The first sharp crack came from near cathead number four, and for a moment the men on the scaffold platforms did not know what it meant. It was just past ten in the morning on Thursday, April 27th, 1978. Fifty-one workers stood on four levels of wooden and steel scaffolding 170 feet above the ground, inside the half-finished shell of cooling tower number two at the Pleasants Power Station in West Virginia. The hoist cable went slack.
Then the cathead gantry began to topple inward, toward the hollow center of the tower. The concrete beneath them was crumbling around the bolt holes where the jump form beams were anchored. The bolts ripped out of the young concrete the way a nail pulls from wet clay. A state police major who arrived later described the scaffolding peeling away from the top of the tower the way you peel an apple.
The failure started at one point near cathead number four, traveled in both directions around the full circumference, and the two fronts met at the point exactly opposite where the collapse had begun. A complete unzipping, a full 360 degrees. The entire event lasted less than thirty seconds. Some eyewitnesses said closer to fifteen.
Every man on the scaffold fell. Most plunged 170 feet into the hollow interior of the tower along with the wooden formwork and the steel scaffolding. At that height, free fall under gravity brought them to the floor of the tower at roughly 65 miles per hour. All 51 men on the scaffold were killed.
The governor’s commission later subdivided the deaths: 45 died from the fall itself, six more were crushed on the ground by falling debris. Seven men who happened to be at ground level inside the tower survived. The only safety net on the job was mounted on the exterior of the shell. It was pulled inward by the collapsing forms and caught no one.
Days later, the head of OSHA was summoned to the White House. President Jimmy Carter had been a nuclear-trained engineer in the Navy. Before she could even open her briefing folder, Carter looked at her across the desk and said five words. “It was green concrete, wasn’t it?
”
He was right. Lab tests later showed the concrete had reached roughly 220 pounds per square inch of compressive strength. The minimum needed to safely hold the crane system above it was 1,000. A rescue worker at the scene reported crushing pieces of the fallen concrete in his bare hands.
But the concrete alone did not kill those men. A routine adjustment to a single wire rope on one of the six cranes, moved six feet closer to the center of the tower, had created a sideways force that the young concrete could not resist. That adjustment had been made dozens of times before on 36 other towers without incident. On April 27th, 1978, the geometry finally caught up with the physics.
To understand Willow Island, you first need to understand what was happening along the Ohio River in the late 1970s. The 1973 oil embargo had panicked every electric utility in the Eastern United States. Oil-fired power plants suddenly looked like liabilities. Nuclear projects were drowning in cost overruns and regulatory delays.
The answer for utilities sitting on top of the Appalachian coal fields was obvious: build more coal plants, build them fast, build them big. New generating stations were rising at Mountaineer, Mitchell, Kammer, and dozens of other sites within barge distance of the West Virginia mines. Pleasants was one of the largest. It was being built on a narrow shelf of land between West Virginia Route 2 and the Ohio River, just downstream from the small town of Saint Marys in Pleasants County.
The owner was Monongahela Power Company, a subsidiary of Allegheny Power System. The total project cost was $677 million in 1978 money. Adjusted for inflation, that is roughly $3. 5 billion today.
The plant was designed with two coal-fired generating units rated at 650 megawatts each, enough combined output to power about 1. 3 million homes. Two massive natural draft hyperbolic cooling towers would handle the thermal load. Each tower would stand 430 feet tall with a base diameter of 357 feet.
These are the hourglass-shaped towers that generate their own updraft through the stack effect, requiring no mechanical fans. The general contractor was United Engineers and Constructors of Philadelphia. Their resident project manager on site was R. S.
Milne, known as Bill. The cooling tower subcontractor was Research Cottrell Incorporated, operating through their Hamon Cooling Tower division out of Bound Brook, New Jersey. Their subcontract for both towers was worth $12 million. The company president, Dennis Carlton Jones, would fly in from New Jersey the day after the disaster and hold a press conference where he cited the system’s perfect safety record.
