Case Study · Dam Failure · 1963
October 9, 1963. 10:39 PM. A 260-million-cubic-meter landslide hit the Vajont reservoir in Italy at 68 mph in 45 seconds. The dam — one of the most sophisticated arch dams ever built — withstood a 250-meter wave overtopping it. The water fell 500 meters into the valley and obliterated Longarone and four other villages. Nearly 2,500 dead. The slope's instability had been identified during construction. The water level had been raised above the safe threshold. By five days before the disaster, the slope was moving at 8 inches per day and local mayors were calling for evacuation. A dam disaster without dam failure.
Longarone, Italy · October 9, 1963
The Vajont Dam stands 250 meters tall in a narrow gorge in the Italian Alps — one of the tallest arch dams ever built, completed in 1959, and an engineering achievement of the postwar era. The AGU Landslide Blog describes the physical setting: "Vajont is located in the south-eastern Alpine region of northern Italy, with the dam immediately upstream of a narrow slot canyon leading down to the wide Piave River valley." The dam's design made use of the gorge geometry: an arch dam 160 meters wide could retain 150 million cubic meters of water. What the dam builders discovered partway through construction was that the valley's geology had been shaped by an ancient landslide — an unstable slope on Monte Toc above the reservoir's future shoreline. The Earth Magazine account of the disaster documents the warning: "A German engineer named Leopold Müller and two Italian geologists, Eduardo Semenza and Franco Giudici, studied the valley and concluded that the slopes above the reservoir were unstable and that the ancient slide could move again. However, their warnings fell on deaf ears; by then, dam construction was nearly completed."
During the filling of the reservoir from 1960-1963, a clear pattern emerged: as the water level rose, slope movement on Monte Toc increased. The Study.com account of the disaster documents the specific pre-disaster data: "By October 4, 1963, these movements were being measured at almost 8 inches per day. Engineers predicted an imminent landslide, and mayors of local towns and villages began to issue calls for evacuation, which were largely ignored." The reservoir level had been raised to 710 meters — 10 meters above what was considered the safe operating threshold of 700 meters. Earlier, on November 4, 1960, a 700,000-cubic-meter landslide had entered the reservoir as a warning sign. The company operating the dam, SADE, had issued assurances that any landslides would be small and manageable.
On October 9 at 10:39 PM, those assurances ended. The AGU Landslide Blog documents the sequence: "A block of approximately 270 million cubic metres detached from one wall and slid into the lake at velocities of up to 30 m/sec (approx. 110 km/h). As a result a wave over topped the dam by 250 m and swept onto the valley below, with the loss of about 2500 lives. Remarkably the dam remained unbroken by the flood." The slide completed in 45 seconds. The displaced water — approximately 50 million cubic meters — rose 250 meters up the opposite valley wall, destroying the village of Casso, then careened over the dam's crest and fell approximately 500 meters into the Piave Valley. The All That's Interesting account of the disaster describes the impact: "The pastoral villages of the Piave Valley were swept away in a matter of minutes, killing families as they slept and leaving those who survived homeless and displaced."
Oct 9, 1963
Date
~2,500
Deaths
250m wave
Over the Dam
Dam survived
Structurally Intact
8 in/day
Slope Movement Oct 4
The Science
Think of a reservoir as a confined body of water under a rigid boundary (the dam) and an open boundary (the valley sides and floor). When a large mass enters the reservoir rapidly, it displaces water volume equal to its own submerged volume — in the Vajont case, 260 million cubic meters of rock entering the reservoir displaced an enormous volume of water virtually instantaneously. This creates a wave of the displaced water that must go somewhere. At Vajont, the gorge confined the water; the only directions were up the opposite valley wall and over the dam. The AGU Landslide Blog documents the physics: the slide "travelled up to 140 m up the opposite bank" before the water overtopped the dam. The wave wasn't a slow-building flood. It was an impact event. From the moment the slide entered the reservoir to the moment the towns of Longarone were struck: approximately one minute. No warning time was possible after the slide began. The only prevention was preventing the slide — which required either not filling the reservoir near the unstable slope, or evacuating downstream communities before the already-accelerating movement reached failure.
The Vajont disaster had a warning system — it just wasn't connected to action. The Study.com account of the disaster documents that slope monitoring was occurring: "Throughout the reservoir's short operational life, a clear pattern emerged whereby as its fill increased, measured earth movements within the Vajont basin increased as well." This is a direct, measurable signal that the reservoir operations were driving slope instability. The critical decision point came when the water level was raised 10 meters above the established safe threshold of 700 meters. If the threshold existed because engineers believed instability increased above that level, raising above it was a direct decision to accept higher slope failure risk. By October 4, the movement rate of 8 inches per day signaled imminent failure to engineers familiar with landslide precursors. The July 1962 Ghetti Report had specifically found that under certain landslide conditions, wave height could exceed the dam crest — but its conclusions were not widely shared with relevant authorities. The mayor calls for evacuation that were "largely ignored" were the institutional last chance. The Vajont disaster is a textbook case of warnings detected, analyzed, partially acted on — and ultimately not followed through to the evacuation that could have emptied the downstream valley before the wave arrived.
The Vajont-type hazard is not unique to Italy. Reservoirs created in mountain valleys throughout the Western US are surrounded by slopes that have their own geological histories of movement. The Army Corps of Engineers and Bureau of Reclamation conduct slope stability assessments for major reservoirs — but not all reservoirs near unstable slopes have up-to-date hazard assessments. Additionally, dam overtopping from extreme rainfall events is a documented and increasing risk: if a dam's spillway cannot discharge as fast as the reservoir is filling during an extreme storm, the reservoir will rise until it overtops. FEMA's National Dam Safety Program maintains inundation maps for regulated dams, but these maps are designed around dam structural failure scenarios rather than overtopping from landslide-generated waves. The Association of State Dam Safety Officials (ASDSO) cites the Vajont disaster in its educational materials specifically because the lesson — that the hazard is the downstream inundation, not the structural integrity of the dam — remains underappreciated in public understanding of dam risk.
Timeline
01
1957: Construction begins on the Vajont Dam, designed by Italian engineer Carlo Semenza. During construction: geologists Leopold Müller (German), Eduardo Semenza (Carlo's son), and Franco Giudici study the valley and conclude the slopes above the future reservoir are unstable — they are sitting on an ancient landslide of Monte Toc. Their report is not sent to authorities. Construction nearly complete by time warnings are raised. February 1960: first filling begins. November 4, 1960: a 700,000-cubic-meter landslide enters the reservoir — warning sign. Engineers lower water level. Seismic monitoring installed.
02
1960-1963: Clear pattern: as water level rises, slope movement increases. SADE assures residents landslides would be small and manageable. July 1962: Ghetti Report finds wave height could exceed dam crest under certain landslide conditions — not widely disseminated. 1963: water level raised to 710 meters — 10 meters above established safe threshold of 700 meters. October 1963: movement rates increasing. October 4: slope movements measuring approximately 8 inches (20 cm) per day — engineers predict imminent landslide. Local mayors call for evacuation of villages. Calls largely ignored. Residents not evacuated. People go to sleep on the night of October 9.
03
October 9, 1963, 10:39 PM: 260-270 million cubic meters of rock detach from Monte Toc and slide into the reservoir at approximately 68 mph (110 km/h). Slide completes in 45 seconds. 50 million cubic meters (13 billion gallons) of water displaced. Wave runs 250-260 meters up the opposite valley wall — destroying the village of Casso at 260 meters elevation. Water crests the dam by approximately 245-250 meters. Water falls approximately 500 meters into the Piave Valley below. Longarone, Pirago, Villanova, Rivalta, and Fae obliterated. Nearly 2,500 dead — most in their beds. One-third of Longarone's population killed. The dam, structurally undamaged, still stands today.
04
SADE executives and engineers tried at the Aquila court. Disaster prompts major reform of Italian dam and reservoir safety oversight. The dam is decommissioned; its bypass tunnel now generates hydroelectric power. The landslide mass, roughly 400 meters deep in the gorge, permanently blocks the former reservoir. The disaster becomes a foundational case study in engineering ethics — multiple professionals knew of the hazard, some raised warnings, the warnings were dismissed for financial and institutional reasons. ASDSO includes Vajont in its dam safety educational materials. The gorge and dam remain accessible; Longarone was rebuilt. The Vajont disaster is the canonical case study demonstrating that dam safety is not only about dam structural integrity — it is about everything that can generate a lethal downstream flood, including the geology above the reservoir.
Human Decisions
The decision chain
The decision to raise the reservoir level above 700 meters — the established safe threshold — was made while slope movement monitoring was already showing elevated movement rates. The mechanism was understood: higher water level meant higher pore pressure in the slope, reducing friction between rock layers, accelerating movement. The safe threshold existed precisely because engineers knew higher levels increased instability. Raising 10 meters above that threshold in 1963 was a decision to increase slope failure risk for the benefit of additional reservoir storage capacity. The Vajont disaster demonstrates the specific danger of incrementally exceeding safety thresholds: each increment appears manageable; the final increment is the one that triggers failure. The decision logic of "we've gone above the threshold before without catastrophe; we can do it again" is the same logic that drives most major infrastructure failures.
By October 4, five days before the disaster, engineers were measuring 8 inches of daily slope movement — a rate that indicated imminent large-scale failure. The Earth Magazine account of the event documents that "mayors of local towns and villages began to issue calls for evacuation, which were largely ignored." This is the specific decision point where the Vajont disaster could have had a very different outcome. The villages of Longarone and its neighbors were in the flood path, not on the slope. They needed only to be moved to higher ground — not necessarily far, just above the flood inundation level. The 45-second slide itself gave no time to act. But the 5-day warning window, had the mayors' calls been heeded, could have emptied the valley before the slide. The Vajont disaster is specifically a case where the preventive action was available and was not taken.
What downstream communities need to know
FEMA's National Dam Safety Program and state dam safety offices maintain Emergency Action Plans (EAPs) for regulated dams, which include inundation maps showing the predicted flood path and depth downstream of each dam in the event of failure. These maps document the areas that would be flooded, how long the flooding would take to arrive, and how deep the water would be. Most states make some form of dam inundation map publicly available. FEMA's National Inventory of Dams (NID) at nid.sec.usace.army.mil lists regulated dams with basic information. The Vajont lesson for US residents: the threat from a dam upstream of you doesn't require the dam to fail structurally — extreme rainfall, an earthquake, or (in mountainous areas) a slope failure adjacent to the reservoir can generate the same downstream flooding. Knowing whether you live downstream of a regulated dam and understanding the inundation scenario is fundamental dam safety preparedness.
The Vajont wave reached the towns of Longarone within approximately one minute of the slide. The 1889 Johnstown flood (covered in this series' anchor) reached Johnstown within 35 minutes of the dam failing. The 1975 Banqiao flood wave moved at 50 km/h. Dam failure events are fundamentally different from riverine floods (which may give days or hours of warning) and are much faster than the public typically appreciates. This speed is why emergency planning for dam failure is so specific about pre-positioning rather than last-minute response: the Emergency Action Plan must be activated based on predictive monitoring (rising water levels, seismic activity, slope movement) — not based on observing the flood wave. Downstream residents who live in dam inundation zones should not plan to wait until they see or hear a flood wave. They should know their evacuation route and leave immediately when a dam emergency alert is issued.
The cascade lesson
The Vajont disaster defines the landslide-into-reservoir dam disaster lesson: that catastrophic downstream flooding doesn't require dam failure. The downstream inundation zone is the hazard zone regardless of whether the dam collapses or is simply overtopped. For US residents, this means understanding dam inundation zones includes not only the dam structure itself but the broader context of what could happen to the reservoir — extreme storms, adjacent slope failures in mountainous reservoirs, and seismic activity. Knowing your dam inundation zone is the prerequisite. Knowing your evacuation route and responding immediately to dam emergency alerts is the action that saves lives.
What You Can Do Now
The Vajont lesson is about hazard awareness downstream of dams — what can cause a flood wave without dam structural failure, how fast it arrives, and what preparation is required before an emergency alert is issued.
FEMA's National Inventory of Dams (nid.sec.usace.army.mil) lists regulated dams across the US with location and basic hazard information. Most state dam safety offices maintain publicly accessible dam inundation maps or Emergency Action Plans (EAPs) that show the predicted flood path downstream. Your state dam safety office (typically under the state department of water resources, environmental protection, or natural resources) is the direct source. If you live in a valley downstream of a major reservoir, search "[state] dam inundation maps" or "[state] dam safety emergency action plans" to find the information for dams near you.
Dam inundation zone lookup guideDam failure flood waves travel fast — typically faster than riverine floods. In a mountain valley like Vajont, arrival time measured in minutes; in larger valley systems, 30-60 minutes or more. A dam failure evacuation route is specifically to high ground above the inundation depth — not just to higher roads within the valley. Identify the elevation of your home or workplace relative to the predicted inundation depth for a dam failure event. Know which roads lead out of the inundation zone and perpendicular to the river valley. Practice the route. The Vajont towns had no realistic evacuation after the slide began — the window was the 5 days of elevated movement warning before it.
Dam failure evacuation planning guideWireless Emergency Alerts (WEAs), which appear on cell phones automatically within an affected area, are used for dam failure events. County emergency notification systems (registerable via your county's official website) provide additional advance notice. For major dams with downstream populations, Emergency Action Plans specify the notification chain — who is responsible for issuing the alert, at what threshold, and through what channels. Registering with your county's notification system before an emergency ensures you receive alerts as soon as they are issued, rather than learning about a dam emergency through secondary sources.
Emergency alert registration guideThe specific behavioral lesson of Vajont and every dam failure case study: the time to evacuate is when the warning is issued, not when the flood wave is visible. The Vajont towns couldn't evacuate after the slide began — the wave arrived within approximately one minute. The Teton Dam failure (covered in this series) gave communities in its path only hours, and the time from first visible distress to complete failure was approximately 4 hours. Treat a dam emergency alert with the same immediacy as a tornado warning: leave immediately, take your prepared go-bag, and move to elevation above the predicted inundation zone. Do not wait for additional confirmation; do not stop to gather belongings; do not attempt to watch the event.
Dam emergency response action guideThe Vajont hazard — a slope failure generating a wave that overtops an otherwise sound dam — is not unique to Italy. Mountainous reservoirs in the Western US, particularly in active tectonic regions (California, the Pacific Northwest, Rocky Mountain states), are surrounded by slopes with their own geological histories. USGS and state geological surveys maintain landslide hazard maps. For communities below mountainous reservoirs in active seismic zones, asking your state dam safety office whether the dam's Emergency Action Plan accounts for landslide-into-reservoir scenarios (as distinct from dam structural failure scenarios) is the specific Vajont-motivated preparedness question.
Landslide-reservoir hazard assessment guideDam Failure case study series
Johnstown 1889 covers the defining US dam failure. Banqiao 1975 covers the largest death toll in dam failure history and cascade failure. Teton 1976 covers engineering defects in new dam first-fill failure. Oroville 2017 covers deferred maintenance and the US aging dam infrastructure crisis.
Full dam failure case study seriesSources