce very large amounts of Sulphur in the salt (around 10 moles/year). This sulphur production has to be compared with the production of Oxygen in the fluoride salt, via the reaction 19F(n,α)16O, which amounts to 88.6 moles/year. In both cases, the element produced is very corrosive. But, while the Oxygen corrosion only affects the surface of metals, Sulphur weakens metals by placing itself on the grain boundaries, being thus much more corrosive. However, as both Oxygen and Sulphur will form compounds with some fission products, the proportion of these elements contributing to the corrosion of the structural materials is not really known. The Sulphur production has also to be compared to the production of Tellurium, which amounts to 200 moles/year in both fluoride and chloride salts. The corrosion mechanisms due to Tellurium and Sulphur are similar, so that the Sulphur production, which is significantly smaller than that of Tellurium does not represent a major drawback of chloride salts.
3. Conceptual design of the heat exchangers One of the main constraints on the design of the fuel circuit of the MSFR is the ability to evacuate the heat generated while restraining the fuel salt volume mobilized for that task. Upon exiting the core, the fuel salt travels through a liquid-gas separator, a pump, a heat exchanger and returns to the core's bottom inlet. The circuit must bypass the fertile blanket and the neutron protections while taking the shortest route so as to minimize the fuel salt volume within these components. Since here the fuel salt also plays the role of the coolant, the dimensioning of the heat exchangers is constrained by the requirement that the heat evacuation is to mobilize the minimal fuel salt volume. Examples of such dimensioning solutions are given in this section to illustrate the problem.
3.1. Characterization of the heat exchanges Suppose a plate heat exchanger made of Hastelloy and where the fuel salt and the intermediate fluid circulate in opposite directions on either side of a set of plates, as illustrated in Fig 3.
Fig 3. Schematic longitudinal view of the heat exchangers considered in this modeling
The heat exchange involves 2 thermal exchange coefficients (hc for the fuel salt, hi for the intermediate fluid) and the thermal resistance of the plate Rp, defined by: λ Nu 1 λ h ,ic = and = D p eR
21 of 96 where λ is the fluid's or plate's thermal conductivity, D the hydraulic diameter, e the thickness of the plate or the equivalent static thickness of the fluid, Nu the Nusselt Colburn dimensionless number defined, via the Reynolds (Re) and the Prandtl (Pr) numbers, by the following equations (for Pr > 0.5): 4 1 5 3 Re = DV µρ Pr = C λµ and D = 4s/p Nu = Re023.0 Pr p With: ρ the mass density, V the flow speed, μ the dynamic viscosity, Cp the heat capacity, s the flow section, and p the perimeter of this flow section. For Pr << 1 (liquid metals), the Nu number is calculated by: m = bNu Re. with b and m depending upon the Pr number as listed in [9].
The overall heat exchange coefficient is then obtained as: 1 1 1 = + R + h h 푝 Given the power P to be extracted and the mean푐 temperatureℎ푖 difference ΔT between the two heat transfer fluids, the necessary exchange surface is calculates as follows: S = P/h.ΔT The salt volume that can be mobilized in the heat exchangers being fixed, the gap between the plates (also called thickness of the fuel salt channel in the following) is predetermined so that the hydraulic diameter can be calculated recursively so as to re-determine the overall heat exchange coefficient h. It is then possible to derive the pressure drop in the exchangers as well as in the pipes that convey the fuel salt from the core to the exchangers and from the exchangers back to the core using the following relationships:
1 2 1 Λ=∆ ρ DVLP with 2 −=Λ Log + ()ε ()7.3Re9.68.1 D 11.1 2 [ ] With L the pipe’s length and Λ a pressure drop coefficient calculated using the Colebrook equation approximated by Haaland for the turbulent case, the roughness of the pipe surface ε is taken equal to 10-5. The singular pressure drops in the pipes due to the bends have been added, equal to ½κρV2 per bend where κ is the pressure drop coefficient equal to 0.35 for a bend of 90°.
3.2 Physicochemical properties of the fluids 7 233 The initial fuel salt is composed of LiF-ThF4(20 mole%)- UF3(2.5 mole%) with 77.5 mole % of LiF, this fraction being kept constant during reactor operation. The fraction of 233U is adjusted initially to have an exactly critical reactor. During reactor operation, fission products and new heavy nuclei are produced in the salt up to some mole % only, they do not impact the salt physicochemical properties needed for our studies. We have used the characteristics of the initial fuel salt [10], as presented in Table 2. The melting temperature of the fuel salt is equal to 565°C. The linear increase of the calorific capacity is limited to 800°C.
Unit Formula A B Calorific capacity Cp J/K/kg A+BT -1111 2,78 Thermal Conductivity λ W/K/m A+BT 0.928 8.40E-05 Density ρ kg/m3 A+BT 4983.56 -0.882 Dynamic viscosity μ Pa.s ρ .A.exp(B/T) 5.55E-08 3689 Tab. 2: Physicochemical properties of the fuel salt as a function of its temperature T in K [10]
The physicochemical properties used for the three intermediate fluids considered here are summarized in Table 3. The exact composition of the fluid labelled « FLiNaK » is LiF (46.5 mole%) - NaF(11.5 mole%)- KF (42 mole%). The composition of the salt labelled « NaF – NaBF4 » is : NaF (8mole%) - NaBF4 (92 mole%). For the liquid lead, the maximal temperature at the surface, in contact with the structural materials, has been limited to 530°C to limit the corrosion rate, if it takes place.
22 of 96
Unit Liquid lead FLiNaK NaF-NaBF4 175.1 - 4,96.10-2.T Calorific + 1,99.10-5.T2 – 976.199 + C J/K/kg -9 3 1506 capacity p 2.10.10 .T – 1.0624.T 1.52.106.T-2 Thermal -4 0.66 - λ W/K/m 9.2 + 0.011.T 0.36 + 5.60.10 .T -4 Conductivity 2.37.10 .T 3 2446.3 - Density ρ kg/m 11367-1.1944.T 2579.3 - 0.624.T 0.711.T Dynamic 4.55.10-4. 2.49.10-5 . 8.77.10-5 . μ Pa.s viscosity exp(1069/T) exp(4476.23/T) exp(2240/T) Melting T °C 327 454 384 temperature m Tab. 3: Physicochemical properties of the intermediate fluids [11,12,13] (with T the temperature of the fluid in K)
3.3 Analytical method and typical solutions A global geometry can be found by setting some parameters and constraints. A configuration that best satisfies these constraints can then be sought by adjusting a list of variable design parameters. The preset parameters, evaluated through previous neutronic studies of the reactor core are: - the total power is set at 3GWth - the core diameter is equal to the core height - there are 16 identical sectors comprising a liquid-gas separator, a pump and a heat exchanger as well as any joining pipes - the fuel salt volume is 18 m3, 50 % of it in core, 5 % in auxiliary volumes (overflow tank, spaces, etc.) and 45 % in the liquid-gas separators, the pumps and pipes, the heat exchangers - the fertile blanket thickness is 500 mm and the neutron protection thickness is 200 mm
Other secondary parameters are prefixed as well, to simulate the liquid-gas separators, the heat exchanger inputs-outputs, the isolated pressure losses (bends, liquid-gas separators), and the pipes conveying the intermediate fluid to the exchangers. This analytical method has been applied for three intermediate fluids: liquid lead and two salts with a melting point lower than that of the fuel salt (LiF-NaF-KF and NaF-NaBF4). The variable parameters of the present studies are: - the diameter of the pipes - the thickness of the plates - the gap between the plates on the intermediate fluid side, also called “thickness of the intermediate fluid channel” in this paper - the fuel salt temperature at core entrance - the fuel salt temperature increase within the core - the temperature increase of the intermediate fluid in the heat exchangers - the mean temperature difference between the two fluids within the heat exchangers
Some parameters are constrained to ensure an acceptable mode of operation. The most important ones are listed in table 4 along with their limiting value and acceptable deviation.
Each set of values of the variable parameters is evaluated using the following quality function:
푃푖 − 푃0푖 � 푒푥푝 � 푖 � 푖 휎
23 of 96
th where Pi is the value of the i parameter, P0i the limiting value of the parameter, and σi the acceptable deviation for the parameter. The set of values of the variable parameters that minimizes this function is then sought. Typical results are shown in table 5 for the three intermediate fluids considered here.
Constrained Parameter Limiting value Acceptable (P0i) deviation (σi) Minimum thickness of the fuel salt channel 2.5 mm 0.05 mm Minimum thickness of the plate 1.75 mm 0.035 mm Maximum speed of the fuel salt 3.5 m/s 0.07 m/s Maximum speed of the intermediate fluid (liquid lead) 1.75 m/s 0.035 m/s Maximum speed of the intermediate fluid (salt) 5.5 m/s 0.11 m/s Maximum temperature of the materials 700 °C 1 °C Minimum margin to solidification of the fuel salt 50 °C 1 °C Minimum margin to solidification of the intermediate fluid 40 °C 1 °C Table 4: Main constrained parameters with their limiting value and acceptable deviation When liquid lead is used as intermediate fluid, we notice that the gap between the plates on the fuel side is quite large, that implies also voluminous connecting parts. This fact is partially linked to the speed limitation for the lead flow. This also results in a higher temperature at the fuel entrance. When a salt fluid such as FLiNaK or NaF-NaBF4 is used as intermediate fluid, the heat exchanger outlet temperature is higher which allows raising the thermodynamic yield. Nevertheless, the heat exchanger temperature drop is smaller in this case and high velocities are required. Nevertheless these quite high intermediate salt flow velocities seems not really acceptable because of the pumping performance and erosion problems. The 3 cases discussed here are only examples: changing the constraints to take into account technological data is possible and will produce other solutions.
Evaluated parameter Pb FLiNaK NaF-NaBF4 Diameter of the fuel salt pipes [mm] 301 283 303 Diameter of the intermediate fluid pipes [mm] 897 507 470 Thickness of the plates [mm] 1.61 1.51 1.65 Fuel salt temperature at core entrance [°C] 754 698 704 Fuel salt temperature increase in the core [°C] 89 106 98 Intermediate fluid temperature increase within the heat exchangers [°C] 99 41 66 Mean temperature difference between the two fluids in the heat exchangers [°C] 382 242 280 Intermediate fluid temperature at the heat exchangers outlet [°C] 466 530 506 Thickness of the fuel salt channel [mm] 3.38 2.17 2.37 Thickness of the intermediate fluid channel [mm] 29.8 4.49 4.38 Fuel salt speed in the pipes [m/s] 3.92 3.97 3.73 Fuel salt speed in the heat exchangers [m/s] 3.85 2.36 2.91 Intermediate fluid speed in the pipes [m/s] 1.94 6.00 5.67 Intermediate fluid speed in the heat exchangers [m/s] 1.92 5.54 5.75 Maximum temperature of the intermediate fluid [°C] 523 622 595 Maximum temperature of the materials [°C] 701 701 699 Margin to the solidification of the fuel salt [°C] 43.7 54.7 46.7 Margin to the solidification of the intermediate fluid [°C] 39.6 34.5 56.2 Pressure loss of the fuel salt in the heat exchangers [bar] 2.56 2.03 2.56 Pressure loss of the fuel salt in the pipes [bar] 0.99 1.02 0.90 Pressure loss of the intermediate fluid in the heat exchangers [bar] 0.09 2.09 1.66 Pressure loss of the intermediate fluid in the pipes [bar] 0.32 0.71 0.57 Table 5: Typical sets of parameters evaluated for the three intermediate fluids considered
24 of 96
An example of the temperature distributions in the heat exchangers can be plotted as in Figure 3, in the case of liquid lead. At the top of the heat exchangers (hot fuel salt inlet and heated intermediate fluid outlet), the maximum salt temperature is significantly higher than that of the plate (labeled “Tmax material” in Fig 3). The salt's poor thermal conductivity functions as a protection for the heat exchanger plates during normal operation. The same holds at the bottom of the heat exchangers (cooled fuel salt outlet and cold intermediate fluid inlet) but here, the fuel salt temperature (labeled “Tmin salt”) draws near to that of its solidification. Thus, it is, indeed, the lowest permissible fuel salt temperature that constrains the minimal plate surface temperature. On the intermediate fluid side, since lead becomes very corrosive beyond 500-550 °C, the maximum plate surface temperature has to be limited, thus lowering the output lead temperature (labeled “Tmax lead”). On the other hand, the lead solidification temperature constrains the lead input temperature (labeled “Tmin lead”).
Fig 3. Qualitative representation of the temperature distributions in a heat exchanger
Four constrained temperatures limit the temperature ranges. It is mandatory that these temperature limits be observed at all times whatever the reactor's operating mode except, possibly, during rare short duration transient states (incidental or accidental occurrences). Conclusions This paper describes two design studies of the Molten Salt Fast Reactor, related to the fuel circuit, namely the choice of the fuel salt composition and a global study of the heat exchangers base on an approximate method developed to take into account the requirements of the entire fuel circuit, since the fuel salt is also used as the coolant in such reactors. Concerning the choice of the liquid fuel, two types of salt have been studied: a fuoride or a chloride liquid salt. No discriminating difference between fluoride and chloride salts can be identified from a chemical standpoint: their characteristics are quite different but none is exclusionary. However, considering the neutronic standpoint, we demonstrate the real limited ability of the chloride salt to ensure breeding while used in a Thorium-based MSFR, together with the unavoidable production of significant amounts of the radiotoxic and unconfinable 36Cl and the irradiation damages in both cases. These studies demonstrate the definite advantage of using fluoride salts in a MSFR versus chloride salts.
25 of 96 Concerning the conceptual design of the fuel heat exchangers, the interdependence of all the reactor components requires a global analysis of the entire fuel circuit. An approximate method, which takes into account all the constraints presently known (physical, chemical, technological…) has been developed to ascertain whether solutions may exist for this multi- parameter problem. Examples of such solutions in the constraints phase have been shown here: these are only indicative but demonstrate that relevant configurations may be identified. These configurations are going to evolve according to new constraints as they appear. A more complete method, including more realistic models, will have to be developed to further assess the fuel circuit of such innovative reactors.
Acknowledgments The authors wish to thank the PACEN (Programme sur l’Aval du Cycle et l’Energie Nucléaire) and NEEDS (Nucléaire : Energie, Environnement, Déchets, Société) programs of the French Centre National de la Recherche Scientifique (CNRS), the IN2P3 department of CNRS, and the European program EVOL (Evaluation and Viability of Liquid Fuel Fast Reactor System) of FP7 for their support. We are also very thankful to Elisabeth Huffer for her help during the translation of this paper and to Olivier Doche for his advice in some of the heat exchanger aspects.
References [1] L.Mathieu et al, “Possible Configurations for the TMSR and advantages of the Fast Non Moderated Version”, Nuclear Science and Engineering161, pp78–89 (2009) [2] E. Merle-Lucotte, D. Heuer, M. Allibert, M. Brovchenko, N. Capellan, and V. Ghetta, “Launching the Thorium Fuel Cycle with the Molten Salt Fast Reactor”, Proceedings of the ICAPP’09 Conference, Paper 11190, Nice, France (2011) [3] E.Merle-Lucotte et al, “Simulation Tools and New Developments of the Molten Salt Fast Reactor”, Proceedings of ENC2010, Paper A0115,Barcelona, Spain (2010) [4] S. Delpech, E. Merle-Lucotte, D. Heuer, M. Allibert, V. Ghetta, C. Le-Brun, L. Mathieu, G. Picard, “Reactor physics and reprocessing scheme for innovative molten salt reactor system”, J. of Fluorine Chemistry, 130 Issue 1, p. 11-17 (2009) [5] E. Merle-Lucotte, D. Heuer et al, “Minimizing the Fissile Inventory of the Molten Salt Fast Reactor”, Proceedings of the Advances in Nuclear Fuel Management IV (ANFM 2009) Conference, Hilton Head Island, USA (2009) [6] E. Merle-Lucotte, D. Heuer et al., “Optimizing the Burning Efficiency and the Deployment Capacities of the Molten Salt Fast Reactor”, Proceedings of the International Conference Global 2009, Paper 9149, Paris, France (2009) [7] C.Renault et al, “The Molten Salt Reactor (MSR) in Generation IV: Overview and Perspectives”, Proceedings of the GIF Symposium, France (2009) [8] M. Brovchenko, D. Heuer, E. Merle-Lucotte, M. Allibert, V. Ghetta, A. Laureau, "Preliminary safety calculations to improve the design of Molten Salt Fast Reactor", Proceedings of the PHYSOR 2012 International Conference, Knoxville, USA (2012) [9] A.-M. Bianchi, Y. Fautrelle et J. Etay, “Transferts thermiques”, Presses polytechniques et universitaires romandes, pp 170 (2004) [10] V. Ignatiev, O. Feynberg, A. Merzlyakov et al., ”Progress in Development of MOSART Concept with Th Support”, Proceedings of ICAPP 2012, Paper 12394 Chicago, USA, (2012) [11] O. Benes, R.J.M. Konings, “Thermodynamic properties and phase diagrams of fluoride salts for nuclear applications”, Journal of Fluorine Chemistry 130, pp 22-29 (2009) [12] http://www.oecd-nea.org/science/reports/2007/nea6195-handbook.html [13] D.J.Rogers, “Fusion Properties and Heat Capacities of the Eutectic LiF-NaF-KF Melt”, Journal Chem. Eng. Data, 27 p.366-367 (1982)
26 of 96 DEVELOPMENT OF MATERIALS TO WITHSTAND THE EXTREME, IRRADIATED ENVIRONMENTS IN ADVANCED NUCLEAR FISSION REACTORS
D.J. SHEPHERD M Eng (Hons) ProfGrad IMMM NNL Fuel Technology, National Nuclear Laboratory Ltd. NNL Preston Laboratory, Springfields, Salwick, Preston, Lancs. PR4 0XJ – UK
ABSTRACT
Six advanced nuclear fission reactors have been identified by the Generation IV International Forum (GIF) for further development in order to be safe, sustainable, proliferation resistant and commercially viable. Benefits are dependent on the individual system but can include significant reductions in the volume and radioactivity of waste, full replenishment of fissile nuclear fuel materials and the generation of significant heat for other applications such as chemical processing (for example hydrogen gas production).
However the environments within these proposed systems would be extreme with high dose irradiation, very high temperatures and corrosive conditions. Hence materials for use in these systems must meet stringent thermophysical, mechanical, chemical and neutronic requirements. Furthermore the material’s technology readiness and availability must be sufficient so as to be commercially viable.
Conventional materials are inadequate and so combinations of advanced materials are needed which must be specifically designed for each application including fuel cladding. The information in existing literature must be carefully evaluated so that preliminary selections can be made. Data gaps for these must then be filled by conducting new experimental trials in order to demonstrate viability and to aid with the development and validation of behaviour models and design codes.
This paper draws on the experience of the National Nuclear Laboratory to discuss the challenges for materials in advanced fission reactors alongside potential solutions and candidate materials including advanced steels and nickel alloys, refractory metal alloys and ceramic composites. The links to materials challenges within the fusion programme are also discussed.
1. Materials Challenges for Generation IV Systems
Six advanced nuclear reactor systems are being developed by the Generation IV International Forum (GIF) in order to be safe, sustainable, proliferation resistant and commercially viable [1, 2]. These systems are detailed in Figure 1. Five would be capable of utilising fast neutrons, which means that they would be theoretically capable of significantly reducing the volume and radioactivity of nuclear waste as well as full replenishment of fissile nuclear fuel materials. Four of the reactors would also have a sufficiently high outlet temperature (>700°C) for chemical processing applications, in particular VHTR and GFR could allow for hydrogen gas co-production.
However the challenges for materials to operate in these reactors are extreme due to the very high temperatures, high dose irradiation and severely corrosive environments that will exist in these systems. Conventional materials are inadequate for many of the proposed components and structural materials. Therefore advanced materials are needed that may need to be specially optimised for each application. Significant international co-operation will be needed through GIF in order to meet these challenges. Data gaps must be addressed in
27 of 96 order to prove material viability and to validate behavioural models as well as to inform design codes. It will also be necessary to link with fusion and space programmes in order to share in their expertise in materials for extreme, irradiated environments [3].
