Ch. 05 – The Grand Experiment (Part 3)

The deputy chief engineer eventually received permission to run his experiment on Reactor Four at the Chernobyl nuclear power generating station. He ran his first experiment in mid-1985. The record is not entirely clear, but apparently one of the fail-safe mechanisms designed into the reactor kicked in and shut down the reactor when it sensed that the primary coolant pumps were malfunctioning.

Although red-faced with embarrassment, the deputy chief engineer petitioned for, and eventually received, a second chance from the subcommittee, but apparently he was also told about dire consequences should he fail again. 

Midmorning on April 25, 1986, the deputy chief engineer gathered the Reactor Four operating engineers and technicians and explained his plan in detail. He told them how the reactor had shut down the previous year and explained that failure was not an option this time around. He then ordered all five safety systems bypassed, and–just to be sure–he also had all the backup electrical systems shut down, including the emergency diesel generators that could have powered the reactor controls in an emergency.

He probably felt safe doing this because he did not intend on running the reactor for more than a few minutes under load. After all, what could possibly happen in a few short minutes? And, not being nuclear trained, he had no idea of what unintended consequences could result from disconnecting these systems. Although we will never know, he may have been thinking that the worst-case scenario would be a complete shutdown of the reactor as would happen if the fuel supply were cut off from a conventional boiler.

As luck would have it, unexpected power demand that afternoon delayed the experiment until late that evening. In order to get the trial under way, the engineers needed to reduce reactor power to minimum, which–when done by the book–is a time-consuming process. Because they were now behind schedule, they reduced the power level more rapidly than this reactor design could handle. This caused a buildup of neutron-absorbing fission byproducts in the reactor core, which poisoned the reaction and threatened to shut it down altogether. Since that would have spoiled the experiment a second time (hello, Siberia), to compensate, they withdrew most of the control rods. Because of the poisoning, this allowed a power increase to barely thirty megawatts, which was just sufficient to bring the reactor into its most unstable range. Something had to be done immediately.

There were only two choices: do absolutely nothing and wait twenty-four hours for the poisoning to dissipate, or increase the power immediately.

With Siberia in the wings, you know what choice they made.

They finally marginally stabilized reactor power at 200 megawatts–one-fifth its design power. In the process, because the reaction was still poisoned, they had pulled all but six of the 211 rods. The absolute design minimum for this reactor was thirty rods left in the core. They had withdrawn twenty-six rods that never should have been pulled, so the immediate situation was dire.

About a half-hour later, as things appeared reasonably stable, they decided to commence the actual experiment, and shut down the turbine generator. Their intent was to see if the turbine could still supply coolant pump power even though it was only coasting–no longer being driven by the reactor. A successful outcome would prove that they did not need to obtain outside power to maintain proper cooling levels when they decoupled a reactor and its turbine. An engineer with nuclear training could have told them the answer without conducting the experiment. But the deputy chief engineer wasn’t a Nuke. 

With reduced electrical power, the pumps slowed, reducing the flow of cooling water.

The modern nuclear reactor used in the United States and the rest of the world controls neutron levels by absorbing them with boron or cadmium rods. The primary coolant acts as a moderator by slowing the neutrons. The RBMK model, however, works in reverse, using graphite rods to moderate the neutrons and the primary coolant to absorb them.

So we have a reactor operating at a significant power level with almost all the moderating control rods pulled. It is still stable, although barely, because the primary coolant is absorbing neutrons as fast as they are being produced. Now slow the coolant pumps so the water moves more slowly. It stays in the reactor core longer, getting hotter, and finally begins to boil. But steam cannot absorb neutrons; suddenly the neutron flux skyrockets.

The reactor operators immediately hit the emergency butt
on that drives the control rods, all remaining 205 of them, and the emergency protection rods into the core. But all backup power had been shut down, even the emergency diesel generators. The only available power came from the slowing turbine, which meant that the already slow primary coolant pumps had even less power. So the entire cycle was exacerbated. 

This is when another design problem of the RBMK became evident. The control rods had graphite tips followed by a one-meter hollow segment (I don’t know why; they just did) followed by a five-meter graphite section. As soon as the rods penetrated the core, they displaced more coolant without themselves absorbing any neutrons, because of the hollow section. The already skyrocketing neutron flux went ballistic. All hell broke loose, and the reactor container exploded–not a nuclear explosion, just a plain, old-fashioned steam-boiler explosion.

But it was a doozy. Red-hot chunks of fuel and graphite fell everywhere. Fifty tons of nuclear fuel evaporated and were ejected high into the atmosphere. Seventy tons were ejected sideways into the surrounding areas. An additional 50 tons of fuel and 800 tons of graphite remained in the reactor vault, smoldering for days. Experts have placed the release of radioactivity at about ten times the amount generated at Hiroshima..