SF-3 1971: The Soviet Union's Worst Criticality Accident

[00:00:00]It was 4:00 PM on May 26, 1971. At the Kacharov Institute of Atomic Energy, a prestigious scientific and research institution in the Soviet Union for nuclear research.

[00:00:14]The experiment is coming to an end. The experiment was a success, and the researchers are preparing to leave at the end of the day. However, just as they were rebooting the experiment and inspecting the equipment, the nuclear fuel suddenly and violently came to life in an uncontrolled explosion of power.

[00:00:38]At least two people died. Several others remained alive, but not healthy. The accident was covered up, classified, buried, and left to ponder for the next almost 30 years. This is the story of the SF3 experiment, the Soviet Union's worst criticality accident. The Kacharov Institute was the center of the Soviet Union's research in the field of nuclear energy, and many reactor designs originated at this complex. RBMK, VVER, VM reactors for submarines, and many others that they worked on. In 1966, the Kachottov Institute even developed its own experimental thermonuclear reactor, the Tokmax 6, for research into plasma physics. And yet, research into the specific dynamics of reactor behavior, particularly their shape, continued. And that brings us to SF3. At the heart of this accident lies the reactor, or more precisely, the critical assembly, as it is, of course, correctly called, and in fact it was quite small, about 40 cm in diameter, and contained fuel rods of the same height, mounted vertically in the shape of a hexagon. The reactor itself was extremely delicate, with the fuel held in place by a cap at the bottom and supported on top by aluminum plates. The critical node in SF3 had a unique design in other aspects as well. The only thing that stands out about this among other reactors or experiments is the enrichment of the fuel. And it had an incredibly high uranium- 235 content, which was 90%.

[00:02:49]Uranium-235 is the actual fuel of the reactor.

[00:02:54]It is the atoms that undergo fission that create more neutrons to propagate the cycle.

[00:03:00]The reactors used to produce the energy that lights your home use fuel enriched to typically about 2 to 5% uranium-235, with the remainder mostly consisting of the non-fizzy and more naturally occurring uranium-238. 90% fuel enrichment, as you can see, is quite a lot. So why did they have such high enrichment? As it turned out, it was related to an experiment they were conducting. When neutrons are released from fissile atoms or from decay byproducts, they travel at such high speeds that their probability of colliding with another fissile atom and actually causing it to split is very low. But if they are slowed down, the probability that a neutron will cause another fission increases. To slow down neutrons, reactors use moderators.

[00:04:01]The best moderators are atoms that obviously do not absorb neutrons, which defeats the purpose of slowing them down for use in the effusion reaction, and are very small. As with elastic collisions, this maximizes the energy transfer from the neutron to the moderator and the lattice, slowing it down significantly.

[00:04:24]The two most popular moderators historically are graphite, which is carbon, and water, or more specifically, two hydrogen atoms that can be found bonded to an oxygen atom in it. And one more thing, the amount of retarder matters. Too much moderator causes the neutrons to slow down so much that they may not reach the fissile atoms with enough energy to cause a fission reaction.

[00:04:52]Too little moderator results in them not being slowed down enough, which increases the likelihood of capturing a neutron without causing a fission reaction. Therefore, the purpose of this experiment was to compare the number of fuel rods required to bring the reactor to criticality at different ratios of uranium-235 to hydrogen. This was achieved by reducing the distance between the fuel rods or pitch, and thus reducing the volume of water containing hydrogen that could act as a moderator. As can be seen from this table of results, as the distance between the fuel rods decreased, the fuel gradually became more and more unmoderated, and the number of fuel rods required to reach criticality increased dramatically.

[00:05:48]Just a step between the rods from the optimal 14.4 mm leads to a seven-fold increase in the number of fuel rods required to reach criticality. Each time, the experiment required assembling a reactor with significantly fewer fuel rods than the calculated number needed to reach criticality.

[00:06:14]The reactor was then filled with water until completely submerged. And then slowly, taking into account each calculation, a small number of fuel rods were inserted into the reactor, very slowly bringing the reactor to a critical state until it reached it. The assembly was then stopped by inserting the control rods, removing the neutron source in the reactor, and draining the water from the core.

[00:06:47]This process lasted from 20 to 40 minutes.

[00:06:52]The reactor was then reassembled for the next experimental core layout. On May 26, 1971, the experimenters conducted a final experiment with the smallest step between the fuel rods of only 7.2 mm. As can be seen from the table, this required installing 1,790 fuel rods in the assembly. The team overseeing the experiment carefully followed the procedure and extrapolated correctly after every few fuel rods inserted to reach the final count. And once they reached the critical point, the experiment was over. This team was not incompetent, nor were they ignored or left to their own devices. I want you to remember this. They followed the rules and constantly communicated with senior staff at the Kacharttov Institute. For example, there was Andrei Yuri Gagarin, who had by then been promoted to an administrative position, but he continued to visit the laboratories where he had once worked to offer advice and ensure high standards of work. There was also a notable man at the controls that day, although no one at the time really knew of his future, albeit largely unacknowledged, fame. This is Valentin Aubzinsky. Aubzynski has a rather rich history. You see, Aubadzinsky is not the last name he was born with. His father, Ivan Fedorovich Kazanenko, was a tank general who was arrested and executed as part of Stalin's Great Purge. This forced Valentin and his mother, Evdzhaneira Viktor, to flee to Odessa, where Evdzhaneira married Volodymyr Ivanovich Abodzinsky and gave birth to Valentin's half-brother, Valery. Valery and Valentin were complete opposites, with Valery Oenzinsky having a natural talent for singing, a talent he used as a child to help tourists in Odessa, suddenly starting to sing in public places while accomplices stole from the pockets of the scattered crowd.

[00:09:30]Valery graduated from high school and then worked many jobs before eventually using his talent for good, becoming a famous pop singer in the Soviet Union.

[00:09:42]However, Valentin achieved success in science and entered the Moscow Institute of Physics and Technology, and after graduation he got a job at the prestigious Kachiv Institute of Atomic Energy. And from there, he eventually rose to the position of supervising physicist, overseeing the critical nodes they were experimenting on.

[00:10:10]After the experiment was completed, the reaction was stopped, and a group of men, including the experimental team leader Vladimir Erfiev and Valentin Aubzynski, entered the experimental compartment to inspect the assembly. At this moment, Yerafiyev makes a fatal decision, although he does not know what it is about. He orders the water to be drained from the reactor. Previously, this was done by activating a slow water release valve, which took 20 minutes to gradually remove water from inside the unit.

[00:10:51]However, when the experiments were completed, and it was 4:00, almost the end of the working day, Erafiev ordered the use of a rapid or emergency water release valve, which allowed the water to be drained from the reactor in just 30 seconds.

[00:11:10]But suddenly, without warning, the critical reactor collapsed. The support plate under the fuel assemblies inside buckled downward, and the fuel assemblies, along with the water in the critical reactor, fell down and exited the reactor core. The fuel was pushed out from the top, taking on a fan-shaped shape.

[00:11:36]This sudden fall tore the control rods out of the reactor, or rather, tore the reactor away from the control rods. And now, with the fuel rods spaced apart, the reactor was probably in the closest possible fuel- moderator configuration.

[00:11:55]Instantly, the fuel reached a critical state. And not just critical, just like critical for a conventional nuclear reactor. And this is where things get a little technical.

[00:12:11]After the event, calculations showed that the reactivity introduced into the assembly was approximately 2 beta f/s. This is very, very bad. When the reactivity exceeds one beta (f, the effective fraction of neutrons coming from delayed release, such as the decay of by-products of the initial fission reaction), the reactor enters a state where the fission reaction is supported by prompt neutrons—those initially released from the fission reaction—and this is actually enough to increase the neutron population as well. In other words, the number of neutrons, and therefore the energy released, increases exponentially, for example in a boiling water reactor or a pressurized water reactor. This is actually impossible.