Sodium Fast Reactor (SFR) Lead Fast Reactor (LFR) Gas Fast Reactor (GFR)
Fast neutrons Fast neutrons Fast neutrons Ceramic or metallic fuel Ceramic fuel Ceramic fuel Liquid sodium coolant Liquid lead coolant Helium gas coolant 550°C 480-800°C 850°C
Very High Temperature Super-Critical Water Molten Salt Reactor (MSR) Reactor (VHTR) Reactor (SCWR)
Thermal neutrons Thermal or fast neutrons Thermal or fast neutrons Ceramic coated particle fuel Ceramic fuel Molten salt fuel Helium gas coolant Super-critical water coolant No primary coolant 900-1000°C 510-625°C 700-800°C
Figure 1: Gen. IV reactor systems [1, 2]
Two of the biggest materials challenges are the development of suitable fuel cladding materials and the development of appropriate material joining technology. A brief list of desirable characteristics for a Gen. IV cladding material would be:
• Low neutron absorption • Advantageous thermophysical properties • Good mechanical properties • Good resistance to high temperatures and high dose irradiation • Good compatibility with fuel, coolant and environmental species • Low cost • High technology readiness • Low biological hazard (toxicity, activation) • Suitable for recycle
28 of 96 Materials selection will inevitably involve ‘trade-offs’ between these characteristics and it is likely that a combination of materials or surface treatment will be needed in order to ensure compatibility with the operating environment as well as during any reactor transient or accident conditions.
2. Generation IV Candidate Materials
2.1 Improvements on conventional materials
The conventional structural and cladding materials used in Gen. II and III reactors still have scope for some limited evolutionary improvement for their continued use in Gen. IV reactors. Though any enhancements are likely to be modest, the vast amount of operating experience for these materials gives them an advantage over the more revolutionary materials that have never been widely employed in reactors.
Reduced Activation Steels There has been a sizeable effort in recent years to develop steels with lower induced radioactivity following neutron irradiation such as the RAFM (Reduced Activation Ferro/Martensitic) and Eurofer grades [4]. The removal of cobalt is particularly desired in order to prevent its activation to Co-60. The replacement of highly activating elements such as Nb and Mo with lower activation elements such as V and Ta is also desired.
Ferro/Martensitic (F/M) Stainless Steels F/M stainless steels are being widely evaluated for use in Gen. IV applications in particular the T91 grade (9wt%Cr, 1wt%Mo) [4]. These steels show greater resistance to irradiation- embrittlement than austenitic stainless steels though resistance to corrosion is lower and there is no operating experience for existing reactors. Al-coating technology has shown promise for improving the corrosion resistance of these steels.
Oxide Dispersion Strengthened (ODS) Steels The formation of a fine dispersion of oxide particles (primarily Y2O3) within steels has been found to have a remarkable strengthening effect that especially improves creep resistance and will allow for higher temperature operation [4]. However, production of these ODS materials (not just steel) is very difficult and has not been proven on a large scale. Currently the primary route is powder metallurgy involving a mechanical alloying (MA) stage where the oxide is finely dispersed within the steel powder by attritor milling. Welding of ODS steels has yet to be demonstrated and is likely to be problematic due to the potential for destroying the carefully engineered ODS microstructure within the heat affected zone.
Nickel-base Alloys Ni-base alloys such as inconels are currently used for key structural components in existing reactors and Ni alloys have also shown the potential to operate as cladding materials in SFRs [5]. Ni alloy development is relatively mature in particular the superalloys. The scope for improvement of Ni alloys might be relatively limited but they may be more widely deployed in Gen. IV as they will allow for higher temperature operation than current steels.
2.2 Semi-refractory metal alloys
Semi-refractory metal alloys are desirable for Gen. IV as structural or base materials for cladding as they offer the potential to operate at higher temperatures than conventional alloys. The bases for these alloys are those stable metallic elements of feasible abundance with melting temperature (Tm) of 1650-2000°C, excluding Zr due to low high temperature strength. Toxicity, activation, neutron absorption and cost are not major concerns for these materials. However there is no significant operating experience for these materials in nuclear systems and irradiation embrittlement below 400°C is a concern as well as irradiation-
29 of 96 enhanced creep at temperatures above 0.5Tm [6]. Some of the key differentiating characteristics for these alloys are listed in Table 1.
Titanium alloys Chromium alloys Vanadium alloys Elemental Tm (°C) 1660 1857 1902 Favoured alloys V-4Cr-4Ti with Si, Al & Y Ti-6Al-4V Cr-5Fe-1Y O (wt%) 2 3 trace additions Experimental alloys for Main nuclear Low but significant Low fusion systems and fast experience aerospace experience reactors Ductility High Low Medium Corrosion Low (needs surface High High resistance treatment) ODS strengthening is ODS strengthening is possible possible Excellent strength to weight Other ratio Needs vacuum production Cr is naturally a relatively as V is vulnerable to brittle metal interstitial embrittlement Table 1: Key differentiating characteristics of semi-refractory metal alloys
2.3 Refractory metal alloys
Refractory metal alloys are candidate materials for Gen. IV applications as they offer the potential to operate at much higher temperatures than either conventional or semi-refractory alloys. For cladding, they would only really be viable as coatings or liners due to their relatively high neutron absorption penalty which would impact negatively on reactor economics. The bases for these alloys are those stable metallic elements of feasible abundance with Tm above 2000°C. Re is so expensive that it can only be considered as a minor alloy constituent. Hf has too great a neutron absorption to be considered for use in cladding except again as a very minor constituent, though it may still be useful for control rod applications. There is currently relatively limited operating experience for these materials in nuclear systems. All of these alloys may be vulnerable to irradiation-embrittlement at comparatively high temperatures (500-800°C) due to lack of thermal recovery (annealing) [6]. Some of the key differentiating characteristics for these alloys are listed in Table 2.
Molybdenum Niobium alloys Tantalum alloys Tungsten alloys alloys Elemental Tm (°C) 2647 2610 2996 3410 Favoured alloys PWC-11 TZM T-111 5 to 20% Re, Ta, V (wt%) (Nb-1Zr-0.1C) (Mo-0.5Ti-0.1Zr) (Ta-8W-2Hf) or Mo Cost Medium Medium High Low Main nuclear Space propulsion Components in Gen Space propulsion Coatings for fusion experience reactor cladding II/III reactors reactor cladding systems High (can be Activation reduced by (medium / long High Low Low expensive isotopic term) tailoring) Very Low (could be Ductility Moderate Low Moderate improved by nano- grains) Low (needs Corrosion protective High High Moderate resistance treatment) High short term Needs vacuum decay heat production as Nb is ODS strengthening High short term Other vulnerable to is possible Poor thermal activation interstitial conductivity for a embrittlement metal Table 2: Key differentiating characteristics of refractory metal alloys
30 of 96
2.3 Ceramic composites and others
Given the significant problems associated with refractory metals, Continuous Fibre-reinforced Ceramic Composites (CFCCs) are understandably receiving much attention as potential high temperature materials. Carbon Fibre-reinforced Carbon (CFC) is quite technologically mature but has been found to degrade unacceptably under irradiation [7]. Silicon Carbide fibre- reinforced Silicon Carbide (SiCf/SiC) has been found to be significantly more radiation resistant and may be a viable materials option for operation above 1000°C [6]. SiCf/SiC has been significantly developed by the fusion programme to produce nuclear grade materials. However there remain a number of issues including large scale production, suitable joining, degradation of thermal conductivity under irradiation and performance as a barrier to fission products as well as coolants. Other possible alternatives such as Cermets and the recently discovered ductile MAX phase ceramics may have too low technology readiness [8] (a MAX phase contains: M - a metal element, A - a metalloid element and X - a non-metal element).
3. Conclusions
The development of materials to withstand the extreme, irradiated environments in advanced nuclear fission systems is a huge challenge but one for which there are a number of exciting candidate materials. It is worth noting that through the significant achievements of the co- ordinated international nuclear materials research programs (especially the fusion and space programs), materials challenges for Gen IV reactors that were thought to be insurmountable in past years are now looking ever more feasible. The recent development of ODS alloys, refractory metal alloys, SiCf/SiC and MAX phase ceramics are testament to these efforts but there is still a long way to go before viability is fully demonstrated and widespread, commercial deployment of such materials becomes a reality.
4. References
[1] A Technology Roadmap for Generation IV Nuclear Energy Systems; GIF-002-00; U.S. DOE Nuclear Energy Research Advisory Committee and the Generation IV International Forum; 2002 [2] GIF R&D Outlook for Generation IV Nuclear Energy Systems; Gen IV International Forum; 2009 [3] Materials needs for fusion, Generation IV fission reactors and spallation neutron sources – similarities and differences; L.K. Mansur, A.F. Rowcliffe, R.F. Nanstad, S.J. Zinkle, W.R Corwin, R.E. Stoller; Journal of Nuclear Materials 329-333, p166-172; 2004 [4] Innovative materials for Gen IV systems and transmutation facilities: The cross-cutting research project GETMAT; C. Fazio, D.G. Briceno, M. Rieth, A. Gessi, J. Henry, L. Malerba; Nuclear Engineering and Design 241, p3514-3520; 2011 [5] Generation-IV nuclear power: A review of the state of the science; T. Abram, S. Ion; Energy Policy 36, p4323-4330; 2008 [6] Operating temperature windows for fusion reactor structural materials; S.J. Zinkle, N.M. Ghoniem; Fusion Engineering and Design 51-52, p55-71; 2000 [7] Structural Ceramic Composites for Nuclear Applications; W.E. Windes, Y. Katoh, L.L. Snead, E. Lara-Curzio, J. Klett, C. Henager Jr., R.J. Shinavski; INL/EXT-05-00652; 2005 [8] The MAX Phases: Unique New Carbide and Nitride Materials; M.W. Barsoum, T. El- Raghy; American Scientist 89, p334-343; 2001
31 of 96
ARCHER:- MATERIAL AND COMPONENT CHALLENGES FOR THE ADVANCED HIGH TEMPERATURE REACTOR
D. E. BUCKTHORPE Specialist Engineering Design, AMEC Booths Park, WA16 8QZ, Knutsford, Cheshire, UK
With contributions from members of the ARCHER Project M Davies1), F. Pra2), P.Bonnamy2), J. Fokkens3), M. Heijna3), A. Vreeling3), M. Kolluri3), F Bourlier4), D Lhachemi4), A. Woayehune 4), S Dubiez 4), P Hahner 5), M Futterer5), J Berka6), W Hoffelner7), C Klotz8) , P. Homerin9) , B Marsden10) , P Mummery10) ,C Bullough11), D Ponca11) , C Feuillette12), M Hoffmann12), F Rondet 13), A Pecherty13), F Baurand 13), F. Alenda13), M Esch14), N Kohlz15), J Reed16), J Fachinger 17), J Klower18).
1) AMEC, Knutsford, UK 2) Commissariat à l’ Energie Atomique (CEA.), Cadarache, France 3) Nuclear Research and Consultancy Group (NRG), Petten, Netherlands 4) AREVA -NP SAS, Lyon, France 5) European Commission, Joint Research Centre (JRC), Institute for Advanced Materials, Petten, Netherlands 6) NRI Rez, Czech Republic 7) Paul Sherer Institute (PSI), Switzerland 8) SGL Carbon, France 9) Graphtech Intern Ltd, France 10) University of Manchester, UK 11) Alstom,UK 12) MPA, University of Stuttgart, Germany 13)Alfalaval, France 14) Westinghouse Electric, Germany 15) TUV, Germany 16) EdF Energy, UK 17) ALD-VT, Germany 18 ThyssenKrupp VDM, Germany
ABSTRACT
A new European Framework Project ARCHER (Advanced High-Temperature Reactors for Cogeneration of Heat and Electricity R&D) started in 2011 as part of FP7 for a period of 4 years to perform High Temperature Reactor technology R&D in support of reactor demonstration. The project consortium encompasses conventional and nuclear industry, utilities, Technical Support Organizations, R&D organizations and academia. The activities proposed contribute to the Generation IV International Forum and will collaborate directly with related projects in the US, China, Japan, and the Republic of Korea in cooperation with IAEA and ISTC. This paper will address the challenges and work activities on materials and component technologies.
1. Introduction
The High Temperature Reactor (HTR) is one of six advanced fission systems of interest for meeting the Generation IV (GIF) goals of attaining highly economic, safe, reliable, sustainable, proliferation-resistant systems. The HTR offers significant advantages for the longer-term development of sustainable energy and in particular for heat and process applications and hydrogen generation. A European 7th Framework Project (FP7) called EUROPAIRS [1] examines this potential bringing together a partnership of (V)HTR nuclear and process heat suppliers.
The FP7 Project ARCHER, which is a four year programme, builds on a solid HTR technology foundation established in Europe, through former national UK and German HTR programmes and EU FP4, FP5 and FP6 programmes, with the RAPHAEL project completed in April 2010 representing the latest, internationally
32 of 96 acknowledged achievement. Former papers [2, 3] describe the HTR materials and components activities and the success story and highlights of the technical achievements of RAPHAEL covering core physics, fuel, waste management, materials and components, system integration and safety.
The ARCHER Project is primarily oriented towards the short and mid-term needs of cogeneration or process heat requirements. Some technologies studied in RAPHAEL have been considered as mature or close to being so, and have not been continued within ARCHER. The ARCHER project consists of five subprojects involving co- ordination, system integration, fuel and fuel cycle, materials and components and knowledge management. The safety work investigates the Key Safety Challenges identified by the Safety Advisory Group in RAPHAEL with the main focus of the work on air and water/steam ingress into HTGR reactor cores and on the role of graphite/carbonaceous dust in the radiological source term analysis chain from the reactor core to the environment. The fuel work capitalizes on the irradiated material available with state of the art PIE, for improved and fundamental fuel behaviour understanding and the further advancement of fuel performance codes. The work includes an essential fuel performance validation test in an international framework, and will strengthen and extend the unique position Europe holds in HTR fuel development. A description of the materials and components work and its challenges is addressed in this paper. The knowledge management activities look at communication and dissemination and appropriate education and training activities for young engineers which is an important requirement for the longer term development of the technology.
2. Materials and components activities in ARCHER.
The aim of the materials and component activities is to study key materials and critical components. The activities are performed within a Subproject (SP4) comprising of four work packages (WP’s) two of which address non-metallic (graphite) and metallic materials and two investigate key components (i.e. IHX and SG) that can be promoted as potential candidates of a demonstrator. The graphite work is a continuation of the investigations started in RAPHAEL and former projects and within ARCHER will address the full post irradiation examinations (PIE) of the RAPHAEL tests and the specification of the graphites that should be used for the reactor core. Additionally there will be irradiation tests at low dose to serve as a basis for investigating changes due to irradiation at a micro-structural level. The activities on the metals will aim to formulate data for designers, and to perform a gap analysis and tests on available nickel based alloys, including welds, for the specific use of high temperature Steam Generators (SG) and a compact plate Intermediate Heat Exchanger (IHX), especially Alloy 800H material, where ever existing codes are lacking in relevant data (ASME and RCC-MR). For the two components work packages two heat exchangers concepts will be addressed: the IHX Plate Stamped Heat Exchanger (PSHE) design through testing of a reduced scale mock-up in a realistic temperature environment and an overview of SG designs as a mature technology to define the most promising lead concept for HTR near term applications. The following sections summarise in more detail the activities and challenges of the technological work involved.
2.1 Graphite
This WP covers the irradiation behaviour and selection of currently available graphite grades for the HTR core. The work includes the full PIE of the high dose experiments performed in RAPHAEL (see Figure 1) and the selection of the best performing graphites for the HTR core. A second objective covers the scientific and micro-
33 of 96 structural evaluation of the graphites through a low dose experiment to help develop better graphite grades with improved irradiation tolerance and a more stable production over a longer period. A primary objective will be to recommend the best available graphites from today’s graphite grades for the next generation HTR core for process heat and electricity generation.
1. INNOGRAPH-1A 750oC, low/medium dose (FP5) 2. INNOGRAPH-1B 750oC, high dose (RAPHAEL) o 3. INNOGRAPH-2A 950 C, low/medium dose (RAPHAEL) 4. INNOGRAPH-2B 950oC, high dose (RAPHAEL) o 5. INNOGRAPH-1C 750 C, low dose (ARCHER)
Figure 1 OVERVIEW OF IRRADIATION EXPERIMENTS
The main objective of the work package is to determine the irradiation behaviour of each graphite tested, and then to down-select from these the better graphites from a design viewpoint. To this end, the graphites are assessed mainly in terms of peak shrinkage, the dose at which the peak shrinkage, or shrinkage “turn-around” occurs, the dose to reach original dimensions/volume, the anisotropy in the dimensional change behaviour and the scatter in the data. The work includes the PIE of the high dose experiments performed in RAPHAEL at 750oC and 950oC (INNOGRAPH-1B and INNOGRAPH-2B) and the assessment of the data for each graphite. The assessment will provide important information on graphite behaviour needed for core design purposes.
In additional, a low dose experiment will be carried out at 750oC (INNOGRAPH-1C). The objective is to obtain data in the low dose region for those properties that are more rapidly affected by fast neutron irradiation (Young’s modulus, strength and thermal diffusivity/ conductivity). It is thought that the micro-structural changes that arise in graphites at low dose have an effect on their medium and high dose behaviour. It is expected that a fuller understanding of these micro-structural changes at low dose might lead to the development of new graphite grades with improved irradiation tolerance at medium to high doses.
The progress of these tasks is well advanced with specific physical properties for each graphite tested becoming available for transcribing into design curves. The results extend the lower dose full PIE results previously reported in RAPHAEL- IP from ~10 displacements per atom (dpa) to ~23 dpa. The dimensional change data showed that for both temperatures most of the samples had gone well beyond shrinkage turn-around, and a significant number had exhibited positive growth/volume change (i.e. swelling beyond original dimensions/volume). The Dynamic Young’s Modulus data showed the expected increase at lower dose to a
34 of 96 plateau, followed by a further increase at medium dose, and finally a decrease at high dose. As expected, the coefficient of thermal expansion was found to have decreased and reached a stable plateau at medium dose, with indications of a slight increase at high dose. The thermal diffusivity and conductivity values, which typically fall rapidly at low dose, had also reached a plateau at medium dose. Some samples also showed a further reduction in thermal diffusivity/conductivity at high dose which is consistent with other graphites irradiated in the past. Unfortunately the number of INNOGRAPH-2B samples for which thermal diffusivity and conductivity were measured was strongly reduced due to sample activity and by the level of swelling of samples. The results from these full PIE experiments are currently being assessed to establish the better graphites from the core design viewpoint. The important information on graphite behaviour needed for the design of future HTR cores will also be derived.
The development of the low dose experiment is well advanced with the samples specified according to their grain orientation. The experiment utilises the experience from the RAPHAEL experiments and consists of 8 drums, each with 3 columns, to contain the samples. The rig contains 24 thermocouples to monitor and control temperature during the irradiation and the sample holders will contain nine neutron fluence detector sets that will be analysed after the irradiation. The sample holder is filled with helium to prevent oxidation of the graphite samples. The nominal temperature of the experiment will be 750°C. The top drum of the experiment is to contain approximately 30 samples which will have a nominal temperature of 650°C (as was the case for the INNOGRAPH-1A experiment). Four major graphite grades have been selected plus some minor grades with their location with respect to the original block from which they were machined identified. The experiment is planned to start in the last quarter of 2012
2.2 Metals
This WP investigates high temperature metallic materials for the HTR and heat exchange circuit including the steam generator (SG), the intermediate heat exchanger (IHX). The activities include data base management and links with organisations such as the European Energy Research Alliance (EERA) plus damage characterisation under creep, fatigue and creep/fatigue interaction, material performance under high temperature corrosion and investigations on specific welds. The WP includes some investigations on instrumentation covering in-pile and out-of pile tests using representative atmospheres on a new thermometry device. The overall objective is to recommend application limits for different materials for use in HTR process heat and electricity generation.
The procurement and transfer of material for the tests was completed during the first year of ARCHER. Two blocks of Alloy 800H (dimension: 500mm x150mm x16mm) plus two WIG-welded plates (dimension: 500mm x150mm x16mm) were provided by ThyssenKrupp VDM to MPA Stuttgart to forward to all partners according to their requirements. The partners are machining their own samples on their respective sites for their test programmes. A survey of corrosion data generated in previous HTR and related programmes has been carried out providing information on candidate materials for V/HTR systems and mechanisms of degradation of these materials in V/HTR helium coolant. Results of tests on high temperature alloys within previous HTR programs have been summarized and properties of selected alloys (e.g. Alloy 800H, Hastelloy X, Alloy 617, Haynes 230) specified in detail. Work is also underway to evaluation data gaps in the JRC Petten’s MATDB data base which contains over 2600 data stress for Alloy 800H. The existing data set contains tensile, creep, low cycle fatigue and creep crack growth data, together with information on welds and
35 of 96 effects of irradiation. Preparation work for the experiments is well underway with creep and environmental testing to commence as soon as practical. The corrosion tests will include out-of-pile tests in a furnace and high temperature helium loop and in-pile tests in a High Temperature Helium Loop (HTHL) in the Reactor LVR-15 (thermal neutron flux 1.5 x 1014 n/cm2s, fast flux up to 2.5 x 1014 n/cm2s).
