Chernobyl disaster

[Pawelk198604]

I wonder why the reactor exploded when it was about to shutdown?

I heard on Discovery channel that the reactor exploded when the technicians started "Emergency Shutdown" by putting back all control rods, i cannot understand, it should stop reactor why it doesn't work.

[lajoswinkler]

Because there was already enough heat liberated inside. Operators had most of the control rods pulled out (xenon buildup tricked them) and they were in fact playing with such an unstable situation.

Rods don't just fall down in this stupid design, and they also have graphite tips. When they are lowered down, there's a period of neutron moderation by those tips, which increases the fission rate.

Too much heat caused too much steam pressure to be generated, and then the construction gave up.

[Tex_NL]

As I recall they where about to run emergency shut down tests. They purposefully created an event to see if the system would perform the emergence shut down. It didn't. And by the time the under qualified staff realized what had happened, or rather; what didn't happen, it was too late.

But you could have found all of this yourself with a simple google or wikipedia search. Really, finding information on the web is very easi. As long as you're willing the search for it of course.

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

[LordFerret]

As I recall they where about to run emergency shut down tests. They purposefully created an event to see if the system would perform the emergence shut down. It didn't. And by the time the under qualified staff realized what had happened, or rather; what didn't happen, it was too late.

But you could have found all of this yourself with a simple google or wikipedia search. Really, finding information on the web is very easi. As long as you're willing the search for it of course.

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

Brings to mind: "If it ain't broke - don't fix it."

[lincourtl]

Brings to mind: "If it ain't broke - don't fix it."

I'd think, "If it ain't broke, don't go trying to break it," would be more the thing. ;-)

[ZedNova]

I'd think, "If it ain't broke, don't go trying to break it," would be more the thing. ;-)

"If it ain't broke, and we don't know when it will break, shouldn't we?"

[Tex_NL]

Brings to mind: "If it ain't broke - don't fix it."

Problem was, it WAS broken but they didn't know.

[Jovus]

QA note closed: product working as designed

[cicatrix]

A nuclear power plant requies energy to run. While there is fuel in the reactor core you have to constantly pump cooling water through it to prevent overheat. Usually, a small portion of turbine electric power is taken for this purpose. If the reactor is shut down for maintenance then it's powered by neighboring reactors or from outside.

There are also diesel generators for emergency purpose (power outage). However, they take time to start (1-3 minutes). So, how can the coolant pumps be powered while the emergency generators are starting up? There was a question how much time since the steam stops powering the turbines they would continue to spin and thus provide power to the pumps. So, there was an experiment planned to see how much time the turbines would spin down.

On the 25th, April 1986 personnel had to test it.

However, some load was needed for taking necessary measurements. The second problem was the emergency shut down system which automatically switched off the reactor if its thermal power dropped down to 700-1000 MW and there would be impossible to repeat the experiment because of the xenon poisoning of the reactor core.

So, the switched off the emergency shutdown system (ESS) and use reserve centrifugal pumps. (First and second mistakes - there was no need to switch off the ESS, and they could use ANYTHING BUT pumps because this led to the tragedy).

1.00 pm: Lowering the power of the reactor

1.05 pm: The thermal power lowered from 3200 MW to 1600 MW. Turbine no. 7 stopped, electrical systems are powered from the turbine no. 8

2.00 pm: ESS blocked. At this time the "KiewEnergo" operator ordered to postpone the further powering down (the end of week, energy consumption raises). The reactor works on its halved power and ESS has not been switched on again. This was a serious operation error but it had no impact on the outcome.

11.10 pm: The operator allows to continue. Personnel continues to power the reactor down.

Apr, 26, 0.28am: Reactor is powered down so low that the personnel had to switch from local control to general (during normal operation they can move the control rods individually, but since the power was so low, all rods had to be moved simultaneously). They didn't do that. It was the third mistake. At the same time the operator makes the fourth mistake - he didn't issue the 'hold power' command to the control mahcinery. As a result the reactor power quickly drops down to 30 MW. Water stopped boiling in the rod shafts and xenon poisoning started. Now, the operators make the fifth mistake (or a crime, since all instructions require to shut down the reactor) - they removed all control rods from the core.

1.00 am: They managed to raise the thermal power to 200 MW (remember, the experiment had to be made at 700-1000 MW). Because of xenon poisoning they are unable to raise the power any higher.

1.03 am: The experiment starts.

1.07 am: The pump no. 8 is used as a dummy load. The system is not designed to work under such load and cavitation begins in the pumps (there is simply not enough water). They suck the water from the separator tanks and the water level dangerously drops there. A huge amount of cold water running through the reactor has lowered the steam generator to critical level. The contol machinery removed ALL control rods from the core.

1.19 am: Seeing the dangerously low water level in the separator tanks the operator increases the water supply (condensed water) and BLOCK the automatic emergency mechanism that had to shut down the reactor by the 'not enough water' and 'insufficient steam pressure' signals. This was the mistake (or crime) no. 6

1.19.30 am: The water level in separator tanks starts to raise but since so much cool water runs through the reactor core the boiling process stopped. The last automatic control rods were removed from the core. At this point the operator removes the last manual control rod from the core (mistake no. 7) thus losing the ability to control the processes that take place in the reactor. The reactor is 7 meter high and it responds well when they are moved in its middle. When the rod is farther from the center the reactor response worsens. The rods move withthe speed of 40 cm/s.

1.21.50 am: The water level in the separator tanks raised over the normal and the operator switches off several pumps.

1.22.10 am: The water level stabilizes, less water runs into the core. The boiling starts again.

1.22.30 am: The reactor control systesm are not designed to run in these conditions and only 2/3 of needed water now is pumped into the reactor, and the computer shows that the reactivity level is dangerously low. The personnel IGNORES that (another crime). All instructions tell that the reactor should be immediately stopped.

1.22.45 am: The thermal power of the reactor starts to raise slowly. The operators assumed that they'd managed to stabilize the reactor and they decided to continue with the experiment (mistake no. 8 - all rods are removed, ESS is blocked, the reactivity is very low, all automatic shut down mechanisms are also switched off).

1.23.04 am: Operators block another shutdown mechanism that is triggered when no steam is supplied to the second turbine if the first one has been previously switched off. (Turbine no. 7 has been shut down 12 hours before). This was the mistake no. 9 (and the fourth crime that night). The instruction prohibits to shut down this system for WHATEVER REASON). Simultaneously the personnel blocks the steam supply to the turbine. The experiment has started. The turbine starts to slow down, the supply voltage drops and the centrifugal pumps start to slow down as well.

The investigation later on has shown that if this last shutdown mechanism had not been turned off the catastrophe wouldn't have hapenned and the reactor would have been shut down automatically.

1.23.30 am: The centrifugal pumps have lowered their rotation speed and the water flow through the core also lowered substantially. Steam generation increased fast. The three groups of automatic control rods started to descend but they were unable to prevent the increase the thermal power buildup since there weren't enough of them. Since the steam was not supplied to the turbine its rotation slowed down and the pumps provided even less water into the core.

At this point the catastrophe was imminent.

1.23.39 am: The chief operator orders to press the shutdown button. At this command the control rods should go down at maximum speed. This should have stopped the reaction.

(see the pic below)

Here's what hapenned: At 1.5 m under each control rod (4) a so called 'displacer' is hanged (5). This is a 4.5 m aluminum cylinder filled with graphite. Its task is to moderate the neutron capture. Graphite also captures neutrons but with less speed than borom or cadmium. When the control rods are raised maximally high the lower end of the displacer is located at 1.25 above the core. In this area the water that fills the shaft is not yet boiling. When all of the rods started to move down when the emergency shutdown button was pressed the parts of the rods that contained borom and candmium have not yet reached the core, but the displacers have displaced all water from the shaft and exposed the fuel pins. The steam generation skyroketed. The steam pressure was so high that the rods could not go any further - the steam pushed them back. They only moved for 2 meters. At this point the operator switches off the electromagnets that held the rods (the rods should free fall into the reactor at this command), but the generated steam pressure held them above the core.

1.23.43 am: The thermal power started to raise up to 530 MW and higher. The last two emergency systems have been triggered, but they did exactly the same that an emergency stop button which had been pressed 3 seconds earlier.

1.23.43 am: Within fractions of a second the thermal power has increased in 100 times and continued to raise. The fuel pins overheated and the casings ruptured. The enormous pressure has pushed the water back into the supplying pipeline, then the steam destroyed the steam pipelines above.

This was a moment of the first explosion. Reactor stopped being a controlled machine.

When the pipelines were destroyed the pressure in the reactor lowered a bit and the water could flow into the core again where it started to react with the nuclear fuel, overheated graphite and zirconium. These reactions led to fast accumulation of hydrogen and CO. The gas pressure in the reactor increased so that it lifted the lid of the reactor weighting 1000 ton (!) and destroyed all the remained pipelines.

1.23.46 am: The gases from within the reactor have mixed with the air oxygen which, due to high temperature, immediately exploded.

This was the second explosion.

The reactor lid flew upwards, rotated 90 degree and dropped down. The walls and the ceiling of the reactor hall were destroyed. Some part of the graphite and reactor fuel from inside flew onto the roof of the reactor building and started about 30 fires.

The chain reaction has stopped.

1.30 The first firefighting squad was dispatched to the site.

Resume: RBMK-1000 is a very good reactor and it takes quite a bit of effort and determination to explode it.

Edited February 21, 2015 by cicatrix

[Laie]

As I recall they where about to run emergency shut down tests. They purposefully created an event to see if the system would perform the emergence shut down. It didn't. And by the time the under qualified staff realized what had happened, or rather; what didn't happen, it was too late.

Actually, it was a little more complicated. While the test(s) took longer and the longer, the reactor was operated at a very low output level; meanwhile, decay products still accumulated at a high rate because of the reactor having been fully operational until just recently. But before I keep telling it in my words, I suggest you look for "Xenon Poisoning".

My physics teacher likened it to someone pushing a stuck door, until the door very suddenly gives way and the person falls down headlong.

[Bill Phil]

Wouldn't removing the rods be the LAST thing to do in that situation?

[Laie]

Wouldn't removing the rods be the LAST thing to do in that situation?

The rods were removed sometime before; this was necessary because the xenon in the reactor caught so many neutrons that they had, so to speak, crank the throttle to full just to keep it going at all. Then came the point when all xenon was gone and the reactor did what it does when the rods are fully out.

When the rods were then inserted, their graphite tips acted as moderator, briefly boosting the power output. That was enough: the situation deteriorated within moments, the rods got stuck with their graphite tips in the middle of the pile, and the whole thing boiled over. Pressure went from still normal to quite extraordinary in nearly an instant, blowing the lid through the roof.

[Brethern]

Problem was, it WAS broken but they didn't know.

But the other reactors in the facility did operate for years after the fact. So chances are it wasn't broken.

[cicatrix]

Problem was, it WAS broken but they didn't know.

It wasn't broken.

[CptRichardson]

I wonder why the reactor exploded when it was about to shutdown?

I heard on Discovery channel that the reactor exploded when the technicians started "Emergency Shutdown" by putting back all control rods, i cannot understand, it should stop reactor why it doesn't work.

One: There was no water left because of what they did. The Chernobyl reactor had what is called a positive void coefficient, meaning that it can easily boil off its coolant and doesn't have enough to cool the rods from reacting further. Additionally, there were the graphite tips, which as stated acted like dropping a lit match into a 500 ton silo full of black powder, and they were designed to be manually lowered instead of the western 'scram' design, which slams them into place at the slightest hiccup.

[lajoswinkler]

These types of reactors (basically graphite pile ovens) are good - if you want in situ production of plutonium and you don't mess with them.

If you mess with them, they get angry, unlike other designs which by their very nature in passive and active fashion try to stop whatever weird you do to them.

Chernobyl wasn't a typical meltdown (reactor shuts down, coolant boils off, fission products' heat builds up, matrix melts, possible leak of volatile radioisotopes), but an angry bull the operators kicked in the nuts. The pile itself exploded. It is by far the worse reactor accident ever in the history of nuclear technology, and the only reason why Fukushima's reactor accidents got in the same INES level is because the scale is poorly designed, leading to a whole lot of unnecessary fear.

Chernobyl eats Fukushima for breakfast. Uncomparable accidents by any means.

