Hacker News new | past | comments | ask | show | jobs | submit login
How the Curiosity Rover's Nuclear Battery Works (about-robots.com)
174 points by peterwwillis on Aug 6, 2012 | hide | past | web | favorite | 96 comments



Many people here in the far north have extensive solar setups and live completely off-grid. Unfortunately in winter, we only get ~4 hours of sunlight, and it's a pain to sweep snow off the panels every day at ~ -30C to -40C

We don't get much wind, and rivers are only liquid for ~4-5 months, so wind and hydro are not popular.

I've been investigating commercial thermoelectric couplings as a source in the winter. Everyone has a BIG wood stove burning 24x7 for ~6 months. My best research shows it shouldn't be hard to see +400C on the top surface of the stove. My plan is to have a radiator outside, run anti-freeze in the system and get an approx 400 deg C temperature drop.

Now, my interest is peaked in this approach from NASA. I wonder how long it will be until I can buy or build such a thing?


I absolutely wouldn't recommend it.

Thermoelectric generators are far far less efficient than simple generators (say a Stirling engine). The benefit to them is that a) they are very compact and b) they need no mechanical parts.

It makes sense to put a thermoelectric generator in a satellite or rover because the mechanical parts found in a traditional engine are typically very heavy and more likely to fail.

So if you aren't concerned with a couple dozen pounds and the occasionally maintenance call, just use a wood-burning generator. You'll save about an order of magnitude on costs.

edit: what you're looking for is micro-cogeneration (http://en.wikipedia.org/wiki/MicroCHP)


Piqued your interest, just as an aside



Thanks.

I looked into that a while ago (and a similar one), and it looks like any power produced from the TEC is considered a bonus, and is in the range of 1-5W, probably only enough to charge a phone, etc.

I'm looking to produce a sustained 150W + to charge deep cycle batteries.


Tech specs show the biolite to have about a 2.5in radius and (approx) 6.5in high fuel compartment, which is about 2.1L (http://www.wolframalpha.com/input/?i=volume+of+a+vertical+cy...). Using its design you can get about 2W continuous power at 5V.

Scaling this up, roughly, you would need about 157.5 liters of fuel space, though probably much less if your fuel was more compact (obv you don't fill up exactly 2L of space in the canister while burning twigs). The realistic output of heat/power based on space is probably much more efficient than something this large, considering they say 46 grams of fuel can boil 1 liter of water, and 46 grams is approx 0.046 liters (based on water density).

Somebody please correct my horrible assumptions, but basically, fill an oil drum up most of the way with wood and build a thermoelectric generator and you should be good to go.

(Also, if you're charging deep-cycle batteries, i'm assuming you're not going camping? Maybe solar would be simpler? http://www.mdpub.com/SolarPanel/index.html) (Edit again: I forgot your original post, no solar)

More edit: Check this page for an adaptable thermoelectric generator: http://www.tegpower.com/

And more edit: https://www.youtube.com/watch?v=bUH1HA3EnZE (i'll stop posting links now!)


Sorry, this is the one I meant to link to (probably the similar one you mention). Might be a better design to scale up. Don't know if the TEC is any more efficient though.

http://www.kickstarter.com/projects/tpod1/tpod1-thermoelectr...


Who knew steampunk inventions were environmentally friendly?


My understanding of the problem is that nuclear sources tend to be nasty stuff and not the sort of thing that one really wants lay people responsible for discarding.


I think the OP is looking at using their stove as the heat source, not a nuclear battery.

On that note, the idea you're talking about was actually considered for a little while for public use until the dangers were found to outweigh the benefits.


Right now I'm thinking about a stove, but this nuclear battery has me thinking.

I am wondering if it would be possible to buy or build such a nuclear battery.


Governments tend to frown on the personal acquisition of radioactive isotopes. Well, there's a tiny bit in your smoke detector, but if you managed to get the eleven pounds that are in the MSL's RTG you'd make the international news.

Pu-238 isn't naturally occurring, and nobody makes it anymore. The US bought the plutonium that went into the MSL from the Russians, who are starting to run out themselves.

The other problem with RTGs is they get less efficient as they get larger. The one in the MSL only produces 125 watts, which will decay to about 100 watts at the end of its 14 year lifetime. That's not a lot of power.


