Chernobyl Chernobyl: Options to Limit Reactor Meltdown

[Randy Subers]

TL;DR Summary

Were there other options for the Chernobyl control room operators?

Once the water began to boil running up the reaction rate in Chernobyl unit 4, was there something different than the AZ-5 switch that the operators could have done which would have saved the reactor, or at least limited what happened to a meltdown?

 

[Drakkith]

I can't imagine that anything reasonable could have been done once steam started forming in the reactor. From start to expected finish, the test was only supposed to take 39 seconds, and the SCRAM button was pressed 36 seconds into the test. At best, the operators had about 30 seconds to recognize a problem, understand what was going on, develop a response, and implement it. I'd guess they probably had far less time, perhaps less than 10 seconds, between the steam formation and the core going critical. This just isn't enough time to respond to a crisis involving a highly complex device like a nuclear reactor.

 

[Rive]

As far as I understand the reactor was in really dangerous condition several (tens?) of minutes before the problems were actually noticed/occurred, and from that time it would have been troublesome (if not impossible/unlikely) to bring it to stable shutdown.
The only (slight) chance I can think of was if somebody between 01:05 and 01:10 start to drop control rods into the core, only a limited number at a time and not all together.
Without authorization, of course. Can you imagine that? I can't.

 

[Drakkith]

The only (slight) chance I can think of was if somebody between 01:05 and 01:10 start to drop control rods into the core, only a limited number at a time and not all together.

With the cooling system still fully functioning they could probably have scrammed the reactor without any issue. My understanding is that the water in the core started to boil because the pumps lost power as the turbines spun down. Interestingly, it appears that 4 of 8 pumps were still being powered by the grid during the test, so even with half the pumps working there still wasn't enough flow to prevent boiling and steam formation in the core.

 

[Rive]

With the cooling system still fully functioning they could probably have scrammed the reactor without any issue.

That weird build of the control rods (the power surge it could cause), together with the 20 sec deployment time and the burning up Xenon... All rods simultaneously in a scram...
I cannot know but have some doubts.

 

[artis]

@Randy Subers I suggest you read up on neutron multiplication and the ratio of prompt neutrons to delayed neutrons. In fission this is the main thing which basically determines whether you have a bomb or a reactor.
Everything else comes in secondary.
Think about it in terms of a ratio based around 1. Neutrons have cycles/lifetimes between when they are emitted to the point they get absorbed and cause new fission or get lost. The prompt neutron to delayed neutron ratio for a U235 fuel reactor running in steady state (no power increase or decrease) is something like 0.9935% of prompt neutrons VS 0.0065% of delayed ones.
Delayed neutrons are nothing more than just neutrons released by the fission fragments formed after initial U235 fission into other lighter elements. It takes time for these fragments to decay so the neutrons they release come later, in fact much later , around on average 13 seconds later.
Prompt neutrons on the other hand are very fast, each new generation life time being some 0.001 seconds.
So as you can see by this alone if the ratio of prompt neutrons gets even slightly over 1 then you get a multiplication of neutrons happening exponentially at a rate that is so fast no mechanical action could keep up.
So normally one tries to keep this ration below 1 and rely on the delayed neutrons to finish the job so to speak.

RBMK control rods before 1986 moved rather slowly , it took about 22 seconds IIRC for them to completely insert themselves from a topped out position , this on neutron lifetime scale is like the age of the universe VS a blink of an eye. It's a dinosaur mechanism that moves at snail speed.
This is the reason why there are safeguards in place so that the reactor is never operated outside it's safe limits where the neutron multiplication speeds become "prompt" and the mechanics can't keep up.
This is true for all fission reactors because all fission fuel is capable of this "prompt" reaction rate.
The difference is how different reactor designs incorporate safety mechanisms both active and passive in order to overcome this and not allow for the reactor to ever "run away".
Western PWR and Soviet/Russian VVER rely for example on the properties of water as a moderator whereby a decreased water density - more steam causes the fission rate to slow down.
The drawback of the RBMK was that it had both water and graphite as moderator, so as water boiled to steam graphite took over the moderation and being a good moderator did it's job splendidly.This is the long answer, the short answer is simple, the reactor was ran that day at half power and fission byproducts had built up in the fuel like Xenon which absorbs neutrons, in order to counteract this they pulled out more rods than was allowed or even SANE! The older RBMK being a rather "tolerant to intervention" beast allowed for this if the operators so chose and they did.
Hard to tell exactly at which point it could have been averted but certainly by the time the reactor was burning it's Xenon poison with just a few rods still left in place it had the ability to then just jump in power the moment the neutrons burned away enough Xenon, this moment happened around the time the water star
it just so happened to be that the beginning of the experiment and decrease in water flow rate in the core channels was that moment due to water forming more steam and neutron flux increase associated with it.

