Hacker News new | comments | show | ask | jobs | submit login
Lawrence Livermore NIF laser fusion experiment delivers record energy output (powersystemsdesign.com)
35 points by UnFleshedOne 1031 days ago | hide | past | web | 39 comments | favorite

I spent a bit of time over the last 3 months typing up & annotating an obscure but excellent SF/literary novel, _Radiance_ by Carter Scholz ( http://www.gwern.net/docs/2002-radiance ). It's all about the Lawrence Livermore SDI scandals and then the process of them getting NIF, so I read up on the historical background.

I was a little flabbergasted at the overall progress of the laser fusion research program: apparently back in the early '70s, the first LANL people predicted that you could achieve laser fusion with just 1kj of input. Millions were spent, ignition was nowhere near, so they got funding for another, bigger laser, which failed too, so they got funding for more lasers, and those failed too, so they got funding for an even bigger multi-billion dollar laser called NIF, and not only has that failed to reach ignition once as promised - despite going up to 1.7mj (quite a long way from 1kj) - LANL managed to run over budget by at least 400% and took something like a decade longer to finish NIF than promised!

In the top paragraph you mention Lawrence Livermore and in the bottom one you use the acronym LANL, which is Los Alamos. Did you mean LLNL?

Yes, my bad. (They're so damn similar acronyms >.<)

NIF has been around so long that the digitizer scene in the original TRON was filmed there. Its purpose is to explore the physics of exploding nuclear weapons; it is an arm of the lab which designs US nukes.

Any similarity between NIF and a practical Fusion Reactor is purely accidental.

That's an oversimplification. Throughout the history of laser inertial fusion in the U.S., most of the scientists involved have been keenly interested in fusion energy, and advocated a program explicitly aimed at energy production.

Congress opted instead to fund it as a weapons-related program, but that doesn't mean that's all they're doing, and for a while they were explicitly spending the majority of their efforts on achieving energy gain. Recently they missed a deadline set by Congress which redirected the program to be only 20% for energy.

A really great history of the U.S. fusion program, from the 1950s to about 2012, is Search for the Ultimate Energy Source by Stephen O. Dean, who was deeply involved with much of it. http://www.amazon.com/Search-Ultimate-Energy-Source-ebook/dp...

>most of the scientists involved have been keenly interested in fusion energy

Can you back this statement up?

My source is the book I linked. Not everything's online.

But I should probably say "many" rather than "most," since he didn't give actual numbers.

So that's a "no".

You challenged the idea that a DoD project is indeed defense-related by citing (but not really) an obscure book that maybe says some scientists involved with NIF really believe it's for energy production. Forgive me if I'm not convinced.

So, first, TRON was shot in the Shiva Laser facility, a precursor to NIF that is also situated at LLNL.

Second, if anyone would like to see an inspirational speech given by NIF Director Ed Moses on how you move from theory to a a working reactor based on this technology, he spoke at Long Now in SF a few years ago:


Their LIFE design for a practical, economical power plant is also worth looking at. For people who don't care about energy they've sure put a lot of work into it, in collaboration with over thirty industry vendors. https://life.llnl.gov/


TL;DR - More money -> likely fewer results.

Large organizations tend toward waste, delay and mediocrity. Without a profit motive and a mission to deliver a specific solution, the modus operandi is to grow as big of a budget pyramid as possible while providing just enough press releases to keep the DOE gravy-train rolling.

This is why high beta and other approaches from smaller groups may likely crack the problem, despite not having all of the toys (and distractions).

What a ridiculous comment. Large budget in comparison to what? So the scientists who have dedicated their academic lives to research are just in it for the money? Does NASA waste the $16.6 Billion they receive? The Saturn 5 was developed for hundreds of billions. There is absolutely nothing that would substantiate your "gravy train" buzz phrase.

Duh. NASA wastes billions for the value delivered, and it's more/less an Amtrak. SpaceX, Virgin Galatic and Stratolaunch deliver much more value because they have to; hence vendor partnerships and [0]. NASA, along with the rest of USG, is also trying its best to manage a multitude of accelerator/incubator programs to promote private, commercial research.

Saturn V might have worked, but justifying costs of one program with historic waste of another is still an invalid argument in the context of cost of success.

[0] http://www.nasa.gov/pdf/169943main_NASA_Virgin_Galactic_MOU....

"Saturn V might have worked, but justifying costs of one program with historic waste of another is still an invalid argument in the context of cost of success." Did you read your comment? You can't justify your argument by bringing up "Amtrak" (not sure what an Amtrak is) a supposed failing enterprise, as much as I can bring up the Saturn V and use it to support my point. Again there is absolutely nothing to back up what you said. The issue of wasted costs comes down to management and little else. Small business and large businesses both can waste money or apply capital in smart ways. You also didn't specify what a large amount of money is, is Apple more wasteful than Boeing because of the disparity of income? The amount that NASA gets in funding is quite small and they do a lot with it. SpaceX or Virgin Galatic aren't actively commanding multiple missions within and outside of our solar system. Your argument is so general and naive. Have you ever worked in industry?

