Do not mistake technical viability with economic viability.
The 1st Breeder reactor went into service in 1962, there have been many after and they have all meet the same fate.
Yes, they can breed their own fuel but the total cost of doing it is wildly prohibitive.
You can get gold/uranium/lithium from Ocean water, try and do it at a price people will actually pay for it. You can get minerals from space, so long as the market rate of $10 million a ton is viable... etc.
As always, if I get proven wrong - that will be a great day!
You are right, I thought Shippingport reactor 3 did but it also was a blanket design.
I don't really think this matters that much. The neutronics is pretty well understood and the principle has been proven. Its just a matter of investing the money to do it.
Putting fertile material near fissile material and some sodium or lead or beryllium has never been the hard part. Every attempt has sunk at the filthy and expensive reloading part and the making something that doesn't catch fire, leak everywhere or melt down when used with the blanket in part.
The entire history of mineral and energy extraction tells us that once dense deposits are exhausted extraction costs substantally increase even in the face of more sophisticated technology.
eg: Oil was once extracted by sucking it out of a surface pool with a pump .. and now we are fracking for gas fractions.
These "there are XXX tones at YY ppm (or ppb) of Z in the crust or ocean" calculations are almost always impractical wishful thinking economically infeasible bullshit.
For example:
Have a shot at guesstimating the tonnage and value of Palladium (used in catalytic converters) in the near vicinity of road surfaces - it falls there as by product waste.
Now have a stab at the cost of ripping up and processing the central north american road surface to extract Palladium.
Worth it?
It'll be cheaper once we abandon cities and roads, of course.
I think gold cyanidation considerably reduced costs for gold extraction compared to traditional mining methods which depended on higher-quality deposits.
By "Traditional methods" I guess you mean pre 1887 methods?
You still need a relatively high (ie greater than mean crustal compisition) gold percentage to make circuit leeching (by whatever method) profitable.
On the matter of the articles discussion of uranium in the ocean - that's a dream of chasing something evermore expensive.
If there was an empty ocean basin the size of the Earths oceans and if there was an efficient cheap extraction method,
then we could pass the entire ocean through that process from our ocean to the empty basin (okay, this already sounds impractical).
Instead (if we had this hypothetical cheap extraction) we'd find ourselves endlessly pumping the same ocean through the same process and forever chasing smaller and smaller concentrations.
Your statement was "The entire history of mineral and energy extraction tells us that once dense deposits are exhausted extraction costs substantally increase even in the face of more sophisticated technology."
That statement should be independent of time, so hold for 1887 too.
> that's a dream of chasing something evermore expensive
Early gold mining (and tin, lead, etc) followed rich veins with good rewards for hand tools.
Throughout history gold mining has had bursts of finding new rich grounds, but the essential trajectory has been more effort (in the sense of moving more material) to extract less target material, often with more complex processing and harmful side effects (leaching trace amounts from paydirt).
Years back I did the computational backend for a mine modelling program (under ground and pit) with application here at the superpit [1].
The dimensions of this hole are .. large - the volume of material removed is large, and the energy requirements to lift that volume free and the sort it for discard, crushing, refining, etc are also large.
This is just for gold, which is mostly useless (aside from some jewellery and some actual essential use in space electronics the bulk of gold goes to bullion and is valuable because, well, it's gold (go figure)).
You can (I have, and others do) plot the per tonne increased extraction costs of target materials against deposit richness as reserves are depleted.
The entire notion of peak oil is predicated against increasing effort for diminished returns.
I'm far from knowledgeable about the topic. Still, I'm twinging on your earlier use of the "costs", which is different than "good rewards".
If something is rare, people may pay a lot for it. Labor-intensive manual mining (and we mustn't forget the use of slave labor hides the economic costs and adds a human cost) might not move as much material, but may still have high costs.
> plot the per tonne increased extraction costs of target materials against deposit richness as reserves are depleted.
I do understand that. But what does 'dense deposit' mean?
I took it to mean gold deposits where manual mining provided good rewards. Gold cyanidation is for low-grade ore, says Wikipedia, and the result gave good rewards for South African mine owners, yes?
What I don't know is the cost per unit production of either method.
I fully understand that new methods may make previously low-grade material economically profitable, but I don't think those should be re-categorized as "dense".
In looking around, I believe iodine production might be another case to consider. As I understand it, the historical production was from sea water through bioaccumulation in kelp, which was then dried and processed.