Certainly to build one the theory behind it wouldn't be tough, it'd operate on the same principles as the stove setup. To buy though? There's a story of a boy scout who did something similar as part of a personal experiment in the 90's. (Before getting raided out of fear by the FBI) Will have to find that story again.

In almost every country the materials required are tightly regulated and getting permission even in a lab environment is tough work. If you can work out the legalities and safety, I'd be interested too!


The book is The Radioactive Boyscout

http://www.amazon.com/The-Radioactive-Boy-Scout-Backyard/dp/...

I read the book. from what I remember he started with smoke detectors and worked his way up from there. Just a few letters to some supply houses saying he was a researcher or something. he effectively mail ordered his radioactive sources I think.


Looking through the Wikipedia page on RTGs, there is a noted risk classified by the US-DOD that Beta-M RTGs could be used as dirty bombs, which is about what I expected. =/ I wonder if this risk extends to other RTGs in the wrong hands? I imagine so, but I honestly don't know.


Basically a RTG converts nuclear decays into power by absorbing the nuclear decay products, generating heat. This is quite similar to the process of radiation poisoning, where chemical bonds are destroyed by the decay products. So for a rather handwaving estimate, we can assume that the processes are the same.

The lethal dose of radiation is about 3-5 Sv, that is 3 to 5 J/ kg absorbed radiation (times some factors, which describe how the radiation is absorbed and which type of radiation). Therefore assuming full body irradiation and gamma particles, the lethal dose for an adult would be received in a matter of minutes ( 60kg irradiated by 1 W would be 1 Sv/minute). Note however, that a more realistic scenario for a dirty bomb from a RTG involves ingestion of alpha emitting particles for a rather large number of bystanders. ( And therefore the effects vary a lot by the details of the bomb and the radio isotopes used. )

So already with the radioactive material from a rather small RTG you can build a quite potent dirty bomb, rather independent of the exact type of the RTG used.


The sensible way to get electricity out of burning wood would be a steam turbine or stirling engine-based micro CHP (https://en.wikipedia.org/wiki/Micro_combined_heat_and_power).


Toshiba makes a small nuclear reactor for your home: http://www.engadget.com/2007/12/19/toshibas-building-a-micro...


This concept was also used in russian lighthouses because it would be cold enough in siberia to create the delta difference needed. http://en.wikipedia.org/wiki/Radioisotope_thermoelectric_gen...


Used as a plot device in a movie called "How I ended last summer":

http://www.imdb.com/title/tt1588875/

I enjoyed it a lot - it's got that real bleak outlook that some of my favorite Russian cinema has, and some _spectacular_ cinematography.


Some interesting info about safety on this one: http://bellona.no/bellona.org/english_import_area/internatio... (from Wikipedia citation)


I'd disagree with the article and say the biggest threat of the RTG's is not being used by terrorists because a dirty bomb isn't all that effective with that amount of material, however, a bystander digging into one and having fun with the material could be a serious problem. Like the Goinia Incident [0].

Then again. It's kind of a funny thing to worry about with all the unexploded ordinance that's around most countries, even the U.S. I was just out sailing in the long island sound last week and my girlfriend thought it extremely odd how many unexploded bombs pockmark the navigation charts.

0: http://en.wikipedia.org/wiki/Goi%C3%A2nia_accident



Thanks. Edited


NASA has almost run out of U238: http://www.popsci.com/military-aviation-amp-space/article/20...

People promoting liquid thorium reactors have been pointing out that they could supply it: http://flibe-energy.com/products/


Plutonium 238 is NOT Uranium 238.

There's shitloads of U238 in the world.


Crap, that's what I meant. Thanks.


There's Pu238 in our nuclear fuel rods, we aren't allowed to reprocess spent fuel to extract it.


That's worthless because it's contaminated with other isotopes of plutonium. Pure Pu-238 is made by irradiating Np-237, which comes from spent fuel but is already separated in very large amounts in the US. I think the main setback is funding.


How much power does the device generate? How much will it generate in 14 years when it is "expended"?

Edit: Initially 2 kW of thermal power and 125W of electrical power, falling by only 20% after 14 years. https://en.wikipedia.org/wiki/MMRTG#Design_and_specification...