Do you have any evidence to back this up? Off the top of my head I can think of several dozen large successful projects, everything from the Mars Rover to CERN's LHC to the great wall of China and the Pyramids. These were all enormous projects requiring thousands of workers.

For anyone that has worked for a large organization, for- or non-profit, for any amount of time with a budget then it's easier to understand how things actually work.

I have seen 9 figure / 7 year internal projects silently scuttled as to not create a "failure", despite delivering nothing. I have seen 6 figure projects surpass 8 figure projects through a combination of talent and leadership. I've seen duplication of systems merely because two depts can't find a way to work together, effectively doubling costs.

Depends on the meaning of "success." "Success" of a political project is rarely based on results; all projects are "successful," because the unsuccessful ones disappear. Real success, as in achieving an novel aim, is another matter.

Success is great, but at too high of a cost, it may make the results bitter (bodies in the Great Wall, pushing $10 billion total spent on NIF without results). This is where private R&D excels: there's a defined goal, time/cost pressures and usually fewer political constraints. This is why high beta has a higher probability of success AND a lower cost of success (better value/$).

There have been hundreds (maybe thousands?) of independently operating research groups at Universities and government labs, working since the 1950's all of them studying fusion. None of them has come as close to attaining controlled self-sustained fusion as the NIF. That is the entire point of large grand projects, they do what small groups aren't capable of. Simply because the NIF hasn't yet attained self-sustained controlled fusion doesn't mean its a failure.

What do you think is reasonable price tag for the development of an entirely new form of clean energy?

I don't know why they're even pretending anymore. Everyone knows NIF is a weapons research facility and always has been. They've milked all the PR they're going to get out of this.

Did this get a positive energy return? If I've understood past ignition attempts correctly, they too delivered tremendous energy output... but required even more tremendous energy input.

No, it's extremely negative.

"In the nanoseconds that followed, the capsule imploded and released a neutron yield of nearly 3x10^15, or approximately 8,000 joules of neutron energy..."


"These promising returns were the result of a laser experiment that delivered 1.7 megajoules (MJ or million joules) of ultraviolet light..."

More numbers: https://en.wikipedia.org/wiki/National_Ignition_Facility#NIF...

To put these numbers in context, 8000 joules is about 2 kilocalories (or about half an M&M candy). It's about $0.0003 of wall electricity. It's about the energy in an AA battery. It's the energy from exploding 2 grams of TNT.

1.7 megajoules is about 400 kilocalories, or a king size Kit Kat bar. This is the energy of a car going 140 miles per hour. This is about 7 cents of AC power.

I'm getting the KE of a car to be roughly double your estimate:

    assuming ~4000lb[1] car, 140mph ~= 62.6m/s,

    0.5*m*v*v = 0.5 * (4000*0.454) * 62.6^2 = 3.5E6Joules[2]

    ~= 830kcal
Did you miss the 0.5, or use a different Spherical Co^Wstandard automobile mass?

Which is still only 2 kitkats :)

The other potentially confusing thing (judging by the sibling comments here) is that this isn't the energy required to accelerate an actual car from 0 to 140mph. It's the kinetic energy embodied in a car traveling at that speed. Wind & tyre resistance losses, engine/transmission, etc, etc. It's really the (upper bound of the) amount of energy you could expect it to impart if it hit something, or the amount of heat the brakes would need to dissipate to stop it.

Finally, for all those commenting on how much energy there appears to be in a kitkat, consider the famous E=MC^2, and just how staggeringly huge that C^2 term really is (9E16J/kg, if you're wondering) Getting all that out in a usable form, however, is left as an exercise for the reader.

[1] http://www.nytimes.com/2004/05/05/business/05weight.html


[2] http://www.wolframalpha.com/input/?i=0.5+*+%284000+*+0.454%2...

I didn't expect so much controversy here, I just wanted to get a feel for the order of magnitude. I went with 2000 pounds for the car weight, which is the weight of a 1976 Honda Accord - 1976 because of a search error on my part. 4000 pounds is the average new car or light truck, so that number is a bit high because of the trucks. In any case, the energy dumped into this experiment is like a big car crash concentrated into a tiny capsule.

Anyone who has used a treadmill at the gym will recognize that it's a lot of work to burn off 400 calories. Not enough to get a car up to 140 MPH because the body isn't 100% efficient.