2.3 IHX
This WP covers the demonstration of the feasibility of an IHX module including the design and manufacture of a mock-up made from Alloy 800H and tested in the CLAIRE loop at operating conditions (750°C – 800°C) to assess the compact IHX (PSHE) lifetime. CFD and FEM calculations are to be performed also tests on Alloy 800H thin sheets to determine behavioural laws for analysis and the creep behaviour of weld joints samples to estimate corresponding weld coefficients for life estimation.
For higher temperature exchanges the gas to gas IHX is likely to give the most practical, cheapest, robust and compact solution and here issues associated with steam transfer such as corrosion have less impact. The Plate Stamped Heat Exchanger (PSHE) design offers the most promising solution regarding robustness in the long term. This work package (IHX) aims to demonstrate the feasibility of one real size IHX module with design and manufacture of one ‘full size’ mock-up made of Alloy 800H and to test it in the CEA’s CLAIRE loop at operating conditions (750°C – 800°C) to assess the compact IHX (PSHE) lifetime. To reach this objective, complex CFD and FEM calculations will be required in support, as well as sets of tests on Alloy 800H thin sheets for behaviour laws and creep tests on weld joints samples to estimate corresponding coefficients. Past experiments and basic review of available data have shown that knowledge on creep and creep-fatigue on this alloy are much needed.
Figure 2 COMPACT IHX ARRANGEMENT
The first task covers an assessment of the lifetime and potential failure modes of the actual size IHX and the mock-up model using the same material behaviour laws for both. The task includes specification of loading conditions and material data, also CFD pre-calculations to determine temperature maps and creep-fatigue calculations to assess the lifetime. The results will feed directly into the work of the other tasks. This work has largely been completed with an initial emphasis on the specification of the temperature maps and calculation of transient stresses based on elastic evaluations. These are being followed up by inelastic analysis to assess strains. The
36 of 96 results have been used to help plan the experiments and the materials tests to be carried out as part of the materials development.
A second task looks at the development of the required welding and machining processes for the mock-up. The manufacture involves the use of machining procedures and the forming of stamped plates which are welded together using specific welding tools at the plate edges and within the port holes. The development of these devices involves a series of welding trials which acts to test the devices, to develop the welding procedures and to establish representative welded samples for tests. Significant progress has been made on the fabrication and machining actions and development of the welding tools. Laser welding has been used which showed some notable distortion initially and the need to optimize the welding procedures for Alloy 800H.
Once the mock-up has been manufactured it will be inserted in the CEA Claire Loop for around 200 days of tests at nominally around 750°C/800°C plus thermal transients with representative number of cycles. The results from the tests will be used along with CFD and FEM post-test calculations for lifetime estimates. The scheduling and availability of the CEA Loop Tests has been identified for 2013.
To support the lifetime assessments a review of data for Alloy 800H design properties for thin and thick sections has been performed. Although a significant amount of material test and property data are available for thicker section Alloy 800H (typically plate or bar material exceeding 15mm section), the amount of testing on Alloy 800H or other variants of Alloy 800 in sections of 5mm or less is low. A series of tests on representative thin section plate and welded test pieces (derived from the second task) including tensile tests at different strain rates, creep tests and relaxation tests for temperatures up to 850°C are planned for comparison with the developed design information to help confirm its suitability.
2.4 SGU
This WP deals with the components associated with the heat transport circuit with a particular focus on the Steam Generator (SGU). The work covers an evaluation of expected key transients and operating conditions, selection of a lead concept and identification of the main issues affecting reliable operation, design, manufacture and inspection and a Cost/Benefit Analysis. The objective is to arrive at a concept that will provide reliable operation for the HTR for process heat and electricity generation.
For the SGU, significant benefit can be taken from the past developments which have seen deployment for temperatures up to 650oC and there are currently studies looking for solutions at 700oC. To this end investigation of the SGU taking benefit from established providers and users is a key requirement for progress. Such investigations are to address SGU concept options, manufacture and inspection issues, thermal hydraulic and structural integrity issues and cost benefit analysis.
Work has progressed in assisting the development of the specification of the Steam Generator operating conditions using information provided by Alstom from their existing experience of SG’s in operation and from AREVA on their HTR Module experience. Thermal data, key requirements and earlier concepts have been described along with material requirements and lifetime issues. The information on the Module contains a description of the primary and secondary side and summarises the conduct of operations and transients following a main heat transfer malfunction to support the development of the Specification and selection of the Main SG Concept. Later tasks address key aspects of manufacture of the selected SGU
37 of 96 Concept (tube bundle, including tube to tube-plate joints, etc.), issues of integrity, failure mechanisms, manufacturing risks, etc. .Detailed work on these is expected to commence once the Lead SG Concept Specification has been identified.
3. Conclusions
The ARCHER Integrated Project commenced in February April 2011 and focuses on HTR short and mid-term needs of cogeneration or process heat requirements. ARCHER technology will use the information from the FP7 Support Action EUROPAIRS involving both end-users and V/HTR promoters.
In former projects important results have been obtained in the areas of core physics, fuel, waste treatment and disposal, materials, components, safety and system integration raising the worldwide interest of the V/HTR community. This paper addresses the key activities and challenges of the HTR materials and components within the ARCHER project. ARCHER results are expected to further the advancement of the HTR for cogeneration application and are an important input to the GIF members who are interested to receive access to the results in order to complete their database. It is anticipated that the results from ARCHER will complete keys areas of understanding and forward the development of the HTR towards the establishment of a demonstrator for cogeneration.
4. Acknowledgement
Acknowledgement is given to the partner contributions from the ARCHER Project. Acknowledgement is also given to the European Atomic Energy Community (“Euratom”), the co-sponsors of this Framework Project. The information provided herein is the sole responsibility of the authors and does not reflect the Community’s opinion. The Community is not responsible for any use that might be made of the data appearing in this publication.
5. References
[1] E. Bogusch, et al., EUROPAIRS: the major Nuclear Cogeneration Project in FP7. HTR2010, Paper 44, Prague, Czech Republic, October 2010
[2] D. Buckthorpe et al. European Project on Materials for Modular HTR’s, ENC2002, Lille, France, 7-9 October 2002
[3] D. Hittner et al. VHTR, a success story, ENC2010, Barcelona, 30 May – 2 June 2010.
38 of 96 Pressure Drop Analysis in a Pressure-Tube SuperCritical Water-cooled Reactor
W. Peiman, Eu. Saltanov, I. Pioro and K. Gabriel1 University of Ontario Institute of Technology Faculty of Energy Systems and Nuclear Science 1Faculty of Engineering and Applied Science 2000 Simcoe Street North, Oshawa, Ontario, L1H 7K4, Canada [email protected], [email protected], [email protected], [email protected]
Abstract Pressure-drop calculations and fuel- and sheath-temperature profiles are important aspects of a nuclear-reactor design. The main objective of this paper is to determine a pressure drop in a fuel channel of a Pressure-Tube SuperCritical Water-cooled Reactor (PT SCWR) and to calculate temperature profiles of a sheath and fuel elements. One-dimensional steady-state thermal- hydraulic analysis was conducted. In this study, pressure drops due to friction, acceleration, local losses, and gravity were calculated at supercritical conditions. Keywords: Pressure Drop, Supercritical Conditions, SCWR, Thermal-Hydraulics, and Nuclear Reactor
1. Introduction SuperCritical Water-cooled Reactor (SCWR) concept is one of the six nuclear-reactor concepts, which are categorized under Generation IV nuclear-reactor program. A generic PT SCWR operates at a pressure of 25 MPa with inlet and outlet coolant temperatures of 350 and 625°C (Pioro and Duffey, 2007). The high outlet temperature and pressure of the coolant make it possible to use supercritical “steam” turbines, which led to high thermal efficiencies at coal-fired power plants. Therefore, among the six Generation IV nuclear-reactor concepts, only SCWR uses water as the coolant. In regard to thermal-hydraulic aspects of the SCWR design, heat transfer and pressure-drop calculations at supercritical conditions are of interest and importance. The main issue for heat transfer at supercritical conditions is related to a prediction of the heat transfer coefficient at normal and deteriorated heat-transfer regimes. Similarly, the primary issue related to pressure- drop calculations is associated with the determination of frictional and minor losses, because of the dependency of these parameters on empirical correlations. In addition, the correlations developed at subcritical conditions might not be applicable to supercritical pressures. Therefore, there is a need for theoretical and experimental research to develop and verify such correlations. The main objective of this paper is to determine a pressure drop of a PT SCWR. This study only includes a pressure drop inside a fuel channel due to friction, acceleration, local losses, and gravity.
1
39 of 96 2. Pressure-Tube Supercritical Water-cooled Reactor 2.1. Core Design
In terms of a pressure boundary, SCWRs are classified into two categories: 1) Pressure Tube (PT) or Pressure Channel (PCh) SCWRs and 2) Pressure Vessel (PV) SCWRs. For the purpose of this paper, a generic PT SCWR with a vertical core and downward flow has been studied. Table 1 provides operating parameters of this reactor design, which consists of a vertical core with a heated length of 5 m (Yetisir et al., 2011).
Table 1: Operating Parameters of PT SCWR.
Parameters Unit PT SCWR Total Thermal Power MW 2540
Coolant/ Moderator ˗ H2O/ D2O Pressure MPa 25
Tin / Tout Coolant °C 350 / 625 Mass Flow Rate per Channel kg/s 3.91
Thermal Power per Channel MWth 7.6 Number of Channels/Core Length ˗/m 336/5
2.2. Fuel-channel designs
This paper focuses on pressure-drop calculations related to a direct-flow fuel-channel design known as the High Efficiency Channel (HEC). Figure 1 shows a 3-D view of this fuel channel. The HEC design consists of 10 fuel bundles or a fuel-bundle string, a perforated liner tube, a ceramic insulator, and a pressure tube (Chow and Khartabil, 2008). In this paper, the 43-element fuel-bundle design was chosen for calculating the pressure drop in an SCW fuel channel. The 43-element fuel bundle consists of 42 fuel elements with an outer diameter of 11.5 mm and a central unheated element, which contains a burnable neutron-absorbing material and has an outer diameter of 20 mm (Leung, 2009). The length of the fuel bundle is 50 cm.
2
40 of 96
Figure 1: 3-D View of High Efficiency Channel (based on Chow and Khartabil, 2008).
3. Pressure Drop at Supercritical Conditions
Pressure drop calculation at supercritical conditions is similar to subcritical pressure drop analysis. However, the only difference is that the thermo-physical properties of a working fluid changes significantly as the fluid passes through the pseudocritical region. Thus, appropriate modifications are required when determining a pressure drop along the length of a fuel channel operating at supercritical conditions. In general, the total pressure drop at supercritical pressures is the sum of four pressure-drop components. As shown in Eq. (1), these pressure drops are due to friction, flow obstruction or local losses, acceleration, and gravity (Pioro and Duffey, 2007).
P Pfr Pl Pac Pg (1)
The frictional pressure drop, which is a function of the fluid density, mass flux, length and diameter of the tube (or hydraulic-equivalent diameter for a fuel channel), and friction coefficient, is calculated based on Eq. (2). The friction coefficient is determined based on empirical correlations.
L V 2 L G2 P f f (2) fr D 2 D 2
Filonenko developed a correlation, shown as Eq. (3), for the friction factor, f, in smooth circular tubes. This equation is valid for Reynolds numbers in the range of 4103 to1012 (Pioro and Duffey, 2007).
3
41 of 96 1 (3) f 2 1.82 log10 Reb 1.64 In addition to the Filonenko correlation, other equations and correlations have been developed, which should be taken into consideration. Mikheev (1956) developed a correlation shown as Eq. (4), which is applicable to flow of water and other fluids in smooth tubes (Pioro and Duffey, 2007). Selander (1978) developed an explicit form of the Colebrook equation, which is shown as Eq. (5). This equation is valid for smooth and rough tubes (IAEA, 2001).
1/3 1 Pr f w (4) 2 1.82 log10 Reb 1.64 Prb
4 f 2 (5) 10 0.2 3.8 log Re D
These friction coefficients correspond to isothermal or adiabatic conditions. Consequently, several researchers have developed correction factors to account for effects of heat transfer on the friction coefficient. Table 2 (Leung et al., 2005; Pioro and Duffey, 2007; IAEA, 2001) provides a list of these correction factors and their uncertainties if applicable. Table 2: Correction Coefficients for Frictional Losses.
Developer Equation, Comments
Leung (2005) - ( )
Popov (1967) ±10% uncertainty ( )
Kirillov et al. (1990) Applicable to normal and deteriorated ( ) heat transfer regimes
Tarasova and Leont’ev (1968) ±5% uncertainty ( )
The pressure drop due to flow obstruction is determined according to Eq. (6) (Pioro and Duffey,
2007). In this equation, the local resistance coefficient, l , is calculated using appropriate correlations or equations, which are developed for different flow obstructions.
V 2 G 2 P (6) l l 2 l 2
4
42 of 96 The acceleration pressure drop is expected to be one of the most significant pressure drops at supercritical pressures, especially, when the ratio of the heat flux to mass flux, q/G, is high. The importance of the acceleration pressure drop is due to the fact that the thermo-physical properties of the working fluid (i.e., coolant), specifically its density, undergo significant changes within the pseudocritical region. Thus, it is necessary to divide the length of a fuel channel into small axial segments (increments) in order to take into account for the density changes when undertaking pressure-drop calculations for an SCWR fuel channel. The acceleration pressure drop is calculated using Eq. (7) (Pioro and Duffey, 2007).
1 1 2 2 2 Pac i1 Vi1 i Vi G (7) i1 i
The pressure drop due to gravity is zero in horizontal tubes or fuel channels. However, in vertical geometries, the pressure drop due to gravity should be taken into consideration. The pressure drop due to gravity can be calculated using Eq. (8) (Pioro and Duffey, 2007). In this equation, H, L, ρ, and ϴ are the fluid enthalpy, tube (or fuel channel) length, density, and the inclination angle to horizontal plane (e.g., 90° for a vertical fuel channel), respectively. In Eq. (8), “+” sign is for an upward flow direction and “-“sign is for a downward flow direction.
H H i1 i1 i i Pg g L sin (8) H i1 H i
4. Numerical Model A code was developed in MATLAB, and NIST REFPROP software was used for retrieving thermo-physical properties of a light-water coolant in order to calculate the following parameters: fuel centerline temperature, sheath temperature, heat transfer coefficient, and pressure drop. In the developed code, a steady-state one-dimensional heat-transfer analysis is adopted to perform the calculations. To complete a set of calculations, the code performs the following calculations in order. First, the heated length of the fuel channel is divided into small segments of one-millimeter length. Second, the temperature profile of the coolant is calculated. Third, the outer- and inner-surface temperatures of the sheath are calculated. Fourth, the temperature of the fuel in the radial and axial directions was calculated. Fifth, the pressure drop is calculated based on the methodology described in Section 3.
5. Results A steady-state one-dimensional thermal-hydraulic analysis was conducted to calculate the pressure drop in a generic SCW fuel channel with the following specifications: a mass flow rate of 3.91 kg/s, an outlet pressure of 25 MPa, a coolant inlet temperature of 350°C, a thermal power per channel of 7.6 MWth. Figure 2 shows the pressure drop and its components along the length of the channel. It should be mentioned that the liner tube was assumed to be a solid tube. In other words, the pressure loss associated with liner being a perforated tube has not been taken into consideration. Table 3 provides a summary of pressure drop due to the three examined
5
43 of 96 correlations/equations with and without correction factors. The surface roughness was assumed to be m. The local pressure losses are very dependent on the geometry of the fuel channel, especially the fuel bundle. Thus, in order to account for the pressure drop due to the presence of fuel bundles, end plates, bearing pads, and spacers local loss coefficients are required. As mentioned, these coefficients depend of the design of a fuel bundle. However, since the corresponding values are not available for a 43-element bundle some assumptions have been made in order to estimate the local pressure drop.
It has been assumed that the loss coefficients associated with end plates, spacers and bearing pads are 0.6, 0.12 (Brasnarof et al 2011) and 0.05 (Munson el al., 2006), respectively. It has also been assumed that there is one plane of spacers at the middle of a fuel bundle. Each fuel bundle has three planes of bearing pads and two end plates. Therefore, the total loss factor for each fuel bundle is 1.47. This approximation would result in a total local pressure loss of 35.5 kPa. It should be noted again that this number is just an approximation and further research is required to obtain more accurate loss coefficients if necessary.
Figure 2: Acceleration, Friction, and Gravity Pressure-Drop Profiles.
6
44 of 96 Table 3: Friction Pressure Drop at SCWR Conditions.
Friction Friction/Total Pressure Drop, kPa Factor Without Leung et al. Popov Kirillov et al. Tarasova and Correlation Correction (2005) (1967) (1990) Leont’ev (1968) Filonenko 39/66 41/67 28/54 32/59 40/67 Mikheev 36/62 37/63 26/52 30/56 37/63 Selander 203/229 210/236 144/170 167/193 208/235 Colebrook 211/237 219/245 150/176 173/200 217/243
6. Conclusions One-dimensional thermal-hydraulic analysis was conducted in order to calculate the pressure drop in a generic SCWR fuel channel. The calculated pressure drop consists of the pressure drops due to friction, acceleration, gravity, and local losses. Among these contributing factors to pressure drop, the pressure drop due to local losses was the highest followed by the friction and acceleration pressure drops. However, there are relatively high degrees of uncertainties associated with the calculated pressure drops due to friction and local losses. These uncertainties are mainly due to dependency of these pressure loss components on empirical correlations. Therefore, there is a need for developing correlations for the prediction of friction factor at supercritical pressures. Moreover, loss coefficients due to a fuel bundle and its components should be determined experimentally.
7. Acknowledgements Financial supports from the NSERC/NRCan/AECL Generation IV Energy Technologies Program and NSERC Discovery Grant are gratefully acknowledged.
7
45 of 96 8. References Abdalla, A., King, K., Qureshi, A., Draper, Sh., Peiman, W., Joel, J., Pioro, I. and Gabriel, K., 2011. Thermalhydraulic Analysis of 43-, 54-, 64-Element Bundles with UO2 Plus SiC Fuel for SuperCritical Water-Cooled Reactors, Proc. 19th Int. Conf. On Nuclear Engineering (ICONE-19), Osaka, Japan, October 24-25. Brasnarof, D. O., C.Marino, A., Bergallo, J. E., & Juanico, L. E. (2011). A New Fuel Design for Two Different HW Type Reactors. Hindawi Publishing Corporation, 15 pages. Chow, C. K. and Khartabil, H.F., 2008. Conceptual Fuel Channel Designs for CANDU-SCWR, J. of Nuclear Engineering and Technology, Vol. 40, pp. 1−8. IAEA, 2001. Thermalhydraulic Relationships for Advanced Water Cooled Reactors. Retrieved December 7, 2011, from IAEA: http://wwwpub.iaea.org/MTCD/publications/PDF/te_1203_ prn.pdf . Leung, L.K.H., 2009. Effect of CANDNU Bundle-Geometry Variation on Dryout Power, Proc. 16th Int. Conf. On Nuclear Engineering, Orlando, FL, USA, May 11-15. Leung, L.K.H., Groeneveld, D.C., Teyssedou, A. and Aub´e, F., 2005. Pressure Drops for Steam and Water Flow in Heated Tubes, Nuclear Engineering and Design, 235 (1), pp. 53-65. McDonald, M.H., Hyland, B., Hamilton, H., Leung, L.K.H., Onder, N., Pencer, J. and Xu, R., 2011. Pre-Conceptual Fuel Design Concepts for the Canadian Supercritical Water- Cooled Reactor, Proc. 5th Int. Sym. SCWR (ISSCWR-5), Vancouver, BC, Canada, March 13-16,. Munson, B.R., Young, D.F. and Okiishi, T.H., 2006. Fundamentals of Fluid Mechanics. John Wiley & Sons, Inc. Nuclear News, 2011. 13th Annual Reference Issue, USA: American Nuclear Society (ANS), March. Pioro, I.L., and Duffey, R.B., 2007. Heat Transfer and Hydraulic Resistance at Supercritical Pressure in Power-Engineering Applications, ASME, New York, NY, USA. Yetisir, M., Diamond, W., Leung, L.K.H., Martin, D. and Duffey, R. 2011. Conceptual Mechanical Design for a Pressure-Tube Type Supercritical Water-Cooled Reactor, Proc. 5th Int.Sym. SCWR (ISSCWR-5), Vancouver, BC, Canada, March 13-16.
