2 billion-year-old African nuclear reactor | Hacker News Reader

2 billion-year-old African nuclear reactor

extremetech.com

141 pointsgivan38 comments

interesting topic, you'd perhaps do better reading the wikipedia article

http://en.wikipedia.org/wiki/Natural_nuclear_fission_reactor

they've found out about it because the uranium ore had smaller ratio of U235 than usual so they suspected that some of it might have been stolen for weapons.

The reactor worked because U235 was more abundant naturally 2 billion years ago, now it has decayed more, so it's no longer possible. The reactor was periodically flooded with water which moderated the reaction. After 30 minutes of reaction all water would boil away.

It is interesting for the study of nuclear waste disposal - this is 2 billion years old nuclear waste still sitting exactly in the place where it was created

U235 is hardly waste it's fuel. Dangerous nuclear waste tends to contaminate useful things but the dangerous bits decay quickly which is what makes them dangerous in the first place. Read up on 'fuel reprocessing' and note that by simply leaving things alone for even just 100 years it becomes far cheaper to reprocess.

In this case the nuclear "waste" in Oklo are the would be the same as you would get in any nuclear fuel process. It is the exact same process that created them. The upside is that the reactor only started and stopped and did not run continuously for long times apparently.

The point about the viability of sequestration at Yucca Mountain is also interesting.

There is also an insightful bit that atoms don't split like legos, its more like a smooth graph, which has all kinds of implications.

1) The isotopic concentrations of waste from 2 billion years ago are about the same as you'd measure if you built it and lit it up today and then in theory let it sit around for two billion years, so to a zillion decimal places nothing has changed either in fission properties or decay rates in billions of years, which is interesting.

2) There exists no natural way to "kick" decay rates higher or lower. This being 2 billion years old, it's probably suffered thru billions of years of every crazy particle and field that can happen naturally in 2 billion years, none of which had any effect. So if you think there exists a way to speed up or slow down radioactive decay, its a minimum factor of two billion "un-natural" (more than two billion times the natural background of "techno-babble" radiation, or whatever) or perhaps most likely, it doesn't exist. So it appears to a zillion decimal places under varying natural conditions that nuclear decay rates are a constant. A pity, think how cool it would be to have a star trek like "thing" that decontaminates things. Or a controllable RTG that varies its power output by varying its decay rate. Or more negatively, if you could take a sphere of Co-60 the size of a softball, and increase its decay rate by a trillion or so, that would be a most impressive bomb. Or shine a magic disruptor beam at enemy troops or democracy activists or whatever and instantly all the C14 isotope in their body cooks off at the same time, that would be interesting.

> There exists no natural way to "kick" decay rates higher or lower.

Some nuclei decay by electron capture, and their decay rates can be changed simply by bringing them in different chemical compounds. http://en.wikipedia.org/wiki/Electron_capture

To a modern physicist, perhaps the most interesting thing about the Oklo reactor is that it gives us one of the biggest levers on the question "Have the laws of physics changed over time?"

So far as anyone can tell, nuclear physics was the same back then as it is today.

I didn't know this was in doubt, do you have any additional information on this debate? It sounds interesting.

It's not ind doubt so much as there is little intrinsic reason to support either side, and scant evidence. Maybe the laws of physics worked differently in the early universe -- how would you know?

The answer of course is astronomical observations -- it's the only window into the past that we have, other than a few historical events like the OP, as the GP points out. But how do you interpret astronomical observations? By using the currently understood laws of physics, of course, so it becomes a bit of a circular argument.

If you assume that certain constants change over time, that does help explain away current mysteries like inflation & dark energy, under some models. However too much deviation would contradict other observable evidence.

> But how do you interpret astronomical observations? By using the currently understood laws of physics, of course, so it becomes a bit of a circular argument.

One of the best ways to measure this is to use quasar absorption spectroscopy to probe the precise energy level spacing of atoms (like iron, magnesium, and so on) from billions of years ago with the energy level spacings found in the same atoms of today.

If the fine-structure constant (for example) were different by some amount in the distant past, its effect is calculable -- and the spectroscopic signature of a different fine-structure constant is distinct from any other physical effect (redshift of the spectrum, etc.).

There are some intriguing results that have come out over the past decade or so, with new information coming in all the time.

So all the stories I've read about how it implies the fine structure constant has changed, those are bullshit?

You misread them, those stories are talking about checking IF the constant has changed. They are not saying that it has. None have shown any change.

> To a modern physicist, perhaps the most interesting thing about the Oklo reactor is that it gives us one of the biggest levers on the question "Have the laws of physics changed over time?"

I don't think it really does, because our interpretation of the causes that produced the effects that we can see now as something which supports the conclusion that the laws of physics were the same then itself rests on the assumption that the laws of physics were the same then.

