Uncle Sam to blow millions on getting fusion power finally working – with the help of AI The US Department of Energy has set aside $21m (£16m) to fund the use of machine learning to advance fusion power. “These awards will enable fusion researchers to take advantage of recent rapid advances in artificial intelligence and machine learning,” said Chris Fall, Director of DOE’s Office of Science. “AI and ML will help …
I reckon that we would have abandoned nuclear fusion research decades ago, if it wasn't for the giant fusion ball in the sky that keeps telling everyone that this technology does work. All science has to do is miniaturise it a little bit. Apparently miniaturizing what the sun does is not as simple as shrinking transistors.
I take it that was meant sarcastically. The problem is in using the Sun's energy in a way that is continuous and efficient, which is currently impossible. Requiring an acre of solar panels to power one house (but only during non-overcast daytime) is not a practical solution to our energy needs.
You couldn’t be further from reality there, I’ve just has 26x 550 watt panels installed on a modest size four bed bungalow and a self designed battery storage system for 23k of off grid storage, even with todays crappy weather in the mid UK I’ve generated 48Kwh and charged my batteries full for the evening run and charged my EV. Acres is a bit of an overstatement.
"Acres" was obviously hyperbole, but is seems that you have about 52 sq meters of panel. OK on a property that has that much roof area in a favourable orientation, but at least 50% of people in the UK will not have access to nearly that amount of sunlit area. Tell me that you could happily disconnect from the grid permanently and I'll take notice. Your 48kW would equate to 4 hours of sunlight with the panels you have - which may be the *average* you get per day over a year. But during Winter you can get weeks of very low output, and your 23kWh of storage just won't last.
That's before we get to the expense - you may be "saving the planet" but you are certainly not saving money. The panels will have cost at least £13000, which for me equates to 10 years of electricity bills - and that's not counting the batteries, electronics and installation costs. What's its predicted annual maintainance cost, and what's the expected lifetime before major replacement becomes necessary?
Again, you’re way off the mark, you’re firstly presuming panel size which is incorrect, probably based on a misconception from panels of years gone by, secondly your costs is over by £5000 and that included my batteries, sure it doesn’t suit everyone’s needs but mine it does, my energy costs per month were in the region of £180 and now I’m down to less than £45, Panels dont need cleaning or maintaining, the lovely UK weather does that for us, every part has 25 years warranty with guaranteed 98% output of the pv in 20 years. My lithium phosphate batteries are good for 8000 cycles before they drop to 80% effiecncy with a DOD of 85% so that’s around 9 years with my current use cycle before they start dropping capacity. All my heavy power use is during the day and my battery storage will happily provide power from sundown to sun up with 60% left in the morning. Yep I did save some cash by designing and installing my ac coupled unit and batteries myself, but that’s the bonus of prior research
" ... your costs is over by £5000 and that included my batteries, ... "
I estimated £13000, which means that you paid £8000 for 13kW of solar panels and 23kWh of storage? That is a huge amount cheaper than a quick online search indicated - My £13000 estimate came by reducing the cheapest online price of 500W roof solar panels I found by 25%. I am now amazed at the price you paid for 23kWh of Lithium Phosphate batteries. On Ebay and Amazon they seem to sell for about £300 per kWh, which would make the battery cost alone almost £7000. Seems that everyone I know with a PV system has paid many times more than they need have done.
> I reckon that we would have abandoned nuclear fusion research decades ago, if it wasn't for the giant fusion ball in the sky that keeps telling everyone that this technology does work. All science has to do is miniaturise it a little bit. Apparently miniaturizing what the sun does is not as simple as shrinking transistors.
I worked at JET for 5 years. Over that time, some of the physicists killed off my hopes for Fusion power, when they told me that the power density (both gravimetric and volumetric) of the Sun, is roughly equivalent to a healthy compost heap. At the Sun's core it's about 300W/m3, and a similar number per tonne mass. The sun is just so mind-bogglingly enormous that we get so much energy radiated from its surface.
That's why fusion power on earth needs to involve orders of magnitude higher temperatures and pressures than exist at the core of the sun.
Really, I think the earth would have been much better off if we'd spent all the money we threw at fusion, on improved Fission designs instead. Like molten-salt Thorium reactors which eat the nuclear waste produced by other reactors.
"Let me guess - with these advances, experts predict we'll have fusion power for the masses in 20 years?"
Thirty years is the "traditional" value. It has held good for all of my life, which started a long time ago.
"Nuclear fusion energy is 30 years away...and always will be."
