Positive Scram Effect
When operators pressed AZ-5 at 01:23:40, the emergency shutdown system increased power instead of decreasing it. This demo shows why.
01:23:40, Before AZ-5
The reactor is at 200 MWt with nearly all control rods withdrawn. The Operating Reactivity Margin is only 6-8 equivalent rods, far below the required minimum of 15. The operators are about to press the AZ-5 emergency scram button.
The 3-Second Window
When all 187 top-entry control rods begin inserting simultaneously, each rod's graphite displacer enters the lower core before its boron absorber. For roughly 3 seconds, the combined positive reactivity from all displacers outweighs the absorption from the few rods that have traveled far enough for their absorber sections to reach the core.
The system-wide positive reactivity spike of approximately +0.5$ was enough to push the already-unstable reactor from delayed criticality into prompt supercritical. Power surged from 200 MWt to over 30,000 MWt in under 4 seconds, a 150-fold increase.
The Cruel Irony
AZ-5 is the emergency shutdown, the last line of defense. It was designed to save the reactor. Instead, because of the graphite displacer design flaw, pressing AZ-5 with the rods nearly fully withdrawn was the direct trigger for the explosion. The operators did exactly what procedure called for in an emergency. The reactor's design turned their correct response into a catastrophe.
After the accident, the RBMK design was modified: the graphite displacers were shortened and the lower water column was extended, eliminating the positive scram effect. But for Unit 4, this fix came too late.