South Fork Dam — the Johnstown Flood: a Plugged Dam Overtopped and Killed 2,209

Summary

On the afternoon of 31 May 1889, the South Fork Dam — an earthfill embankment some 22 metres high and 284 metres long on the Little Conemaugh River, 14 miles above Johnstown, Pennsylvania — was overtopped by floodwater and eroded away in roughly an hour, releasing Lake Conemaugh and its 14.55 million cubic metres of water down the valley. The flood reached Johnstown about an hour later as a churning mass of water and debris and killed 2,209 people. It remains the deadliest dam failure in United States history. The embankment itself did not fail by piping, sliding or foundation defect; it failed because it had no margin left to pass a flood. A succession of owners had lowered its crest, removed its low-level discharge pipes, and screened its single spillway to keep fish in the lake. When extraordinary rain raised the reservoir, the water ran over the top and cut the dam open.

The dam was not new and was not the work of careless hands at the moment it broke. It had been built by the Commonwealth of Pennsylvania between roughly 1838 and 1853 as the Western Reservoir, a feeder for the state's canal system, and was for its day a competent embankment with a cut-stone spillway and five cast-iron discharge pipes that let an operator draw the lake down at will. The railroad made the canal obsolete; the reservoir was abandoned, partially breached in 1862, and left derelict for nearly two decades. In 1879 the site was acquired by the South Fork Fishing and Hunting Club — a private retreat for Pittsburgh's industrial elite, including Henry Clay Frick and, later, Andrew Carnegie — which rebuilt the embankment as a resort lake without a civil engineer in charge. Every modification the club made to suit a pleasure lake subtracted from the dam's ability to survive a flood: the discharge pipes were never replaced, the crest was cut down by about a metre for a carriage road, and iron fish screens were fixed across the spillway. None of these acts was individually unthinkable; together they converted a sound embankment into one that could not pass its design storm.

The storm came on 30–31 May 1889, when six to ten inches of rain fell on the catchment in 24 hours. The lake rose toward the lowered crest, the debris-clogged spillway could not pass the inflow, and by early afternoon water was running over the embankment. Earthfill has almost no resistance to overtopping: the overflow cut a notch in the downstream face, the notch deepened and widened by headcut erosion, and the dam unzipped. The 1891 investigation by the American Society of Civil Engineers concluded the dam would have failed regardless of the club's changes — a finding that protected the club from liability and that modern hydraulic analysis has since overturned. South Fork is the canonical American lesson that a dam is only as safe as its spillway and its freeboard, and that an embankment with no way to release water is a flood waiting for a date.

---

Timeline

1838–1853

Built as the Western Reservoir

The Commonwealth of Pennsylvania constructs an earthfill dam on the Little Conemaugh to feed the Western Division Canal at Johnstown. It is fitted with a cut-stone spillway through the right abutment and five cast-iron discharge pipes in a stone culvert that let an operator draw the lake down.

1850s–1860s

Canal obsolescence

The Pennsylvania Railroad supersedes the canal. The reservoir loses its purpose; the state sells it to the railroad, which has no use for it either. The discharge pipes are eventually sold for scrap.

1862

Partial breach

The neglected dam fails partially, emptying the lake with little damage because no large settlement lies directly below. The remnant embankment is abandoned and left exposed for nearly two decades.

1879

Club acquires the site

The South Fork Fishing and Hunting Club, a private retreat for wealthy Pittsburgh industrialists, buys the derelict reservoir to create a resort lake renamed Lake Conemaugh.

1879–1881

Rebuild without an engineer

Benjamin Ruff supervises reconstruction with no qualified civil engineer in charge. The 1862 breach is backfilled with earth, rock, mud and debris rather than properly reconstructed; the discharge pipes and their culvert are not restored; and the crest is cut down by about three feet (roughly one metre) and widened to carry a carriage road, sacrificing freeboard and lowering the level at which the lake will overtop.

1880s

Fish screens added

Iron screens are installed across the spillway to keep the lake's stocked game fish from escaping. The screens present a surface on which storm debris can lodge and block discharge.

1881–1889

Recurrent leaks, no relief

The embankment leaks repeatedly in the years after reopening. With no discharge pipes, the club cannot draw the lake down to inspect or relieve the dam; leaks are patched superficially from the surface.

30–31 May 1889

Extreme rainfall

A severe storm drops an estimated six to ten inches of rain on the catchment in 24 hours. The Little Conemaugh rises rapidly and the lake climbs toward the lowered crest.

31 May 1889, morning

Spillway chokes

Debris collects against the fish screens and grating in the spillway, throttling the only relief path. Workers try to clear the screens and to raise the crest with earth and a hastily cut secondary channel, without success.

31 May 1889, ~midday

Overtopping begins

Water rises to and over the lowered crest. Overflow begins to cut into the downstream face of the earthfill embankment.

31 May 1889, ~15:10

Catastrophic breach

The overtopping erosion deepens into a headcut that unzips the embankment; the dam gives way and releases Lake Conemaugh's ~14.55 million m³ of water. The flood reaches Johnstown about an hour later and kills 2,209 people.

1891

ASCE investigation

An American Society of Civil Engineers committee reports that the dam would have failed even if maintained to its original design — a conclusion that absolves the club and that later analysis disputes.

The Build: A Canal Reservoir Reduced to a Pleasure Lake

The South Fork Dam began life as serious infrastructure. Built by the Commonwealth of Pennsylvania over roughly 1838 to 1853 as the Western Reservoir, it stood about 72 feet high and 931 feet long across the Little Conemaugh, impounding a lake to feed the Western Division Canal at Johnstown, 14 miles downstream and some 400 feet lower in elevation. As state-built dams of the period went, it was a properly conceived embankment. It had a cut-stone overflow spillway carved through rock at the right abutment, and — crucially — a battery of five large cast-iron discharge pipes set in a stone culvert at the base, controlled from a wooden tower, by which an operator could lower the reservoir whenever the dam needed inspection or relief. In its original state the dam could both spill its surplus and empty itself on command.

Two things then removed the reason for the dam and, with it, its care. The Pennsylvania Railroad made the canal obsolete within a decade of the reservoir's completion, and the structure passed from the state to the railroad to private hands with no one needing the water; the discharge pipes were sold for their scrap iron. In 1862 the embankment partially breached and drained, and because no large town then sat directly in its path the failure passed almost unnoticed. For close to twenty years the broken dam lay open in the valley, its purpose gone and its outlet works stripped away.

When the South Fork Fishing and Hunting Club acquired the site in 1879 it wanted not a reservoir but a lake — a sheet of water for boating and stocked bass, ringed by cottages for the families of Pittsburgh's steel and rail magnates. The reconstruction that followed, supervised by Benjamin Ruff with no civil engineer in responsible charge, treated the dam as scenery rather than as a hydraulic structure. The 1862 gap was simply backfilled with earth, rock and whatever was at hand, never engineered as a homogeneous, keyed-in embankment, and the discharge culvert and pipes were not restored, so the new lake had no bottom outlet at all. The dam the club called finished could spill water over its lip and do nothing else: it could not be drawn down, inspected from within, or relieved when it leaked — and it leaked repeatedly in the years that followed.

The Failure Sequence: No Freeboard, No Outlet, a Blocked Spillway

The club then made the two changes that decided the outcome. To run a carriage road along the crest, the top of the embankment was cut down by about three feet and widened — sacrificing the freeboard that is an embankment dam's primary defence against overtopping. And to keep the stocked fish in the lake, iron screens were fixed across the spillway — converting the dam's one safety valve into a trap for the debris a flood carries. The dam now had no low-level outlet, reduced freeboard, and a relief channel that would clog precisely when it was most needed. Each deficiency was survivable alone; the combination left no margin.

The storm of 30–31 May 1889 supplied the load the modified dam could not carry. An estimated six to ten inches of rain fell on the catchment in a single day, the heaviest in regional memory, and the Little Conemaugh poured into Lake Conemaugh faster than the choked spillway could release it. As the lake rose, storm debris piled against the fish screens and the spillway grating, throttling the discharge still further. Club men and labourers worked through the morning to clear the screens, throw up earth on the crest, and open a rough secondary channel at the far abutment, but the inflow outran every effort. By around midday the water reached the lowered crest and began to flow over the top of the earthfill.

Overtopping is the mechanism that destroys embankment dams, and it works quickly. Compacted earth has little resistance to water flowing over and down its unprotected downstream face. The overflow cut a notch in that face; the notch eroded backward and downward as a migrating headcut, steepening and widening until the breach reached the reservoir and the full head of the lake drove through it. At about 15:10 the dam gave way. Roughly 14.55 million cubic metres of water — some 20 million tons — left the lake in under an hour and ran down the narrow valley, gathering houses, trees, rail cars and a stone-arch railroad bridge into a moving wall of debris that struck Johnstown about an hour after the breach. The recorded death toll was 2,209, with hundreds never identified and many never found, the worst dam-failure death toll in United States history.

The Reckoning: An Investigation That Cleared the Club, and the Verdict Reversed

The American Society of Civil Engineers convened a committee that reported in 1891. Its conclusion, much quoted afterward, was that the dam would have failed in that storm even had it been maintained to its original specifications — that the rainfall was so extreme the embankment was doomed regardless of what the club had done to it. The finding had consequences beyond engineering. The club's members were among the wealthiest men in America, and survivors' suits failed: the courts treated the flood as an Act of God for which no individual or corporation was liable. By locating the cause in the weather rather than in the modifications, the 1891 report gave that exoneration a technical foundation.

Modern hydraulic analysis has dismantled it. A 2016 dam-breach hydrology study by Neil Coleman, Uldis Kaktins and Stephanie Wojno reconstructed the original and the modified dam and showed that the club's changes roughly halved the structure's discharge capacity — from a configuration with higher crest, an auxiliary spillway and five discharge pipes able to pass on the order of 197 cubic metres per second at overtopping, down to about 96 in the rebuilt state. Their conclusion is the opposite of 1891's: a dam restored to its original height, with its pipes and an unobstructed spillway, would very likely not have overtopped, would probably have survived the storm, and thousands would have lived. The mechanism was never mysterious — overtopping of an earthfill embankment whose owners had stripped away its freeboard, its outlet and the use of its spillway. South Fork is the case in which engineering at first agreed to call a designed-in deficiency an act of God, and in which the calculation, done properly a century later, named the deficiency instead.

---

Contributing Factors

01

Freeboard sacrificed for a road

Lowering the crest by about three feet to carry a carriage road removed the margin between full pool and overtopping. Freeboard is an embankment dam's primary defence against being overrun by a flood; cutting it down directly reduced the storm the dam could survive and set the level at which it would fail.

02

No low-level outlet

The five cast-iron discharge pipes that originally let an operator draw the reservoir down were sold for scrap before the club rebuilt the dam and were never replaced. With no bottom outlet, the lake could not be lowered to relieve the dam under threat, to inspect the recurring leaks, or to reduce the head during the storm — a dam that can only spill over its top has surrendered control of its own water.

03

A blocked spillway

Iron fish screens fixed across the single spillway, to keep stocked fish in the lake, turned the dam's safety valve into a debris trap. In the flood the screens and grating clogged, throttling the only relief path exactly when maximum discharge was required, negating the structure's capacity to pass its design flood.

04

Reconstruction without engineering

The club rebuilt the embankment under a supervisor with no engineering training and no civil engineer in responsible charge. The 1862 breach was backfilled with earth, rock and debris rather than reconstructed as a homogeneous, keyed embankment, and no analysis tied the lowered crest and screened spillway to the flood the dam had to survive. The deficiencies were never assessed because no one qualified assessed them.

05

Overtopping erosion of earthfill

Once water ran over the lowered crest, the dam's fate was set by material behaviour: compacted earth has almost no resistance to overtopping flow. The overflow cut a headcut in the downstream face that migrated back into the embankment and unzipped it within about an hour. The initiating event was hydraulic — lost freeboard and blocked relief — but the destruction was the inevitable erosion of an unarmoured earthfill being overrun. ---

Aftermath

The Johnstown Flood killed 2,209 people and destroyed much of the city in minutes; it remains the deadliest dam failure in United States history and one of the worst civil disasters of the nineteenth century. The Red Cross, under Clara Barton, mounted one of its first major domestic relief operations at the site. No liability was ever assigned and the club's wealthy members paid no compensation, with the 1891 ASCE report serving as the technical shield for that outcome. The lasting engineering legacy was the principle the failure burned into American practice — that a dam's safety is governed by its spillway capacity and its freeboard, that an embankment must retain a usable low-level outlet, and that no decorative or commercial alteration may reduce its ability to pass a flood. The disaster fed the slow rise of state dam-safety regulation and inspection in the decades that followed, and modern hydraulic reassessment turned South Fork from an act of God into the textbook byword for a dam whose owners removed its defences one convenience at a time.

---

Lessons

Never trade a dam's freeboard for anything: the margin between full pool and the crest is the structure's first defence against overtopping, and lowering it for a road, a view, or a wider top directly lowers the flood the dam can survive.
Keep a working low-level outlet on every reservoir, so the lake can be drawn down to inspect the embankment, relieve it under threat, and reduce head during a flood — a dam that can only spill over its top has lost control of its own water.
Treat the spillway as untouchable: it is the safety valve, and screening, grating or otherwise obstructing it for fish, debris or appearance converts the one relief path into the cause of failure when the flood arrives.
Put a qualified engineer in responsible charge of any dam, old or rebuilt, and require that every modification be analysed against the design flood — deficiencies invisible to a layman are obvious in the calculation.
Reject "act of God" as a verdict before running the numbers; the storm that overran South Fork was survivable for the dam as originally built, and the failure was the sum of choices, not the weather. ---