It provides one more data point that constrains just how different the laws of physics could have been in the past -- no matter how different the laws of physics were, it must have still enabled an event which created the isotopic concentrations we see today.

Yes, that is a much looser constraint than the GP described, but it is still a significant data point.

> laws of physics were the same then itself rests on the assumption that the laws of physics were the same then

No, that's definitely not true. Here's a good (somewhat technical) overview of these kinds of scientific investigations:

http://relativity.livingreviews.org/Articles/lrr-2011-2/

The pdf should be free to download and there's an entire section on the Oklo reactor.

This topic has been making the rounds the last few weeks around the various technical-oriented sites, with mentions on HN at https://news.ycombinator.com/item?id=7687948 and https://news.ycombinator.com/item?id=7683888 .

I don't believe the author of this Extreme Tech piece quite understands the deep time involved. The article says:

> This boil-seep-boil system would have made the area extremely volatile — perhaps some remnant of that violence explains why the area features so heavily in local lore and has even grown to prominence in modern day religions such as Falun Gong.

It also points out that the natural reactor was 2 billion years ago.

That's long enough for a mountain chain to form and erode back down into a flat surface. Any physical remnants from that era should be much less interesting than innumerable more recent geological events.

I find it hard to believe ancient Falun Gong practitioners were operating geological nuclear reactors in Gabon 2 billion years ago. I wouldn't put it past a Pak Protector(1), but the timeline indicates the Tnuctipun(2) are more likely suspects ;)

1) http://en.wikipedia.org/wiki/Pak_Protector 2) http://en.wikipedia.org/wiki/Tnuctipun

It was likely part of the caretaker's operations. The reactor was during the Great Oxidation event, when the yeast was first oxidizing the air and before the whitebeasts were introduced.

Indeed. Spoilers here.

http://www.larryniven.net/stories/downinflames.shtml

DIF is fun, but it changes too much established canon for my liking. I have my own grand unified Known Space theory, originally in two forum posts which a friend has compiled together:

http://news.larryniven.net/concordance/content.asp?page=The%...

Ooh, a fellow Niven fan!

I have a problem with your theory that the Outsiders built the ring as a doomsday Protector breeding ground: why would they create copies of Known Space worlds, and then seed them with the inhabitants of those worlds?

Cop out answer: The same reason anyone else would, i.e. that's not a particular reason to exclude them as against anyone else.

Best answer I can come up with: they are laboratories for experimenting with known space sentient species. After all these are species that evolved from programmed good yeast, so the outsiders may want to figure out how that works and if there are any nasty surprises hidden in their genomes.

Which is an interesting idea I think and I wouldn't have though of it if you hadn't asked, so thanks!

Yes, also the part about humans producing more complexity than nature is ridiculous. Nature wins at complexity hands down with what amounts to an overly complex collection of hacks on hacks over billions of years. It's like if instead of fixing design flaws in your code you kept designing new systems to alleviate the worst flaws of the existing systems. For billions of years. Sort of like evolving an OOM killer.

> A huge, naked pile of nuclear waste has sat beneath Oklo for billions of years; why not put a bit more under the Nevada desert, in shielded canisters no less?

I think this is a bad case of survivorship bias.

I think you can't reasonably claim that without any evidence.

How?

If there are piles that didn't survive wouldn't that imply that such leaks are basically harmless, since we're all fine?

Or do you mean survivorship of the entire planet?

Regarding the comment toward the end of the article questioning why the west doesn't sell CANDU reactors to Iran; I would guess that's probably to do with the political situation in Canada with the ruling Conservative Party. Some parts of the party tend to be quite jingoistic, and I'm sure would never support building in Iran, despite there being CANDU reactors in India, Pakistan and China. Also, despite heavy water reactors not being a good source of enriched uranium, they still do produce tritium which can be used to do nasty things.

I don't think you can do very much that is too nasty with tritium unless you already have a working fission bomb (i.e. boosting the fission weapon or using the tritium as part of a multi-stage weapon).

India, Pakistan, and China are already nuclear weapons powers. Iran is too, for all intents and purposes, but that doesn't necessarily mean that people want to support that.

Could someone more knowledgeable about nuclear physics than I am comment on the safety aspects of the positive void coefficient of the CANDU design? If the water boils, the reaction could run away.

That said, the fact that a CANDU runs on unenriched natural uranium and the way it can be refuelled without shutting down power generation are seriously cool features. And they claim that the long time constant makes avoiding a Chernobyl-style excursion unlikely if the operators are awake. But I still worry about a design that isn't statically safe.

Chernobyl would have been safe if the operators were "awake" and doing their job.

> A huge, naked pile of nuclear waste has sat beneath Oklo for billions of years; why not put a bit more under the Nevada desert, in shielded canisters no less?

