How Does an RBMK Reactor Explode

[Dartz]

Or the weird things that you come across when doing your research for a fic exactly thirty people have read.

----Pt 1---------

Everybody knows about Chernobyl. Everybody knows either one of two things. The accident was caused by gross incompetence. Or the accident was caused by a gross design error. Or the accident was caused by both. Yes, that's three things - but this isn't the Spanish Inquisition.

Anyway, to figure out why an RBMK reactor can explode, first we have to understand how a nuclear fission reactor works in the first place. It's not actually that complicated. A uranium atom splits - small particles called neutrons fly out from the atomic shrapnel and they each find another uranium atom, collide with it and break it apart too, letting more neutrons find more Uranium atoms in a chain reaction. And out of each split we get energy - once split, the fragments snap apart as the energy holding them together is released - like cutting an elastic band. These fragments explode away, run into the atoms next to them at high speed and all that speed is turned into heat. Heat is used to boil water. This steam turns a turbine. This turbine turns an electric generator and, eventually you get electric power out of it.

Now, it's not quite that simple. Because big, heavy atoms like Uranium come in multiple different versions - called Isotopes - sort of like different models of the same car. They're all Uranium, but they're all slightly different at the same time. The most well known, are Uranium 235, and Uranium 238. U238 is by far and way the most common - on the order of 99% or more of Uranium on Earth is U238. It's a big, heavy fat atom that doesn't really like to split - it takes more energy to break it apart than it gives up when it breaks. On the other hand, U235 is much happier to break apart - unfortunately, it's about as rare as common sense. So, out of a large block of Uranium, only a small amount can actually split. Only the U-235 is actually fissile.

Out of the ground, less than 1% of your Uranium fuel is actually fuel.

More than that, in order to have a chain reaction, the neutrons released by one atom splitting have to be able to cause further atoms to split. Otherwise everything just runs down. It turns out, that the neutrons released by fission are extremely fast - too fast in fact to actually find their way to the next atom and split it. For a fast neutron, the probability that it will cause another fission is really low. Either you need a lot more U-235 around it to get the probability up, or you need to find someway of slowing it down. If you slow it down - the probability of fission goes way up. This is called a Moderator. 

For most reactors, this moderator is water. Ordinary - albeit extremely pure - water. Water is a good moderator, but it has one slight drawback - it absorbs neutrons. Absorbed neutrons do not get to go and make another fission happen - they just turn the water radioactive. Using water to moderate a reactor absorbs so many neutrons, that the quantity of U235 in the reactor fuel has to be increased. This process is called enrichment. It's expensive and energy intensive - and it turns out if you enrich Uranium to about 80% U235 it can be used to make an atomic bomb.

Which, naturally, is why Isreal got so pissy about Iran having the ability to enrich Uranium. The difference between safe reactor fuel and weapons-grade bomb fuel, is time in the centrifugal oven to bake.

Now, doesn't it seem eminently sensible to find a way to build a reactor that will be happy on regular, non-enriched Uranium?

Canada did it with the CANDU reactor. Instead of regular water, a CANDU reactor uses 'heavy' water. 'Heavy' water is like ordinary 'Light' water, except the Hydrogen atom that's the H in H2O is a little different. It has one neutron and one proton, rather than just a single proton. It absorbs less neutrons, which means more neutrons are free to cause more fission. In fact, Heavy Water is so effective that CANDU reactor doesn't need enriched fuel. The Nazis appearred to have tried a similar appoach, and the destruction of their Norwegian Heavy Water factories stalled their weapons program.  On the other hand, heavy water - while common - is still fairly expensive.

What if you could build a reactor that ran on natural uranium, that didn't need heavy water?

The first Hanford Reactors, built in the United States for the Manhattan project used Graphite - pencil 'lead' - as a moderator. They also used regular, light water as a coolant to keep the reactor from melting down with its own heat. This hot water was dumped merrily into the local river. Of course, somebody had worked out that if the reactor lost cooling water, it would very quickly begin to run out of control so the Hanford reactors were built miles from anywhere inhabited. They never generated a watt of electricity-  but they did create the Plutonium for your nuclear weapons.

The British Government, aware of this risk, built the Windscale Piles to be Air cooled - with giant fans blowing air over hot graphite and metal. These then went and caught fire. In the end, the solution was to use graphite and an inert gas, like carbon dioxide, to cool the core.

This was still extremely expensive.

So, the Soviet Union looked at this and though; We can build a graphite moderated reactor, cool it with regular light water and so long as we don't fuck up, we'll have a shitton of free energy.

We know the result already. But that's being a little bit churlish. These people weren't fools.

This why an RBMK reactor is different to every other reactor built anywhere else in the world. The majority of modern nuclear reactors are basically big pressurised kettles, filled mostly with pressurised water. This water either boils in the kettle-  or it is under such a high pressure that it remains liquid and is used to boil water in a second circuit. In an RBMK reactor, the fuel is contained in more than 1600 vertical channels cut through the graphite moderator. Inside these channels, light water flows as coolant. It enters a pair of drums high above the reactor where any steam bubbles in the water seperate and are drawn off to the turbine to generate power - while the liquid water is recirculated, being mixed with cooler water coming back from the turbine. Interspersed within these channels are more than 200 others - each contaning a control rod. These control rods are also cooled by liquid water - but at a much lower temperature.

The main steam circuit on an RBMK reactor operates at somewhere around 270 degrees and 60-odd Bar of pressure. The control rod circuit operated at 70 degrees.

The light water in the channels still absorbs neutrons sure - but because there's so much less of it, the reactor will still run on natural unenriched Uranium. It also means that, since each fuel assembly has its own individual channel, it can be removed, moved and replaced without shutting down the reactor. This is a feature few other reactors have - most reguire a shutdown to open the reactor vessels to refuel. This is goof for fuel economy, good for efficiency, and good for creating weapons grade plutonium on the sly if you were that way inclined.

The people who designed the Chernobyl reactor weren't fools. There were compelling reasons for making the decisions they did. It made a big, powerful reactor cheap and easy to build, while improving the reactor's fuel economy and general uptime. And you could potentially fuel a weapons program with it.

The one clear drawback with this should be obvious. When the cooling water boils, its replaced by a bubble of steam. This steam absorbs far fewer neutrons than liquid water - meaning more neutrons availably for fission, which means more fission, more heat, more steam, more neutrons, more fission.

This is called a Positive Void Coeeficient. It is an example of positive feedback - an action creates a stronger action in the same direction - like setting a ball rolling at the top of a hill, it's only going to start rolling faster as it gets further down. Engineers love positive feedback. It usually results in entertaining explosions.

This is potentially problem for an RBMK reactor specifically because the water does not act as a moderator - more correctly, it provides little to no moderation. On a conventional reactor, the water also provides moderation - so if it is boiled away by heat, the moderation in the reactor reduces, neutrons get faster, the probability of fissions gets lower, less fission happens and the problem self-corrects. On an RBMK reactor - even if all the water in the core somehow is removed, the moderator is still present in the form of the graphite to keep the reactor going.

It would be dangerous to have a reactor which behaved like this. The engineers who designed Chernobyl were, of course, aware of this. But real physics is not that simple. As the fuel heats up and gets more energetic, it responds to neutrons differently. The hotter it gets, the harder it is for a neutron to cause a fission. Hotter fuel is less likely to fission - so an increase in power will actually reduce the ability of the fuel to fission and create power - in effect an automatic brake provided free by simple physics. This is called Negative feedback, and is basically the same as you feeling a tug in the steering wheel of your car, and steering the other direction to compensate.

Positive feedback acts to destabilise. Negative feedback acts to stabilise.

But - if the negative feedback from the fuel heating is stronger than the positive feedback from the steam boiling, the reactor's power level will self stabilise and everything will be fine.

For a large part of the reactor's life this was true. But this changed as the reactor got older, and more and more fuel was used up. the negative reactivity from the fuel heating up, no longer counterbalanced the positive reactivity from the steam boiling.

This balance changed with old fuel. Where a reactor had been running for several years- the fuel gets more and more depleted. In addition, more and more poisons are added, each of which absorbs neutrons differently or introduces additional hazards into the reactor. The reactor in Chernobyl had been running for about three years. After this time, changes in the fuel meant that the negative feedback from the fuel heating was no longer enough to counterbalance the effect of the void coefficient. The reactor operated in a positive feedback loop.

An increase in power, left unchecked would create a further increase in power. Only the reactors control rods kept the reactor under control, The majority of these control rods inserted from the top of the reactor. Some inserted from the bottom. They served to absorb any excess neutrons in the core and act as the final brake on the reactor to keep it in control, to keep the reactor critical.

Fission reactors are at their happiest when they're critical. A critical reactor is a reactor running in a balanced steady state at a constant power. It's the desired state of being. Every fission is creating one further fission and that's it. A reactor that is supercritical, is a reactor that's accelerating - each fission creates more than one further fission. A reactor that is subcritical, is a reactor that's decelerating - each fission creates less than one further fission.

The reactor is moved from state to state by adding or removing reactivity. Reactivity is like the throttle and brake on the reactor. It's not really the current power level - it's sort of the potential change in power level. Positive reactivty means fission is more likely to happen than it is now - which will cause an increase in power. Negative reactivity, means making fission less likely to happen - which will cause a decrease in power.

In theory, there is no limit to the amount of negative reactivity you can add - all it does is stop the reactor faster. But there is a limit to the positive reactivity.

When an atom fissions, the vast majority of neutrons are released instantaneously - at the moment of fission. The neutrons fly away, get themselves moderated, and in the space of microseconds find more atoms to collide with and split. The scientists of the Manhattan project called this a 'Shake' and it is an extremely short interval of time - from nanoseconds to microseconds. These are called Prompt neutrons.

Fission with prompt neutrons happens so quickly, that there is little to no mechanical process capable of controlling and regulating it. If the universe had been created in such a way that there were only prompt neutrons - controllable fission power would likely be impossible.

But, luckily, a very small fraction of the neutrons released by a fission event are delayed - they happens seconds, to minutes later. It is this small fraction of delayed neutrons which enables every nuclear fission reactor to be controlled. It is possible therefore, to have a reactor which is critical on the combination of the Prompt, and the Delayed neutrons. In fact, this is how things normally are. Even a supercritical reactor, will take seconds to minutes, to change power output. There's time there for the process machinery of the reactor to respond to changes and stabilise.

But, if the reactor is pushed to the point that it is capable of achieving criticality on the Prompt neutrons alone - before any of delayed neutrons are emitted - then things get interesting. Instead of a power increase that happens on the order of seconds to minutes - now the only limiting factor the the reactor's power increase is the time it takes for a neutron to find an atom to fission, and however long the reactor manages to stay together in a critical assembly against the energies that are being very rapidly liberated. A reactor which has going prompt critical, has become, in effect, a really, really shit nuclear bomb. The big difference being that bombs take advantage of physics, inertia and a dozen other things to keep the reaction going that few nanoseconds longer it takes to go from 5 tons of TNT, to 15 Kilotons of TNT.

Scientists at the Manhattan project, for whatever reason, called this interval a 'Dollar' of reactivity. Once you get a reactor past that point - unless it's a type specifically designed to go there and self recover - the reactor will be destroyed. Importantly, this does not have to happen within the entire reactor - it can be limited to a very small part of the core where conditions align like the stars.

At the Chernobyl reactor, Reactivity was added by fresh fuel, by removing control rods and by boiling water to make steam. Reactivity was removed by increasing water flow, adding control rods, and by another factor.

The shrapnel left over from fission creates what's known as 'fission products'. Most of these are hideously radioactive. Many of these are effective at absorbing neutrons. Absorbed neutrons reduce reactivity, which has to be compensated for either by withdrawing control rods, or by removing the used fuel and replacing it with fresh fuel. One of the most effective neutron absorbers is an isotope of Xenon, called Xenon-135.

It starts to appearr about 6 hours after the fission events that effectively 'created' it. The amount of it that's created, is in direct proportion to the quantity of fission that happened six hours ago. So if a reactor is run at full power for a long time, and then throttled down, Xenon will continue to appearr according to that fuel burned six hours previously. It's a bit like the exhaust from your cars engine magically taking longer to form after the combustion in the cylinders. Normally, with the reactor in a steady-state, Xenon is created as quickly as it is consumed - the physics balances out. It can make it very difficult to increase or reduce power - if power is reduced too quickly, and the Xenon continues to build, the reactor might even be stalled by it.

It can also mean that, if the fuel in the reactor has been burned for a long time - there may not be sufficient reactivity in the remaining fuel to overcome this Xenon pit - the reactor is stalled and effectively impossible to start.

This is important. Because after a few hours more, the Xenon goes away. More than that, Xenon which absorbs a neutron also 'goes away' - it's no longer Xenon-135 and it's massive ability ot hoover up neutrons is suddenly gone.

Keeping all of these positive and negative reactivities in balance is the job of the Senior Reactor Engineer, who manipulates the reactor core's systems and control rods to achieve the required stable power output. The Engineer has only so much control as the rods will give them.

Finally, there is the concept of the Reactivity Margin. And that's basically the count of control rods left inside the reactor, which are required to maintain criticality. The higher the reactivity margin, the more control rods remain in the core and the more reactivity can be added to the core. A reactor with fresh fuel will have a very high reactivity margin. A reactor with old fuel, or with xenon poisoning, will have a low reactivity margin. Other factors like Xenon can push the reactivity margin down. It may seem that a low reactivity margin might be 'safer', because now there's less reactivity that can be added by the control rods (which are already out of the reactor at low margin). At a high reactivity margin, the control rods are inside the reactor. More of them can be withdrawn, to push the reactor  further into the supercritical - more positive reactivity can be added to the reactor.

the Chernobyl reactor was happy around about 30 Rods of reactivity. At this point, fresh fuel was being added frequenctly enough to keep the reactor stable, but not so frequently as to be uneconomical. The official limit, was somewhere around 15 Rods of reactivity. This wasn't thought to be a safety limit - it was simply economic. The reactor was far more likely to stall at low reactivity margins, resulting in downtime and lost energy production - or surprise blackouts.

But, below 30 rods of reactivity, another insidious effect began to occur.

The Control rods of an RBMK reactor are manufactured from boron. Boron absorbes neutrons, which reduces power. The deeper they go into the reactor - the more neutrons are absorbed - the slower the reactor goes. They can also be moved independently of each other - which changes where and how power is produced throughout the reactor, to compensate and balance for old and fresh fuels and how they're distributed through the reactor.

But, at the tip of the control rod on a telescoping extension, is a single slug of graphite and the end of a telescopic rod. The graphite tip of the control rod acted as a displacer. Its purpose was to push water out of the control rod channel, to remove it and its neutron absorbtion effect after the rod was withdrawn. In effect, instead of giving the control rod an action of -1,0 - they are something like -1,+1. They graphite displacer gives the control rod a stronger control action. It makes it more powerful by adding reactivity after the rod is withdrawn.

This led to an interesting effect.

If a number of rods in the same area of the reactor were inserted at the same time, and were in the same vertical position as they moved, a small amount of postivie reactivity could be momentarily added to the bottom of the reactor. This would cause an uptick in power for a few seconds before the boron control rod travelled the entire height of the core and finally quenched the reaction.

This was not thought to be much of a concern - power changes in the reactor after all, take longer than it takes for the rod to travel. It was just something the RBMK reactor did. Methods to mitigate it had been known and discussed for a decade prior Chernobyl Disaster, but were not seen as too much of a big deal. An RBMK reactor cannot explode, after all.

It would also reflect badly on the director of the Kurchatov institude if the reactor he had overseen were found to have a potentially fatal flaw. It was quietly buried in the documentation.

We do not yet know how an RBMK reactor explodes. But we know what we need to know

[classicdrogn]

The reactor in Chernobyl had been running for about three years and over this time - and the negative feedback from the fuel heating was no longer enough to counterbalance the effect of the void coefficient.

looks like you might have an editing artifact around the dash? At least, it would make more sense with a couple less words and/or no dash.

Another one a bit later is cutting things off:

More of them can be withdrawn, to push the reactor

the Chernobyl reactor was happy around about 30 Rods of reactivity.

The very last line is also missing at least a period.

I also found a typo:

This is important. Because after a few hours more, the Xenon goes away. More than that, Xenon which absorbs a neutron also 'goes away' - it's no longer Xenon-135 and it's massive ability ot hover up neutrons is suddenly gone.

ot hover -> to Hoover

This may not be the sort of comment you were looking for, but since you mentioned it being for your fic I figured a little help with polishing wouldn't be too out of line. It's an interesting piece, at any rate.

edit: Also, either an extra or missing carriage return at

after the combustion in the cylinders.
Normally, with the reactor in a steady-state,

[Black Aeronaut]

Dartz, this is utterly fascinating. Will there be more, or is this all of it?

[Star Ranger4]

I have only one question, Dartz... Did you start digging into all of this as a result of research for "The Anatiomy of a Nuclear Oops" or was Anatomy a way of working out your own take on what your research had found?

[Dartz]

A little of both - but to have a story about how a nuclear reactor explodes, it's important to know how a nuclear reactor explodes. These things don't happen because of one person's decision - but instead a culmination of factors that make it impossible not to have an accident. There'll be a second part definitely - maybe a third depending on how it goes.

[Star Ranger4]

*Nods*

True, the best stories are a blend of knowledge and questions.

[Dartz]

---

The Chernobyl Nuclear Power Plant, and the city of Pripyat beside it, had been the brainchild of Viktor Bryukhanov. Bryukhanov had taken the city and its power plant form a paper concept, to a living, thriving thing. By 1986 Pripyat had become a model for the best in Soviet life. The shelves were well stocked, the apartments comfortable and the amenities accessible. There is a hotel, a swimming pool, multiple schools and a 'Palace of Culture', where residents can indulge in their hobbies.

Pripyat is what the Soviet Union wishes it was.

By 1986, four reactors had been completed at Chernobyl, with two more under construction. Pripyat and Chernobyl are planned to be the largest nuclear power complex in the entire world. Chernobyl reactor 4 was finished in 1983, a few weeks ahead of schedule, earning Bryukhanov an award from the government. On the other hand, there have been some corners cut. Fireproof materials for the roof were not available, so conventional materials were used. Some safety tests were failed when initially attempted - but the deadline was looming, so the reactor was put into service anyway.

The tests however, still had to be completed.

Over the next three years the test is attempted a further two times - failing each time. Finally, the fourth time, was April 26th, 1986. The test was succesful, and the reactor operators shut down the reactor.

Then it exploded.

The test was of a safety-critical system, but was not considered safety-crticial in and of itself. It was not really thought of as a test of the nuclear systems. The test program had been drawn up by an electrical engineer and from there it made its way to the desk of Nikolai Fomin, the station Chief Engineer. Fomin - only just back from work after a serious car accident - reviewed it, and not seeing it as potentially hazardous to the reactor, signed off on the proceedure. The proceedure was less than clear about what was required from the reactor.

The test, was an attempt to answer a question. If offsite power was lost due to accident or grid failure, and no other sources of power were available to the reactor, how would the electrically driven cooling pumps be operated? These pumps had to be kept running - they could not be permitted to stop. The reactor did have emergency diesel generators, but these took a full minute to start up. The pumps themselves each required 30MW of power. They can each move 12 Megalitres of water per hour. And there are 8 of them - of which 6 were normally in use.

If, at full power, the reactor were to loose cooling water and the control rods were to somehow fail to fully insert, it was calculated that it would take just 40 seconds for the fuel channels in the core would begin to buckle and warp. They would loose all resiliance. A breach in just three fuel channels out of the more than 1600 in the reactor could introduce enough steam pressure into the reactor chamber to lift the lid off the reactor - like the lid on a boiling pot hopping. Everything attached to the lid - the fuel channels, the control rods, would then be pulled out of position, potentially to the point where they would jam in place, leaving the reactor endlessly accelerating itself while all the remaining water boiled off.

Suffice to say, it was necessary to find another source of sufficient energy to keep the pumps running.

This was the reactors own turbine generator. At the first moment of an accident, the throttle valve to the turbine would close, but sheer force of inertia would keep the turbine and its generator spinning. Thousands of tons of steel rotating at 2500rpm may have just enough remaining energy to keep the reactor pumps circulating. Doing this, involved careful switching and regulation of the generator, both to keep the turbine from slowing too quickly, and to keep enough power flowing to run the pumps.

This test is important enough that Deputy-Chief engineer Anatoly Dyatlov chooses to supervise it personally. Dyatlov was not well liked - by all accounts he was an abrasive man, and very strict about professionalism. He was the teacher who would not let you have your pudding, without eating your meat. But at the same time, Dyatlov was fair - if you ate your meat you got your pudding. It didn't matter - the appearrance of him in the room always ratcheted the tension up. Everyone was afraid of making a mistake under his watch because he would be merciless. Nothing would be overlooked.

Also in the control room of Reactor 04 that night was the team operating the reactor. There are three main areas of responsibility. The Reactor engineer is responsible solely for maintaining the balance of reactivity in the core, keeping the core alive and operating at a stable power point. A pump engineer, is solely responsible for maintaining the flow of water through the reactor, and balancing it against the power produced to keep steam flowing to the turbine. The turbine engineer, monitors the turbine, condensers and generator, to make the best use ot steam generated.

The groundwork for the test is laid during the previous dayshift. The emergency core cooling system is disabled - a process which takes hours manually cranking valves closed - to prevent it from being triggered accidentally by the test. The dayshift operators also wind the power of the reactor down from its normal 3000 MW, down to 1500MW. The reactor operators measure core power based on the heat energy produced inside the core - the thermal power - a figure which is typically three times higher than the electrical energy output.

The test proceedure calls for the reactor to be at a self-sustaining power level. This is set around and about 700Mw thermal. Before they can do this, the day-shift are told to maintain power by the local grid controller.

It isn't until after midnight, and a shift change, that Chernobyl is given permission to reduce power. The dayshift has long gone home, and the night shift has taken over.

Aleksandr Akimov, the Shift supervisor, and Leonid Toptunov - who at age 25 was a Senior reactor engineer - will now be responsible for the reactor. Controlling the pumps, will be Boris Solyarchuk, overseen by Yuriy Tregub who stuck around from the dayshift to see what happened. The Turbine and its generator, are to be the responsibility of Igor Kirchenbaum. Observing, are two trainees - Viktor Proskuryarkov and Aleksandr Kudryavtsev. Elsewhere in the plant, a team from the turbine's manufacturer will take the opportunity to run vibration tests on the turbines. There are other hangers on in the control room, watching, curious about the test.

Just after midnight, Dyatlov orders the reactor's power be brought down to 700.

Toptunov begins to work the power down. As he winds the reactor down, Xenon is still being produced according to the high power levels from 6 hours previously. Toptunov has to precariously balance a reactor that suddenly seems to be trying harder and harder to shut itself down, as more and more negative reactivity is added to the core by the Xenon. It gets harder and hard to reduce the power safely.

Under the gaze of Dyatlov, a mistake is made. To try and stabilise Power, a setting is switched. The setting has the opposite effect of that intended.

The chain reaction collapses. The reactor stalls.

With power below 30MW, the test is fucked. There isn't even enough energy to spin up the turbine. Proceedure calls for the reactor to be shut down. This would allow any Xenon in the reactor to decay, and make restarting it easier. Dyatlov overrides proceedure, commanding Power be raised immediately. This is not necessarily unsafe, provided the reactivity margins of the reactor aren't compromosied.

According to some, Akimov objects to Dyatlov. There is a discussion between the pair. Dyatlov overrules him. What is certain, is that Akimov restarts the reactor.

But the reactor doesn't want to start.

The Xenon poisoning is so strong, the reactor is effectively completely poisoned out.

Akimov and Toptunov withdraw control rods. They withdraw almost all the control rods they can but still the reactor does not want to start. They disconnect some of the emergency rods from the computer control, and withdraw them from the core. Of the two hundred and eleve control rods, over two hundred are taken clear out of the reactor. They are all at the exact same, zero, position. The graphite displacers at the tip of the rod are now dead centre in the reactor, with a meter either side of the displacer being filled with liquid water.

The power level in the reactor grinds up to 200MW.

It's barely enough to spin up the turbine to the required speed, but the reactor is self-sufficient.

Dyatlov is satisfied. It's enough to complete the test.

On the other hand, the low power levels are creating problems for Stolyarchuk running the pumps. Water is moving too fast through the reactor to boil. Water levels in the steam seperators are getting too high - risking liquid water being ingested into the steam turbine. This would destroy the turbine. An automatic emergency shutdown is triggered, threatening to terminate the chain reaction and finish the test again.

Dyatlov orders the emergency shutdown to be overriden. This is not a violation of proceedure. It's permitted to disable the automatic emergency shutdown when the reactor is at low power levels.

The low power level still causes trouble. In order to make enough steam to keep the turbine spinning, hot water is being recirculated back to the reactor from the steam seperators, faster than it can be cooled. It is hot enough to begin to flash to steam inside the circulating pumps. Still, Stolyarchuk is able to stabilise flow and keep steam flowing to the turbine. The reactor is now stabilised - the difficult part is over.

At 1:40:03, Dyatlov orders the test be begun.

Kirchenbaum closes the throttle valve to the turbine. The turbine begins to coast. The generator is disconnected, and switched over to directly supply half of the cooling pumps. For safety reasons, half of the pumps are left running at full power using grid electricity. Normally, the reactor would have shut itself down as soon as the throttle valve closed. But the automatic shutdown has been disabled. The reactor continues to operate at 200MW of power.

As the generator slows, so do the pumps. Less and less water moves through the reactor. More and more steam is generated. The positive void coefficient begins to take effect The reactor control system automatically compensates for this. Power level remains constant throughout the test.

For forty seconds, everything proceeds as normal. Everything looks normal.

The test is succesful. The stalling generator is able to maintain the pumps for long enough for the generators to start.

There is no moment of forboding. There is no sense of impending doom.

With the test almost completed, Akimov calmly chooses to shut down the reactor - instructing Toptunov to smother it. There is a button on the reactor operator's control panel, hidden behind a wax-sealed guard labelled AZ-5. AZ-5, in this case, being translated as Emergency Protection System 5. The button immediately forces all control rods to be inserted into the reactor at once.

This should terminate the reaction.

Toptunov puts his finger on the button and holds it. It has to be held in place, or else the rods will stop moving.

The control rods begin to move into the reactor core, sliding down their channels. The graphite displacers push forward, pushing water out of the reactor ahead of them. This water, which aborbs neutrons, is replaced by graphite, which moderates neutrons. Reactivity is added to the bottom of reactor, even as it is being removed from the top.

If multiple control rods are moving together, enough reactivity could be added to momentarily cause an overall increase in power.

At this moment, a two hundred control rods are moving in unison, pushing reactivity down into the bottom of the core. More and more water is being converted to steam, creating voids which further increase reactivty.

So long as the reactor stays within the delayed-critical regime, this is not a problem. It will take seconds for the control rods to push through this regime - not long enough for the power to even start to run away with itself. This is what the designers expect will happen. This is what has always happened.

This is not what happens.

Power begins to rise - it rises far faster than it should. Steam forms. Reactivity increases. Heat increases. More steam forms. In a small section of the core - either through a unique combination of fresher fuel, or slightly less poisoning, or a large steam void - the reactivity added by the control rod tips passes that one dollar mark for one brief instant. It becomes Prompt Critical. Instead of taking seconds or minutes to increase, the reaction is now only limited by the time it takes for each neutron to find the next atom. Fuel elements shatter.

In one moment, the reactor power is in the hundreds of megawatss. The next, it exceeds 22000MW - eight times it's normal operating limit. It is drawn by the datalogger as a vertical line on the graph, running straight off the top. For one brief instant, Chernobyl Reactor 4 has become a nuclear bomb.

Water in the reactor surrounding this point is flash boiled. The steel structure of the reactor itself is melted and boiled by the intense release of energy. Above this pocket of rapidly expanding steam is a slug of liquid water. As far as the rest of the reactor is concerned, nothing unusual is happening - it is currently shutting down.

This changes in an instant. This bubble of steam rapidly expands, ramming the remaining water in the reactor out of its way. The fuel rod channelsburst open, flooding the core chamber with high pressure steam. In moments, the reactor chamber overpressurises, blasting the reactor lid clean off the top of the reactor, tearing every single control rod and fuel rod channel free of the reactor stack. The lid drops back down onto its side, a tangle of control and fuel rods trailing behind it like some hideous medusa. The bottom plate of the reactor blasts downwards to the foundations, bursting every single pipe beneath. Any remaining water in the core immediately flash-boils to steam.

The explosion rattles the men in the control room. But they have no idea what's really happened - there is no 'Core exploded' indicator light. Stolyarchuk sees this on his control panel as a water hammer. Trugub sees every single pressure relief valve at once open - something that should theoretically be impossible. The datalogger records the moment as a surge in water level in the steam drums, and as a momentary push back against the force of the main circulating pumps.

For a few seconds, these pumps continue to run normally, picking up speed. There is still water enough in the remains of the cooling circuit to keep them supplied, even if they are now pumping water into an open reactor pit.

All of this water pours across superheated steel and graphite. It doesn't even boil - the sheer heat of the graphite cracks it apart, splitting water into oxygen and hydrogen. A spark from any one of a hundred broken cables lights this off in one massive explosion - more powerful than the first

The second explosion blows the building clean open, collapsing one of the pump rooms, killing technician Valery Khodemchuk instantly. The core of the reactor is ejected to the sky, fragments of burning graphite lighting spot fires across the remaining roof of the building. Radioactive fuel rods fall onto the turbine hall below. The remnants of the graphite in the reactor ignite, burning like the devils own barbeque, spewing hot radiation high into the atmosphere.

The radiation from the reactor core is so powerful, it is ionising the air itself splitting the molecules of the atmosphere apart. A bright laser-beam of blue light reaches like a searchlight to the heavens, marking the birth of a new and terrible Godzilla.

That is how an RBMK reactor explodes.

--

[classicdrogn]

An article about a bit of the following activity, which has a current-ish picture of one of the control rooms:
https://www.businessinsider.com/chernoby...ion-2016-4
It reminds me of a Star Trek bridge in several ways.

only one editing bobble stuck out to me this time, right around "reactivity was added to the bottom of the reactor."

And taken as a whole... this, gentlemen, is why we do not override safety systems, and when one must be overridden for the sake of a test we especially do not keep disabling more of them. Even then it would have been fine if they'd just put the control rods back in in smaller and more separated groups, rather than hitting the button to do them all at once.

[Black Aeronaut]

Man, and it all happened so quickly.  The stuff that I've read all made it seem like there was this slow and steady buildup where they knew they were screwed but were trying to regain control.

Now I know differently - that they thought everything was fine and honestly had no reason to suspect something might be wrong... and then it suddenly blows up in their faces.

No wonder why there was so much confusion.  Not even the people in charge knew for sure what had just happened, let alone why.

It really is a shame.  If not for this one rather esoteric safety flaw, the RBMK Reactor seems like the platonic ideal of fission reactors.  Powerful, runs off of a fuel that is cheap and plentiful, and can breed even more powerful fuel during normal operation (if, ya know, you were up to running a reactor off of weapons-grade fuel...  Which the US Navy does, IIRC - this is how they make such powerful reactors in that small size).

Annnnnd I just checked up on the status of the remaining RBMK reactors in Russia.  Most of them are - surprise surprise - still operating now.  Of course, there's been some rather extensive refits for safety reasons, and it seems like those will be phased out midway this century in favor of the new RBMK-derived design, the MKER, which incorporates more redundancies and passive safety systems.

We can't do anything like that here in the US due to a combination of NIMBY and corporate enterprise not seeing any profit in it anymore.  Though I'll leave that particular topic for the Politics forum.

[classicdrogn]

Well, in this particular case the NIMBYs have a pretty glaring example of why they wouldn't want one in their back yard, and all the lecturing and info-graphics and professional assurances in the world aren't going to reassure them between the literal and figurative fallout of Chernobyl and the more recent Fukushima incident. Add the coal lobby to fund opposition (never mind that burning coal is one of the major things that releases radiation into the atmosphere, along with sulfates and so on) and the pushback gets severe.

[robkelk]

only one editing bobble stuck out to me this time, right around "reactivity was added to the bottom of the reactor."

How is that a bobble?

[classicdrogn]

How is that a bobble?

The actual line trails off with a comma and an incomplete clause, not a period.

[hazard]

It really is a shame.  If not for this one rather esoteric safety flaw, the RBMK Reactor seems like the platonic ideal of fission reactors. 

Esoteric?

There's a positive void coefficient in the reactor design. Which means it's not passively safe nor does it throttle itself down as it grows more active and starts to enter an uncontrollable state. That's not an esoteric design flaw, that's a disaster waiting to happen in any industrial process. That's why you shove as many safety systems onto it as you can fit to prevent it from exploding.

The use of graphite lead control rods only makes the problem worse. I get wanting to use a moderator to get a bit more reactivity going on start up, but it'd have been safer to use a neutron reflector (fast neutrons as you noted earlier are much less likely to split another atom compared to thermal neutrons, and RBMK reactors are thermal neutron based reactors) or another neutron absorbing material as the water displacing channel plug, even if that does mean you can't target specific channels with a bit of extra moderator to make it more reactive.

Part of the problem is that they designed a reactor where one of the safety systems also doubled as a reaction catalyst to make it go faster. And they did it deliberately. It's utter stupidity to give a safety system any other role other than being a safety system, as then it's no longer working or employed as a safety system.

[Bob Schroeck]

How is that a bobble?

The actual line trails off with a comma and an incomplete clause, not a period.

I spotted another -- "ratched" for "ratcheted".

Regardless, an amazing piece of work, Dartz.

[Dartz]

I don't have the creativity to write Science-Fiction at the moment - so some Science-Fact will have to do. Even though a little bit of creativity tends to be required - I had to go back and correct a few things I misremembered. There're a dozen shifting accounts of what happened - even from those in the room that night.

With Chernobyl the truth is churned up and spat out in a thousand different versions - of which I'm just churning up and spitting out the least dramatic one I can think of . There're so many competing agendas, quite a few in power who'd like to remain that way, and so many convenient scapegoats.

Akimov and Toptunov gave statements before their deaths. Their testimony is still classified, as is the entire subsequent trial proceeding. Tregub and Stolyarchuk survived to speak publically - which they do from time to time. Bryukhanov is still alive, and living in Ukraine. Fomin went back to working in the nuclear industry after serving a prison sentence. Radiation finally caught up with Dyatlov in the early 90s - but not before he wrote a book that very much disagrees with the official account in many places.

There is a need for those in the room to be responsible - it's far easier to point the finger at the engineer who pushed the button. It was known within a month what had caused the explosion - however the Scientist who discovered the flaw was quietly fired for his troubles years before. Not by Gorbachev, or the KGB - but by his own boss and his colleagues who had every interest in their designs for the RBMK not being found to be responsible for disaster.

The KGB of course, knew this, and knew how utterly necessary it was to figure out what really happened to make sure it never happened again. This is why Valery Legasov was appointed to head the commission - he was a radiochemist, a specialist in how monatomic gasses behaved in ionising radiation - and was most definitely not an expert on RBMK reactors. He was unconnected with the reactor's designers, didn't move in their circles and was most likely to give an honest response. That he did - and more. While the effectiveness of some of his decisions during the managment of the disaster are debatable - Legasov and his team of scientists also faced something which had never occurred in human history and had to make it up as they went along. That most of Europe is still habitable suggests something was done right.

He still toed the party line when faced with the world's glare at Vienna - to avoid a massive international outcry - but at home he knew the truth and pushed for reform of the scientific community that had created the disaster. His 'secret' memoirs were widely circulated and were printed in extract in the newspapers. He was villified not by the Soviet state, but by his own colleagues, who had nothing but their prestige to lose.

Legasov ultimately blamed the disaster on the ability of the operators to override the automatic safety systems - and on the simple fact that the reactor was never equipped with a fully enclosed containment system.

Nevertheless, the reactors were still rapidly modified after the disaster to reduced the void coefficient, and eliminate the tip-effect from the rods. Because reality always wins - and it's one thing to have an annoying power increase, and quite another to have an explosion occuring with no warning. It seems nobody believed Chernobyl was possible until it actually happened. It would take a certain level of insanity to run a reactor knowing that kind of explosion was possible.

It was far easier to rewrite the documentation after the fact. The reactivity margin limit at 15 did exist, and it was violated - but the consequences of violating the limit were never made clear to anyone involved. What consequences were known - were vaguely defined. It was like the speed limit on a wide road.

Ignalina - a similar reactor, but more powerful - in Lithuania was modified heavily after Lithuania joined the EU to the point where it was mostly compliant with European norms. The modern RBMK is a very different beast - it now runs on partially enriched fuel, enabling enough fixed absorbers to be used in the core to force a negative reactivity coefficient in all circumstances. The control rod channels were modified - either with redisigned tips, or replacing liquid water with gas. Their speed was increased, and some of the computer control systems were changed. Ignalina was modified, so that the lid will not lift until fifteen channels are burst. A second shutdown system was added, given control over a specific group of control rods seperate from AZ-5, and was designe to be non-overridable. The a third shutdown system was added after this - injecting liquid boron into the cooling shutdown circuit to kill the reaction dead.

Lithuania was happy with it. Ignalina's two reactors produced enough power to send two Deloreans back to 1955 every operating second, with enough to spare to keep the local town's lights on.

The EU still noped the fuck out and paid them to shut it down.

[Black Aeronaut]

It really is a shame.  If not for this one rather esoteric safety flaw, the RBMK Reactor seems like the platonic ideal of fission reactors. 

Esoteric?

Yes.  Esoteric by the layman's standards.  Which is kind of the problem these days.  People simply look at Chernobyl, Three Mile Island, and Fukushima and go NOPE when the topic of nuclear energy comes up.

At any rate, Dartz said it better than I did.  They thought they had the positive void coefficient handled.  And when Chernobyl happened?  They made modifications to fix that.  The reactors that are operating right now are still very much RBMK reactors.  They're just a lot safer now at the cost of having to enrich the fuel a tiny bit.

That said, I would love to see how well the new MKER reactors perform.

[Dartz]

CANDU reactors do most of the same basic thing - except they use heavy water rather than graphite. They're a fair bit more expensive to build and operate - but you get all the benefits with none of the kaboom.

I may turn this into a Useful Notes page on ATT.

----

Froom INSAG-7 - https://www-pub.iaea.org/MTCD/publicatio...3e_web.pdf - Page 87


"The Operating Procedures permitted operating conditions similar tothose prevailing at Chernobyl Unit 4 on 26 April 1986 and they might haveoccurred without any intervention on the part of the personnel. We only needto assume a perfectly possible situation in which triggering of EPS-3 occurswhen the reactor is operating initially at rated power with an ORM of26 manual control rods. Under these conditions, approximately one hour aftertriggering of EPS-3 the ORM could have fallen to less than 15 manual controlrods at a reactor power of 200-300 MW(th), and any further action, whetherautomatic or remote, to shut down the reactor could have led to a similar repe-tition of the events of 26 April 1986."

No violation of proceedures was required. The conditions that caused the explosion had the potential to occurr automatically in the normal operation of the reactor (EPS-3 is a different automatic protection mode which was a fast power reduction rather than a total stop.)

[Black Aeronaut]

Well that's a "YIKES".

I know that the reactor explosion was unexpected, but even so, I can only imagine what it would have been like for one of the regular shifts expecting a normal day's operation.

[Star Ranger4]

Lithuania was happy with it. Ignalina's two reactors produced enough power to send two Deloreans back to 1955 every operating second, with enough to spare to keep the local town's lights on.

*blink* *blinK* *blink*

So, 2.42Gigawatts, PLUS the local load????

*sigh*

[hazard]

Look, when you are operating in the multiple gigawatt range, a few megawatts just make less of an impression.

[Black Aeronaut]

Lithuania was happy with it. Ignalina's two reactors produced enough power to send two Deloreans back to 1955 every operating second, with enough to spare to keep the local town's lights on.

*blink* *blinK* *blink*

So, 2.42Gigawatts, PLUS the local load????

*sigh*

And to think, the NIMBYs never seem to understand why the proponents of nuclear power push so hard.  This is the real benefit of nuclear power - the sheer ridiculous energy density.

Nuclear reactors can be so powerful that you can keep them all in a centralized location - safe from harming any established community* - and transmit power over incredibly long distances through an alternate means such as superconducting power lines... which can be cryogenically cooled with hydrogen from a cracking plant fired by the reactors.  Hey!  That's TWO energy mediums in ONE piece of infrastructure!  How awesome is that?

* For a given value - there will be at least one community close at hand to such a reactor complex, and that will be solely dedicated to the people that work in this complex.  Of which there would be a fair number.

It also helps if you build containment buildings the way we do here in America.  Make no mistake: even if it was only a "partial meltdown", Three Mile Island was pretty much the worst-case scenario for a boiling-water reactor - what really sets it apart from Fukushima and Chernobyl is the American fixation on The Worst Case Scenario... So we build our reactor containment buildings so strong that you'd need an actual bunker-buster bomb to breach one.  Thus, TMI #2 was contained, and the only danger was a small radiation leak from the stuck relief valve as it vented steam laced with radioactive noble gasses into the air.  (To put it into perspective, people in Denver, CO get a higher dose from being out in the sun.)

[classicdrogn]

US uranium mining happens in Wyoming, Nebraska, Utah, and Texas, according to a quick search, all places famous for having lots and lots of elbow room. With the existing reactors aging toward retirement or heavy required refits and the not-impossibly-blue-sky prospect of those superconducting grid trunks putting a bunch o fgeneration capacity out thataway would make quite a bit of sense... permitting and politics are the biggest hurdles no matter where you look, though.

[hazard]

IIRC the world's biggest (and only) long term nuclear waste storage is supposed to be build in some far off corner of the USA in an inhospitable landscape. The people living there love the idea of oh so many federal dollars for keeping an eye on the world's deepest pit. Every community around that would see the nuclear transports roll past? Not so much, which is why that nuclear waste storage facility isn't operational.

[ECSNorway]

IIRC the world's biggest (and only) long term nuclear waste storage is supposed to be build in some far off corner of the USA in an inhospitable landscape. The people living there love the idea of oh so many federal dollars for keeping an eye on the world's deepest pit. Every community around that would see the nuclear transports roll past? Not so much, which is why that nuclear waste storage facility isn't operational.

Meanwhile research on Thorium reactors is stalled because - despite the fact that their waste doesn't require that kind of long-term storage and they're far safer and less likely to have issues than any uranium/plutonium design - nobody will let them build any to test it.

[hazard]

To be fair, thorium reactors still need long term storage, just not 'let's wait a few thousand years' long term storage.

Another reactor option is KNa cooled reactors, using a sodium and potassium alloy. Under the right ratios that mixture is liquid at room temperature and has a boiling point well above 1000 degrees celsius. The USA had a test reactor cooled this way, it ran for years (decades IIRC) without issue and at one point they yanked every control rod out of the reactor for a test and just let it come up to whatever temperature it wanted.

Funny thing about nuclear reactors, they do tend to become less active as temperatures increase. And by the time it reached some 1100 degrees celsius, temperatures ceased to increase. No need for a pressure vessel to contain the reactor, very little risk for a meltdown.