I was a fusion physicist 30 years ago. My son hopes to study physics at Cambridge and did his summer essay on fusion. The conclusion; 30 years away. But in 30 years a lot has changed. The International community plans to build the DEMO reactor in 2050 (after ITER) and the plan is based on solid physics and engineering. However there is a small chance that a private company can do a SpaceX and get there first using novel technology such as high temperature superconducting magnets. For that reason I would bet a pint on <=30.
From the article: fusion seems to be one of those technologies that is always a decade away.
Snarkily I might suggest that when you pay for perpetual RESEARCH, you get what you pay for.
But I suspect it will be more like "privatized outer space" in that, once you encourage private industry to do its OWN research [instead of having DoE and/or DoD own it all at one level or another] then you'll suddenly have some "for profit" reactors being built, and then the floodgates will open.
But if you pay for RESEARCH, you may only get what you pay for.
Q: how come ONLY tokomak is being used at the moment? I understand there are a couple of other very interesting designs, some with magnetic confinement, that aren't making headlines...
A: Maybe ONLY tokomak is getting funded for RESEARCH ???
But we have a lot of cheap oil right now so the urgency and financial factors aren't quite "there". When tech finally makes it possible to produce CHEAP fusion power, expect wacky environmentalists and other protest groups to show up and do things to frustrate its development...
This post has been deleted by its author
Saying a fusion reactor would be safe ... doesn't make it safe.
See https://thebulletin.org/2017/04/fusion-reactors-not-what-theyre-cracked-up-to-be/
The saving grace for fusion reactors is that they are unlikely to ever exist, happily obviating all safety concerns.
IMO, the main reason governments fund fusion research (with paltry budgets) is to keep their nuclear genii busy and satisfied at home. This to forestall them being lured to go work in Pakistan or Russia on their nuclear projects, ones that are actually practical and nothing to do with cutting carbon.
A secondary reason is to keep punters and the Readers Digest set lulled into a sense of security about the future. Like my uncle who smoked all his life, saying "By the time I get cancer, they'll have found a cure for it" .
That article is one hell of a wet blanket on my fusion enthusiasm. If what he says is true, we're better off with thorium reactors.
The article doesn's say "safe". It says "safer than nuclear [fission] plants". That's a simplification, of course, but it's a very reasonable assertion. A fusion plant cannot melt down, and the link you posted does not dispute that. It is not "safe" in an absolute sense, which is also true of pretty much every large-scale power generation method.
> The article doesn's say "safe". It says "safer than nuclear [fission] plants".
It's not difficult to be safer than a nuclear tea kettle. Water plus radioactives are a spectacularly bad conbination and Alvin Weinberg had exactly the right idea in not only keeping them well separated, but ensuring that substitutes were also chemically/environmentally safe if not inert
Liquid sodium coolant - burns furiously on exposure to air. Like that's never going to happen, right (monju)
Molten lead coolant - doesn't burn, but the oxides floating around eat people's brains
Gas coolants - leak
water - dissolves everything in sight
Fusion might be viable in 50 years time. MSRs were proven viable 50 years AGO and they don't need heavy containment buildings that take decades to construct
I didn't forget about it exploding in water, I'm just referring to the accidents which HAVE happened.
"Honestly guys, it won't catch fire this time. Promise"
It took 15+ years to clean up the non-radioactive mess from the secondary cooling loop leak and sodium fire at Monju. It wasn't just the fire and 20+ tons of sodium in the basement, it was that the fire was so hot it destroyed the structural integrity of the building by melting most of the steel girders holding things together
Just because something is a "perfect coolant", doesn't make it a "practical coolant", any more than a solid gold frying pan is the best way to achieve the perfect fried egg, despite its wonderful thermal conductivity
More to the point the failing of the molten metal systems is that they _still_ rely on fuel rods and the intrinsic disadvantages of that method such as neutron poisoning/inability to deep load follow.
Alvin Weinberg knew a thing or two about building nuclear reactors when he built the MSRE experiment at Oak Ridge - after all he's the guy who came up with the ORIGINAL DESIGN used on the Nautilus that had been scaled up to obscene Rube-Goldberg sizes for civil power and speciically came up with a better, safer method because of his concerns at the engineering stresses being imposed on steam vessels and the fact that enriched uranium fuelling _RELIES_ on the weaonsmaking proces (plus is 89% wasteful before the reactor even starts, let along the 98% waste of input fuel on the output side - his better mousetrap method reduces output waste by 99%)
His reason for the Nautilus design: It wasn't the best design. It was what could be done with what was available, was entirely self-contained, had an ocean available for cooling and used boilers/turbines so that it could be safely operated by navy guys who understood boilers & turbines. It was not intended for large scale power generation and he was deeply disturbed when industry scaled it up.