According to Wikipedia, 125 watts when it was launched, down to about 100 watts in 14 years.

The JPL guys claim that the RTG is really just a trickle charger for the batteries that actually handle the load -- which likely has transients that the thermocouple in the RTG couldn't handle.

I don't know the lifetime of the lithium-ion battery pack but I'm guessing it'll degrade way before the RTG power decreases below the point where it can effectively charge the battery.


There's a LiPo technology called "thin film lithium" which is claimed good for 10,000 cycles - assuming one charge/discharge cycle per day (and incorrectly using earth-days for simplicity), that'd give you ~27 years lifetime from the battery. (I wonder if this'll end up with the same "over engineered" reliability that let the Spirit rover last over 6 years when intended for a 90 day mission?)


The Opportunity rover is still up and running, which means it has exceeded its planned duration of activity more than thirty times over.


[In the voice of old "Gauntlet" game] "About-robots.com needs editing badly!"

> The first time I heard about the Curiosity Rover nuclear battery, I was thinking this must be top notch technology. What a surprised I had when I saw it has been used for more that 40 years already.

√Čvariste Galois came up with mathematics in the early 1800's which only in just the past several decades had widespread applications in cryptography. Carbon electric arcs were developed right at the start of the 1800's, and not only have we been finding new applications for them ever since, we'll probably keep doing so!

Also note that the mid-90's trappings of the digital office were running at Xerox PARC in the early 1970's.

As is often said: "The future is here, it just hasn't been evenly distributed yet." There is a lot of "top-notch technology" that the general public doesn't know about yet, and may not for a decade.


e-ink is another good example. The technology was originally developed in the 70's.


this article is wrong on many things.

one being that the rover can drive during night.

the small power output of the "nuclear battery" is not used to drive the rover. it used to recharge batteries during day and night so that the rover can operate on the batteries during the day.


That's a good point... while in theory is probably could drive at night, it probably isn't a good idea to have a rover on a different planet driving around in the dark.


Is radio communication with the rover even possible at night? We only have line of sight to mars during the martian daytime.


Our line of sight to Mars is independent from Mars seeing the sun.


Not really. If the rover is on the side of Mars facing away from the Sun, it's almost always facing away from the Earth, too, since the Earth is deeper inside the solar system.

Facing the Earth and facing the sun are highly correlated. Seeing one xor the other is possible, but unusual.


He's still right in a very technical, strict sense. I'm too lazy to look up numbers and do the math, but I expect there's a window near the Martian sunrise/sunset when the sun isn't visible over the horizon and we have line of sight.

I expect that window is small to the point where no one actually cares. It would be an interesting applied-math problem for a grade-schooler, though.


The geometry of viewing Earth in the Martian sky is the same as viewing Venus in our own Terran sky. The interior planet periodically approaches a certain maximum angular separation from the Sun. Just as we see Venus but not the Sun sometimes in the early morning or evening twilight, Martians (and rovers on Mars) could see Earth but not the Sun at corresponding intervals of Mars' rotation. It's a significant amount of time, up to several hours per day at maximum separation.

The angular separation between the Sun and the Earth as seen from a space probe is significant as far out as Cassini at Saturn. The probe can receive commands from Earth without the signal being overwhelmed by solar radiation, except for a few days each (Earth) year when Earth is too close to the Sun as seen from the spacecraft. (Earth doesn't literally go behind the sun often, thanks to inclination of the planetary orbits plus Cassini's own inclined orbit at Saturn.)

> Our line of sight to Mars is independent from Mars seeing the sun.

Our line of sight to Mars as an entire body is independent, but our line of sight to a particular point on Mars is indeed correlated with that point facing the Sun.


Do you happen to have illustrations and/or numbers? I was planning to actually look everything up and work it out when I got home from work, but it'd be nice to have confirmation of my results, too.


It's also not possible to directly-drive the Rover anyway, with latency varying between 4 and 20 minutes depending on Mars/Earth orbit.

Curiosity drives itself.


Not sure about this specific mission but I think we have had satellites around mars in the past that have helped relay messages when line of sight isn't available.


Yup. The 3 satellites will be used for primary communication along with an X-band transmitter on the rover itself to talk to earth.