Approximately 2,700 men from 20 different subcontractors worked on the overall Pleasants site, a small city of workers crammed onto a narrow strip of riverbank between a two-lane highway and the Ohio River. The concrete supplier was Chris Concrete Company. The concrete testing firm was Pittsburgh Testing Laboratory. The site safety and communications officer was Phil Corvino.
Every morning, workers checked in by dropping a numbered metal token into a collection box at the gate. Each man had a unique number. When you dropped your token, the company knew you were on site. When you picked it up at the end of your shift, they knew you had left.
That simple brass tag procedure would become grimly important before the day was over. And here is a detail that almost nobody mentions when telling this story. Research Cottrell paid their tower workers a flat eight hours of wages per day, regardless of how long the actual concrete pour took. If the crew finished the day’s lift in six hours, they still got paid for eight.
But if they needed nine hours, they still only got eight. Investigators later concluded that this pay structure created an enormous financial incentive to rush the pour, jump the forms early, and start the next day’s work before the concrete had properly cured. It rewarded speed, and speed was the enemy of safety on this project from day one. The towers were built using a patented system called the lift form method, developed by the Belgian engineering family behind the Hamon company.
The founder, Achille Hamon, started the firm in France in 1904. His nephew, Maurice, licensed the cooling tower technology to Research-Cottrell in 1958. Maurice held multiple patents on the system, including United States patent number 3,731,461 for drift eliminators and patent number 3,700,051,017 for the tower design itself. Research-Cottrell’s corporate roots went back even further.
The company had been founded on Frederick Gardner Cottrell’s 1908 electrostatic precipitator patent, one of the foundational inventions in industrial air pollution control. By the 1970s, Research-Cottrell had built or was building 61 natural draft cooling towers in American power plants. Most of the natural draft cooling tower fleet on the United States power grid traces back to this single Belgian-American technology license. The lift form system worked like this.
Every morning, carpenters arrived at 6:30 and unwedged the previous day’s form work. Iron workers followed at 7:30 to set the reinforcing steel. By 8:30, the entire scaffolding and form work rig was hydraulically jacked upward by exactly 5 feet to create the mold for the next ring of the tower shell. By 10:00, concrete started flowing up from the ground.
By 3:00 in the afternoon, pouring was complete. The men descended, and the fresh concrete cured overnight. One 5-foot lift per working day. Six cathead gantry cranes sat on top of the scaffolding ring, spaced at 60-degree intervals around the tower’s circumference.
Each cathead was powered by twin drum electric hoists and had its legs bolted to aluminum jump form beams. Those beams were anchored into the previous day’s cured concrete by bolts and cones cast into the shell about 12 feet apart. A wire rope, called the static line, ran from each cathead down through a pulley to a ground anchor point. The static line served two purposes.
It prevented the rising concrete bucket from swinging into the thin shell wall, and through the pulley arrangement, it provided mechanical advantage that reduced the force needed to hoist the bucket. As the tower grew taller and the hyperbolic shape curved inward, the ground anchor point of each static line had to be physically moved to accommodate the changing geometry. This was considered a routine adjustment. Nobody wrote procedures for it.
Nobody calculated the forces it generated. And for 36 towers and 7 years, it had never caused a problem. Tower number one had been completed in August 1977 using this exact system. Dennis Carlton Jones bragged at his press conference that the patented system had been used safely for 7 years.
That track record was a major reason nobody questioned what was happening on tower number two. By April 27th, 1978, tower number two had reached lift 28. It stood 166 feet tall, about 39 percent complete. Above lift 10, the entire rig and every man on it were supported by nothing except the concrete shell below them.
There was no independent structure. There was no backup. The scaffolding literally stood on the building it was building. The workforce on the cooling tower was drawn almost entirely from the hollows and small towns of Pleasants County, West Virginia, and the Ohio River communities just across the state line.
