Who Destroyed Three Mile Island? - Nickolas Means |

[@mmettler42]

I'm a tech lead at a medical device company and I first watched this a few years ago. This is still my favorite developer talk, years later.

[@InekoBK]

This is like "Seconds from Disaster" just without the dramatic visuals and voice over . Nice calm informative way of hearing what happend

[@marianmarkovic5881]

Seconds from Disaster, u watch it, u know what someone mom, some driver and some technician were doing when things happends, what u dont learn is what actualy happend,..

[@elamentri]

I was actually thinking the same thing, I'm a big fan of SFD, but would love to hear it delivered in this fashion

[@drstrangelove09]

Best discussion of TMI that I've ever seen!!! Hats off!!!

[@nitehawk86]

This was, in fact, not Too Much Information.

[@NoobNoobNews]

@@nitehawk86 Too much information of what you don't need drowning out what you do need.

[@deusexaethera]

The anecdote about the shift supervisor being a former Navy reactor technician, and thus being preoccupied with concerns specific to submarine nuclear reactors, highlights an extremely important point that unfortunately isn't mentioned in the lecture: *PEOPLE WHO MAKE IMPORTANT DECISIONS **_MUST_** EXPLAIN **_WHY_** THEY ARE MAKING THOSE DECISIONS.* Otherwise nobody can double-check their logic. Managers hate being second-guessed, but second-guessing nonetheless prevents logical errors from turning into figurative (or sometimes literal) meltdowns.

[@deusexaethera]

@John Buick: Source?

[@deusexaethera]

@John Buick: I was quoting the video. Who are you quoting?

[@deusexaethera]

@John Buick: I am aware that there was a hardware fault that caused the initial problem. But as the video explains, the shift manager made a decision to compensate for the initial problem that would've made sense for a submarine reactor but not for a civilian reactor. So despite the now-obvious chip on your shoulder about what I assume is your own prior Navy service, the final outcome of the accident WAS partially the fault of the shift operator's overconfidence in his prior experience.

[@deusexaethera]

@John Buick: Oh, well, if the official report doesn't mention it, then the information doesn't exist, right? books.google.com/books?id=tf0AfoynG-EC&pg=PA78&lpg=PA78&dq=Brian+mehler+three+mile+island&source=bl&ots=Orw_DbMNuH&sig=ACfU3U22mOmQwJQjZp8arAcS0cX2n7ZZdw&hl=en&sa=X&ved=2ahUKEwjN6vLSxr7pAhWVZDUKHZbYBd8Q6AEwAnoECAYQAQ#v=onepage&q=Brian%20mehler%20three%20mile%20island&f=false. The president at the time was also a Navy officer and thus had a potential conflict of interest in any situation involving other Navy officers.

[@deusexaethera]

@John Buick: No, if it wasn't in the official report then the people who wrote the official report omitted it for reasons that are known only to them -- maybe because it was hearsay, maybe because they were told to omit it, maybe because they are humans and simply thought it was insignificant without even consciously realizing they were making that decision.

[@l-cornelius-dol]

This guy's an incredible presenter.

[@charleswinn4289]

This is an excellent presentation and discussion and certainly applies to other industries like aviation and medicine. Carter as a former Naval officer understood the the ultimate accountability is with the person at the top. Solve the problem not lay blame. The investigation of Chernobyl sought to lay blame on the operators since those at the top could not be blamed. This laid the ground for future disasters. As we go forward with nuclear power we need to remember this presentation.

[@Kannot2023]

Both 3 Miles Island and Chernobyl accidents took place in early morning. It seems that URSS didn't learn anything from 3 Miles Island

[@lancetekk]

Yeah, The US were quite eager to share their nuclear power plant procedures with the east back then. How could they have missed it?

[@marianmarkovic5881]

True that in Chernobyl they(Soviet goverment) looked for scapegoats, but also made changes to RBMK reactors and procedures. Legasov also did a lot trying clear Akimovs and Toptunovs names.

[@infernosgaming8942]

From what I've picked up, from watching, listening and reading about nuclear incidents, you're in for a bumpy ride whenever the narrator/speaker pauses

[@CinemaDemocratica]

"You closed the block-valve, right?" Homer Simpson: "Uh ... yeah, I did that."

[@lloydsumpter7735]

Working in Process Control and Instrumentation since 1976, I can fully understand the problems. Instrumentation is KEY - for instance, position sensors MUST be used for any critical valve. And the Alarm System must inform the operators IMMEDIATELY of any new alarm - at BC Hydro, we had a system that flashed any unacknowledged alarm and set off a buzzer, even if there were 1000 alarms. Those two items would have prevented a disaster.

[@PAVANZYL]

Brilliant, brilliant, brilliant. When the was happening i was working at the Koeberg Nuclear PS in South Africa. I understand the functions of the various items mentioned. I have never seen an expiation of the events this clear. Thanks, I now have a pretty clear idea. One of the best ever technical videos I have ever seen.

[@fredrosse]

Having spent 40 years in the nuclear power industry, I fully agree with your statement here, far better, and far more clear than I have reviewed in other explainations.

[@lancetekk]

I would like to add: i am impressed how good the explanation of a quite complicated incident is so that non nuclear engineers (the audience: software lead developers) get it without hassle. Abstraction does the rest. The talk explains the issue and engages the audience, because they are always looking for the "what can i do to make sure such stuff does not happen" lesson. Beautiful.

[@billleach3396]

This is, IMHO, an excellent analysis of the TMI-2 accident. However, he did misscharacterize the Navy Nuclear training a bit. I am a former "Navy Nuke." I was trained that a solid pressurizer for an OPERATING reactor was indeed a disaster in the making. We were also trained that a solid pressurizer for a shutdown reactor with the steam stops closed was NOT a disaster. This is the same training that the TMI operators should have received. Nickolas reminds me of a PhD engineer from GE that was one of the design engineers for the S5G reactor. Usually about a week after an event at the prototype (where he was then working in Idaho) he would very informally "chat" with all of the crew personnel that had been involved in the incident. It was well known and understood that he would not reveal the sources for the information that he obtained. The result of this one man's way of dealing with issues led to corrections that actually fixed problems. The success of that man stuck with me throughout my career and served me well. I see now that he was a man "ahead of his time." While individual managers and supervisors are still a problem using the "who to blame" mode, the nuclear industry as a whole has been quite good about serious "root cause analysis" and what can be done to help operators from making errors. I don't want to try to write a book here but Brown's Ferry and TMI alone resulted in design and "Human Factors Engineering" changes numbering in the thousands.

[@johneeadbl5876]

Everything you say is your opinion. (From Common Law, which we inherited primarily from England and to a lesser degree from Europe, only God can know fact. Everyone else is simply, always, just stating their opinion as to what is fact that only God can know.) I'm not a former "Navy or other Nuke", but, I still understood his explanation well enough to think your opinion is not only wrong but potentially dangerous. I strongly suggest you go back, view it again, find where he discusses why the naval reactor training was/is inappropriate, and see if you're remembering your own training correctly and if you've "jumped to..." an inappropriate "... conclusion" You might have said, "that a solid pressurizer for a [properly] shutdown [naval] reactor with the steam stops closed was [less likely to (insert proper phrase, i.e. meltdown, vent radioactive steam, etc.)]." Again, some study online about the differences between military naval reactors and commercial reactors might be appropriate. Oh, and the word you were probably looking for is "misstated" and not the misspelled "misscharacterize", which spelled properly is "mischaracterize", which you used above.

[@billleach3396]

@@johneeadbl5876 Of course I'm expressing my opinion. That opinion however is based upon over 10 years of experience that took place during the time that those operators were Navy nuclear operators. First, the pressurizer performs two important functions. The first is to limit pressure excursions during normal and accident conditions (assuming a physically intact cooling system). It is also necessary to ensure that no over-pressurization event that occurs can threaten the physical pressure limits of the coolant system. When the reactor is operating, the pressurizer is required for this function as relief valves are not sufficient. Once the reactor is shutdown, relief valves are sufficient. Once the reactor is shutdown there is nothing that is capable of heating the coolant fast enough to exceed the relief valve capacity. The REAL root problem is management attitude toward "procedural compliance." Operators are afraid to do anything that is not called out in a procedure for fear of punitive action on the part of management. I don't know the solution to that problem other than to allow an occasional TMI event.

[@johneeadbl5876]

@@billleach3396 -You need to watch the video again, Bill, until you understand why your experience, like those operators at Three Mile Island, doesn't directly apply. Nickolas Means explained it very well. So well, in fact, that I understood it with one viewing. I'm thinking--very strongly--that you may be suffering from the thinking-problems he talked about. Those procedures--which you should know from your naval experience--are put in place so that people can override the disastrous effect panic can have on thinking. Anti-nuclear-plant activists correctly point out that nuclear energy production is too complex, not well enough known. Certainly, being able to take decisive action contrary to the "manual" may be valuable when some makes the correct decisions, i.e. the Russian who refused to initiate a nuclear counter-attack when their early warning radar system malfunctioned and it appeared we had initiated a nuclear first strike. But, from watching the video above, it seems to me--I simply don't have time at the moment to re-watch it to verify what I've been saying to you--that a failure to follow the manual is why the Three Mile Island nuclear accident occurred. All too often, the owners and operators of nuclear reactors take unjustifiable risks, i.e. Chernobyl and Windscale (it's a beautifully simple design, in theory. In practice, it released huge amounts of radioactive fallout, and, the operators followed the manual until it caught fire, then had to be proactive in creating procedures to put out that fire). Again, I think you need to re-watch the video until you understand why the accident was partially caused by the operators' bias from their naval experiences. Until then, please don't waste any more of my time.

[@billleach3396]

@@johneeadbl5876 OK we're done wasting each other's time. However, I'm afraid that there is one place where I do have to agree with you! The ultimate root cause of TMI was indeed failure to follow procedure and comply with design requirements. The accident would never have been initiated had not the plant been in an operating condition that was not permitted by design (and procedure). The best thing that came out of TMI was INPO. INPO instituted significant improvement in how utilities operated their plants. INPO and the NRC must not "let up" on demanding excellence from the highest levels of plant management... and not just "lip service."

[@johneeadbl5876]

@@billleach3396 OOP (: a single mistake. It takes several mistakes to require an "OOPS" :) Forgot to unsubscribe. You might take a look at the other comments above and below to see how others have begun to change their thinking patterns after watching this video. That's the point of the video, not an analysis of the accident, but, how analyzing the accident can help to change the defective thinking pattern(s) that caused it, Bill, i.e. your "single think" where you can't get past former naval reactor operators relying on that experience when it didn't apply at Three Mile Island. OK. I have rectified my single OOP and am now unsubscribed.

[@Inquisitor6321]

OMG! 40 years after the TMI incident, I finally have an understanding as to what happened there. We were concerned at the time as I lived in Philadelphia in which we were down-wind from the TMI plant. Most of this stuff was kept from us, thanks to complicit news media. Thanks Patrick!

[@LewisCowles]

That book he showed was probably available 10-20 years ago. Maybe blame the media some, but you have to seek knowledge, not expect it to wash over you like a storm.

[@mennol3885]

The job the media has given themselves is to entertain you with frightening stories and speculations, not to be a reliable source of useful information.

[@zenunderground]

Ive been watching this guy for 4 hours now.. He's a great speaker, chooses topics that are engaging and complex, but is able to dissect them if need be, and apply real-life birder-defying lessons to his presentation. The complete opposite of my college teachers in 2003!

[@langdons2848]

I rewatch this talk every few years. On top of teaching great lessons about team management it's an excellent demonstration of the inherent and unknowable vulnerabilities of complex systems and therefore the importance of the precautionary principal - which is applicable in so many areas of the modern world.

[@VuPhamMusic]

I am a new fan of Nickolas! One of the best presenters I have ever seen!

[@violabso]

This lecture finally provides all the information I have been curious to learn for decades, but could never find outside a scientific treatise.

[@zpetar]

There is something very important not mentioned in this video. this accident happened during night. Human mental capabilities are greatly diminished during night. After millions of years of evolution every cell in our body wants to take rest during night. I have experience of 15 years of working night shifts and trust me I know I know what I'm talking about.

[@OrigamiMarie]

Yeah, even if you consistently work nights, so that your rhythms are all based on nighttime wakefulness, it still doesn't really work out well. We have hormone systems that are based on light!

[@deusexaethera]

I'm a full-time software developer and it is _vastly_ easier for me to write code all night than to write code all day. I'm more awake and my mind is much clearer without having the outside world distracting me. As it happens, I have a MRI scan of my brain that shows the gland responsible for regulating my circadian rhythm is hollowed-out and probably non-functional. Apparently the pineal gland enlarges during puberty and shrinks again in adulthood, but for some people it turns into a water-filled bubble instead of shrinking properly. So my circadian rhythm is, in fact, _not_ properly governed by daylight as it would be for most people. Your rule, as with all rules, is true in general, but it's not true for every single person without exception.

[@fukpoeslaw3613]

@@deusexaethera int'resting

[@jek__]

Humans are adapting to being able to be productive regardless of the time of day, especially as global communication becomes ubiquitous and time of day become less relevant to socialization. Humankind will adapt to new opportunities and responsibilities as is always has. However, the fact that human mental capacities wane and wax is relevant; they should have just trusted the robots

[@alaskanalain]

I must be nocturnal then. Good point though.

[@jimbaranski4687]

I hope everybody involved at three mile island has seen this, especially the second half!

[@wilcalint]

I have, read my post.

[@nextlifeonearth]

@@wilcalint Care to repeat it? Can't find it.

[@verschissmuss3171]

@@nextlifeonearth - please find Bill Kenney's post below (thanks to "copy and paste"). Bill Kenney For the record. This guy was likely a teenager or younger in 1979. Three Mile Island (3MI). I wanted to point out that there is a Company in La Jolla CA called “General Atomics”. At the time of the accident GA had an employee count of about 2000 people. More then half of which had advanced Degrees. One of those employees was me. GA did not design, install, build or operate the controlling systems at 3MI. But, it did design, build and install 50 “PIG” monitors. Particulate, Iodine and Gaseous. The PIGs were scattered around the site and fed the gathered data back to a central computer(s) designed and built by GA ( and me ). At the time of the accident the GA system was operating correctly, being monitored locally, and by remote connection at GA in La Jolla. One of the Computer Engineers on the GA design team was yours truly. The causes of the 3MI accident have been gone over and over and over again for more then 40-years. Wild ass'd claims and conspiracy theories have been floated over and over again. All of them debunked. There was an accident, a bad one, it was fully contained, no one was injured, there was no widespread release of radiation. Inside the containment building was a nuclear hell. There was little to no detectable radiation or release there of at any time. Radiation at the site fence was at "background" levels all during the event. I saw the green bar paper print outs. Almost in real time. There were 28Kb modems in 1979.

[@remalm3670]

... Outstanding presentation, thanks ...

[@castirondude]

It's bizarre that the control room would be designed with a mishmash of warning lights, some critical, some trivial, all mixed together. People can only process a few pieces of information at a time so segregating and highlighting important information is key. I also don't understand why there wasn't a water level sensor in the reactor. You can't just look at a giant vat from the outside (or control room) and know how much water inside. You need sensors to tell you that. You need flow sensors to know what is flowing where, just having a green/red light about what a valve *should* be doing doesn't cut it, you need to measure what is *actually* happening.

[@KingOfKings34]

we are coming from an era where we have 5 cameras on our phones. these people lived in 1970 ies

[@lloydsumpter7735]

@@KingOfKings34 In 1976 at BC Hydro we had MUCH better position indication and alarm system. I do agree that flow sensors were uncommon back then (DP cells across an orifice), but water level, valve position and an alarm annunciator that accomodated NEW alarms were definitely easy to implement.

[@verschissmuss3171]

Well, I think that were simply some design failures.

[@neon-john]

There was no water level indicator because a PWR is always full. Too much information is as bad as not enough. Had that operator tapped that pressurizer level gauge, he would have known the reactor was dry. The water level system all PWRs were forced to install cost $millions and only confused things. If that indicator ever comes off of 100%, the operator has a heart attack. An electrical glitch or a nearby 2-way radio can do that. After the accident, the NRC published this inches-thick document called RegGuide 1.97. It contained thousands of mandatory modifications all plant operators were forced to make. Most were inane and a few years later, the NRC rescinded many of them. One of the more egregiously idiotic requirements was to design and install post accident radiation monitors so deaf that if the entire core appeared under the main vent duct monitor, it would stay on-scale. They hired my company to design, fabricate and calibrate these monitors. It was thrilling working with thousands of Curies of Kr-85 and Xe-133 but it was all so unnecessary. The indicators should always remain at zero but they didn't. A nearby walkie-talkie could cause an upscale reading. So could a routine radioactive gas leak that allowed the gas to diffuse through the seals, directly contact the detector and cause an upscale reading. Every time there was an indication, the operators had to initiate an emergency plan until the indication could be proved an anomaly. They caused so much trouble across all plants that a few years later, the NRC relented and allowed them to be removed. That held true for many of the RegGuide requirements. This knee-jerk document, mostly prepared by academics, cost $billions, only to be mostly scrapped. John

[@wilcalint]

For the record. This guy was likely a teenager or younger in 1979. Three Mile Island (3MI). I wanted to point out that there is a Company in La Jolla CA called “General Atomics”. At the time of the accident GA had an employee count of about 2000 people. More then half of which had advanced Degrees. One of those employees was me. GA did not design, install, build or operate the controlling systems at 3MI. But, it did design, build and install 50 “PIG” monitors. Particulate, Iodine and Gaseous. The PIGs were scattered around the site and fed the gathered data back to a central computer(s) designed and built by GA ( and me ). At the time of the accident the GA system was operating correctly, being monitored locally, and by remote connection at GA in La Jolla. One of the Computer Engineers on the GA design team was yours truly. The causes of the 3MI accident have been gone over and over and over again for more then 40-years. Wild ass'd claims and conspiracy theories have been floated over and over again. All of them debunked. There was an accident, a bad one, it was fully contained, no one was injured, there was no widespread release of radiation. Inside the containment building was a nuclear hell. There was little to no detectable radiation or release there of at any time. Radiation at the site fence was at "background" levels all during the event. I saw the green bar paper print outs. Almost in real time. There were 28Kb modems in 1979.

[@neon-john]

TMI had no GA equipment. Everything was Victoreen. Victoreen called them Continuous Air Monitors. TVA's Westinghouse reactors have GA systems. A good friend was the project manager on the changeout from analog channels to a digital system at Sequoyah and Watts Bar, identical plants. John

[@dave5194]

If you feel this video is too long, then please at least watch from 23:38 onward. The taake-away & lessons learned from this that separate this semi-disaster from the clusterfuck that was Chernobyl are so important.

[@TS6815]

honestly though, if you feel this video is too long, take whatever medicine you take for your attention span, make time, and sit down to watch the whole thing. it's all valuable stuff end-to-end

[@cat637d]

Stark and compelling presentation of truth and analysis!

[@himanshuanand9048]

I would like to know how he makes the presentation slides. They're awesome.

[@MrBanzoid]

Great analysis. Thank you.

[@KPearce57]

What I got from this was, If the polisher had a by-pass system to isolate any of them from the system the whole accident could have been avoided.

[@hene193]

K.D. Pearce There was so many more times this could have been avoided. People should know how to operate nuclear reactor.

[@hene193]

@@lederp42 Yes of course. But people who operate nuclear reactors need to be held to a higher standard.

[@tonypegler9080]

@@hene193 You learned nothing from the video then.

[@TrueHolarctic]

@@hene193 You need to develop that standard before the accident happens, otherwise you are saying that people should be perfect and never make mistakes. You can't change standards after the incident and expect actors to follow it

[@hene193]

MegaMedia5 Oh I never expect people to be perfect. That is a big key in many areas of design. You need to assume that people make mistakes. And desing systems so that when people make mistakes they can see it and hopefully correct them before anything too bad happens. But in hindsight things always seem so easy. Bad things happen and you learn from them. Like maybe the reactor engineers don't need to the elevator warning light.

[@johnyoungieyoung123]

Absolutely fantastic!

[@MikeJones-rk1un]

Wasn't President Carter a US navy nuclear engineer?

[@simonm1447]

He was one, so he was familiar to this technogy.

[@jamiefeitler6416]

President Carter had experience with Naval Submarine nuclear power plants. There are several key differences between those and the commercial generating plants. The submarine plant was built with little regard to cost. It was designed with robust, conservative engineering to sustain battle damage and continue operating. The commercial plant was designed to make money for the shareholders and many systems were combined or used fancier, less expensive elements that compromised the operators understanding of their characteristics. It’s the difference between a F-16 and a 737.

[@lancetekk]

@@jamiefeitler6416 it seems a bit far fetched that operators of a 737 would have less understanding of their platform than operators of a f-16.

[@jamiefeitler6416]

lancetekk you’ve missed the point. My point wasn’t about the relative operator understanding of their machines. It was about the intended function of those machines. One is intended to make commercial profits and the other is a war fighting machine.

[@lancetekk]

@@jamiefeitler6416 i agree

[@mininoturbusiness2286]

Did you know that the river where they take the cooling water has a total different fauna&flora on both sides of the plant? Just cause the water has a tiny difference in temperature when it comes back out of the pipe then before it had before they took it out of the river. Complete different fish, plants and just everything that lives in there...

[@synthclub]

Excellent talk. Just what I needed. Outcome bias, and hindsight bias. Wow.

[@tIhIngan]

Great video, great leadership lesson!

[@admiralcapn]

The one missing link in his explanation is WHY the water level in the pressurizer continued to rise despite the fact that they were hemorrhaging primary coolant. This is because the decay heat from the reactor was creating steam in the loop (which Nickolas explains is not supposed to happen in a pressurized water reactor) which increased the volume of the system faster than the water was being vented through the PORV. This highlights why using the cheaper instrumentation on the pressurizer alone to measure water level in the primary coolant loop was not sufficient to cover all failure possibilities. Considering how vitally important it is to keep the core covered with adequate coolant at all times, it was a huge design flaw not to directly report this information to the operators.

[@eddiej.g.3998]

You don’t need vessel level indication if you are operating within the pressure/temperature constraints to maintain a subcooled liquid. In reality, a level indicator for the vessel would be useless as it would cease to function accurately in a steam/ water mixture. The coolant parameters were outside of those required to assume subcooled liquid existed everywhere.

[@admiralcapn]

@@eddiej.g.3998 True. I am certainly not versed in what other instrumentation was available (total primary coolant loop pressure? a temp reading somewhere near the core?). If you had those you could infer that your system had entered a mixed state. And of course if you had one of those and KNEW you had a mixed state you could calculate the other. The flaw at TMI was operators making assumptions of reactor state that were wildly incorrect, and the system doing a very poor job of showing how those assumptions were wrong.

[@windsaw151]

@@eddiej.g.3998 So basically this was a really big design flaw of the reactor. Or was there a way for the operators to know if it was just water or a water/steam mixture they were measuring that they just missed?

[@Leonardo-ok3nz]

Hi everyone, I wrote some key takeaways about this talk, hope can help someone and receive feedback about the content :) - Hindsight bias → Basically is super-estimate the capacity to prevent/predict an accident when analysing it in retrospective(all information about the incident available). "I knew it all along". - Outcome bias → When you know the outcome you're more willing to judge every decision made, however who did the decision at the time it happened probably didn't know about the outcome. - First stories vs second stories - First stories is focused on humans involved at the accident and what they should done differently. It's focused in blame the humans involved at the accident. - Second stories focused first in discovery what the humans involved at the accident know about the situation at the time that the event happened. Human error is seen as systematic vulnerabitliy deep inside the organization, not someone fault. - People involved in the incident PoV is important - Understand why a decision made sense during the incident. People usually don't go to work with intention to do a bad job. - Assume positive intent from people involved at the incident. They did the best they could with the information they had at time. Baseline, human error is never the cause. Mostly cases blaming human error prevents to really understand what happened and where improvements can be made. - Seeks to understand what is responsible for an accident vs whose fault is it. - ⭐ In the end what is really important is to provide as much as possible **needed(quality over quantity)** information to people who is facing an incident to make informed decisions. - What is responsible over who. - Seek forward accountability, not backward. - Removing punishment frees people to share their incident stories and learnings. - This promotes knowledge sharing across an environment and this can prevent the same incident happens again - Good monitoring - Filtrate what is creating noise from what is important. On the three mile island example the alarmy that indicates that the SAMP is full(really important one) is right near the one that represent an elevator stuck(no important at all). - Should not be common people get used to a lot of alarmys being triggered and these represents nothing. People will have a lot of noisy and it will be confused about what is really happing, which increases the chance they miss something important. - UX/efficeny of monitoring systems are important for who is monitoring the production system. On the three mile island they don't have(at least in timely manner) a way to differentiate which alarmys are new from the already triggered ones. - The book "The field guide to understanding 'human error'" seems to be a good source of knowledge to improve outage "handling" culture .

[@stevee8884]

We need Thorium & Molten Salt!

[@thatclintguy]

This guy is an awesome story teller. I've watched his other two already!

[@Flash1857]

So it’s very interesting that the printer was slow, no visible valve indicators, alarm panels with randomly placed lights, got through all design meetings and all mock ups. Wonder what else got through? Mmmmm

[@totemictoad4691]

again second story im sure if you asked the people who designed the alarm panel they would have a perfectly rational and logical reason why the panel was wired up that way and the printer, could simply be that's the best tech they could use for the printer, remember when TMI was built tech was nowhere near as high bandwidth it could be as simple as the computer controlling the reactor could have enough bandwidth for all the things or you could have a fast printer,,,,,,, which do you choose,,,, (im just throwing out a guess)

[@jorgenfischer]

@@totemictoad4691 Besides: System Design and the understanding of user interfaces have greatly improved since the 1970s.. :)

[@iamdave84]

A basic schematic of the workings of the reactor with warning lights in the relevant location on that schematic is simple to design and quickly understand

[@darthkarl99]

@@iamdave84 & @TotemicToad A year late but from what i recall, 3 mile island was one of the big industrial accidents that made everyone aware of the importance of good UI design in safety critical equipment, it still catches people out in subtle ways from time to time. But it's something thats actually considered in the design phase of stuff now, it really wasn't once upon a time.

[@pohjoisenvanhus]

The talk is about a nuclear power plant but the analogies to running a high load distributed system with uptime guarantees can not be more obvious.

[@pohjoisenvanhus]

@kcotte59 Thanks. I'll look it up. In hindsight things tend to seem so obvious.

[@williamcorcoran8842]

The second story concept has a lot in common with crew resource management. Combining the two could be a major benefit:

[@Lozzie74]

The second story perspective was incredible. Each issue made me realise that they didn’t do something as stupid as it seemed. In fact, they did the right thing with what they knew.

[@jonathanwest6564]

The story of Chernobyl runs similar to 3 mile. They had many opportunities to have a different outcome. There were many people making bad decisions that made sense to them at the time. Plus the Soviet system was all about blame, and blindly following orders.

[@jorgenfischer]

The operators at 3MI thought they made good decisions at the time. Did you watch the 2nd half of the video to the end?

[@jhyland87]

I love these detailed post accident technical analysis review of accidents. There's one for Chernobyl as well (not from this guy) which is also very interesting.

[@alienworm1999]

When is this man gonna get a youtube channel

[@juliankandlhofer7553]

This is amazing!

[@rcsontag]

The excuse about Navy trained Nuclear Operator's concern about "going solid" is unfounded. Naval nuclear training has always emphasized core cooling as the critical activity. If that excuse is to have any weight, then Where is the blame on the plant specific training? Given that the design of TMI is drastically different than the Westinghouse PWR found in the Nuclear Navy. Plant specific Operator training or absence thereof had much to do with this accident, including the attitude that plant procedures were "optional".

[@fredrosse]

Some key events left out in this presentation, and serious ones. The radiation inside containment rose to about 80,000 R per hour, a horrifically large number. Never seen these numbers published, but that high a radiation level did exist at TMI. Then there is the Hydrogen explosion that spiked containment pressure, turned on the containment spray system, and could have easily ruptured containment if the initial conditions (containment pressure) was higher. The avoidance of a Hydrogen explosion was (and is) one of the "Holy Grail" criterion for nuclear plant containment designs, because the rupture of containment here could have potentially made habitation from Harrisburg to Philadelphia and NYC impossible for decades. We were lucky on that one, but the news reports conveniently leave this scary stuff out of the story. WTF

[@MegaLokopo]

I would love to see what he thinks of the new netflix documentary.

[@stevee8884]

Nickolas Means! Please do a presentation on Deep Water Horizon!

[@my.own.devices]

I am _really_ looking forward to your review of PIA8303 ... once all the facts are in, of course.

[@DaveStewartLondon]

He's so good isn't he. I wonder what it would be like to work with / for him?

[@Eric_Hutton.1980]

So basically if they'd have let the reactor do its thing it would have been ok? That is what I am getting. If I'm not understanding correctly I do beg pardon for my misunderstanding.

[@marcelb6214]

He is incredible!

[@redfern_mike]

The closed captions on this vid are terrible

[@jimmuo9286]

Software hell...! This is about leading all types of teams and the individuals inside each. People are the only thing that gets work done and done correctly. And within these team are individuals who must be treated on an individual basis - its where the knowledge resides. There is nothing more nonsymmetrical, inadequate or incommensurate than the equal treatment of unequal people. This holds true in life and career. Everyone is different and requires an individualistic approach. Yes, overriding top level rules apply at the macro level, but deeper down, once you pass the higher structures , policies and protocols, each person is unique and it’s these leaders that understand the concept are the ones that identify, acquire, build, train, motivate, mentor and help drive (help, not micromanagement) the individuals in the team so the entity of the team perform in such a way that exceptional achievements can be produced. Out the empowerment with the members of the team and not simply shout edicts downwards to be followed, rather go to the people that touch things on a daily basis for the answers to each of the programs problems. Kelly knew this decade ago and it was a big factor in the company’s success. Didn’t work for all teams on all programs but the basics apply everywhere. Taylorism, Fordism, Ohnoism, and Peterism had their moments in the sun (and at different times in history) and helped in small ways to move team dynamics and management styles forward, yet practice applications of Kelly (and others, not only him) remains the best functional deployment to this day. .

[@rcsontag]

The reactor scram should have been coincident with the turbine trip(design error #1). Power Operated Relief Valve indication shows demand, not actual valve position (design error #2). Rx vessel drain tank was not instrumented for level or temperature.(Design error #3). Annunciation system design is non-intuitive and conducive to human error(design error #4). Boiling occurred in the core because the pressurizer was not actually the highest point in the primary system, the steam generators were (design error #4). No mechanical safety valve was provided for the pressurizer (design error #5).

[@donaldrandall5628]

I wonder whether this sort of analysis has been applied to the 737MAX problems, and in particular to the FAA's part in the creation of these. Organizations analogous to the FAA almost always say things like "Solve the problem, not lay blame" but their usual practice (SOP) is in fact "Solve the problem by laying blame elsewhere".

[@Lucien86]

Hindsight - all you need is working precognition that is 100% reliable and in a situation you never encountered before. So when someone says why didn't you do this x-x-x? .. What they're really asking is how did your 100% reliable crystal ball fail?

[@clearingbaffles]

Raising pressurized level WILL raise primary pressure insignificantly 33’ will get you 1 atm or about 14.69# turn on more heaters if you want pressure to rise Thanx from the left coast near the Krapitol of California

[@robmorgan1214]

Missing assumption: Bad faith changes everything. Detecting and eliminating narcissists and premadonas from teams is a prerequisite to healthy management. With bad faith or egotistical actors the drama they create will destroy team cohesiveness and eliminate the possibility of enacting CRM. Step 0 is to identify and solve this problem step 1 is establishing CRM competency... at this point domain competence becomes relevant and decisions can be pushed out to the leaves of the tree. Lots of human factors being taken for granted: good leader, good followers, good culture, good team work, effective communication, informed decisions. The correct information must flow to the person making a decision...that person must have the knowledge capabilities and skills to both make the decision and act on it. This is not easy to do in a multidisciplinary team. Ignorance arrogance and inexperience combined with human nature and propensity for conflict and competition create a recipe for many disasters. Look at the worst civil aviation accident in Pakistan: cause of crash was literally an insecure captain who's personality made him unfit to fly. He was psychologically unstable. Human factors matter blame is not about creating a "safe space" to "listen" to everyone's input on the team. Blame is the team's immune system. Teams are extremely fragile one bad apple can irrevocably kill a healthy stable group dynamic. This is not uncommon. I've seen it happen both first and second hand in the form of bad leadership (out of touch, incompetent, victim of dunning Kruger effect etc.) and bad followers (disgruntled, resentful, ambitious, victims of dunning kruger effect etc.). Blame must be assigned early and often to maintain a healthy team. Ironically, the best team I ever worked on had a terrible boss who nevertheless allowed a good team to function underneath him because he genuinely cared about safety and effectively weeded out the people who couldn't hack it were insecure, cavalier, unsafe or otherwise bad for the team dynamic. People didn't respect him but they followed his instructions and effectively/efficiently operated well together despite having personality conflicts, widely divergent political beliefs and the other usual BS. But it was checked at the door. We always rose to the challenge we always got the job done. A couple of people who couldn't do that caused problems but were readily unmasked and he eliminated them. Blame was not something I feared it was awesome and highly useful. If someone is playing fast and loose with their own safety that's terrible but even worse is when they do it with yours. This is far too common. Blame is essential. People who fear blame simply shouldn't be working on teams. It's not about "accountability" or other meaningless managerial double speak nonsense. It's about self awareness integrity and responsibility. If you drop the ball and let the team down it puts everyone at risk. If no one has the balls to "hold you accountable" how do you improve? Social pressue is a huge force that is destructive in unhealthy teams...but in healthy teams it can deliver miracles. Hiding a problem or pretending it doesn't exist is never a good idea. Seeing how people react to criticism and the social stress it creates is a much better way to evaluate how they will handle themselves when TSHTF, as opposed to waiting for an actual high stakes situation. This is the only way to operate when teams include highly specialized professional members who literally can't be managed by the boss. The experts better be humble and self aware because everyone is depending on them to do their part and often they have zero common background with another expert and need to sort things out amongst themselves without disrupting the team, it's leadership, the priorities of the job or its constraints (sometimes it's your livelihood thats at stake sometimes it's your body or life). It's the foundation of trust: i can focus on my job because he's focused on his etc. I can concentrate on this dangerous task because he's monitoring the overall safety of the situation etc. I know we will probably finish on time because everything we need has been documented and organized ahead of time etc. I know if we have to work late no one will be distracted checking their watch or phone etc. I want to succeed so that the team will succeed. I want to be a part of something much larger than myself or any individual etc.

[@dw300]

Hi V-Sauce, Michael here!

[@lisaschuster9187]

Why are there two options for captions? English... or English. I’ve lived long enough to see, in many cases, the moment unnecessary complexity is injected into our lives. In whose mind did this seem necessary and why?

[@personzorz]

Hey, Vsauce! Nicolas here.

[@rabidbigdog]

Many of these power plants were controlled by PDP-11 systems. Not sure about TMI.

[@neon-john]

No power plant of that era had digital controls. Too unreliable, very difficult to make redundant and unlike analog instruments, software cannot be made provably safe. Even if digital controls were allowed, nothing off the shelf could do the job. The Westinghouse/Xerox Sigma 7 data logging computer at the Sequoyah and Watts Bar plants, a 60s design, remained in service up until recently because nothing off the shelf could replace it, though many vendors tried. It has 64k of core memory and a giant 10Mb hard drive, exposed behind a plastic door. By giant, I mean disks several feet in diameter. One head per track If one absolutely had to protect some data, he simply flipped the toggle switch on that head. Clock speed - 5MHz. No OS. One allocated his own sectors in his FORTRAN program. We had a large notebook where every program's sector map was recorded. No boot loader. At startup, someone had to fat-finger in a minimal bootstrap program using the front panel toggle switches, just enough to get the paper tape reader to run. Despite these mundane specs, it scanned 2500 contact closures and 500 analog inputs every second. And it drove 6 DECWriters, 4 in the control room and 2 in the computer room. It also did the core neutronics calculations and printed out a map of how the Rod Block Monitor was to be programmed. This is a matrix of bananna jacks and is programmed with jumpers. It controls the sequence in which the control rods are allowed to be pulled. The other programmer had a nice little side company which manufactured replacement PCBs for the computers (one per unit). Anyone who thinks that a PDP-11 can run a nuclear plant is smoking some mighty fine stuff. John

[@rabidbigdog]

@@neon-john When I worked at Digital we had contracts working with GE on PDP-11 equipment, so I'm unsure where your info comes from. Not myself personally I must say. But, correct, the PDPs are not in full authority control of the chain reaction but plenty of the systems.

[@neon-john]

@@rabidbigdog in many plants, a PDP-11ran the gamma spectroscopy in the radiochemlab. More often I saw HP 9845 "calculators" because the software was better and easier to modify. In one plant I saw a DEC running the water treatment plant. Peripheral applications like that. John

[@tomibach9712]

excellent....

[@fredrosse]

Good presentation, but at 26:36 the term "Going solid" in the pressurizer is mentioned. This is nonsense jargon in the engineering field. There are three states of H2O, solid, liquid, and gas. Solid does not occur unless temperature is less than or equal to 32F (0 C). There was no "going solid" here. The real statement is that the pressurizer came to a state of becoming fully liquid, thus losing the elastic (compressibility) property generally attributed to a gas, steam in this case.

[@jorgenfischer]

The presenter did not invent or propose that this was an appropriate statement or jargon to use for that event. He merely stated that this was the terms/jargon that the operator himself used to describe what was happening / what he wanted to avoid.

[@fredrosse]

@@jorgenfischer Understood. However, using clearly incorrect jargon, even if its use is commonplace, should have a comment to correct the words. Just a correction from an old engineer who has spent over 40 years in analysis of fluid transients.

[@zergreenone8111]

I don't think it's really meant to be taken literally. Although according to Wikipedia, some parts of industry refer to it as "going hard" rather than "going solid", which is probably a better phrase to use, but other parts of industry do refer to it as "going hard" by convention, despite the fact that it's not literally true that that is what is happening.

[@neon-john]

I have a psychology paper that applies here. Titled "People who don't know enough to know what they don't know". "Going Solid" may not fit your delicate engineering sensibilities but the term is universally used and everyone knows exactly what it means. If that bothers you then "tripping the reactor" must drive you insane. I mean, what did it trip over, some loose fuel rods laying around?

[@kj9219]

If you can't operate the machine, you have no business being near it. Or even building it in the first place. We got sold a bill of goods here in Washington state. None that were built operate today. And Most of all the containers to "contain" the waste are over in the Columbia river leaking.

[@thanksfernuthin]

The Legacy of Too Much Information.

[@TlalocTemporal]

And they missed important alarms because there was too much information on the alarm board!

[@sebastianhaas5863]

Moral of the story: Being unable to question and check your core beliefs about a situation, (and thus admiting to yourself, that those might, in fact not be true), leads to ugly outcomes.

[@neon-john]

Reactor Operators are required to follow written procedures which have been heavily vetted using "Verbatim Compliance". An operator who deviates from the procedure better have a good reason or the NRC will fine him and if bad enough, he will lose his license. I took 2 of the 3 years of training to become a reactor operator at Sequoyah. I decided that it was far more stress than I could endure so I moved over to the engineering track.

[@mambagr]

There is an error in the presentation. Human error is always the cause of bad decisions BUT this human error or series of errors were made during the reactor instrumentation design and initial maintenance.

[@robertsipes7391]

I wish the government could see this. It says a lot toward our wanting to blame China for COVID-19.

[@johnsauer8446]

WHYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY

[@benchapple1583]

This sounds a lot like NTSB philosophy.

[@Matticitt]

@@rlsweeneyjr1 yes, just like it was human error with the TMI incident. However they all list secondary causes and list ways to prevent accidents in the future by indroducing systemic changes.

[@stevecarter8810]

@@rlsweeneyjr1 Sure.. but then you can make a systemic change to detect the pilots / operators who are not up to snuff, and take them out of the system.

[@zzodr]

Still a better result than a certain RBMK 1000 reactor.

[@MikeJones-rk1un]

Luck?

[@Key_highway]

Mike M better redundancy in design

[@simonm1447]

@@Key_highway RBMK is one of the worst designs someone can develop, without any containment. It's Soviet, their safety standards were way lower than the one in the west.

[@Key_highway]

Simon Maier oh for sure I was saying the American ones had redundancy haha

[@tenminutetokyo2643]

The bigger question is why does the Hershey’s plant down the river take in water for chocolate production that 3MI dumps out of its reactors into the same river.

[@haveaniceday7950]

Are you serious? This is still going on today?

[@tenminutetokyo2643]

@@haveaniceday7950 Yes. The truth is all nuke plants leak to some degree all the time. 3MI sucks in Susquehana River water to cool its reactors then discharges the water back into the river. Then the Hershey's plant sucks the same water in to make chocolate and we eat them. Hence they all have trace amounts of Plutonium in them - the most deadly substance known to man. One single atom of Plutonium can kill a person.

[@haveaniceday7950]

@@tenminutetokyo2643 if one single atom kills then how could they have Trace amounts? You can't have less than one atom of plutonium?

[@jamesmonahan1819]

The age of books

[@allmhuran]

Works great with highly trained nuclear technicians or career programmers. But some people are, simply, not competent for the job. So, to answer the question "what is responsible" rather than "who is responsible": A "management policy" of hiring unqualified staff because they don't want to pay the amount required for someone with expertise - but really, that's just managers. It's a who, not a what. "A lack of passion in the work environement". Aka, lazy people who care about a shallow veneer of "seeming to be productive" rather than "actually being productive". This story would have been quite different if it had been about Chernobyl. You'have no chance of finding a "second story" to excuse some of the people there. To put it another way, everything we're talking about are systems created by humans, here. So if you're blaming "a system", you're just pushing the blame back to the person or people who designed the system. Some*one* or some*ones* are always ultimately responsible, The universe doesn't just spawn these things without human input. Is this kind of management false economy? Yeah, of course. But short term budget savings look good on a manger's scorecard, by the time it matters and things are falling apart they'll have already failed upwards. This is reality, denying it doesn't change it.

[@djkoenig4716]

Is this long explaination cause us to think: oh I see, I really can trust all these fallible human beings & Nuclear reactor problems will never happen again. WRONG! Regarding the men who cleaned up unit 2 for 14 years. How many of them died as a result of that very dangerous job? Do I expect to get the truth? Of course NOT!

[@Silverhand290]

If "there are two families of reactors in The U.S" and he is going to talk to people later, about the reactors in "this" country...Where was this filmed? EDIT Just spotted "Lead Developer London" right at the start of the video. School boy error. D minus, can do better.

[@KidCorporate]

Shakedown 1979, error codes never had the time On a pilot valve stuck up off the steam, you and I will meme

[@glendooer6211]

Anyone else get the wrong story when you click on a title??

[@392redienhcs]

A comedy of systemic errors. You can imagine they wouldn't allow reporters inside the control room or there'll be a bunch of "Oh, why all the blinking red lights on that lightboard?" "That's normal. Next question."

[@392redienhcs]

And also, it seems the best action they could have taken is inaction.

[@elchinator]

I disagree with the idea that there is no person responsible, everything "just happened" and we have to look for systematic reasons only. I agree, that the crew running things is rarely responsible! They don't get the information to do their job, they seldom have the liberty to make their own decisions and they are mostly underpayed. But there are people who first made the decision that things will happen this way! These "leaders" (who never lead, but just trail behind) are responsible! On Three-Mile-Island as well as in this Covid-19 crisis! And I think *those* are to be held responsible!

[@tomstech4390]

Information, Not dramatisation.

[@DKTAz00]

Who destroyed Three Mile Island? The intro in this video did

[@Vikingarrogance]

Nicholas means what? I hate that! Means what!!!!!!!

[@chair547]

I did

[@jona_archi]

almost american chernobyl

[@garybratton7250]

Blame Trump! You have gone a long way to understanding why out country is so caught up in the blame game.

[@neon-john]

I thought I smelled the stench of Rasmussen. This video is what you get when you try to sound authoritative based on what we at TMI called the "joke book". How about some input from a nuclear engineer who was there? I don't know what that little thing you showed is but it is NOT the condensate polishing system at TMI. The tanks were huge and had ladders going up the sides. Yes, there was a clog in one but that was only indirectly responsible for starting the event. What started the event was that an Auxiliary Unit Operator (AUO) accidentally cross-coupled the service water system with the instrument air system, forcing water into the air system. Easy to do because both systems used the same connector, a Chicago Pneumatic hermaphroditic quarter turn coupling. This water in the control air system shut down the pneumatic controls on the polisher. They could have bypassed the polisher except the water also shut down the feedwater heaters control systems. BTW, your estimate of a nuclear sub's reactor power is several orders of magnitude low. The operator turned off the safety injection pumps because the pressurizer went solid despite the PORV being open. All critical systems have at least one level of redundancy. The level was indicated by 2 vertical scale analog meters, one from safety system train A and one from B. Train B was tagged out for routine calibration. That left only train A indication and it indicated 100% water level in the pressurizer. The operator made a seemingly trivial but critical mistake. One always taps an analog gauge before reading it, in case the needle is stuck. He didn't. Another operator came over later to check the gauge and properly tapped it. The needle had been stuck. When tapped, it dropped to zero. The pressurizer and the reactor were dry and the fuel cladding had started to burn. About that time, the Chief Engineer of the plant, Ivan Porter walked into the control room to begin his shift. He later told me "you know it's going to be a bad day when the entire radiation monitoring system was in alarm and lit up like a Christmas Tree. BTW, that piece of sh*t geiger counter you showed is NOT a power plant grade instrument. The Victoreen system at TMI used ion chambers and read out directly in mR/Hr. If you want to see an actual radiation monitoring channel from TMI, just ask. When I had to leave 4 years later to run my company, they presented me with the channel for RM-L7, the channel that monitored the water discharge to the river, to show their appreciation for my work. I had designed and had fabricated a very low background and self-cleaning system that replaced the one they gave me. I still have it and it still works. Of course, they turned the safety pumps back on but it was too late. The pressurizer developed a satisfactory level indication fairly rapidly. Too rapidly. When the water hit the white hot core, it formed a steam bubble that blocked further water entry. When they figured out what was happening, the operator, as ordered by Ivan, cranked the injection pressure up beyond tech specs and that blew through the steam bubble. The rumble of the water deluging the hot core could be felt in the control room. They knew inside an hour how bad the damage was when the radio chemlab sampled the offgas system and found it saturated with Kr-85, Xe-133 and I-131. These fission products are normally contained within the zirconium fuel rod. When there is massive amounts of these coming out of the reactor and to the off-gas processing system, there has been massive fuel failure. Neutrons radiolytically decomposes water into hydrogen and oxygen in a stoichiometric ratio just perfect to burn or explode. So, whipped up by the media who sat across the river and monitored our walkie-talkie traffic (we quickly obtained encrypted radios), we had the Hydrogen Bubble Crisis. Actually 2. Someone postulated that there might be a hydrogen/oxygen bubble under the head of the reactor. The Chief Engineer from B&W, the reactor manufacturer and one of the most brilliant engineers I've met, quickly came up with a way to tell. The found a positive displacement pump that discharged 5 gallons a stroke. The modified it by blocking the intake and removing the outlet check valve, turning into a water pulser. A Valedyne (best there is) precision pressure transducer was plumbed into the primary system and its output drove a Brush (also the best there was) strip chart recorder. They brought the primary system absolutely solid. The pump pushed 5 gallons of water into the primary system and pulled it back out on the back stroke. A very pretty sine wave of the pressure variation appeared on the strip chart. If there had been a bubble, the pressure would have varied little. No bubble and the pressure oscillation was large. Charles took the chart home (he lived in the apartment next to me in the building I bought) and by about midnight had written a simulator program on his Apple ][ and calculated that there was little if any hydrogen in the dome. He called the results in and then we had a wonderful dinner together that my wife had cooked. There was a hydrogen burn in the containment. Not an explosion but from judging by the burn patterns on chairs, desks and paperwork in the containment, more like a swamp will-o-wisp. The containment pressure was monitored by a balance beam, inverted cup oil gauge that someone dug out of the 1800s. Its time constant was measured at 8 seconds. The containment is designed for 30 psi. The chart recorder pegged at 60. Charles wrote another program that calculated the probable containment pressure based on the instrument's slow response. The calculation was near 100 psi! The containment building is wrapped with a series of steel cable tendons that reinforce the concrete. Each tendon had a strain gauge. When they were later read, there was no indication that any had been stressed at all. This demonstrates just how over-designed nuclear plants are. THERE WERE NO CONTROL COMPUTERS AT TMI! All analog. The ancient Baily mainframe in the control room was strictly for data logging. It had 32 kilo-words (odd word size, 18 bits if I recall correctly) of core memory and a 5 megabyte rotating drum storage. The printers you pointed out were NOT dot-matrix printers. They were IBM Selectric typewriters, modified by IBM by adding solenoids to operate the keys. It printed at 120 characters a minute. There were 4 of them. When they got about 2 hours behind, the Baily core dumped and halted, losing that amount of data. The engineers I talked with said that this wasn't any great loss, as anything important had an analog strip recorder connected to it. I could pick your presentation apart further but I think I've written enough to show you don't know what you're talking about. Maybe you should talk to some actual nukes who were there and modify your presentation accordingly. John

[@markpasieka]

Why do we need to see the narrator.? Why not make the images full screen and please, minimize your mic sensitivity when your gulping water

[@tripackdroned4626]

Another camera operator without a clue...jeez, so distracting.

[@shawngreenforussenate6277]

Sabotage.

[@andrewallen9993]

What can we learn from this? If you want a safe nuclear reactor buy it from Britain, France or Canada. From Russia, Japan or America? Not so much.

[@andrewallen9993]

@@adiabolicalliberty2614 Yep, that was an atom bomb factory not a power plant.

[@TheeRocker]

,,, and about the bias concepts ? lol ,,, nah you know your good, lol

[@MikeJones-rk1un]

Oh great, nukes boil water and run a steam generator. Who knew.

[@MikeJones-rk1un]

@Colin Cleveland Why don't you put a pressure cooker on your stove and seal the vent? That will be efficient.

[@natalyrivera2042]

The responsible wound acly bomb because oak intermittently hum near a cautious screw. grubby gruesome, living ornament

[@fredafox5602]

The impossible twilight bailly owe because rose intrestingly suggest amongst a industrious birth. massive, depressed surgeon