> The natural nuclear reactor formed when a uranium-rich mineral deposit became inundated with groundwater, which could act as a moderator for the neutrons produced by nuclear fission. A chain reaction took place, producing heat that caused the groundwater to boil away; without a moderator that could slow the neutrons, however, the reaction slowed or stopped. The reactor thus had a negative void coefficient of reactivity, something employed as a safety mechanism in human-made light water reactors.
> After cooling of the mineral deposit, the water returned, and the reaction restarted, completing a full cycle every 3 hours. The fission reaction cycles continued for hundreds of thousands of years and ended when the ever-decreasing fissile materials, coupled with the build-up of neutron poisons, no longer could sustain a chain reaction.
Interestingly, the -oid suffix usually refers to things that only resemble the thing, so an andr-oid is not really a man, a meteor-oid is not really a meteor, and fact-oid... might not really be a fact.
However the really amusing bit is that that everybody can agree--for different reasons--that:
I wonder if anyone's done studies to measure and quantify the "distance" of slang (in years or generations) before an older person feels instinctive linguistic revulsion...
My recollection from many years ago is that Norman Mailer coined "factoid" to refer that what looked like facts, were offered as facts, but weren't facts.
approximately 1.7 billion years ago, during the Statherian period of the Paleoproterozoic, and continued for a few hundred thousand years, probably averaging less than 100 kW of thermal power during that time.
Nuclear criticality safety was one of my favorite courses back during my nuclear engineering degree. The failure modes, like the article mentions, can be incredibly non-intuitive. Swirling vortices, firefighting foam, and even the mere presence of a human body can lead to a fatal power excursion. A lot of complex computational modeling and simulation goes into determining safety limits for fissile systems, as well as a lot of sheer imagination.
The most important thing I took away from that experience is the concept of engineered compliance. People will break administrative rules and regulations all the time if they get int the way of work, so you need to engineer the physical system to account for that (e.g. flat or skinny geometry, or lots of neutron absorbers). This can apply to any compliance activity related to safety and security, so the lessons are universal.
Los Alamos published a good survey on various worldwide criticality accidents (PDF): https://www.nrc.gov/docs/ML0037/ML003731912.pdf. Worth a read for all the various ways things can go wrong.
> People will break administrative rules and regulations all the time if they get in the way of work
The example I like to use is that the more paperwork you require for something like "obtain an internal use certificate for API endpoints" then the direct consequence will be less encryption of internal API endpoints.
Note that I didn't say "there may be", or "I predict that". No. It's a guaranteed consequence, like gradient descent making AIs better at meeting loss function. Except that it's gradient ascent, and you're making it steeper and expecting people to climb harder instead of just going around your pointless obstacle.
That's almost an axiom in some circumstances but with criticality accidents there's a difference—first, the stakes don't get much higher, at close quarters death's a certainty and it's about as unpleasant as it gets; second, it's all over and irreversible in milliseconds.
There is no time to think, the body-brain reaction time is far too slow to take corrective action no matter how effective it may be when deployed.
That's different to giving consideration to a course of action based on the difficulty of obstacles in one's way—choosing the path of least resistance, etc.
The issues with that infamous photo are different, there was no obvious logical if decision to be made—it was all very straightforward, lay the Pu out and photograph it. The problems, however, are multifold: (a) those involved didn't have sufficient understanding of the possible consequences (of the physics involved), (b) they were poorly trained in safety procedures, and (c), the work environment wasn't 'engineered' to ensure compliance—that is, nothing stopped them from doing it.
As I pointed out in my earlier post, training is crucial and it cannot be just theoretical. When the stakes are this high training has to be ingrained to such a degree that certain procedures are automatic, similarly, dangerous situations don't just become obvious after thinking about them, they are so the instant the brain responds. For instance, one recognizes the danger as quickly as one would if one tripped over a snake.
That's what the military does with training, one automatically reacts to a situation with the correct response without thinking. And that takes practice until one's perfect.
It works, it's a damn long time since I did any military training and I can stil carry out certain actions and procedures I learned decades ago and do so automatically and without a moment's thought. The people who set up and took those photos should have been trained to a similar level of discipline.
Another typical example is charging internal applications for IDPs, guess what, people are just gonna use LDAP/AD directly, at a not-so-small detriment to application and user security.
