How this series doesn't have more views and likes is baffling. You have a tenured professor giving away hours of very useful, easy to understand and factually correct knowledge, for free.
Because as my 12 year old superasian says the information is not complete. This is too simplistic and does not cover the initial failure of the condensate polisher system and the involment of the pressurizer water level indicator still indicating the system as full even though pilot operated relief valve system was evacuating water from the system.
100% Worked in a plant with Ammonia, lots of it. Well I worked on the operations side which was monitored with thousands of alarms. One night the accumulator built up pressure and the relief valve should have opened but didn't. Que a little bit of panic, followed by massive amounts of phone calls that went unanswered for a pretty good amount of time (30-40 minutes) 1 AM. Easy to spot where the problem existed by the piping not having frost built up on it. Hard to figure out why though. Short story was a whole fucking 2x4 left inside some piping that was earlier locked out for another reason and worked on. Still don't know how or why the 2x4 was in there for any logical reason. Some contractor got fired for sure though.
There is another valuable lesson here for Electrical Engineers and mission critical monitoring systems. At TMI the "valve closed" light in the control room merely monitored the power to the motor or solenoid that was to close the valve. There was no feedback that the valve actually closed. You need a separate sensor to indicate if the mechanical action actually happened, not just that it was told to happen. This idea was nothing new at the time. Every aircraft with stow away landing gear had/has such a system. "The gear down and locked" light actually monitors the mechanical action, not just that power was sent to the motors or hydraulic valves. I am really surprised this was overlooked in the TMI design.
I agree that no physical feedback on a safety value, to prevent a major nuclear accident, is nothing short of negligence on the part the the design team, the oversight team and the nuclear regulatory commission. Can you imagine being in a plane with "3 greens" from the landing gear without a physical feedback mechanism? There would be hell to pay for the 1,000's of people killed in planes with stuck landing gears. A major nuclear accident would kill millions of people. What would we get from the design teams after millions of people die from their oversight, "Oh I am sorry, I forgot about that!". That is such bullshit.
Its worse than that, if I understood a version I read elsewhere correctly the valve wasn't rated for mixed phase flow, that is to say water and steam mixed which is something that might happen under emergency conditions. To get around this a manual block valve was added, which might have been OK if anyone had been able to determine that it needed closing. Also I think there is a more general anti-pattern in industrial plant design where the instrumentation fitted turns out to be sufficient to monitor normal operation but leaves dangerous blind spots whenever it shifts into abnormal operation. For examples of this watch CSB reconstruction videos and listen for the words "what the operators did not know was" @@richardcommins4926
System was believed to be reliable to the point that such a confirmation system was not nessessary, however there was in a place a procedure to ensure PORV closure, a thermocouple was installed within the PORV outlet, this would indicate as to whether the valve was closed or open.... this check was carried out and the results ignored, misinterpreted and put aside as a result of the previous operation of the valve. Like many other steps if they had followed procedure and closed the manual outlet valve down line of the PORV the loss of coolant would have been prevented.
I just figured out that the way he writes on the board is that he doesn't!!! His wedding ring and watch are on what would be his right hand. Thus, this video is actually mirrored!
It turns out the big valve was designed as a scaled-up milk line valve. In a creamery, the valve is continually lubricated by milk and milkfat. Not so in a reactor, there the water is corrosive. A hard-learned lesson. A billion-dollar lesson.
GREAT VIDEOS... Questions. There were indicators telling the operators the holding tank was filling up (I watched the second video) but because of it's location NOBODY saw it for almost 2 hours? No indicators telling them the core area was losing water, especially a totally uncovered core? If EITHER of those things happened and surely they had that.. They would have known to restart the ECCS either slowly or much sooner and trace the water problem to the stuck valve. I'm still amazed that even in the 70's they didn't have backup indicators on everything such as the valve or people to run physical checks on everything during a problem. Also wouldn't they have had Hydrogen (the gas) indicators as yet another fail safe and gas exhaust ports to prevent build up ?