Here's a hint on using water in a nuclear reactor: The engineering stresses increase with the cube of power generation capacity. A 10,000hp (8MW) steam boiler as used on a railway locomotive could (and occasionally _did_) level half a city block when it blew and modern 800MWe nuclear boilers are closer to 3200MWt - that's why the containment buildings are so big/solid/expensive and why steam boilers were dumped across industry in favour of internal combustion engines or electric motors wherever possible.
Getting rid of water from the nuclear loop improves safety by several orders of magnitude because you're reducing the steam generator part to a non-nuclear section which can run at 800C instead of 450C (which means you can superheat your steam and run efficient turbines, but also only generate just enough steam as needed, with just enough water as needed and any steam explosion is just a steam explosion, as seen at any convetnional power station). Water is the common factor in just about every civil nuclear reactor accident and incident (the stuff is corrosive as all hell at high temperature and pressure - Diablo Canyon was a very near miss - then there's things like the Nuclear Ouchi incident - handling processes not needed for a MSR, therefore not able to be abused by inexperienced workers not following the manuals)
Getting rid of enriched uranium adds a couple of safety extra factors because reactor fuel/used fuel is incredibly difficult to weaponise (russians and americans both tried making U233 bombs and failed miserably)
Getting rid of water as dangerous means it makes zero sense to introduce a substitute which is dangerous when exposed to air or water or which may leak and shows a reckless mindset on the part of advocates that justifies antinuclear campainers protests. That's the point of Alvin Weinberg's design - it's walkaway safe.
of course thumbs downing MSRs and screaming about how they _might_ be weaponisable seems to be the order of the day because it looks like China's going to be the country cmmercialising them - never mind that existing civil nuclear technology is a PRODUCT of the nuclear weapons industry and by definition if you can build a conventional civil nuclear reactor, you already have the materials onhand to easily build several dozen nuclear weapons should you choose to do so....
This post has been deleted by its author
I've not looked in huge detail at the F1 analysis, but I think it's based on qualifying 'raw' speed, the main comparison being against teammates. I have my doubts about the accuracy and usefulness of the outcome.
Not all teammates are equal, but even if this is taken into account there will still be large areas of bias. Not all teams treat their drivers equally, not all cars are produced or specified equally (in the same team), some teams technical development direction will be specifically to suit their number 1 driver, and some drivers (those experienced and talented enough) will sacrifice their qualifying setup for better race pace. The list goes on I'm sure. It's as much a talking point generator than a meaningful analysis.
And even after all that, raw qualifying speed is only one part of being the best 'complete package' F1 driver. I've watched it for a long time, and I would put Hamilton right at the top of this list. It's a privilege to witness the levels of performance he has reached.
I have my doubts about the accuracy and usefulness of the outcome.
I don't: it's not very accurate nor useful, but maybe it's entertaining. There are simply too many variables, especially unknowns, to make this kind of approach any more useful than any other. Indeed, as you say, comparing against teammates is almost bound to lead to overfitting.
Senna was considered by many to be the best driver of his generation but I think that's about as much consensus as there's ever likely to be.
I have a brilliant idea. How about discovering an already existing fusion reactor that's a safe distance away, then place photovoltaic cells facing it? If we could get enough such cells converting that fusion energy into electricity, we'd be able to generate a useful amount of power.
"How about discovering an already existing fusion reactor that's a safe distance away, then place photovoltaic cells facing it?"
1: Extensive hydrofluric acid pollution lakes in China which has been threatening agriculture and potable water for tens of millions of people for the last 15 years. No, it's not "someone else's problem"
2: Those panels are a substantial E-waste problem when they reach end-of-life - which substantial volumes are now doing - again, this is not "Someone Else's Problem", as they're difficult to recycle
3: With the best will in the world and despite the factor that "renewables can slightly outproduce carbon-sourced electrical generation" - there's a VERY LARGE GAP between total electrical demand if carbon is totally eliminated and what renewables can produce (electricity only accounts for ~1/3 of carbon emissions. Eliiminating the other 2/3 requires an increase of around 6-8 fold in electrical production - and then there are the developing/less developed countries playing catchup)
Because in most cases using PV cells to provide the energy you require is completely impractical for 2 reasons. 1) the amount of energy produced per unit area is far too small for most high-energy requirements and (2) they do not work well on cloudy days and not at all after sunset.
Staying vaguely on-subject with trouncing future green energy dreams... is anyone else mildly perturbed by the knowledge that 30% of the UK's 'renewable' energy mix comes from burning "low carbon" wood-chip? Sounds very much like a cop-out to me. But don't worry, I'm sure someone in HMG has worked through all the life-cycle CO2 emissions (amongst other stuff) and figured out it all makes sense comparatively.