These are things where cost center accounting just plainly hurts the company. (There are many cases where it works well and aligns incentives).
Same thing with problem reports and maintenance requests. The more work you make them, the more people will put up with things that kind of halfway work. That's bad for the productivity of the organization as a whole, though, no matter whether the more cumbersome process makes things more convenient for IT or management or auditors.
If actually enforced this way, then the next consequence is that people will use a credit card to spin their service up in some random cloud that you don’t control and won’t break.
… where all API endpoints are encrypted by default with zero paperwork required.
Get it? Forcing people to climb steeper slopes never ever works.
You need to make sure they fall into the pit of success even if they’re stumbling around in the dark with their eyes closed.
It should not have been possible for that photo to have been taken anytime or anywhere since 1946, even during Manhattan with war deadlines it should not have been possible. That it was and as late as 2011 is of grave concern. Frankly, I find it very difficult to fathom.
Years ago when I approached my first reading of the full accounts of Slotin and Daghlian I did so as with any other info, since then I have forced myself to reread them several times as a reminder—as both the accounts and photos are grim. They ought to be part of compulsory safety training for anyone vaguely associated with the nuclear game.
Perhaps, as part of that training there should be more acting out roles with dummy materials similar to military exercises where adjudicators tag one as dead for making a 'lethal' mistake and can play no further role in the exercise. This may seem trite but it's not, being forced to play dead on the battlefield whilst the exercise continues because of one's incompetence is both formative and embarrassing (I speak from experience).
Moreover, like military training, the exercises should be practiced until one's actions are automatic.
That said, such training ought to be a second-line defense—as fastneutron rightly points out—as such work environments ought to be engineered to force compliance.
Is the problem that plutonium doesn't look particularly scary on its own? It's just a chunk of gray metal. It's not fuming, it's not warm (I assume), it's not glowing, it's not particularly ugly, it's not a bright color. Very few people have experienced radiation poisoning and have a visceral understanding of the horror involved. None of our self-preservation instincts understand the danger.
Your question implies I've firsthand experience working with the stuff but that isn't so. It's true I've worked in the nuclear industry and keeping the stuff safe and out of harm's way was in my job description but I wasn't alone, many hundreds of us were (and are) involved in doing so. Luckily there were people between me and those who worked with the stuff and I'm very happy about that.
Still, you're right it's a silvery gray metal much like many others that oxidizes quickly in air to give it that dullish look. If you were to pick it up you'd notice it's very dense. It's marginally heavier than gold by about 0.5g/cc^3 at about 19.8g/cc^3. That it looks so boring could easily generate complacency when handling it.
Edit: the 238 isotope is much more radioactive than 239, so much so it'd be hot to touch. But you'd be mad to do so.
That's a nice example of the prevention paradox at work. You do your work well. Nothing happens for a while. You get laid off. Then eventually something happens...
And that's why you shouldn't just be testing manually but you should be writing automated tests that run again and again, even long after you have moved on.
It looks like amount wise they got lucky. I wonder how they were transported for that photograph. And whether or not the person that made it knew how far away from something really bad they were. That gives me the creeps.
It’s worth reading about refineries and chemical plants. We’ve all heard of Bhopal and other disasters but it’s worth bearing in mind just how few disasters happen in these fearsomely complex and dangerous plants. Almost all are managed successfully by the private sector.
There was an interview with Jim Radcliffe recently in which he described Ineos’ approach to safety. It’s a variation on what’s described my Steven J. Spears in his case study of Alcoa. The truth is that the private sector has innovated much of what’s best in modern safety practice in a way that has made monumentally dangerous industrial processes routine.
There are perhaps good reasons to keep nuclear weapons management in the public sector. Canards about the safety record of the private sector are not though.
> Almost all are managed successfully by the private sector.
Yes, because of a mix government regulations forcing them to not cut too many corners for the sake of saving expenses, and people like the author of the article pushing back against higher-ups who try to do so anyway.
Let's not pretend that workers didn't die needlessly in factory fires a century ago because the factory owners locked the doors.
The Bends were discovered when mysteriously, during the construction of the Brooklyn bridge, the underwater welders kept dying when coming out of the water.
After scientists found what happened and saturation diving was theorised - companies kept their protocols of not decompressing when coming back, and kept killing their divers.