 

[artis]

My understanding is that the water in the core started to boil because the pumps lost power as the turbines spun down. Interestingly, it appears that 4 of 8 pumps were still being powered by the grid during the test, so even with half the pumps working there still wasn't enough flow to prevent boiling and steam formation in the core.

I believe this is not the case, they had grid power and sitting at around 200MW thermal at the beginning of the test , 2 pumps out of the 8 would have been enough to give adequate cooling.
The reason steam started to form was different, not being able to go beyond the 200MW mark due to reactor core still being xenon poisoned they (Dyatlov) decided to begin the test anyway. Original papers stated to begin at 700MW.
When they began the test they shut off the steam supply to the turbines. Now what happened was the steam was shut to turbines and the turbine rundown began, but the reactor was still at 200MW and not shut down as it should have been if they began the test at 700MW.
The additional steam having nowhere to go went to the steam separators and then mixed back to the "downpipes" going back to the pump inlets. RBMK being a one loop reactor means it ran the steam from the reactor directly to the steam separator drums and then directly to turbine.
Shutting off the steam to turbine would automatically imply shutting down the reactor as they have nowhere to dump the huge amount of heat.
This heat now being in the small loop was being run through the core, overall flowrate being small the water started to boil rater quickly , I believe they noticed the boiling but given rods were out and xenon was about to be overtaken by neutrons any moment they simply didn't have enough time on their hands to safely shut down the reactor, it seems the water boiling was accompanied almost immediately by the rapid spike in neutron activity and so the prompt criticality began, seeing the reactor power skyrocket in seconds they pressed the AZ-5 but at this point it was a futile attempt and made no real change.
The rest as we know is history, this is the moment of the explosion.For many I believe it is complicated to understand how a simple boiling event could have started a rapid immediate reactor power increase that shot through the roof not just literally (the block exploding with the reactor cover physically flying through the reactor hall roof) but also figuratively. It seems the answer is the fact that the reactor at that point was held back only by excess xenon and some little water density, xenon rapidly being burned away and water starting to boil happening at the same time was enough to then give the moderation over to graphite in the core and this then causing a bomb condition.
The last reading on the screen before reactor core detectors were annihilated was I believe some 30GW, a frightening amount of power

 

[Drakkith]

I believe this is not the case, they had grid power and sitting at around 200MW thermal at the beginning of the test , 2 pumps out of the 8 would have been enough to give adequate cooling.

Well, they had 4 and this apparently wasn't enough given the reactor conditions. Not sure why exactly. Perhaps the extra coolant flow was actually too much? Lower coolant flow might have allowed some steam voids to exist in the core, which might have kept the operators from pulling additional control rods to keep the reactor at around the 200 MW range.

When they began the test they shut off the steam supply to the turbines. Now what happened was the steam was shut to turbines and the turbine rundown began, but the reactor was still at 200MW and not shut down as it should have been if they began the test at 700MW.

What were the procedures to shut down the reactor? Did they differ between the 700 and 200 MW states?

Shutting off the steam to turbine would automatically imply shutting down the reactor as they have nowhere to dump the huge amount of heat.

Indeed, the test procedures appear to have been to shut off the steam to the turbine and power down the reactor. I wonder how far they got into the shutdown process, if at all. The test took less than a minute, so I wonder if the shutdown procedures had even been started at the time of the incident.

For many I believe it is complicated to understand how a simple boiling event could have started a rapid immediate reactor power increase that shot through the roof not just literally (the block exploding with the reactor cover physically flying through the reactor hall roof) but also figuratively. It seems the answer is the fact that the reactor at that point was held back only by excess xenon and some little water density, xenon rapidly being burned away and water starting to boil happening at the same time was enough to then give the moderation over to graphite in the core and this then causing a bomb condition.

It would seem the near-complete removal of the control rods, the high coolant temperature due to high coolant flow, and the ability to quickly burn off neutron-absorbing xenon-135 in the event of a power spike made the reactor essentially a bomb waiting to go off as soon as the cooling water started to boil.

 

[artis]

Well, they had 4 and this apparently wasn't enough given the reactor conditions. Not sure why exactly. Perhaps the extra coolant flow was actually too much?

No I don't think so, because a reactor that has a design operating thermal capacity of around 3000MW thermal would have no problems whatsoever to cool down the core under 200MW thermal.
Remember that before the prompt criticality began the reactor was operating at 200MW.

The accident happened because the core was poisoned and they removed far more control rods than was even allowed, because that was the only way to let the core "breathe" because they were rushing to complete the test and sitting and waiting for the xenon to burn under "safe conditions" would have taken them all night and probably half of next day.
So now the very little flux they had in the core was effectively burning away the accumulated xenon. So essentially the only thing "controlling" the reactor at this point was the amount of xenon and the density of water in the channels. Given then Xenon concentration was depleting but at an unknown rate as there is no direct way to measure it's amount in the core , the only thing they knew and could control is the water flow and temperature in the core channels. But the moment they shut the turbine steam valve is when they lost this ability because without first shutting down the core shutting it's steam exit means that the extra steam now heats up the small loop (core-steam separator- downpipe/pump inlet - core) this then caused the water to start to boil the in channels, the water normally boils at the very top of the channels but as the heat was accumulating the level of boiling began to go lower deeper into the core effectively increasing the flux within the core.
200MW may be a tiny amount of energy for the RBMK core but only if that energy is dumped to turbine, if that is simply circulated around in a loop then it is enough to heat up the core rather quickly.

From what I understand they managed to "hit the jackpot" as the steam began to increase in channels flux also was increasing and at this very time the Xenon was also burned away enough to allow for a sudden increase in activity, and it began.

I also don't know why they did not shut down the core the very moment they shut the turbine steam valve as that was per procedure, the test was always to shut the turbine flow and also the core.
The test was meant to be started from the minimum safe power level of 700MW thermal but they couldn't squeeze that level due to the Xenon poison so they managed to get to 200 and began from there.
In any way they should have shut down the core immediately after cutting the steam.

It could be that the power jump was so sudden that it happened within few seconds and they were about to shut the core but before they managed to push the button it already began.
I don't think they had a reason to keep the core working once they tripped the turbines.

 

[artis]

What were the procedures to shut down the reactor? Did they differ between the 700 and 200 MW states?

The 200 state was outside the limits of what was allowed , they simply decided to go from that because that was all they could get given the conditions they were at.

Lower coolant flow might have allowed some steam voids to exist in the core, which might have kept the operators from pulling additional control rods to keep the reactor at around the 200 MW range.

Yes maybe it would have been better because unlike steam in the channels the control rods have greater absorbing capacity and having more around could have averted the consequences possibly.
But then again having those rods in would have probably killed the core flux as it was already weak. IIRC even with the rods pulled out they barely managed to get to those 200MW from a couple of MW at one point. And this happened because someone inserted too many rods in instead of getting them out by mistake.
You know it's almost unbelievable but the amount of experimenting and mistakes they made makes it even impossible to remember all the things that went on before the accident.
I believe no one in the history of nuclear technology has made so many wrong decisions at one place in a continuous manner.

 

[Drakkith]

Wow. Some more info on the reactor that I didn't know until now. From this article: https://www-pub.iaea.org/MTCD/publications/PDF/Pub913e_web.pdf

The normal core inlet temperature is 270°C and the core outlet temperature is 284°C at a pressure of 7 MPa (approximately 70 atm).

Only a 14 C difference in temperature between inlet and outlet. Wow.

The temperature of water flowing into the suction header of the main circulating pump depends on the rate of steam production of the reactor, since the steam is condensed after passage through the turbine and becomes the cooler feedwater component of the coolant entering the pumps and the core. When the flow of this feedwater component of the coolant is decreased through reduction of the reactor power, the temperature of the coolant at the pump intake and the core inlet increases accordingly.

So, as the reactor power is reduced, the coolant temperature at the inlet increases to keep the water boiling as it passes up through the core so there is steam to power the turbines.

The use of all eight pumps increased the coolant flow rate above that under nominal conditions at full power, reducing the already low steam content of the core. This low steam fraction reduced the friction factor for coolant flow. In addition, owing to the lower reactor power level at that time, the core inlet coolant was only slightly subcooled and, depending on the exact settings for feedwater flow and recirculation flow and the pressure distribution in the system piping, may not have been subcooled at all.

Normally there are only 6 pumps used, three for each side, with the others remaining as emergency pumps.

These conditions led to the onset of boiling at or near the bottom of the core.

Boiling normally occurs near the middle of the core I believe.

If I understand things correctly, the reactor being kept in a low power state around 200 MW for an extended amount of time had the seemingly paradoxical effect of raising the coolant temperature above what it would be at a higher power level. This is a normal function of the way the coolant system works, but it was not designed for the core to stay in such a lower power state for long. Thus, when either the core power spiked, or the coolant flow rate was reduced, the coolant was already very near boiling and could flash to steam very quickly. In the incident both of these happened.

The more I read, the more flabbergasted I am. Nearly every step in the improvised test was a mistake.
Lower power level - bad for reasons stated above.
Turning on extra pumps - bad because it initially reduced the steam voids and caused the operators to pull more control rods before the test.
Pulling control rods to bring the power level up - Bad Bad BAAAD!
Hitting the scram button - BAAAA-*BOOM*

 

[artis]

The temperature of water flowing into the suction header of the main circulating pump depends on the rate of steam production of the reactor, since the steam is condensed after passage through the turbine and becomes the cooler feedwater component of the coolant entering the pumps and the core. When the flow of this feedwater component of the coolant is decreased through reduction of the reactor power, the temperature of the coolant at the pump intake and the core inlet increases accordingly.

Just to add, this is a logical result because the RBMK is a single loop design, so your core outlet mix goes directly to the steam separator where any steam rises up and escapes to turbine while anything less than steam falls back down and then directly flows down under gravity in the so called "down pipes" and into the intake of the pumps.
The steam is normally hotter than the water falling back down but the steam also rises up and then travels a much larger distance through the turbine and because of this it eventually manages to cool off more than then water directly falling back to the pumps, even though the steam was at first hotter, it eventually manages to become cooler as it had more time and surface area to do so while the water did not.

If I understand things correctly, the reactor being kept in a low power state around 200 MW for an extended amount of time had the seemingly paradoxical effect of raising the coolant temperature above what it would be at a higher power level. This is a normal function of the way the coolant system works, but it was not designed for the core to stay in such a lower power state for long. Thus, when either the core power spiked, or the coolant flow rate was reduced, the coolant was already very near boiling and could flash to steam very quickly. In the incident both of these happened.

Yes I believe you got the basic idea right. The small loop was not designed to work as a heat exchanger. That being said It radiated off heat rather well compared to other types of reactors due to it's large surface area. Think of the RBMK core as a giant pipe organ the size of a highrise building , then add the distribution pipes, down pipes, steam separator drums which were the size of two end to end railway locomotives and the pumps and connecting pipes , this whole structure I believe was enough to waste the heat as long as some of it was allowed to escape as steam to the turbines.
The real problem started the moment they shut off the turbine but did not shut down the core...

They were operating at low power that night for quite a while before the test if that alone could cause the core channels to flash to steam it would have happened earlier I think. Yet it happened just moments after shutting the turbine exit, of all the mistakes I believe this was their biggest.
As far as I am aware no reactor ever anywhere is allowed to be operated if the turbines are completely shut off, and they were as that is the very idea of a "rundown test" .

So yeah these few peculiar conditions like not enough neutron absorbing reserve and Xenon rapidly decreasing and then killing the steam exit causing a sudden water flash to steam event all came together just in the right time. If just one of those was avoided the whole outcome could have been different.

 

[Astronuc]

If I understand things correctly, the reactor being kept in a low power state around 200 MW for an extended amount of time had the seemingly paradoxical effect of raising the coolant temperature above what it would be at a higher power level. This is a normal function of the way the coolant system works, but it was not designed for the core to stay in such a lower power state for long. Thus, when either the core power spiked, or the coolant flow rate was reduced, the coolant was already very near boiling and could flash to steam very quickly. In the incident both of these happened.

It was the reduction in coolant flow that lead to an increase in coolant temperature. In the core, the thermal energy from the fuel is transferred to coolant. The thermal power, P, must be balanced by the coolant flow rate and change in specific enthalpy $\dot{m}$ * Δh from inlet to outlet. As the coolant flow rate decreases, the change in enthalpy increases, and below a certain value of flowrate, the liquid becomes vapor (where specific enthalpy exceeds the saturated value at a given pressure).

As I recall, a number of safety systems were disabled. The operators attempted to insert control rods, but they had graphite tips (or followers) such that they added more reactivity to the system. Prior to prompt critical, they would have seconds to respond, but as one approaches prompt critical, the response time decreases. At some point, the core went prompt supercritical, and the transient occurred in milliseconds. Prompt critical or prompt supercritical are prohibited states for a power reactor. Only certain special reactors (with pulsing capability) are allowed to achieve prompt critical or prompt supercritical.

Example of a reactor pulse -

 

[Drakkith]

It was the reduction in coolant flow that lead to an increase in coolant temperature.

Before, or during the incident?

 

[Astronuc]

Before, or during the incident?

During the incident, i.e., fateful (or fatal) experiment, leading up to the transient.

Chernobyl Unit 4 was taken so far outside the domain of safe operation that it was not an accident but an act of criminal negligence, IMO.

 

[anorlunda]

During the incident, i.e., fateful (or fatal) experiment, leading up to the transient.

Chernobyl Unit 4 was taken so far outside the domain of safe operation that it was not an accident but an act of criminal negligence, IMO.

That's a much better way of expressing it.

Many people struggle with the idea that the design and operation could be so bad as to allow that accident. That a plant could be vulnerable to criminal sabotage is more understandable.

 

[artis]

Chernobyl Unit 4 was taken so far outside the domain of safe operation that it was not an accident but an act of criminal negligence, IMO.

But truth be told this reactor was especially capable of becoming a "monster" if taken out of safe limits just by a little bit. I also partly blame the designers , because if the brake pedal in certain situations (transients) functions as a gas pedal for a short while before it becomes a brake pedal then that alone is "asking for trouble"
I remember reading one Russian former senior operator describe the handling of transient modes of RBMK like "playing a piano concert" by which I suppose there is emphasis on the attention to detail in every step.

They saw this back in december 1983 when Ignalina 1 came online. They performed some tests but nothing outside safe limits and they too noticed the flux increase effect of automatic shutdown rod insertion.A much less known accident and at the time entirely concealed was 11 years before Chernobyl , in 1975 on 30th November at the Leningrad NPP first block, also the first ever block of RBMK-1000, they scheduled to take one of the two turbines offline for repairs. They decreased the power to the turbine according to plan and dropped reactor power to half, then by mistake the control room operators disconnected the running turbine from the grid. The turbine valves shut closed to prevent the turbine from catastrophic overspeed and failure, this in turn tripped the reactor itself automatically.
The administration was mad surely since all generating capacity was lost and from what I understand there was a push towards the operators for them to try to immediately restore reactor power at previous level.
So they did , while doing so going outside the safe limits ofcourse.
The senior reactor operator at the time was Mikhail Karrask, who before LNPP worked at one of the Soviet plutonium reactors in Siberia.
I am translating this all to English from Russian since I can't find good English sources.
So basically the core was reluctant to come back on after the shutdown, so they managed to pull out almost all the rods (much like in Chernobyl) some sources say they at one point removed all but few.
They managed to get the power back to about 800MW thermal, but not without a fight, as the reactor shutdown again two times in the process after which they disabled the automatic shutdown and went purely manually. From what I read they even managed to "hack" the core detectors as they signaled of high distortions in axial and radial flux in the core which tripped the core shutdown the first 2 times.
Then finally getting to 800MW as said, they experienced a sudden climb to 900MW in a few seconds, at this point Karrask took the decision to abandon all efforts and shutdown the core for good.
But as he did so he did not use the AZ-5 button instead he went manually and slowly , he inserted rods in pairs but not all at once. Only after the rods were almost inserted he pressed the automatic button and let the automation insert all leftover rods at once.
He later said that his intuition made him do this slowly as he had seen power jumps doing transients before in the RBMK.

But all was already too late as during that 100MW jump the flux was highly uneven and one of the core channels burst spraying water/steam inside the sealed reactor chamber/barrel. neighboring channels were affected. The pressure build up necessitated the venting of gas overpressure through the stack , this released radioactive aerosols into the nearby areas, the fuel rods in the affected channels burst , so in total one pressure tube burst and the fuel elements inside that tube while also fuel elements in other intact tubes were damaged even though the tubes themselves survived.
I can't recall whether it was after this incident but I know they at some point opened up the reactor at LNPP 1st block and changed graphite stacks as the graphite had swollen too much and started pressurizing the core channels.
From what I understand one of the drawbacks of the huge physical size of the RBMK core was that it even at normal operation modes could develop an uneven Xenon concentration and neutron flux within the core, it seems that in this 1975 incident this was also the case, so after restart of the core the flux was uneven, it came back better at some parts of the core while was almost dead at others.

All in all this 1975 incident was basically the same background as Chernobyl , they got lucky by the fact that the core had different fuel burnup/composition (from what I read the fuel was new) and the operator was cautious and intuitively moved slow to a condition where others might have become scared and "jumped the gun" by dropping all rods as fast as possible. Also their water situation might have been better without the large steam content as was in Chernobyl , but I'd say they were close from what I read.But info on this is scarce as it was largely and successfully concealed. Not sure whether IAEA agency was every briefed about it.

 

[anorlunda]

Designs of all kinds have properties. We can debate the merits ad nauseum. But the presence of a person on site with so much excess authority that he can brush aside all safeguards, all restrictions is itself an unacceptable risk.

We recognize that risk in other fields. For example protocols that prevent a rogue head of state impulsively launching a thermonuclear war that would end the world. On a less dramatic scale, we see it in the doctrine of cockpit resource management in airliners, where the first officer is expected to challenge or even override the captain in case the captain is endangering the aircraft.

 

[Astronuc]

A much less known accident and at the time entirely concealed was 11 years before Chernobyl , in 1975 on 30th November at the Leningrad NPP first block, also the first ever block of RBMK-1000, they scheduled to take one of the two turbines offline for repairs. They decreased the power to the turbine according to plan and dropped reactor power to half, then by mistake the control room operators disconnected the running turbine from the grid. The turbine valves shut closed to prevent the turbine from catastrophic overspeed and failure, this in turn tripped the reactor itself automatically.

If one has some sources, I'd be interested. It is not well-known in outside of Russia.
A Wikipedia article on Leningrad NPP gives a poor summary.
https://en.wikipedia.org/wiki/Leningrad_Nuclear_Power_Plant#Incidents_and_accidents

On 28 November 1975, a fuel channel in Unit 1 suffered a loss of coolant, resulting in the degradation of a nuclear fuel assembly that led to a significant release of radiation lasting for one month. Immediately after the accident, the radiation level in Sosnovy Bor, 5 kilometres (3.1 mi) from the affected power unit, was 600 mR/h; in total, 1.5 MCi was released into the environment. The exposed inhabitants of the Baltic region were not notified of the danger. The accident was not reported in the media.

Leningrad NPP Unit 1 began commercial operation 1974-Nov-01, so it only had one year of operation, but long enough to build a modest inventory. They did manage to obtain 44 years of operation.

But all was already too late as during that 100MW jump the flux was highly uneven and one of the core channels burst spraying water/steam inside the sealed reactor chamber/barrel. neighboring channels were affected. The pressure build up necessitated the venting of gas overpressure through the stack , this released radioactive aerosols into the nearby areas, the fuel rods in the affected channels burst , so in total one pressure tube burst and the fuel elements inside that tube while also fuel elements in other intact tubes were damaged even though the tubes themselves survived.

Yes, the RBMK is inherently, potentially unstable, hence the concern about lack of a reinforced concrete containment.

I can't recall whether it was after this incident but I know they at some point opened up the reactor at LNPP 1st block and changed graphite stacks as the graphite had swollen too much and started pressurizing the core channels.

That happens in graphite moderated reactors, but probably was much later. It shouldn't happen in one year.

From what I understand one of the drawbacks of the huge physical size of the RBMK core was that it even at normal operation modes could develop an uneven Xenon concentration and neutron flux within the core, it seems that in this 1975 incident this was also the case, so after restart of the core the flux was uneven, it came back better at some parts of the core while was almost dead at others.

During steady-state operation, Xe-135/Xe-135m reaches an equilibrium concentration. When power (and neutron flux) is reduced, Xe-135/135m build up through decay of I-135 (half life ~ 6.58 hrs) while a reduced burnout (by transmutation) of Xe-135. It takes many hours of the Xe-135 (half life ~9.14 hrs) to decay. If a portion of the core has reduced power (by control rod insertion), then Xe-135 will accumulate in the fuel nearest the control rod (at lowest power). When the control rod is withdrawn, the increase in neutrons absorbed by Xe-135 will burn out the Xe. A similar phenomenon occurs in BWRs.

All in all this 1975 incident was basically the same background as Chernobyl , they got lucky by the fact that the core had different fuel burnup/composition (from what I read the fuel was new) and the operator was cautious and intuitively moved slow to a condition where others might have become scared and "jumped the gun" by dropping all rods as fast as possible. Also their water situation might have been better without the large steam content as was in Chernobyl , but I'd say they were close from what I read.

The Leningrad event in 1975 is significantly different in some important ways despite any similarities. For one, the fuel at Chernobyl was older, so it had a great fissile-Pu inventory. Leningrad didn't disable critical safety systems as happened at Chernobyl. Chernobyl had a much greater power reduction and a very different reactivity distribution.

I believe Mikhail Karrask was much more experienced and cautions, and more professional than Dyatlov. Safe operation of a reactor should not rely on anyone individual.

Original papers stated to begin at 700MW. When they began the test they shut off the steam supply to the turbines. Now what happened was the steam was shut to turbines and the turbine rundown began, but the reactor was still at 200MW and not shut down as it should have been if they began the test at 700MW.

The test at Chernobyl 4 should have been postponed or canceled at that point, since they were outside the test plan.

 

[artis]

If one has some sources, I'd be interested. It is not well-known in outside of Russia.

Yes of course I can give you the sources, but before I do I want to clear some stuff. Not only is it not well known outside of Russia but it is also not well known inside of Russia... Honestly nobody really knows, this makes it very hard to get a grip on the real chain of events.
Pretty much every secret accident that happened in the USSR before the mid 80's was effectively concealed (Chernobyl being the first large pioneer in terms of how much info got out). The combination of "climate change" and political ice melting in USSR with the scale of the accident and European detectors going off in alarms are to thank for that we know anything at all about Chernobyl, HBO Chernobyl series would never exist if not for this.

Let me clear out what I think we know and what we can only speculate about.
So first a list of stuff I am sure about, things we do know.
1) Leningrad NPP was the the first large scale commercial RBMK reactor , namely block 1 being the first, opened in the 1973, and put into continuous operation from 1974.
2) There was an accident on the night of 30 November in 1975.
3) Some radiation was released , but how much and over what area we can only speculate as that was never disclosed.
4) All sources , even the ones that disagree on other details , agree that the release was due to a burst in some fuel assemblies and one technical channel pipe. Contamination of the sealed core barrel happened. Excess gasses were vented to atmosphere.
5) Although unclear what specific tasks were undertaken to remedy the cause,(possibly the broken channel was permanently closed via it's manual valve and the surrounding channels were repaired or closed, graphite stack was vented with nitrogen and the moisture content vented) it is know that this 1st block was put back into operation and eventually served it's full design life and beyond and was closed only recently, last year or in 2019 IIRC. The graphite replacement happened at 2 of the 4 blocks (not sure which ones) but not in the 70's I checked.
6) Being the first block of an RBMK-1000 reactor the LNPP block 1 underwent testing , various modes were tried out on it. The staff was assembled from former plutonium reactor workers or industry professionals. It is unclear whether the 1975 accident happened due to an operator fault as I said earlier or during one of the tests of the reactor at low power operation.
What is interesting is that the Soviets made changes to the RBMK "on the go", there was no specified test facility, the LNPP 1st block as used both for power production as well as for various tests for necessary updates. Much like Soviet VVER designs were all built and tested at the "Novovoronezh NPP facility" were every new model first was built and operated.
https://en.wikipedia.org/wiki/Novovoronezh_Nuclear_Power_Plant
7) After the accident , changes were made for RBMK models , including those already in existence. Control rod count was increased from 179 to 211. The minimum rod in core at all times count was increased to 15.
Also from what I read a system was introduced for automatic core flux shaping , it seems before 1975 on the first block it was done manually.
8) Based on the known information one can reasonably suspect that the burst of the channel pipe and rupture of fuel assembly rods indeed happened due to local overload/power surge in the core

Things I am not sure about (not sure whether anyone is...)

1) The reason for the reactor accident in 1975, whether it was really due to the accidental disconnection of the working turbine or part of a low power stability test
2) Method by which the sudden excursion was suppressed , whether it was done automatically as some suggest or indeed manually with great care to avoid runaway.
3) It is possible that due to the slow and labor intensive early controls of the RBMK the Xenon effect on flux was very uneven assuming low power mode and manual flux shaping which upon core power increase might have created the necessary conditions for a runaway situation in a isolated part of the core were conditions were met for such event.
The similarity to Chernobyl would be that in 1986 the conditions for such runaway were met instead in the entire core at once as opposed to at some local part of the core as in 1975. This would clearly explain the similarity of the rupture but the difference in the energy and scale of it (couple of channels VS entire core).Here are the sources. For an English speaker a good translator is needed, except for one where there is an English version.
http://accidont.ru/ENG/LAES.html
https://ru.wikipedia.org/wiki/Авария_на_Ленинградской_АЭС

Here is a blog about the comparison between Chernobyl
http://nuclearno.ru/?id=11302
https://cont.ws/@Kamenski/1526274

Here is Karrask's supposed personal memories from it.
http://memory.biblioatom.ru/persona/karrask_m_p/karrask/

Here is an anonymous Russian video with english subtitles retelling the story but with some inconsistencies , although the part at the end about one of the workers remembering the staff being in radiation protection suits after the accident I have read also elsewhere.


Here is a official video from Rosatom (Russia nuclear authority) in the beginning there are exempts from Leningrad 1st block closure, and new VVER openings and other stuff , also with English subtitles


And a 360 video of LNPP second block closure.
Both 1st and second LNPP blocks of the RBMK -1000 type have now been closed permanently after 45 years of operation.