They did have a backup indicator in the form of a temperature sensor on the valve outlet piping. The problems were that that particular valve had been weeping (a very slow leak) that caused a higher than normal temperature even without the valve being stuck open. Also, the expected temperature would not have been as high as one might expect, given the behavior of steam as pressure is reduced. Finally, the backup indicator was not in the main control room benchboard area, and was not in their procedures to check. After TMI, plants went to a vibration method, such that if the valve is open, the tailpipe will be vibrating.
Do you subscribe/have you heard of the theory that there was sabotage involved? I have no "dog" in this theory either way .. more can be found on the theory here: atomicinsights.com/was-three-mile-island-an-accident/ Also, there is quite a bit in the comments section on that page with additional material and observations by others from that era.
uploadJ I had not heard of such a theory. I was analyzing the performance of the men in the control room from the standpoint of organizational psychology.
Jim Lahey everything I was able to get my hands on said that on March 28, 1979 at 4:36am a crew was trying to clean biological build up out of the filters for the main feed water pumps. And for some unknown reason the main pumps failed. I always thought maybe they dislodged something. Then a bypass valve failed to open as well. I have never come across any information as to the cause of that.
@@jimlahey5354 The accounts that I heard shortly after the accident are rather involved. When water got trapped in some small instrument air lines that controlled valves for condensate filters, the filters isolated. The main feed pumps that pump water into the secondary side of the steam generators then shutdown (not 'break') because of a loss of suction pressure (pumps often have low-suction pressure trips, to PREVENT them from breaking). When this feedwater was interrupted, the primary temperature rises very quickly, the expanding primary coolant surges into the pressurizer and raises pressure. The PORV valve is designed to automatically open and relieve this pressure increase.
At IEP, I’ve been watching your content back to back and I absolutely love it. A huge thank you for your sharing of knowledge. With regards to this video, you mentioned water being a combination of hydrogen and oxygen which when super heated has the potential to break bonds and release hydrogen (as per on of the Fukushima incidents (I’m no expect but I believe I heard that somewhere)).... my question is, is there a moderator that would be better suited than water. The capability of releasing a volatile/explosive chemical seems like a bad idea if in the most unlikely of situations. Thanks again for your work. My greatest respect
A good moderator has a low atomic mass so that the fast neutrons (coming directly out of the splitting U-235) are slowed down by 'elastic scattering'. The lower the mass of the atoms where the neutrons 'collide' with, the more energy is lost with each collision. Hydrogen (in form of water) is a ideal moderator, because is consists only of one proton (and an electron of course). Another important feature of a good moderator is that it only slows neutrons down, but does not (or nearly not) absorb neutrons in the nucleus. You could use craphite (like in RBMK, Magnox..), but this has other disadvantages of course. CANDU reactors use deuterium as moderator (because deuterium absorbs less neutrons compared to hydrogen), in form of D2O (= 'heavy water'), but in case of an accident it will also will react with the zirconium fuel rod cladding at high temperatures (because its chemically comparable to hydrogen). The disadvantages of water as a moderator and coolant don't outweigh the advantages. Anyway, the zirconium-steam reaction will only occur in case of LOCA where the fuel cladding has the chance to heat up. If the core is/stays submerged in liquid water (by the ECCS), there is little chance this reaction will occur. And in case there is some hydrogen formation (during an accident it's not only formed by zircaloy-steam reaction) there are recombiners in the reactor building, which will recombine hydrogen with the oxygen in reactor building to form water, without chance of explosion.
@@dvraemdo it's always amazing to me when you see the entire point being betrayed by Propaganda engineering. Here avoiding splitting water molecules into two is argued to be a good thing and the point of the cooling in part with you double Downing on the propaganda by not mentioning that you can safely obtain hydrogen Which is far better than electricity because of the ability to store it including at twice the density of liquid without the energy tax of liquid. Criticisms of hydrogen are based upon the cost of compressing it but modern hydrogen storage does not have that cost. So but using helium until of course the powers-that-be talked LOL JAPAN out of this you can produce hydrogen which can prevent the use of fossil fuels entirely. So when you're definition a good moderator does not go outside the reservation and provide energy outside of the electric red grid. I believe that a good moderator does the opposite oh, it makes hydrogen so cheap that it itself can liquefy it because saving half the space is not a reason to use hydrides. Using hydrides is an excuse to delay the Banning of all not merely fract but all petroleum for purposes of combustion being granted leases to extract. Water is what the uranium is 4, water is not there to moderate water is there to be turned into gold. I don't know what color of liquid hydrogen is but I know it's better than black. Black lives matter Black Gold is a Star Trek showed it to be the ultimate Evil.