Most of our nuclear waste is nothing like the nuclear "waste" in Oklo [1]. The primary danger comes not from the reactor but from the enrichment process and uranium hexaflouride (UF6). During enrichment, fissile uranium atoms are extracted to make "enriched" uranium while the rest (over 90%!) of the UF6, which is very toxic and reacts with air and water to form extremely corrosive compounds, is stored in gas cylinders as "depleted" uranium.

Think about that again: we use metal gas canisters to store toxic waste, waste that reacts with some of the most abundant molecules to form more chemicals that slowly destroy the canister. Depending on how these containers are stored and maintained, the UF6 leaks into the air or, even worse, into our ground water reservoirs, taking the non-fissile but still radioactive uranium with it.

Injecting our nuclear waste into the Earth like we do carbon sequestration would be great, but we just don't have the millions of years to wait for the uranium to pass through the ground water system or the technology to figure how and where our reservoirs interconnect (or even how big most of them are). Until we do, storing waste like Oklo is just a step shy of giving everyone some old Fiesta Ware.

[1] http://web.ead.anl.gov/uranium/

http://web.ead.anl.gov/uranium/faq/storage/faq22.cfm

It says they're busy converting the UF6 to uranium oxides as we speak. Uranium oxides are the primary uranium ore. So, what you're getting out is the same stuff that came in from the mine, but considerably less radioactive (since the U235 and all the fission products (radium, etc.) have been removed).

"but we just don't have the millions of years to wait for the uranium to pass through the ground water system"

Those uranium oxides sat in the ore bodies for billions of years without causing trouble.

There are almost certainly gigantic uranium ore deposits underneath the ocean. No one is worrying about them.

That deconversion process will take another 15-20 years [1] if the Paducah and Portsmouth plants are at the planned utilization, which AFAIK hasn't been true since at least the 2013 budget sequestration. Our UF6 isn't going away for another few decades.

> Those uranium oxides sat in the ore bodies for billions of years without causing trouble.

We've found ground water that is contaminated not only by uranium from coal plant ash ponds [2], uranium mines [3], ore processing [4], and flooding [5] but from the untouched ores lining natural underground aquifers as well [6].

> There are almost certainly gigantic uranium ore deposits underneath the ocean. No one is worrying about them.

Any ore deposit is compressed for eons under pressure that is hundreds, or thousands, of times greater than anything we experience on a day to day basis, or even at the extremes of our technology. There's a huge difference between nature's long term storage and our temporary tin cans.

[1] http://www.nrc.gov/materials/fuel-cycle-fac/ur-deconversion....

[2] http://en.wikipedia.org/wiki/Uranium_poisoning_in_Punjab

[3] http://www.denverpost.com/ci_14894696

[4] http://www.epa.gov/radiation/docs/tenorm/402-r-08-005-voli/4...

[5] http://energy.gov/lm/articles/doe-releases-data-summary-repo...

[6] http://www.sciencedirect.com/science/article/pii/08832927949...

"Any ore deposit is compressed for eons under pressure that is hundreds, or thousands, of times greater than anything we experience on a day to day basis"

Sorry, that's simply not true. Ore can be in any form from hard rock to sand. In fact a significant amount of uranium ore occurs as placer deposits.

Are you disputing that there are almost certainly very large uranium ore deposits on the sea floor, exposed directly to (uber corrosive!) saltwater? Why aren't you worried about those?

"There's a huge difference between nature's long term storage and our temporary tin cans."

Oh, please.

1) There are about 4.5 billion tons of uranium already in the ocean.

http://spectrum.ieee.org/energy/nuclear/nuclear-fuel-from-th...

2. This 4.5 billion tons represents a concentration of 3.3 parts per billion.

3. The U.S. has 470,000 tons of uranium (combined into 700,000 tons of UF6).

http://web.ead.anl.gov/uranium/faq/storage/faq16.cfm

4. As an exercise for the reader, figure out how much the ocean's concentration of uranium in ppb would change if we just dumped all 470,000 tons directly into the sea (i.e., the worst case scenario to end all worst case scenarios).

It's a non-issue, dude.

Would make for an interesting scify novel. What if humanity came on the scene hundreds of millions of years earlier and U235 was at 1%.

Hundreds of millions? It depends precisely what number. Terrestrial macroscopic life is barely 500mil years old and 100mil years earlier the atmosphere would have been too low in oxygen to breathe. Land animals with larger brains are rather recent too, just 250mil years or so. But that still leaves a large margin for sentient life to develop.

Steven Baxter wrote about this in his sci-fi novel Manifold:Origin. Terrific trilogy, I highly recommend them. And he’s ex-NASA, so the sci (of which there’s plenty) is above-par.

Reply on news.ycombinator.com