8
46 of 96
ADVANCED REACTORS II
47 of 96 SPES-3: THE INTEGRAL FACILITY FOR SAFETY EXPERIMENTS ON SMALL AND MEDIUM SIZED REACTORS
R. FERRI, A. ACHILLI, G. CATTADORI, F. BIANCHI, A. LUCE, S. MONTI SIET S.p.A Via Nino Bixio 27c, 29121 Piacenza, Italy
P. MELONI, ENEA, UTFISSM v. Martiri di Monte Sole 4, 40129 Bologna - ITALY
M.E. RICOTTI Politecnico di Milano, Department of Energy Via La Masa 34, 20156 Milano, Italy
ABSTRACT From the technological viewpoint the advanced/innovative small and medium sized reactors are not different from large reactors. They differ for the higher degree of innovation implemented in their designs to achieve competitiveness and reliable performances as well as to enhance safety by incorporating passive and inherent safety features. To address the potential concerns related to the use of intrinsic and passive safety features and in particular to the primary-to-containment coupling during a LOCA, large experimental facilities are needed for demonstrating the performance and intervention sequence of safety systems during postulated design accidents and for providing data for code validation. To this end an integral test facility, called SPES-3, was designed and is under construction at SIET laboratories in Piacenza (Italy). The SPES-3 facility reproduces all parts and components of the IRIS plant with 1:100 volume scaling factor, 1:1 elevation scaling factor, prototypical fluid and thermal-hydraulic conditions. This paper describes the SPES-3 facility, its features and performance, besides the status of its realization. It also highlights its capability in simulating phenomena occurring in any kind of integral small modular reactor under development in the world.
1. Introduction Today, there is a renewed interest in the development and near term deployment of advanced/innovative small and medium sized reactors (SMRs)1. This technology is the most suitable option for deployment in countries with small electrical grid capacity or electricity demand, as well as for non-electrical nuclear energy applications, i.e. seawater desalination, district heating, hydrogen production and other process heat applications. Several SMR designs are currently at different stage of development through the world, such for example the Nuscale 45 MWe (USA) [1, the mPower 125 MWe (USA) [2, the Westinghouse SMR (W-SMR) 225 MWe (USA) [3, the CAREM 25 and 200 MWe (Argentina) [4], etc.. From the technological point of view the SMRs are similar to large reactors, but they differ for the higher degree of innovation implemented in their designs to achieve competitiveness and reliable performances. Moreover they make use of passive and inherent safety features to enhance safety. To address the potential concerns of using inherent and passive safety features to achieve targeted safety goals, large experimental facilities are needed for demonstrating the
1 According to the classification currently used by the IAEA, small reactors are the reactors with an electric power less than 300 MWe, medium-sized reactors are the reactors with a power between 300 and 700 MWe. Since 2008 this acronym is also used more in general for indicating Small Modular Reactors with power lower than 300 MWe.
48 of 96 performance and intervention sequence of safety systems during postulated design accidents. In particular, the primary-to-containment coupling during Loss Of Coolant Accidents (LOCA) is one of the most relevant issues to be investigated, especially for SMR with an integral primary system layout. This evaluation is of paramount importance not only for the licensing aspects, but also for the validation of stand-alone and coupled thermal hydraulic system codes, like RELAP5, TRACE, CATHARE, GOTHIC, etc.. To this end an integral test facility, called SPES-3, was designed to study the primary-to- containment coupling phenomena in an integral type SMR and to provide data for code validation. The construction of the SPES-3 facility is in progress at the SIET laboratories in Piacenza (Italy), under ENEA responsibility, within an R&D programme supported by the Italian Ministry of Economic Development (MSE), devoted to innovative reactor concepts and related fuel cycles. The SPES-3 facility is based on the IRIS reactor design, of which it reproduces the primary, secondary, containment and safety systems. The IRIS reactor, sketched in Fig. 1, is an advanced, medium size, modular nuclear reactor with an innovative integral configuration and safety features suitable to cope with LOCA through a dynamic coupling of the primary and containment systems [5]. For its integral layout, the reactor pressure vessel (RPV) includes all the main primary components: core, control rod drive mechanisms zone, pressurizer, helical coil steam generators and primary pumps. Two Emergency Boration Tanks (EBT) are directly connected to the Direct Vessel Injection (DVI) lines, high pressure side. Two stages of Automatic Depressurization System (ADS) help the RPV depressurization in case of accident. Each of the secondary loops is provided with an Emergency Heat Removal System (EHRS) rejecting the decay heat to the Refuelling Water Storage Tank (RWST), when the reactor is isolated. The containment includes the Dry Well (DW) and Reactor Cavity (RC), the ADS Quench Tank (QT), the Pressure Suppression System (PSS) and the Long-Term Gravity Make-up System (LGMS). A Passive Containment Condenser (PCC) system allows the plant long term cooling in case of EHRS unavailability.
EHRS ADS PCC EHRS
RWST ADSAD S 2 1 2 1 MSIV Steam Line PSS vent
QT EBT È B T RPVRV SGMT LGMS PSS
PSS LGMS MFIV Feed Line
Start - up FW
Fig. 1. IRIS containment and safety systems
An international consortium, initially led by Westinghouse, including industries, universities and research organizations, designed the IRIS reactor. Notwithstanding Westinghouse decision to leave the IRIS consortium in 2010, the Italian organizations decided to continue the development of IRIS with particular attention to the SPES-3 facility and the general safety features of integral type SMRs currently under development worldwide. The SPES-3 has important capabilities to offer a critical insight on several innovative solutions, adopted in some SMR designs (integral type, PWR based concepts).
49 of 96 This paper describes the SPES-3 facility, its features, performance and realization status. Moreover it highlights its capability in simulating phenomena occurring in other SMRs under development in the world.
2. SPES-3 facility overview
2.1. SPES-3 facility description The SPES-3 facility reproduces all the parts and components of the IRIS plant with 1:100 volume scaling factor, full elevation, prototypical fluid and thermal-hydraulic conditions [6]. The flow diagram and axonometric layout are shown in Fig. 2 and Fig. 3. The power channel (RV) includes the internals consisting of the electrically heated core simulator, the riser with control rod device mechanisms zone, the pressurizer, the pump suction plenum, the helical coil steam generators (SGs) wrapped around the central riser, the downcomer and lower plenum. Three SGs simulate the eight IRIS SGs (representing 2, 2, 4 out of 8), while a single pump simulates the eight IRIS pumps.
RWST
EHRS
DW
EBT RV LGMS
QT
PSS
RC
Fig. 2. SPES-3 facility flow diagram Fig. 3. SPES-3 facility layout
For room reasons, a single SPES-3 pump is located outside of the RV and connected to it by a piping system. The two emergency boration tanks are simulated and connected to the DVI lines. The four IRIS secondary loops are simulated up to the main isolation valves by three secondary loops (representing 1, 1, 2 out of 4). Each secondary loop includes a feed line (FL), a SG, a steam line (SL) and an emergency heat removal system (EHRS) with a vertical tube heat exchanger immersed in a refuelling water storage tank (RWST). The EHRS heat exchanger connected to the double secondary loop is in a pool, the other ones are in the other pool. Separated tanks connected among them and to the RV by pipes simulate the different IRIS spherical containment compartments: DW, RC, PSS, LGMS and QT. Such pipes do not exist in IRIS, consequently they are designed in terms of size and layout to limit their influence on the flow. Tank shape and dimensions are fixed in order to reproduce the trend of IRIS compartment volumes versus height. The SPES-3 PCC consists of a U-tube horizontal heat exchanger located at the DW top and it is dimensioned to remove specified power. Two trains connected to the PRZ top simulate three IRIS ADS trains: a single and a double train. Each train consists of a safety valve, a line to the Quench Tank (Stage-I), ending with a sparger to enhance the steam condensation under the water level, and a line to the Dry Well (Stage-II). The break line systems are designed to simulate split and double ended guillotine (DEG) breaks of different lines: DVI line, EBT top line, ADS Stage-I line, FL and SL. The existing auxiliary systems provide fluids (water and air) to the experimental
50 of 96 facility at the required temperature, pressure and mass flow. Direct current generators provide power to the fuel bundle, to the PRZ heaters and to the containment tank wall preheating system. A large set of instruments (about 600) is foreseen on SPES-3 to provide data both for the facility operation and test assessment. It consists of conventional and two- phase flow instrumentation used for direct measures, such as temperature, pressure, velocity, etc., and derived measures, i.e.: level by differential pressure, density, mass flow- rate, etc.. The rod bundle is instrumented with 120 wall thermocouples distributed at different levels, with a greater density at the upper levels. They provide the rod cladding temperature as well as the signals for core protection against superheating. The facility is suitable to perform both integral and separate effect tests, i.e. to simulate Design Basis Accidents (DBA) and Beyond Design Basis Accidents (BDBA) and to verify the EHRS and SG heat transfer capabilities.
2.2. Status of realization activities The detailed design of all components has been completed, including the thermal-hydraulic analysis of the main transients of the test matrix considered relevant for design activities. In support to the component design and procurement, an experimental campaign to investigate thermal performance of four heated rods of the bundle provided by two suppliers (ROTFIL and THERMOCOAX) has been completed, but the choice of the reference solution has not been done as the thermocouple reliability must be increased. The performed experimental campaign has required the design and construction of an ad hoc loop. Also the development and characterization of a spool piece for two-phase flow measurements has been done in order to limit intrusive devices and to cope with the different flow regimes occurring in the break lines, during the SPES-3 transients. It consists of a set of heterogeneous instruments: vortex for velocity, drag disk on load cell for momentum flux and capacitive detector for void fraction, [7] [8. A Venturi meter upstream of the break valves provides the mass flow until single-phase conditions persist. Up to now the procurement and construction of the SPES-3 components have concerned the load bearing structure, all the tanks simulating the containment, the EBT, the RWST pools and the EHRS. Furthermore some component replacement of existing auxiliary systems have been performed, such as the 130 kV transformer, switches, disconnecters and related devices of the electrical sub-station, the power line of fuel rods alimentation. The procurement and construction of other components, namely the power channel, piping, instrumentation, data acquisition and control system will be subjected to the funds made available each year by the MSE R&D programme devoted to innovative reactor concepts and related fuel cycles.
3. SPES-3 simulation capabilities of advanced SMR The SPES-3 facility is a simulator of IRIS, originally designed to carry out experimental campaigns in support of the IRIS design certification and to answer specific issues, such as the coupling between primary and containment during a SBLOCA. It was designed with a scaling factor of 1:100 in order to have a significant representation of thermal hydraulics phenomena and to allow also separate-effect tests. The scaling of the facility was performed with “power-to-volume” scaling method and verified with the “Fractional Scaling Analysis” for reducing the design distortions introduced in the conceptual design phase. The advanced Small Modular Reactor designs (integral type, PWR based concepts), such as NuScale, mPower and W-SMR, share a common set of design principles adopted also for the IRIS reactor to enhance plant safety and robustness: incorporation of primary system components into a single vessel, increased relative inventory in the primary reactor vessel, more effective heat removal, increased relative pressurizer volume, vessel and component layout that facilitate natural convection cooling for core and vessel [9. From this viewpoint, the SPES-3 facility has also important capabilities to offer a critical insight especially for the innovative solutions adopted in the SMRs under design in USA, such as the sequence of intervention of safety systems or the heat transfer between containment and the outer pool after a LOCA event, besides the opportunity to validate the best-estimate stand-alone or coupled thermal hydraulic system codes concerning the containment-primary coupling during
51 of 96 a loss of coolant accident, phenomenology common to the above-mentioned reactors. Thanks to the great scaling factor, the SPES-3 facility could be exploited to perform separate-effect tests in support of the development of these reactors, such as steam generators, pumps, two-phase flow instrumentation, in-pool heat exchangers, etc. Considering that IRIS and the above-mentioned SMRs have in common particular features: the same know-how on innovation concepts and technologic advances on AP1000® which the W-SMR is based on, helical coil steam generators and in pool heat rejection in NuScale reactor, etc., SPES-3 can be used as an integral facility, with proper specific design modifications, for such reactor testing and provide qualified experimental data for system code benchmark.
4. Conclusions This paper reports the description of the SPES-3 facility under construction at the SPES3 laboratories in Piacenza (Italy), its features and performance, besides dealing with the status of its realization. It also highlights the capability in simulating specific features and phenomena occurring in other small modular reactors under development or realization in the world. In particular, the SPES-3 facility could be utilized to carried out: - integral tests in support to the development of SMR with common features with IRIS; - separate-effect tests to investigate specific phenomena in support to the component development of the SMR under development or realization worldwide. Moreover, even if SPES-3 facility is IRIS reactor design oriented, the experimental data could be used for the qualification of best-estimate thermal hydraulic system codes (RELAP5, TRACE, CATHARE, GOTHIC, etc.) on physical phenomena common to many SMRs, like primary-to-containment system coupling, natural circulation, in-pool heat transfer, etc. The completion of the SPES3 construction could be accelerated in case the Italian government funds were sided with private funds of companies or organizations interested in new SMR reactor development. SIET wishes to complete the facility construction in the interest of the whole nuclear community.
5. References [1] E. Young et al.: The NuScale Advanced Passive Safety Design, Proc. of ASME 2011 Small Modular Reactors Symposium SMR2011, Sept. 28-30, 2011, Washington, DC, USA, paper N° SMR2011-6658 [2] J.A. Halfinger et al.: The B&W mPowerTM Scalable, Practical Nuclear Reactor Design, Nuclear Technology. Vol.178, N° 2, pages 164-169, May 2012 [3] R.J. Fetterman et al.: An overview of the Westinghouse Small Modular Reactor, Proc. of ASME 2011 Small Modular Reactors Symposium SMR2011, Sept. 28-30, 2011, Washington, DC, USA, paper N° SMR2011-6597 [4] Himénez M. et al.: Self pressurization and natural circulation: RELAP assessment with experimental data for CAREM CAPCN. Proc. IAEA-CRP on nat. circ. phenomena, modelling and reliability of passive systems that utilize natural circulation. Vienna 2007. [5] M.D. Carelli et al.: The Design and Safety Features of the IRIS Reactor, Nucl. Eng. Design, 230, pp. 151-167 (2004) [6] Ferri, et al.: SPES3 facility and IRIS reactor numerical simulations for the SPES3 final design, ENC 2010 European Nuclear Conference, Barcelona, Spain, May 30 – June 2, 2010 [7] M. Greco et al.: Two-phase flow measurements studies for the SPES3 integral test facility for IRIS reactor simulation, Proc. of the 18th International Conference on Nuclear Engineering ICONE18, 17-21 May, 2010, Xi'an, China [8] S. Gandolfi, et al.: SPES3 - Two-phase mass flow measurements: technical specifications, report RDS/2010/69, available on www.enea.it [9] D.T. Ingersoll: An overview of the Safety Case for Small Modular Reactors, Proc. of ASME 2011 Small Modular Reactors Symposium SMR2011, Sept. 28-30, 2011, Washington, DC, USA, paper N° SMR2011-6586
52 of 96
A REVIEW OF THREE FAMILIES OF SMALL MODULAR REACTORS (SMRS): LAND-BASED; FLOATING; IMMERSED
D. VIGNON ; M. LECOMTE; J.Y BEON; J.M POIMBOEUF NucAdvisor – 168/172 boulevard de Verdun - 92408 Courbevoie - Cedex France
ABSTRACT
Small Modular Reactors (SMRs) have drawn a continued interest from the nuclear industry, since its inception. None have been put in commercial operation so far, as their generation costs have never been competitive with larger reactors. However SMRs have recently regained interest, and several projects are actively developed around the world, with national government funding supports. Most of these projects are land based; a few projects, namely the Rosatom barge (Russia) fitted with two KLT-40S reactors , and the Flexblue immersed concept developed by the French manufacturer DCNS do not follow the main stream, and are truly “transportable”. A market analysis shows the strong potential of all SMR types, provided their generation price is less than 100€/MWh. Furthermore, transportable reactors have additional merits, as they allow the implementation of new business models, where investment, generation, and operation, are truly decoupled. Finally, immersed reactors can benefit of an infinite resource of water, both for cooling and shielding, and allow technical breakthroughs, which may change the nuclear landscape.
1. Introduction When analyzing the PRIS data base maintained by the IAEA [1], it seems at first glance that there is a regular distribution of the 4371 operational reactors, from very low outputs (12MW for the four Bilibino units on the Artic ocean, to much larger sizes up to 1500 MWe.
Size of Operational reactors
2000 1500 1000 500 0 Electrical Electrical output MW
Figure 1 - Size of operational reactors – All countries and types included
But this global view miss the important fact, which is the steady increase of unit size over years.
1 December 2011
53 of 96 1800
1600
1400
1200
1000
800 Power (Mwe) 600
400
200
0 17/7/69 7/1/75 29/6/8020/12/8512/6/91 2/12/96 25/5/0215/11/07 Commissioning date
Figure 2 - Size vs. commissioning date (Western type reactors – PWR and BWR only)
Some atypical reactors, built without much economical consideration, are worth mentioning: the four Bilibino reactors, still operating on the Baltic sea, with a unit size of 12 MWe Eighteen Indian reactors2, still operating around the country, with a unit size of 200 MWe
The unit power of the 294 Western type operating reactors3 has steadily increased over time from an average of 800 MWe in the late sixties, to 1500 MWe and more nowadays. Both PWR and BWR technologies faced the same trend.
Number of units Average output PWR 215 978 MWe BWR 79 947 MWe Total 294 970 MWe Table 1 - Average output per technology
Reasons for this trend are overwhelming and well known: economy of scale which governs all energy projects4 is an incentive for larger units. It is all the more true for nuclear projects, since nuclear safety costs are essentially fixed, or slowly variable with size (cost of the safety infrastructure (licensing body, waste agency; cost of security (fences, guards); cost of safety features (containment, safeguard systems), , etc.). And nuclear vendors offer very large units (more than 1700 MWE for the GE ESBWR).
2. A renewed interest for small and medium reactors
This trend lets many countries or operators out of the nuclear landscape. Their finance are not robust enough, or their grid is too small to consider nuclear as part of their energy mix.
2 16 heavy water reactors and 2 boiling water reactors 3 Gas cooled, and Heavy Water reactors are excluded from the chart 4 Windmills follow the same pattern ; maximum size was in the range of 700 kWe some fifteen years ago; 4 MWe or more is a current standard nowadays
54 of 96 And finally, nuclear technology gets into a deadlock: new reactors are safer and safer; their Leverage Cost of Electricity (LCOE) keeps being competitive with alternate sources, but only when considering larger units5 that most of the operators cannot afford.
Tab.1 - SMRs development projects (2011)
Power Technology Credib Acronym Type Country Comments (Mwe) provider ility CNEA & CAREM 27 PWR ARG - INVAP INET & HTR-PM 105 HTR CHN - Huaneng Collaboration Eskom, South PBMR 165 HTR ZAF - Africa, AREVA – TA 100 to 200 Technological and GT - SMR PWR DCNS—EDF - FRA + MWe market assesment CEA FlexBlue 160 MWe PWR DCNS France + MRX 30-100 PWR JAERI, Japan JPN - ITHMSO, FUJI 100 MSR Japan-Russia- JPN - USA KAERI, S. Generic PSAR SMART 100 PWR KOR ++ Korea submitted in 2012 Initially Internatio IRIS-100 100 PWR Westinghouse- - W pulled out in 2011 nal led SMR 225 225 PWR Westinghouse USA + Oregon All four designs are NuScale 45 PWR USA - University candidate to the Babcock & DOE sponsored mPower 180 PWR USA + Wilcox program SMR-160 160 PWR HOLTEC Intl + G.Atomics + GT-MHR 280 HTR USA - Minatom BREST 300 LMFBR RDIPE RUS -
Atomenergopr VK-300 300 BWR RUS - oekt KLT-40 35 PWR OKBM, Russia RUS ++ Under construction Therefore, and whatever the trend for large size reactors, there is a growing interest for Small Modular Reactors (SMRs) targeting smaller countries and smaller grids. Their technical breakthroughs and modularity are expected to offset the disadvantage of a lower scale. In 2006 according to the International Atomic Energy Agency (IAEA), more than fifty innovative SMR concepts and designs have been, or were being, developed by national or international programs involving Argentina, Brazil, China, Croatia, France, India, Indonesia, Italy, Japan, Republic of Korea, Lithuania, Morocco, Russian Federation, South Africa, Turkey, USA, and Vietnam [2, 3]. Innovative SMRs are under development for all principal reactor lines and for some non-conventional combinations. The target dates when they would
5 Even the Westinghouse AP series followed the same trend. It started with a promising 600 MWe model (AP 600). The present AP 1000 output is above 1100 MWe and has nearly doubled.