One extended family dominated the crew. The Steel and Blue Ear families had been iron workers for generations. Locals joked that you could not get into the union unless your last name was Steel or Blue Ear. On the morning of April 27th, four Steel brothers drove to work together in a pickup truck down through the steep hollows.
Larry Gail was 32. Ronald was 30. Ernest was 29. Miles was 26.
Their uncle Emmett Steel, 61 years old and the most experienced man on the crew, drove separately that morning. His children had been begging him to retire for months. Their cousins James Blue Ear, 40, and Robert Blue Ear, 36, were also on the tower crew. So were three of their nephews.
Stephen Blue Ear was 26. Chet Payne was 19. Brian Taylor was 18 years old, one of the youngest men on the job. The eldest Steel brother, Robert, 35 years old, was working on a completely different section of the power plant that morning.
He was the only one of the five brothers who would survive the day. Thursday, April 27th, broke fair and chilly along the Ohio River. The overnight temperature had hovered near freezing at about 40 degrees Fahrenheit. Heavy rain had hammered the site the night before.
The concrete for lift 28 had been poured the previous afternoon, with the final placement finishing around 3:30 p. m. on Wednesday, April 26th. By the time the Thursday morning crew arrived, that concrete was roughly 18 to 20 hours old.
And it had spent that entire curing period in cold, damp conditions that dramatically slow the chemical process that gives concrete its strength. By 7:00, the men were checking in, dropping their number tokens into the collection box. By 8:30, the rig had been jacked up to position for lift 29. By 10:00, cathead gantry number four was hoisting a concrete bucket, and number five was lifting rebar.
Carpenter foreman Kenneth Hill of Midland, Pennsylvania, was on the working deck. He had been living out of a camper across the river for the duration of the job. Electrician William Van Vlack, Jr. , was also up there.
Van Vlack had previously told a state inspection panel that he considered the system extremely dangerous. Around 10:00 that morning, as the third bucket of concrete was being lifted, workers heard the sharp crack near cathead number four. Laborer John Peppler, 38 years old from Murphytown, was one of the survivors. He had just sent a concrete basket up when he looked over his left shoulder and saw people and scaffolding falling through the air.
He said they started dropping like dominoes. An unnamed operating engineer in a hoist shanty on the south side also survived. He was a 30-year construction veteran, and this had been his first cooling tower job. The first units on the scene were from the St.
Mary’s emergency squad, all three trucks with 24 trained personnel. Squad members Richard McCulloch and Harry Morgan borrowed a flatbed truck from a nearby Quaker State plant to haul in additional equipment. Within an hour, fire departments from Belmont, Parkersburg, Vienna, and St. Mary’s in West Virginia, plus Marietta and New Matamoras in Ohio, had arrived.
State Police Captain Joe Trupo took charge of the operation. Trupo was a veteran emergency coordinator who had also managed the response at the Silver Bridge collapse in 1967, which killed 46 people, and the Farmington mine disaster in 1968, which killed 78 miners. He said he had never seen a community pull together like Belmont and St. Mary’s did that day.
Cherry picker boom trucks and cutting torches were used to extract bodies from the tangle of twisted steel and broken concrete. Operating engineer Sam Smith of Newport, Ohio, drove a cutting torch truck in, but could not fire it for hours because of leaking acetylene. The last body was pulled from the wreckage shortly after noon. The Belmont Volunteer Fire Station garage and the basement of the Belmont United Methodist Church across the street were converted into a temporary morgue and family identification center.
Doctors Michael Lewis, Richard Hamilton, and E. E. Maxon worked with nurses from the Colin Anderson Center to attend to family members as they arrived. Families walked across the street one at a time, accompanied by clergy and Red Cross workers, to identify their dead.
Many men could only be identified by the wallets, watches, and pocket knives in their pockets. Others were confirmed by their numbered tokens from the morning check-in box. One victim was never positively identified by family members. Co-workers confirmed his identity only from the token he had dropped into the collection box that morning.
The final identification was completed after 10:00 that night. Of the 51 dead, 40 lived in West Virginia. Six were Ohio residents. Five came from Pennsylvania.
Fifteen of the dead were from St. Mary’s alone, a town of fewer than 2,400 people. Sixteen lived in Pleasants County, with a total population of roughly 8,200. President Jimmy Carter and First Lady Rosalynn personally telephoned the Steel family.
Governor Jay Rockefeller flew to the site immediately, riding in a state police car at 115 miles per hour. The mayors of Belmont and St. Mary’s declared the following Sunday a statewide day of mourning. Now the engineering analysis must explain exactly why this happened, and it reveals something that every existing account of this disaster either gets wrong or leaves out entirely.
OSHA inspectors, led by David H. Rowe, were on site within hours. Two days later, on April 29th, 1978, a five-man field team from the National Bureau of Standards arrived. The lead investigators were H.
S. Lew and S. G. Fattal.
They produced two reports that remain the definitive engineering record of the collapse. Report NBSIR 78-1178, published November 1979, ran 188 pages with appendices. Report NBSIR 80-2010, published July 1980, contained the critical finding that changes the entire understanding of this disaster. Their lab tests immediately cleared the hoisting system.
The cables, the chain hoist, the static line, and the cathead anchors were all stress tested at NBS facilities. None had failed before the collapse. They had all been snapped during the collapse as everything came down together. The concrete bucket never struck the cathead beam.
NBS calculated that the bucket was approximately 18 meters below the cathead at the moment of failure, making a strike physically impossible. The wire rope, the pulleys, the gantry steel, and the electric hoists were all within their rated capacities. The equipment was not the problem. The investigators divided the rubble into three suspect systems and tested each one independently: the hoisting and crane system, the scaffolding and formwork system, and the tower shell itself.
They built physical models, ran computer simulations on the SHORE 3 and SAP 4 finite element programs, and reconstructed the loading sequence hour by hour from the moment the previous day’s pour had been completed. The failure was in the connection between the scaffolding rig and the concrete shell. It had two components that worked together to create a disaster. First was the concrete itself.
The National Bureau of Standards team made test cylinders using the same Chris Concrete Company ready mix and cured them in environmental chambers that replicated the rainy 40-degree Fahrenheit overnight conditions that the actual lift 28 had experienced. The result was devastating. At the time of collapse, roughly 18 to 20 hours after placement, the concrete of lift 28 had achieved approximately 220 pounds per square inch of compressive strength. The minimum required to safely support the construction loads above it was 1,000 pounds per square inch.
And that 1,000-pound figure already included no safety margin whatsoever. The strength of the concrete was at roughly 20 percent of the strength it needed to hold. It was at roughly 5 percent of its eventual 28-day design strength of 4,000 pounds per square inch. OSHA area director Stan Elliott later told the Charleston Gazette that this concrete had no business holding anything, let alone 51 men and thousands of pounds of scaffolding and steel.
But here is the finding from the second National Bureau of Standards report that changes everything. The concrete weakness was necessary, but not sufficient to cause the collapse. What actually triggered the failure was the geometry of the static line on cathead number four. As the tower grew taller and the hyperbolic shell curved inwards toward the throat, the ground anchor point of each cathead static line had to be physically relocated to maintain proper alignment.
On cathead number four, the anchor had been moved closer to the center of the tower to accommodate the tightening curve. This seemingly minor adjustment shifted the angle of the static line from nearly vertical to a substantial sideways angle. That new angle generated a horizontal side force on the cathead beam and the jump form beams below it. When the concrete bucket swung during hoisting, the dynamic load combined with that sideways pull to create a shear force that the immature concrete at the bolt anchorages simply could not resist.
The National Bureau of Standards stated it explicitly in their 1980 report. If the base anchor point of the static line had been kept at its previous location before the last relocation toward the center, the effects of the construction loads would have been reduced to such an extent that failure of lift 28 would probably not have occurred. In plain language, a routine geometric adjustment that untrained workers had been making without written procedures on tower after tower for 7 years finally encountered concrete too young and too cold to absorb the forces that adjustment created. OSHA also found that anchor bolts and cones that should have been left in the lower portion of lift 27 had been pulled early so the carpenters could reuse them on the next lift.
Those missing bolts removed a critical load path that might have redistributed the forces. There was also a warning that nobody had acted on. Within 13 days of the collapse, attorney Robert B. Stoolberg of Ralph Nader’s Public Citizen Health Research Group released a citizen’s investigation report.
Stoolberg had pulled OSHA’s pre-disaster inspection files and found that as early as March 1977, 13 months before the disaster, an OSHA inspector had visited Willow Island while tower number one was still being finished. That inspector had documented scaffolding badly in need of repairs, only one temporary stair tower as the sole emergency exit for the entire rig, repairs being made without knowledge of the engineering department, and a specific written warning of potential disastrous consequences if the system was improperly used. That memo was filed and forgotten. OSHA had performed 14 inspections of the Pleasants Power Station site since 1973.
Not a single one had examined Research-Cottrell’s tower work. When Stoolberg presented his findings to Congress on June 30th, 1978, Representative Robert Mollohan of West Virginia dismissed him on the record, saying, “Any flannelmouth can get up here and make a charge. ” OSHA administrator Eula Bingham called the report untrue and misleading. In hindsight, Stoolberg was right, and every official who dismissed him was wrong.
OSHA Assistant Secretary Eula Bingham announced the citations in Washington on June 8th, 1978. Research Cottrell received 16 citations. Ten were classified as willful violations, the most serious category. The willful citations included failure to field test concrete before removing forms, improper scaffolding anchoring, improperly designed specifications, no erection instructions on site, and inadequately trained employees.
United Engineers and Constructors received two serious citations. Pittsburgh Testing Laboratory, the firm responsible for verifying concrete strength, received two serious citations. The total proposed fines came to $108,300. The cases were eventually settled for $85,300 total.
Dividing that by 51 workers who died results in roughly $1,700 per life lost. In today’s money, that is about $5,600 per worker. Joe Powell, president of the West Virginia AFL-CIO, told Congress that judged as a percentage against profits, such fines would not constitute a tenth of a percent, and most assuredly would not serve as a deterrent against future negligence. Under the Occupational Safety and Health Act, a willful violation causing death carried potential criminal penalties of up to a $10,000 fine and 6 months in jail per count.
The Labor Department referred the case to the United States Department of Justice. A federal grand jury was convened in West Virginia. After hearing evidence, no indictments were returned. Not a single individual at Research Cottrell, United Engineers, Pittsburgh Testing, or Chris Concrete was criminally charged for the 51 deaths.
The general contractor hired the prestigious New York engineering firm Lev Zetlin Associates to perform a counter investigation. Lev Zetlin Associates argued that the premature removal of anchor bolts from lift 27 was the dominant cause, not concrete strength or static line geometry. The National Bureau of Standards responded by running the analysis both ways, using SHORE 3 and SAP4 finite element computer programs, and concluded that even with all the lift 27 bolts properly in place, failure was inevitable at the measured concrete strength combined with the static line force vector. Both firms produced complete sets of physical models and computer simulations.
The engineering community sided with the National Bureau of Standards. The regulatory response was painfully slow. In April 1978, there were only seven OSHA construction inspectors covering the entire state of West Virginia. Nationally, OSHA had roughly 1,500 inspectors for 5 million work sites.
After Willow Island, OSHA deleted the generic formwork removal schedule that had allowed contractors to decide for themselves when concrete was strong enough. They added the requirement that concrete must be field tested before any forms or shoring are removed. They banned tying all formwork sections together in a continuous ring after a follow-on accident at the Satsop nuclear plant in Washington state killed two more workers. At Satsop, the formwork sections had been connected, and when one section failed, it pulled the adjacent sections down with it.
Had Satsop used the same continuous ring design as Willow Island, the death toll would have been far higher. OSHA also commissioned a 1983 publication by the National Institute of Standards and Technology titled “A Guide for Safety Inspection of Jump Form and Slip Form Systems Used in Concrete Cooling Tower and Chimney Construction. ” The Nuclear Regulatory Commission later adopted the Willow Island disaster as their foundational case study in safety culture policy, publishing it as Nuclear Regulatory Commission Safety Culture Communicator Case Study Number 3 in September 2011. But it took more than 7 years to even propose new comprehensive concrete and masonry construction rules.
Those rules were finalized only in June 1988, and they were finalized only because of a second disaster. On April 23rd, 1987, exactly 4 days short of the 9th anniversary of Willow Island, the L’Ambiance Plaza building in Bridgeport, Connecticut collapsed during lift slab construction and killed 28 more workers. It took a second mass casualty event to force the regulations that Willow Island should have produced a decade earlier. New scaffolding rules did not arrive until 1990, 12 years after 51 men fell.
The eldest surviving Steel brother, Robert, 35 years old, drove to the Willow Island Baptist Cemetery on the morning of April 28th, 1978 to arrange burial plots for his four brothers. The cemetery had stopped selling plots 2 years earlier because it had completely run out of room. Cemetery trustee Jim Brammer had to find space in what had been a walkway between existing graves. Robert paid $680 in cash for the plots.
The four brothers, Larry, Ronald, Ernest, and Miles, were buried side by side. The family deliberately positioned the graves so the brothers would face the tower that killed them. Uncle Emmett was placed a short distance away. A single memorial service was held on Sunday, April 30th, 1978 at the Pleasants County Middle School gymnasium for the five Steels.
More than 2,000 mourners attended, roughly one quarter of the entire population of Pleasants County coming to one funeral service. Reverend Amos McVay kept the eulogy short. The Washington Post headline on April 29th, 1978 read, “Four boys, I lost four boys. ” That was the patriarch Lee Steel sitting in his hollow watching his only surviving son make funeral arrangements for the other four.
Angie Colvin was a fifth grader at Pleasants County Middle School the morning of the disaster. Her father was Larry Gail Steel. She lost 13 family members that day. In a 2018 interview with West Virginia Metro News, she recalled being outside during gym class when she heard the sirens.
Everyone in the county heard the sirens. She still passes the tower and the memorial almost every day on her way through town. Her son, Anthony Lawer, the grandson of Larry Gail Steel, began working on a memorial project as a middle school social studies assignment years later. He completed it as his Eagle Scout service project.
With $73,000 raised from the community, on April 27th, 2002, the 24th anniversary of the collapse, a memorial shaped like a cooling tower and bearing all 51 names was dedicated about 200 yards from the actual collapse site along West Virginia Route 2. Tower number two was eventually completed using modified procedures and stricter concrete testing. The Pleasants Power Station entered commercial service in 1979 and 1980. For decades after, workers and local residents could see a visible color line at the 166-foot mark on tower number two, where the old concrete from before the collapse met the new concrete poured afterward.
A permanent scar marked exactly where the scaffold had peeled away that April morning. The power station operated for over 40 years before being mothballed in June 2023. The towers still stand along the Ohio River, visible from the road, and the memorial still faces them. Willow Island remains the deadliest single-event construction accident in United States history.
Fifty-one men went to work on a Thursday morning in April 1978. They dropped their numbered tokens into a box, climbed 170 feet into the air, and stood on concrete that was not ready to hold them. A wire rope was anchored 6 feet from where it should have been. A warning memo sat in a filing cabinet in an OSHA field office, 13 months old and unread.
And the total penalty that the federal government collected for all 51 of those lives came to $1,700 per worker.