In this comment I again attacked those who cater to the existing Market in essentially a kind of institutionalized assumed dysfunctional sloth. Whether you produce electricity or not is a direct consequence of the cost of the energy. The energy obtained from fission is not expensive enough to justify making it even more expensive for little reason. When your primary energy source is so low in cost only corruption explains you converting it into electricity. The use of electricity is the sort of perfectionism but what is being optimized is is neither profit efficiency or really anything except a kind of protection of that which is obsolete which is directly against the health of the planet and the interests of the people. There is a rhetoric concerning efficiency that is the way the fossil fuel industry stays alive. It is not what they practice but it is what the feed store alternatives to fossil fuel the mischief and Malice imposing devastating efficiency costs up on the competition. When there are no externalities, it is very important to maximize cost engineering which means you're maximizing waste. That's how you stop all internal combustion that's how you stop all burning for the purpose of obtaining expansion only without benefiting from the heating much or vice versa. The fossil fuel industry has has as a rule waste oh, so you see pressure being harvested as you see an internal combustion, or when you have boilers, historically and with almost no exception in the marketplace of existing operating aggregate of destroying systems he'd only being harvested. It's the way they roll, but they have prevented us from Rolling that way even when it would be defendable. When you don't do what is defendable sometimes it's suicidal. This is one of those times. Delivering energy where it's needed, doing work, using more of it, but polluting less
Two key misunderstandings about plant behavior at TMI. Tail-pipe temperature indications of an open PORV valve were not fully understood. With 600+ degree F steam upstream and downstream temperature several hundred degrees cooler, some operators miss-interpreted this to indicate the valve was shut. The phenomenon of 'shifting the bubble' from the pressurizer to the reactor vessel. A "steam-space LOCA" can quickly lower plant pressure below the saturation pressure for temperatures normally found in the reactor outlet. When this happens, the water in the vessel starts to boil and force water into the cooler, lower pressure pressurizer. Plant design at the time only had level instruments on the pressurizer to indicate primary system inventory Operators were confronted with ECCS systems running, and the pressurizer level indication rising rapidly. Fearing the dangers of OVER-filling the primary, they shut off ECCS. Operators are trained differently and Emergency Procedures now guide them how to properly deal with this eventuality. Since TMI, MANY changes have been made. But three key ones are: better flow indications for PORV valves, separate reactor vessel level instrumentation, and Emergency Procedures to guide operators.
Even today not all PWR's are equipped with reactor vessel level instrumentation (or they do, but its only designed to use during shutdown conditions), but with the completely rewritten post-TMI emergency procedures (and training) every reactor operator knows not to rely on pressurizer level indictors when the water in the reactor reaches satured conditions. Another important post-TMI change is that the reactor protection system now inhibits the manual shutdown of the ECCS during a certain period after automatic start (there's a timer built-in).
Thanks for this explanation. I was left wondering why they didn’t notice water levels decreasing in the core, but duh. They didn’t have any way to monitor directly!
An excellent description of why pressurised water reactors are potentially so dangerous and therefore so expensive to make safe. They are indeed very safe and the industry is the safest power source we have. Molten salt reactors especially the Stable Salt (SSR) remove all of those hazards. There is no pressure and no water in the reactor core. They can even lose cooling at full load and remain safe even if the control rods do not drop. That makes them much cheaper to build in fact cheaper even than natural gas plants. In normal use the control rods would drop and the decay heat is removed by convection only.
David Elliott If It were so easy, it would be happening everywhere. Do a lot more research and you will see the enormous difficulties and drawbacks inherent in the use of salt. You conveniently cite two or three advantages as though they exist in a bubble of some sort. Engineering does not work that way.
We have done research here at our lab, and it seems the way is done is that you set a large mirror, then the camera (filming the mirror), then the Glass, where you write, then the person. And so the person appears to be left handed because they and the text are mirrored. We have surmised this from other lecturers online that all appear to be left handed in the videos. This is how we think that this great mystery is solved. You are welcome.
I live here in Middletown all my life. Born in 85. Most my family died of cancer from high raydon in soil. My grandma got a call from relative that day that works there saying " they don't know what to do or what's going on"
Not only is this a great lecture but I'm also stoked by how his writing appears the correct way when written from behind glass. I assume some trickery is involved (or I'm being dumb, either is a probable explanation)
I sat through a quick presentation about TMI at a local training center and I remember it very close to what he is describing. Their explanation was that at that time there wasn't a single person in charge and that everyone had focused on the affected instruments and were not seeing the whole picture. I would be interested to hear from others more qualified to address it.
Probably the most logical explanation I've seen about TMI. Koodos. Only part that gets me is how nobody noticed the decreasing water levels within the reactor? I'd imagine there had to have been some type of indicator for that.
A similar incident happened at the Davis-Besse NPP on Lake Erie outside Toledo, Ohio but the operators discovered the coolant leak and corrected it before the core became uncovered. Had it gone undiscovered for another forty minutes it would have melted down just like TMI2. For a complete list of accidents there see en.wikipedia.org/wiki/Davis-Besse_Nuclear_Power_Station
I think there also was a thermocouple on the tailpipe of the relief valve to indicate the temperature. I high temperature would indicate a leaking or stuck valve.
One thing I'm not clear about because I've only seen block diagrams but the way the pressuriser was connected to the reactor vessel appears to leave a dead space above the vessel outlet that was above the connection to the pressuriser. Now I'm guessing that under normal or low pressure circulating coolant conditions this didn't matter as any traces of gas or vapour would be forced out but under passive cooling there doesn't seem to have been a way for vapour at the top of the vessel to reach the pressuriser. Had there been such a path maybe the pressuriser level would have dropped and instruments would have reported this.
great explanation of the event , without propaganda etc and the "professor" left out the difficult part about pressure not pressure isothermics , but that makes sence
Can you please do a video on the "Demon Core". I have read a lot but still dont quite understand what happened.. A screwdriver slipped. You are great at explaining these things..
What happened is that a pair of scientists, at seperate times, were both handling two hemispheric halves of a uranium core, and during the demonstrations they were conducting, they both made the same mistake of losing their grip, allowing the halves to come together and reach criticality.
Nuclear power plants should be maned like NASA mission control - a separate controller for EACH major system at their own console next to each other and overseen (physically and operationally) by a Master Controller. One monitors the pumps to the heat exchanger, another monitors the pumps to the reactor. Another monitors the reactor and control rods. Another monitors the safety systems. Another monitors pumps to the cooling tower. Another monitors system wide temperature, radation levels, and flows (redundant). Then each controller has an operator at the control panels (which are too large to be in front of the controllers/monitors). Every time a nuclear accident occurs, it's appears to be a bunch of people scrambling about, no one person has any focus. The controllers should remain seated and at their consoles... their respective operators should be at their control panels and never leave their stations. Another way to think of this this is how a US Submarine works. I hear that in plants like TMI, many warnings lights are normally on at any time. That is a major flaw... warnings MUST be corrected promptly or the computer shuts down the reactor - What I call: "Penalized Safety"... sort of like a child getting a "timeout" in the corner for doing something wrong. The computer just spits out warnings and alarms, back then by visual lights (annunciators) and a hard copy printout for record. But it's needs to PRIORITZE the warnings and alarms, not just "sea" of lights and items listed in the order they occur, but a printed PRIORITIZED list as well. Example: 100 warnings and 20 alarms have gone off. A 2nd printout (now a Computer monitor screen) shows the Alarms and Warnings IN ORDER of root effect on the system. In the TMI case, that should have been: All have time stamps down to the millisecond when they first occurred as well as a time stamp the entire "summary" was updated. 1. Main Pump to Heat Exchanger Off Line 1a. Failure mode: Top Brearing at 250 C overheated by 10 C 2. Valve from Backup Pump to Heat Exchanger Closed 2a. Failure Mode: Manually Closed 2b. Condition: Low Flow Rate in Coolant 3. Pressure Relief Valve Open 3a. Failure Mode: Commanded to Close but Close Not Comfirmed [note, the Valve actually did not have a Comfirmed Closed signal - a design flaw] 4. Reactor Water Pressure and Temperature Anomoly: 4a. Condition: 250 PSI 10 PSI High, 280 C 20 C High And so on... but not in the order they occur or how "critical" they are in general (one would think the 4th item would be first), but in which order they are in importance TO CORRECT as a fault/recovery tree. This way item 2 stands out in EVERY update... it would have been impossible to ignore and thus corrected in seconds... thus items 3 and up would have never occurred.
What is the cause of this accident due to issues with Control Room usability and ergonomics, because of the way the human interface where is the reactor Control System was such that there was so much complexity, that the alarm for the backup pump was missed?
They didn't have a way to measure water level or presence inside the reactor pressure vessel either. You ALWAYS want to be able to know how much is inside a tank, at the very least at 3 4 levels and you ALWAYS want some sort of alarm or automation before hitting full.
There is actually a lot more technical detail not covered here. Events leading up to this, mis-reporting of the relief valve temperature, Navy training about pressurizer levels, B&W's lack of coolant level instrumentation in the reactor, and so much more.
How to make an illusion of backwards writing: 1) Set up a glass plane 2) Position a presenter with a white-board marker on one side and a camera on the other. 3) Record the presenter through the glass pane while he does his normal writing on his side of the glass pane. When viewed from the camera side, this will appear mirrored 4) Digitally mirror the recording in post-production so it will now appear the right way around to the viewer. There, now we know how they do it, it's very simple really. No need to get all mind-blown over a simple camera trick
@@troelshansen6212 too complex. Put small mirror in front of objective at 45° and rotate whole camera setup to 90deg. You will get mirrored image immediatly without any furthe post-processing.
I worked for 40+ years in the nuclear industry. Reactor Operator and Senior Reactor operator at 5 different nuclear facilities. I was 2 weeks away from my NRC exam date when Three Mile Island accident happened. Because of the accident we attended over days upon days of lecture on the accident and Heat Transfer and Fluid Flow classes. I worked with the Reactor Operator that experienced the same event at Davis Bessie and mitigated that leak. I operated a reactor that was a sister plant to Three Mile Island. I have been trained by people who were Reactor Operators and Senior Reactors at Three Mile Island at the time of the accident. With all due respect to the Professor, his explanation is both confusing and technically incorrect. First I would call TMI which in this case stands for Too Much Information. The event was initiated in the secondary system feedwater due to operator error but this had nothing to do with the reactor accident. There are many things in the secondary side and electrical distribution that can result in a reactor trip but discussing these details distracts from the reason for the reactor accident. In addition, while the story behind why a reactor trip is called a SCRAM is interesting again it has no bearing on the accident. Concentrate on the Primary Reactor Coolant system in discussing this accident. The important facts of the event include not only that the relief valve stuck open and indicated closed but why it indicated closed. There were other indications that the operators used that made them believe that the relief valve was closed but the Professor never addressed those issues. The operators manually shut off the emergency cooling system and they did this because they were taught as was I that there was a more overriding concern and something again that the Professor never addressed. I can assure you that at no time after the reactor trip did the operators believe anything was normal, slapped themselves on the back for a job well done, or were filling out any files or paperwork. My facility was contacted 2 hours after the event asking for Radiation Control Technicians because they knew they had a really big problem. In an event such as this, there are literally hundreds of panel alarms and they had a pin fed paper printout of all the alarms that was constantly printing out but 30-60 minutes behind due to the sheer number of alarms. The best way to understand this event is not by writing some words on a board but drawing the reactor primary system on the board and using that to explain what happened. In my experience, the best instructors are those that understand exactly what happened but have an average intelligence If an average intelligent person can understand the event then they are more likely to be able to explain it to the general population that are also of average intelligence.
Why was the water level in the reactor falling, after the valve for the back up pump was opened ? Cold water must have been circlulating the reactor agian, keeping the 60 MW's of thermal power in check
Because it's only circulating the water already in the reactor, not adding new water. Because the relief valve was open, all of this circulated water was being drained.
Three Mile Island is so named because it is located three miles downriver from Middletown, Pennsylvania...........I also thought the same ...so i had a look in the Net....the professor is right..!
I don't get it. Water level drops, hydrogen is created... (Video ends) and what happened next? Was there an explosion? Meltdown? Why the three mile island just couldnt be repaired and the gas and water safely pumped out?
Anyone notice he's writing backwards, on the back side of a transparent sheet, with his left hand? And wears his wedding right on his right? I bet he's writing forwards with his right hand, and the video just flips L-R so we can read what he wrote. Cool technique.
What happened to the guy that signed off on the false routine maintenance that started this whole mess? Was he billed for the 5 billion loss, jailed or hanged? Who did end up footing the bill, the electricity customers, a little bit every month? I don't believe for a moment that the investors paid that 5 billion dollar bill.
Gives you confidence in professionals. I have always felt very uncomfortable listening to the explanation for the TMI incident because it makes me realize how stupid professionals can be. Fukushima is just more of the same, one bad decision piled on top of another bad decision. I'm a supporter of conventional, fission, nuclear energy, but it's hard to feel confident about it when the people who are the experts repeatedly make stupid mistakes. It can be safe if they use their heads, but these types of incidents leave me scratching my head.
re: "Gives you confidence in professionals. " Quote "Professionals" (white collar) unquote, or the plant (blue collar) operators (technicians)? I have nothing against either group, I'm just asking, uh, for a friend ...
@@uploadJ I am skeptical of the skill of both the engineers who design the plants, and the technicians who run them. The designers may be hindered by orders from the company bean counters who want to save money, so maybe that's a factor that should be counted for. However, if I was on a job that could kill loads of people if it wasn't done right, and my boss told me to cut corners, I'd quit. I've walked away from a number of jobs when the money people wanted me to do things I didn't want to do.
How do I write backwards? I don't. I am right handed, and wear my wedding ring on my left hand. I am writing on a glass board and the camera is filming me through the board. Then they "mirror-image" the video they took before presenting it to you. Everything is flipped right to left. The images you see are added after the fact, so that is why they are not flipped.
One reason, because it screws up the downstream eco system. Next to Madison, WI, where I grew up, there was a big heat exchanger for the MG&E powerplant during the early 1960s. It slowly raised the average temperature for the whole lake (Lake Monona), which suffered increased algae blooms, stank, had major fish kills and never froze completely in the winters anymore.
It decrease efficiency ! Let understand why? 1)At beginning you have use river water at normal temp let say 28 °c 2)you heat the water at above saturation temp to turn in gass is about 100℃ at 1atm, using nuclear reactor heat. 3) Now you have steam at high pressure , flow it from turbine generate electricity ,an end of turbine you got very low pressure steam and moistures. 4) you need water in liquid form to again put in nuclear reactor and heat up, but you have moist steam. So , you use surrounding water in heat exchanger to just *condensed the low pressure moist steam to Saturated liquid water*. And condensed liquid temp is higher than the river water which is at 28℃ *So, this time you already get water little bit below to 100℃ ready to convert in steam with minimum heat gain* and cycles repeat...
I had the same thought. In most of the video he appears left handed, but If you look at the bits in a classroom he appears right handed (he holds a hammer, and an ax in his right had), so I suspect that the person who suggested the video is flipped left->right is correct.
