DEAD DROP ARCHIVES
Episode 4 — Chernobyl: The Forty-Second Flaw
AZ-5, the positive void coefficient, and INSAG-7 — reconstructed from the primary record.

Transcript


April 26th, 1986.

01:23 in the morning.

Chernobyl Nuclear Power Plant. Reactor Number Four. Ukrainian SSR.

A safety test is running.

The shift running this test had not been briefed on the procedure.

01:23:04.

Test phase initiated. Reactor power: 200 megawatts thermal.

01:23:40.

Emergency shutdown ordered. The AZ-5 button is pressed.

Four seconds later —

Reactor power surges to approximately 100 times the reactor's nominal design capacity.

01:23:44.

First explosion. Reactor vessel destroyed.

Forty seconds.

AZ-5 is the emergency shutdown button.

Pressing it is designed to stop the reactor.

The Soviet investigation blamed the operators.

The IAEA's revised report, published five years later, did not.

What the documents show is the subject of this file.


This is Dead Drop Archives. A forensic documentary series.


PHASE 1 — THE MACHINE

The RBMK-1000.

A Soviet graphite-moderated channel-type reactor. Boiling light water. Direct steam cycle. A nominal capacity of 3,200 megawatts thermal — roughly the output of a large coal power plant.

By April 1986, sixteen RBMK reactors were operating across the Soviet Union. Chernobyl Unit Four had been generating electricity for the Ukrainian grid since 1983. It was not experimental. It was not unusual.

It was a workhorse.

The RBMK had a design characteristic that set it apart from most Western reactor types.

In a standard pressurized water reactor, if the coolant water begins to boil — if steam voids form in the core — the chain reaction slows. The reactor corrects itself.

In the RBMK, the opposite was true.

When coolant boiled, when steam voids increased, the chain reaction accelerated. More steam. More power. More steam.

The technical term is the positive void coefficient. INSAG-7 uses it directly. It is not a malfunction. It is how the reactor was designed.

INSAG-7 — the IAEA's post-accident report, published in 1991 — contains a specific passage about what the reactor's designers knew.

"The reactor designers were aware that the dangerous property of the reactor they had developed could be a cause of nuclear instability."

That is a direct quotation. From INSAG-7. Section 2.1.2.

The report continues: they failed to estimate quantitatively the possible consequences. They attempted to protect operators by imposing operating limitations. INSAG-7's assessment: the protections were, in its words, extremely poor.

The designers were aware.

The operators at Chernobyl were not told.

Three years before Chernobyl, the positive void coefficient produced a direct, observable consequence at another Soviet RBMK reactor.

1983. The Ignalina nuclear power plant. Lithuania.

During startup testing, operators directly observed what INSAG-7 calls the positive reactivity insertion upon scram — the positive scram effect.

When the shutdown signal was sent, the control rods began to descend. As they entered the core, the graphite tips at their base inserted positive reactivity before the neutron-absorbing section arrived.

Power did not fall. It rose.

INSAG-7 records this directly. Section 2.2. It calls the Ignalina observation a direct precursor to what happened at Chernobyl.

The Chernobyl operators were not informed of the Ignalina observation. Not in their training. Not in their procedures. Not before April 26th.


PHASE 2 — THE SETUP

On April 25th, 1986, Chernobyl Unit Four was scheduled to conduct a safety test.

The test had a specific purpose: to verify whether, during a power loss, the residual rotation of the turbines could generate enough electricity to power the emergency cooling pumps for the approximately 60 to 75 seconds it would take the backup diesel generators to come online.

It was a gap in the plant's verified safety margins. The test was the fourth attempt to close it.

The first three attempts, since 1982, had not been completed.

At approximately 14:00 on April 25th, the reactor's Emergency Core Cooling System was manually disconnected.

This was part of the test protocol. The ECCS was switched off so its automatic activation would not interfere with the turbine coastdown measurements.

At the same time, Kyivenergo — the regional electricity grid controller — denied permission to reduce the reactor's power.

There was a shortage on the Ukrainian grid. Unit Four was needed at full output.

The test was postponed.

The reactor ran for nine hours and ten minutes with its Emergency Core Cooling System physically locked off.

Not because of the test. Because the grid needed the power.

At approximately 23:10 on April 25th, Kyivenergo granted permission to proceed. Power reduction resumed.

Shortly after midnight, the night shift took over.

Shift foreman Alexander Akimov. Reactor operator Leonid Toptunov. They had not been assigned to run this test. The test had been scheduled for the previous shift — the afternoon crew, who had been briefed.

The night shift received a brief handover.

INSAG-7 is specific about what this meant: the operators who would execute the test had, in the report's assessment, minimal time to review the test procedure before being required to execute it.

This is not a judgment. It is what the primary record states.

During the nine-hour delay, the reactor had continued to operate.

This created a problem.

At the power levels the reactor ran during the delay, a fission byproduct called xenon-135 was accumulating in the core. Xenon is a neutron absorber — it competes with the chain reaction, suppressing it. This is called xenon poisoning.

When power reduction resumed after midnight, the reactor fell faster than expected. At approximately 00:28 on April 26th, reactor power dropped to roughly 30 megawatts thermal. Near shutdown.

Dyatlov ordered the operators to withdraw the control rods and recover power.

They stabilised at approximately 200 megawatts thermal.

The planned test window was 700 to 1,000 megawatts. 200 megawatts was, by any engineering measure, dangerously low for the conditions the reactor was now in.

Dyatlov's own account, published after the accident, states he was unaware that operating the test at this power level was prohibited.

The manuals did not explicitly forbid it. INSAG-7 confirms this.


PHASE 3 — THE FORTY SECONDS

01:23:04.

April 26th, 1986.

The turbine emergency stop valves are closed. The safety test begins.

Reactor power: 200 megawatts thermal. The xenon-poisoned core. The night shift. The untested procedure.

In the reactor core, the positive void coefficient is present. It has always been present. The graphite tips are on the control rods. They have always been there.

The forty-second clock is running.

The control rods in an RBMK reactor serve two functions.

When inserted into the core, the neutron-absorbing section of the rod slows the chain reaction. This is what a shutdown rod is supposed to do.

But the design of the RBMK control rod included a graphite tip — approximately one metre of graphite at the base of each rod, ahead of the neutron-absorbing section.

Graphite is a neutron moderator. When it enters the core, it increases the chain reaction, not decreases it.

When AZ-5 is pressed and the rods begin to descend, the graphite tips enter the core first.

Before the neutron-absorbing section arrives, the graphite inserts positive reactivity. In a core already operating with a positive void coefficient, in a xenon-poisoned low-power state, the effect is not corrective.

INSAG-7 names it the positive scram effect.

01:23:40.

Deputy Chief Engineer Anatoly Dyatlov orders the AZ-5 button pressed.

The control rods begin to descend.

The graphite tips enter the lower core.

Positive reactivity is inserted. The xenon-poisoned core — already unstable at 200 megawatts — responds not with shutdown, but with excursion.

Four seconds.

Reactor power surges to approximately 100 times the reactor's nominal design capacity. The prompt criticality threshold is crossed.

01:23:44.

First explosion. Steam. The reactor vessel is destroyed.

Seconds later, a second explosion — graphite and fuel ejected from the building. The graphite moderator catches fire.

Forty seconds.


PHASE 4 — THE RECORD

Inside the plant, Shift Foreman Akimov's team began assessing the damage.

Standard plant instrumentation had failed. The explosion had destroyed the systems that would normally report reactor state.

The crew turned to their field dosimeters — handheld radiation survey meters, standard-issue Soviet equipment. The model issued to plant workers was the DP-5V.

Akimov reported a reading of 3.6 roentgen per hour.

That reading went to Plant Director Viktor Bryukhanov. Bryukhanov reported it to Moscow.

3.6 roentgen per hour is a manageable radiation level — elevated, but not immediately life-threatening with appropriate protective measures.

The DP-5V dosimeter had a physical maximum of 1,000 microroentgen per second. Which is 3.6 roentgen per hour.

When the needle reaches the right side of the scale, that is the only reading the instrument can produce.

The needle had reached the right side of the scale.

The actual radiation levels near the reactor core on the night of April 26th were not 3.6 roentgen per hour.

INSAG-7 documents indicate the actual levels in the vicinity of the open reactor were in the range of 10,000 to 30,000 roentgen per hour.

Not 3.6. Ten thousand to thirty thousand.

The DP-5V could not express that number. The instrument's range stopped at the first digit.

The KGB was present in Pripyat.

On the morning of April 26th, KGB memorandums from the Ukrainian State Archives record radiation levels on Pripyat's streets in the hours after the explosion.

The city of Pripyat — 49,000 people — was three kilometres from the reactor.

The KGB knew the streets were contaminated before dawn.

The formal evacuation of Pripyat was not ordered until 14:00 on April 27th.

Thirty-six hours after the first explosion.


PHASE 5 — THE REVISION

On August 25th–29th, 1986, the IAEA convened a post-accident review meeting in Vienna.

The Soviet delegation presented a detailed technical account.

The resulting report — INSAG-1, published in 1986 — concluded that the accident was caused by a conjunction of violations of operating procedures by the operators and a specific operating state that, combined with a design deficiency, led to loss of control of the reactor.

The primary cause, in the INSAG-1 account: the operators.

The design was described as sound.

In 1991, the IAEA published a revised report.

INSAG-7. The Chernobyl Accident: Updating of INSAG-1.

The revision addressed five primary points.

Primary cause: INSAG-7 concluded that fatal reactor design flaws were the principal mechanism of the accident. The positive void coefficient was the primary driver of the prompt criticality excursion.

Operator culpability: INSAG-7 concluded that the operators were placed in an impossible situation by a design they did not know was dangerous. Blame was redistributed from individuals to institutions.

Design flaw acknowledgment: What INSAG-1 had described as sound, INSAG-7 named as a central mechanism of the disaster.

Prior knowledge: INSAG-7 stated explicitly that the reactor's designers were aware of the dangerous property before 1986, and failed to implement fixes or warn operators.

The fifth point: INSAG-7 confirmed the Ignalina observation.

The Ignalina observation in 1983 was not an obscure technical detail.

INSAG-7 is explicit: the positive scram effect — the exact mechanism that destroyed Chernobyl Unit Four — had been directly observed at another RBMK reactor three years before the accident.

The observation was made. It was documented internally. And the Chernobyl operators were never informed.

This is not the narrator's interpretation. It is the finding of the IAEA's own post-accident report. Section 2.2. Published 1991.

INSAG-7 also confirms — in Section 2.1.2 — that the reactor's designers were aware the positive void coefficient was a dangerous property, failed to quantify its consequences, and attempted to compensate with operating restrictions that, in the report's own words, provided extremely poor protection.

The designers knew. The operators did not. Three years of documented prior knowledge.

INSAG-7 contains a finding about inevitability.

Section 3.2: the accident was structurally inevitable given the reactor design operating under those test conditions. Operators were not in a position to prevent it.

This is a specific finding. It does not mean no one is accountable. It means the design created a system in which accountable actors were not given the knowledge or the tools to act differently.

The Soviet state blamed nine men.

The IAEA's own revised record concluded that the conditions for the disaster had been engineered into the reactor before any of those nine men arrived for work.


CONCLUSION

The question the cold open posed was simple.

What do the documents show?

They show a reactor with a known design flaw — known before the first operator arrived for the night shift. A nine-hour window in which the primary safety system was offline. A shift that executed a procedure it had not been fully briefed on. And a shutdown command that, by the physics of the design, made the reactor worse.

The documents show the sequence. They do not resolve the question of accountability.

INSAG-7 draws a distinction the Soviet account did not.

There is the question of who made the decisions on the night of April 26th. That record is clear: Dyatlov ordered the test to proceed. Akimov and Toptunov executed it.

And there is the question of what those decisions were made inside of.

A reactor design whose instability was known. Procedures that did not reflect that instability. An institution that had documentation it did not distribute.

INSAG-7 holds both questions open. The primary record holds both questions open.

The reactor is sealed beneath a concrete sarcophagus on the Ukrainian steppe.

The documents are not sealed.

INSAG-7 is publicly available. The IAEA website. Free.

What they show is in the record.

The file on Chernobyl Unit Four is not closed. The forty-second flaw is in the archive.


SOURCES

International Atomic Energy Agency. 1992. The Chernobyl Accident: Updating of INSAG-1. Safety Series No. 75-INSAG-7. Vienna: IAEA. (Referred to throughout as INSAG-7; published 1991, IAEA issue date 1992.)

International Atomic Energy Agency. 1986. Summary Report on the Post-Accident Review Meeting on the Chernobyl Accident. Safety Series No. 75-INSAG-1. Vienna: IAEA.

Legasov, Valery. 1988. Audio testimony, transcribed. Multiple published English transcriptions.

USSR Academy of Sciences. 1986. Post-accident report on the Chernobyl accident. State technical report.

KGB memorandums, April 26–27, 1986. Declassified. Ukrainian State Archives / SBU declassified collections.

Shcherbina / Legasov government commission records. 1986.

Dyatlov, Anatoly. 1995. Chernobyl: How It Was. Plus trial testimony.

Akimov, Alexander, and Leonid Toptunov. Operator testimony. INSAG-7 appendix testimonies; Soviet trial records.

Bryukhanov, Viktor. Trial transcripts. Soviet trial records, partially published in English.

Chernov, Dmitry, and Didier Sornette. 2016. "Examples of Risk Information Concealment Practice." In Man-made Catastrophes and Risk Information Concealment, 9–245. Springer.

Higginbotham, Adam. 2019. Midnight in Chernobyl. New York: Simon & Schuster.

Plokhy, Serhii. 2018. Chernobyl: History of a Tragedy. New York: Basic Books.

Brown, Kate. 2019. Manual for Survival. New York: W. W. Norton.


The archive is open. deaddroparchives.com