Historical research guide · 0965 · 2001
An Engineering Test Facility for Heavy Ion Fusion – Options and Scaling
Wayne R. Meier · D. A. Callahan-Miller · J. F. Latkowski · B. G. Logan · J. D. Lindl · Per F. Peterson
In one minute
Between an experiment that proves fusion works and a power station that sells electricity there is one more building. It is called an engineering test facility, and its job is to make every subsystem run together for the first time. Wayne Meier and colleagues at Lawrence Livermore, with Per Peterson at Berkeley, ask what that building looks like if the driver is a beam of heavy ions. Their list of what it has to demonstrate is a whole power plant in miniature: making the targets, injecting and tracking them, propagating and focusing the beams, getting gain and yield out of the target, letting the chamber absorb the shot and recover before the next one, removing the heat, recovering the tritium, and staying safe throughout. They then fold target physics, thick liquid wall chambers and a heavy ion driver into a single scaling study, so driver energy, target gain, yield per shot and repetition rate can be traded against one another as one system rather than optimised one at a time.
Connection
Chapter 12 keeps the ledger of fusion routes, and this is the step the heavy-ion route has to cross before a power plant exists: not another physics record, but the first facility where all of the engineering runs at once. Chapter 9 gets the beam and chamber physics — focusing an intense ion beam onto a millimetre target, and letting a thick curtain of liquid metal take the shot and reform in time for the next one.
From the source
The engineering test facility is the development step immediately preceding a demonstration power plant for inertial fusion energy — the building whose purpose is integration rather than any single new physics result.Abstract, first two sentences
What it must demonstrate, in one integrated facility, is the performance of every key subsystem required for fusion energy production: target production, injection and tracking, beam propagation and focusing, target gain and yield, chamber response and recovery between shots, heat removal, tritium recovery, and plant safety.Abstract, second sentence
The chamber carried into the study is the thick liquid wall: the surface facing the target is flowing liquid rather than structure, which is what lets the chamber take the shot, shield what is behind it and clear itself before the next one.Abstract, closing sentence
The method is integrated system scaling. Current understanding of the target physics and technology, of thick liquid wall chambers and of a heavy ion driver are combined so that scaling and operating scenarios can be investigated for the facility as a whole, instead of each subsystem being optimised on its own.Abstract, closing sentence
The budget the trade is made against, from Meier’s own earlier Livermore scoping study: a total cost goal of two billion dollars for the test facility against three billion for the demonstration plant, with the test facility characterised by yield per shot rather than by net power, and carrying none of the electric generating equipment a demonstration plant needs because it dumps its thermal power. A small-scale chamber test at about 30 megajoules of yield needs 0.8 to 2 megajoules of driver energy depending on the driver and the gain curve, and at that yield the allowable total driver cost is above a thousand dollars per joule, still above five hundred at 60 megajoules.Wayne Meier, Scoping the Parameter Space for Demo and ETF, UCRL-ID-132991, 19 January 1999: Introduction and the section ETF Results, with Figures 11 and 12
What to watch: for a heavy ion driver the binding trade is repetition rate against driver energy. In the same scoping study a 300 megawatt electrical demonstration peaks in allowable driver cost near 2 megajoules but needs 17.7 shots per second there; hold the rate to five per second and the driver grows to 3.4 megajoules while the allowable cost falls from about 730 to about 510 dollars per joule. Accelerator driver efficiencies of 15 to 30 per cent, against 5 to 10 per cent for a laser, are what buy the higher allowance in the first place. The measurement that settles the design is target gain at the energy a driver can actually be built to deliver.UCRL-ID-132991, sections Driver Efficiencies and Allowable HIB Driver Cost for Different Net Powers, Figures 8 and 9
The way in
https://doi.org/10.13182/fst01-a11963316How to cite it
Wayne R. Meier, D. A. Callahan-Miller, J. F. Latkowski, B. G. Logan, J. D. Lindl, Per F. Peterson (2001) An Engineering Test Facility for Heavy Ion Fusion – Options and Scaling doi:10.13182/fst01-a11963316