[00:13:15]But high fuel enrichment and optimal fuel configuration created a perfect storm. Aubadzinski described hearing a sudden roar as the power rapidly increased, and then losing consciousness.

[00:13:32]Inside the assembly, a sudden release of energy caused structural damage. The fuel rods in the two outer rings around the fan quickly heated up and began to fragment.

[00:13:45]Their remains were later described as resembling the remains of welding rods. This destruction actually knocked them out of the optimal core configuration that had formed, but it was much less significant than the sudden heating of the water in the core, which quickly boiled or splashed out. In a few seconds, the critical situation was over. One of the rooms adjacent to the SF3 installation was called the birdhouse. A usually quiet room on the third floor of a building where a single worker was performing some calculations. He noticed that for several hours it was quiet there, no one entering or leaving the room. This was a bit strange, he thought, since the room had been used by the SF3 experimental team for the past few days. Just then, one of his colleagues burst into the room and headed straight for the safe. The worker asked what his colleague was doing, and he told him that the building had been evacuated for the past 3 hours and the evacuees were sitting outside.

[00:15:05]The worker and his colleague left. He took a colleague, clutching several bottles of alcohol, from the safe.

[00:15:15]Two workers went to the apple orchard on the grounds of the Kachtov Institute, where his namesake, Yegor Kachtov, regularly ate apples straight from the trees, to prove the safety of the facility. This time everything was different.

[00:15:34]The accident resulted in the formation of radioactive aerosols that lingered in the area. At each entrance and exit to the institute there were checkpoints where dosimetrists measured everyone's radioactivity. Those who had their dosimeters triggered had to go through the institute to special decontamination facilities, where they would have to wash again and undergo a scan. However, the Institute's junior staff, members of the Kshot team, did not want to deal with this. They preferred to just wait and drink alcohol. So instead they all sat together in the garden, and every now and then someone was sent to the checkpoint.

[00:16:23]If they were sent back, they knew the aerosols were still here. At 11:00 PM, they sent another one of these so-called messengers. And when he didn't return, they realized that the aerosols were gone and there would be no need for decontamination. So everyone left. It is difficult to say exactly how many people died in this accident. We can safely say that there were at least two of them, and both received a terrible dose. The technician closest to the installation, I. Vasily, received a dose of approximately 60 Severs.

[00:17:08]This is one of the highest doses of radiation I have ever read about.

[00:17:14]Very few people have actually received such a high dose that I can compare it to. Louis Slottin and Hashi Vi received a dose of about 20 sever from the demon's core and settling tank, respectively. The highest radiation dose received by any of the victims of Chernobyl was received by Volodymyr Paravik – just 16 Severs. Boris Kchelov, who stood atop a nuclear reactor that spontaneously went critical on the Soviet submarine K19 in 1961, described seeing blue flames erupting from the reactor and receiving a dose of 54 severi. I have only seen two cases of accidents with the same or higher dose. The dose of 60 Severs received by Cecil Kelly in 1958, and the dose of Robert Peabody, who received a dose of 70 to 260 Severs in 1964. But how long did Vasily live after that?

[00:18:27]We don't know for sure. In Versailles in 1999, one of the witnesses to the accident, who later became a senior researcher at the Kacharov Institute, Dmitry Pavovich, said that Vasyl died the next day of a heart attack.

[00:18:45]That sounds reasonable, considering the dosage.

[00:18:51]However, in the report on this accident, written by the same Pavovich, Vasylv actually lived 5 days after exposure. If that's the case, I can only imagine the agony of those final days, especially knowing that your death was imminent.

[00:19:12]The head of the experimental group, Vladimir Erfiev, was also near the critical assembly and received a dose of 20 Severs. He lived another 15 days before radiation poisoning claimed his second life. If you read the official accident reports published by Los Alamos and the Idaho National Environmental Engineering Laboratory, you will only see these two fatalities. However, there may have been more of them in the works of one of Andrei Gagarin's colleagues, who, as mentioned at some point, worked in this same laboratory and died in 2023.

[00:19:57]Two more names are mentioned, with whom Gagarinsky worked together at some point, and who also died in the accident.

[00:20:05]Sergey Garanin and Rustam Sadikov. It is not stated what dose they received or their roles, and there is no evidence of their existence. Are they real? Did the author mention them incorrectly? We just don't know. And Valentyn Aubzynski, who was in the room, mentioned eight people who received a lethal dose, but here we don't know their names.

[00:20:37]Of course, Aubzynski incorrectly stated the month of the accident in his memoirs, which gave me a big headache when I found out. So it's entirely possible that he's wrong here too.

[00:20:53]Speaking of Aubertsinsky, he and another employee are the only two people known to have been in the room when the critical situation occurred and survived.

[00:21:05]But the dose they received was horrific, somewhere between 7 and 9 units. Even today, this dose is considered almost always fatal. And yet, after months of treatment, both men were discharged from the hospital. But with that dose came restrictions on their future work.

[00:21:27]None of them returned to work on these experimental reactors.

[00:21:33]Aubertsinsky, for example, took a job in the visual information department, and both struggled with health problems for the rest of their lives. So why did this accident happen? Like many other things, it comes down not so much to the hasty actions of the operators, but to the poor design of a critical node. You may have noticed that I never actually mentioned what material the SF3 critical node structure was made of.

[00:22:07]While the upper support structure for the fuel rods was made of aluminum, the support plate that supported each fuel rod was made of Plexiglas. This is a somewhat unconventional choice for a base plate for a nuclear reactor, but the designers had their own considerations. The hydrogen content in Plexiglas is similar to the hydrogen content in water. However, this means that by the end of the experiment, there were 1,790 fuel rods on the plexiglass and a water tank at least 60 cm deep. This is normal, as long as the base under the plexiglass support plate does not disappear. Yet the structural designers failed to consider what would happen if the quick-release valve were opened to remove all water from the structure in the event of an emergency. This relief valve was essentially a giant hole in the bottom of the tank. A Plexiglas support plate almost completely covered this bottom. It almost completely covered the entire bottom of the tank, leaving only a small gap for water to flow around and drain through the drain valve. So when the valve opened, it not only essentially opened a large hole under the plexiglass, but it also created downward pressure on it. This, along with the weight of 1,790 fuel rods, caused the plexiglass plate to suddenly sag downward so much that the fuel rods simply fell out of the upper support plate.

[00:24:00]The rest is history.

[00:24:04]Yerafi could not have known that opening the quick-release valve would lead to the actual collapse of the structure. No one ever warned him, and there was no sign of it.

[00:24:19]Although the scientists who slowly brought the structure to a critical state performed careful calculations, no calculations were ever made on the structure itself. No one had ever tested a quick-release valve before. And so the obvious flaw in the foundation...And so the obvious flaw in the fact that the foundation of the critical node would sag was never noticed. The quick release valve activation action was incorrect. There was no real need for this. But imagine if the same thing happened in an emergency. The causes of the accident were inevitably classified by the Kachhatov Institute, and with it almost all information about the accident itself. In the absence of reports, myths and legends about this accident take their place, and some of these rumors are fascinating.

[00:25:26]According to rumors, the SF3 accident was actually an explosion similar to a nuclear bomb, and it caused the release of a radioactive cloud that covered Moscow, reaching as far as Arbat Street, about 9 km away.

[00:25:44]Others claimed that the explosion was so powerful that it cracked the foundation of the building in which it was located. And yet these rumors never went beyond the Cachart Institute.

[00:26:00]The organization kept aloof and tried not to share its problems with the public or other organizations. These rumors were also not true. It was nothing like a nuclear bomb in terms of structure, timescale, or energy release. Even if something could be compared to a nuclear explosion. The sudden release of energy ruptured the effervescent material before it could reach its potential maximum energy release. There was also no significant release of radiation, except for aerosols that dissipated very quickly.

[00:26:45]The building's foundations also did not crack.

[00:26:48]Of course, the secrecy of this accident ended in September 1999.

[00:26:57]You see, a little over a week before the Tokyo nuclear power plant accident completely overshadowed it, the Sixth International Conference on Nuclear Criticality Safety was held in Versailles, France. There, two senior scientists from the Kachadov Institute, Dmitry Parvanovich and Vladimir Azimolof, revealed information to the West, explaining what happened and how it happened. This sudden release of information was not necessarily about transparency on the part of the then relatively young Russian Federation. No, there was actually a political motive behind it.

[00:27:48]At least one of the two researchers believed that Russia's nuclear safety continued to deteriorate, the work was becoming increasingly dangerous, and better regulations were needed to prevent future accidents. Vladimir Asmolov later explained his opinion with what one journalist later called nostalgia for the totalitarianism that once surrounded the Soviet nuclear program, and lamented its transition from the militaristic Ministry of Mechanical Engineering, or Minseredmash, to the Ministry of Energy. quote: " The construction and operation of our own nuclear power plants requires a level of knowledge that Soviet nuclear scientists lacked in the 1950s and 1960s." This lack of information was compensated for by strict rules, deviation from which could easily lead to death.

[00:28:53]During the 22 years of development of Soviet nuclear power plants, not a single serious incident occurred. And this was possible solely because all nuclear power was subject to the spread of congestion. A defense department with iron discipline and absolute adherence to rules.

[00:29:21]Whether Asimoloff knew about the extremely poor rules and discipline, for example, at Shredmash-controlled Chelyabinsk, which contributed to the excessive exposure of countless people and may even have been linked to the death of Igor Kachtov, the nuclear scientist after whom the Kartov Institute is named, or whether he deliberately ignored these and other places remains a mystery. There remains one more undetermined question.

[00:29:55]Valentin Aubbertsinsky, as mentioned, was removed from experimental work and took on a new role in the visual information department, essentially documenting the work at the Kachottov Institute in video and photographs. And when the Chernobyl explosion occurred in April 1986, Aubadzinsky was sent to the scene before the end of the month, despite the fact that he was initially forbidden from visiting due to radiation exposure from the aforementioned accident.

[00:30:27]Aubzinsky would later say that there was no one who could replace him.

[00:30:34]For months, Aubzinski documented the disaster site and the surrounding area, even flying over the stricken reactor during its active phase, when smoke was still rising from the building. He was on the ground photographing the heroes of the liquidation, among whom was Valery Lagassov, and he was there when the shelter object, better known as the sarcophagus, took shape over the remains of block four. In December 1986, Aubzinski completed another important mission, descending through the intricate network of dark corridors, narrow staircases, and flooded ravines of the sarcophagus. Eventually, he reached his destination and installed the lighting equipment. Finally, when the subject of the photograph was properly visible, he took photographs, dismantled the equipment, and returned to the surface. You may be wondering who or what was the subject of the photograph. None other than an elephant's foot. Years later, a copy of this photo was given to Bill Zuller at the University of Washington, where he was told that this photo had cost a man his life. But this did not happen, because Aubadzinsky did make it back to safety and, although he was forbidden to return to the Chernobyl nuclear power plant, he lived for a long time. His radiation dose during his lifetime was enormous, but he continued to live.

[00:32:20]In fact, Obodzinski continued to take photographs around the exclusion zone for 3 months a year until 1993, when the effects of his high radiation dose finally began to manifest.

[00:32:34]He developed arrhythmia in his right leg. His blood vessels were damaged, and despite several surgeries, there was no way to repair them. In 2005, Aubadzinsky became an amputee, relying on a prosthesis he purchased just before the anniversary of the Chernobyl disaster.

[00:32:56]In fact, he was very proud of it. And now, about 20 years after the Chernobyl disaster and 35 years after the SF3 accident, the story ends. Not by his death, although Aubrzynski is probably already dead today, about 20 years after the last article was written about him. No, this story ends with his legacy and some of the amazing photos he took.