It's not just government regulations. It's also a functioning judiciary and legal system where individuals can sue corporations and win.
As a random example where a non-functioning judiciary resulted in a catastrophe, take a look at the 2020 Beirut port explosion, which a lot of people saw coming a mile off, but were unable to do anything about because of inept and corrupt judges that couldn't make a decision.
For the chemical plants and refineries, I really recommend looking at the videos of the CSB[0]. They are doing postmortem analysis of all the majors incidents in the US. They are extremely well done.
You made me realise that Canada does not seem to have something similar - we just have a transportation safety bureau that handles trains, pipelines and the likes. And the Titan thing that happened a few weeks back.
I mean, there are incidents in refineries and plants. It's just hard to know how common they are because many go unreported or just don't reach wide public knowledge.
Seeing as we still see tiktoks of mud-covered men manually assembling or dissembling drilling equipment with no PPE, (intended to goad women about how the genders are unequal), I think additional government scrutiny can't hurt in this space.
Probably pushed by some recently rotated manager or new CEO dropped out of a recent acquisition...sometimes the code pushes the system over critically, sometimes it's just a code broken arrow event. Something to tell in the privacy of your next off the record interview or in your biography...
Is there a specific variety the author knows that's toxic? ... Because this (manathakkali or "fragrant tomato") features in South Indian cuisine and is rather yummy.
> Atropa belladonna, commonly known as belladonna or deadly nightshade, is a toxic perennial herbaceous plant in the nightshade family Solanaceae
> The foliage and berries are extremely toxic when ingested, containing tropane alkaloids. These toxins include atropine, scopolamine, and hyoscyamine, which cause delirium and hallucinations
When I was a kid my Dad spotted it growing in my school grounds...
To put this in perspective 22.54 per 100,000 would die due to motor vehicles in the same year that Slotin died. The average car (especially on US roads) is much more deadly and can become deadly in a blink, much faster than plutonium.
How many people died of not just plutonium, but all criticality events? Less than 22, total (most of those Soviet). There is little reason to believe that even if these were household materials that they would be more dangerous than cars in the USA.
> To put this in perspective 22.54 per 100,000 would die due to motor vehicles in the same year that Slotin died.
This is an incorrect perspective. If the same number of people as those driving cars had been playing with enriched uranium or plutonium, the casualties would have been much worse.
That's because there are more people who drive cars than people who handle plutonium. A more sensible statistics would be the number of death per hour of usage (which might still be in favor of handling plutonium).
You're comparing a small number of objects handled by expert technicians in highly controlled environments against a large number of objects driven by average people in much more chaotic environments.
By the same logic, many more kids have died of chocking on toys than of radiation poisoning by plutonium, so we really should give kids plutonium to play instead of toys.
Did you not realise how ridiculous your comparison was?
Whilst I agree with your conclusion that cars are involved in more loss of life, I don't think it's fair to only look at criticality events and ignore other radiation based deaths. Also, it's the drivers, not the cars that cause the danger.
> it's the drivers, not the cars that cause the danger
"People kill people, not guns".
You can kill me by throwing a person at me at 30mph; but it's easier to kill me by throwing a ton of iron at me at 30mph.
If you're right, then there are no driving accidents, and all injuries result from negligence or incompetence. In fact, the driving environment is so complex and chaotic that accidents are inevitable, even for the competent and attentive.
It's recommended to not call them "accidents" anymore as that implies a completely unpreventable incident which carries no blame. Road Traffic Incidents or Road Traffic Collisions are the preferred neutral terms.
The vast majority of RTCs are indeed due to the inattention of a driver or not looking sufficiently (e.g. not taking care to move your head so that the A-pillar does not obstruct your view of pedestrians/cyclists etc).
I would consider accidental RTCs to be mainly due to either mechanical failures (which could be argued is due to lack of maintenance) or previously unknown medical issues.
It's interesting the difference in use of language between RTCs and shooting incidents. News reports will often make no mention of the driver and will make statements such as "car hits a building", but to use the same style with shootings would be risible e.g. "multiple bullets hit man resisting arrest". Of course, the car industry has been exerting control over media reporting of RTCs for about a century now, so it's very ingrained.
Those aren't the same. A suspect is someone who has been charged, or who is being questioned under caution, because he might be charged. The example given was of a man arrested for possessing a knife - therefore a suspect - who was shot by police. Since a policeman shooting a civilian isn't usually charged or cautioned on the spot, it's not surprising that there is no suspect in the shooting case.
