Chernobyl + 22
22 years ago, on a late April evening near the city of Pripyat in northern Ukraine, Reactor #4 at the V.I. Lenin Memorial Chernobyl Nuclear Power Station exploded. This disaster, the only instance to date of a "Level 7" on the International Nuclear Event Scale, resulted in a complete breach of the containment dome accompanied by a severe nuclear meltdown. (By comparison, the Three Mile Island accident in Pennsylvania was a Level 5: "Accident with Off-Site Risk".)
The four reactors at Chernobyl were each GigaWatt-output "reaktor bolshoy moshchnosti kanalniy" (Reactor-High Power-Channel Type, or RBMK). Together, these four reactors provided 10% of Ukraine's electricity.
RBMK reactors rely on light water (i.e., non-deuterium or tritium hydrogen in the water) for cooling, and graphite rods for moderation (akin to the world's first-ever nuclear reactor, the Graphite Reactor in Oak Ridge, Tennessee). Given the size of RBMK reactors and the methods for moderating fast neutrons, they can use natural uranium (vice enriched) for fuel. This also makes the design prone to instability, where boiling coolant can create a very large void coefficient: coolant that is supposed to act as a neutron absorber is boiled away, which increases the reactivity of the core, creating more energy in a positive feedback loop. The added energy can quickly lead to a "Loss of Coolant Accident", which in a large system like the RBMK leads to a catastrophic failure. (BTW: The word "scram" -- which Webster's defines as "a rapid emergency shutdown of a nuclear reactor" -- is actually an acronym coined by legendary physicist Enrico Fermi.)
In late April 1986, while Chernobyl's Reactor #4 was shut down for maintenance, technicians decided to test the ability of the reactor's turbine to power the safety systems should external electrical power be lost. The key question was whether or not the turbines, as they wound down from the reactor, could power the reactor's water pumps while the backup diesel generators came online. Though earlier tests had failed, the technicians wanted to check if recent modifications were sufficient to achieve positive results.
Before the reactor power had been decreased for the test to be conducted, a regional power station went off-line. The grid controller from the Ukrainian capital of Kiev asked that further power reductions from the reactor be postponed to allow sufficient electricity to meet evening demands.
But the night shift was not informed of the postponement. So, when the Kiev grid controller allowed the reactor shut-down to continue, the technicians at Chernobyl instead followed the original test protocol and powered down too quickly. This led to a build-up of Xenon-135 -- a highly-effective neutron absorber that can "poison" a reactor.
Technicians who began to increase the reactor power for the test saw far less power than expected. Unaware of the Xe-135 build-up (and commensurate "burn-off" as neutron flux increased), they removed the graphite rods to increase the reactivity. The increased power and decreased moderation from the graphite rods created steam bubbles in the coolant -- increasing the void coefficient described above. And though the technicians began to "scram" the reactor at 01:23:40 local time, the spike in energy caused the control rods to fracture and jam.
At 01:23:47 local time, as the last of the Xe-135 was burned off, the reactor jumped to 30 GigaWatts thermal: more than twenty times normal operating output. The fuel rods began to melt and the build-up of steam pressure created an explosion that blew off the reactor lid, resulting in a surge of oxygen that caused the graphite to ignite. The loss of containment and the graphite fire exacerbated the spread of radioactive debris throughout the region.
Today, Pripyat and the surrounding area (within a 30km radius of the reactor) is abandoned. Chernobyl has been shut down. However, there are still several RBMK reactors in operation in the Former Soviet Union (in St. Petersburg, Kursk and Smolensk).
Having spent this week at a 5-day "Team Leader Training Course" for the TapRooT Root Cause Analysis methodology, I have a new appreciation for the consequences of "human performance difficulties", particularly with highly complex systems. (Maybe Vinay Gupta and John Robb have a point re: "Resilient Communities" and "simple solar"!)