Who Destroyed Three Mile Island? - Nickolas Means | The Lead Developer Austin 2018
On March 28, 1979, at exactly 4 o’clock in the morning, control rods slammed into the reactor core of Three Mile Island Unit #2, halting the nuclear reaction because of a fault in the reactor cooling system. At 4:02, the automated emergency cooling system activated as the reactor core temperature continued to rise. At 4:04, one of the plant operators made the befuddling decision to switch off the emergency cooling system, dooming the reactor to partial meltdown.
Why?
When something bad happens, it’s easy to just blame someone and move on. Taking the time to find the systemic causes, though, will not only help keep the problem from repeating, it will enable you to build the psychological safety necessary for your team to truly collaborate. Let’s let the story of Three Mile Island teach us how to make our teams stronger through systems thinking and just culture.
Came in for a nuclear dissaster story and came out with a life and management lesson. DAMN YOU NICKOLAAAAAAAAAS
exactly lol
who is nicholassssss
@@bryankirk3567 He said Nickolaaaaaaaaas. ;)
If you need videos to show you how to live you'll be DEAD if internet and tv goes dark.
Legit
Came to the video interested in nuclear accidents, left it with a new way of looking at mistakes. Awesome talk!
Yes and no ... Problem is, that those 4 guys who operated the reactor were basically competent, they just were stuck in their submarines and the reactor design left something to be desired (especially monitoring) as already said in the video. Now often enough you are dealing with incompetent or downright stupid people. Imo it's not as clean cut as the presenter wants to make you believe at the end of the video.
@@folterknecht1768 Yes, this design has heat exchanger redundancy but not coolant pump redundancy. WTF.
@@folterknecht1768 This is true, and I would agree that this method works best with competent employees. I need to read the book mentioned in the video but I can see ways of using the methods discussed in the video to make sure there was no underling issues that can cause issues in the future, especially if the incompetency comes from a company culture or training issues. After this fire the people who are stupid or are grossly incompetent and which these underlying issues would not explain the mistake made.
Folterknecht You are completely stuck in the blame paradigm which as everyone except you knows gets you exactly nowhere. Didn’t you learn anything from the presentation? Watch it again and you might actually discover a better model for understanding and not repeating complex technical/human systems failures than the blame game of calling it human error or in your comment, incompetence. Dalepsych
@@DaleTyler-rq3cr Not really. If you watch the talk the base assumption is that those people tried their very best to stop the accident. However this is not always the case. If you have people that are just 'working for the paycheck' your systematic problem lies in your hiring process and while you should fix this you also have to blame the workers that you either have to "fix" or replace
Another problem was that although the same valve had gotten stuck 18 months earlier at another reactor of the same design, resulting in the same problem (but discovered by the operators in time) Babcock & Wilcox had failed to notify other users of this reactor design of the flaw.
That's the issue with complex systems: you can't scream about every little problem until it becomes a trend.
@@UncleKennysPlace I would think that we might want to scale our level of caution with the size of the repercussions. Don't we determine whether a problem is big or little by the size of the potential consequences?
Secrecy caused this accident, just like it did with Chernobyl!
Sounds like Boeing 737max sticking angle of attack sensors. Not to mention the fact it only used one sensor not both. Plus the simulator felt nothing like a real trim level take over by the computer.
@@plateoshrimp9685 sounds like outcome bias. Kenny's point stands because every little problem could be a big problem in a reactor. We just don't know how things will play out. The speaker even made a joke about the trivial "elevator stuck" alarm. Well if those men had been stuck in the control room because the elevator was stuck, we would all be sitting here complaining about how obvious it is that the elevators are so important because elevators get people out of a dangerous area quickly.
As a human factors facilitator and Error Occurrence investigator within the aviation sector I found this to be the best explanation of both human and environmental factors contributing to an accident. Indeed we have come a long way with a no blame culture and a positive attitude towards open reporting. Excellent presentation thank you.
It's not a human factor, it's a greedy factor
Still pilots and other people working in aviation are often fired and sometimes charged and imprisoned if they make grave errors. And I think this is a good thing, as not all people care about other people, so some people only avoid errors if they have consequences for themselves. This is especially important for ATC: if a pilot causes a severe incident he might get killed himself, if an ATC operator causes it, he will be physically unharmed.
This has been the most amazing explanation of 3 Mile Island I have heard to date. Thank you for this!!!
I live and work in the shadow of this plant, and I can tell you, that as neighbors of Three Mile Island, we have never, in 40 years, been give such a clear and detailed accounting of what happened inside that plant.
@@macgto It is one of those thing where, if things can be hidden, things will be hidden, and we (the general public) will never really know what goes on with things like this in full. While this video is by far the best explanation of what happened I have seen to date, I still think we will never know the full truth.
I'm sure you are correct. My father is a retired engineer who worked at the Calvert Cliffs Nuclear Plant in Maryland, I forwarded this video to him to watch. I'm interested on hearing his take.
Car, you may not have noticed this, but normally it takes decades on most any high-level screw-up like this before the ~truth will come out,,, Usually most or all the people in question are dead,,
@@macgto howdy, neighbor! I rather miss seeing the plume from the remaining unit, but alas, its cost was double natural gas electrical production. The major part of the entire hot mess was human factors engineering. Seriously, whoinhell looks on the far side of their console for a critical indicator? I'll not go into ignoring a submarine reactor SCRAM decay heat being trivial, it's close to a stick of TNT going off, nobody in their right mind ignores that! Per scale, both were equally important, the sub having a lot simpler number of systems. And of course, more technical geared indicators, of every part of the operational components of the system, some Rickover SOB insisting on them, as well as precise engineering documentation and methods for nuclear submarines.
A LITTLE KNOWN TMI STORY: My father was one of the B&W Engineers on the first (and subsequent) conference phone calls. Unlike today these were a big deal, relatively expensive and it took some time to setup. On the first conference call an AT&T operator breaks into the call and asks, "Who's paying for this call?". Everyone at the table in three locations looks at each other dumbfounded and there is absolute silence. When there is no response the operator hangs up disconnecting the conference call. And in three places all there is to hear is the hum of the dialtone. My father was one of the B&W engineers who knew immediately what the problem was. . . albeit too late. What the current generations born after TMI need to know is that this was the END of nuclear power in the US. Old plants continue to run but many new plants under construction at the time (some actually complete) were abandoned and no new plants were approved since that time. Nuclear power has many problems (such as waste transportation and disposal), that have not been solved technically OR politically. The technical is always possible, but the political?