55 of 96 be ready for deployment range from 2010 to 2030. The main developments which are underway are listed in Table 16. In this paper, we will focus at small and medium size reactors, from ~50 MWe to 300 MWe, and more specifically consider the following developments, which are the most mature:
The KLT-40S developed by the OKBM institute in Russia. Its Nuclear Steam Supply System is based on ice breaker technologies, with the consideration of passive systems for safety functions. The concept calls for mounting the reactor and auxiliary systems on a floating barge, allowing full shipyard construction. The SMART project developed by KAERI in Korea, which generic safety analysis report was filed to the Korean Safety Authority. According to Korea officials, construction of a first-of-a kind could start in the near future The four designs which are competing in the US for getting a DOE grant. Indeed, at the beginning of 2012, the US Department of Energy announced that the government plans to fund of a program that will pay up to half the cost of developing and licensing up to two SMR designs. A total of $452 million would be allocated over five years to the project. Projects should be commercial by 2022. Four companies in the US announced that they applied for benefitting of this program : Babcock & Wilcox, headquartered in North Carolina, is developing a 180-MW concept called mPower; Pennsylvania-based Westinghouse is designing a 225-MW unit named Westinghouse SMR; Oregon startup NuScale Power (initially affiliated to the Oregon University) is the developer of the 45-MW NuScale; and Florida firm Holtec International is developing a 160-MW reactor concept called SMR-160. Two winners of this program should be announced by the US-DOE before end of 2012. The FlexBlue concept developed by the French company DCNS. It features a civil nuclear reactor, and turbo-generator installed in a submersed vessel. Based on proven technologies, it should be marketable at the beginning of the next decade. After a quick review of the potential market, a comparison of the three main concepts (floating barge, stationary reactor, submersed) will be sketched.
3. A promising market According to the 2010 outlook of The Energy Information Administration (EIA), a bipartisan U.S. Agency controlled by the Congress, global energy demand will continue to grow at least until 2030, with an overall rate of about 1.6% a year. However, this growth will be highly uneven between the different regions: 0.5% per year in developed countries (OECD), and more than 2% per year in developing countries where access to energy is still very limited. Nuclear energy should also continue to grow, at a rate of 1% per year in OECD countries, and more than 4% per year in developing countries where this industry is at an early stage; overall, the average growth of nuclear power would then be about 2.4% per year. As a result, approximately 10 GWe should be placed on the network each year, which represents from six to ten large size commercial reactors. To asses the potential of either to floating or submersed reactors, NucAdvisor performed a specific study of coastal countries having a too small grid to afford large size reactors; however, only grids with an installed capacity of 1500 MWe or more were considered. Excluding major networks eligible for high-power nuclear power plants, over fifty coastal countries have been identified. The total installed capacity of these coastal Countries is over 400GWe and according to their growth rate, their total installed capacity should exceed 1,000 GWe in 2030.
Currently and still for long the cost of electricity on networks is benchmarked by gas turbines, and is around 80 € / MWh. However the long-term trend is strongly upward, and this value could double in the next twenty years, resulting in an increase of 60% of the electricity price
6 Some promising developments based on new technologies are not listed, because their basic feasibility is not yet proven (Hyperion, Bill Gates foundation development, etc.)
56 of 96 on these networks7. This suggests that nuclear power plants of small size (about 150 MWe) which could operate below 100 € / MWh (all costs included) would be significantly competitive. Assuming that the nuclear market share is 20% in the group of countries identified in the study (as the average of major developed countries), the market could be of 200 GWe SMRs in twenty years, that is more than 60 units 100/200 MWe per year.While this study focused coastal countries, it should be remembered that all in all, coastal regions represent 70% of World GNP. Adding inland sites should multiply these figures by ~1,4. Furthermore, small or medium size reactors should also be considered as candidates to supply electricity to larger networks, when either electricity production is fragmented between multiple operators, or aging small size fossil units have to be replaced. NucAdvisor also performed a systematic market survey of all world countries having presently an installed capacity higher than 800 MWe. The aim of the study was to figure out the nuclear market potential considering the size of the country, its per capita GNP, its growth, financial stability, political outlook, etc. Essentially this study confirmed the above figures. Specific results for year 2030 are provided in Table 3. Therefore and with a high degree of confidence, it can be stated that a large market exists, even at higher generation costs than targeted by large fossil plants.
Figure 3 - SMRs capacity in 2030 4. Common technical features and specificities of SMR lines Small Modular reactors have some key common features in common. They can easily accommodate passive safety features, with strengthened defense in depth; and they can be built under several business model, not accessible to large plants. Furthermore, transportable nuclear power plants have additional benefits, due to the mere fact that they can be rather easily moved. 4.1. Similar technical features The specific size of small reactors allow to adopt simpler designs, and improve the overall safety level. Reviewing the key features adopted for the traditional three safety barriers, some general observations can be made: They all feature standard PWR fuels (first barrier), sometimes with lowered performances (temperature, linear flux, etc.) which should allow improved performances. All aim long fuel cycles (at least two years).
7 Gas prices are subject to huge regional variations, and this short summary cannot provide the detailed analysis NucAdvisor performed on this subject.
57 of 96 Cooling of the first barrier in case of an accident is made by mean of passive systems in all considered designs. Safeguard systems make use of concepts developped for W AP600/AP1000 designs (systematic depressurization in case of loss of coolant accidents). The low size of SMRs, and the benefit from an improved surface to volume ratio makes such approach extremely effective/ Some designs (NuScale ; mPower) also eliminate primary pumps and rely on natural convection in normal operation. Several designs (SMART, Nuscale, mPower, W –SMR) consider integrated reactor concepts, eliminating primary loops, and therefore reducing the LOCA risk and improving the robustness of the second barrier. All stationary designs feature underground reactors, improving the protection of the third barrier. Furthermore, some designs feature a direct cooling of the containment, continuously immersed in water (Nuscale), improving the robustness of the third barrier. In total, small reactors benefit from an overall safety level which seems to be higher than reached by large reactors.
4.2. Multi-unit operation Operating costs of a nuclear unit, which has to be permanently manned, may be high when their size is quite small. To offset this factor, in the NusCale concept (45 MWe output), twelve modules, grouped in three clusters of four, would be controlled from one single control room. While the licensing of such approach remains to be done, its positive result could be used even at larger sizes.
4.3. A new licensing approach The cost of licensing small reactors would place an unacceptable burden on their competitiveness if licensing rules would remain unchanged. In particular, under the present scheme of things, each country is responsible for licensing. Clearly, such concept which clarifies the responsibility line, should be kept. However, the standardization of SMRs would allow to make benefit of a standard licensing package; and only changes, and adaptation to local conditions should be relicensed for a new project. Such approach should allow the possibility for newcomers to mobilize the expertise from countries having already licensed an SMRS reactor, and to subcontract the licensing expertise.
5. Small Modular Reactors: an avenue to new business models Small Modular Reactors are often blamed for their potentially high capital costs. While this might be true if SMRs were designed as downsized large reactors, this critic does not consider the new, simplified design features which can be implemented in smaller reactors, with an easier, more efficient implementation of passive systems. Small Modular Reactors share some common features, which allow envisaging completely new paths to the market, which large reactors cannot dream of. More specifically, they mobilize less capital, and they provide much more certainty to investors: construction time should be drastically cut (two years being an accessible target); complete standardization should result in greatly reduced licensing risks (ref. §4.3), and high capacity factors through sharing of operating feedback and maintenance expertise; setting operating companies specialized in operating a certain type of reactors would allow uncoupling the roles of operator, investor, and power producer. These factors, combined with a lower cost of capital (due to a higher certainty in nuclear projects) will allow meeting the generation price target of 100 €/MWh, as shown in Figure 4.
58 of 96
Figure 4 LCOE comparison between large and small reactors (first of a kind)
Large commercial reactors can only be owned and operated by giant utilities, having the strength to take the risk, and a balance sheet large enough to convince banks that they are trustable borrowers. SMRs on the opposite will be considered by investors like gas turbines: independent from the site, bought “off the shelf”, and built in series and sold under flexible schemes: Independent Power Producer; Built Own and Operate under a concession scheme; Built Own and Transfer, etc. All in all, SMRs have the potential to make the nuclear world – presently a nightmare for investors – comparable to the rest of the generation business.
6. FlexBlue: the quintessence of Small Modular Reactors FlexBlue combines all main merits of Small Modular Reactors. With a power output of ~160 MWe, Its module (~150 m long and 14 m diameter) is factory built, by assembling skid mounted modules. Therefore, quality, and construction time can be more easily controlled. It is easily transportable and made available where and when needed. Being submersed into the sea, FlexBlue benefits from an unlimited cooling capacity. In operation, and in accidental conditions (even severe accidents), the shell, acting as a containment, would remain leak tight. When considering severe accidents as occurred at Chernobyl or Fukushima, the main environmental consequences were the atmospheric and land pollution from the transfer of Iodine and Cesium. With a submersed reactor, such transfer of contamination to human beings is eradicated. FlexBlue is unmanned, except for periodic visits, which should result in reduced operating costs, and in a high reliability. Seawater allows to easily implement and easily operate passive safety systems capable of removing residual heat both from the secondary and primary system, increasing defense in depth.
59 of 96
Figure 5 - The FlexBlue concept
Combining essentially all merits of small modular reactors, from the technical, safety and business standpoint, FlexBlue has the potential for a major breakthrough in the nuclear world.
60 of 96
REFERENCES
[1] The Power Reactor Information system (PRIS) data base is accessible at www.iaea.com
[2] INTERNATIONAL ATOMIC ENERGY AGENCY, Innovative Small and Medium Sized Reactors: Design Features, Safety Approaches and R&D Trends, Final report of a technical meeting held in Vienna, 7-11 June 2004, IAEA-TECDOC-1451, Vienna (2005).
[3] INTERNATIONAL ATOMIC ENERGY AGENCY, Status of Innovative Small and Medium Sized Reactor Designs 2005: Reactors with Conventional Refuelling Schemes, IAEA- TECDOC-1485, Vienna (2006)
61 of 96 PRELIMINARY EVALUATION OF A SEVERE FLOODING EFFECTS ON AN INNOVATIVE SMR
R. LO FRANO, V. BAUDANZA, G. FORASASSI Department of Mechanical, Nuclear and Production Engineering (DIMNP), University of Pisa L. Lucio Lazzarino 2, 56126, Italy
ABSTRACT
The aim of the present study is to evaluate the structural effects induced by a severe flooding/tsunami on a SMR outer containment (e.g. the one of Westinghouse proposal) in agreement with the “stress tests” suggestions, presently foreseen by the European and International Associations and on basis of the lesson learnt from Fukushima accident. Therefore to evaluate the consequences induced by the breaking and splashing water waves on a SMR containment building, non linear analyses (by using adequate FEM codes) were performed simulating the water impact forces in terms of pressure, calculated on the basis of the water velocity and flow depth (according to the ASCE rules). The obtained results highlighted the tsunami effects are mainly dependent on the wave elevation, flow velocity, hydrodynamic phenomena and geometry of the reference structures. Finally they showed that the considered containment building seemed capable to withstand the flooding/tsunami applied forces without unbearable loss of the integrity of SMR containment structure(in the assumed geometries and material characteristics).
1. Introduction
The severe accident occurred at the Fukushima Daiichi plants in March 2011, induced by the 9 Richter magnitude earthquake and 14 m tsunami, has attracted the worldwide interest (and concern) due to the observed dramatic consequences (e.g. large core melt down and rather radioactivity releases) and extensive damages suffered by the reactors’ structures and important components. The lessons learnt from Fukushima highlighted that the reactors proved to be seismically robust but vulnerable to the tsunami, that makes the plant isolated and inaccessible for long time. The tsunami waves overtopped the plant seawalls and damaged diesel backup power systems, leading to severe problems including three large explosions and radioactive leakage (Figs.1).
a) b) c)
Figs. 1: Flooding effects in the Fukushima plants.
Therefore on the base of the observed consequences, it is comprehensible the reason why the safety margin of existing/future NPPs, like the small-medium reactors (SMR) [1-2], subjected to an extreme natural (whose intensity may exceed that one considered in the design stage) event like the tsunami/flooding should be evaluated.
62 of 96 In the present study the safety margin of a SMR was evaluated assuming that the safety systems were not operating as well as that the plant was isolated and in station black-out conditions: in particular the performances of the outer containment building, which is the last safety barrier capable to withstand a severe flooding or tsunami consequences, were analyzed by adopting a deterministic approach (in agreement with what suggested by the WENRA in the safety track of the stress tests [3]). To the purpose the water waves impact hydrostatic and hydrodynamic forces on the walls of nuclear and non-nuclear buildings were adequately considered in order to evaluate the effects induced by the mentioned flooding/tsunami event. The methodology used in the present study, as clearly described in what follows, is in agreement with the international requirements codified in the rules ASCE-USA.
2. Description of tsunami/flooding phenomenon
The tsunami, as it is well known, is usually originated by an undersea earthquake: the sudden shift and vertical displacement of water determine the arising of waves that initially propagate from offshore to the coast (propagation phase). During this propagation they become shorter and, due to the variation of sea bed depth, may grow in height amplitude (run up phase characterized by the elevation of the wave). Afterwards the water waves may reach the shore line initiating the inland inundation phase (Fig. 2). In this phase “train wave” may be characterized by periods ranging from minutes to hours and heights that could reach tens of metres [4-5].
Fig. 2: Representation of the tsunami phases [3].
It is important to note that the inland propagation of water waves may also include land draining off dislocation/transport of all type of mobile and floating objects with it, such as boats, cars etc., that could impact and damage to a various extent the infrastructures even at several kilometres from the coast. In this study the attention was focused on the inundation phase during which the water waves, exerting hydrostatic, hydrodynamic and impact forces on the walls of nuclear buildings (the impact of debris, assumed like missiles, [4-6] was not considered in this study), may determine the most safety relevant damages.
2.1 Evaluation of wave height force
First of all it is important to stress that the tsunami is a hydraulic phenomenon of submersion characterized inland by a non-Newtonian flow with a heavy sediment charge [6]. The potential damage effects of tsunami/flooding are mainly determined by the front impact and lateral pushing of the first waves that are responsible of the smashing force of a water
63 of 96 wall traveling at high speed and of the destructive power of a large volume of water, even if the wave did not look large. According to Leone et al. [7], the tsunami phenomenon may be characterized by three basic processes from upstream (transit zone) to downstream (flood zone) and breaking of the wave (Fig. 3): each one described by physical parameters (magnitude criteria) that can be theoretically measured and used for modelling purposes. In particular as for as the wave height elevation concerned it is important to stress that to calculate it in a chosen site several information, related to the hydrodynamic processes, are necessary, such as the seabed vertical displacement, the wave period, the velocity, the ground elevation, the height during inland run-up phase, the water pressure, etc.
Fig. 3: Processes and associated damage modes (second wave) [7].
The available theoretical formulations and analytical solutions, for each of the mentioned phases characterizing the tsunami, are based on simplifying assumptions as described by various authors like Sharma et al., 2005, Cho et al., 2004. In addition the magnitude scales of Imamura (1949), Iida (1970), Soloviev (1970), Abe (1981), Hatori (1986) and Murty and Loomis (1980), based on maximal wave heights (Hmax in meter) on coast [8] were used. In agreement with the “stress tests” suggestions, presently foreseen by the European and International Associations and on the basis of the lesson learnt from Fukushima accident a deterministic approach based on the evaluation of the hydrodynamic loads, calculated according to the ASCE/SEI 7-10 rules [9], and by using FEM model was adopted.
2.2 Approach to tsunami analysis
According to the ASCE/SEI 7-10 [9] the methodological approach to be used to design buildings and structures, in area characterized by a high tsunami risk, is based on the evaluation of the design flood elevation (DFE) and of the hydrostatic and hydrodynamic loads induced by the flowing water. The water waves elevation or breaking wave height (Hb) may be calculated as:
Hb = 0.78 ds (1)
In Eq. 1, ds indicates the local still water depth that shall be calculated in agreement with the Eq. 5.4-3 of [9]. This values may be determined applying the following formula: ds = 0.65 (BFE – G) (2) where BFE is the assumed design flood elevation, while G is the ground elevation.
64 of 96 The hydrostatic force, that acts laterally on structures, may be determined by the standing water resulting from tsunami inundation or from the interaction between the water and structures. Instead the hydrodynamic one results from the rapidly moving water and its interaction with structures (impact of the waves on the building walls). The maximum pressure and breaking wave force, assuming a normally incident wave onto a rigid wall and for a limited depth in size, may be calculated as [9]:
Pmax = Cp∙γw∙ds+1.2∙γw∙ds (3)
2 2 Ft = 1.1∙Cp∙γw∙ds + 2.4∙γw∙ds (4)
In the above Eqs. 3 and 4, Pmax is the maximum waves pressures (dynamic and static); Ft is the net breaking wave force, Cp is the dynamic pressure coefficient, ranging from 1.6 to 3.5; γw is the weight of the water and ds the water depth at the base of building. This procedure assumes also that the vertical walls may cause a reflection of the waves against the water ward side and that the space behind them are dry (therefore with no fluid on the outside of the wall). In the case of inclined walls the horizontal component of the breaking wave force shall be calculated taking into account the vertical angle between non vertical surface and the horizontal one. Finally after having determined all the hydrostatic and hydrodynamic loads, the behaviour of a SMR containment building, like the IRIS one, was analyzed (adopting a suitable FEM code), taking into account adequately the geometrical and material non linearities as well as the reflection of waves effects.
3. Global response of a SMR outer containment
An example a SMR containment like the IRIS reactor one (having about 335 MWe that could be scaled down up to 100 MWe) [10] was considered in order to evaluate the flooding structural effects. This containment building (Fig. 4) has about 1 m thick walls and is partially embedded (-20 m below the ground) level.
Fig. 4: IRIS layout
The FEM models of the IRIS reactor building (RB), shown in Fig. 5, was implemented using 3-D solid and thick shell elements, adopting suitable material properties for the concrete and steel components and adequate initial and boundary conditions.
65 of 96 Structural effects induced by the wave of flooding/tsunami have been evaluated assuming different wave heights (BFE) and representing the input breaking waves loads, by means of the equivalent pressure values (calculated by means of the previous Eq. 3). The BFEs have been considered ranging from 5 m to about 25 m.
Fig. 5: FEM model.
All the preliminary non linear dynamic simulations were carried out setting up suitable boundary conditions and features capable to represent the progressive damaging and failure processes of the concrete. The assumed failure criteria were based on the maximum stress value. In addition each simulation was performed considering a transient duration of 0.5 s; this time interval (determined by means of sensitivity analyses) seemed to be sufficient to represent the effects of the pressure loads and is related to the global dynamic behaviour of the RB structure. The application directions of the water pressure are assumed orthogonal and lateral to the auxiliary buildings and outer containment walls, as represented in Fig. 6;
Fig. 6: Directions of the wave impact in terms of pressure.
66 of 96 4. Analysis of the obtained preliminary results
The obtained results, considering the made assumptions and SMR geometry and material properties adopted, highlighted in general that the stress level is function of the wave height: the stress level in the RB walls, of course, increases with BFE like the exerted hydrostatic and hydrodynamic pressure. In the following Figure 7 and 8, the obtained results are represented in terms of stress and displacement carried out in the case of wave height elevation equal to 10 m. The BFE equal to 10 m, without the breaking of water waves on the dome of the plant, showed that the stress values were not so relevant to determine an excessive deformation and damages of the RB walls (Fig. 7) and impair the integrity of the RB itself.
Fig. 7: Von Mises stress distribution for BFE=10 m.
Fig. 8: Overview of the horizontal displacement (x axis) distribution for BFE=10 m.
67 of 96 Moreover the obtained maximum localized displacement resulted equal to about 4 mm, as shown in previous Fig. 8, although the mean value was about 2.5 mm. The obtained results, in the case of BFE equal to about 25 m, that involves the beginning of the submersion of the dome, in terms of Von Mises stress (σmax≈ 23 MPa, like shown in Fig. 9 b) indicated that the concrete containment walls were suffering local damage and initial failure phenomena (such as cracking, etc.) due to the hydrodynamic and the hydrostatic pressure loads, although the stress mean value was lower (about 9MPa).
2,5E+07
2,0E+07
1,5E+07
1,0E+07
5,0E+06
0,0E+00
Equivalent Von[Pa] Mises stress 0,00 0,10 0,20 0,30 0,40 0,50 Time [s] Dome Cylindrical Wall Auxiliary Building
(a)
(b)
Figs. 9 a, b: Behaviour and distribution of the Von Mises stress for BFE=25 m.
In addition it is important to stress that further future study developments of the flooding effects, considering a more in depth analysis of the hydrodynamic behaviour of the water waves (e.g. the influence of buoyancy, the fluid-structure interaction due to the splashing
68 of 96 waves, etc.) seem necessary to correctly evaluate the performances of a outer containment building as resulted in the present preliminary study.
5. Conclusions
In this work has been preliminarily analyzed the dynamic behaviour of an SMR, with reference to the IRIS reactor (as an example) subjected to a severe flooding / tsunami event using a deterministic approach (by means of FEM simulations) in order to assess preliminary the safety margin of considered structure, according to the new and/or updated International requirements and in agreement with the lesson learnt from the Fukushima 2011 accident. In particular, the attention has been focused on the dynamic response of the outer containment system which is the last safety physically barrier capable to prevent radioactivity release to the external environment and population in the reference accident conditions. Structural effects induced by the water breaking waves have been preliminary analyzed considering also different wave height elevations and, of course, the resulting pressure values, which were calculated assuming a high risk site plant. In the performed simulations it was also assumed that the rear part of the reactor building exposed to the wave front remained “dry” (not immersed in water) when the inundation and the first waves impact occurred. Moreover in the set up FEM analyses failure criteria capable to represent the progressive damaging and failure processes of the concrete were assumed. The preliminary obtained results highlighted that the resulting stress values were higher in some localized part of outer containment, even if the stress mean value was lower than those ones: particularly in the case of wave height equal to about 25 m, the considered reactor building, with the assumed hypotheses and geometry, seemed to be undergoing local structural damage and progressive failure, even if in general the containment integrity seemed to be ensured. Further future developments are necessary to study in depth the influence of buoyancy and of the fluid-structure interaction on the hydrodynamic forces, arisen from the splashing and travelling waves, and subsequently on the pressure loads exerted on the outer containment building walls.
6. References
[1] Nuclear Energy Agency, Current Status, Technical Feasibility and Economics of Small Nuclear Reactors, OECD NEA, 2011. [2] ENSREG, Stress tests performed on European nuclear power plants, Stress Test Peer Review Board, 2012. [3] US NRC, “Tsunami Hazard Assessment at Nuclear Power Plant Sites in the United States of America “, Final Report, March 2009. [4] A.K. Ghosh, Assessment of earthquake-induced tsunami hazard at a power plant site, Nuclear Engineering and Design 238 (2008) 1743–1749. [5] A.K. Chopra, “Dynamics of Structures: Theory and Applications to Earthquake Engineering”, New Jersey (1995 [6] Chanson, H. (2005). Le Tsunami du 26 décembre 2004: un phénomène hydraulique d’ampleur internationale. Premiers constats. La Houille Blanche, Paris, 2, 25-32. [7] F. Leone et al., ” A spatial analysis of the December 26th, 2004 tsunami-induced damages: Lessons learned for a better risk assessment integrating buildings vulnerability”, Applied Geography 31 (2011) 363-375. [8] Papadopoulos, G., & Imamura, F (2001). A proposal for a new tsunami intensity scale. ITS 2001 Proceedings, Session 5, Number 5-1, 569-577. [9] ASCE, Minimum design loads for buildings and other structures, 2010; ASCE/SEI 7-10. [10] IAEA, “Status of Small and Medium Sized Reactor Designs” (2011).
69 of 96 EUROPEAN DESIGN STUDY ON SUPERCRITICAL WATER COOLED REACTORS
T. SCHULENBERG, A.G. CLASS IKET, Karlsruhe Institute of Technology 76021 Karlsruhe – Germany
J. STARFLINGER IKE, University of Stuttgart Pfaffenwaldring 31, 70569 Stuttgart – Germany
ABSTRACT
Since about 10 years, the long term future potential of water cooled reactors is being studied and assessed by the Generation IV International Forum. Aiming at a once through steam cycle with superheated steam at supercritical pressure, the nuclear power plant could be simplified significantly, while increasing its thermal efficiency compared with conventional light water reactors. In this context, a supercritical water cooled reactor, called the High Performance Light Water Reactor, with a core inlet pressure of 25 MPa and a core outlet temperature of 500°C has been developed by a European consortium, producing electric power of 1000 MW with an efficiency of around 44%. The size of the reactor pressure vessel is comparable with pressurized water reactors, but steam generators or reactor coolant pumps will no more be required. Instead, the superheated steam is fed directly to the steam turbine like in a boiling water reactor, and steam generators and dryers could be reduced to a small start-up system outside the containment. As a consequence, the containment size could be reduced significantly, still leaving enough volume for 2000 m3 of water, providing a heat sink for depressurization of the reactor and for pressure suppression inside the containment. Long term heat removal is foreseen with active and passive heat removal systems.
1. Introduction Following the trend of coal fired power plants in the last 20 years, the evolutionary development to higher temperatures and pressures, which are meanwhile exceeding even the critical pressure of water, has been considered by the Generation IV International Forum (GIF) as an option also for future water cooled reactors. The higher steam enthalpy could enable a direct, once through steam cycle such that neither steam generators nor steam separators and dryers would be required, and even primary coolant pumps could be omitted. Thus the coolant mass flow through the core is only driven by the feed-water pumps. Moreover, steam turbines and re-heaters could be significantly smaller than today, while the steam cycle efficiency would even be higher. As fossil fired power plants with supercritical steam conditions have been operated during the last 20 years, reaching 600°C live steam temperature or even more, this nuclear plant concept can benefit from proven design of turbines, feed-water pumps and most other components of the steam cycle except the re- heater. On the other hand, the containment design can basically be derived from the latest boiling water reactors.
In Europe, a consortium of 12 organizations from 8 European countries started in 2006 to address this challenge by working out a design concept of such a reactor, which they called the High Performance Light Water Reactor (HPLWR), with a core exit temperature of 500°C at a supercritical system pressure of around 25 MPa. The design objectives were a core with a thermal neutron spectrum, a net electric power of 1000 MW and a net plant efficiency of around 44%. Schulenberg and Starflinger [1] are summarizing results in a recent textbook. The following chapter shall provide an overview of what has been achieved
70 of 96 2. Reactor design These target data differ from conventional light water reactors not only by the higher pressure and core outlet temperature, but also by a significantly higher enthalpy rise in the core. Indeed, the difference between life steam enthalpy and feed-water enthalpy of 1936kJ/kg exceeds the one of pressurized water reactors by around a factor of 8. Assuming an overall hot channel factor of 2 between the peak and the average coolant heat-up, this enthalpy rise would result in peak coolant temperatures of 1200°C in a conventional core design, which is far beyond the target temperature limit. A strategy to overcome this issue can be learned from fossil fired boiler design. These boilers are characterized by multiple heat-up steps with intensive coolant mixing between them to eliminate hot streaks.
Closure Head Control rod extensions
Bolt
Control rod connection tubes Nut
O-ring sealing
Spring Cold leg
Control rod guide tubes
Hot leg Backflow limiter
Steam plenum
Outlet pipe assembly
Hot steam pipe
Reactor pressure vessel Fuel assembly cluster: Evaporator
Core barrel
Fuel assembly cluster: Super heater 1 Steel reflector
Fuel assembly cluster: Super heater 2
Core support/base plate
Lower mixing plenum
Fig. 1: Reactor pressure vessel and core components
The core layout of the HPLWR, as shown in Fig. 1, is based on such a strategy. It includes a first heat-up of the coolant as moderator water between the fuel assembly boxes or inside water boxes, comparable with the economizer of a fossil fired boiler. The second heat-up should be in the evaporator assemblies in the centre of the core, followed by coolant mixing
71 of 96 in a plenum above the core. From there, the coolant is directed downwards in assemblies of the first superheater, surrounding the evaporator, to be mixed again in an annular chamber underneath the core. Final heat-up to the envisaged core outlet temperature of 500°C is foreseen in a second superheater stage with upward flow again in assemblies at the core periphery. Assuming a hot channel factor of 2 for each heat-up step as a target number, the power ratio of evaporator to superheater 1 to superheater 2 should be around 4:2:1 to reach the same peak coolant temperature of 600°C in each region. The proposed core layout is trying to reach this power ratio by placing the second superheater at the core periphery where the neutron leakage is reducing the neutron flux anyway.
The fuel assemblies are designed with 40 fuel pins each and with a single moderator box in their centre enabling a small wall thickness of moderator and assembly boxes. To ease handling during maintenance, 9 of these assemblies are grouped each to a cluster with common head and foot piece. Wire wraps were proposed as grid spacers to improve coolant mixing in both flow directions. The cluster can be disassembled at its foot piece to exchange single fuel rods for repair. Control rods shall be inserted from the top of the core. They run inside 5 of the 9 moderator boxes of each cluster.
Automatic depressuri- 4 upper zation pools system
Containment isolation Emergency valves condenser
Pressure Pressure suppression suppression tubes pool
Low pressure coolant injection and RHR system
Fig. 2: Containment and safety systems
72 of 96 3. Containment and safety systems Even though the HPLWR plant concept looks similar to a boiling water reactor (BWR), at a first glance, it differs fundamentally by the missing recirculation pumps. Whereas a control of water inventory in the reactor pressure vessel is sufficient for the BWR to ensure the residual heat removal even in case of severe accidents, a continuous coolant mass flow rate through the reactor is required for this once through steam cycle as there is no closed coolant loop inside the reactor. This can be achieved either with redundant feed-water pumps or by depressurization of the pressure vessel such that the residual heat is removed by vaporization. For the unlikely case of a severe accident, these functions must also be provided inside the containment. With this respect, most safety systems of the containment can indeed be derived from latest BWR containment concepts, with the exception of passive flooding and emergency condensers for reasons explained above.
A design proposal for the containment is shown in Fig. 2. It is made from reinforced concrete, equipped with an inner steel liner and a pressure suppression system. The design pressure of the containment is considered to be in the range of about 0.3 to 0.4 MPa. Containment isolation valves for each of the 4 feed-water and steam lines, inside and outside of the containment, close automatically in case of a feed-water or steam line break inside or outside the containment. The reactor is scrammed and the depressurization valves release steam through 8 spargers into 4 upper pools, removing the residual heat until at least one of the 4 redundant, active low pressure coolant injection pumps in the basement of the containment becomes available. In case of a steam line break inside the containment, any pressure increase by steam release is limited by a large pressure suppression pool in the lower half of the containment into which 16 open pressure suppression tubes are running. Long term passive residual heat removal (RHR) from the containment can also be provided by containment condensers to the spent fuel pool above the containment.
4. Steam cycle components Fig. 3 shows the design concept of the HPLWR steam cycle. Superheated live steam leaves the reactor with a temperature of 500°C at 24 MPa pressure and 1179 kg/s mass flow rate. Before it enters the high pressure (HP) turbine, some live steam is extracted to reheat steam in the counter-current reheater. Most of the steam (82.2 % of the total mass flow rate) is expanded through the HP turbine and reaches the shell side of the reheater with a temperature of 260.2°C at 4.25 MPa pressure and, due to the steam extractions in the HP turbine, with 824 kg/s mass flow rate. There it is reheated with the live steam from the reactor (494°C; 22.6 MPa; 209.5 kg/s) to 441°C, before it is expanded in the intermediate pressure (IP) and low pressure (LP) turbines to 32.9°C at 5 kPa pressure with a steam quality of 0.87. The technology of the turbines is based on the turbines of supercritical fossil-fired power plants. Like there, full speed turbines and generator can be applied for the HPLWR concept.
The steam cycle is operated at a fixed pressure of 25 MPa at core inlet. A sliding pressure start-up with two-phase flow at sub-critical pressure has been avoided because of the risk of high fuel cladding temperatures under post dryout conditions. Instead, a constant pressure operation from start-up to full load has been foreseen. If the thermal power of the core is decreased below 50% load, the mass flow rate is kept constant to avoid flow reversal in the three pass core, and the reactor outlet temperature decreases, which would cause the steam to expand into the two-phase region of the HP turbine with the risk of erosion of the HP turbine blades. Therefore, the steam turbines are disconnected in this load range below 50% and the steam is expanded instead in a combined start-up and shut down system, formed by a battery of steam separators and dryers in the turbine building.
73 of 96 Turbine building Reheater HP Turbine IP Turbine LP Turbine Generator
Preheater Start-up system Drain tank and pump Feedwater pumps Feedwater tank
Fig. 3: Steam cycle components inside the turbine building
5. Status on current research and development To avoid misunderstandings, we have to mention here that this power plant will never be built. It was intended to serve as an example of the future potential of water cooled reactors. The design study was performed by around 50 students of European universities, guided by senior experts of nuclear engineering, who were highly motivated to design such a brand new concept from the very beginning. We are proud to say today that more than 90% of these students joined the nuclear business afterwards.
As a follow-up project, a small scale fuel assembly with just 4 fuel rods, but operated at 25 MPa and at the maximum linear heat rate of the HPLWR of 39 kW/m is planned to be installed and operated inside a pressure tube, replacing an ordinary fuel assembly of the research reactor LVR-15 in Řež, Czech Republic. Design and licensing of this in-pile test is just being performed in the European project SCWR-FQT. The young European consortium is cooperating with a Chinese consortium as a joint training project on nuclear engineering.
6. Acknowledgement This work has been funded by the European Commission as part of their project HPLWR- Phase 2, contract number 036230.
7. References T. Schulenberg, J. Starflinger, High Performance Light Water Reactor – Design and Analyses, KIT Scientific Publishing, Karlsruhe 2012, ISBN 978-3-86644-817-9
74 of 96 `
HEAT-TRANSFER CORRELATIONS FOR SUPERCRITICAL WATER AND CO2 WITH UPWARD FLOW IN VERTICAL BARE TUBES
IGOR PIORO SARAH MOKRY SAHIL GUPTA
EUGENE SALTANOV
Faculty of Energy Systems and Nuclear Science University of Ontario Institute of Technology 2000 Simcoe Str. N., Oshawa ON L1H 7K4 Canada E-mail: [email protected]
Keywords: Forced Convective Heat Transfer, Supercritical Pressures
ABSTRACT
This paper presents an analysis of new heat-transfer correlations developed for supercritical water and carbon dioxide flowing upward in vertical bare tubes. A large dataset within conditions similar to those of SuperCritical Water-cooled nuclear Reactors (SCWRs) was obtained from the Institute for Physics and Power Engineering (Obninsk, Russia). Previous studies have shown that existing correlations, such as the Dittus-Boelter, Bishop et al., and Jackson, deviate significantly from experimental Heat Transfer Coefficient (HTC) values, especially, within the pseudocritical range. The Swenson et al. correlation provided a relatively better fit for the experimental data, as compared to the previous three correlations within some flow conditions, but deviates from data within other conditions. Also, HTC and wall temperature values calculated with the FLUENT CFD code might deviate significantly from the experimental data, for example, the k-ε model (wall function). However, the k-ε model (low Reynolds numbers) shows better fit within some flow conditions. Therefore, new empirical correlations based on two approaches in terms of the characteristic temperature: 1) Bulk-fluid temperature and 2) Wall temperature; were developed. Analysis of uncertainties for these correlations showed that calculated wall temperatures were within ±10% and HTC values were within ±25% for the analyzed dataset. These correlations were also compared against data from other datasets, and it was found that they are the most accurate correlations compared to many other correlations developed so far. The proposed correlations can be used: (1) for calculations of supercritical-water heat-transfer in heat exchangers; (2) for a preliminary heat-transfer calculations in SCWR fuel channels as a conservative approach; (3) for future comparisons with other independent datasets and with bundled data; (4) for the verification of computer codes on thermalhydraulics; and (5) for the verification of scaling parameters between water and modeling fluids (CO2, refrigerants, etc.). However, when these correlations were applied to various SC-carbon-dioxide datasets it was found that they cannot predict these data with the same accuracy as the supercritical-water data. Therefore, a new heat-transfer correlation based on wall-temperature approach was developed for supercritical CO2 data obtained by I. Pioro. Analysis of uncertainties for this correlation showed that calculated wall temperatures were within ±20% and HTC values - within ±30% for the analyzed dataset.
1. Introduction
In the 1950s, the idea of using supercritical water appeared to be rather attractive for steam generators/turbines in the thermal-power industry. The objective was to increase the total thermal efficiency of coal-fired power plants. Work in this area was mainly performed in the former USSR and in the USA in the 1950s – 1980s. Within the same timeframe SuperCritical Water-cooled nuclear Reactor (SCWR) concept were also developed. However, due to some difficulties these ideas were abandoned till 1990’s. Currently, several countries are working to develop SCWRs, which will have significantly higher thermal efficiencies (up to 45 - 50%) compared to current fleet of water-cooled reactors with thermal
75 of 96 ` efficiencies up to 36 – 38% [1, 2].
2. Thermophysical properties of supercritical fluids
In support of further developing SuperCritical Fluids (SCFs) applications, heat-transfer analysis at supercritical conditions is very important. However, at supercritical conditions the task of calculating HTC is very complicated, because heat transfer is influenced by significant changes in thermophysical properties at these conditions [1-4]. SCFs have unique properties. Beyond the critical point, the fluid resembles a dense gas. Figure 1 shows transition of water and CO2, respectively, through various phases as the temperature and pressure are increased. At critical and supercritical pressures the transition from a single- phase liquid to a single-phase gas does not involve a distinct phase change at these conditions. Phenomena such as dryout (or critical heat flux) are therefore not relevant]. However, at supercritical conditions, deteriorated heat transfer, i.e., lower Heat Transfer Coefficient (HTC) values compared to those at normal or regular heat transfer may exist.
General trends of various properties near the critical and pseudocritical points can be illustrated on a basis of those of water and carbon dioxide . Thermophysical properties of water and another 105 fluids and gases at various pressures and temperatures, including critical and supercritical regions, can be calculated using the NIST REFPROP software [5].
Figures 2-7 show thermophysical-properties profiles vs. reduced temperature (i.e., Tr = T / Tcr) for 3 fluids: water (25 MPa), CO2 (8.4 MPa (equivalent of 25 MPa water pressure)) and R-134a (4.6 MPa (equivalent of 25 MPa water pressure)), for reference purposes. Properties of water are taken at 25 MPa - the most common pressure for supercritical steam generators and turbines and the proposed operating pressure of SCWRs. Pressures for CO2 and R-134a, as modelling fluids, were scaled through the reduced pressure.
In general, an analysis of graphs in Figs. 2-7 shows that trends in properties profiles for all 3 fluids are quite similar. However, by absolute values, SCW properties are significantly different from those of supercritical CO2 and R-134a. The only exceptions are: 1) dynamic viscosities within the pseudocritical point (Fig. 6) and 2) Prandtl number for gas-like fluids, i.e., beyond pseudocritical regions (Fig. 7).
(a) (b)
Fig. 1. Pressure vs. Temperature diagrams for water (a) and for carbon dioxide (b).
76 of 96 `
R134a Temperature, oC R134a Temperature, oC 100 125 150 175 200 225 250 275 300 325 100 125 150 175 200 225 250 275 300 325
o o CO2 Temperature, C CO2 Temperature, C 25 50 75 100 125 150 175 200 25 50 75 100 125 150 175 200 80
Water - 25.0 MPa Water - 25.0 MPa CO - 8.4 MPa 2 CO - 8.4 MPa 800 R134a - 4.6 MPa 2 R134a - 4.6 MPa 60
3 600
40
, kJ/kg.K ,
p
400 C
Density,kg/m
20
200
0 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.0 1.1 1.2 1.3 1.4 1.5 1.6
Temperature Ratio, ( T/Tcr ) Temperature Ratio, ( T/Tcr )
350 400 450 500 550 600 650 700 750 350 400 450 500 550 600 650 700 750 Water Temperature, 0C Water Temperature, 0C
Fig. 2. Density vs. reduced temperature. Fig. 3. Specific heat vs. reduced temperature. R134a Temperature, oC R134a Temperature, oC 100 125 150 175 200 225 250 275 300 325 100 125 150 175 200 225 250 275 300 325
o o CO2 Temperature, C CO2 Temperature, C 25 50 75 100 125 150 175 200 25 50 75 100 125 150 175 200 0.5 4000
Water - 25.0 MPa
CO2 - 8.4 MPa 0.4 R134a - 4.6 MPa 3000 Water - 25.0 MPa
CO2 - 8.4 MPa 0.3 R134a - 4.6 MPa 2000
0.2
Enthalpy, kJ/kg Enthalpy,
1000 ThermalConductivity,W/m.K 0.1
0.0 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.0 1.1 1.2 1.3 1.4 1.5 1.6
Temperature Ratio, ( T/Tcr ) Temperature Ratio, ( T/Tcr )
350 400 450 500 550 600 650 700 750 350 400 450 500 550 600 650 700 750 Water Temperature, 0C Water Temperature, 0C
Fig. 4. Specific enthalpy vs. reduced Fig. 5. Thermal conductivity vs. reduced temperature. temperature.
77 of 96 `
o R134a Temperature, C R134a Temperature, oC 100 125 150 175 200 225 250 275 300 325 100 125 150 175 200 225 250 275 300 325
o o CO2 Temperature, C CO2 Temperature, C 25 50 75 100 125 150 175 200 25 50 75 100 125 150 175 200
9 Water - 25.0 MPa Water - 25.0 MPa
CO - 8.4 MPa CO2 - 8.4 MPa 80 2 8 R134a - 4.6 MPa R134a - 4.6 MPa 7
Pa.s 6 60 5
4
Prandtl number Prandtl 40
DynamicViscosity, 3
2
20 1 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Temperature Ratio, ( T/Tcr ) Temperature Ratio, ( T/Tcr )
350 400 450 500 550 600 650 700 750 350 400 450 500 550 600 650 700 750 Water Temperature, 0C Water Temperature, 0C
Fig. 6. Dynamic viscosity vs. reduced Fig. 7. Prandtl number vs. reduced temperature. temperature.
At critical and supercritical pressures a fluid is considered as a single-phase substance, in spite of the fact that all thermophysical properties undergo significant changes within the critical and pseudocritical regions (Figs. 2-7). Near the critical point these changes are dramatic. In the vicinity of pseudocritical points with an increase in pressure these changes become less pronounced [2]. Also, it can be seen that properties such as density and dynamic viscosity undergo significant drops (near the critical point this drop is almost vertical) within a very narrow temperature range (see Figs. 2 and 6), while the specific enthalpy undergoes a sharp increase (see Fig. 4). The specific heat, thermal conductivity and Prandtl number have peaks near the critical and pseudocritical points (see Figs. 3, 5 and 7). The magnitudes of these peaks decrease very quickly with an increase in pressure [1-4]. Also, “peaks” transform into “humps” profiles at pressures beyond the critical pressure. It should be noted that the dynamic viscosity and thermal conductivity undergo through the minimum right after the critical and pseudocritical points (see Figs. 5 and 6).
The specific heat of water (as well as of other fluids) has the maximum value in the critical point [1-4]. The exact temperature that corresponds to the specific-heat peak above the critical pressure is known as the pseudocritical temperature (see Fig. 3). At pressures approximately above 300 MPa for water] a peak (here it is better to say “a hump”) in specific heat almost disappears. Therefore, the term pseudocritical point does not exist anymore. The same applies to the pseudocritical line. It should be noted that peaks in the thermal conductivity and volume expansivity may not correspond to the pseudocritical temperature.
In early studies, i.e., approximately before 1990, a peak in thermal conductivity was not taken into account [1]. Later, this peak was well established and included into thermophysical data and software. The peak in thermal conductivity diminishes at about 25.5 MPa for water (see Fig. 5).
78 of 96 `
In general, crossing the pseudocritical line from left to right (see Fig. 1) is quite similar as crossing the saturation line from liquid into vapour. The major difference in crossing these two lines is that all changes (even drastic variations) in thermophysical properties at supercritical pressures are gradual and continuous, which take place within a certain temperature range. On the contrary, at subcritical pressures properties experience discontinuities on the saturation line and have one value for liquid and another for vapour.
3. Forced-convection heat transfer in bare vertical circular tubes at supercritical conditions.
3.1. Water
A number of empirical generalized correlations have been proposed to calculate the HTC in forced convection for various fluids, including water and CO2, at supercritical pressures. However, differences in calculated HTCs can be up to several hundred percent [1].
The most widely used correlation at subcritical pressures for forced convection is the Dittus- Boelter correlation [1]. In 1942, McAdams proposed to use the Dittus-Boelter correlation in the following form:
Nu 0.0243 Re0.8Pr0.4 (1) b b b .
However, it should be noted that Eq. (1) does not include the effect of heat flux on HTC, i.e., does not account for variations of properties near the wall if the temperature difference (Tw – Tb) is quite significant due to high heat flux. Therefore, it produces quite unrealistic results within some flow conditions (see Fig. 8), especially, near the critical and pseudocritical points, where properties variations are drastic. In general, experimental HTC values show just moderate increase within the pseudocritical region. This increase depends on flow conditions and heat flux: higher heat flux – less increase. Therefore, the determination of the characteristic temperature is an important task.
Bulk Fluid Enthalpy, kJ/kg 1400 1600 1800 2000 2200 2400 2600 2800 Dittus - Boelter correlation 36 28
K 20 2 Hpc 16 12 Heat transfer coefficient 8
HTC, kW/m
Normal HT 4 Normal HT
pin=24.0 MPa 2 600 G=503 kg/m s DHT Improved HT 2 Q=54 kW 2
C 550 o qave= 432 kW/m 500 Inside wall temperature
450 tout 400 Bulk fluid temperature
Temperature, t 381.1 o C 350 tin Heated length pc 300 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Axial Location, m Fig. 8. Temperature and HTC profiles along heated length of vertical tubes with upward flow [1]: Water, ID 10 mm; Symbols - experimental data, lines – calculated data).
79 of 96 `
In general, several approaches based on various characteristic temperatures are known: 1) bulk-fluid-temperature approach; 2) wall-temperature approach; and 3) film-temperature approach in which Tf = (Tw + Tb) / 2. Moreover, at supercritical pressures the cross-sectional averaged Prandtl number ( Pr ), which accounts for thermophysical properties variations within a cross section due to heat flux, was proposed to be used in many supercritical heat- transfer correlations instead of the regular Prandtl number. Nevertheless, this classical correlation (Eq. (1)) was used extensively as a basis for various supercritical heat-transfer correlations.
In 1964, Bishop et al. [6] conducted experiments in supercritical water flowing upward inside bare tubes and annuli within the following range of operating parameters: P = 22.8 – 27.6 2 2 MPa, Tb = 282 – 527ºC, G = 651 – 3662 kg/m s and q = 0.31 – 3.46 MW/m . Their data for heat transfer in tubes were generalized using the following correlation with a fit of ±15%:
0.43 0.66 0.9 w D Nu 0.0069 Re Prb 1 2.4 b b x b . (2)
Equation (2) uses the cross-sectional averaged Prandtl number, and the last term in the correlation: (1+2.4 D/x), accounts for the entrance-region effect.
In 1965, Swenson et al. [7] found that conventional correlations, which used bulk-fluid temperature as a basis for calculating the majority of the thermophysical properties, were not accurate. They suggested the following correlation in which the majority of thermophysical properties are based on a wall temperature:
0.231 0.613 0.923 w Nu 0.00459 Re Pr w w w b (3)
Equation (3) was obtained within the following range: P = 22.8 − 41.4 MPa, Tb = 75 − 576ºC, 2 Tw = 93 − 649ºC and G = 542 − 2150 kg/m s; and predicts the experimental data within ±15%.
An analysis performed by Pioro and Duffey [1] showed that correlations of Bishop et al. [6] (Eq. (2)) and Swenson et al. [7] (Eq. (3)) were obtained within the same range of operating conditions as those for supercritical coal-fired thermal-power plants and SCWRs.
The majority of empirical correlations were proposed in the 1960s – 1970s, when experimental techniques were not at such an advanced level as they are today. Also, even thermophysical properties of water have been updated since that time (for example, a peak in thermal conductivity in critical and pseudocritical points within a range of pressures from 22.1 to 25 MPa was not officially recognized until the 1990s). Therefore, recently new correlations based on modern sets of heat-transfer data and the latest thermophysical properties of water [5] have been developed and evaluated within the power-engineering operating range (see Figs. 9 and 10):
1) Bulk-fluid-temperature approach [8]:
0.564 0.684 0.904 w Nub 0.0061Reb Prb (4) b
Equation (4) has uncertainties of ±25% for HTC values and about ±15% for calculated wall
80 of 96 ` temperatures.
2) Wall-temperature approach [9]:
0.398 0.156 0.764 0.941 w w Nuw 0.0033Rew Prw (5) b b
Equation (5) has uncertainties of ±25% for HTC values and about ±10% for calculated wall temperatures (see Fig. 10).
It should be noted that all heat-transfer correlations presented in this paper are intended only for normal and improved heat-transfer regimes.
(a) (b) Fig. 9. Comparison of data fit (Eq. (5)) with experimental data: (a) for HTC and (b) for Tw.
Fig. 10. Comparison of HTC values calculated through Eqs. (1-5) with experimental data: Vertical bare tube, upward flow, D=10 mm, Lh=4 m, Pin=24.1 MPa and G=500 and 1500 kg/m2s.
81 of 96 `
Figures 10 and 11 show that the latest correlations (Eqs. (4) and (5)) properly represent experimental data1 and follow trends closely even within the pseudocritical range. Also, the Swenson et al. correlation showed good predictions with the current experimental data. Computational Fluid Dynamics (CFD) codes are a nice and modern approach. However, not all turbulent models are applicable to heat transfer at supercritical pressures, plus these codes should be tuned first on the basis of experimental data and only after that used in similar calculations (see Fig. 11) [10, 11].
A recent study was conducted by Zahlan et al. [12] in order to develop a heat-transfer look- up table for critical/supercritical pressures. An extensive literature review was conducted, which included 28 datasets and 6663 trans-critical heat-transfer data. In their conclusions, Zahlan et al. [12] determined that within the supercritical region the latest correlation by Mokry et al. (Eq. (4)) showed the best prediction for the data within all three sub-regions investigated (1 -- close to the critical or pseudo-critical point region, 2 -- the high-density or liquid-like region, and 3 - the low-density or gas-like region). Based on an analysis of the data discussed in [12], the Mokry et al. correlation (Eq. (4)), which is based on the bulk-fluid temperature, was found to be the most accurate supercritical-water correlation within a wide range of operating conditions compared to other ones. The correlations by Gupta et al. (Eq. (5)) and Swenson et al. (Eq. (3)), which are based on the wall-temperature approach, closely follow the Mokry et al. correlation. Additional analysis conducted by Zahlan et al. [12] showed that the Mokry et al. correlation (Eq. (4)) is the most accurate superheated-steam correlation within a wide range of operating conditions compared to other ones. Moreover, the Mokry et al. correlation is just slightly less accurate (by 1% of RMS error) compared to the Gnielinski correlation [13], which is the most accurate for subcritical water.
Bulk-Fluid Enthalpy, kJ/kg
1600 1800 2000 2200 2400 2600 2800 Proposed corr. Proposed corr. k-model (wall function) k- model (wall function) k- model (low Reylonds) k- model (low Reynolds) k- model (SST) 500 k-model (SST) P = 23.9 MPa in Hpc H 2 pc P = 23.9 MPa G = 1002 kg/m s in 34 2 2 q = 681 kW/m G = 1002 kg/m s ave 30 2
2 C
q = 681 kW/m o q = 688 kW/m ave 26 450 dht
K q = 688 kW/m2 2 dht 22 Heat transfer coefficient 18 Inside-wall temperature Tout 14 400
Temperature, Temperature, 10 HTC,kW/m
C 6 o o 450 Tpc = 381 C T 2 out Tin 400 Bulk-fluid temperature 350 Bulk-fluid temperature o Tin T = 381 C 350 pc Heated length Heated length
Temperature, 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Axial Location, m Axial Location, m
(a) (b) Fig. 11. Comparison of HTC (a) and wall-temperature values (b) calculated through Eq. (4) and FLUENT CFD-code [20] with experimental data: Vertical bare tube, upward 2 flow, D=10 mm, Lh=4 m, Pin=23.9 MPa and G=1000 kg/m s.
3.2. Supercritical carbon dioxide
In support of developing an SCW reactor, studies are being conducted into heat transfer at supercritical conditions using carbon dioxide as a modeling fluid as a less expensive alternative to using supercritical water. In the same way as with supercritical water, the majority of empirical correlations for supercritical carbon dioxide were proposed in the 1960s
1 Data of Professor P.L. Kirillov were used [1].
82 of 96 `
– 1970s, when experimental techniques were not at such an advanced level as they are today. Also, thermophysical properties of carbon dioxide have been updated since that time. The correlation developed by Mokry et al. was compared to a set of SCCO2 data, to determine its applicability to this modeling fluid. In general, it was found that the correlation under-predicted the inside-wall temperature. Thus, it was determined that this correlation was not applicable to SCCO2.
Therefore, a new correlation based on a modern set of heat-transfer data2, the latest thermophysical properties of SCCO2 [5] and the wall-temperature approach has been developed: 0.222 0.836 0.754 -0.139 0.957 wwkw Nu 0. 0038 Re Prw ww k bb b (7)
Equation (7) showed uncertainties of ±30% for HTC values and about ±20% for calculated wall temperatures. It should be noted that this heat-transfer correlation is also intended only for normal and improved heat-transfer regimes. Analysis of Gupta et al. correlation (Eq. (7)) [14] showed that Eq. (7) properly represents experimental data and follow trends closely even within the pseudocritical range.
4. CONCLUSIONS
Several heat-transfer correlations have been developed for supercritical water flowing upward in vertical bare tubes. The Mokry et al. correlation appeared to be the most accurate one compared to other correlations within supercritical conditions (with uncertainties of ±25% for HTC values and about ±15% for calculated wall temperatures) and for superheated steam (with an average error of -5% and RMS error of 20%). Also, it is just 1% (RMS error) less accurate compared to the best correlation for subcritical water by Gnielinski [13].
It was determined that the Mokry et al. correlation (the same applies to many other supercritical water correlations) is not applicable for supercritical carbon dioxide. Therefore, a new correlation was developed, which generalized data with uncertainties of ±30% for HTC values and about ±20% for calculated wall temperatures
The proposed correlations can be used: (1) for calculations of supercritical-water heat- transfer in heat exchangers; (2) for a preliminary heat-transfer calculations in SCWR fuel channels as a conservative approach; (3) for future comparisons with other independent datasets and with bundled data; (4) for the verification of computer codes for thermalhydraulics; and (5) for the verification of scaling parameters between water and modeling fluids (CO2, refrigerants, etc.)
5. ACKNOWLEDGEMENTS Financial support from the NSERC Discovery Grant is gratefully acknowledged.
6. NOMENCLATURE cp specific heat at constant pressure, J/kg∙K H H c p average specific heat, J/kg∙K, w b Tw Tb D diameter (usually inside diameter), m G mass flux, kg/m2s H specific enthalpy, J/kg
2 Data by I. Pioro [8].
83 of 96 ` h heat transfer coefficient, W/m2/K k thermal conductivity, W/m/K L length, m m mass-flow rate, kg/s P pressure, Pa q heat flux, W/m2 T temperature, ºC x axial location, m
Greek letters dynamic viscosity, Pa∙s density, kg/m3
Dimensionless numbers h D Nu Nusselt number k c p Pr Prandtl number k Pr average Prandtl number c p k G D Re Reynolds number
Subscripts ave average b bulk calc calculated cr critical dht deteriorated heat-transfer exp experimental fl flow h heated in inlet out outlet pc pseudocritical r reduced w wall
Acronyms CFD Computational Fluid Dynamics CHF Critical Heat Flux HTC Heat Transfer Coefficient RMS Root Mean Square SCF SuperCritical Fluid SCWR SuperCritical Water-cooled nuclear Reactor
7. REFERENCES
[1] Pioro, I.L. and Duffey, R.B., 2007. Heat Transfer and Hydraulic Resistance at Supercritical Pressures in Power Engineering Applications, ASME Press, New York,
84 of 96 `
NY, USA, 328 pages. [2] Pioro, I., 2011. The Potential Use of Supercritical Water-Cooling in Nuclear Reactors. Chapter in Nuclear Energy Encyclopedia: Science, Technology, and Applications, Editors: S. Krivit, J. Lehr and Th. Kingery, J. Wiley & Sons, Hoboken, NJ, USA, pp. 309-347 pages. [3] Pioro, I. and Mokry, S., 2011. Thermophysical Properties at Critical and Supercritical Conditions, Chapter in book “Heat Transfer. Theoretical Analysis, Experimental Investigations and Industrial Systems”, Editor: A. Belmiloudi, INTECH, Rijeka, Croatia, pp. 573-592. [4] Pioro, I., Mokry, S. and Draper, Sh., 2011. Specifics of Thermophysical Properties and Forced-Convective Heat Transfer at Critical and Supercritical Pressures, Reviews in Chemical Engineering, Vol. 27, Issue 3-4, pp. 191–214. [5] Lemmon, E.W., Huber, M.L. and McLinden, M.O., 2010. NIST Standard Reference Database 23: Reference Fluid Thermodynamic and Transport Properties-REFPROP, Version 9.0, National Institute of Standards and Technology, Standard Reference Data Program, Gaithersburg, USA. [6] Bishop, A.A., Sandberg, R.O. and Tong, L.S., 1964. Forced convection heat transfer to water at near-critical temperatures and super-critical pressures, Report WCAP-2056, Westinghouse Electric Corporation, Atomic Power Division, Pittsburgh, PA, USA, December, 85 pages. [7] Swenson, H.S., Carver, J.R. and Kakarala, C.R., 1965. Heat transfer to supercritical water in smooth-bore tubes, Journal of Heat Transfer, Transactions of the ASME, Series C, 87 (4), pp. 477–484. [8] Mokry, S., Pioro, I., Farah, A., et al., 2011. Development of Supercritical Water Heat- Transfer Correlation for Vertical Bare Tubes, Nuclear Engineering and Design, Vol. 241, pp. 1126-1136. [9] Gupta, S., Mokry, S. and Pioro, I., 2011. Developing a Heat-Transfer Correlation for Supercritical-Water Flowing in Vertical Tubes and Its Application in SCWR, Proc. ICONE-19, Makuhari, Japan, May 16-19, Paper 43503, 11 pages. [10] Farah, A., Kinakin, M., Harvel, G. and Pioro, I., 2011. Study of Selected Turbulent Models for Supercritical Water Heat Transfer in Vertical Bare Tubes Using CFD Code FLUENT-12, Proc. ICONE-19, Makuhari, Japan, May 16-19, Paper #43492. [11] Vanyukova, G.V., Kuznetsov, Yu.N., Loninov, A.Ya., et al., 2009. Application of CFD- code to calculations of heat transfer in a fuel bundle of SCW pressure-channel reactor, Proc. 4th Int. Symp.on Supercritical Water-Cooled Reactors, March 8-11, Heidelberg, Germany, Paper No. 28, 9 pages. [12] Zahlan, H., Groeneveld, D.C., Tavoularis, S., et al., 2011. Assessment of Supercritical Heat Transfer Prediction Methods, Proc. 5th International Symposium on SCWR (ISSCWR-5), Vancouver, BC, Canada, March 13-16, Paper P008, 20 pages. [13] Gnielinski, V., 1976. New equation for heat and mass transfer in turbulent pipe and channel flow, International Chemical Engineering, 16 (2), pp. 359–368. [14] Gupta, S., McGillivray, D., Surendran, P., et al., 2012. Developing heat-transfer correlations for supercritical CO2 flowing in vertical bare tubes. Proc. ICONE20- POWER2012, July 30–August 3, Anaheim, CA, USA, Paper #54626, 13 pages.
85 of 96 SAFETY ANALYSIS OF A SODIUM-COOLED FAST REACTOR WITH TRANSMUTATION CAPABILITIES
SARA PEREZ-MARTIN, ALEXANDER PONOMAREV, REGINA KRUESSMANN, DANKWARD STRUWE, MICHAEL SCHIKORR, WERNER PFRANG, WOLFGANG HERING Karlsruhe Institute of Technology Institute for Neutron Physics and Reactor Technology Hermann-von-Helmholtz-Platz 1 76344 Eggenstein-Leopoldshafen, Germany
DAVID LEMASSON, SIMONE MASSARA, DANIELE VERWAERDE EDF - R&D SINETICS - I28 1, avenue du Général de Gaulle 92141 Clamart Cedex, France
ABSTRACT
In this work we analyse the behaviour of the so-called Reference Oxide core of the CP ESFR project under an ULOF transient considering transmutation of minor actinides (MA). We compare such results with those obtained for the original MOX fuel case. Power, reactivity and two-phase flow for boiling onset and pin failure time are presented. Even though reactivity feedbacks are more critical in the transmuter core, grace times up to boiling onset and pin failure are similarly small for MOX and MA-MOX fuelled cores.
1. Introduction
The level of safety and fuel efficiency are two key-points that are intended to be improved in next generation of nuclear reactors. The way to achieve it will be based on new nuclear systems, advanced core designs and the use of innovative nuclear fuel. One of the innovative systems is the sodium-cooled fast reactor using MOX fuel bearing minor actinides. The influence of such transmuter core on the plant safety has to be investigated as a major topic. We report here on results of investigations related to safety aspects of an innovative SFR taking consequences of an unprotected loss of flow (ULOF) transient as indicative for the achieved safety level. In addition, we explore the impact of minor actinides in the fuel on the evolution of the core during such an ULOF transient. We consider thus their impact on reactor physics characteristics (feed-back reactivity coefficients and kinetic parameters) as well as on thermo-mechanical properties.
One of the advantages of fast neutron spectrum is the efficient transmutation of minor actinides generated in current light water reactors. The reduction in the amount of transuranium elements would reduce the radiotoxicity of nuclear waste as well as the heat load and size of the final repository. However, the integration of Am, Np and Cm in MOX fuel can deteriorate the response of the core under transient conditions. In order to study this effect, we have simulated a ULOF transient in two different cores: one using MOX fuel and another one using MOX fuel with MA.
We consider a ULOF simulation in the worst conditions that is at End of Equilibrium Cycle conditions (EOEC) when the important parameters as Doppler coefficient and sodium void effect (SVE) worth are deteriorated in comparison with a core loaded with fresh fuel.
86 of 96 The steady state and transient analysis has been done using the SAS-SFR code Ref2011 Error! Reference source not found.]. SAS-SFR is able to analyse the initiating phase of core disruptive accidents, in particular, resulting from unprotected under-cooling loss of flow (ULOF) or overpower (UTOP) conditions. Recently material properties of minor actinides fuels have been implemented in SAS-SFR (thermal conductivity, heat capacity, linear expansion coefficient and melting temperatures).
All neutron physics calculations are prepared with the modular code system KANEXT [2][3] on basis of the JEFF 3.1.1 cross section data library [4]. KANEXT is intended for broad spectrum of core neutron physics problems and allows to provide specific data for calculations with SAS-SFR code.
2. Core description
The core design used in this study is the Reference Oxide core of the CP ESFR project [5][6]. The main characteristics are summarized in Table 1. Figure 1 shows the radial and axial cross section of the core model.
Outlet sodium
DSD CSD Upper steel shield 90 % 20 % boro boro n Sodium Plenum n Upper Gas Plenum Upper Axial Blanket
Radial Radial steel Outer core Inner core Outer core steel reflector reflector
Follo Follo wer Lower axial blanket wer
Sub-assemblies of inner core Control rods (225 SAs) (24 CSDs) Lower gas expansion volume
Sub-assemblies of outer core Shutdown rods (228 SAs) (9 DSDs) LPP LPP LPP
Radial steel reflector sub-assemblies Inlet sodium and central dummy channel
Figure 1 Radial and axial cross section of the Reference Oxide core
Total reactor power (MWth) 3600 Core inlet/outlet temperature (ºC) 395 / 545 (U,Pu)O or Core fuel 2 (U,Pu,Np,Am,Cm)O2 Pu mass content in fresh fuel in inner/outer core (%) 14.05 / 16.35 Minor Actinides mass content in fresh fuel (%) 0 or 5 Height of the core (m) 1.00 Number of SAs in inner/outer core 225 / 228 Number of control rods (CSD/DSD) 24 / 9 Number of pins per SA 271 Pin outer diameter (mm) 10.73 Pellet outer diameter (mm) 9.43 Clad thickness (mm) 0.50 Table 1 General data on ESFR Reference Oxide core design
Core Geometry The core includes inner and outer sub-core with different Pu content (225 SAs with 14.05% and 228 SAs with 16.35%, respectively) in order to flatten the radial core power shape. The fuel sub-assembly (SA) consists of a hexagonal wrapper tube that contains a triangular bundle arrangement of 271 fuel pins with helical wire wrap spacers. The hexagonal wrapper
87 of 96 tube is made of ferritic martensic steel EM10. The fuel pin contains either (U,Pu)O2 or (U,Pu,Np,Am,Cm)O2 fuel pellets and has ODS steel cladding. The control rod system is composed of 24 CSDs (Control and Shutdown Device) and 9 DSDs (Diverse Shutdown Device).
For the MA case we consider uranium to be partially replaced by MA thus 5% MA mass content is included in all SA (both in inner and outer core) while the plutonium content is kept unchanged. The MA isotopic vector in fresh fuel is given in Table 2 [7]. The core has no fertile breeder zones and is surrounded by three rows of steel reflector sub-assemblies.
Isotope % w Isotope % w
237Np 16.86 242Cm 0.02
241Am 60.61 243Cm 0.07 242 244 Amm 0.24 Cm 5.14
243Am 15.7 245Cm 1.26
246Cm 0.1 Table 2 Minor Actinides Vector
Burn-up In accordance with the Reference Oxide core concept the targeted fuel residence time corresponds to 2050 equivalent full power days (EFPD) [6][8]. The same fuel residence time is considered for the MA-MOX case. In order to simulate detailed burn-up the reloading pattern proposed in [8] is taken (with some minor modifications) for both MOX and MA-MOX cases. Every batch is 410 EFPD length and considers the load of 1/5 of fresh fuel sub- assemblies. The target average and peak core burn-up are respectively 100 GWd/t and 145 GWd/t for an estimated average power density of about 200 W/cm3 [5]. For simplicity, all CRs are considered to be withdrawn and located at top core level during the whole burn-up simulation.
Power-to-flow ratio scheme Since representation of all individual sub-assemblies is a heavy computational task, we group SAs having similar characteristics in thermal, hydraulic, neutron physics and fuel pin mechanics behaviour into one representative SA (so called SAS-SFR channel) with characteristics averaged over all SAs belonging to the group. In order to represent accurately the core state with different levels of burn-up and SA power 30 groups were selected. The grouping scheme is chosen such that fuel pin failure sequences and / or boiling onset sequences in the different SA groups represent as closely as possible the sequences to be expected as a consequence of the overall core design specifications and the considered transient. This makes it necessary to model the different enrichment zones, the different batches of the core load and the number of cooling groups of the core design concept in the necessary detail. Moreover, the current detailed burn-up simulation allows us to follow the “real” power history of the channel and to provide it to SAS-SFR code to simulate the power operation phase appropriately. Figure 2 (left hand side) shows the allocation of the SAs to the different channels. The hexagon colours reflect different burn-up levels of SA in accord with reloading pattern and the two numbers give the channel number and power-to-mass flow ratio in the corresponding channel. The coolant mass flow per SA is adjusted in such a way that the average coolant heat up over the core yields 150 K.
88 of 96
Figure 2 Allocation of the 30 SA groups (left hand side) and the reactivity feedback sets (right hand side)
3. MA-MOX Fuel Reactivity Feedbacks
Reactivity feedback effects are discretised spatially in 4 radial zones and 21 axial nodes (13 within the fissile core). Figure 2 (right hand side) shows the radial zones considered according to the ring wise SA-grouping approach. Spatial distributions are constructed based on direct reactivity difference after the perturbation is inserted locally in the considered region. The material reactivity effect (for sodium, fuel and clad) supplied to SAS-SFR is defined as the ratio of the corresponding reactivity change to the variation of material mass. Doppler effect is defined as the ratio of corresponding reactivity change to the variation of fuel temperature. The sodium reactivity effect is derived for all regions of lower axial blanket, fissile height of the core, upper axial blanket, upper gas expansion plenum and sodium plenum. The clad and fuel reactivity effects are derived in all regions except in the sodium plenum.
In Table 3 the Doppler constant and sodium void effect are given. Wet and dry Doppler constants of the MA-MOX case are ~25% lower than in the MOX case. SVE in the core is about 200 pcm higher for MA-MOX case than for MOX whereas for both cases the sodium plenum gives the similar reduction of SVE by about 130 pcm.
Reactivity Coefficients MOX case MA-MOX case Doppler constant (pcm) Wet -808 -613 Dry -623 -472 Sodium void effect (pcm) in core 2206 2397 in core and plenum 2068 2273 Table 3 Reactivity effect values calculated in KANEXT
4. MA-MOX Fuel Thermal Properties
The integration of MA in the MOX fuel will modify not only the reactivity performance of the core but also the thermal properties of the pins.
89 of 96 Due to the similar thermal behaviour of americium oxide, curium oxide and MOX fuel, we 1 have derived new thermal conductivity correlations for AmO2-x and CmO2-x based on similar fundamentals as used in Philipponneau’s work [9]. Therefore dependencies on burn-up, porosity, O/M ratio and self-irradiation effects are considered, besides temperature. Such correlations are based on experimental and model results available in the literature. Using the Bruggeman method [10][11] to estimate the thermal conductivity of the MA-MOX fuel, we have found that 5% MA content does not affect the thermal conductivity very much.
Following the same procedure for the thermal expansion coefficient and heat capacity we have seen that the inclusion of 5% MA only has a minor influence too. Probably a larger fraction of MA would affect the MA-MOX thermal properties more distinctly, but in the range of 5-10% of MA, the differences are importantly small. Figure 3, Figure 5 and Table 4 show the thermal conductivity, heat capacity, expansion coefficient and melting temperatures for 5% MA-MOX fuel compared with MOX fuel.
Figure 3 Thermal conductivity as a function of the fuel temperature for 5% MA-MOX, AmO2-x, CmO2-x and MOX
Figure 4 Linear Expansion Coefficient as a function of the fuel temperature for 5% MA-MOX, NpO2, AmO2-x, CmO2-x and MOX
1 Americium and curium oxides have different oxygen to metal ratio depending on the fuel temperature. The O/M can vary from stoichiometry (AmO2) to hypo-stoichiometry (Am2O3)
90 of 96
Figure 5 Heat capacity as a function of the fuel temperature for 5% MA-MOX, NpO2, AmO2-x, CmO2-x and MOX
To estimate the melting temperature of the MA-MOX fuel, we have used the average of the individual actinide oxide melting temperatures as a first approximation. The estimated melting temperature of 5% MA-MOX mixture is 3003 K with an uncertainty of ~20 K. It is slightly lower than the corresponding MOX melting temperature (around 3050K).
Melting Temperature (K)
UO 2 3120±20
PuO2 2674±20
NpO2 2836±50
AmO2-x 2481±15
CmO2-x 2543±25 5%MA-MOX 3003±20 Table 4 Melting temperatures of actinide oxides
5. Results
Steady-state SAS-SFR code is used to determine the steady-state conditions after start-up and up to the specified EOEC core conditions. For the start-up, a time of 6 h was assumed before steady- state is reached, followed by a 2050 EFPD operation time period; the transient calculations start subsequently. We have used the same nominal coolant inlet temperature (668.15 K), pressure at the coolant outlet plenum and coolant mass flow rates as were given in the CP ESFR project. An average coolant heat-up of 150 K between coolant inlet and outlet plenum was also specified. The calculated pressure drop over the core is 4.8 bar including contributions from SA inlet throttling and the pressure drop along the grid plate flow path.
ULOF transient We have simulated an unprotected pump coast down to 0% of rotational speed with a pump flow halving time of 10 s according to the following equation for the coolant flow:
Qt() ;where 10 s Q( t 0) t
91 of 96 The reactivity feedback effect due to the expansion of the control rod drive lines and the reactivity feedback effect of the diagrid expansion/contraction were calculated with the SAS- SFR built-in models. However, neither the model for the control rod expansion feedback nor the model for the diagrid expansion feedback is experimentally validated. Therefore, input values were adjusted to what was specified by the CP ESFR project.
SAS-SFR code is able to simulate the entire initial phase of the ULOF transient. It only ends when 5 hexcan nodes in the hottest SA are molten provoking radial spreading of the mobile core materials which is not modelled in SAS-SFR. Most system codes such as RELAP, TRACE or CATHARE stop their calculations at coolant boiling onset.
Figure 6 Normalised power and reactivity contributions during the single phase heat-up. MOX case on the left and MA-MOX case on the right
The transient variation of the normalised power and the net reactivity up to boiling onset are shown in Figure 6 together with the individual contributions of Doppler and sodium density variation for both cores.
MOX case MA-MOX case Boiling Onset Time (s) 25.2 23.8 Normalised power 1.04 1.04 Net reactivity (c) +0.2 +0.4 Sodium density Reactivity (c) +31.8 +32.4 Doppler effect (c) -20.3 -21.2 Fuel expansion reactivity (c) -8.7 -8.7 Clad material expansion reactivity (c) +2.2 +2.2 Control rod drive line expansion reactivity (c) -4.8 -4.4 Table 5 Normalised power and reactivities (in cents of $) at boiling onset
The normalised power, net reactivity and the important terms of net reactivity are shown in Table 5 for boiling onset. For MA-MOX fuelled core the time of boiling onset is 1.4 s earlier than in the MOX core reaching the same normalised power in both cases. The reactivity feedback due to sodium heat up is 0.6 c larger for MA-MOX than for MOX and the Doppler effect is 0.9 c larger for MA-MOX then for MOC fuelled core. The average fuel temperature for the channel that boils first is 1714ºC for MA-MOX case and 1675ºC for MOX case. The higher temperature in the MA-MOX fuel explains the larger Doppler effect compared to the MOX fuel. The reason for that higher fuel temperature in MA-MOX fuel can be found in its slightly degraded thermal properties. However, nuclear reactions of minor actinides can also play a role in power production and therefore in fuel temperature.
The extension of the coolant two-phase zone in the SA group where the boiling starts and the associated sodium void reactivity feedback are shown in Figure 7. The extension of the
92 of 96 two-phase flow zone ends at the reactor outlet plenum because this is the axial location for sodium vapour condensation and thus the location determining the upper pressure boundary condition for the two-phase flow development in time.
Figure 7 Axial extent of the two-phase flow region in the SA group that boils first. Sodium reactivity feedback is plotted in red dotted line. MOX case on the left and MA-MOX case on the right. Axial height is zero at the Bottom of the Fissile Column (BFC).
In the initial phase of boiling the two-phase flow region extends in the upward direction only on a delayed time scale as the rather cold hexcan structure at the level of the upper sodium plenum provides an effective heat sink for the initially produced vapour. Similar observations but on a different time scale can be made when the two-phase flow zone reaches the upper shielding region. For the MA-MOX case the upper end of the two phase region reaches the SA outlet at 247 cm (BFC) only almost 2.1 s after boiling onset and reaches the SA sodium outlet level (350cm BFC) at 2.2 s after boiling onset. However for the MOX case the upper limit of the two-phase flow region reaches the SA outlet 1.7 s after boiling onset and reaches the SA sodium outlet level 2.2 s after boiling onset. The lower two-phase flow front enters the fissile core region after more than 1.5 s boiling time in both cases. Only then it expands further down into the fissile core. The release of the positive void reactivity due to sodium density variation (red dotted line) takes place in accordance with the two-phase flow extension into fissile core height. As shown in the Figure 7 sodium reactivity feedback becomes larger and explains the evolution of the two-phase front.
The transient characteristics of the core behaviour during the boiling phase for both cores are plotted in Figure 8 and Figure 9. In this latter figure boiling onset and pin failure calculated to occur in the different SA groups are indicated. A power excursion takes place after boiling onset leading to fuel melting and to fuel pin failures of a break-up mode around 4s after boiling onset in both cases.
Figure 8 Normalised power and reactivity contributions during the boiling phase. MOX case on the left and MA-MOX case on the right.
93 of 96
Figure 9 Normalised power as a function of boiling time extending into the fuel failure time phase. MOX case on the left and MA-MOX case on the right.
Table 6 compares the results for the two cores showing the differences in the boiling onset and pin failure times as well as the maximum normalized power and net reactivity. MA-MOX core reaches the boiling onset 1.4 s earlier than MOX core. However pin failure occurs in both cases ~4s after the boiling onset.
MOX MA-MOX Boiling onset (time into the transient, s) 25.2 23.8 Boiling time until first fuel pin failure, s 3.9 4.0 Max. normalized power, - 913 4265 Max. net reactivity, $ 1.023 1.058 Table 6 Comparison of times for boiling onset and pin failure as well as power and net reactivity for the different fuelled cores (MOX vs. MA-MOX)
Figure 10 shows the normalized power after pin failure together with the sodium void, clad and fuel relocation effects for both cores. Void reactivity effect increases more pronouncedly in the period 50 ms to 75 ms in the MA-MOX core than in the MOX core. Due to that higher slope the maximum value of the normalized power is 4265 in MA-MOX and 913 in the MOX core.
Figure 10 Normalised power and reactivity contributions during the post failure phase of the accident for MOX core (on the left) and MA-MOX core (on the right).
After pin failure the strong fuel relocation feedback controls the net reactivity and the power reduces drastically. According to these results it can be stated that reactivity feedback effects due to Doppler and fuel relocation are effective in preventing the ULOF transient from
94 of 96 becoming super-prompt critical during the primary phase even for cores where minor actinides are present.
6. Conclusions
We have summarized SAS-SFR results simulating the initiation phase of an unprotected loss-of-flow (ULOF) transient of the CP ESFR Reference Oxide core using MOX and 5% MA- MOX fuelled cores.
The effects of 5% minor actinides content in MOX fuelled core are shown to be rather small for this specific core design: coolant boiling onset is 1.4s earlier compared to MOX fuel and the time delay until pin failure is the same. After pin failure the Doppler effect and the fuel relocation are the dominant contributions of the net reactivity. The normalized power is then reduced. These conclusions should not be extended to any other core designs, because the ULOF transient response is strongly dependent on the core and plant design particularities.
7. References
[1] “Representative transients within the design basis” Deliverable CP ESFR SP3.3.1.D1 2012.
[2] “KAPROS-E: Modular Program System for Nuclear Reactor Analysis, Status and Results of Selected Applications” C.H.M. Broeders, R. Dagan, V. Sanchez, A. Travleev Jahrestagung Kerntechnik, Düsseldorf, May 25-27, 2004
[3] KANEXT description: http://inrwww.fzk.de/kanext.html
[4] “Processing of the JEFF-3.1 Cross Section Library into Continuous Energy Monte Carlo Radiation Transport and Criticality Data Library” NEA/NSC/DOC(2006)18
[5] “CP ESFR Working Horses Core Concept Definition” Deliverable CP ESFR SP2.1.2.D1 2009.
[6] “Synthesis of options to optimize feedback “ Deliverable CP ESFR SP2.1.3.D1 2012
[7] “Transmutation options assessments” Deliverable CP ESFR SP2.1.4.D1 2011
[8] “Working Horses ESFR Core Concepts Characterization: Neutronic and Thermal- Hydraulic Characteristics” Deliverable CP ESFR SP2.1.2.D2 2010.
[9] “Thermal conductivity of (U, Pu)O2 mixed oxide fuel” Y. Philipponneau Journal of Nuclear Materials 188 (1992) 194-197
[10] “Predicting thermo-mechanical behaviour of high minor actinide content composite oxide fuel in a dedicated transmutation facility” S.E. Lemehov, V.P. Sobolev, M. Verwerft. Journal of Nuclear Materials 416 (2011) 179–191
[11] Appendix B, Deliverable 36, WP 4.2 PDS-XADS issued by KTH (2004)
95 of 96