Edit: Link (communications): http://en.wikipedia.org/wiki/Curiosity_rover#Specifications


But the rover can drive at night, right? It just doesn't for safety reasons.


Correct. If the writer had only watched the NASA broadcast he would have heard it right from the horse's mouth. Because of the low temps at night, the power requirements to heat up the rover and devices would not make it worthwhile. In addition, the total output of the generator and batteries on a particular day is not enough for it to run continuously for the entire day.

Another interesting point brought up was that the rover was tested to 3x life but not to failure. It's certainly possible for it to run for many years longer than the stated life.


They talked about this during a press conference today. Theoretically it could operate at night, but the power draw to heat the motors and joints would be excessive since it's so cold after sunset. So they don't and allow the MMRTG to charge the battery (and keep the heater for the electronics) while most of the machine is idle.


Relevant: INL helps power mission to Mars http://www.idahostatesman.com/2012/08/06/2219647/inl-helps-p...

"In simple terms, Curiosity runs, in part, on a $100 million nuclear battery developed at the lab, said Stephen Johnson, division director of Space Nuclear Systems and Technology.

The 2-foot-tall, 2-foot in diameter cylinder aboard Curiosity is packed with radioactive isotopes generating heat. That thermal energy is converted into the electricity fueling Curiosity's wheels, arms and other gadgets, as well as recharging its bank of lithium-ion batteries."

"Fuel cells in the energy source are 1 inch tall and 1-inch diameter cylinders. Each puts out a mind-boggling 9,000 to 10,000 degrees of heat, shift supervisor Dave Hendricks said."


Wow, that's super cool. I wonder if this technology is being used on any earth-based applications? If not, I'm assuming there are harmful side effects that don't necessarily affect the space missions. I'd like to learn more about the negatives of this technology basically.


If this is the same tech I remember, Soviet Russia used them a lot for far-flung lighthouses that couldn't be easily refueled. Many of them are still out there, or have been stolen, vandalized, stripped of metal by unwitting thieves, spreading hazard all around. One was even found underwater.

edit: links:

http://bellona.no/bellona.org/english_import_area/internatio...

http://englishrussia.com/2009/01/06/abandoned-russian-polar-...


There is a movie on Netflix streaming that has some small plot feature tied to one of these these nuclear batteries in a far North Russian weather station:

http://www.imdb.com/title/tt1588875/


For a short while they were used in implantable pacemakers as 'lifetime' power sources[1]

The main issue is that they're really rather inefficient. The efficiency of the thermocouple at converting thermal->electrical is only about 5-10%, and combined with the cost of shielding, expense of the radioisotope to begin with, and security/safety considerations, they're really only suitable for niche aerospace/defense applications.

Edit: I was curious if anyone had considered a stirling or other heat-engine driven by decay heat, and found https://en.wikipedia.org/wiki/Stirling_Radioisotope_Generato... which looks like it can hit 20+% efficiencies. The downside is that unlike thermo-electric/Seebeck effect converters, they have moving parts that could be a threat to reliability, which is the major issue when you're a planet away from the nearest repair tech.

[1] https://www.orau.org/ptp/collection/Miscellaneous/pacemaker....


The bigger issue using it in pacemakers is it's actually beneficial to have a battery that needs replacing every 10 years or so - pacemaker tech gets better so fast that a 30 year old device would be crap compared to the current generation.


How much better really, if the current one wasn't working well you would find out real quick.


More sophisticated programs (fewer false positives and false negatives), smaller devices, batteries that last longer between charges, etc.


There's probably a radiation issue that wouldn't make it suitable for close to human contact. I wonder how much energy the rover needs to move around since the gravity there is considerably weaker.


Rover mass is same on Earth or Mars. So acceleration costs the same. Low gravity also means less friction with the ground, making it somewhat harder to move on Mars.

Climbing hills would be easier on Mars.


Isn't there also an issue of impulse though? There must be an acceleration dip compared to earth since I'm basically pushing a mini cooper on earth but only a bicycle on mars. Note: I'm a scientist so I'm embarrassingly bad at physics.


It would be easier to lift the rover perhaps, but F=ma, so since the mass of the rover hasn't changed, neither has the acceleration you can develop for a given force applied.


Isn't the friction dependent upon the normal force which on Mars would be less. Therefore, shouldn't the force required to overcome that also be less?