Of course, you're talking about the chemical energy in a Kit Kat bar. Were it completely converted to energy, it would produce around 5.8 Terra-watt years of power assuming a lossless conversion.

"...about 400 kilocalories, or a king size Kit Kat bar. This is the energy of a car going 140 miles per hour."

Wow. How can those two things be equivalent? I know there's a lot I don't know about physics...

1) Realize that you are a highly articulated animatronic robot, built with inefficient meat actuators;

2) think about all the things you could do in the hour or so that a king sized Kit Kat bar could keep you energized;

3) visualize the trillions of cell-sized actuations required to perform each step of each task previously listed, and the neural computation required to organize these actions;

4) compare that with the relative simplicity of turning wheel axels to propel a car down the road;

5) marvel at the power of a Kit Kat bar powering the magnificent machine that is you.

Think about it, do you think you could move a car to 140 mph by burning a kit kat bar (in space, with no friction)? Consider ion propulsion drives, which use a teeny tiny bit of force (less than the weight of paper) but are highly efficient. What percent of a kit kat bar do you think it takes to get an astronaut on the ISS (say, 1/30 the mass of a car) to bound across the ISS at 10 mph?

But yes, the comparison is supposed to be to put in natural contexts, and so this is flawed example since we don't really live in space.

Kinetic energy is the speed, squared, times the mass, halved.

140 mph is not a small number; to the power of two it's rather large. Now multiply the result by the mass of a car, which is still pretty huge even halved.

A racing car barreling down the speedway packs a hell of a punch.

I think the GP's question was more like "how the hell can a single Kit Kat bar get a whole car to 140 mph?".

I see.

Well, a kitkat bar actually has a decent amount of energy in it. Very roughly, the same energy as the same mass of wood, or gasoline. Chemical fuels don't vary too much in terms of energy density. They stay within the same order of magnitude, at most.

And yes, food is chemical fuel.

No, a Kit Kat bar has way more energy per gram than wood, or even TNT. A king size Kit Kat bar packs basically the same energy as 0.5kg of TNT. That's a lot of power.

Probably only a straight-up chemical fuel like gasoline or methane packs more punch per gram.

Strange. I did some research and found some numbers.

The energy density of Kit Kat is 22 MJ/Kg (5.2 Kcal/g) http://www.wolframalpha.com/input/?i=%28Kit+Kat+calories+to+...

TNT has 2.8 MJ/Kg in an explosion and 4.2 MJ/Kg in a combustion. It has too much Carbons. In an explosion, it doesn’t react with the oxygen in the air, because it’s too fast. There is an interesting discussion in http://en.wikipedia.org/wiki/Trinitrotoluene#Energy_content

In general, the explosives have to explode without using the Oxygen in the air, so they have a lot of Oxygen and Nitrogen atoms inside. The fuels for slow combustions can have almost only Carbon and Hydrogen atoms to store the energy more efficiently and get the Oxygen from the air.

Nitroglycerin has less Carbons, so has more energy density. Mixed in dynamite has 7.5 MJ/Kg and alone has 6.4 MJ/Kg (Is this correct? I expected a bigger value).

I found a list of these values for a wide range of materials. In particular it says that green wood has 10 MJ/Kg, air dry wood has 15 MJ/Kg and oven dry wood has 20 MJ/Kg. http://physics.info/energy-chemical/

Other values: Gasoline ~45 MJ/Kg, Methane 55.5 MJ/Kg, Hydrogen 142 MJ/Kg

with extremely low efficiency of lasers, my bets is on Sandia Z machine.

Not that these guys are in any rush to produce practical reactor as they are also mostly nuke weapons research, yet as a side effect they have been making very interesting and meaningful progress during last 15 years. The machine itself isn't for fusion research, it is for X-ray generation, and it just happens that it generates so much and so efficiently that it seems they couldn't resist and finally started trying mini Teller-Ulam :)

Link to the original press release:


Cultural observation: some serious govspeak interleaved with respectful capitalisation of 'Mother Nature'.

(At least if the world disappears due to a crazy physics experiment one day, metaphysical archaeologist sensors in some parallel dimension who became aware of the event might see that we weren't entirely disrespectful!)

WTF are you talking about?

TL;DR - 3 orders of magnitude from economic viability.

in 1.7 MJ

out 8 kJ

Q = 0.00444_

Compare that to 5 MJ output from JET, at an efficiency of 65%.

Yup. JET is delivering much than NIF, in terms of reaching the goal of maximizing Q.

NIF has a program cost (1997-2013) on the order of $8B USD to reach this unimpressive number. (Based on $450m/yr not including fixed/one-time costs.)

JET is in for about $20B USD ($15B €). [0]

JET might be around thrice as costly, but the results speak for themselves.

[0] http://www.efda.org/fusion/why-should-society-invest-in-fusi...

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact