1장
When Atoms Go Wrong: The Spectacular Failures That Shaped Nuclear History
Nuclear power has always captivated our imagination with its promise of limitless clean energy, yet lurking behind this promise is a history of spectacular failures that have shaped our relationship with atomic technology. James Mahaffey's "Atomic Accidents" takes us on a journey through the often bizarre, sometimes tragic, but always fascinating history of nuclear mishaps. The book has become a cult favorite among both nuclear enthusiasts and critics, with Bill Gates naming it one of his must-reads for understanding energy challenges. Its cultural impact extends beyond science circles - references to incidents like Three Mile Island and Chernobyl have permeated everything from "The Simpsons" to HBO's award-winning miniseries. What makes this book particularly compelling is Mahaffey's insider perspective as a nuclear engineer who can explain complex technical failures while acknowledging the very human errors behind them. As we face climate change and debate the role of nuclear in our energy future, understanding these accidents becomes not just historically interesting but critically important for charting our path forward.
2장
The Allure and Danger of the Spectacular
Our fascination with disaster runs deep in human psychology. Just as crowds gathered to witness the Russian locomotive's derailment in 1954, we're drawn to spectacular failures. This principle was brilliantly exploited by William "Bill" Crush of the Missouri, Kansas & Texas Railway in 1896, who staged a deliberate locomotive collision as a publicity stunt. With train travel both essential and dangerous in that era (7,029 Americans died in rail accidents in 1891), Crush's psychology was ingenious: rather than falsely assuring safety, show people the worst possible outcome, replacing the dread of random accidents with the excitement of controlled spectacle.
Despite Crush's confident assurances that "these old engines are tough" and wouldn't explode, at least one boiler did exactly that after impact, sending deadly shrapnel into the crowd of 40,000 spectators. Three people died instantly and six were seriously injured, including the official photographer who lost an eye. Yet remarkably, this tragedy didn't dampen public enthusiasm-spectators rushed the wreckage, burning their hands to collect souvenirs.
The era of staged train wrecks continued with "Head-On Joe" Connolly making a career of locomotive crashing, eventually staging 73 wrecks without killing a spectator. He perfected the science: 4,000 feet of track for optimal 45 mph impact speed-fast enough for spectacle without dangerous boiler explosions.
This specter of exploding steam engines profoundly influenced nuclear power plant design, where preventing steam catastrophes remains the primary safety focus. Nuclear plants are essentially sophisticated steam engines using uranium instead of coal as a heat source. After World War II, engineers recognized that nuclear reactors could increase power output instantaneously rather than gradually, potentially causing catastrophic steam explosions. This became the critical safety concern in nuclear power development.
Through rigorous testing and robust engineering, the explosive steam release risk was eliminated. In over seven decades of commercial nuclear power in America, not one steam explosion has occurred nor life been lost to radiation exposure. Yet ironically, nuclear accidents have happened for seemingly insignificant reasons-stuck valves, operator errors, or backup systems failing sequentially.
What makes nuclear accidents uniquely terrifying isn't just their technical complexity but the invisible threat of radiation-not just immediate harm but long-term contamination. While concrete shielding keeps radiation contained, public perception remains disconnected from actual risk probabilities, creating a unique challenge for nuclear power acceptance.
3장
The Dawn of the Atomic Age: Early Radiation Pioneers
The earliest encounters with radiation came before the phenomenon was even understood. Marie Curie, who discovered radium in 1898, had no concept of its dangers as she carried test tubes of radioactive materials in her pocket. By the 1920s, radium had become a wonder substance, used in everything from watch dials to health tonics.
The tragic case of the "Radium Girls"-factory workers who painted luminous watch dials with radium paint and were encouraged to point their brushes with their lips-revealed radiation's dangers. Many developed horrific radiation poisoning, with their jawbones literally disintegrating. Yet even as these women suffered, wealthy industrialist Eben Byers was consuming "Radithor," a radium-infused health tonic, at his doctor's recommendation. By the time he died in 1932, his bones were riddled with radiation and literally falling apart.
Despite these tragedies, the public's fascination with radioactive substances persisted in surprising ways. Radium dial watches continued being manufactured until 1963, while US Radium operated until 1980 before becoming Safety Light Corporation. Even more astonishing is the enduring popularity of radioactive mineral water. The 1980s mineral water craze unwittingly revived the same principles that killed Eben Byers-spring water naturally contains dissolved uranium oxide and radon gas bubbles.
Thousands of "healing" radium springs still operate worldwide. Japan maintains 1,500 mineral spas despite its sensitivity to radiation issues. Austria's Badgastein spa, developed from a former Nazi gold mine where enslaved workers mysteriously "got healthier," now treats patients in radon-filled tunnels, claiming to cure everything from arthritis to infertility.
The radiation poisoning story contains paradoxes too. William Bailey, who killed Eben Byers with Radithor, lived to 64 despite consuming more radium water than anyone. When his remains were studied, they showed significant radioactivity, yet he never suffered the horrific symptoms of his customers. These inconsistencies raise questions about individual radiation tolerance. Could humans eventually evolve radiation resistance?
Meanwhile, radiation's pioneers paid the ultimate price. Marie Curie died in 1934 of aplastic pernicious anemia, her bone marrow destroyed by radium and X-ray exposure. Her daughter Irene Joliot-Curie, also a Nobel Prize-winning radiation scientist, died in 1956 at age 58 from leukemia after a polonium capsule exploded in her face.
4장
When Playing with Fire Becomes Playing with Atoms
World War II brought nuclear dangers beyond comprehension. By 1939, the radium scandals had left the public fearing radiation as an invisible, deadly threat. But nuclear fission would make radium contamination seem minor by comparison. The atomic bombing of Japan served as both military action and scientific experiment, providing the first hard data on radiation effects on humans.
The nuclear explosion's effects were devastating. Burns from intense thermal energy, not radiation, injured many survivors. The fireball created temperatures of millions of degrees with air pressure in millions of pounds per square inch, causing everything within 12 miles to ignite. People were vaporized, crushed by the blast, or severely burned.
The Manhattan Project was a marvel of rapid development without fatal accidents, though close calls occurred. Under General Leslie Groves, the remote Oak Ridge facility grew to 75,000 people focused on uranium-235 enrichment, with workers kept ignorant of their purpose for security reasons.
The enriched uranium-235 presented unprecedented dangers. If stacked improperly, it could spontaneously generate heat and radiation in a runaway chain reaction. At Oak Ridge, Richard Feynman discovered alarming conditions: uranium solutions stored in wooden drums positioned where they could couple into accidental reactors, with building layouts designed without nuclear physics considerations. The entire Clinton Works complex was a potential disaster. Feynman convinced Oppenheimer that workers needed basic nuclear physics training, as security measures keeping workers ignorant were now endangering the project.
After Japan's surrender in 1945, Manhattan Project work continued. Scientists wondered if substituting tungsten carbide for uranium in the tamper would increase yield. Testing this required dangerous "tickling the dragon" experiments-manually stacking tungsten carbide bricks around a plutonium core until approaching criticality.
These experiments claimed their first victim when physicist Harry Daghlian, working alone against regulations, accidentally dropped a brick onto a plutonium sphere. As blue light flashed, he instinctively knocked the brick off with his right hand, feeling the tingling of direct neutron exposure. His hand absorbed 20,000-40,000 rem, eventually dying as circulation collapsed. After brief improvement, severe nausea returned on the tenth day. Despite transfusions and medications, he slipped into a coma after 25 days, dying on September 15, 1945-history's first accidental acute radiation poisoning death.
Nine months later, Louis Slotin, one of the investigators of Daghlian's accident, was demonstrating the same experiment using a screwdriver instead of proper wooden spacers. When the screwdriver slipped, the assembly went prompt critical. The blue flash filled the room as Slotin instinctively jerked the tamper away, but he received 2,100 rem-twice the lethal dose-and died nine days later.
The accident ended manual bomb assembly experiments forever. All further work was conducted remotely, using the "Godiva" systems positioned a quarter-mile away.
5장
Experimenting with Disaster: The Idaho Nuclear Testing Station
In the 1950s, the United States developed atomic technology on a grand scale at the National Reactor Testing Station (NRTS) in the barren desert between Arco and Idaho Falls-chosen specifically because accidents there would cause minimal damage. With over 50 prototype reactors, the NRTS experienced several significant accidents, including meltdowns and steam explosions.
The first accident involved the Experimental Breeder Reactor One (EBR-I) in 1955, designed to produce more fuel than it consumed. On November 29, 1955, while testing the reactor's positive temperature coefficient of reactivity, disaster struck. With coolant flow stopped and safety systems disconnected, power surged from 50 watts to one million watts in just three seconds. The thin fuel rods buckled under high heat, bending inward and increasing reactivity. The United States had experienced its first meltdown-the core completely trashed, with its top half liquefied.