> It's recommended to not call them "accidents" anymore
Perhaps I've misread you; you seem to be asserting what I said you were - that there are no accidents.
Also, why the passive voice? Who makes this recommendation?
[Edit]
> The vast majority of RTCs are indeed due to the inattention of a driver
Arguably, all RTCs are the result of inadequate attention from someone or other. But attentive, well-trained drivers have collisions too. Perhaps their attention lapsed? They're human - human attention wanders, they're not focused 24/7 on the task, because they're not machines.
Cars are all very well on roads that carry light traffic. But not many roads are like that nowadays: most driving environments are way too complicated for even an expert to negotiate completely safely.
> Arguably, all RTCs are the result of inadequate attention from someone or other. But attentive, well-trained drivers have collisions too. Perhaps their attention lapsed? They're human - human attention wanders, they're not focused 24/7 on the task, because they're not machines.
I'd say most of them. It's possible to be driving along with a clear road ahead and have a person or object suddenly appear in front of you. I recall a case where a child was riding a bike on the pavement (sidewalk in the U.S.?) and for some reason lost control and fell into the road. The driver had been adhering to the speed limits and was paying attention but unfortunately was unable to stop in time and the child ended up dying.
There's lots of edge cases with driving which is why there's a bunch of recommendations regarding speed and leaving space so that when something unexpected happens, there's more chance for a careful driver to be able to stop or avoid the incident.
Pedantically (i.e. according to the Highway Code) the part reserved for pedestrians is the 'footway'; the part for carriages ('things with wheels', roughly) is the carriageway.
I suppose that between them they comprise the pavement.
I've taken to referring to them as carriageway and footway; it's unambiguous, and it even makes sense if you're shouting across the Great Pond.
Well, I'm not a British driver, but am a British cyclist. I'm very familiar with some parts of the Highway Code (not so much the motorway sections), but those terms just aren't commonly used, whereas pavement and road generally are.
Well, it depends on exactly how you define accident. In general, it would mean an unintended consequence, but arguably, driving whilst drunk and hitting something may be unintended, but not surprising - I wouldn't classify that as an accident as it was preventable and a known consequence of driving under the influence. If a careful driver suddenly has a heart attack and loses control, then I would classify that as an accident.
Numerous "self driving" cars have been in crashes. I believe under defined law it could be considered industrial accident when that occurs but is still open for litigation and definition.
A better comparison might be fatalities among NASCAR drivers during races - i.e. professionals operating in a dangerous environment using tools engineered for safety.
Google claims a total of 29 drivers have died during races, so the number of deaths in comparable, but I don't know how the total person-hours compare between NASCAR drivers and nuclear materials handling worldwide.
The difference is that the Newtonian physics of driving a car are mostly intuitive and that cars serve an obvious purpose making the risk both understandable and acceptable.
Handling highly radioactive materials serves no obvious immediate purpose and the dangers are very abstract until you get it wrong.
In the summer of 2011, Los Alamos National Laboratory presented a happy self-portrait to the public.
But that year something alarming happened at Los Alamos that the lab kept quiet. A pair of workers with cavalier attitudes nearly doomed a room full of colleagues by stuffing so much plutonium into a small space that they came close to triggering an accidental nuclear chain reaction, all to get some photos.
Scribing the rods probably releases plutonium particles, a vicious carcinogen, as would more professional looking laser engraving. A Sharpie would be safer.
https://en.wikipedia.org/wiki/Natural_nuclear_fission_reacto...
> The natural nuclear reactor formed when a uranium-rich mineral deposit became inundated with groundwater, which could act as a moderator for the neutrons produced by nuclear fission. A chain reaction took place, producing heat that caused the groundwater to boil away; without a moderator that could slow the neutrons, however, the reaction slowed or stopped. The reactor thus had a negative void coefficient of reactivity, something employed as a safety mechanism in human-made light water reactors.
> After cooling of the mineral deposit, the water returned, and the reaction restarted, completing a full cycle every 3 hours. The fission reaction cycles continued for hundreds of thousands of years and ended when the ever-decreasing fissile materials, coupled with the build-up of neutron poisons, no longer could sustain a chain reaction.
What a wild story!