We've since moved to richer sources, either mineral (caliche) or brine.
"Dense deposit" means there's a lot of gold per tonne of not gold.
Manual mining produces good rewards in nugget rich grounds with dense primary rewards, if you move (say) 10,000 tonnes of material you find a lot of gold, even without extra processing (such as gold cyanidation).
When you hit low grade regions there simply isn't as much gold present - not only do you still need to move 10,000 tonnes of material, you know also need to chemically bind and extract in order to get less gold overall.
The pattern is, the easy is cheap (in terms of effort), the harder stuff costs more (in terms of effort), and minor advances in technique aside .. everything ladders upwards to cost more in extraction effort for less return.
It's been true for gold, for copper, for fossil fuels, etc.
Historically you can see hard data for this in something like [1] which is sadly a subscription service.
I interpret this as meaning that with the technology of the 1880s it was not considered a dense deposit, and earlier dense deposits were being exhausted.
This is tied back to your original statement "once dense deposits are exhausted extraction costs substantally increase even in the face of more sophisticated technology."
If it wasn't a dense deposit, then did the costs substantially increase with cyanidation? (Not total cost, but cost per unit production.)
> If it wasn't a dense deposit, then did the costs substantially increase with cyanidation? (Not total cost, but cost per unit production.)
I made a broad long term statement that's true over multiple decades and centuries - if you take a keyhole view there will be times when the long term trend is bucked.
I don't specifiaclly know the exact answer to your question (although it can be worked out by a research student with a month or two to spare) but I would hazard that profits from gold mining were dwindling with a high cost of getting some value from fines .. and then cyanidation made things profitable again.
It's a market with supply | demand and a finite amount of gold in the crust - nuggets are no longer laying aboutto be picked up, and now many tonnes of sand and grit need to be centifuged | screened | shaken to get a concentrate .. and as the profit from that dwindles and price/kilo rise due to limited supply - it become possible get more gold from the concentrate with a little additional cost (in time + chemical) and profits rise again.
Whatever specifically happened in a short time window in a specific location though; the long term trend remains, more effort for less return of product.
I understand how you can think of it as a keyhole view.
Instead, I think it's that I want "dense deposit" to mean something fixed, so we can look at a Roman gold extraction operation and say "yes, that is a dense deposit" or at a South African mine and say "no, that is not a dense deposit" independent of the technology in use.
Here's a thought experiment for that research student - which would cost more using current wages:
- extract 1 ton of gold from a deposit as rich as (say) the Dolaucothi Gold Mines when it used by the Romans, and using only Roman techniques.
- extract 1 ton of gold from a deposit equivalent to a South African mine in 1900, using cyanidation techniques of that era.
(I don't know if 1 ton is too low or too high to be reasonable.)
> nuggets are no longer laying about to be picked up
Oh, and as a really edge case, argon gases is a renewable resource which is extracted from the air. It's cost has almost certainly gone down over time as we have improved methods for refrigeration.
Breeder reactors are extremely fuel efficient. The cost is in building an actual breeder reactor itself, as it is an experimental technology. And the cost is not "widely prohibitive". The Wikipedia page says they are 25% more expensive than non breeder reactors.
Breeder reactors have no bearing on the costs of primary mineral extraction from the crust - it's a seperate line of discussion altogether to this sub thread.
Perhaps you intended your comment in reply to someone discussing the pros|cons of fuel generation via breeder reactors rath than to my comment which addresses the inevitable rising effort required to extract more resource from the crust (or ocean) over time.
The reason for this is that commercial adoption happened with PWRs and after that the nuclear companies had no need to commercialize anything else.
By the time Breeder research by government was happening the anti-nuclear movement of the 70/80 was already in full effect and research money was being cut and very few nuclear plants were being built so there wasn't much reason invest in commercial breeders.
Specially because fuel isn't that expensive in the first place and waste isn't actually a big problem either.
There are still other good reason to create breeders and if we are gone develop next generation reactors, we might as well go in that direction.
Whilst the setup `may` sustain itself fuel wise for that kind of duration, I'm not aware of any building able to last 4B years, let alone a nuclear plant, which generally has a lifespan of a few decades.
Do isotopes of any element exist with all three of these properties?
