Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance

Comprehensive Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance Comprehensive nuclear materials 3 07 TRISO coated particle fuel performance nuclear materials 3 07 TRISO coated particle fuel performance

3.07 TRISO-Coated Particle Fuel Performance D A Petti, P A Demkowicz, and J T Maki Idaho National Laboratory, Idaho Falls, ID, USA R R Hobbins RRH Consulting, Wilson, WY, USA Published by Elsevier Ltd 3.07.1 Introduction 153 3.07.2 3.07.2.1 3.07.2.1.1 3.07.2.1.2 3.07.2.1.3 3.07.2.1.4 3.07.2.1.5 3.07.2.1.6 3.07.2.1.7 3.07.2.1.8 3.07.2.1.9 3.07.2.1.10 3.07.2.2 3.07.2.2.1 3.07.2.2.2 3.07.2.2.3 3.07.2.2.4 3.07.2.2.5 3.07.2.2.6 3.07.2.2.7 3.07.2.2.8 3.07.2.2.9 3.07.2.3 3.07.2.3.1 3.07.2.3.2 3.07.2.3.3 3.07.2.3.4 3.07.2.3.5 3.07.2.3.6 3.07.2.3.7 3.07.2.3.8 3.07.2.3.9 3.07.2.3.10 3.07.2.3.11 3.07.2.3.12 3.07.2.3.13 3.07.2.4 3.07.2.5 3.07.2.6 3.07.2.7 3.07.2.7.1 3.07.2.7.2 3.07.2.7.3 3.07.2.7.4 Irradiation Performance Overview of Irradiation Facilities and Testing BR-2 IVV-2M HFR Petten HFIR ATR SAFARI TRISO-coated particle fuel irradiation testing Thermal and physics analysis considerations Gas control system considerations FPMS considerations German Experience R2-K12 and R2-K13 BR2-P25 HFR-P4 SL-P1 HFR-K3 FRJ2-K13 FRJ2-K15 FRJ2-P27 HFR-K6 and HFR-K5 US Experience F-30 HRB-4 and HRB-5 HRB-6 OF-2 HRB-14 HRB-15B R2-K13 HRB-15A HRB-16 HRB-21 NPR-1 and NPR-2 NPR-1A AGR-1 European Experience Chinese Experience Japanese Experience Irradiation Performance Summary Heavy metal contamination In-service failures Failure mechanisms Acceleration effects 154 154 154 154 155 155 155 155 156 156 158 159 160 160 161 162 162 163 163 164 164 165 165 166 167 169 170 171 173 174 174 175 176 177 178 178 180 181 184 185 185 186 186 187 151 152 TRISO-Coated Particle Fuel Performance 3.07.3 3.07.3.1 3.07.3.1.1 3.07.3.1.2 3.07.3.1.3 3.07.3.1.4 3.07.3.2 3.07.3.3 3.07.3.3.1 3.07.3.3.2 3.07.3.3.3 3.07.3.3.4 3.07.3.4 3.07.3.4.1 3.07.3.4.2 3.07.3.5 3.07.4 References Safety Testing Facility Overview KuăFA at ITU INLs FACS ORNL’s Core Conduction Cooldown Test Facility KORA German Experience European Experience AVR 73/21 AVR 74/18 HFR K6/3 HFR K6/2 US Experience and Future Plans Past experience Future plans Japanese Experience Conclusions Abbreviations AGR ATR AVR BAF BISO BOL BR-2 CCCTF CVD DOE EFPD EOL FACS FIMA FPMS FRJ GETR HEU HFEF HFIR HFR HRB HTGRs HTR-10 HTTR IFEL IMGA Advanced Gas Reactor Advanced Test Reactor Arbeitsgemeinschaft Versuchsreaktor Bacon anisotropy factor Bi-structural isotropic Beginning of life Belgian Reactor Core Conduction Cooldown Test Facility Chemical vapor deposition Department of Energy Effective full-power day End of life Fuel accident condition simulator Fissions per initial metal atom Fission product monitoring system Research Reactor Juelich General Electric Test Reactor Highly enriched uranium Hot Fuel Examination Facility High-Flux Isotope Reactor High-Flux Reactor HFIR Removable Beryllium High-temperature gas-cooled reactors High Temperature Reactor 10 High-temperature test reactor Irradiated fuel examination laboratory Irradiated microsphere gamma analyzer INET 189 189 189 190 192 193 193 199 199 199 200 200 202 202 205 206 209 212 Institute of Nuclear and New Energy Technology INL Idaho National Laboratory IPyC Inner pyrolytic carbon ITU Institute for Transuranium Elements JMTR Japan Material Test Reactor KuFA Cold finger apparatus (in German) LEU Low-enriched uranium LHTGR Large High Temperature Gas Reactor LTI Low temperature isotropic MOL Middle of life NE-MHTGR Commercial version of NP-MHTGR NGNP Next Generation Nuclear Plant NP-MHTGR New Production Modular High-temperature Gas-Cooled Reactor ORNL Oak Ridge National Laboratory ORR Oak Ridge Research Reactor PIE Postirradiation examination R&D Research and development R/B Release to birth ratio SiC Silicon carbide TRIGA Training research and isotope production, General Atomics TRISO Tristructural isotropic UCO Uranium oxycarbide Uranium dioxide UO2 VHTR Very-high-temperature reactors VXF Vertical experimental facility WAR Weak acid resin TRISO-Coated Particle Fuel Performance 3.07.1 Introduction For all high temperature gas reactors (HTGRs), tristructural isotropic (TRISO)-coated particle fuel forms the heart of the concept Such fuels have been studied extensively over the past four decades around the world, for example, in countries including the United Kingdom, Germany, Japan, United States, Russia, China, and more recently, South Africa In early gas-cooled reactors, the coated particle fuel form consisted of layers of carbon surrounding the fissile kernels Highly enriched uranium (HEU) and thorium carbides and oxides were used as fissile and fertile kernels Ultimately, the carbon layer coating system (termed BISO for bistructural isotropic) was abandoned because it did not sufficiently retain fission products, leading to the development of the current three-layer coating system (termed TRISO for tristructural isotropic) In TRISO-coated fuel, a layer of silicon carbide (SiC) is sandwiched between pyrolytic carbon layers This three-layer system is used to both provide thermomechanical strength to the fuel and contain fission products In addition, for operational and economic reasons, the fuel kernel of choice today is low-enriched uranium (LEU) uranium dioxide (UO2) for the pebble bed design and uranium oxycarbide (UCO) for the prismatic design In both pebble bed and prismatic gas reactors, the fuel consists of billions of multilayered TRISOcoated particles ($750–830 mm in diameter) distributed within fuel elements in the form of circular cylinders (12.5 mm in diameter and 50 mm long) called ‘compacts’ or spheres called ‘pebbles’ (6 cm in diameter) The active fuel kernel is surrounded by a layer of porous carbon, termed ‘the buffer’; a layer of dense carbon, termed ‘the inner pyrolytic carbon layer’; a layer of SiC; and another dense carbon layer, termed ‘the outer pyrolytic carbon layer.’ These collectively provide for accommodation and containment of fission products generated during operation The buffer layer is designed to accommodate fission recoils, volumetric swelling of the kernel, and fission gas released under normal operation The inner pyrolytic carbon layer protects the kernel from reactive chlorine compounds produced during SiC deposition in the chemical vapor deposition (CVD) coater The SiC layer provides structural strength to the particle The outer pyrolytic carbon layer protects the particles during formation of the fuel element Under normal operation, radiation damage causes shrinkage of the pyrolytic carbon layers, which induces compressive stresses in the SiC layer to counteract tensile stresses associated with fission gas release All three layers of 153 the TRISO coating system exhibit low permeability These fuel constituents are extremely stable and are designed not to fail under normal operation or anticipated accident conditions, thereby providing effective barriers to the release of fission products Figure is a montage of TRISO fuel used in both prismatic and pebble bed high-temperature gas reactors Rigorous control is applied at every step of the fabrication process to produce high-quality, very lowdefect fuel Defect levels are typically on the order of one defect per 100 000 particles Specifications are placed on the diameters, thicknesses, and densities of the kernel and layers; the sphericity of the particle; the stoichiometry of the kernel; the isotropy of the carbon; and the acceptable defect levels for each layer Statistical sampling techniques are used to demonstrate compliance with the specifications usually at the 95% confidence level For example, fuel production for German reactors in the 1980s yielded only approximately 100 defects in 3.3 million particles produced This remains the standard for gas-cooledreactor fuel production today.1,2 Irradiation performance of high-quality, lowdefect coated particle fuels has been excellent In Section 3.07.2, a detailed review of the state of the art in irradiation testing, capabilities of existing fission reactors worldwide to irradiate TRISO fuel, and the irradiation behavior of modern TRISO-coated particle fuel around the world will be discussed Testing of German fuel under simulated accident conditions in the 1980s has demonstrated excellent performance Section 3.07.3 describes the accident behavior of TRISO-coated particle fuel largely on the basis of the German database and the plans to perform similar testing for the current generation of TRISOcoated fuels Additional limited testing of TRISOcoated particle fuel performed under air and water ingress events and under reactivity pulses has been reported elsewhere3 and will not be repeated here The outstanding irradiation and accident simulation testing results obtained by German researchers form the basis for fuel performance specifications used in gas-cooled-reactor designs today Specifications for in-service failure rates under irradiation and accident conditions are very stringent, typically on the order of 10À4 and  10À4, respectively Significant research and development (R&D) related to TRISO-coated fuels is underway worldwide as part of the activities of the Generation IV International Forum on Very-High-Temperature Reactors (VHTRs) The focus is largely on extending the capabilities of the TRISO-coated fuel system for higher 154 TRISO-Coated Particle Fuel Performance Pyrolytic carbon Silicon carbide Uranium dioxide or oxycarbide kernel Prismatic Pebble Particles Matrix Compacts Fuel element TRISO-coated fuel particles (left) are formed into fuel compacts (center) and inserted into graphite fuel elements (right) for the prismatic reactor Kernel Buffer layer mm graphite layer Coated particles imbedded in graphite matrix Inner PyC-layer Fuel-free shell SiC-layer Fueled zone Outer PyC-layer Fuel sphere Dia 60 mm Half section TRISO-coated fuel particles are formed into fuel spheres for pebble bed reactor Figure TRISO-coated particle fuel and compacts and fuel spheres used in high temperature gas reactors burnups (10–20%) and higher operating temperatures (1250  C) to improve the attractiveness of hightemperature gas-cooled reactors as a heat source for large industrial complexes where gas outlet temperatures of the reactor would approach 950  C.4 Of greatest concern is the influence of higher fuel temperatures and burnups on fission product interactions with the SiC layer leading to degradation of the fuel and the release of fission products Activities are also underway around the world to examine modern recycling techniques for this fuel and to understand the ability of gas reactors to burn minor actinides.5,6 3.07.2.1.1 BR-2 The Belgian Reactor (BR-2) reactor is a materials test reactor in Mol, Belgium7 that produces very fast (3.5  1014 neutrons cmÀ2 sÀ1 [E > MeV]) and thermal neutron fluxes (1012 neutrons cmÀ2 sÀ1) The facilities have irradiation test rigs ($15 mm ID and 400 mm long) that can be used to irradiate coated-particle gas reactor fuel forms They have adequate flux, fluence, and temperature characterization for the capsule, and have the infrastructure needed for capsule disassembly and postirradiation examination (PIE) The capsule size precludes irradiation of pebbles; however, it could handle approximately six to eight fuel compacts 3.07.2 Irradiation Performance 3.07.2.1 Overview of Irradiation Facilities and Testing This section provides a brief overview of irradiation facilities that are available today to perform TRISOcoated particle irradiations 3.07.2.1.2 IVV-2M The IVV-2M is a 15-MW water-cooled reactor that has been used in Russia for a variety of coatedparticle testing.8 Four different test rigs have been used to test specimens ranging from particles, to compacts, to spheres The coated particle ampoule TRISO-Coated Particle Fuel Performance is a noninstrumented rig that can hold 10–13 graphite disks (15 mm in diameter and mm thick), each of which can hold 50 particles The rig can also hold coated particles