TRACE THE
FAILURE
NUCLEAR ENGINEERING

Chernobyl’s shutdown button.
It made things worse.

The emergency system was supposed to stop the reaction. Under the accident conditions, its first movement could make part of the core more reactive.

Follow the evidence
AI-generated episode illustration juxtaposing a reactor scene and an emergency shutdown button
AI-generated illustrative artwork from the episode package; not an archival photograph or a verified reconstruction of every detail.
THE COMPANION FILMChernobyl Disaster: The "Safety" Button That Made It Worse Watch episode 3:25

Why could shutdown make the reaction worse?

Chernobyl’s RBMK control rods included graphite displacers below their neutron-absorbing sections. With a rod fully withdrawn, water remained below the displacer in its channel. As insertion began, graphite could replace that water in the lower core and initially add local reactivity, before the absorber’s shutdown effect developed there. INSAG-7, section 2.2.

The size and consequence of this effect depended on the reactor’s condition and power distribution. It was not a rule that every shutdown of every RBMK would produce an accident. INSAG-7 describes a dangerous concurrence of reactor physics, rod design, operating conditions and failures in the wider safety system. INSAG-7, conclusions.

Watch the explanation

See the distinction between the neutron-absorbing part of a control rod and its graphite displacer.

The absorber and the graphite displacer had different jobs

A neutron absorber reduces the chain reaction by capturing neutrons. The graphite displacer was a different part of the assembly. Its position changed how much water occupied the channel as the rod moved. Calling the assembly merely a “graphite-tipped rod” can hide the geometry that made its first movement dangerous.

INSAG-7’s diagram shows the fully withdrawn pre-accident rod with the graphite displacer within the fueled region and water at its ends. As the assembly moved down, the displacer replaced water at the bottom. Because water absorbed neutrons in this arrangement, replacing it with graphite increased reactivity locally. INSAG-7, section 2.2 and Figure 1.

The emergency rods took time to travel into the core. Reactivity could rise during that movement, before the absorber completed its intended action. The assembly’s behavior throughout insertion determined how it affected the reactor.

Why the reactor’s condition mattered

Steam formation supplied another possible positive feedback. In the accident configuration, the reactor had positive void and power coefficients: an increase in steam voids could increase reactivity, encouraging more power and more boiling. The graphite moderator continued doing its work while the amount of neutron-absorbing water changed. INSAG-7, section 2.1.

The reactor’s operating state and control-rod configuration made this feedback especially dangerous. INSAG-7 found that important safety information was poorly communicated and that procedures and instrumentation were inadequate. It also criticized operator actions, identifying failures in both the equipment and its operation. INSAG-7, sections 5.8 and 6.

INSAG-7 says it is not known with certainty what started the destructive power excursion. It describes the added reactivity from rod insertion as probably a decisive contributing factor. INSAG-7, conclusion 4.

What the investigation said needed to change

INSAG-7 reports changes aimed at reducing positive void feedback, improving shutdown speed and replacing the control-and-safety rods with a design that eliminated the lower water columns. It also records revised operating instructions and additional emergency protection. These are reported post-accident measures in the 1992 document, not a claim that every safety problem had been solved. INSAG-7, appendix.

A shutdown system needs to be assessed throughout its operation, including its first effects on the reactor. That assessment also depends on the reactor’s state, what operators know about it and how known defects are handled.

The source record

IAEA INSAG-7: The Chernobyl Accident, Updating INSAG-1

1992 · Printed pp. 3–6 / PDF pp. 13–16: void feedback and rod geometry. Printed pp. 16, 22–25 / PDF pp. 26, 32–35: assessment and uncertainty. Appendix pp. 27–28 / PDF pp. 37–38: reported modifications. Figure 1, printed p. 5 / PDF p. 15, inspected visually.

Research checked September 21, 2026. Illustrative episode images are not primary evidence.