Samuel Untermyer II's BORAX-I experiment culminated in a spectacular finale on July 22, 1954. After 200 successful tests proving that boiling water reactors could safely self-regulate, the team decided on one final experiment-a deliberate prompt supercritical excursion. With important visitors gathered at the observation post, physicist Harold Lichtenberger hit the EJECT CONTROLS button. The result exceeded all expectations-instead of the predicted 80 megajoules, they got 135 megajoules of energy as the reactor ramped from 1 watt to 19 billion watts in 2.6 milliseconds. A blue flash lit the assembly, the 2,200-pound control mechanism flew 30 feet upward, and the core melted instantly into a column of black smoke.
Rather than frightening observers, the controlled destruction actually calmed fears by demonstrating that even the worst-case nuclear accident produced a manageable steam explosion.
The most tragic accident was the SL-1 incident in January 1961. Three military operators were working the night shift, reconnecting control rod mechanisms after maintenance. Against explicit instructions, operator Jack Byrnes suddenly yanked the main control rod completely out of the core-pulling it 23 inches when just 4 inches would cause criticality. The reactor went prompt critical with a 2-millisecond period, creating an unprecedented steam explosion. The water above the core slammed into the vessel head, launching the 13-ton vessel nine feet upward, shearing pipes and sending shield plugs flying like cannon shells.
Byrnes and Richard Legg died instantly from the explosive shock wave. Richard McKinley died two hours later from massive head trauma. All three bodies were deeply contaminated with fission products, requiring them to be treated as high-level radioactive waste.
The cleanup took 13 months and $2.5 million, requiring remote manipulation, gamma cameras, and meticulous planning. This tragedy affected attitudes toward small nuclear power, demonstrating that no nuclear system could be truly "foolproof."
6장
Britain's Nuclear Ambitions and the Windscale Disaster
Post-war Britain found itself cut off from American nuclear secrets despite their wartime collaboration. While British scientists had contributed to the atomic bomb project, they were never allowed access to plutonium production reactors. After the 1946 Atomic Energy Act forbade sharing atomic secrets, Britain was forced to develop its own nuclear program with only Canada as an ally.
Britain's first experimental reactor, GLEEP (Graphite Low Energy Experimental Pile), was assembled at Harwell in 1947, but before significant experimental work could be done, construction began on the first plutonium production pile at Windscale. The two Windscale reactors were massive air-cooled graphite structures with 3,440 fuel channels holding approximately 70,000 uranium slugs. Unlike American water-cooled designs, the British opted for air cooling, using eight enormous blowers to push air through the reactor and up 410-foot chimneys.
The Windscale piles experienced mysterious temperature rises due to Wigner Energy, a phenomenon where neutron bombardment stored potential energy in graphite that could suddenly release as heat. This required periodic "annealing" by heating the core to 250C, a delicate procedure.
By 1957, with Prime Minister Harold Macmillan desperate to restore relations with America through nuclear achievements, the Windscale piles were operating in emergency mode at maximum power with flammable materials throughout.
On October 10, 1957, disaster struck when a routine annealing operation went wrong. Temperatures soared wildly, with one thermocouple reading 412C despite attempts to cool it. When workers removed an inspection plug, they discovered a bright red glare-the pile had been burning since Tuesday. Men worked in rubber suits and respirators as temperatures passed 1,200C and flames shot from fuel channels.
Tom Tuohy, summoned from his flu-stricken family, arrived to find men struggling with fuel rods glowing yellow and dripping molten uranium. After carbon dioxide proved ineffective, they resorted to pumping water directly into the fuel channels-a desperate measure never before attempted in a graphite reactor. The fire continued raging until Tuohy had a brilliant insight: turn off the blowers. Without forced air, the flames quickly died out.
The core was a total loss with over 10 tons of uranium melted and five tons burned. Though 20,000 curies of iodine-131 escaped over 200 square miles of territory, the filters (once mockingly called "Cockcroft's follies") fortuitously captured 30,000 curies. Despite statistical predictions of 240 thyroid cancers, follow-up studies through 2010 found no evidence of health effects from the radiation release.
7장
The Liquid Metal Dream: Sodium-Cooled Reactor Failures
Admiral Hyman Rickover's passion for nuclear submarines resulted in the groundbreaking USS Nautilus, launched in 1955. While early nuclear submarine development faced concerns about steam explosions and uranium availability, Rickover's experience in leaky submarines informed his rejection of liquid sodium cooling-which would react violently with inevitable water leaks-in favor of pressurized-water reactors. His water-cooled design prioritizing crew safety became the standard for both military and civilian nuclear power.
Despite Rickover's warnings, engineers pursued sodium-cooled breeder reactors that could produce more fuel than they consumed. The Sodium Reactor Experiment (SRE) began providing 6.5 megawatts of electricity to the Moorpark community by April 1957-the first civilian nuclear power in the United States.
The reactor went completely rogue on July 13, 1959, with power increasing uncontrollably. Over the following days, it continued operating despite alarming signs: radiation readings of 14,000 counts per minute, multiple scrams, temperature fluctuations reaching 1,465F, and fuel assemblies becoming jammed. When operations finally ended on July 24, operators discovered the reactor core was wrecked-13 of 43 fuel rods had disintegrated.
Investigation revealed tetralin had leaked into the sodium coolant, decomposing into a black substance that blocked coolant flow. The resulting high temperatures created a uranium-steel eutectic that melted at just 1,340F, causing fuel assemblies to collapse.
Even more ambitious was the Fermi 1 reactor near Detroit, featuring remarkable engineering complexity. On October 5, 1966, disaster struck. At 3:05 PM, with the reactor at 34 megawatts, engineers noticed erratic neutron activity and dangerously high temperatures. By January 1967, they confirmed fuel had melted, and by May they'd removed the damaged assemblies. After draining the sodium and using a specially-built periscope, they discovered a piece of zirconium blocking coolant flow-a modification they had never properly documented on the final plans.
Despite repairs, Fermi 1 continued to struggle. In 1970, during a restart with AEC inspectors present, 200 pounds of sodium suddenly broke loose, causing an explosion when it mixed with water and air. The plant finally reached its designed 200 megawatt power level in October 1970, but could only remain online 3.4% of the time. Its license extension was denied in 1972, and it was decommissioned in 1975.
Admiral Rickover had accurately predicted the fate of sodium-cooled reactors in 1957, calling them "expensive to build, complex to operate, susceptible to prolonged shutdown as a result of even minor malfunctions, and difficult and time-consuming to repair."
8장
The Human Cost: Radiation Accidents in Fuel Processing
At 64 years of age, Harold McCluskey was nearing retirement but still eager to pass along his skills as a chemical operator. On August 30, 1976, at 2:45 a.m., McCluskey entered the Americium Recovery Room to process americium-241, a valuable byproduct of plutonium production used in smoke detectors. Just minutes after starting the process, he heard hissing and found the glove box filled with dense brown smoke. As he tried to check the drain valve, the resin column disintegrated in a heavy blast.
The explosion blew out the glove box windows and knocked McCluskey to the floor. Covered in blood and blinded, he crawled toward the door with help from his junior colleague. Geiger counters went off-scale when held to his body-he was contaminated with between 1 and 5 curies of americium, billions of times more than previous incidents.
Treatment began immediately with chelation therapy using calcium diethylene-triaminepentaacetate (Ca-DTPA) to capture americium in his bloodstream before it could reach his bones and liver. By day's end, his contamination level had dropped by a factor of 1,000.
Though McCluskey recovered physically, he faced psychological challenges as the "Atomic Man." Despite living near the Hanford plutonium plant, his friends and church members feared his radioactivity, refusing to visit his home. His despair eased somewhat when his pastor gave a sermon convincing the congregation it was both Christian and safe to be around him.
McCluskey ultimately died of coronary artery disease on August 17, 1987, at age 75-not from radiation effects. To the end, he maintained his injuries were merely an industrial accident, remaining supportive of nuclear power.
The Rocky Flats Plant, located near Denver, transformed from grazing land into a critical Cold War nuclear weapons facility producing plutonium "pits" for nuclear weapons. Working with plutonium presented unique challenges, particularly its pyrophoric nature-like wood shavings that ignite more easily than logs, plutonium machined into thin pieces can spontaneously combust in air.
On May 11, 1969, a catastrophic fire erupted when a pile of plutonium-contaminated rags spontaneously ignited beneath a briquette press. With no workers present, the fire spread to an uncovered can containing a plutonium briquette, which burned white-hot. The flames engulfed Plexiglas windows and rubber gloves in the glove boxes, leaving arm-holes open that allowed air to rush in and intensify the plutonium fire.
When firefighters arrived at 2:27 PM, they found the north plutonium foundry fully engulfed. Carbon dioxide extinguishers proved useless, forcing Captain Wayne Jesser to order water hoses deployed despite the risks of hydrogen explosion or criticality. The scene was apocalyptic-plutonium and magnesium burning furiously, molten lead dripping from gamma-ray shielding, and even the glue holding the Benelex together aflame.
By 8:00 PM the fire was declared contained, though plutonium continued flaring up until Monday morning. The $70.7 million in damages set a record for U.S. industrial fire losses.
9장
From Three Mile Island to Fukushima: Modern Nuclear Disasters
The Three Mile Island incident in March 1979 began with a seemingly minor fault but escalated when operators misinterpreted confusing signals. Following their training to prevent the pressurizer from "going solid," they fatally throttled back the emergency cooling pumps. As the core began uncovering, the fuel reached temperatures of 5,000F, causing the zirconium cladding to react with steam and produce hydrogen gas. Within hours, the core had partially melted, creating a ceramic pool at the bottom of the vessel.
When news spread about potential hydrogen explosions, panic ensued-42,000 people fled Harrisburg immediately, growing to 135,000 evacuees by Sunday. Meanwhile, radiation leaked into the auxiliary building and even the control room, forcing staff to wear respirators.
The 1986 Chernobyl disaster occurred during a safety experiment to test whether a spinning turbine could provide enough residual power to run coolant pumps during an emergency shutdown. The experiment began with power reduction and safety systems disconnection. When resumed at 11:10 PM, shift changes brought in operators who struggled to control the reactor manually. They accidentally dropped power to 30 megawatts, trapping the reactor in the "iodine valley" where xenon-135 buildup prevents easy power increase.
At 1:23:04 AM, the turbine was throttled and the MPA button pushed, causing water in the reactor to boil furiously as power spiked. After 36 agonizing seconds, the emergency shutdown button was pressed, but the control rods jammed. Within seven seconds, the reactor reached 30 billion watts and began disintegrating. A massive explosion followed as pressure-relief valves blew, steam separators tore through the roof, and the reactor's 500-ton cap lifted off. Superheated steam mixed with hydrogen gas detonated, vaporizing the core and sending radioactive material 36,000 feet into the air.
In March 2011, Japan was struck by a devastating 9.0 magnitude earthquake followed by a 49-foot tsunami that overwhelmed the 18.7-foot wall protecting the Fukushima Daiichi nuclear plant. The tsunami destroyed water intake structures, collapsed pumps, and shorted all external electrical connections. Within six minutes, all underground diesel generators were flooded and disabled.
In Unit 1, operators worked in darkness with only flashlights. Within three hours after the earthquake, steam-relief valves had opened and water boiled away from the reactor core. By four-and-a-half hours, the fuel began melting, generating hydrogen gas. On March 12, Unit 1's reactor building exploded in a spectacular geyser of debris, setting back recovery efforts as all subsequent work required bulky radiation suits and respirators.
The disasters cascaded: Unit 3 exploded on March 14, Unit 2's fuel melted, and most surprisingly, the seemingly safe Unit 4 exploded on March 15, despite having no fuel in its reactor, when hydrogen from Unit 3 backflowed through the shared vent stack.
10장
Breaking Free from the Rickover Trap
The current state of American nuclear power is bleak: our operating reactors are oversized versions of Rickover's submarine plants with structural vulnerabilities; we lack any fuel reprocessing capability; and reactor waste sits in dry casks with nowhere permanent to go.
The real danger facing nuclear engineering is the "Rickover Trap"-continuing with established technology simply because the infrastructure exists, rather than pursuing potentially superior alternatives. Just as we burn billions of gallons of gasoline annually not because it's optimal but because the infrastructure is entrenched, we've remained wedded to pressurized water reactors despite better options.
Yet nuclear engineers aren't idle. At least five new power reactor designs are in development-all small, modular units that can be factory-built and trucked to installation sites. This approach offers tremendous advantages: small reactors have proportionally small problems, explosions, coolant leaks, and investment requirements.
Most promising is the Generation IV International Forum, a nine-country coalition formed in 2000 to develop next-generation nuclear plants by 2030. Their research includes the molten-salt thorium-fueled reactor-potentially reviving this nearly forgotten concept for clean energy production.
While nuclear power will always carry risks, we've learned valuable lessons from each incident. As long as nuclear engineering continues to innovate and learn from its history of accidents, nuclear power can deliver substantial benefits for our economy, society, and environment. The spectacular nuclear disasters that make compelling narratives may now be behind us, replaced by a safer future built on hard-won experience.
