Crews cranked a Folsom Dam gate open about 2 feet, and within seconds nearly 40 percent of the lake behind it emptied down the American River

A crew at Folsom Dam in California cranked open a spillway gate to keep water moving during a scheduled shutdown.

The lake behind it sat at full capacity.

At an opening of about 2.4 feet, the operator felt a strange vibration and heard grinding.

Within five seconds, the gate swung completely open.

There had been no warning before it happened.

So what actually gave way inside that steel arm, and why did decades of normal openings hide the flaw until that one morning?

What actually broke inside the gate

Folsom’s spillway gates are curved steel radial gates, each riding on a pair of pin hinges called trunnions, with struts and diagonal braces holding the curved face rigid against the reservoir’s weight. As the gate opened that morning, the failure began at a diagonal brace between the lowest two struts, right next to the trunnion.

The struts mostly carry compression, but friction at the pin hub interface added bending stress as the gate moved. The diagonal braces take that stress as axial load, and the forces finally exceeded what the bolted connections could hold. The first thing to fail was that bolted brace connection, not the steel face itself.

Once the connection released, the load had nowhere to go. Investigators traced the collapse to the entire bracing assembly giving way at once. The gate stopped being a gate and simply became an open door.

Inside the flood itself

The gate released roughly 40,000 cubic feet per second, close to 300,000 gallons every second, and uncontrolled flows continued for about seven hours down the American River toward Sacramento.

That was large, but not unmanageable. The rated safe channel capacity downstream was 115,000 cubic feet per second, and a second dam downstream helped moderate the surge. Recreationists were evacuated, but no significant downstream damage followed.

The reservoir absorbed the real loss. Nearly 40 percent of the storage drained from the 18 square mile lake before crews could close the breach, and repairs eventually cost about $20 million.

Why the design missed it for decades

After a multi year investigation, the Bureau of Reclamation pointed to a design flaw: friction at the pin hub interface was never factored into the original gate analyses, a conclusion laid out in the agency’s record of the failure. That omission matched standard engineering practice when the gates went in during the mid 1950s.

The slow part is what makes this strange. Corrosion on the trunnion pin’s bearing surface gradually raised friction there, which in turn raised the bending stress in the strut and the axial force in the brace. That was a condition that changed gradually while the gate kept working normally for years.

Engineers later reproduced the exact failure in finite element models using a trunnion friction coefficient of 0.25, a figure confirmed separately by physical friction tests on the failed pin. Routine openings had always stopped automatically at small increments, but that stop was overridden as normal procedure, and the gate failed as the opening neared 2.4 feet.

What the sudden river did downstream, and what changed after

The surge did not stay a dam story for long. The sudden pulse of cold freshwater reaching San Francisco Bay was atypical for summer, and a public broadcasting account of the dam reports that salmon and striped bass were fooled into starting their fall migrations weeks early. That effect sits outside the engineering reports, which deal only with steel.

It is the kind of ripple that rarely shows up in a failure analysis, yet it echoes other places where moving water meets wildlife nobody designed for, the same way a juvenile sturgeon slipping through a turbine wheel or an eel pushed through spinning blunt blades shows how fish read infrastructure built with no thought for them at all.

Because the fix went further than the one broken part, the gate was replaced entirely, and the other seven radial gates at the dam were retrofitted to updated standards.

What changed everywhere else because of one bolt

The fix did not stay local. Federal regulators soon required an analysis of all Tainter gates nationwide, using a friction coefficient of 0.3 wherever no measured value existed. Reclamation, the Army Corps of Engineers and federal power regulators all launched programs to evaluate radial gates under their watch.

State regulators moved too, directing dam owners across California to fully inspect and evaluate the structural integrity of their own radial gates. That single failure triggered a national effort that reshaped design practice for radial gate structures everywhere.

Nobody can promise a hidden joint like this will never slip again somewhere else. But the lesson the investigation left behind is plain: measure the friction at the hinge instead of assuming it, and inspect the pivot as carefully as the parts everyone already watches.