1. Radioactive with a half-life short enough to be dangerous (e.g., not bismuth-209) but long enough that waiting for it all to decay isn't a feasible way of getting rid of it (e.g., not francium)
2. Produced by nuclear reactors
3. Not usable as a fuel source in any breeder reactors
If not, then why is there such a thing as nuclear waste?
No it isn't. The cost isn't high at all. Even the current spent fuel is tiny in volume and tiny in storage cost. They are literally a bunch of cask on a parking lot. All of Switzerlands fuel for 50+ years is literally in one storage building. The US has already collected a huge amount of money that could be invested in solution but they just leave it on a parking lot.
And if we are going to use nuclear for the long term, using that 'waste' as fuel is clearly the thing to do, and we can simply burn all that stuff up and the remaining fuel only needs to be stored for 300 years.
Thorium is not needed for this, just better reactors.
To me any technological projection that goes beyond 200 years is a bit of non-sense. 200 years ago trains did not exist. Steam power existed, but just in a tiny corner of the world economy.
If we don't run out of nuclear fuel in 200 years, then we'll never do.
And we certainly have enough uranium to not run out of it for 200 years, with the current technology. No breeder reactors, or anything fancy needed.
CANDU reactors run on unenriched uranium [1]. This instantly gives a multiplier of 10. If the current reactors can run for a few decades, then switching to CANDU reactors we'd have fuel for a few centuries.
Why aren't we switching to CANDU design? Some new builds are projected [2], but overall they appear to be too capital intensive, compared to the more traditional light water reactors. Still if fuel availability were a concern, we'd switch to CANDU reactors and stop having any scarcity for hundreds of years.
All current nuclear reactors run on U235 or the barely significant dregs of Pu239 left over.
They do varying amounts of breeding to produce up to 50% of their energy from Pu, but the fuel source is always U235. Leaving more U238 with it in a HWR doesn't change this.
> And we certainly have enough uranium to not run out of it for 200 years, with the current technology
At the current (insignificant) portion of final energy they produce. Increase it 10x for just the current electrical grid and it's <20 years. Switch to SMRs which are inefficient and it's <10.
If you want people to take nuclear seriously, try telling the truth about anything at all at least once, ever.
Uranium will run out long after the sun's rising luminosity will reduce atmospheric CO2 levels to below the level plants need to live (600 million years) and boil the oceans away (1 billion years).
39 comments
[ 1.8 ms ] story [ 82.7 ms ] threadYes, they can breed their own fuel but the total cost of doing it is wildly prohibitive.
You can get gold/uranium/lithium from Ocean water, try and do it at a price people will actually pay for it. You can get minerals from space, so long as the market rate of $10 million a ton is viable... etc.
As always, if I get proven wrong - that will be a great day!
You're being too generous. It's could theoretically not can. An actually closed fuel loop has never happened.
[1] https://www.asme.org/wwwasmeorg/media/resourcefiles/aboutasm...
What has not happened is that that fuel was removed from that reactor and then inserted into another reactor.
This is a lie just like the rest. EBR 1 had a separate core and blanket. It didn't run on the bred fuel any more than a PWR does.
I don't really think this matters that much. The neutronics is pretty well understood and the principle has been proven. Its just a matter of investing the money to do it.
eg: Oil was once extracted by sucking it out of a surface pool with a pump .. and now we are fracking for gas fractions.
These "there are XXX tones at YY ppm (or ppb) of Z in the crust or ocean" calculations are almost always impractical wishful thinking economically infeasible bullshit.
For example:
Have a shot at guesstimating the tonnage and value of Palladium (used in catalytic converters) in the near vicinity of road surfaces - it falls there as by product waste.
Now have a stab at the cost of ripping up and processing the central north american road surface to extract Palladium.
Worth it?
It'll be cheaper once we abandon cities and roads, of course.
You still need a relatively high (ie greater than mean crustal compisition) gold percentage to make circuit leeching (by whatever method) profitable.
On the matter of the articles discussion of uranium in the ocean - that's a dream of chasing something evermore expensive.
If there was an empty ocean basin the size of the Earths oceans and if there was an efficient cheap extraction method,
then we could pass the entire ocean through that process from our ocean to the empty basin (okay, this already sounds impractical).
Instead (if we had this hypothetical cheap extraction) we'd find ourselves endlessly pumping the same ocean through the same process and forever chasing smaller and smaller concentrations.