Absolutely, if you were pushing a block along the ground. Fortunately modern "wheel" technology reduces the bearing friction so much that it is no longer a factor. The only friction left is the friction holding the wheel to the ground. This is reduced in reduced gravity, such that wheels are more likely to slip.


Wheels need friction to develop acceleration. So its a negative not a positive.


It's only a negative if the vehicle needs to accelerate quickly. So sufficient friction is sufficient, not a negative or a positive.

Given that the batteries store roughly 2.4 kW-h, it's probably safe to assume that the max drive power is less than 500 watts (2.4 kW-h / 10 hours -> 240 watts). Earth side electric vehicles commonly have drive power over 100 kilowatts.


Sometimes used in earth-based applications too. See http://en.wikipedia.org/wiki/Radioisotope_thermoelectric_gen....


Calling it a "battery" is disingenuous. An RTG doesn't fit the typical definition of a battery -- it doesn't store electrical energy. A better name would be "power supply". </pedant>


Well it does - it stores nuclear binding energy that's unstable enough to be extracted and stable enough to last for x years. It's a nuclear battery just like convential ones are chemical batteries.


You're not clipping leads to the ends of it and getting current, though -- you're using thermocouples. It's more of a "heat battery", I suppose.


And now draw a black box around that whole system. What do you get? A voltage source, right?

Engineers usually don't care much about the principles of how something is made. As long as it can be abstracted to a known concept you're all good and you can integrate it with your given tools, e.g. a combinational circuit plan.


Is a thermocouple more efficient than a Sterling engine, or just more reliable because of the lack of moving parts?


Don't forget more compact. A sterling engine would require a generator, which in turn means a regulator and a rectifier. Plus, R&R's aren't very reliable either. When the alternator in your car dies, it's usually the R&R that failed.

The main disadvantages of thermocouples are cost and efficiency, but cost isn't a big deal for mega government projects and efficiency isn't a big deal when the goals are modest (i.e. not flying about like a helicopter) and the power source is amazing (i.e. nuclear decay)


Given the MSL has a constant power source does that mean it can operate 24hrs and 40 minutes a day? Do they have driving lights or can drive with IR or other sensors? If so they could get a lot more science done than the MER mission. Or do they have to channel the power into heaters each evening?


> does that mean it can operate 24hrs and 40 minutes a day?

Not if it consumes electrical energy from the batteries faster than the MMRTG replenishes it. Then it needs to "rest" to give the MMRTG a chance to recharge the batteries.

> Or do they have to channel the power into heaters each evening?

Besides 125 watts electric, the MMRTG also continuously outputs 2000 watts of heat. Heat can be pumped around the rover (either for cooling or heating) to keep the instruments at optimal temperature [1][2]

[1] https://www.youtube.com/watch?v=fBtXnug-rgM#t=89s

[2] http://www1.nasa.gov/mission_pages/mars/images/20081209_msl....


>Or do they have to channel the power into heaters each evening?

AFAIK, No. The battery generates 2kw of heat in addition to the electricity, so I dont think the rover even has any heaters.


What are the safety implications of launching 10 pounds of plutonium-238 on a rocket that could malfunction or explode before gaining escape velocity? The article says the plutonium would not explode, but what about plutonium particles or radiation entering the atmosphere and ocean?


The same safety implications as launching 10 pounds of anything twenty miles in the air, over an ocean, and subjecting it to an explosion. 10 pounds divided by any realistic footprint is negligible.

Plutonium is dangerous stuff, sure, but it isn't 10 orders of magnitude more dangerous than anything else the way some people act like it is. It's just dangerous, not imbued with an evil malevolent spirit that wants to irradiate your soul.

Bear in mind the Earth is covered in radioactives; it doesn't take all that much division before you've got less radioactivity per acre than already exists naturally, which contrary to apparently popular belief is not 0.


Thanks for dousing a small outbreak of nuclear hysteria.

It's depressing that the article has to insert (non explosive) next to the plutonium. Clearly some people must think that this is a bomb waiting to go off at any moment.


Per wikipedia[1]:

  A commonly cited quote by Ralph Nader, states that a pound of plutonium dust spread into the atmosphere would be enough to kill 8 billion people. However, the math shows that one pound of plutonium could kill no more than 2 million people by inhalation. 
So no need to worry.