One such completed plant is the one I live near. Shearon Harris Nuclear Power Plant, in NC. I believe it was the last new reactor to go online in the US, in May of 1987. But getting a hold of new parts must be getting difficult with the ban, because they've had to refurbish and install the turbine from TMI Unit 2.
"Being built" is not "operating". We've (partially) built many new plants over the last 30 years. As far as I know, none of them have ever been licensed and turned on. (Cherokee, SC was never finished - sold to E.O..Studios and he filmed the Abyss in the flooded reactor building. They started building a new site near there several years ago; I doubt that one will ever produce power.)
Can't say to the rest of Your story but the DEAL ABOUT the AT&T operator is on par with some stuff that happened at my place of employment years ago. One of the VICE PRESIDENTS had a call that involved a Spanish speaking caller. A stupid operator cancelled the call because She was NOT BI-LINGUAL . The officer at our company called a supervisor to rip them a new one and who ever He spoke with thought that he was talking about bi-sexual instead of BI-LINGUAL and got all offended. There was another deal where The company lost BIG MONEY because Something went wrong in the AT&T phone system.One of my Supervisors was ready to bite nails in two!
Watts Barr unit 1 came online in 1996. Admittedly the license was approved in 1973. Unit 2 went critical for the first time in 2016.
What about a new nuclear plant in Georgia, which is supposed to become operational in a near future? Or is it?
Superb account of both technical and human factors in this incident, but Nickolas's greatest achievement in this presentation was to make the "lessons to learn" process and the attribution of blame issue applicable to a vastly greater range of technical and human endeavours. Thank you.
Mark Cowell. Well said. I heartily concur with all you have written. I, too, really enjoyed the clear development of both technical and human factors. The most important ‘take-away’ for me was the importance of creating a blame-free environment in trying to understand complex interactive systems failures of all sorts. For without a clear and accurate time-line, we can never learn from the past or devise corrective measures for future improvements. Mistakes are for learning, apportioning of blame gets us nowhere. I remember stressing the benefits of making mistakes in the educational process to my children, recasting mistakes as good rather than shameful as long as one learnt something from each one and didn’t keep repeating the same error. I will be trying to locate more of his lectures as I found this one so enjoyable. A most articulate and informative speaker. Dalepsych
Reminds me of the Challenger disaster. The O-rings were never supposed to leak as failure could result in the loss of the spacecraft, but since the spacecraft managed to survive (somehow) multiple launches, leaks were ignored, until a particularly cold day arrived. A culture of deviance set in. Likewise, at TMI the valve wasn't supposed to leak but was ignored because things continued to run even though the water temp exceeded 200 degrees. If the valve problem had been addressed at the start, the reactor wouldn't have been subsequently lost.
"multiple lunches" so it indigestion?
@@garyhoffmann1615 Over weight
As far as the challenger goes, if they would have noticed the fire coming out of the side of the booster rocket, the boosters could have been jettisoned early to save the challenger.
@@fredgarvin9493 Assuming Challenger would have survived the jettison of the SRB's operating at full power (questionable), it would still have to gain enough altitude with its main engines to reach the Azores for a safe landing, else it would have made one giant splashdown in the Atlantic.
Fineman has discovered the cause of a spaceshuttle which id is the deformation of a o riing. he is a excellent genius even if doesn't work at the Nasa. 🌻
Really glad to have stumbled across this excellent talk ✌️
Comrade Legasov sent me here.
Yes
Did you get your 3.6 Roentgen today? Not too good....not too bad
YOU DID DENT BECAUSE HES NOT THERE!
you're here all because of a piece of rock?
How does an RBMK reactor explode? LIES. How does a Babcock and Wilcox reactor explode? SLOW PRINTER.
As an electrical engineer with 45 years of experience I throughly enjoyed the story and lesson he brought to the accident investigation. It was very well thought out as was the real cause of the accident.
he's an idiot. I am a Nuclear Power Plant Operator and Engineer. These Condensate Polishers are on the secondary side of the Reactor Plant...and not on the Primary Side within the Reactor Plant Shielded Compartment (room)...where the Reactor Core, Primary Pumps, Pressurization Systems, and the Electro Mechanical Control systems that either raise or lower the Hafnium Controls are. If it is on the secondary side; then it cannot cause a "core meltdown". He even says that "one of the engineer looked into the viewing port of No. 7 polisher". You cannot look into something that is inside the Primary Reactor System...unless the Primary is completely shut down for many hours or days. They will not run the risk. There are 7 Polishers and they operate independently...not in series. This means that you can shut down one...and still keep the plant operational. This is called "double and triple back-up redundancy" ... and it exist in EVERY POWER PLANT. These Polishers remove any possible contaminates on the "Steam Side" of the Secondary Steam Generator Systems. They are not actually "necessary" to operate a plant. They are a overly redundant "cleaning system". You can read about them here....en.wikipedia.org/wiki/Condensate_polisher A condensate polisher is a device used to filter water condensed from steam as part of the steam cycle, for example in a conventional or nuclear power plant (powdered resin or deep bed system). It is frequently filled with polymer resins which are used to remove or exchange ions such that the purity of the condensate is maintained at or near that of distilled water. There is ZERO STEAM IN THE PRIMARY SYSTEM. This does not matter if it is a conventional or nuclear powered system. IN other words...there is no "condensate" on the primary side; because this is where the heat is generated and the water inside all components and piping...MUST REMAIN WATER ALWAYS. This is true in a conventional plant. A Steam Generator has U-tubes on the inside that carry hot water "internally"...which heat the pipes to ~600 degrees. The outside of the pipes are surrounding by water at ambient temperatures...usually 80 degrees; depending on the water source. The water "flashes" to steam due to temperature differential and "rises" to the top of the steam generator; where the piping directs this steam to a turbine; which spins an AC/DC motor generator that "creates Electricity". The AC motor generator is connected to the Transmission lines that distribute power to "where you want it". Condensate is the very last thing to happen to the steam....has it cools down stream of the Turbines that the steam moves and therefore...CANNOT AFFECT THE PRIMARY REACTOR CORE...in anyway; whatsoever. If the "steam generator relief valve" were to lift...the Reactor core would have already scrammed...and the control rods would have gone into place. Therefore, since the entire system's "Fission Heat Reaction" would have been stopped in 2 seconds; then there is NO EXCESSIVE HEAT. As a matter of fact...the Polishing system can be completely bypassed and you can dump regular city water into the Steam Generator to "cool down the system". There would have been no reason to "lift the steam generator Steam Relief Valves". He says that "the system stopped pumping water into the Secondary Steam Generator side of the Secondary system. Well; if this is true....NO WATER MEANS NO STEAM....which means that the Steam Relief Valves would have released NO STEAM into the atmosphere. And, unless you have an "over pressure situation" in the Steam Generator; you can continue to spin the turbines with the residual steam...as the Reactor Core continues to Cool Off...due to the SCRAM EVENT; where ALL FISSION and ALL HEAT GENERATION STOPS. So...this entire "official" narrative is a bunch of propagandized bullshit...and the NOVICE on the video; whose dad gave him a book on "how things work"....does not know what he's talking about or how a Power Plant...nuclear or otherwise....actually work. He has ZERO business talking about things; when does not know...HOW IT WORKS in the Real World of Power Generation and/or Power Distribution.
