3 ms·
Fusion is a materials science problem and materials informatics (machine learning methods for improving properties and / or devising new alloys) will help immen
by DrNuke 8y ago
Fusion is a materials science problem and materials informatics (machine learning methods for improving properties and / or devising new alloys) will help immensely once a new generation of digital-native scientists comes into the working force. Fusion might not become economically viable for a long time yet, though, so investments could very probably be moved elsewhere sooner than later. Healthcare for aging populations in the West and a new space race from China come quickly to my mind as powerful “distractions” from ITER part II, in the case part I ignition goes well by 2030. Smaller, more efficient reactors would not solve the energy problem even if they go successfully past the laboratory stage, in that being out of question imho.
- pfdietz 8y agoDistractions from ITER would be welcome. The power density of ITER is ~0.05 MW/m^3, some 400x worse than the power density of an LWR reactor vessel (and 2000x worse than the power density of the LWR core inside that reactor vessel). It's a project where research money goes to die, and new materials are not going to help that.
- DrNuke 8y agoPower density is irrelevant for fusion, though: you will never exhaust fusion “fuel” on planet Earth. It will never be an efficiency problem, like it is for fission aka uranium.
- pfdietz 8y agoNo, power density is very important, since it directly affects how expensive fusion reactors will be. The cost of a machine is a function of its complexity and size; fusion reactors will be both much larger and much more complex than fission reactors of the same power output.
- DrNuke 8y agoPossibly yes but, even the case, the US, the UK, Japan and France would switch to fusion immediately, if feasible, to get rid of fission and its waste. Renewables & nuclear are the perfect mix to contain greenhouse gases emissions, that’s why the debate is so hot.
- pfdietz 8y agoNo, nuclear and renewables do not mix well at all. Intermittent low cost sources of power shoot nuclear's baseload business case in the gut. The reactors cannot sell power at high enough price enough of the time to make sense.
- DrNuke 8y agoYou know, it is renewables to lose, apart of local places blessed by any sort of competitive advantage (water, silicon, etc.). However, I am not here for politics and you simply cannot beat nuclear by numbers, though.
- pfdietz 8y agoNuclear is being beaten by the numbers as we speak. That's why renewables are being installed worldwide and, with a few exceptions, reactors are not.
- neolefty 8y agoYes, fission creates waste, but fusion reactors as currently envisioned would also create radioactive waste — not the fusion products themselves — but materials of the containment vessels which have been bombarded with neutrons.
- foota 8y agoDoesn't it matter for construction costs though?
- DrNuke 8y agoFusion can’t be done now so construction costs are unknown at this stage. Graphene is the first of many wonders we can expect from materials science this century imho and it is extremely cheap.
- pfdietz 8y agoIn what fantasy scenario does a much larger, much more complex, made from much more advanced materials, system end up CHEAPER than the smaller, simpler, low tech one?
- DrNuke 8y agoFusion output would be unlimited (almost), you would then have solved any other problem apart of obsolescence of materials. Any other tech would be finite or more wasteful in comparison.
- pfdietz 8y agoI'm sorry, but that argument makes no sense whatsoever. "Finite"? What matters for decisions we make now is availability within relatively short term planning horizons. "Wasteful"? WTF? Something MORE expensive would be LESS wasteful?! What sort of doublethink is this?
- foota 8y agoI'm not a super believer in the possibility of fusion, but you can imagine that there might be benefits from doubling the size of a reactor in terms of decreased operational costs or even decreased construction costs (less need to duplicate infrastructure). I'm by no means an expert though.
- 8y ago