Contributing factors towards Chernobyl accident

[Methias]

Hi, recently I'm doing some reading up on the Chernobyl accident, then I came across what INSAG-7 reported in their findings to be contributing factors leading up to the accident. I've understood that mainly the design of the reactor core and the control rods played a huge part in it. However there were also mentions of violation of safety procedures like conducting the test at lower levels than planned. Was this part crucial in the build up of thermal runaway?

Your insight is much appreciated, thanks

 

[.Scott]

Yes, it way a huge factor.

Unlike US (and most other) reactors, the Soviet reactors could not be shutdown and cooled down rapidly without thermal stresses breaking the piping.
The experiment they were conducting was to investigate this realm of the reactor operation. In particular, the point of the experiment was to determine if residual turbine spin could generate enough electric power to keep the cooling system running as the reactor was shutting down.

The standard for the planning of such an experiment should have been quite detailed and conservative. It was not. It was much more of a "see what happens" type experiment. This was the Soviet's most decorated reactor control group - and they were acting like cowboys. In many descriptions, you will hear terms like "unexpected power surge". Everything that happened was within their abilities to model and predict. And because the reactor design was less conservative that western designs, that type of modelling and predicting was absolutely necessary for safe operations. But design engineering assistance was not requested and no criteria was established ahead of time for when to abort the experiment.

Finally, in order to keep the reactor from shutting down completely, they shut down several safety systems. Then, when the remaining alarm signaled, they ignored it for 20 seconds.

BTW: Soon after that accident, a group from MIT and another US university went to Russia and met with the investigators there. On return, the MIT group put together a very detailed technical presentation that was open to anyone who wished to attend. Only two people attended, myself and a professor from Lowell Tech (now UMass Lowell). As we worked through the sequence of events, we were able to explain based on the reactor design every detail of what happened - including the double explosion.

 

[QuantumPion]

They had limits to the amount of control rods they could have withdrawn and the amount of coolant flow, which they willingly violated, instead of aborting the test due to xenon poisoning. These were the main factors causing the accident.

 

[.Scott]

I just read the wikipedia article. As you read through it, look at all of the unplanned, on-the-spot decisions that were made to keep the reactor operating and push the experiment forward. Every one of those made the reactor less stable. Look at how many alarms were sounded and ignored.

As I said, they were acting as cowboys.

When evaluating the relative importance of the reactor design and the reactor operation to causing the disaster, consider whether any reactor design is capable of safe operation when the operators start disabling safety systems and manually retracting the most critical control rods.

 

[Methias]

Yes, it way a huge factor.

Unlike US (and most other) reactors, the Soviet reactors could not be shutdown and cooled down rapidly without thermal stresses breaking the piping.
The experiment they were conducting was to investigate this realm of the reactor operation. In particular, the point of the experiment was to determine if residual turbine spin could generate enough electric power to keep the cooling system running as the reactor was shutting down.

The standard for the planning of such an experiment should have been quite detailed and conservative. It was not. It was much more of a "see what happens" type experiment. This was the Soviet's most decorated reactor control group - and they were acting like cowboys. In many descriptions, you will hear terms like "unexpected power surge". Everything that happened was within their abilities to model and predict. And because the reactor design was less conservative that western designs, that type of modelling and predicting was absolutely necessary for safe operations. But design engineering assistance was not requested and no criteria was established ahead of time for when to abort the experiment.

Finally, in order to keep the reactor from shutting down completely, they shut down several safety systems. Then, when the remaining alarm signaled, they ignored it for 20 seconds.

BTW: Soon after that accident, a group from MIT and another US university went to Russia and met with the investigators there. On return, the MIT group put together a very detailed technical presentation that was open to anyone who wished to attend. Only two people attended, myself and a professor from Lowell Tech (now UMass Lowell). As we worked through the sequence of events, we were able to explain based on the reactor design every detail of what happened - including the double explosion.

From what I can understand, does it mean that instead of the originally planned power level, they agreed on a lower power level (which should be 200MW if I'm not mistaken) to conduct the test. Then at this lower power level, they shut down the reactor and with the residual power output they were supposed to test if that power output would power the cooling system enough so that the reactor can be safely shut down? Would I be correct to say that in the end the cooling system was insufficient to cool the reactor, and coupled with factors like the positive scram effect and high positive void coefficient, the fuel channels ruptured?

Would there be any way I can get hold of any publications or papers by the group of MIT regarding this presentation you mentioned online?

 

[QuantumPion]

They didn't "agree to" conduct the test at a lower power level. They were originally going to conduct the test at 20% power, but due to delays, xenon built in and reduced the core power to 1%. They managed to get back up to about 6% power by bypassing all safety limits on coolant flow rate and control rod withdrawal. They did not have enough excess reactivity to overcome xenon, but tried to anyway, resulting in an unstable and extremely dangerous condition.

 

[.Scott]

Would there be any way I can get hold of any publications or papers by the group of MIT regarding this presentation you mentioned online?

I don't know. I did walk away with a half-inch thick stack of reports from that MIT meeting. I'm sure I never threw them away - so they are probably stored away in a box somewhere. Of course, all this predated the regular use of the internet. That meeting was originally described as a review of Three Mile Island, but because of Chernobyl and their visit to the SU, it became a Chernobyl meeting. It was part of MIT's annual open house, and this presentation was held in MIT's own reactor building.

I read through the wiki article (https://en.wikipedia.org/wiki/Chernobyl_disaster) this morning and have since looked at the INSAG-7 document (http://www-pub.iaea.org/MTCD/publications/PDF/Pub913e_web.pdf) that you already have. Between the two, there are only a few things that I can remember that I do not see in either the wiki article or the INSAG-7 report:
1) It left out that this was the award-winning, top-performance reactor site in the Soviet Union.
2) At the meeting we convinced ourselves that the neutron surge would have generated enough hot Xenon to power that second explosion. But there have been subsequent computer models that should be more authoritative in this point.
3) There may be some reactor design items that are not include INSAG-7 report. I have not read that whole thing, but so far I don't see anything about the reasons that prevented that reactor from being shutdown as quickly as western reactors.

On the other hand, there is quite a bit in the wiki article that I don't recall from the meeting. For examples:
1) any debate about whether the AZ-5 button was hit;
2) The fact that there had been similar experiments in the past.

From what I can understand, does it mean that instead of the originally planned power level, they agreed on a lower power level (which should be 200MW if I'm not mistaken) to conduct the test.

Basically yes. Their plan was to start with the reactor at 700MW or more. But when the reactor dipped way below this they did crazy things to get it back up - eventually to 200MW.

Then at this lower power level, they shut down the reactor and with the residual power output they were supposed to test if that power output would power the cooling system enough so that the reactor can be safely shut down?

No. The first part of the experiment involved changes to coolant flow. So they were trying to hold that 200MW while the experiment was underway. This is when they manually removed all but 18 of the control rods - even though 28 was the absolute "fail-safe" minimum. This was unimaginable insanity. During my pilot training I was introduced to the often fatal disease of "go-itis". It's when you make pilot decision based on something other than the safe conduct of the flight. It appear to me that, at this stage, they were suffering acute, terminal, go-itis.

Would I be correct to say that in the end the cooling system was insufficient to cool the reactor, and coupled with factors like the positive scram effect and high positive void coefficient, the fuel channels ruptured?

No, not at all. The cooling system was important because it was operated in a way that exaggerated the instability of the reactor. But by the time the cooling system was overwhelmed, the reactor was already fully beyond control.

 

[Methias]

Regarding the cooling system, from what I read that is, the slowing down of the cooling system due to the lower power output facilitated the void formation, thus the positive feedback loop?

 

[.Scott]

Yes. That was an aggravating factor.

 

[Methias]

Thank you! As an electrical engineering undergrad, it took me quite some time to make sense of all the new fancy nuclear terms and procedures.

 

[Astronuc]

Regarding the cooling system, from what I read that is, the slowing down of the cooling system due to the lower power output facilitated the void formation, thus the positive feedback loop?

In a graphite moderate reactor, like the RBMK, void of the coolant reduces absorption of neutrons by the water, and that adds positive reactivity to the core.