[1]https://en.wikipedia.org/wiki/Plutonium#Toxicity


You know what the problem with pulling something from Wikipedia is? It has citations. Specifically, for your pulled quote it cites this: http://www.phyast.pitt.edu/~blc/book/chapter13.html , and the section on "Plutonium Toxicity" about halfway down the page.

In particular, I commend to you the paragraphs starting with "In response, I offered to inhale publicly many times as much plutonium as he said was lethal." But more to my point:

"In summary, a pound of plutonium dispersed in a large city in the most effective way would cause an average of 19 deaths due to inhaling from the dust cloud during the first hour or so, with 7 additional deaths due to resuspension during the first year, and perhaps 1 more death over the remaining tens of thousands of years it remains in the top layers of soil. This gives and ultimate total of 27 eventual fatalities per pound of plutonium dispersed."

I can't quite get a direct cite, but I'm pretty sure the claim that a pound of plutonium dust can kill two million people is for a pound of plutonium dust being carefully doled out to two million people in precisely the quantities that will just barely kill them, because the previous paragraph is the description of what happens if you just sort of fling it at a city.

And an explosion from a rocket would actually be dispersed over a much larger area than merely a large city if it were going to hit anybody at all, because we don't launch anything immediately upwind of large cities. Think state-sized dispersal and you'd be closer.

"There have been fears expressed that we might contaminate the world with plutonium. However, a simple calculation show [26] that even if all the world's electric power were generated by plutonium-fueled reactors, and all of the plutonium ended up in the top layers of soil, it would not nearly double the radioactivity already there from natural sources, adding only a tiny fraction of 1% to the health hazard from that radioactivity."

"I have been closely associated professionally with questions of plutonium toxicity for several years, and the one thing that mystifies me is why the antinuclear movement has devoted so much energy to trying to convince the public that it is an important public health hazard. Those with scientific background among them must realize that it is a phony issue. There is nothing in the scientific literature to support their claims. There is nothing scientifically special about plutonium that would make it more toxic than many other radioactive elements. Its long half life makes it less dangerous rather than more dangerous, as is often implied; each radioactive atom can shoot off only one salvo of radiation, so, for example, if half of them do so within 25 years, as for a material with a 25-year half life, there is a thousand times more radiation per minute than emissions spread over 25,000 years, as in the case of plutonium.

"No other element has had its behavior so carefully studied, with innumerable animal and plant experiments, copious chemical research, careful observation of exposed humans, environmental monitoring of fallout from bomb tests, and so on. Lack of information can therefore hardly be an issue. I can only conclude that the campaign to frighten the public about plutonium toxicity must be political to the core. Considering the fact that plutonium toxicity is a strictly scientific question, this is a most reprehensible situation."

And thanks for leading me to the awesome link to post next time this comes up.


The quote was pulled to put the "same risk as 10 pound of anything" claim into perspective, not to give an exhaustive treatment of plutonium ( and specifically Pu 238) dispersement.

The claim of 2 million cancers per pound of Pu is sketched out just at the start of the subchapter "Plutonium Toxicity" in your reference, it depends on a model how long the Pu remains in the lung. ( I suspect Naders number of 8e+9 death is obtained by dividing 1 Pound by an estimate of the lethal dose.)

Additionally the source talks presumably of Pu 239 (at least the numbers in the appendix are for Pu 239) while the activity ( and therefore the dosage) of Pu 238 is about a factor of 200 higher. So we can estimate 4000 death per pound of Pu 238 dispersed in the atmosphere over an city. ( Dispersing over an Ocean instead of a city would lower this estimate of course considerably. What could possibly go wrong ...)

And you are welcome, this seems to be one of the better texts one can cite about the dangers of Pu. (I will be happy to point out the several best case estimates the text makes.)


You know coal typically contains 1-4ppm of both Uranium and Thorium, right? http://pubs.usgs.gov/fs/1997/fs163-97/FS-163-97.html

And that the US burns around 1 billion tons of coal a year? https://en.wikipedia.org/wiki/Coal

So coal fired powerstations send something like a few thousand tons of Uranium and Thorium into the environment every year (he said, blithely simplifying ppm into "proportion by weight").

