Brevity
The colloquial name for the V.I Lenin Nuclear Power Plant near the town of Chernobyl in Ukraine. More commonly refers to the catastrophe that occurred in reactor 4 of the plant on April 26, 1986.
A poorly-planned systems test, reactor design flaws, senior staff who were unfamiliar with nuclear reactors and junior staff who were unwilling to challenge their superiors combined to create a very volatile situation, with the reactor prone to rapid and 媚藥 心得 uncontrollable increases in power.
When such an increase did occur, an attempt to shut down the reactor caused its power output to skyrocket. The fuel melted, causing at least two explosions that partially destroyed the reactor building and set the graphite stacks of the core on fire. The core burned for ten days, releasing millions of curies of radiation into the atmosphere, contaminating thousands of square miles of land inside and outside the USSR, of which Ukraine was then a part.
The cleanup that followed, and attempts to cover up the truth of the incident cost the Soviet Union dearly in lives and international image. The accident and its aftermath are frequently cited as an important demonstration of the internal political rot that eventually caused the USSR to collapse. The effects are still felt today: a 30km exclusion zone encircles the plant, surrounding the worst contaminations which will remain at dangerous levels for tens of thousands of years. The effects of the accident on the health of former inhabitants are still not fully understood.
The concrete shelter currently covering the remains of unit 4, hastily built by remote control on top of the surviving containment structure, is increasingly unstable. Fears persist that the shelter is on the verge of collapse, which would be a fresh radiological and ecological disaster.
Verbosity
Sit down, get comfortable. Get very comfortable. Look how small the scroll bar is, for god’s sake.
Chemistry and Nuclear Reactors 101, or: Understanding As Little As archiewood
If you know about mass defect, radioactive decay, nuclear fission/fusion and all that good stuff, skip to the next heading. Otherwise, continue: a cursory understanding of the chemistry at work in nuclear reactors should help in understanding why things happened the way they did with this particular reactor.
Nuclear power harnesses energy stored in atoms. Atoms are made up of three types of particles: protons, neutrons, and electrons. The number of protons that an atom has determines what element it is. A hydrogen atom, for example, has one proton. A Helium atom has two.
Many elements have several different ‘versions’, called ‘isotopes’. Isotopes of an element have the same number of protons but different numbers of neutrons. An atom of protium—the most common isotope of hydrogen—has one proton and one electron. An atom of deuterium has one proton, one electron and one neutron. Protium is the only naturally-occurring element which does not have any neutrons.
The rather odd aspect of atomic properties and the underpinning for nuclear power is that the weight of an atom and the total weight of its component parts are different.
Pour l’example:
An atom of 1-hydrogen is made of one proton and one electron. We can safely ignore the latter because its mass — about 1/2000th that of a proton — is insignificant here (as an aside, the elusive neutrinos, which fly through just about everything with gay abandon, have a mass about one ten-millionth that of electrons).
One proton has a mass of 1.007277 atomic mass units, so that is close enough to the ‘expected’ mass of 1-hydrogen. 1-hydrogen actually has a mass of 1.00794 AMUs, which is about 0.06% greater than expected. This difference between actual and expected masses is called ‘mass defect’ and is posessed by all elements in varying amounts. 1-hydrogen weighs more than expected, but all of the other isotopes weigh less. A couple more examples:
2-Helium:
Expected atomic mass: 3.0232170 AMUs
Actual atomic mass: 3.0160293 AMUs
Mass defect: 0.0071877 AMUs, or about 0.24% below the expected mass.
24-Chromium:
Expected atomic mass: 50.3999140 AMUs
Actual atomic mass: 49.9460496 AMUs
Mass defect: 0.4538644 AMUs, or about 0.91% below the expected mass. This ‘missing’ mass represents the energy that holds the atom together, which is called binding energy or the strong nuclear force. If you go through all elements, from the lightest to the heaviest, the mass defect increases as well. It peaks at about 0.91% with isotopes of iron and nickel, then tails off slowly, down to about 0.75% for the heaviest isotopes.
Nuclear power works because different elements have different mass defects. Here are a couple of crude, unrealistic examples:
Split an atom of plutonium into two silver atoms, and compare them with an intact plutonium atom (congratulations! You’ve just performed your first fission reaction). Plutonium’s mass defect is 0.79%, but silver’s is 0.89%. So silver — the product of this reaction — has more ‘missing’ mass than the plutonium that fuelled it. This extra missing mass is released as energy.
Now, do it the opposite way. Take two helium atoms and combine them into one beryllium atom. This, to the vexation of humanity so far, is rather more difficult. You will need to smash the atoms together really hard: so hard that their speed overcomes the electromagnetic repulsion of their component parts and they get close enough for the strong nuclear force to stick them together. To make the atoms move this fast you will need to heat them to at least one hundred million degrees celsius. Good luck with that.
All of that nausea taken care of, you should now have a beryllium atom. Compare it to an intact helium atom and you’ll see beryllium’s mass defect is 0.42% greater, and this difference in mass is released as energy. Note that the energy gain of this reaction — nuclear fusion — is far greater than that of the fission example.
So why is uranium such a common fuel for fission reactions but Aluminium, say, isn’t? It might be easier if I use not-particularly-realistic examples on a conveniently-exaggerated binding energy curve:
Iron/Nickel
| / \
| / \
| Aluminum -> / \ <- Tellurium
| / \
| / \
|\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
+——————————————- The x-axis of the graph represents elements ranked by ascending atomic number (number of protons). The y-axis represents the mass defect of each element. Iron is at the peak, with a mass defect of 0.92%. Aluminium has half as many protons as Iron, and a mass defect of 0.86%. Tellerium has twice as many protons as Iron, with a (average) mass defect of 0.88%.
Now, producing energy (or not) using these elements:
If you fused two atoms of Aluminium into a single atom of Iron, you would get an atom with a mass defect of 0.92% from atoms with a mass defect of 0.86%. 0.06% of mass is lost as energy.
If you split (fissioned) an atom of Tellurium into two Iron atoms, you would get atoms with mass defect of 0.92% from an atom with a mass defect of 0.88%. 0.04% of mass is lost as energy.
These energy amounts may seem like much (and they’re not: 0.04% of the mass of an atom is an unimaginably tiny amount), but realise how many billions of atoms are in a small sample of any element and it should be clear how much energy could potentially be released through selective fission or fusion of the right elements.
On the other hand…
If you split an atom of Iron into two Aluminium atoms, you would get atoms with a mass defect of 0.86% from an atom with a mass defect of 0.92%. 0.06% of mass is gained, thus energy is lost.
If you could persuade two atoms of Iron to fuse into a single Tellurium atom, you would get an atom with a mass defect of 0.88% from atoms with mass defect of 0.92%. Again, mass is gained at the cost of energy. So Iron’s pretty crap as nuclear fuel. In fact, it’s as crap as it gets.
Although all of the above examples are inaccurate for one reason or another, the jist is that a nuclear reaction will only produce energy if its products have a mass defect greater than that of the fuel (in other words, the products are more tightly bound elements than the fuel): the reaction releases this extra, ‘missing’ mass as energy.
In principle, nuclear power is quite similar to other common power generation techniques: water is heated to create steam, which drives turbine generators to produce power. The only difference is the manner in which the water is heated.
Current nuclear power generation uses nuclear fission. Humans have yet to achieve stable nuclear fusion in a form that produces usable amounts of energy, although smart people are working on it. With nuclear fission, atoms of fuel are split (fissioned) to release energy. In theory any element heavier than Iron could be used as a fission fuel but in practice, isotopes of unstable heavy elements are used because they are comparatively easy to fission (Uranium is the heaviest naturally-occurring element, and it is this that fuels most nuclear reactors). The nucleus of an unstable atom is so large that it stretches beyond the sphere of influence of the strong nuclear force, so the latter is insufficient to overcome the natural tendency of the particles of the nucleus to fly apart. The result is that unstable elements gradually decay over time, losing particles on their way to becoming more stable elements.
Uranium is actually very stable in its most common form. U-238 has a half-life of 4.5 billion years (or: in 4.5 billion years a sample of U-238 will lose half its mass to radioactive decay), which is greater than that of many much more common elements. U-235, on the other hand, will readily absorb a passing neutron, become unstable and release energy as it splits into lighter elements and neutrons. If U-238 absorbs a passing neutron it becomes U-239, which decays to Neptunium-239, which then decays to Plutonium-239. Pu-239 itself can absorb a passing neutron, upon which it becomes unstable and splits to produce energy.
This is quite handy in using Uranium as nuclear fuel. You see, only a small proportion (2-3% on average) of all of the Uranium fuel needs to be U-235. Enriching a sample of Uranium will split it into a portion that contains more U-235 and a portion that contains less; the process of enrichment can be repeated to produce samples with higher concentrations if needed. The U-238, which makes up the majority of the fuel, can be converted into Pu-239 then fissioned; about a third of the energy produced by a Uranium-fuelled reactor actually comes from fission of Plutonium.
A nuclear reactor relies on a chain reaction in its fuel: an atom of fuel decays, releasing a neutron which is absorbed by a neighbouring atom. This atom then becomes unstable, splitting into other elements releasing energy and neutrons, which then go on to strike other atoms, in turn causing further fission events and so on. This gives a self-sustaining reaction, though it is tempered by various factors as we shall see.
Very broadly, a typical nuclear reactor has the following components:
Core – this is the vessel for the fuel, coolant and control rods of the reactor.
Fuel – most commonly slightly-enriched Uranium, or Plutonium. Usually in the form of rods, bundled together in containers.
Control rods – these are rods made out of neutron-absorbent material such as boron or cadmium. These can be inserted into the core to inhibit or stop the nuclear reaction, absorbing neutrons that would otherwise cause fission events.
Moderator – this is a substance that aids the nuclear chain reaction. Neutrons released by nuclear fission (called ‘prompt neutrons’) are travelling so fast they would fly straight through other fuel atoms and not cause any other fission events. The moderator is a substance that slows down neutrons to increase the likelihood they will be captured by fuel atoms, cause further fission events and keep the reaction going. The moderator in modern reactors is commonly light water (‘regular’ water) but can be graphite, deuterium (aka ‘heavy water’) or beryllium.
Coolant – this is pumped around the reactor core, typically immersing the fuel assemblies, transporting the heat generated by the nuclear reaction away to be used for power generation, and to prevent the fuel from overheating. If the coolant is water, it may be converted into steam or it may, by means of a heat exchanger, heat water in a separate circuit which is then converted into steam and used for power generation. This method keeps radioactive steam separate from the power-generation side of the plant, and is also the way nuclear plants cooled by substances like sodium operate.
Steam Turbine – this is run by steam generated by the heating of coolant. Steam rotates the turbine, which then drives a
Generator – this is connected to the turbine and actually produces the elastic trickery. Power output is usually about one third of the reactor’s thermal output.
RBMK Reactors
Now, the Chernobyl reactor. This was an RBMK-1000, which is a graphite-moderated light water reactor, that outputs 1000MWe (megawatts electrical). The design was originally intended for producing weapons-grade plutonium. Four reactors had been built at the site and two more were under construction at the time of the accident. An RBMK consists of the following:
Core – a cylinder 12 metres in diameter and seven metres high, built from blocks of graphite, which also serves as a moderator. The core is surrounded by a metal containment structure filled with inert gas; this is encircled, topped and tailed with thick concrete biological shields. The core is threaded with about 1700 3.5″ pressure channels for fuel, coolant and control rods.
Fuel – pellets of 2%-enriched uranium dioxide in 3.5cm zirconium alloy tubes. 18 of these tubes are bundled together to form a fuel assembly, and two fuel assemblies are inserted end-to-end in a fuel channel. The top of each fuel channel is covered by a 350kg metal cap. Individual fuel channels in the RBMK can be isolated, so it is possible to refuel the reactor while it is still running.
Coolant – this is light water (H2O) which is pumped through the pressure channels. There are two cooling circuits and four coolant pumps. A separate, emergency system operates automatically under certain circumstances, flooding the reactor with about 350 tonnes of cold water.
Control Rods – 211 boron carbide rods which absorb neutrons, either inhibiting the nuclear reaction or stopping it completely. 179 of these descend into the core from above (139 are manually controlled; the rest are automatic) but the others enter the core automatically from below, so power generation is evenly-distributed throughout the core. The core is so large that without this extra control it would behave more like multiple reactors with multiple critical masses. The upper control rods are lowered by servo motors (complete lowering takes about 18 seconds) but in an emergency, can be disconnected from the motors to fall under their own weight.
Turbine Generators – the RBMK reactors at Chernobyl have two of these, rated at 500MWe each. They are fed steam directly from the reactor’s cooling system, which rotates turbine blades attached to generators to produce electricity. The larger version of the RBMK, the RBMK-1500 (the largest nuclear reactor ever built), has two 800MWe turbine generators.
Design Problems
The RBMK design has been widely criticised, with good reason. It is unique in being the only water-cooled, graphite-moderated reactor design. This is not a good thing.
Moderator
Modern reactors take effort to keep running. The majority use water as coolant and as a moderator, since it has neutron-absorbent and neutron-slowing properties. The benefit this arrangement is that, to a degree, it makes the reactor self-governing.
When water boils it turns into steam (reactors use pressurised water, which boils at a much higher temperature than normal), which neither absorbs neutrons nor provides any moderation. If a water-moderated reactor starts to overheat, some of its moderator turns into steam and is temporarily lost, so the reaction will tend to slow down because fewer neutrons are absorbed by the fuel. This is all very cool.
However, an RMBK reactor has a moderating substance that is separate from the coolant. If an RBMK were to start overheating for any reason—producing excessive steam in its coolant—neutron absorbency would be lost but the moderator would be unaffected. The nuclear reaction would continue, with reduced cooling capacity, and would perhaps even intensify: less neutrons would be absorbed by coolant, meaning more neutrons could be absorbed by fuel atoms. To make matters worse, graphite slows more neutrons the hotter it gets. This brings us neatly to:
Positive Void Coefficient
There are few articles about the Chernobyl mishap that don’t mention these three magic words. Up until it was exposed what serious problems it could cause, all RBMK reactors exhibited this characteristic.
A “void” is a pocket of steam in reactor coolant. Again, steam does not absorb or slow neutrons well compared to water. A nuclear reactor has a positive void coefficient if too many voids in the coolant increase the intensity of the reaction. A reactor with a negative void coefficient—i.e. most reactors currently operating—will produce less power if excess steam forms in the coolant.
If your water-cooled, graphite-moderated reactor starts to produce excessive amounts of steam in the core, then cooling capacity is reduced because steam doesn’t absorb heat very well. You’ve also lost the neutron-absorbing property of the coolant that has boiled, but the graphite moderator remains intact so more neutrons get captured by fuel atoms, more nuclear reactions occur and the reactor produces more power. As it gets hotter, more of the coolant turns into steam, reducing neutron absorbency and cooling capacity even more. As the graphite gets hotter it slows more neutrons, increasing the likelihood they will be captured by fuel atoms. I guess you can see where this is going. A feedback loop of sorts builds up and can do so in a very short time, making the reactor difficult to control.
Control Rods
This design problem was a major contributor to the incident, although it has since been fixed. In the original design, a five-metre length of graphite called a ‘follower’ or ‘rider’ is suspended on steel cables below each manual control rod. When the control rod is completely withdrawn, the rider occupies its space in the core. Above and below the graphite is a one-metre cavity that fills with coolant:
— +—+
^ |\\\|
| |\\\| <-- Boron
|5m /\/\/\/
| |\\\|
v |\\\|
— +—+
^ | |
|1m | | <-- H2O
v | |
— +—+
^ |///|
| |///| <-- Graphite
|5m /\/\/\/
| |///|
v |///|
— +—+
…
+—+
|\\\|
|\\\| <-- Raised control rod;
/\/\/\/ Graphite portion in core
|\\\|
|\\\|
+——+—+——+—+——+
|——| H |——| H |——|
|——| 2 |——| 2 |——|
|——| O |——| O |——|
|——+—+——+—+——|
|——|///|——|\\\|——|
|——|///|-CORE-|\\\|——|
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
|——|///|——|\\\|——|
|——|///|——|\\\|——|
|——+—+——+—+——|
|——| H |——| H |——|
|——| 2 |——| 2 |——|
|——| O |——| O |——|
+——+—+——+—+——+
|///|
|///| <-- Lowered control rod;
/\/\/\/ Boron portion in core
|///|
|///|
+—+
The problem this causes is that as the control rods are lowered from their fully-raised position, the graphite rider pushes coolant out of the channel below it: moderator takes the place of neutron absorber. Although boron is beginning to enter the top of the core as this occurs, the event can sometimes concentrate power generation in the bottom of the core.
Most of the time this doesn’t matter. When the reactor is running at full or high power, there is more steam than coolant in the core anyway. The volume will change little when the control rods are lowered, and won’t significantly affect the running of the reactor.
However, when the reactor is running at low power, particularly with high coolant flow, the situation is different. There will be much more coolant in the core; much less steam is being produced because the fuel is producing less heat. If coolant is flowing quickly through the reactor, even less steam will be produced because coolant doesn’t stay in the core long enough to boil. In these circumstances the coolant will play a much more significant role in absorbing neutrons, simply because there is so much more of it. It is quite possible that control rods will have to be withdrawn to keep the power output steady.
In a situation like this, if you not only displace lots of coolant at once but replace it with graphite, the effects could be significant. If you were to do it with all of the manual control rods at once, you could get quite a nasty power spike.
This design problem was actually known about well in advance of the Chernobyl accident. Unfortunately, the Soviet culture of concealment in which all nuclear accidents were state secrets (because, y’know, nuclear power—and particularly Soviet nuclear power—was completely safe so accidents could never happen) prevented this information being disseminated so problems could be fixed or learnt from.
Containment
The other major criticism of the RBMK design is that it incorporates no secondary containment structure. In most nuclear reactor complexes a steel and concrete building, commonly spherical or dome-shaped, encloses the reactor containment vessel, to contain any radiation escaping from the core. Such a structure almost completely prevented radiation leakage during the Three Mile Island nuclear meltdown in 1979.
Having said that, it’s possible a secondary containment structure may not have been able to withstand the explosion of the Chernobyl accident. It may have lessened the environmental effects and made the aftermath somewhat easier to deal with, but this is all speculation from someone who knows relatively little.
The Cast
The following are some of those involved in the accident and those on duty on the night (spellings taken from Piers Paul Read’s book).
Alexander Fomin – Chief Engineer, V.I. Lenin Nuclear Power Plant
Anatoli Dyatlov – Deputy Chief Engineer, units 3 & 4
Alexander Akimov – Shift Foreman, unit 4 (night shift)
Yuri Tregub – Shift Foreman, unit 4 (afternoon shift)
Valeri Perevozchenko – Shift Foreman, Equipment Maintenance Department, units 3 & 4
Leonid Toptunov – Senior Reactor Control Engineer
Piotr Stolyarchuk – Senior Unit Control Engineer
Igor Kirschenbaum – Senior Turbine Control Engineer
Razim Davletbayev – Deputy Head, Turbine Section
The Test
On April 25, 1986 it was planned to shut down reactor 4 of the Chernobyl for maintenance and refuelling. As the construction of the plant had been rushed to meet a schedule, some tests on the reactor’s systems had been neglected in favour of putting the plant into operation as early as possible.
One such test was on the plant’s turbines, intended to test whether they would work as a temporary source of electricity during a power failure. A nuclear power station consumes electricity from the grid as well as producing it; if there is a power cut it doesn’t just turn off like a light. Coolant pumps, servo motors and control systems must be kept running for hours afterwards until the reactor has completely shut down. Normally a reactor can tap its own production to power these systems but if it is in the process of shutting down, an alternative is needed.
The RBMK reactor has backup diesel generators for this purpose, but they take 40-50 seconds to start up (another design flaw). During this time a further power source is needed. The intention was to use the power produced by the reactor’s own turbines as they spun down; this had actually been attempted during a previous shutdown but had been unsuccessful, because the power produced by the generator dropped too quickly. New voltage regulators had since been fitted to the generators and it was these that were to be tested.
The test proposal seemed very simple: the reactor would be run at low power, then the valves to the turbines would be closed and the reactor shut down. The output of the turbines would then be monitored as they slowed down.
The proposal also required the reactor’s Emergency Core Cooling System (ECCS) to be disabled. The reasoning was that if the ECCS was triggered during the test it would shut down the reactor; it would be impossible to repeat the test, as the next scheduled shutdown was not for another year. There were also concerns that so much cold water entering the reactor would cause a heat shock that would damage, possibly even destroy the core. It is not agreed whether disabling the ECCS would have made much difference to the accident, but it contravened plant regulations nonetheless.
It was not thought necessary to clear the programme with Boris Rogozhkin, Chief Of Shift for units 3 & 4; a proviso was merely added to the test procedures that all switching operations be cleared by the Shift Foreman. All safety measures could also be overridden by the Deputy Chief Engineer (Dyatlov), who intended to be present for the test. Mikhail Lyutov, Scientific Deputy Chief Engineer responsible for matters of nuclear safety and who was supposed to have been consulted, was away having a medical check at the time. Chief Engineer Fomin, a turbine engineer, knew very little about nuclear reactors. He approved the test proposal.
Timeline
This is the result of consulting multiple fragmentary, sometimes inconsistent sources of event sequences and triggers. It is worth noting here that much of the physical evidence of what happened in this accident was destroyed so by necessity there is an element of uncertainty in any account. This writeup represents what I believe to be the prevailing timeline and contributory factors. Timings are taken from Engineering.com for consistency’s sake but again, are approximate.
April 25, 1986
01:00 – reactor 4 is stable at 3000MWt (megawatts thermal). Akimov begins to gradually reduce reactor power output in preparation for the test. This reduction must be gradual because Xenon and Iodine – short-lived Uranium decay products present in the core – have to be allowed time to decay themselves. Xenon and Iodine absorb neutrons so if power is reduced too quickly, these elements could excessively accelerate the power reduction, possibly shutting down the reactor prematurely.
03:47 – power reduction is halted at 1600MWt.
08:00 – shift change. Dyatlov and others remain to supervise the test as it continues.
~13:00 – reactor 4 reaches half power at 1500MWt. No. 7 turbine generator is disconnected from the grid (each reactor has two turbines – unit 4 has turbines 7 and 8).
Almost everything is now ready to start the test on turbine 8. The Emergency Core Cooling System is disabled before continuing, as some of the test conditions would trigger it (the reduction in power to the coolant pumps from the slowing turbine, or the triggering of the diesel generators). The valves allowing water from the ECCS into the reactor are also closed, and the diesel generators are disabled to ensure all electricity comes from turbine 8.
14:00 – a call is received from the Load Dispatcher in Kiev saying power demands require turbine 8 to continue operating until 23.00. The test is postponed. Dyatlov goes home, leaving instructions to the following shift to maintain the reactor at half power. The ECCS remains disabled.
16:00 – afternoon shift takes over with Foreman Yuri Tregub. It is expected the experiment would have been complete by this point so none of the operators have been briefed, and are unsure whether to follow the instructions left by the previous shift.
23:00 – night shift staff begin to return. The Load Dispatcher telephones to advise that turbine 8 can now be disconnected from the grid.
23.10 – under orders from Dyatlov, Tregub now begins to reduce power from 1500MWt to ~700MWt so the test on the turbine can begin.
April 26, 1986
00:00 – night shift takes over. Tregub remains behind to watch the test. Turbine Engineer Igor Kirschenbaum prepares for the test but is concerned that the turbine may be damaged by disconnecting it from the reactor while it is still running. As is common in the Soviet Union, he defers to his superiors.
00:28 – Akimov orders Toptunov to disable the Local Automatic Control (LAC) system to give the operators more precise control of the reactor.
The LAC system is designed to even the distribution of neutron flux throughout the reactor core. It monitors neutron production in each of several circular ‘zones’ throughout the core, and inserts control rods into particular zones if neutron levels get too high.
The LAC system is intended to hold reactor power at or above 700MWt. According to Engineering.com (which has the best explanation of what happened next), Toptunov either fails to give the ‘hold power at required level’ signal or the system doesn’t respond to the command. Reactor power then slumps to 30MWt.
The reactor has now fallen into an ‘iodine well’, poisoned by the neutron-absorbent byproducts of its own reactions. These are accumulating in the top half of the core, which causes power generation to concentrate in the bottom half. The test should have been aborted here: the reactor shut down, the fission products allowed to decay and the reactor restarted. The only “safe” alternative would have been to hold power at that level for 24 hours while the Xenon and Iodine decayed, because trying to pull the reactor out of the iodine well would be difficult and dangerous.
Toptunov evidently understands this because when Dyatlov—very angry at this development—orders him to immediately increase reactor power, he refuses. Shift Foreman Akimov backs him up and they explain that to raise power from this level they would have to withdraw many more control rods, possibly putting the reactivity reserve (the effective number of whole control rods in the core) below the mandatory minimum of 15. Plant regulations also forbid increasing power if it has fallen from 80% of capacity with the current reactivity reserve of 28. Raising power would clearly make it very risky to continue running the reactor.
Dyatlov ignores this argument, saying that as reactor power has only fallen from 50% of capacity the regulations do not specifically prohibit what he is ordering. He threatens to replace Akimov with Tregub if he does not increase the reactor’s power. The two men cave and Toptunov starts withdrawing control rods from the core.
01:00 – reactor 4’s power level has increased to and ‘stabilised’ at 200MWt. The reactor is virtually out of control at this point and will explode if an emergency shutdown is attempted, due to the control rod design problems mentioned earlier and the positive void coefficient property.
01:03 – as part of the test, an extra pump is switched on in the left-hand cooling circuit.
01:07 – as part of the test, an extra pump is switched on in the right-hand cooling circuit. These two acts reduce reactivity, because it increases the amount of water in the core. More water means more neutrons are absorbed, meaning fewer are ‘available’ to cause fission reactions. Thus, more control rods have to be withdrawn to stop power levels dropping. Also, the low power levels of the reactor make the hydraulic resistance of the core low, which increases the flow of water through it.
This increased flow has three effects: first, steam production reduces because water is not in the reactor for long enough to boil. This reduces the steam pressure and water levels in the steam separator drums.
Second, with so many cooling pumps operating at such low reactor power, coolant flow exceeds normal levels by about 30%. Coolant pumps start to vibrate and there is increased risk of pumps failing.
Third, there is so much coolant in the core that it becomes much more significant to the stability of the reactor. The coolant is absorbing neutrons the control rods should be absorbing, which is why more control rods have had to be withdrawn. Now, if the amount of coolant in the core decreases (and this could happen much faster than any control rods could be lowered to compensate), fewer neutrons would be absorbed and a rapid increase in power would be difficult to prevent.
01:15 – the steam pressure in the separator drums continues to decline, approaching a point at which it will automatically trigger a shutdown of the reactor. Akimov, with Dyatlov’s consent, disables the trip signal for this so the reactor can be kept running. Dyatlov then calls Perevozchenko, the Shift Foreman of the maintenance department to the control room.
Could the disaster have been averted in this situation? The answer is yes. All they needed to do was categorically scrap the experiment, switch on the emergency core cooling system, and start up the emergency diesel generators, thereby securing a reserve supply of electricity in case all power was lost. Operating manually, one step at a time, they should have lowered reactor power until the reactor was completely shut down, while taking great care not to press the [emergency shutdown] button, which would have been the equivalent of an explosion.
-Grigori Medvedev, former Deputy Chief Engineer, Reactor 1, Chernobyl NPP
01:19 – to address the problems of water levels and steam pressure in the separator drums, Toptunov withdraws more control rods to increase power. By this point the reactivity reserve is around eight (plus automatic control rods).
01:21:40 – to further address the steam pressure problems, Toptunov reduces coolant flow rate. This raises steam pressure but means heat is removed from the core more slowly.
01:22:10 – the computer monitoring the reactor produces its last printout. It shows power as 0.2GW and an operating reactivity reserve of eight rods, just over half the mandatory minimum value. It takes about ten minutes for this data to be collated so it may not reflect the actual value. Nevertheless, a reactivity reserve of eight rods would normally require an immediate shutdown of the reactor (of course, any attempt at an emergency shutdown now would just make the reactor explode earlier). Toptunov reports this to Akimov and Dyatlov just as the test is about to start.
01.23:04 (0:56 to go) – the throttle valves to turbine 8 are closed, starting the test. The turbine starts to coast down, monitored by the Donenergo representatives. The shutdown of both turbines would normally trigger an automatic scram of the reactor but this trip signal, too, is overridden.
01:23:10 (0:50 to go) – because half (or all, depending which source you check) of the coolant pumps are running from the power produced by turbine 8 (whose power output is falling as it slows down), coolant flow to the core begins to decrease. Steam production starts increasing in direct correlation, as coolant is in the core for longer; the coolant begins to boil.
01:23:21 (0:39 to go) – steam generation reaches a point where the reactor’s positive void coefficient means any further increases in steam pressure would lead to a rapid power increase. Every new steam formation reduces the coolant’s capacity for absorbing neutrons and heat, so it is just a matter of time before this happens.
01:23:35 (0:25 to go) – Toptunov notices reactor output beginning to rise. Knowing the dangers that were considered earlier he immediately tells Akimov the reactor must be shut down.
01:23:40 (0:20 to go) – after a few seconds of consideration and watching the power gauge rise from 200MWt to 530Mwt, Akimov presses the ‘AZ’ button on his control panel to insert all control rods into the core and “shut down” the reactor. Dyatlov later claimed in his book that this was simply shutting down the reactor on completion of the experiment. All control rods begin to move, but some of the remaining coolant is displaced by the graphite tips of the control rods as they enter the core. Instead of a drop in reactivity, there is a positive reactivity surge as neutron-absorbing coolant is replaced with graphite. This coincides with the rise in power already occurring from increased steam generation and the reactor’s positive void coefficient.
Because of the power increase, the hydraulic resistance of the core increases sharply, reducing coolant delivery further still. Film boiling occurs in the fuel channels, cutting the amount of heat the coolant is absorbing from the rapidly-overheating fuel rods to almost nothing.
Immediately after Akimov presses the AZ button Perevozchenko runs into the control room, reporting that as he walked across the catwalk at the top of the central hall (about 20 metres above the top of the reactor) he saw the heavy caps to the fuel channels jumping in their sockets. Akimov checks his readings on the position of the control rods and sees that they have stopped descending after only moving about 2.5m of their 7m travel; a number of deep thuds are felt from within the building.
01:23:44 (0:16 to go) – reactor power reaches 1.4GWt, the core becoming prompt critical at some points, meaning it needs no moderation to continue running and is impossible to control. Akimov disconnects the control rods from their servo motors so that they will fall into the reactor under their own weight, but they don’t move. Sudden heating has probably caused thermal expansion, distorting the reactor channels and jamming the control rods in position. More strong shocks are felt from the central hall.
01:23:49 (0:11 to go) – power levels in the reactor, still running but steadily boiling away its own coolant, now register at about 3GWt. The positive feedback mechanism that has built up from the knock-on factors of steam generation, increased reactivity and increased heat is unstoppable.
01:24 The power increase rapidly accelerates, from doubling every second to doubling roughly every millisecond. In the space of one or two seconds reactor power soars to an estimated 1.2 Terawatts (thermal), some four hundred times design capacity.
The fuel channels in the core rupture from this sudden heating and the fuel disintegrates.
Water directly contacting Uranium through the shattered fuel cladding produces a tremendous amount of steam, which immediately destroys relief valves and reacts with the white-hot graphite in the core to produce Carbon Dioxide. This rapidly accumulates at the top of the core with hydrogen, produced by zircalloy reacting with superheated water. The reactor, still running but now uncooled, detonates the mixture and a gigantic explosion partially destroys the core, blasts its three-metre thick, two-and-a-half thousand tonne steel and concrete lid into the air and blows the roof off the reactor building. The lid flips onto its side and jams back into the top of the core.
The explosion disperses about 170 tons of fuel, graphite and fission products into the surrounding countryside. A large amount of hot fuel and graphite lands on the flammable roof of the adjacent turbine hall, starting around 30 small fires and causing part of the roof to collapse. Air rushes into the remains of the central hall and sets the graphite remaining in the core on fire.
Aftermath
The events described above are actually the minor aspects of the accident; the explosion had very little effect on the environment and only killed two people. No, the real damage was done by the radioactive emissions of the graphite fire which burned for ten days afterwards. Humans have very little experience of fighting graphite fires, and still less fighting graphite fires seething with radioactivity.
Many firemen who arrived at the scene were soon cut down by the huge levels of radiation they were unwittingly exposing themselves to. The only radiation detectors on hand at the plant could not register more than 1 millirem per second – 3.6 rems per hour – and they were permanently fixed to their maximum reading. There was only one detector in the whole complex that could register more but it was locked in a safe that got buried under rubble by the explosion. Nonetheless, for some time 3.6 rems/hour was the figure those in charge gave to anyone that asked about radiation levels.
It was a while before anyone of seniority even accepted that the reactor had been destroyed. The unquestioning belief in Soviet nuclear power (and the culture of secrecy that prevented them from learning from, or of, other accidents) assured those in charge that nuclear reactors were completely safe and certainly could not explode. Their subordinates, who received horrendous doses of radiation from the quivering mass of twisted metal and burning graphite that now constituted the reactor core, were ignored when they reported their discovery.
When the Scientific Deputy Chief Engineer Lyutov finally arrived on site, Victor Smagin (who had arrived to replace Akimov as shift foreman) showed him graphite littering the surroundings of the central hall, but Lyutov did not believe the reactor had exploded. Even when the exasperated Smagin pointed out the holes bored through the graphite for the fuel channels, Lyutov remained unconvinced it was core material. Perevozchenko, who had gone to find two trainees sent to manually lower the control rods, was similarly ignored by Akimov when he reported that the reactor had been destroyed.
The report of Andreyevich Sitnikov (Deputy Chief Engineer of units 1 & 2), sent by Fomin to make an impartial assessment of the damage also fell on willfully deaf ears. Sitnikov went around the unit, as close to the central hall as he could get and up onto the roof of the neighbouring water treatment plant to get a better view. He received a huge dose of radiation as he looked directly down into the radioactive volcano of the core, his head alone receiving a dose of up to 1500 roentgens. When he reported what he had seen to Fomin and plant director Brukhanov, retching and carrying a fresh nuclear tan, the two men angrily dismissed and ignored his report.
The result of this blinkered incomprehension was a series of futile attempts to cool the now-nonexistent reactor. Men were exposed to radioactive air, steam and water as they tried to find their way through the remains of the reactor building to coolant valves and control rod servos. When the coolant valves were finally opened, the water did not flow into the core but instead to the spaces and electrical conduits underneath the building. Not that it would have helped, since the temperature of the graphite fire was high enough to split the water into explosive hydrogen and oxygen and make things worse, but these efforts badly depleted the reserves of clean water from unit 3 and risked short-circuiting the electricity supply to all three remaining reactors. The shift foreman of unit 3 asked Fomin for permission to shut his reactor down, but Fomin refused. Shortly afterwards the foreman began to do so anyway, and by 3am unit 3 was safely shut down.
Damage Control
The refusal by those in charge to accept that the reactor was mostly destroyed (what remained of the core was still running) was also the reason that it took so long for serious efforts to begin putting out the fire; until this began, the belief was that if water could just be pumped into the “core”, everything would be solved:
It is a strange thing, but during those truly weird hours, the overwhelming majority of the operational staff, including me, believed what they wanted to believe, and not what was really happening.
The nonsensical but extremely comforting idea that the reactor was intact mesmerized a great many people here, in Pripyat, in Kiev, and also in Moscow, which sent forth a stream of increasingly rigid and vehement commands: “Feed water into the reactor!”-Victor Smagin, Shift Foreman, unit 4
It took helicopter flights over the reactor by designers of the RBMK and the same from the chairman of the government commission to Chernobyl before there was any real appreciation by those in charge of how serious the situation was:
They were horrified by what they saw. With binoculars they looked down on the burning graphite, the red-hot biological shield, and the sinister blue glow in the core. It was terrible, but it was also awesome; they realised they were facing a catastrophe of a historic kind, like the eruption in Pompeii, or the earthquake and fire in San Francisco.
-Piers Paul Read
By the 27th, fission in the reactor was believed to have ceased but the graphite fire was doing all the damage, releasing an expanding cloud of radioactivity which was being blown over thousands of square kilometres. The first suggestion outside the Soviet Union that something may have gone wrong was not until the 28th, when very high levels of radioactivity were registered on plant workers at a Swedish nuclear plant. When no local leaks were found, eyes turned squarely towards the East.
The graphite in the core of unit 4 was burning at a rate of about one ton per hour, and with a core weighing 2500 tons (some of which had been thrown out by the explosion) clearly could not just be left to burn out. However, once graphite starts burning it is pretty difficult to extinguish. Do a web search for the words ‘graphite’, ‘burning’, ‘put’ and ‘out’ and you get articles on this accident.
Bags of sand mixed with boron, lead and dolomite dropped by helicopters were used to attempt to smother the fire and cut the emission of radiation. The sand cut off oxygen to the fire, the boron would absorb neutrons and inhibit the chain reaction, the lead would absorb heat and cool the core, the dolomite would do the same but also generate inert gas to smother the fire. Getting the sacks to land on target was exceptionally difficult because they were being quickly dropped by hand, and the concrete shield jammed into the top of the core left only one small, crescent-shaped opening.
It was also difficult to procure the thousands of tons of sand needed to do the job, as few people available to help collect it from the banks of the Pripyat river. People from a nearby collective farm were approached to help, but initially just laughed as commission members drafted for the work explained what had happened to the reactor. Eventually the point was established, as well as the fact their land was probably permanently ruined, so the farmers joined in.
Mi-8 helicopters made repeated sorties over the reactor, dropping several bags of material with each one; on the 28th 93 runs were made, on the 29th this doubled. The emission of radionuclides had dropped from an estimated 12 million curies immediately after the accident, to 4 million on the 27th, 3.75 million on the 29th and 2 million on the 30th. The downside to the ‘bombing’ was every bag of sand dropped threw a cloud of radioactive dust into the air, but the net release of radioactivity was dropping.
Unfortunately on May 1st the level of emissions jumped to 4 million curies again, indicating the core was getting hotter. It is not clear why this was, but it was estimated that the vast majority of the Uranium fuel had remained in the reactor. There were growing concerns that smothering the reactor with so much material (over 5000 tons by the time ‘bombing’ sorties stopped) was restricting the flow of oxygen to the core, which was previously having some cooling effect. No oxygen is needed for fission, so it was quite possible the fuel was still critical but now was being insulated and heating up. Uranium melts at 2900°C and at that temperature could burn through the two metre-thick concrete base of the core.
At this point people started getting worried about the water in the bubbler pools. The bubbler pools are areas under the reactor core where excess or leaking water collects for feeding back into the heat transport system. Much of the water from the initial misguided attempts at cooling the reactor after the explosion had ended up in them. If molten uranium fell into the bubbler pool it would immediately split the water into explosive elements of hydrogen and oxygen, which would then detonate with catastrophic results. The same risk was posed by the water that had collected in the basements under the complex, and the water table in the ground under that. The latter carried the further risk of contaminating drinking water.
After hazardous missions under the reactor to discover the amounts of water present, divers were sent into the radioactive bubbler pools to open the valves and let out the water. This went without a hitch but the water in the basements remained; this had to be drained by the local fire brigade. Hoses inadvertently being damaged by a passing vehicle hindered attempts at pumping water out considerably, but by May 7th all of the water had been pumped out of unit 4 to a neighbouring reservoir.
This averted the immediate danger but in the meantime, the commission had been focusing on the problem of the core that still seemed to be getting hotter. At first the idea was floated to build a heat exchanger underneath the reactor building and use water pumped from the emergency core cooling systems of units 1-3 to remove heat, but this was too long-term a solution for the majority staffing the government commission. Their plan was to pump nitrogen into the building, the thinking being that it would freeze the earth underneath the structure and smother the fire. When the nitrogen eventually arrived, it only took a day of pumping it into the building and it simply escaping into the atmosphere before those who dreamt up the idea realised how stupid it was.
On the 6th of May the emission of radionuclides dropped from 8 million curies to 150,000, so it was believed the fire had gone out and the fuel was cooling down. A further run over the reactor showed it was still glowing a little, so another 50 tons of lead was dropped which seemed to solve this.
In the mid 1990s, the UK documentary series Horizon made an expedition into the reactor. They found warm, smouldering concrete. It was found that molten fuel had burned right through the base of the core. It had mixed with the sand, ending up in the cavities below the core and setting in a somewhat crystallised state. One notorious deposit of this was the so-called ‘elephant’s foot’ – a clump of melted fuel and debris – which was first seen on said documentary, filmed by a remote camera. It was so radioactive, emitting more than ten thousand roentgens per hour, that it was impossible to get near it to collect a sample. A small chunk of it was shot off with an AK-47 rifle.
Back in 1986, work on the heat exchanger under the reactor began shortly after the attempt at cooling it with nitrogen and, about halfway through May, the building of a ‘sarcophagus’ to cover the reactor remains was announced on Soviet television.
[Designing the sarcophagus] was a complex task, not simply because [it] had to be built in such hazardous conditions but because the corpse was still twitching; plutonium 239 has a half-life of 24,360 years. The graphite had burned out and the temperature in the core had declined to about 270°C, but the fuel was still there in an unknown condition. It required no oxygen for fission so could not simply be buried. The sarcophagus would have to contain the radiation, yet have apertures for ventilation and observation.
Piers Paul Read
The sarcophagus was constructed very hastily and mostly by remote control due to the dangerous environment, in some cases using the remaining structure of the reactor building for support. This is the main reason for the concerns about the deteriorating state of the structure. A replacement arch structure will shortly begin construction close to the plant; this will slide over the top of the entire complex (the last of the four reactors closed down on December 15, 2000) on rails and contain decommissioning work. If the original sarcophagus collapses, as is the concern, the resulting fallout should all be contained within this new structure.
So, whose fault was it?
Needless to say there are many, many people better placed to answer this than me. My officechair view is that the Chernobyl accident was one of those event cascades where if one thing had not happened the way it did, the whole disaster might have been averted. There were many factors.
It was the fault of the designers for creating a reactor with multiple severe design flaws.
It was the fault of the Soviet state for building some.
It was the fault of the Soviet state for allowing the advancement of the well-connected, rather than those who were well-suited for particular roles. It was this attitude that allowed a turbine engineer to become the director of a huge nuclear power complex, a turbine engineer to be its chief engineer and a man with a background in small, research reactors to be deputy chief engineer of two of what were, at the time, the largest nuclear reactors in the world.
It was the fault of the turbine test planners for not fully understanding the reactor that was key to their test, and what an unusual and hazardous state their test could put it into.
It was the fault of Dyatlov for ordering Akimov and Toptunov to try to burn out of the iodine well, rather than to shut the reactor down. This is a good example of the letter of the law (the plant regulations, in this case) being used to defeat the spirit of the law.
It was the fault of Akimov and Toptunov for not having the courage of their convictions and standing up to Dyatlov, despite knowing that following his orders could result in disaster.
It was the fault of the Soviet state for conditioning citizens to unquestioningly respect and obey all forms of authority.
But if I had to cite one of these? Dyatlov, by ordering the Akimov and Toptunov to bring the power of the reactor back up after it had dropped to 30MWt was probably the last straw. Although the accident could possibly have been averted after this point it would have been much more difficult and time-consuming than simply shutting down the reactor here. Dyatlov maintained for the rest of his life that the later attempt to shut down the reactor followed the successful conclusion of the test, but the other evidence and testimony does not seem to bear that out.
Please /msg me with any questions or corrections. I have tried to understand this and make it as clear, accurate and comprehensive as possible, but as always I expect to be wrong somewhere. There are too few good narratives on the aftermath of the accident to make it worth noding; instead I point the reader to the first two sources which are both excellent (though an occassionaly tabloid-esque style and patchy translation in the former case), respectively for technical detail and imagery.
Sources and references:
Medvedev, G; “The Truth About Chernobyl”; Printed word, published by I.B. Tauris, ISBN 1850433313
Read, P; “Ablaze: The Story of Chernobyl”; Printed word, published by Mandarin, ISBN 0749316330
Environmental Toxicology and Chemistry, Vol.19, No.5, pp.1231-1232, 2000; “THE CHERNOBYL NUCLEAR DISASTER AND SUBSEQUENT CREATION OF A WILDLIFE PRESERVE”; website
various authors, newsgroup archive; website
BELLONA; “Chernobyl – the accident”; website
Engineering.com; “Chernobyl”; website
(author unknown); “Reactor accidents – Chernobyl”; website
various authors; “RBMK”; website
Adams Atomic Engines; “The Accident at Chernobyl”; website
(author unknown); “THE MYTHS OF CHERNOBYL”; website
Nuclear Engineering International; “Chernobyl, 26 April 1986”; website
Cheney, Glenn A; “Chernobyl: The Ongoing Story of the World’s Deadliest Nuclear Disaster”; website
Dukelow, Jim; “The Chernobyl Affair”; website
Uranium Information Centre Ltd;
“RBMK Reactors, Nuclear Issues Briefing Paper # 64A, February 2002”; website
“Chernobyl Accident, Nuclear Issues Briefing Paper 22, March 2006”; website
“WHAT IS URANIUM?”; website
“The Chernobyl site and accident sequence”; website
(author unknown); “Time Line of the Disaster”; website
Physlink.com; “Why is uranium fissionable and not, say, aluminum?”; website
Nave, C. R; “Chernobyl”; website
The Ukrainian World Congress; “Chornobyl Commission Report”; website
Jasiulevicius, A, Department of Energy Technology, Royal Institute of Technology, Stockholm; “Analysis Methodology for RBMK-1500 Core Safety and Investigations on Corium Coolability During a LWE Severe Accident”; website
Lithuanian Energy Institute; “Ignalina Source Book; Reactivity Control System”; website
Opening Up: A Guide to Creating and Sustaining Open Relationships by Tristan Taormino