in axial holes, 1.2 mm in diameter, and a uniform volume of coated particles, 12–18 mm in diameter and 20–255 mm high, in a graphite matrix Another rig, termed a ‘CP hole,’ is 27 mm in diameter and that can handle six to eight capsules A third rig, identified as ASU-8, is a 60-mm hole that can handle three compacts The largest channel available is Vostok, which is 120 mm in diameter and contains four cells All of these rigs can irradiate fuel at representative temperatures, burnups, and fluences for HTGRs There is a large degree of flexibility in the testing options at IVV-2M Their rigs can handle particles, compacts, and spheres 3.07.2.1.3 HFR Petten The High Flux Reactor (HFR) in Petten, Netherlands, is a multipurpose research reactor with many irradiation locations for materials testing.9 The HFR has two different types of irradiation rigs/locations in the facility: one that can accommodate compacts and another that can accommodate spheres Rigs for spheres are multicell capsules, 63–72 mm in diameter that can handle 4–5 spheres in up to separate cells For compacts rigs/locations are $32 mm in diameter and 600 mm in useful length They can handle three or four parallel channels of compacts For the threechannel configuration, approximately 30 compacts could be irradiated in the rig There is a large axial flux gradient across the useable length (40% spread maximum to minimum) that must be considered in the design of any experiment 3.07.2.1.4 HFIR The High Flux Isotope Reactor (HFIR) at Oak Ridge National Laboratory (ORNL) is a light-water-cooled, beryllium-reflected reactor that produces high neutron fluxes for materials testing and isotope production.10 Two specific materials irradiation facilities locations are available for gas reactor fuel testing: (a) the large RB positions (eight total) that are 46 mm in diameter and 500 mm long, and can accommodate capsules holding up to 24 compacts (three in each graphite body, eight bodies axially) in a single swept cell; and (b) the small vertical experimental facility (VXF) positions (16 total) that are 40 mm in diameter and 500 mm long, and can accommodate capsules holding up to 16 compacts (eight in each graphite body, two bodies axially) in a single swept cell Capsules can be irradiated in the lower flux small 155 VXF positions and then moved to the higher flux removable beryllium positions Neither of these positions can accommodate pebbles A third facility, the large VXF positions (six total), are farther out in the reflector (and therefore have lower fluxes), but are 72 mm in diameter and also 500 mm long As with the HFR, there is a large axial flux gradient that must be considered in the design of any experiment in any of these facilities 3.07.2.1.5 ATR The Advanced Test Reactor (ATR) at Idaho National Laboratory (INL) is a light-water-cooled, berylliumreflected reactor fuel in a four-leaf clover configuration to produce high neutron fluxes for materials testing and isotope production.11 The clover leaf configuration results in nine very high flux positions, termed ‘flux traps.’ In addition, numerous other holes of varying size are available for testing Several positions can be used to irradiate coated-particle fuel The 89-mm-diameter medium I position (16 total) and the 100–125-mm-diameter flux traps can accommodate pebbles Specifically, the use of a medium I position early in the irradiation, required because of the enrichment of the fuel, followed by transfer of the test train to the northeast flux trap can provide irradiation conditions representative of a pebble bed reactor Approximately 10–12 pebbles in five or six individually swept cells can be envisioned in the test train The large B positions in ATR (four total) are 38 mm in diameter and 760 mm in length They can accommodate six individually swept cells, with two graphite bodies per cell, containing up to three 2-in long compacts per body Thus, 36 full-size US compacts can be irradiated in this location Of special note, here is the very flat burnup and fluence profile available axially in the ATR over the 760 mm length This allows for nearly identical irradiation of large quantities of fuel 3.07.2.1.6 SAFARI The SAFARI Reactor in Pelindaba, Republic of South Africa, is an isotope production and research reactor.12 The core lattice is an  array, consisting of 28 fuel assemblies, control rods, and a number of aluminum and beryllium reflector assemblies The reactor is cooled and moderated by light water and operates at a maximum power level of 20 MW In-core irradiation positions include six high-flux isotope production positions: two hydraulic, two pneumatic, and two fast transfer systems that are accessible during operation Several other irradiation 156 TRISO-Coated Particle Fuel Performance positions can also be accessed when the reactor is shut down A large poolside facility allows for a variety of radiation applications An intermediate storage pool and a transfer canal allow for easy and safe transport of activated materials to a hot cell 3.07.2.1.7 TRISO-coated particle fuel irradiation testing The historical experience in irradiation testing of coated particle fuels suggests that multicell capsules wherein fuel can be tested in separate compartments under different temperature, burnup, and fluence conditions allow for tremendous flexibility and can actually save time and money in an overall fuel qualification program Although there are differences in details of the test trains used in each of the reactors, they share a number of important similarities in the state of the art with irradiation testing of this fuel form In this section, these important similarities are presented to highlight the technical considerations in executing this type of testing Because of the differences in neutron flux spectrum between a gas reactor and a light-water materials test reactor, simultaneous matching of both the rate of burnup and the rate of accumulation of fast neutron fluence is difficult to achieve In addition, the traditional 3-year fuel cycle of high-temperature gas reactors makes real-time irradiation testing both timeconsuming and an expensive part of an overall fuel development effort To overcome these shortcomings, irradiations in material test reactors have historically been accelerated relative to those in the actual reactor Usually, the time acceleration is focused on achieving the required burnup in a shorter time than would be accomplished in the actual reactor, with the value of the fast fluence left as a secondary variable that must fall between a minimum and maximum value The level of acceleration can also impact the potential for fuel failure during irradiation The level of acceleration at a given test reactor power, coupled with fuel loading in the experiment, results in a power density for the fuel specimen in the experiment The power density peaks at the beginning of the irradiation when the fissile content is highest and decreases as the fissile material is burned out of the fuel As the level of acceleration increases, the temperatures in the fuel kernels increase above that in the fuel matrix because of the thermal resistances associated with the coatings of the particle,13 and the potential for high temperature, thermally driven failure mechanisms to play a deleterious role in fuel performance becomes more important As discussed in Section 3.07.2.7, the irradiation performance database suggests that modest levels of acceleration (1.5–3Â) appear to be acceptable without jeopardizing fuel performance in the irradiation, and should be a baseline requirement for future gas reactor irradiations This acceleration level can be translated into a maximum power per fuel body or power per particle that can be used by experimenters in the design of the irradiation capsule Given the limitations of materials test reactors around the world, the TRISO-coated particle irradiation database contains results from tests conducted under a range of accelerations Successful German TRISO-coated particle fuel irradiations in the European HFR-Petten reactor were conducted using an acceleration of less than a factor of three By comparison, other German irradiations in the Forschungzentrum Reaktor Juelich (FRJ) reactor at Juălich had a neutron spectrum that was too thermalized This resulted in the fuel receiving too little fast fluence to be prototypic of a high-temperature gas reactor Similarly, historic US irradiations in ORNL’s HFIR reactor had too high a thermal flux resulting in significant burnup acceleration of the irradiation On the basis of these considerations, the large B positions (38 mm diameter) in the ATR (see Figure 2) were chosen for the US Department of Energy’s (DOE) Advanced Gas Reactor (AGR) Program fuel irradiations because the rate of fuel burnup and fast neutron fluence accumulation in these positions provide an acceleration factor of less than three times that expected in the hightemperature gas reactor 3.07.2.1.8 Thermal and physics analysis considerations Given the complexity of the capsules currently being designed, the extensive review by safety authorities of the thermo-mechanical stresses, and the importance of each capsule in terms of irradiation data for fuel qualification, three-dimensional physics and thermal analyses are essential in irradiation capsule design These analyses are critical to ensure that the fuel reaches the intended burnup, fluence, and temperature conditions To achieve high burnups with these fuels requires detailed physics calculations to determine the time to reach full burnup Given the concerns about severely accelerated irradiations, it is not uncommon for such irradiations to take approximately years to reach full burnup in LEU TRISO-coated particles In addition, because thermocouples should not be attached directly to the fuel, thermal analysis is used to calculate the fuel temperature during the irradiation TRISO-Coated Particle Fuel Performance 157 North ON-8 ON-9 ON-10 ON-11 ON-12 ON-3 ON-4 ON-5 ON-6 ON-1 Fuel elements ON-7 ON-2 I-19 I-20 I-1 I-2 H positions I-3 I-4 I-18 I-17 Small B position I-5 I-6 I-16 I-15 I-7 I-8 I-14 I-13 I-12 I-11 I-10 OS-6 OS-1 I-9 East large B position location for AGR-1 In-pile tube OS-2 OS-3 OS-4 OS-5 OS-7 OS-8 OS-9 OS-10 OS-11 OS-12 Control drum OS-13 OS-14 OS-15 OS-16 OS-17 I positions OS-18 OS-19 OS-20 OS-21 OS-22 Figure Schematic of ATR showing fuel and select irradiation positions Examples of a test train for fuel compacts used in INL’s ATR and the pebbles used in HFR-Petten are shown in Figures and respectively These irradiation capsules have extensive instrumentation to measure temperature, burnup, and fast fluence at multiple locations in the test train Traditional commercial thermocouples have been used extensively in past irradiations, but thermocouples can suffer from drift and/or de-calibration in the reactor Redundancy in thermocouple measurements is another consideration in light of the low reliability of thermocouples at high temperatures and long times in neutron fields typical of TRISO-coated particle fuel irradiations Melt wires are inexpensive and have been used as a backup to thermocouples where space was available in the capsule However, melt wires only indicate that a certain peak temperature has been reached, and not the time of that peak Direct temperature measurements of the coated particles are problematic because direct metal contact (e.g., thermocouple wires or sheaths) with the fuel element is not recommended as the metals can attack the TRISO fuel coatings Thus, temperatures must be calculated on the basis of thermocouples located elsewhere in the capsule Thermocouples are generally located as close as possible to the fuel body to minimize the uncertainties on the calculated fuel temperatures related to irradiationinduced dimensional change and thermal conductivity changes of the materials in the capsule Encapsulating the fuel element in a graphite sleeve or cup and inserting thermocouples into the graphite has been used successfully in many designs The high conductivity of graphite minimizes the effect of irradiation-induced dimensional changes on the calculated fuel temperature 158 TRISO-Coated Particle Fuel Performance Gas line Fuel stack Thermocouple SST holder Thermocouple Purge gas pipe Radiation shields Hafnium shield Graphite cup Test fuel element Capsule spacer nub Figure Schematic of capsule used in US INL AGR program Historically, metal sleeves have not been allowed to touch fuel elements because of past experiences in which SiC was attacked by transition metals (Fe, Cr, and Ni) Although quantitative data on transport rates of such metals into the fuel element and corrosion rates of the SiC are unknown, or mm thickness of graphite between the fuel element and the metallic components (e.g., graphite sleeve) has been found to be effective in minimizing the potential for interaction These irradiation experiments typically include both thermal and fast fluence wires A number of different fluence wires have been used successfully to measure thermal and fast neutron fluences in coated particle fuel irradiations The specific type of wire to be used will depend on the measurement need (fast or thermal), the temperature it will experience during the irradiation, and compatibility with the material of encapsulation Quartz encapsulation is not recommended for high-temperature, high-fluence applications Neutronically, transparent refractories (e.g., vanadium) are a good alternative material of encapsulation Inert gas filling of the flux wire Figure Schematic of pebble irradiation experiment used by the Germans encapsulation is recommended to ensure no oxygen interaction with the flux wire Although fission chambers and self-powered neutron detectors have been used extensively in other reactor irradiations, they may not be practical in the space-constrained capsules expected for TRISO-coated particle fuel qualification tests 3.07.2.1.9 Gas control system considerations Automated gas control systems – designed to change the gas mixture in the experiment to compensate for the reduction in fission heat and changes in thermal conductivity with burnup – minimize human operator error and have proven to be a reliable method of thermal control during these long fuel irradiations The temperature of each experiment capsule is controlled by varying the mixture of two gases with differing thermal conductivities in a small insulating gas jacket between the specimens and the experiment containment A mixture of helium and argon has been used in the past and provides a wide temperature control band for the experiments Unfortunately, TRISO-Coated Particle Fuel Performance argon cannot be used in fuel experiments where online fission product monitoring is used because the activated argon will reduce detectability of the system Therefore, helium and neon are used instead Computer-controlled mass flow controllers are typically used to automatically blend the gases (on the basis of feedback from the thermocouples) to control temperature The gas blending approach allows for a very broad range of control Automatic gas verification (e.g., by a thermal conductivity analyzer) has been implemented in some experiments to prevent the inadvertent connection of a wrong gas bottle Gas purity is important and an impurity cleanup system should be implemented during each irradiation Flow rates and gas tubing should be sized to minimize transit times between the mass-flow controllers and the experiment, as well as between the experiment and the fission product monitors 3.07.2.1.10 FPMS considerations In addition to thermal control, sweep gas is used to transport any fission gases released from the fuel to a fission product monitoring system (FPMS) A number of techniques have been used historically to quantify the release of fission gases from the fuel in these irradiation capsules Techniques include gross gamma monitoring, online gamma spectroscopy, and offline gamma spectroscopy of grab samples Online gross gamma monitoring of the effluent gas in the experiment using ion chambers and sodium iodide detectors is an excellent means to capture any dynamic failures of the coated particles associated with the instantaneous release upon failure Grab samples can provide excellent noble gas isotopic information The temporal resolution and the number of isotopes that can be measured depend on the frequency of the grab samples and the delay time between acquisition of the grab sample and offline analysis Weekly grab samples are typical in most irradiations, although daily or even hourly samples are possible if failure has occurred, assuming operation and associated analysis costs are not too high Typical isotopes to be measured include 85mKr, 87Kr, 88 Kr, 131mXe, 133Xe, and 135Xe Measurement of verylong-lived isotopes (e.g., 85Kr) would be useful in elucidating fission product release mechanisms from the kernel, but would also require waiting for the decay of the shorter lived isotopes in the sample Online gamma spectroscopy, although the most expensive in terms of hardware costs, can provide the most detailed real-time information with detailed isotopic spectrums as often as necessary subject to data storage limitations of the system An example of the system used for the US AGR program is shown in Figure With such systems, transit times from the experiment to the detector should be minimized to allow measurement of short- and medium-lived isotopes, but must remain long enough to allow decay of Temperature control gas mixing system Vessel wall He Filter Silver zeolite Capsules in-core Fission product monitoring system Grab sample Figure Integrated fission product monitoring system used in US AGR program irradiations H and V exhaust Ne Particulate filters 159 160 TRISO-Coated Particle Fuel Performance any short-lived isotopes associated with the sweep gases ($2–3 min) With this delay time, 89Kr, 90Kr, 135m Xe, 137Xe, 138Xe, and 139Xe should also be capable of being measured online Measurements of xenon gas-release during reactor outages are recommended to provide information on iodine release behavior from the decay of xenon precursors Multiple options for fission gas-release measurements should be considered for long irradiations where reliability of the overall fission gas measurement system can be a concern Redundancy is also recommended for online systems so that failure of a spectrometer does not jeopardize the entire experiment On the basis of the online concentration data, a release-to-birth ratio (R/B), a key parameter used in reactor fuel behavior studies,14 can be calculated and provide some insight into the nature of any particle failures Because these instruments are online during the entire irradiation, a complete time history of gas release is available Gas release early in the irradiation (i.e., from the start of the irradiation) is indicative of initially failed particles or contamination outside of the SiC layer Release later during the irradiation is indicative of in situ particle failure The timing of the failure data can then be correlated to temperature, burnup, and/or fluence that can be used when coupled with PIE to determine the mechanisms responsible for the fuel failure 3.07.2.2 German Experience Previously, particle fuel development was conducted by German researchers in support of various HTGR designs that employed a pebble bed core These reactors were intended to produce process heat or electricity The sequence of fuel development used by German researchers followed improvement in particle quality and performance and was largely independent of developments in reactor technology German fuel development can be categorized according to the sequence of fuels tested as provided in Table German irradiation test conditions generally covered projected fuel operating conditions, where fuel was to reach full burnup at fast fluences of 2.4  1025 n mÀ2 and operate at temperatures up to 1095  C for process-heat applications and up to 830  C for electrical production applications With the exception of irradiation duration, the various experiments performed bounded expected nominal conditions or were purposely varied to meet other test objectives In order to obtain results in a Table German particle fuel development sequence Date of design consideration Fuel form 1972 1977 BISO coated (Th, U)O2 Improved BISO coated (Th, U)O2 TRISO-coated UCO fissile particles with ThO2 fertile particles TRISO-coated (Th, U)O2 LEU TRISO-coated UO2 1981 timely manner, tests conducted by German researchers were generally accelerated by factors of 2–3 The following sections present irradiation experiment summaries for fuels of ‘modern’ German design.1 For these experiments, this definition extends to high-enriched (Th, U)O2 TRISO-coated particles fabricated since 1977, and low-enriched UO2 TRISOcoated particles fabricated since 1981 Table provides the physical attributes of the fuel used in these tests Mixed oxide fuel test summaries are presented first, followed by the UO2 tests 3.07.2.2.1 R2-K12 and R2-K13 The R2-K12 and R2-K13 cells were irradiated in the R2 reactor at Studsvik, Sweden The main objective of the R2-K12 experiment was to test mixed oxide (Th, U)O2 and fissile UC2/fertile ThO2 fuel elements, whereas for R2-K13, the main objective was to test mixed oxide (Th, U)O2 fuel elements and supply fuel for subsequent safety tests In R2-K12, four full-size spherical fuel elements were irradiated in four independently gas-swept cells Two cells contained mixed oxide fuel spheres, while the other two contained fissile/fertile fuel spheres As the German researchers did not develop the twoparticle fissile/fertile system further, only the mixed oxide results were reported R2-K13 was a combined experiment with the United States Four independently gas-swept cells were positioned vertically on top of one another The top and bottom cells each contained a full-size German fuel sphere The middle two cells contained US fuel and will be discussed in Section 3.07.2.3 Configuration and irradiation data from both experiments are given in Tables and Cold gas tests on each fuel sphere during PIE indicated that all the particles had remained intact in both R2-K12 and R2-K13 These tests are conducted after the fuel has been stored (for $14 days) at room temperature and a quasi-steady-state release of fission gas has been reached The fuel is then swept with a carrier gas that is monitored for various fission TRISO-Coated Particle Fuel Performance 100 1800 ЊC the range 1900–2500  C is thermal decomposition of the SiC layer.50 10–1 1700 ЊC 3.07.3.3 fractional release 10–2 137Cs 199 1600 ЊC 10–3 Five compacts 10–4 1600 ЊC 10–5 Five fuel elements 10–6 10–7 50 100 150 200 250 Heating time (h) 300 350 Figure 36 Cesium release during heating of spherical fuel elements (1600  C) and compacts (1600–1800  C) Reproduced from Schenk, W.; et al J Nucl Mater 1990, 171(1), 19–30 European Experience Several high-temperature accident tests have been performed at the new KuăFA installation at ITU.51 Prior to these tests, the upgraded system was tested to verify furnace operation, fission gas system performance, and condensation plate collection efficiency It was found that the collection efficiency of the plate for Cs was 70% (Ỉ16%), which is in agreement with the efficiency determined previously at Juălich during earlier work with the furnace.38 The new accident tests on irradiated fuel include several fuel elements from the AVR reactor, from the HFR K6 irradiation test (proof test for HTR Modul reactor fuel element design), and from the more recent EU1bis irradiation test Table 60 provides information and the irradiation history of the AVR and HFR K6 fuel elements tested to date Because of the age of the fuel tested, no 110mAg data were collected in the latest accident tests Testing on four HFR-EU1bis spheres has been completed, but data from these experiments have not yet been published 3.07.3.3.1 AVR 73/21 cesium at times greater than 200 h in Figure 35, and temperatures above 1600  C) Ceramographic sections shown in Figure 37 show evidence of increasing degradation in the SiC layer for longer times at 1600  C and higher burnup, the most degraded being the SiC in sphere HFR-K3/1 Microprobe profiles through particles after heating, as shown in Figure 38, show the buildup of fission product palladium at the IPyC/SiC interface in spheres HFR-K3/1 and HFR-K3/3 It is hypothesized that corrosion by palladium degrades the SiC, leading to accelerated diffusion of cesium through grain boundaries.45 The distribution of metallic fission products averaged over a number of UO2 TRISO fuel element spheres examined after accident testing is shown in Table 59 At 1600  C,49 the kernel and the coatings are equally important in holding up cesium, while the kernel is the principal reservoir for strontium and silver At 1800  C, cesium and silver are contained principally in the coatings, whereas strontium is retained in the kernel The primary mechanism for coating failure and fission product release at extreme temperatures in Testing of the AVR 73/21 fuel element was the first use of the upgraded KuăFA with irradiated fuel and was primarily conducted as a shakedown test of the equipment under hot conditions The sphere was heated to 1600  C for h, followed by a ramp to 1800  C and hold for an additional h The 85Kr release during this relatively short test was below the detection limit of the system Two cold plates were exchanged during the test, but a problem was encountered with 137Cs contamination of the plate container from the hot cell, which interfered with analysis of released 137Cs on the plate and prompted a change in procedures to eliminate the carry-over contamination issue on subsequent tests 3.07.3.3.2 AVR 74/18 The second accident test was performed on the AVR 74/18 sphere The test took place in two parts: the first phase involved heating at 1600  C for 100 h, followed by a second phase of heating at 1800  C for an additional 100 h Unplanned temporary furnace shutdowns were encountered during each heating cycle (one per cycle) due to operational issues with the furnace system, but 200 TRISO-Coated Particle Fuel Performance 100 mm 1600 ЊC, 160 h 50 mm FCs1374ϫ10–5(FRJ2–K13/2;8%FIMA) 100 mm 1600 ЊC, 500 h 1600 ЊC, 500 h 20 mm 20 mm FCs1372ϫ10–5(AVR71/22;3.5%FIMA) FCs1371ϫ10–4(HFR–K3/1;7.7%FIMA) Figure 37 Ceramographic sections through particles heated at 1600  C (complete particle followed by enlarged views of SiC layers) showing increasing degradation of the SiC layer with increasing burnup Reproduced from Schenk, W.; et al J Nucl Mater 1990, 171(1), 19–30 these not appear to have significantly affected the test results No particle failures were detected during this test Fractional releases at the end of the test were $6  10À6 for 85Kr and $8  10À6 for 137Cs 3.07.3.3.3 HFR K6/3 This heating test was carried out in four stages: 1600  C for 100 h, 1700  C for 100 h, 1800  C for 100 h, and 1800  C for 300 h The temperature was returned to room temperature in between stages 85Kr fractional release was low ( 0.1 MeV (cmÀ2) GLE-3 A3-27 16 400 10 UO2 9.82 2.5 235 February 1984 $700  1021  1020 GLE-3 A3-27 16 400 10 UO2 9.82 4.8 480 February 1985 $700 4.15  1021  1020 GLE-4 similar A3-27 14 600 9.44 UO2 10.6 9.7 633 May 1993 $1140 2.5  1021 4.8  1021 GLE-4 similar A3-27 14 600 9.44 UO2 10.6 9.3 633 May 1993 $1140 2.5  1021 4.6  1021 Source: Freis, D.; et al Post irradiation test of high temperature reactor spherical fuel elements under accident conditions In 4th International Topical Meeting on High Temperature Reactor Technology, Washington, DC, Oct 1–4, 2008 3.07.3.4 US Experience and Future Plans 0.1 2000 1800 1400 1E-3 1200 1E-4 1000 800 1E-5 600 Temperature (ЊC) Kr 137 Cs 400 85 200 0 100 200 300 400 500 600 700 Fractional release 0.01 1600 Temperature (ЊC) room temperature at the end of the 1600  C test, but several interruptions of the experiment during the first stage occurred because of system malfunctions The results of this heating test are shown in Figure 40 The 85Kr fractional release was below the detection limit of the system until $90 h into the 1800  C cycle, when the release increased up to $10À7 Approximately 120 h into the 1800  C cycle, another continuous release was observed that reached approximately 10À5 by the end of the test 137Cs was released fairly continuously throughout the test up to a  10À3 fractional release at the end of the test In general, these initial tests demonstrated excellent fuel performance at 1600 and 1700  C with zero particle failures Higher 137Cs and 85Kr releases were observed at 1800  C, with some particle failures occurring for the sphere HFR K6/3 1E-6 1E-7 800 Time (h) Figure 39 Results of heating test HFR K6/3 at ITU Reproduced from Freis, D.; et al Post irradiation test of high temperature reactor spherical fuel elements under accident conditions In 4th International Topical Meeting on High Temperature Reactor Technology, Washington, DC, USA, October 1–4, 2008 3.07.3.4.1 Past experience Very little work has been done on accident testing of US fuels as most irradiations produced significant irradiation-induced particle failure fractions However, postirradiation anneals of long durations at temperatures up to 1500  C were performed to accelerate diffusive fission product releases from a variety of fuel types.52 UO2, UC2, UCO, UO2*(1), and UO2*(2) fuel particles were irradiated in the HRB-15B capsule in HFIR.52 In the UO2*(1) fuel, the kernel was coated with a ZrC layer, while in the UO2*(2) fuel, ZrC was dispersed in the buffer layer surrounding the kernel These changes were intended to control free oxygen released during fission, which should improve fission product retentiveness The fuel burnup was in the range of 21–25% FIMA and the fast neutron fluence was in the range of 3.4 – 5.5  1025 n mÀ2 The irradiation was accelerated with a residence time of 169 effective full power days Only the fission product release data at 1500  C are discussed here, as cesium was not released at the lower temperatures No fission product releases were measured at any temperature from UO2*(1) fuel particles Ten particles of each fuel type were annealed for 11 866 h at 1500  C Integral releases for each 10-particle batch were measured from individual particles by gamma counting each particle before and after the test and by periodic gamma monitoring of fission product collectors TRISO-Coated Particle Fuel Performance during the anneal as a function of time The agreement of the integral releases from each 10-particle batch by these two methods was excellent Cesium was released from only the UO2 and UC2 fuel particles as is shown in Figure 41 These same two fuel batches released the greatest fractions of silver as illustrated in Figure 42 The time signatures 2000 0.01 1800 Temperature (ЊC) 1400 1E-4 1200 1E-5 1000 800 1E-6 600 400 Fractional release 1E-3 1600 1E-7 200 1E-8 0 50 100 150 200 250 Time (h) 300 350 Temperature (ЊC) 85 Kr Cs 137 Figure 40 Results from heating test HFR K6/2 at ITU Reproduced from Freis, D.; et al Post irradiation test of high temperature reactor spherical fuel elements under accident conditions In 4th International Topical Meeting on High Temperature Reactor Technology, Washington, DC, USA, October 1–4, 2008 203 of the releases of cesium and silver from the UO2 fuel particles in Figures 41 and 42 indicate a diffusion release mechanism through the SiC layer However, the release of cesium from the UC2 fuel batch is sudden in Figure 41, and the release of silver shows a rapid increase at the same time as the sudden release of cesium, as pointed out in Figure 43 The distribution of fission product releases among particles within the fuel batches in Table 61 indicates that the release of cesium from the UO2 fuel particles is from two of the ten particles and from only one particle in the UC2 fuel batch This same table shows that the release of silver was 100% from the UO2 fuel batch, and 82% from the UC2, with of the 10 UC2 particles releasing 100% of their silver inventories, two particles releasing 85–95%, one particle releasing 50%, and one particle retaining 100% Particle-toparticle variations in fission product release dominate the behavior implied by the data shown in Table 61 The microstructures in Figure 44 show that the SiC layer in the UO2 batch exhibits large columnar grains, whereas the UCO batch exhibits a strong laminar grain structure in the SiC The UC2 and UO2*(1) batches exhibit laminar structures in the SiC that are somewhat weaker than in the UCO batch The results in Table 61 indicate that silver release at 1500  C is greatest (100%) in the case of columnar SiC, least (3%) for strongly laminar SiC, and intermediate (82%) for somewhat less strong laminar SiC Although Cs was released from only three of the 50 particles annealed at 1500  C, two Cesium release (%) 102 101 100 134Cs 137Cs 10−1 10 20 30 Annealing time (Ms) 40 50 Figure 41 Release of Cs from various types of TRISO-coated fuel particles at 1500  C Reproduced from Bullock, R E J Nucl Mater 1984, 125(3), 304–319 204 TRISO-Coated Particle Fuel Performance 102 UO2 UC2 Silver release (%) UO2* (2) 101 Release from UO2* (1) = UCO 100 10−1 10 20 30 Annealing time (Ms) 50 40 Figure 42 Release of 110Ag from various types of TRISO-coated fuel particles at 1500  C Reproduced from Bullock, R E J Nucl Mater 1984, 125(3), 304–319 102 Ag Cs Release (%) 101 100 10−1 10 20 30 Annealing time (Ms) 40 50 Figure 43 Abrupt 10% increase in 110Ag release from UC2 particles at 1500  C when one of the ten test particles released its entire Cs inventory particles had columnar SiC and one had a somewhat weak laminar SiC The sensitivity of cesium release to SiC grain structure was recognized in Myers53 where the diffusivity of cesium through columnar SiC was given as an order of magnitude greater than through laminar SiC The diffusion equations from Myers53 are accessible in Table A-3 of IAEA.3 As shown in Figure 45, releases of europium are greatest (37–46%) for the fuel batches containing UC2 in the kernel, compared with fuel batches containing only UO2 in the kernel (9–16%) As shown in Table 61, cerium release is 45% in UC2, only $1% in UCO, and nil in UO2 particles These behaviors are related to the thermodynamics of rareearth oxides and carbides, according to Homan et al.,54 where oxides formed in UO2 (such as Eu2O3 and Ce2O3) are less likely to escape from the kernel than are the more mobile rare-earth carbides formed in UC2 In UCO, europium forms a carbide and cerium forms an oxide.54 TRISO-Coated Particle Fuel Performance Table 61 205 Distribution of fission product release within particle batches during postirradiation annealing Annealing temperature ( C) TRISO particle typea 1500 UC2 with laminar SiC ¼ 0% ¼ 99% 10 ¼ 12% 1500 UO2 with columnar SiC 1500 UC0.4O1.6 with laminar SiC ¼ 0% ¼ 99% 10 ¼ 24% 10 ¼ 0% 1500 UO2*(2) with laminar SiC 10 ¼ 0% 1350 1350 UC0.4O1.6 with laminar SiC UO2*(2) with laminar SiC 10 ¼ 0% 10 ¼ 0% 1200 1200 UC0.4O1.6 with laminar SiC UO2*(2) with laminar SiC 10 ¼ 0% 10 ¼ 0% Release breakdown from the ten particles within a test batch for: 134 137 Cs Cs 110m Ag ¼ 0% ¼ 50% 85% < < 95% ¼ 100% 10 ¼ 82% 10 ¼ 100% ¼ 0% 10% < < 20% 10 ¼ 3% ¼ 0% 70% < < 80% 10 ¼ 27% 10 ¼ 0% ¼ 0% 45% < < 55% 10 ¼ 19% 10 ¼ 0% 10 ¼ 2%b 154 144 15% < < 25% 45% < < 55% ¼ 100% 10 ¼ 46% Uniform release of 16% 12% < < 18% 18% < < 25% 70% < < 80% ¼ 99% 10 ¼ 45% 10 ¼ 0% Uniform release of 37% 10 ¼ 1%b ¼ 0% 0% < < 10% 15% < < 25% Uniform release of 23% 10 ¼ 4%b 10 ¼ 0% 10 ¼ 0% 10 ¼ 0% Uniform release of 6% 10 ¼ 0% 10 ¼ 0% 10 ¼ 0% Eu Ce There was zero release within about Ỉ5% as determined from individual particle counting before and after annealing for all isotopes from each of the ten particles in all test combinations not listed, that is, UO2*(1) at all temperatures, and UC2, UO2, and UC0.4O1.6 at 1350 and 1200  C As no release on collectors was detected at the 0.01% level from the combined ten particles within each of these test batches, it can be assumed that release from any one of these particles was certainly