The last 10 min was a massive eye opener for me, thank you
No, he’s wrong. 1) killing someone in a car crash because you made a misjudgment is fine and doesn’t require punishment 2) but killing someone in a car crash because you were texting is not fine and does require punishment People need to 1) tell the truth regardless 2) be responsible 3) accept responsibility 4) have fear of punishment as a motivation to do your best. We generally don’t punish people for making honest mistakes. If there is a punishment, there’s usually is reward of less punishment for being upfront and honest and that’s because they accept responsibility. Tell the truth always regardless of the outcome, and do not take a knee to amorality or immorality to prevent lying. If a person lies about the 3 mile disaster to save himself, then he’s not taking responsibility of himself and the lives of others. Tell the truth always.
@@thomaspayne6866 imagine confusing systemic design problems with negligence. The talk conflates neither, you are spreading misinformation - either intentionally as a bad actor or to be an edgy contrarian. Either way your comprehension is terrible if that is what you take away from this.
@@thomaspayne6866 it turns out you simply can not demand from people to tell the truth despite they know they'll be punished based on what they're saying. It just does not work that way. You can ask, you can order, but the most of them just won't comply
jSkrat Nyarlathotep -- No, not for amoral and chaotic society they won’t. There’s a reason why religion exists. It instills morality, logic, truth, rules, self-restrictions from birth. When we lost our religion, or “when god has died” in our society, so does morality. And that is how we get a society full of upside down people. People who live in a world of chaos and inversion. Also known as “clown world”, or hell. Matriarchal societies live in chaos, and the men of today are becoming rapidly feminine, chaotic, amorality, demoralized, degenerate. THEY ARE LIVING IN UPSIDE DOWN WORLD.
@@thomaspayne6866 we've never lived in right world, no matter how religious the society were. There always were a sins, there always was the need of mercy from god. I think it means that this is just not the way. It's just not working. It is better to accept the sinners and deal with them, than trying to convert them. All our history tells us that you can not create and maintain for long a so to say "holy" society. And I really doubt that religious "eye for the eye, blood for the blood" is better and more effective, than modern judging system. (I'm not saying ours is perfect)
Thanks for clearing that up. I've been wondering what caused it since it happened. I was a welding inspector on the Waterford 3 plant at the time and only heard rumors of incompetence which I now feel were untrue. My brother made a bundle helping to clean it up. They needed men who were knowledgeable and hadn't been exposed too much and he was a superintendent at the Fulton power plant who seldom was exposed to radiation. The men had only a very short time to work in the environment before they needed to get out. Less than ten years later my brother died of leukemia.
My condolences and thanks to you and your brother.
I am currently in charge of RCA and failure analysis in out QA department. This video is a real eye opener in dealing with people involved in a defect, and I hope others will take these lessons to heart as well.
5-why
What happened to RCA? My family had a business in York Pa. We did very well selling their TVs. They made some great products . I'm curious because they were big in our area . We bought them from D&H distributors in Camp Hill PA. They sold to 100 dealers and we sold more top end than anyone. I miss their flat screen 27 and 31 inch.
I keep coming back to this video over and over to recommend it to people. SUCH a great talk.
Wonderful presentation. So much to reflect on about system design, monitoring and control systems, the importance of checklists and protocols.
Wow, wow. Thumbs up. I have never seen any nuclear "experts" explain the whole saga so clearly, let alone by an outsider.
Seriously. I've been on a Nuclear Reactor Accident binge, and just watch an hour-long doc about TMI. The first 20 mins of this video already had way more information than that whole documentary.
Sometimes you want someone who's an expert in storytelling, rather than an expert in the field which the story is about.
He got the event 95% right. He skipped items that wouldn't add to your understanding but an engineer would have spent time going over. Trust me, I worked the industry for 20 years. They don't skip anything.
@@nawdawg4300 I came here from HBO🤔 I've been on a 5 day binge 🤪🤪.
@@minnesota7010 KZhead: Hey kid, I've got some more of that _Nuclear Reactor_
My husband a former seasoned plant operator says " fear causes mistakes that otherwise be avoided " " as long as critical decision making is made under dire circumstances you can be sure mistakes will assuredly be made "
Arguably one of the most important videos I've seen in a long time. Studying engineering failures is like that. And then there's the important character lessons for organizations...
Wow, what an incredible video! Very educational from a nuclear standpoint (what exactly happened) but was rewarded with a management and engineering and life lesson. Thank you for this!
Very good. As a life long control systems engineer I can confirm this presentation is accurate and is in complete alignment with my own experience. TMI had a big impact not just on the nuclear industry, but industrial systems everywhere. President Carter's Commission did a fantastic job and much credit is owed to them.
i think accident like these are important ... We are lucky we had so few .. and only one has been a real big catastrophe .... I rather have an accident like that then chernobal happening again ... I would almost say accident like that are the best thing that can happen ... we can learn from them and make sure that with nuclear power the world is not just a greener space switching out dirty coal power ... but make sure that nuclear is as save as possible by learning from mistakes ... We learn from mistakes ... and when a mistake happens without major consequences ... then this is a good thing ... Sure in the end it was en expensive mistake .... but a good one. Mistake is maybe the wrong word ... but i cant find a better word ... maybe accident ... but i kinda feel accident is kinda wrong too to use ... Anyway .. with only 3 major accidents ... where this one here was not really that bad in the end .... nuclear power seems to be still a save power source .. and we can learn from japans latest accident too and make nuclear even safer
@@noblackthunder The problem is that now we are on the cusp of having perfectly safe reactors the damage is already done and getting new plants commissioned is a nightmare. Not to mention the folks who overestimate the danger and magnitude of nuclear waste by peddling the myth that it's everywhere and can never be processed. The west should educate people to the truth about nuclear like they do in Korea and Japan but big oil lobby against it. Fukushima happened in Japan and the Japanese are still pro-nuclear, they are a very clever people with a low level of ignorance. I hope North America and Western Europe become like that some day.
@@krashd yea fake news about nuclear power is just sad... on one side we want green energi, on the other side people dont want nuclear... meaning we have no power source that can meet the demands... its inpossible without nuclear. We have learned so much from the 3 dissasters...
No Fred. We will eventually go to even bigger monoliths of human ingenuity. Fusion is actually really likely within the thirty years often joked about. There have been some remarkable breakthroughs over the last five years. There is now a massive international collaboration to build the biggest yet Tokamak type device called ITER, that may be the first actual step in producing more energy than the process consumes. Though this still will be a test device for determining how better to proceed from here. Suggesting Batteries is the answer simply illustrates how clueless you are regarding the actual realities and engineering problems. You seem to be mixing up scales somewhat. To imagine batteries and solar, wind, and tidal is the answer is preposterous. I think you need to actually go and check out how much energy we are talking about JUST in the USA. Let alone the planet. With all the cells needed for the future electric vehicle revolution that is definitely coming, we are going to have to manufacture some tidy number of Lithium Ion cells. Or someone come up with another entirely new storage medium. We can't have high level reservoirs everywhere now can we?
@@krashd But Japan is also a very conformist and high trust society which leads them to be overly trusting or deferring to authority.
This has to be one of the best talks i've ever watched! Kudos!
I take it you haven't come across Yuri Bezmenov's video presentations on YT?
Water hammer in the primary is way scarier on a sub than just "loss of propulsion and disabled [boat]." You're talking about the potential for a steam explosion in a very confined space, possibly under the surface of the ocean.
It wouldn't be a steam explosion but could cause a leak, but that's what the reactor compartment is for 😁 But yeah "loss of propulsion" while underwater is kind of a big deal 😓 There's a chance to never come back up... Also, as a Navy "nuke" I can verify about the obsession with never "going solid". Navy reactors have lots of transient power operation, which is another difference he doesn't mention. Levels all over the sub plant could be changing by the minute or hour whereas a land plant is more steady.
Kursk. Just saying Exploding submarine
@@redgemon Pretty much, a steam leak maybe . Our Navy nukes can go from a cold start to underway in 30 minutes or less. Civilian plants... 9 months. SSBN 626 Gold and 644 Gold. US Navy has never had a nuclear accident. Multiple redundancies and the worst best training on the planet! Did these civilians have have an ORSE board?
@@timavery99 😮wow I was on the 626 in prototype 😅
@@redgemon Or a carrier, where do you think catapult steam comes from anyway...
That was such an exceptional presentation on so many levels, I think I'll be revisiting this over time. Thanks!
Extraordinarily-well done talk, sir!
Not only did I find an excellent summary on the TMI accident, I also found a new perspective for analyzing problems in the future. Thanks for the video.
Been working at the gas station all your life? Glad you are finding out something useful,, The point he failed to bring to the forefront is, AUTOMATION,, IFF (if and only if) the software is written correctly, you can remove human-faults that screwed things up here,, -- It's really no different than automated cars driving up around,, Automation removes the human-errors that were injected into the process as shown here
Excellent conference! Thank you! As an engineer, I appreciate a little learning like this.
That was an excellent presentation! Not only a detailed, second by second account of what happened, but how the "blame" should be handled and what should be accounted for, and not who!
Great presentation, really explained the reactor, systems, events, and people that were involved.
Amazing talk! I work in DevOps and I have already been applying some of this during outages. But there is a lot I haven’t thought of before. I am passing this around to my team because I hope it improves our incident response.
This was one of my favorite talks from this conference. So interesting!
Nickolas is one of the best technical speakers I've ever encountered. It's a joy to listen to him give a talk.
I was one of the egghead engineers who designed the pressure vessel for the reactor. We did radiographs for every weld and each radiograph was inspected by 2 technicians at the same time
Congratulations on a perfect job considering the massive stress test.
Nice! History is preserved!
Hi There, It seems that the pressure vessel was fine so Your team did their job and the pressure vessel was NOT THE PROBLEM . The tech error happened elsewhere in the loops. Not sure if anybody ever asked the critical question about WHAT HAPPENS if the WATER CLEAN-UP SYSTEM GOES SOUTH?
Well, Since you are being so defensive, let us all tell you that this mess was NOT your fault. It was a series of bad choices made by guys who had bad information and training in the worng time and kind of reactors. The pressure vessel held as far as anyone knows. You can stop being defensive now. Relax, lean back, take a few deep breaths..... Ahhhh....... doesn't that feel much better after so many years in denial. You're cool, dude. Peace, Baby!!!.
@@organbuilder272 Civilian plant owners fell into the habit of trusting the excellent training military personnel received...the problem was when the water polishers shut down and the instrumentation started giving out misleading information, the operators fell back on their military training, because thats all they had. This led to a fuel element failure. It would be akin to a police department hiring Navy Seals and saying "You guys are already highly trained , here's your gun and badge" and then being shocked when their officers engage in "reconnaissance by fire" when raiding a crack house. ("Reconnaissance by Fire" is when you shoot through doors and walls to see if anyone screams or shoots back.)
Absolutely incredible talk, Nikolas. I came to the video wondering what about TMI could be learned and used in software development (and indeed, everything else, too), and left inspired. I never really knew much about TMI, and I learned tons about that. Never really thought about "first stories" and "second stories", but that's eye-opening, too. I've heard/thought of the concept of not focusing on punishment, but rather focusing on information gathering and processing, but hadn't really ever heard of it used in a non-textbook way. Seriously good stuff here!
Thank you Nickolas. I worked at a place where chronic communication issues plagued the office and nobody ever took the time to address the underlying problems behind them. This is exactly the talk I needed to hear. It isn't about not talking responsibility and growing, it is understanding that you can try your best and still mess up. We are human and it happens.
A superb analysis and presentation. Thanks.
This presentation is fantastic! 👍
In telecoms, we had "compelled signalling". One end sends a signal and waits for the distant end to respond. It does not assume that the other end has acted appropriately until one of a range of acceptable signals is sent back and this is in a non safety critical situation. In the case of the valve, "I told you to close" is simply not good enough. Only " I have definitely closed and here is my signal confirming that this has happened" is good enough in a safety critical environment )and it's good practice anyway).
Excellent explanation of what happened. Thank you for your analysis.
Thank you for the excellent video. Applause 🎉. I was a senior manager of control departments in large coal fired power plants for a 35 year career. That meant every blip in the power plant usually came my team's way for solution. I have seen operators admit to fault that I absolutely knew was not their fault but upper plant management being compliant with having the root cause and a solution nearly immediately available for Corporate who also often have regulator reporting requirements in mind. Some unit trips or equipment trips have causes so subtle or unexpected that the right question is readily ovelooked. Sneak circuits can lay dormant for years until an operator may need to adjust something to exactly the same setting as always but in combination with any of a thousand other settings uniquely in place at that time, drives the system unstable. Luckly those are rare occurances and redundantly designed tripping systems take over. Finding the needle in the haystack was both frustrating and rewarding and absolutely necessary. I would go on looking for the real cause long after solutions had been implemented. Nuclear power plants are highly regulated as you know. The control room layout is essentially frozen at the time of commissioning. THUS the alarm management panel layout can be a relic of regulation, leaving the control operators more vulnerable through the life of the plant.
thank you for this excellent report.
Hell of a story, pretty incredible
That was so good I have to watch it again, thank you!
This is one the best talks on KZhead ever. I just wish I could give a KZhead video two thumbs up! Great presentation!
This is very timely considering the current Boeing 737 max saga, and can be applied with pilot error being story 1, and corporate cost cutting decisions being story 2.
That's actually just shifting blame from one group of people (pilots) to second group of people (corporate executives).
I feel it'd be more appropriate to put Boeing's rushing of production, and what seems to be neglect within the process as the first story. However, the second being Boeing's own backstory of how these planes were designed, the pressures from airlines such as AA, and competitor Airbus and the latent threats that had evolved into the current scenario. It is a really interesting demonstration of this factor though! I'm in pilot training at the moment and believe this kind of mindset could be extremely critical within my future!
@@gtpk3527 Yes. Though I'm impressed by the presentation, I take issue with "it's never human error." It's *always* human error. The crucial task is to identify all the human decisions and actions that led to the failure, including design and construction, and correct them going forward.
@@joesterling4299 What he means though is specific meaning in systems operation in relation to causes. Instead of saying this is Bob's fault or Dave didn't do this job well, we need to look at why the system in place allowed Bob to commit that mistake or why Dave was in that position even though he as was not at acceptable skill level. That's the systemic vs human error. Obviously there's secondary question of negligence in criminal sense (if someone is drunk for example) but that still doesn't change anything on investigating the event. Those are two separate universes, so to say.
The FAA being both the promoter and regulator of air travel is story 3?
I found this to be an informative talk. I'm not getting into the controversy around nuclear power plants ... save to mention I am intrigued by the thorium cycle/molten salt technologies being developed recently. I just like the clear summary of the chain of events. And your discussion of the second story. I have a new entry on my reading list thanks to this presentation. Good job.
What a great break down on the 3 mile accident, and then an even more important lesson in looking for the second story. Really thrilled I caught this!
Absolutely superb video - so much food for thought here.
I watched your talk on the UAL DC-10 and now this, you're amazing and the stuff you talk about is really interesting!
This is an excellent description of what happened at TMI-2. Thanks for sharing. I would add only a small comment, pertaining to why steam is condensed before being pumped back to the primary loop instead of just piping it there as steam, or more likely pumping it there as steam. Typically gases do not "pump" very well, thus it takes much larger equipment to deal with the huge volumes of gases involved - certainly much, much more than it would with liquid water. When the steam is condensed back to water, the volume of materials becomes a small percentage of what would have been required as steam. So, economics and several other considerations require conversion back to something that is manageable to handle and transport. Again, thanks.
The last 5 minutes of this video is incredible. The summation is true life wisdom. Very well done!
Wow. This was exceptionally well-done! I watched the entire thing!
next time my manager give me crap, I'm going to be like "Hey, you are outcome biased"
"Hey, you are fired"
This doesn't even make sense regarding what outcome bias means...
You are assuming he/she is the problem… you are stuck in 1st story.
@@GabrielPettier for it to even make sense that he's outcome biased, he would have to know that giving him shit would improve his production, or he wouldn't waste time doing so and just fire him. This person clearly doesn't understand outcome bias as a concept.
Yep, maybe you can be promoted to working inside and off the gas pumps outside
Very good presentation. Its also interesting to compare this with the decisions that were made at Chernobyl.
What an excellent presentation! The point you've made here is so important that I refer back to it whenever the blame game comes up. I'm about to send it someone right now because the topic is fascinating as well. Thank you!
When I took a Health & Safety course at work about 8 years ago, this same point was made over and over: look deeper for the true causes of an accident, which are most likely in the way a system works, or more importantly, the way a system fails. Thank you, Mr. Means, for a deeper lesson in life!
The actual length of the island is 3.4 miles. Not great, not terrible.
Took me three years to get the joke.
@@Markstrosity 3.4 years actually. Not great, not terrible.
@@darthkek1953😂 👏🏻
😉
Wow, what an excellent, EXCELLENT presentation. I have just two minor comments: First, there is another bias that can come into play, and that's normalcy bias. When things go outside our experience and training, we have difficulty recognizing the actual problem as serious. Second, I was very happy when the accident occurred, that President Carter was himself a former Navy nuclear engineer. I'm sure he asked some very pointed questions during his visit, and would have seen through any B.S.
now the US has trump. His biggest bias seems to be ego.
Carter knew his shit! kzhead.info/sun/YJpxmbiun4mEdWg/bejne.html
@@JasonHoningford Yes he did and still does. Still sharp as a tack at age 94 with what is going on in our world today and often speaks out about it. An amazing man.
Fabulous talk! All managers and bosses should watch this! Great lessons to be learned!
Outstanding explanation of the Three Mile Island accident. Best I've heard on yet on KZhead.
During the last few minutes, I imagined it was the Galactic Empire conducting a "Who Destroyed the Death Star?" presentation.
I work in Harrisburg PA and can see 3 mile island from my desk at work.
This is a fundamentally brilliant talk, how have I only just seen this! This will be being shared around my organisation!
Very watchable, thanks! Going to catch up on the rest of these!
My former employer had a saying "It's not the people, it's the process!" Bad design of that leaky valve, system of error lights, and a single slow-ass printer were some of the causes. If you had good design, every possibility would have been thought of in advance. Redundant systems, extra sensors and automatic processes for every contingency would have been designed and installed beforehand. Human error might have been a small factor since it was humans who designed this crappy system. And that one submarine captains advice was certainly wrong. But that's it.
In the end, it's all human error. Humans designed and built the thing. I agree that it shouldn't be about blame, but about learning from our mistakes, and fixing what we did wrong going forward. There's much more to this presentation than I expected. It really needs a better title.
I agree. The system wasn't designed to make it impossible for humans to screw things up. That's also the fear of AI... Automation helps, but there are so many unknown circumstances that cannot be automated. For example, the Max Air AOA probes were part of an automated system; which stalled the planes that crashed. The automatic system overpowered the operators because it's impossible to regulate all of these automatic processes without an AI or human operator. There's a learning curve for every process as well. The submariners had their learning curve; and were also never properly trained on their system.
Once its automated people stop paying attention.
@@routtookc8064 Well I find it hard to believe people will "stop paying attention" in a nuclear reactor, however that IS the entire point of automating something! Not that you can really pay attention to an error in the buffer of a printer that won't even be visible for another couple hours.
Why FMEA is so important. Especially a complete one where, in addition to going through ever conceivable failure mode, you also explore the inconceivable...
This is made of 100% solid awesome. I was 13 at the time, and my dad was a physicist in a completely different field, but understood enough about nuclear power to know that events like this didn't represent a problem with the safety of nuclear power. The point he kept hammering on is that taken as a whole, the system worked. No leak occurred. They experienced an unpredictable series of corner-case events and *still* did not exceed the overall inherent design safety. The series of events is one that in a conference room talking about design, someone would say "but that's absurd. Next you'll say that we have to safeguard against aliens coming in and sabotaging the system." Among other things I learned from my dad is that Murphy's Law is not simply a cynical joke about how unlucky human beings are. It is a design principle that is baked into the way a federal government should design things that must not fail. From the Apollo missions to nuclear power to the $500 hammer in the toolkit on FMC's Bradley Fighting Vehicle. It only looks like waste when someone with an agenda wants to make someone else look bad in hindsight.
> corner case You should read about the Davis Besse incident. They experienced essentially the same scenario: secondary loop problem, stuck PORV, high water level in the pressurizer, manual override of the injection pumps, the primary coolant even began to boil in the pumps and reactor vessel. The only major difference in the scenario was theirs happened during startup and the core had much less decay power. Davis Besse was 18 months before Three-Mile Island. B&W didn't fully understand that their procedures and training were not preparing operators for a leak from the pressurizer, and that's the second story I take from this presentation and a little more reading. Mike Derivan was the shift supervisor at Davis Besse and has written a deeply fascinating (but more technically detailed) presentation. ansnuclearcafe.org/2014/04/23/tmi-operators-did-what-they-were-trained-to-do/ (One of the systems he mentions which was missing from this talk is the volume control system. Some water is drained from the primary loop - "letdown flow" - then cooled, filtered, demineralized. This cool water is injected at the coolant pump seals so that they don't overheat and so that the small amount of water that weeps through the seals is relatively clean. Water can also be returned to the loop through a heat-exchanger which reheats it. You can read an overview of all the systems on Gen II BWRs in the US here : www.nrc.gov/reading-rm/basic-ref/students/for-educators/04.pdf)
@@jordanrodrigues8265 The pdf link from nrc.gov leads to a "page not found" error. Apparantly the NRC doesn't want us to see that info. jerks
never mind, there was a right parentheses that auto. showed up after clicking the link. it works.
Brilliant. Thanks for posting.
This was simply an excellent presentation. Thank you.
Can verify that in Navy Nuke world having the pressurizer go solid is considered one of the most dangerous things you can do to a shipboard plant. It's drilled in our heads from day 1 and is brought up in every test, every qualification.
28:00 Well there's your problem. The reactor was being run by former reactors! They should have had qualified human operators!
[three] _former naval nuclear reactors_ 🤔😆😉
@@oubrioko it sounds like you are trying to blame and shame them, but its not their fault, look for the second story
@@hindugoat2302 In the talk, the presenter forgets the word "operators" at the mentioned timestamp, giving the phrase "[the 4 guys] were all former naval nuclear reactors."
@@hindugoat2302 The: "🤔😆😉," makes it obvious that we were _poking fun_ at what the presenter _said._ As @CodeBurger explained, the narrator misspoke by calling the workers, nuclear *reactors.* _People_ are not and cannot be former nuclear *reactors.* Since you feel compelled to troll comment sections to fight against imaginary grievances which never occurred, you should at least have the common courtesy of actually _understanding_ what you read before replying impulsively. It sounds like *you* carelessly looked for the FIRST story here . . . and completely missed it.
yeah reactor operators. Which were Electronics Technicians in the Navy. But usually the rating isn't used to describe their job (my rating was nuclear machinist mate but I had a further NEC classification). Presumably at least 1 of those guys was at least an officer rank who supervised Maneuvering (on surface ships we called it EOS).
I've been back to watch this through a few times now, such a great talk.
Excellent! Blaming people is unhelpful. Finding systemic flaws and working to correct them, that's the magic.
And this is kinda the flip side of a saying Pres. Reagan kept on his desk. It said: "It's amazing how much can be accomplished when no one cares who gets the credit." Change "accomplished" and "credit," to "fixed" and "blame," and you have the nutshell of this talk. Fred
President Carter was also a Nuclear Engineer trained by Admiral Rickover.
Carter was part of the reaction/cleanup team for the first nuclear disaster/accident. I think that was the small base reactor that cracked open and killed three... navy men I think? In canada.
@@dragonsword7370 SL-1 was in Idaho.
@@dragonsword7370 Carter was qualified on A1W in Idaho. He NEVER served on a nuclear powered submarine as he got out of the Navy shortly afterwards.. He was on a diesel boat and then entered nuclear training. I was NOT aware that he worked on the SL-1 reactor after the three workers were killed. Further research shows Carter got out of the Navy in 1953 after the death of his father. The SL1 reactor accident was in 1961. Therefore those dates kill the rumor that he was involved in the SL-1 reactor decon effort. The SL-1 reactor was an ARMY effort, not Navy. If you look at the SL-1 accident, it should be clear to anyone who was in Rickover's Navy program that RIckover would NOT have let that operation fly. Three men died unnecessarily due to incredibly poor practices. Rickover worried about details. The Army nuclear program apparently did NOT.
@CaliforniaCheez It says "Close Cover Before Striking Match" on a matchbook cover. He kicked a guy out of the program for not doing that when he lit a cigarette. (Failure to follow a written instruction)
@@SCGATOR2001 I was pointing out that "base reactor" accident that killed three operators was SL1 and wasn't in Canada , he was mixing details from two seperate accidents. Carter directed the cleanup of the NRX reactor meltdown in Canada in 1952 (Nobody died in that one)
OTOH Three Mile Island worked as designed failure and all. There was no crisis, the containment vessel was the bottom line solution for unkown problems and it worked. It was costly, but the design did NOT fail.
Don Adams - It did fail, the purpose of TMI was to generate electricity and a few months after it was turned on half of it melted and never generated a single spark of electricity again. Sure, it was a safe design and Chernobyl didn’t happen but so much money and resources was put into building that plant all to just go to waste, I don’t think it’s that hard to design a safe nuclear power plant, it takes an incredibly STUPID nation with a severely flawed government (Soviet Union) to build a reactor that can blow up as spectacularly as Chernobyl did. They took a plutonium producing reactor and modified it (RMBK) to make electricity and then didn’t cover it with a containment dome and then was like “hey, let’s do some tests and see if we can blow it up”. Wow... That control room at TMI was built by idiots and that PORV should of had better indicators on if it was open or closed. I think if that plant was built today the accident would of never happened due to computers and cameras automating every aspect of TMI. TMI proved that even doing all the wrong things nuclear power can still be safe, it’s just a damn shame a new reactor was turned into a nuclear waste dump instead of generating power. That incident ruined nuclear power in America when we could be like France and had clean affordable electricity. That PWR proved to be safe but not good, I grew up next to the same reactor in Northern California called Rancho Seco, it was shut down because it had nothing but problems and the public voted to scrap it, another waste of so much potential.
Would you believe if it's meant to generate electricity and it can't then it's failed?
Peter Allan - I would. Such a waste of resources by dumb people. I want nuclear power to work in America, we just need smarter people in charge.
This is an excellent talk, easily in my top 5. It's interesting throughout with an eye-opening lesson at the end.
This talk is amazing. You did a great job telling the story Three Mile Island, but the skillful way in which you used it to illustrate systemic problems is even more inpressive.
I love how he did not vilify the engineers who had to make these decisions. A little perspective goes a long way. With a little hindsight bias, you could even say the events of three mile island were ultimately a good thing.
Nahhh, surely doing better is a better way then only blaming people. But that horrible midtakes were made and we ciuld havr done without all this is the core fact that shouldnt be forgotten also. Otherwise your consequence is glossing over not real fair critical forward thinking analysis.
I remember as a teenager when I discovered that nuclear power plants were basically hi-tech steam engines. My mind was blown. Up to that point, I'd always assumed that they somehow converted radiation directly into electricity or that maybe there was some sort of small-scale version of a nuclear explosion going on, with the explosion turning a turbine. I felt kind of stupid when I found out how it really worked.
This is just popular extant power designs, alternate options exist. RTGs are common and aren't big steam engines. For more theoretical options, one could put a nuclear source at one end of a sterling engine to create power.
Do not feel stupid. "Stupid" is shorthand for "not being curious enough to question assumptions."
beautifully explained- thanks
This is an excellent presentation if you ask me. Thanks for sharing Nick!
I am a quality professional who has worked in the aerospace and defence sphere for 12 years, as well as life safety equipment. I never accept human error as a root cause. How that error was allowed by the system to wreck the reactor was the root cause.
Alastair Archibald Which is why human error is always the cause. But not just the error of the operators - especially the error of the ones who made the system. In Three Mile Island, the error made by the ones who made the system is that they didn't include redundant valves in the design and that they gave the operators the possibility of turning off the water injection pump despite the fact that it was an extremely important piece in the system. If you know humans do oopsies because of some bias or something don't let them do it - it should have been in the design. Besides this, those operators weren't even trained to run this specific type of reactor. Reactor designs are as many and diverse as there are types of cheese. Plus, humans under stress should be considered mistake generating machines. If you don't believe me, there is a study that tested groups of students by giving them a bunch of problems to solve in a fixed time interval. If for each group you decrease the time linearly, because we assume students to have some average computing power that tends to be constant in large groups, you would expect the percentage of solved problems to drop linearly as well - except the results drop much faster - it is because the students who had less time felt more stressed. The number of bad answers in the solved problems also correlates with the stress level. The lesson is simple - as a system designer, don't assume that a bunch of scared unprepared operators desperately searching the light panels and scrambling for solutions is an effective mechanism for preventing accidents. By the point we reach that phase, it is already too late.
Refusing to accept human error as a root cause is a human error.
@@cezarcatalin1406 Have you got statistics for that cheese comparison? I think you may have made the error of massive exaggeration.
A backwash filtration system that clogs up the works didn't design and built itself. A leaking valve, known to create unreliable and erroneous information, didn't go un-repaired all by itself. The first issue may or may not have been understood by the design engineers, the second issue about the leaking valve was known and given a contemptuous hand wave. That, my friend, was human error.
Can confirm, as a professional working in medical laboratory. The best way to eliminate human error is to eliminate human involvement.
As a mechanical engineer, specializing in monitoring and control systems, I can see first some design faults in the monitoring system, thus the operating faults caused.
outcome bias how are you doing ?
joy to watch. thanks for this!
This is really informative. I really liked this presentation.
This was a good talk. So.. 1st story is told by someone who was around for the event but not actually there. 2nd story is told from a first person point of view by actual participants during an event. 3rd story is how we choose to document history. 4th story is an evaluation, analysis and interpretation of the history that we had chose to document.
This is why I left the Sub force after serving 13 years. The nuclear Navy is constantly seeking heads to roll. Every day on board a ship was another day wondering if you were going to Captains Mast. I finally had enough. Not only did I refuse to re enlist, On my departure interview, I removed my Dolpins and threw them onto the XO's desk. I ended up "Finishing my 20" Actually did another 13 years) In the Reserves as an EOD Diver. One drill weekend, we had a uniform inspection. The CO stops in front of me and asked "Weren't you on submarines?" "Yes Sir! I WAS!" Came my response. "Where are your Dolphins?" "I choose not to wear them sir!" "Either you put them on or submit a request to have to Submarine qualifications stricken from your record!" Aye sir! It will be on your desk by 1630!"
snipe69 The PEB was a copy of the ORSE nukes go through. Extending nuclear regulations into non nuclear areas was referred to as “Creeping nukism”
@snipe69 Yep, and it got worse during the post Cold War draw down. The Sub Force was literally cut in half and all those Submariners were assigned to other communities bringing their "Nukism" With them.
@snipe69 The XO of my last boat sealed my decision to leave active duty when he gave the following speech on the 1MC "Do not let Glasnost fool you! We are still at war! Only now that the Soviets are bankrupt, the enemy is dirt! Commence Field Day!" I submitted a 1306 to Go EOD and my detailer shot it down declaring "If we let everyone who wanted off the boats to transfer off, we would have to tie up half our ships for lack of crew!"
What a wuss and a fucking cry baby!! You should have never join the military px warrior! Now fuck off!
@@11B30Inf Stockholm Syndrome much?
Wow, that was reeeealy interesting! Not just the details about the accident, also the paradigma of asking for the second stories. I will take this thoughts and include it into my daily work. Thanks alot.
Wow that was a great explanation that made it simple to understand what happen. Awesome talk.
I had those same "How Things Work" books growing up
I swear our civilization will end with an assumption of being correct.
i have a feeling at the end it will be a bunch of people trying to point the finger and place blame.
Look around you.
Like AOC?
@@allgrainbrewer10 like your mom
eaglesclaws8 are you 12? Go watch a Justin Bieber vid loser
Wow, this was strong. I was not expecting what came in the last 5min. Stellar!!
Between @BecauseScience and this, I've learned a hell of a lot about Nuclear Power and now a new thought process as I return to work on Monday.
as someone who was involved in the TMI recovery effort, I think this is an excellent summary. we all thought the operators were screwups until we looked at things from their perspective. don't blame the operators - they did what their training and experience told them to do. I think Mr Means could've talked more about "setting people up to fail" - the very design of the BW Nuclear Steam Supply System with once thru steam generators that contain very little water set operators up for failure - the PORV is designed to open with loss of condensate. the other PWR designs (Westinghouse and CE at the time) do not. BW did this to try to distinguish themselves from the other vendor (the one thrus can superheat steam slightly, others cannot) which was driven by their desire for economic success. so should we revamp capitalism also. well beyond my pay grade.
Well, with your experience, it's possible You might be able to answer this question I've been developing... How did the pressurizer have water in it when the main reactor tank was running low? Is there something I'm misunderstanding about the design?
@@caseyriley1014 IMPORTANT - Recognize that the PRZ is very much isolated from the Rx vessel - only being connected to the RX vessel by a small surge line that is large enough to permit pressures in the PRZ and Rx Vessel to be balanced. the question you pose is exactly the issue that the "Post TMI lessons learned" modifications addressed - the operators were trained to maintain pressurizer level (with heaters and sprays internal to the PRZ) and had no idea what the water level was in the Rx vessel (they all do now!). they were trained that if the PRZ water level is good then the Rx vessel water level is good - except when you forget about what the "P" in PWR stands for. (Note - they did not recognize the PORV was stuck open because the PORV was almost always leaking and the temp gage on the other side always indicated high temp - so a stuck open PORV did not look much different to them then during normal operations. OUCH!) ultimately
@@caseyriley1014the pressurizer is supposed to be the only place in the primary coolant system that isn't completely full of liquid water. It's maintained at a higher pressure and temperature to be sure that the steam bubble remains there and only there. That's the reason pressurizer level is the proxy for vessel liquid level; the vessel is supposed to be completely full all the time, and any change in total volume of water will show as a change in pressurizer level. As the steam bubble was released through the relief valve the pressurizer pressure lowered and allowed reactor pressure to basically push the level up and a steam bubble to form in the vessel.
The Anatoly Dyatlov second story would make musical... or ballet even...
Yeah...you could definitely see him in a different light after this, he still was an arse though...apparently.
I can imagine one of Dyatlov's songs... _"Not great, not terrible, but either way, you're delusional! It's like, a chest X-ray, so get yourself, to the infirmary!"_
@@ChrisView777 Knowing his history before Chernobyl has to be taken into account, and the simple structure of the system he was working in. He still was totally at fault for doing the test the way he did the problems of the reactor being kept from him, and others, is what blew it up.
@@Swarm509 But even forcing the test to go ahead was still a systemic failure elsewhere as in the USSR you either did your job or you had no job, do you think the only reason he rushed the test is because there was a football game on he wanted to see? No, the test had to be done on that day or it meant he was incompetent and not fit for his job. The only thing excuses got you over there was a bullet or the gulag. A systemic failing.
there is in fact a long interview of him and is quite interesting if you hear his perspective
So glad I got watched this video! Facts about the accident were good, but the unexpected message at the end was a truly awesome eye opener! Thank you!
This video is fantastic. I admire your objectivity and desire to portray the big picture. Well done.
I feel spoiled with my 9600 baud Arduino printout
Baud is a vastly misused term. What you really mean is bits per second (bps). Baud is not the same, except at a few very low numbers, such as 300, which is where the misnomer originated.
TMI is often viewed as a way of increasing accuracy. I have seen this over and over in many different places. The people who design a system frequently do not understand the mental implications of constant data overload on operators.
Nuclear Energy is really complicated this is a official Training Simulator for Nuclear Powerstation kzhead.info/sun/ZZyBk8Wkg5GIeas/bejne.html
@@user-si5fm8ql3c When you state that "Nuclear Energy is really complicated", are you referring to the principles of nuclear physics or the actual operation of a nuclear reactor (as a part of electrical energy production)?!? Regardless, "Nuclear Energy" is _not_ particularly complicated, although individuals of average intellect (or lower) will likely fail to adequately comprehend much of the science involved.
the BEST story explaining on what happened and why awesome presentation
My good sir this is the first talk I've ever seen from you, but I'm already a fan