> is anyone else mildly perturbed by the knowledge that 30% of the UK's 'renewable' energy mix comes from burning "low carbon" wood-chip?
You'll be even more perturbed when you learn that wood chip is "old growth" North American forests being clearfelled tens of square miles at a time, chipped and shipped across the Atlantic to feed the burners
The word you're looking for is "Greenwash"
Thats right, because they are clearing forests just for woodchip, ignoring the far more profitable timber market. They was me thinking it was just the waste.
You have a valid point about the shipping, just like how Almond Smilk is no better than dairy when you take into account the water, shipping and packaging.
There's an argument for burning wood from sustainable sources: it can now be done efficiently and reasonably cleanly and you're essentially recycling the CO2 temporarily stored in trees. But this comes with a lot of caveats as properly sustainable forests, as opposed to grow fast but fuck the ecosystem pine and eucalyptus, tend to grow slowly thus limiting their potential.
Wood-chip powered generators are in fact a very good solution IF the wood chips are sourced from sustainable farming (i.e. trees are replaced as fast as they are used). In effect it is a way of converting the Sun's energy to electrical energy without having the problems of direct conversion methods, because trees act not only as a means to convert the Sun's energy to chemical energy, but also as an energy store so that we can use it on overcast days and at night. Net CO2 generation is zero (the trees absorb the exact same amount of CO2 as the generating station produces). The big downside is that sustainable forestry need huge areas of land, which are not usually found in the same place as the electricity is needed, meaning that there is enegy wasted in transporting the wood from the forests to the generating station. Probably a better solution would be to build a factory that converts wood to a more energy-dense chemical fuel (e.g. artificial oil), then ship that fuel instead of the wood.
"Trees are replaced as fast as they are used" is an interesting challenge in time travel.
In effect you have to plant your forest and let it grow to maturity (capturing the carbon) before you burn it and return the carbon to the atmosphere.
Each power station would need its own dedicated new timber source planted long before the power station entered production.
Burning ancient timber doesn't fit this. Even burning logging and wood processing waste doesn't fit this; it just returns carbon to the atmosphere faster than decomposition.
I think there are some interesting stats re. biomass as a fuel source – whether it's wood for burning or plant oil – for anything like our current use of fossil fuels we'd basically be planting nothing else!
But I do like the currente wheeze of turning oil into fertiliser for maize that gets turned into petrol additive! This is a nice bit of arbitrage that is making some investors very rich, while driving up farm land prices beyond the the reach of many farmers.
The best intentions and all that!
Some fairly recent stuff on this topic on YouTube about ITER.
https://www.youtube.com/watch?v=JCpWPJrH7TA
https://www.youtube.com/watch?v=bAw-rdl9YZg
And lets not forget we learned to fly because we never stopped trying.
About Tokomak:
"A tokamak (Russian: Токамáк) is a device which uses a powerful magnetic field to confine a hot plasma in the shape of a torus. The tokamak is one of several types of magnetic confinement devices being developed to produce controlled thermonuclear fusion power. As of 2016, it is the leading candidate for a practical fusion reactor.[1]
Tokamaks were initially conceptualized in the 1950s by Soviet physicists Igor Tamm and Andrei Sakharov, inspired by a letter by Oleg Lavrentiev. Meanwhile, the first working tokamak was attributed to the work of Natan Yavlinsky on the T-1".
https://en.wikipedia.org/wiki/Tokamak
A fusion reactor that supplies electricity to the power grid seems only slightly more likely than a genuine quantum computer, in other words not at all likely.
So far, from a massive energy input, the main output is a huge neutron flux - no use to anyone.
Considering the amount of money that has been (and is being) spent, e.g. on ITER and similar projects, space-based solar power could be a better bet. Though still years away, the problems involved are at least tractable. The same can't be said for fusion power, sadly.
The best solution, anathema to the world economic system, is simply make less stuff, use less energy. Then, maybe, renewables could supply our needs.
We also need to make less people. The planet has bee suffering ever since we started agriculture. It's been guessed that 10 million hunter gatherers existed in (stressful) harmony with the rest of the ecosystem. Using nuclear fusion suddenly seems like a tractable problem compared with my proposal.
The 'too many people' argument. I agree the trouble might have started with agriculture, by which accumulation of surplus = accumulation of wealth/property etc. = roots of capitalism. It's too simplistic to say there are too many people. More like there is too much wealth in the hands of a tiny minority of people, an 'elite'. Without wishing to sound 'political', a system that has the likes of Bezos, Gates & co worth more than the GDP of very many countries must be a tad suspect.
(Btw, what is your proposal?)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