Still worried about 10lbs of Plutonium?


You're right that the amount of radioactive material is too small to warrant serious worry, but please distinguish between isotopes; Plutonium-238, the isotope used in the rover, is a very powerful alpha emitter. You definitely don't want to inhale more than a trace amount of it. (On the plus side, a piece of paper can shield you against alpha particles, which makes this one of the easier-to-handle isotopes -- a good choice for radiothermal generators like this.)


Well, there's been a few accidents involving these things, including Apollo 13. They're designed to survive reentry. http://en.wikipedia.org/wiki/Radioisotope_thermoelectric_gen...


Somehow I stumbled on the exact PDF you want to see: http://www.nasa.gov/pdf/604332main_APP%20MSL%20Launch%20Nucl...


Thanks. The relevant bits:

Like previous generations of this type of electrical- power generator, the MMRTG is built with several layers of protective material designed to contain its plutonium dioxide fuel in a wide range of po- tential accidents, verified through impact testing. Each MMRTG carries eight individually shielded general purpose heat source modules (compared to 18 modules in the previous generation). The thickness of the protective graphite material in the center of the modules and between the shells of each module in the MMRTG has been increased by 20 percent over previous modules.

However, the same report says there is about a 3% chance of an accident with no release, and a .4% chance of an accident "with release".


Not significant. iirc, NASA's nuclear batteries are designed to survive re-entry and the resulting collision, meaning you aren't going to have to deal with vaporized plutonium in the air, and cleanup in the event of an accident wouldn't be too bad.

We are not talking about nuclear reactors here. Only basic nuclear decay. No criticality. The key thing is the packaging, to make sure you aren't contaminating an area when something goes wrong.


>> The Curiosity Rover Nuclear Battery will supply the system with constant power, allowing it to work as much as needed, all year long for as long as 14 years.

Very cool. Massive improvement over the solar panels that only worked during the day and non-winter times.


Burning question... can a miniature version of this ever be ported to consumer electronics devices? :-/


There's a prototype betavoltaic battery which can sustain 50 microwatts on 20 Curies of tritium [1] -- an efficiency of about 2%. Scaling linearly by 10^4, current technology could get 500 milliwatts on 200 kCi, or 20 grams, enough to charge a 5 watt-hour phone battery in 10 hours. The tritium cost would be on the order of $600,000 today [2]. If you could do this, the battery charge time would be a few decades.

[1] http://www.citylabs.net/index.php?option=com_content&vie...

[2] http://fire.pppl.gov/fesac_dp_ts_willms.pdf


Yes, you would be burning with such a power source in a consumer electronic device.


So, when can I get a nuclear car? And what's the insurance gonna cost?


when we start building "breeder" reactors again. Plutonium is a byproduct of the breeder reactors when they create the type of uranium that can be used for nuclear bombs (this is one of the reasons that breeder reactors aren't popular). I was talking with one of the scientists at Ames last night, and apparently this is pretty much the last of NASA supply, since the US has significantly cut down on their production of weaponizeable uranium.


That's not really accurate, but rather than explaining why, I'll just give the abbreviated rundown on plutonium:

Plutonium 238 is a powerful alpha emitter with a half-life of 87.7 years, making it a great element for powering mars rovers. It's produced by exposing Neptunium 237 to neutron flux. You can get Np-237 out of nuclear waste from ordinary reactors. The US has mostly been buying Pu-238 from Russia, but we're running out, and starting up our own production again is kind of expensive. We can do it, though.

Plutonium 239 is the kind that gets used in bombs. It's fissile. It's produced by exposing Uranium 238 ("depleted uranium") to neutron flux in a nuclear reactor. It's tricky to make weapons-grade Pu-239, because it tends to be contaminated with Pu-240.

Plutonium 240 is annoying and nobody likes it.


well, i guess i learned not to blindly trust NASA scientists today. thanks for the info; that was pretty interesting.


The type of Uranium used in nuclear weapons (fissile U235) isn't created in reactors; it's purified from natural Uranium, which is almost entirely non-fissile U238 but contains a small proportion of U235.

Most commerical enrichment of Uranium uses gas centrifuges to separate the heavier isotopes from the lighter.




Applications are open for YC Summer 2019

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

Search: