Video compilation here of several nuclear reactors starting up:
The light is obviously Cherenkov radiation:
What interests me is the extreme transient nature of it. Why does the light output ramp up (or down) with such extreme rapidity?
From what I understand, which ain’t much, a nuke plant is unstable. It has to be continually adjusted as if it’s just a bit too high, it grows exponentially, too low and it dies exponentially. Upon startup it goes from near zero to noticeably ‘on’ which is already runaway zombi-apocalypse heading to a melt down - warp cord breach but not really observable till it’s too late and it happens so fast that it’s hard to design a mechanical system to limit it with a feedback loop.
Production nuclear reactors are certainly not designed to be dynamically unstable, at least not against the sorts of conditions the reactor is expected to see. They’re stabilized by physics, geometry, and general core design before any mechanical or electrical control systems are considered.
For instance, the pressurized water reactors I operated were stabilized against runaway fission by the fact that the cooling water would heat up, which caused the water to be less dense, which caused it to reflect fewer moderated neutrons back into the core, causing the fission rate to slow down, causing the water to cool down, …
All of this happens far faster than any electromechanical control process and certainly faster than any human could control.
My reactors were stable against boiling, as steam moderates and reflects neutrons much worse than water. Other reactors designs (notably Chernobyl and Fukushima) are not stable against boiling and other aspects of the plant need to be designed for stability. Chernobyl was moderated by carbon blocks; Fukushima was designed to have boiling; there is nothing fundamentally wrong with either of those choices. For any energetic system, if it goes outside its design envelope, all bets are off.
All of those clips in the OP appeared to be research reactors, so who knows what they were designed for.
My cores were not in swimming pools and did not have video cameras pointed at them, but based on the detectors we monitored, the fission rate increased smoothly (albeit exponentially) during startup. There was no clank-glow. I could easily be wrong, but if I had to guess, those were prompt-criticality or power burst experiments or something similar.
If the reactor is stabilized by the cooling water, then starting it up will release a large amount of radiation, which continues until the water heats up enough to slow the reaction rate down. Depending on the time constant, this could take fractions of a second, so there is a flash of light, and then the light dims to a study glow.
TRIGA-type reactors “flash” because they are designed for pulsed operation. As soon as the fuel temperature goes up, the reactivity goes down, shutting off the reaction. Pulsed operation is possible at up to 9000 MW, compared to 5 MW steady-state operation.
For transient (pulsing) operation the reactor is first brought to criticality at a low power level (usually 10 to 300 watts). This is accomplished by drawing only the standard control rods and leaving the transient rod in the core. Some facilities bring the reactor critical with all controls rods and then lower the transient rod to produce a subcritical pulse. Reactivity insertion is calculated by evaluating the differential transient rod worth curves and selecting the transient rod height. The pulse is then produced by rapidly firing the transient rod out of the core pneumatically within a fraction of a second. The reactor power will increase sharply to a value that will result in a fuel temperature increase, which will compensate for the excess reactivity inserted. At this point the peak reactor power will occur. The fuel temperature will continue to rise resulting in an increasing loss of reactivity, and the reactor power will decrease to a comparatively low level. This final power level depends on the reactivity insertion and the heat transfer characteristics of the fuel element.