Your statement was "The entire history of mineral and energy extraction tells us that once dense deposits are exhausted extraction costs substantally increase even in the face of more sophisticated technology."
That statement should be independent of time, so hold for 1887 too.
> that's a dream of chasing something evermore expensive
I have no issue with that point.
Early gold mining (and tin, lead, etc) followed rich veins with good rewards for hand tools.
Throughout history gold mining has had bursts of finding new rich grounds, but the essential trajectory has been more effort (in the sense of moving more material) to extract less target material, often with more complex processing and harmful side effects (leaching trace amounts from paydirt).
Years back I did the computational backend for a mine modelling program (under ground and pit) with application here at the superpit [1].
The dimensions of this hole are .. large - the volume of material removed is large, and the energy requirements to lift that volume free and the sort it for discard, crushing, refining, etc are also large.
This is just for gold, which is mostly useless (aside from some jewellery and some actual essential use in space electronics the bulk of gold goes to bullion and is valuable because, well, it's gold (go figure)).
You can (I have, and others do) plot the per tonne increased extraction costs of target materials against deposit richness as reserves are depleted.
The entire notion of peak oil is predicated against increasing effort for diminished returns.
[1] https://www.youtube.com/watch?v=8Wykx-_RWDw
I'm far from knowledgeable about the topic. Still, I'm twinging on your earlier use of the "costs", which is different than "good rewards".
If something is rare, people may pay a lot for it. Labor-intensive manual mining (and we mustn't forget the use of slave labor hides the economic costs and adds a human cost) might not move as much material, but may still have high costs.
> plot the per tonne increased extraction costs of target materials against deposit richness as reserves are depleted.
I do understand that. But what does 'dense deposit' mean?
I took it to mean gold deposits where manual mining provided good rewards. Gold cyanidation is for low-grade ore, says Wikipedia, and the result gave good rewards for South African mine owners, yes?
What I don't know is the cost per unit production of either method.
I fully understand that new methods may make previously low-grade material economically profitable, but I don't think those should be re-categorized as "dense".
In looking around, I believe iodine production might be another case to consider. As I understand it, the historical production was from sea water through bioaccumulation in kelp, which was then dried and processed.
We've since moved to richer sources, either mineral (caliche) or brine.
Manual mining produces good rewards in nugget rich grounds with dense primary rewards, if you move (say) 10,000 tonnes of material you find a lot of gold, even without extra processing (such as gold cyanidation).
When you hit low grade regions there simply isn't as much gold present - not only do you still need to move 10,000 tonnes of material, you know also need to chemically bind and extract in order to get less gold overall.
The pattern is, the easy is cheap (in terms of effort), the harder stuff costs more (in terms of effort), and minor advances in technique aside .. everything ladders upwards to cost more in extraction effort for less return.
It's been true for gold, for copper, for fossil fuels, etc.
Historically you can see hard data for this in something like [1] which is sadly a subscription service.
[1] https://www.spglobal.com/marketintelligence/en/campaigns/met...
I do understand that.
In 1880s South Africa (before cyanide was used), was that remaining ore considered a dense deposit? I don't know.
https://en.wikipedia.org/wiki/Gold_extraction#Industrial_era informs me "mining ... began to slow down ... as the new deposits being found tended to be pyritic ore. The gold was difficult to extract from such ores."
I interpret this as meaning that with the technology of the 1880s it was not considered a dense deposit, and earlier dense deposits were being exhausted.
This is tied back to your original statement "once dense deposits are exhausted extraction costs substantally increase even in the face of more sophisticated technology."
If it wasn't a dense deposit, then did the costs substantially increase with cyanidation? (Not total cost, but cost per unit production.)
I made a broad long term statement that's true over multiple decades and centuries - if you take a keyhole view there will be times when the long term trend is bucked.
I don't specifiaclly know the exact answer to your question (although it can be worked out by a research student with a month or two to spare) but I would hazard that profits from gold mining were dwindling with a high cost of getting some value from fines .. and then cyanidation made things profitable again.
It's a market with supply | demand and a finite amount of gold in the crust - nuggets are no longer laying aboutto be picked up, and now many tonnes of sand and grit need to be centifuged | screened | shaken to get a concentrate .. and as the profit from that dwindles and price/kilo rise due to limited supply - it become possible get more gold from the concentrate with a little additional cost (in time + chemical) and profits rise again.
Whatever specifically happened in a short time window in a specific location though; the long term trend remains, more effort for less return of product.
Instead, I think it's that I want "dense deposit" to mean something fixed, so we can look at a Roman gold extraction operation and say "yes, that is a dense deposit" or at a South African mine and say "no, that is not a dense deposit" independent of the technology in use.
Here's a thought experiment for that research student - which would cost more using current wages:
- extract 1 ton of gold from a deposit as rich as (say) the Dolaucothi Gold Mines when it used by the Romans, and using only Roman techniques.
- extract 1 ton of gold from a deposit equivalent to a South African mine in 1900, using cyanidation techniques of that era.
(I don't know if 1 ton is too low or too high to be reasonable.)
> nuggets are no longer laying about to be picked up
This isn't entirely true. People do still fund nuggets by happenstance. A news search finds things like https://www.yahoo.com/entertainment/family-finds-24k-gold-nu... from a few years ago.
But yes, they aren't the types which kick of a new gold rush.
In any case, Roman gold extraction wasn't just from picking up nuggets either, so I'm not sure that's quite the right comparison.
Helium was first extracted as gas from rock. Wikipedia says it was first isolated from cleveite. Onnes's 1908 paper on liquefying helium says his helium came from monazite sand. https://web.archive.org/web/20180809111624/https://babel.hat...
Oh, and as a really edge case, argon gases is a renewable resource which is extracted from the air. It's cost has almost certainly gone down over time as we have improved methods for refrigeration.
(Oxygen is likely also in the same category.)
Perhaps you intended your comment in reply to someone discussing the pros|cons of fuel generation via breeder reactors rath than to my comment which addresses the inevitable rising effort required to extract more resource from the crust (or ocean) over time.
"It might be cost-prohibitive now, but is there any reason to think that it can't get cheaper?"
By the time Breeder research by government was happening the anti-nuclear movement of the 70/80 was already in full effect and research money was being cut and very few nuclear plants were being built so there wasn't much reason invest in commercial breeders.
Specially because fuel isn't that expensive in the first place and waste isn't actually a big problem either.
There are still other good reason to create breeders and if we are gone develop next generation reactors, we might as well go in that direction.
But of course this articles assume you would build new reactors over time.
1. Radioactive with a half-life short enough to be dangerous (e.g., not bismuth-209) but long enough that waiting for it all to decay isn't a feasible way of getting rid of it (e.g., not francium)
2. Produced by nuclear reactors
3. Not usable as a fuel source in any breeder reactors
If not, then why is there such a thing as nuclear waste?
I don't know, but to me "not usable" and "not economically usable" don't sound quite the same.
They're just a fiction used for marketing and to buy social license for the plutonium separation facilities.
Fuel recycling and alternate fuels such as thorium might reduce storage burden by being able to burn up plutonium waste from traditional nuke plants.
And if we are going to use nuclear for the long term, using that 'waste' as fuel is clearly the thing to do, and we can simply burn all that stuff up and the remaining fuel only needs to be stored for 300 years.
Thorium is not needed for this, just better reactors.
If we don't run out of nuclear fuel in 200 years, then we'll never do.
And we certainly have enough uranium to not run out of it for 200 years, with the current technology. No breeder reactors, or anything fancy needed.
CANDU reactors run on unenriched uranium [1]. This instantly gives a multiplier of 10. If the current reactors can run for a few decades, then switching to CANDU reactors we'd have fuel for a few centuries.
Why aren't we switching to CANDU design? Some new builds are projected [2], but overall they appear to be too capital intensive, compared to the more traditional light water reactors. Still if fuel availability were a concern, we'd switch to CANDU reactors and stop having any scarcity for hundreds of years.
[1] https://en.wikipedia.org/wiki/CANDU_reactor
[2] https://www.world-nuclear-news.org/Articles/Romania-adopts-d...
Check the average discharge burnup.
They do varying amounts of breeding to produce up to 50% of their energy from Pu, but the fuel source is always U235. Leaving more U238 with it in a HWR doesn't change this.
> And we certainly have enough uranium to not run out of it for 200 years, with the current technology
At the current (insignificant) portion of final energy they produce. Increase it 10x for just the current electrical grid and it's <20 years. Switch to SMRs which are inefficient and it's <10.
If you want people to take nuclear seriously, try telling the truth about anything at all at least once, ever.
I'm not sure HN is the place for this type of language.