Inside the Factory for Nuclear Pebbles
X-energy’s TX-1 plant will turn a liquid uranium feed into billions of ceramic-coated particles, then press them into graphite spheres built to circulate through a reactor.
TRISO particles can be embedded in graphite spheres for pebble-bed reactors. These are generic TRISO fuel pebbles, not documented TX-1 production. Credit: U.S. Department of Energy. Seven Grams of UraniumA finished Xe-100 fuel pebble is a black graphite sphere about six centimetres across and roughly 200 grams in weight. It contains around 18,000 coated fuel particles and about seven grams of uranium, obtained by dividing TX-1’s planned five tonnes of annual uranium throughput among 700,000 pebbles. Discharge burnup measures the cumulative heat extracted per tonne of uranium before the fuel leaves service. At the Xe-100’s expected 165 gigawatt-days per tonne, those seven grams correspond to about 28 megawatt-hours of heat and roughly 11 megawatt-hours of electricity. On a heat-content basis, one pebble is close to four tonnes of bituminous coal. TX-1 is under construction in Oak Ridge, Tennessee. Clark Construction is beginning the interior build-out and equipment installation inside a 214,000-square-foot process and administration building, while an adjacent building will prepare the graphite matrix powder. X-energy expects the plant to make 700,000 pebbles containing five metric tonnes of uranium each year. At 18,000 particles per pebble, the coating lines would handle about 12.6 billion fuel particles a year. Building a Fuel ParticleFabrication begins with high-assay low-enriched uranium, or HALEU, in uranyl-nitrate solution. X-energy specifies 15.5% uranium-235 for steady reactor operation, compared with less than 5% in conventional light-water-reactor fuel. In the French fuel chain described in France: The European Nuclear Powerhouse, plants press uranium oxide into pellets and stack them inside metal tubes. In the process developed for X-energy at ORNL, a vibrating nozzle meters a mixture of uranyl nitrate, finely dispersed carbon and other ingredients into droplets, which gel into spheres as they fall through heated silicone oil. The wet gel sphere already sets the geometry of the eventual kernel, although it will contract as water leaves and its chemistry changes under heat. Oak Ridge National Laboratory sorted the spheres into narrow size bands with roller micrometers and modelled the relationship between their starting diameter and the finished kernel. It also tested carbon dispersions because an uneven mixture can produce different oxide-to-carbide ratios inside kernels of the same external size. Washed and dried spheres pass through a controlled furnace sequence. Uranium trioxide is reduced to uranium dioxide, then part of the dioxide reacts with carbon to form uranium carbide. In the finished oxycarbide kernel, that carbide binds oxygen released during irradiation that could otherwise form carbon monoxide at high temperature. ORNL’s development rig combined controlled gas flow, carbon-monoxide monitoring and temperatures up to 2,000°C, in batches of up to six grams, about 30,000 gel spheres at a time. Its crucibles were designed to keep carbon from leaving the kernels during conversion. Trials also found an occasional uranium carbide skin and oxide rind on converted kernels, artefacts whose timing and cause ORNL set out to trace by pulling kernels out at intermediate points in the sequence. The kernels next enter a fluidised-bed chemical-vapour-deposition furnace. Gas rising through a conical graphite chamber carries them in a hot particle fountain while four layers are deposited in sequence: a porous carbon buffer, inner pyrolytic carbon, silicon carbide and outer pyrolytic carbon. The layers grow from gases reacting on each moving kernel at 1,250 to 1,500°C, so time, temperature, gas concentration and particle circulation jointly determine their thickness. A production-scale US reference batch used 425-micrometre kernels and produced coated particles about 0.85 millimetres across, the size of a poppy seed. X-energy describes its proprietary particles more broadly as roughly one millimetre in diameter.
Reference geometry from ORNL acceptance testing of production-scale AGR-5/6/7 fuel. X-energy’s final commercial coating specification is proprietary.
A generic TRISO particle places its containment layers around an individual fuel kernel. Credit: U.S. Department of Energy. Uniform coating depends on the motion of the particles through the furnace. Every kernel must cross comparable gas concentrations and temperatures, while dead zones or an unstable fountain can leave abnormal layers on a small fraction of a batch. During pilot development, ORNL prepared a scaled-up coater to raise the batch from 150 grams to about 1.5 kilograms. Its 2017 report for X-energy lists a multi-inlet gas distributor in development for the wider chamber, with computational-fluid-dynamics models of particle velocity, pressure drop and dead spots under way. A wider chamber changes residence times and the shape of the fountain.
Individual TRISO particles at a 500-micrometre scale. “Individual particles”, Idaho National Laboratory, CC BY 2.0. Around the coater sits a compact chemical plant. In a representative production unit, a resistive graphite heater surrounds the chamber inside a water-cooled shell. Acetylene and propylene feed the carbon layers, while methyltrichlorosilane, carried in hydrogen, supplies silicon and carbon for the silicon-carbide layer. Gaseous hydrogen chloride is the main by-product of that deposition, so the off-gas requires corrosion-resistant handling and scrubbing. After coating, the apparatus cools and the finished particles are drained through the bottom of the chamber. Turning Particles into PebblesBefore pressing, a flow coater rolls each particle through fine graphite powder and resin. The ceramic shells are hard but thin, so the softer overcoat prevents direct particle-to-particle loads and gives each one a surface that can bind into the wider graphite matrix. Representative pebble processes use an overcoat about 200 micrometres thick. Pilot work for X-energy adjusted the powder feed and liquid spray to control adhesion, powder loss and thickness variation. Graphite conducts heat away from the particles and moderates neutrons inside the reactor, slowing them to the speeds at which uranium-235 fissions most readily. On the production line, the powder must flow around thousands of particles, consolidate evenly and retain enough strength to protect them. ORNL screened at least ten varieties each of natural graphite, synthetic graphite and resin for impurities, density, dimensional change on firing and compressive strength. Compressive strength carried the most weight, because the matrix’s first job is to protect the particles embedded in it. In the pebble design X-energy has published, the overcoated particles and additional matrix material first form a fuelled core about five centimetres across, pre-moulded in silicone rubber moulds at 5 to 30 megapascals. More graphite then extends the sphere to six centimetres, and isostatic pressing at about 300 megapascals consolidates a fuel-free shell roughly five millimetres thick. ORNL’s development work for X-energy compared full isostatic pressing, which applies pressure around the whole body, with the German-style quasi-isostatic method and its flexible mould, to find which is more economical at high throughput. ORNL made tooling quickly by 3D-printing negative patterns and casting mould components around them; trapped air was the first problem it had to solve. X-ray tomography of early surrogate cores then located each of the roughly 30,000 particles inside and checked for damage or displacement after pressing.
Cross-section of a cylindrical fuel compact containing TRISO particles. Pebble manufacture uses a different final geometry, with the same requirement to distribute particles through a carbon matrix without damaging them. “Fuel pellet cross-section”, Idaho National Laboratory, CC BY 2.0. The pressed sphere remains a resin-bound green body. In the representative process, a lathe first brings the green sphere to its dimensional specification. Carbonisation at about 800 to 900°C in inert gas then turns the resin into carbon, followed by heat treatment at roughly 1,900 to 1,950°C under vacuum to remove residual gases and impurities. Public sources confirm these operations; X-energy’s final sequence and set-points stay proprietary. How the body is supported while its resin changes and how quickly it is heated can decide whether the green sphere comes through firing intact... Continue reading this post for free in the Substack app
|
Posts
0 Comments