[LordFerret]

Shutting a reactor down isn't like turning a key to turn a car off. Sure, there are stop-gap measures to ramp stuff down quickly, but really all those measures do is to allow a little steam to bleed off (pardon the pun). Case in point: There's a reactor not too far from where I live which is in the process of shutdown. Environmentalists won their case and the thing is now forced to shut down. I won't go into the politics of it, but thanks to them, we've essentially shot ourselves in the foot for it - the taxes paid by the facility eased an enormous burden on the public, which the public is only just now waking up to the fact that they will have to back-fill the loss (meaning: they pay more taxes). In any case, the shut-down has begun. I have an engineer friend who works at the facility; He's laughing, because, as he tells me, it will take 20 years to shut the reactor down fully ... and during that time, it will continue to generate (and thus need cooling) throughout the process.

The facility is known as Oyster Creek. Note they've been given an 'operating license extension'... the 20 years required.

[lajoswinkler]

Shutting a reactor down isn't like turning a key to turn a car off. Sure, there are stop-gap measures to ramp stuff down quickly, but really all those measures do is to allow a little steam to bleed off (pardon the pun). Case in point: There's a reactor not too far from where I live which is in the process of shutdown. Environmentalists won their case and the thing is now forced to shut down. I won't go into the politics of it, but thanks to them, we've essentially shot ourselves in the foot for it - the taxes paid by the facility eased an enormous burden on the public, which the public is only just now waking up to the fact that they will have to back-fill the loss (meaning: they pay more taxes). In any case, the shut-down has begun. I have an engineer friend who works at the facility; He's laughing, because, as he tells me, it will take 20 years to shut the reactor down fully ... and during that time, it will continue to generate (and thus need cooling) throughout the process.

The facility is known as Oyster Creek. Note they've been given an 'operating license extension'... the 20 years required.

It doesn't take 20 years. It can take a week. The pressure vessel is flooded, top is opened, fuel assemblies are lifted up and transferred by a water channel into the cooling pond. The reactor itself can be dismantled afterwards.

4-5 years after the fuel assemblies have been decaying their fission products, they can be stored in so called dry casks - large concrete barrels, because the heat output is no longer enough to cause melting.

But what you're describing is a PWR or a BWR type reactor which are like water heaters, and RBMK is like a brick oven through which channels for coolant and fuel assemblies pass. Radically different designs.

My condolences for the plant shutdown. The stupid have won and now you're all gonna pay for it. Pun intended.

[cpast]

Shutting a reactor down isn't like turning a key to turn a car off. Sure, there are stop-gap measures to ramp stuff down quickly, but really all those measures do is to allow a little steam to bleed off (pardon the pun). Case in point: There's a reactor not too far from where I live which is in the process of shutdown. Environmentalists won their case and the thing is now forced to shut down. I won't go into the politics of it, but thanks to them, we've essentially shot ourselves in the foot for it - the taxes paid by the facility eased an enormous burden on the public, which the public is only just now waking up to the fact that they will have to back-fill the loss (meaning: they pay more taxes). In any case, the shut-down has begun. I have an engineer friend who works at the facility; He's laughing, because, as he tells me, it will take 20 years to shut the reactor down fully ... and during that time, it will continue to generate (and thus need cooling) throughout the process.

The facility is known as Oyster Creek. Note they've been given an 'operating license extension'... the 20 years required.

There's a difference between "shut down" meaning "decommission a plant" and "shut down" meaning "stop the nuclear reaction." The latter isn't like turning a key, because it's even easier - you press one button, the control rods are immediately fully inserted into the reactor, and if it is a decent reactor design (not the RBMK, with its graphite-tipped control rods) your reaction has stopped in seconds. That doesn't mean you're home free, as fuel rods are still decaying and releasing heat, but the reaction is shut down extremely quickly, and decay heat is much less than the heat caused by the reaction. Even that doesn't take too long to be no longer a problem assuming your cooling systems were working properly enough (and most reactors have additional emergency cooling systems that draw water from outsidefor exactly this reason). Decommissioning takes time not because it takes time to stop the reaction, but because it takes time to deal with waste disposal regulation, bring enough capacity online to offset the loss in generation, and wind down everything that's going on at the plant; however, decommissioning is irrelevant here, and the relevant kind of shutdown happens initially in seconds, and then all that's left is decay heat (there's no chain reaction anymore).

[sgt_flyer]

Nevertheless, you still need active cooling for quite some time after a scram to deal with the decay heat (most reactors are maintained below 600°C during normal operation - to preserve the rods structural integrity.

A scrammed reactor thermal output generally drop below 7% of it's normal output within seconds after a scram - (not enough to sustain steam generation - so you can't use this power to drive coolant pumps - you need external power)

As a reactor's envellope is designed to limit thermal losses (so they can transfer most of their heat to the coolant) without coolant flow, decay heat would make the core temperature rise up to damaging points (even if it takes some time, without extractîg this decay heat, the temperature could rise to levels capable of damaging the reactor's components))

So scram systems still require careful monitoring for a few weeks (residual decay heat thermal power is still around 0.5% after 1 week)

Now, in case of a problem with the scram system (ex, some rods broke and got stuck, or the reactivity of the core became too high to be suppressed by the rods) modern / modernised designs can also inject neutron poisonning agents directly into the core to stop the reaction (though that a bit more time than the 4 seconds control rods take to be fully inserted during a scram)

Edited February 23, 2015 by sgt_flyer

[peadar1987]

But the other reactors in the facility did operate for years after the fact. So chances are it wasn't broken.

It wasn't broken.

Not broken as such, but a positive void coefficient is a profoundly stupid thing to have, and completely avoidable.

A bit like the air-cooled piles that caused the Windscale fire. They work fine when the reactor is at nominal conditions, but become very dangerous when you have a certain set of problems. In the RBMK's case, it's that if you have any boiling of coolant, you have a sudden and massive problem. With PIPPA, if your reactor goes on fire, increasing coolant flow fans the flames, and makes your problem worse.

[magnemoe]

Nevertheless, you still need active cooling for quite some time after a scram to deal with the decay heat (most reactors are maintained below 600°C during normal operation - to preserve the rods structural integrity.

A scrammed reactor thermal output generally drop below 7% of it's normal output within seconds after a scram - (not enough to sustain steam generation - so you can't use this power to drive coolant pumps - you need external power)

As a reactor's envellope is designed to limit thermal losses (so they can transfer most of their heat to the coolant) without coolant flow, decay heat would make the core temperature rise up to damaging points (even if it takes some time, without extractîg this decay heat, the temperature could rise to levels capable of damaging the reactor's components))

So scram systems still require careful monitoring for a few weeks (residual decay heat thermal power is still around 0.5% after 1 week)

Now, in case of a problem with the scram system (ex, some rods broke and got stuck, or the reactivity of the core became too high to be suppressed by the rods) modern / modernised designs can also inject neutron poisonning agents directly into the core to stop the reaction (though that a bit more time than the 4 seconds control rods take to be fully inserted during a scram)

Yes, as I understand the wast heat was the problem in Fukushima, because the power and generators was down they was unable to circulate water for cooling.

Now the problem is limited as long as you have water as you can went steam, and add more water, this is not something you want to do as the steam contains radioactive particles.

Smaller reactors like the ones used in subs are far easier to shut down fast as they are small and don't contain much stored heat.

[lajoswinkler]

A scrammed reactor thermal output generally drop below 7% of it's normal output within seconds after a scram - (not enough to sustain steam generation - so you can't use this power to drive coolant pumps - you need external power)

It is possible to use this energy to drive coolant pumps. That's what BWR reactors at Fukushima used when deprived of external electrical energy and diesel generators. However it can't be used for long. Few hours perhaps.

Now, in case of a problem with the scram system (ex, some rods broke and got stuck, or the reactivity of the core became too high to be suppressed by the rods) modern / modernised designs can also inject neutron poisonning agents directly into the core to stop the reaction (though that a bit more time than the 4 seconds control rods take to be fully inserted during a scram)

4 seconds is too long for modern PWRs. SCRAM actually occurs very fast, often by almost a free fall of the rods. In Chernobyl, however, the graphite pile SCRAM takes a lot longer, perhaps ten seconds or more. That, combined with their graphite tips, ensured the heat to build up tremendously.

Smaller reactors like the ones used in subs are far easier to shut down fast as they are small and don't contain much stored heat.

Don't be so sure about that. Submarine reactors contain extremely enriched uranium so the density of fission products can be even higher.

It's not the leftover heat that needs to be conducted away. That would take a few minutes. It's the fission products which decay produces heat.