Welcome to our new Geotechnical History Blog! This inaugural post, by lead blogger and G-I member Michael Bennett, will be presented in several parts. All notes and image credits will be in the final part.
The Johnstown Flood of 1889: A Catastrophe of Civil Engineering
by Michael D. Bennett, E.I.T., A.M.ASCE
Dedicated to the memory of the victims of the Johnstown Flood of 1889.
On May 31, 1889, the deadliest dam disaster in American history played out in western Pennsylvania. After a torrential storm the previous night had drenched the area around Lake Conemaugh, a mammoth reservoir held back by the massive yet dilapidated South Fork Dam, the rain-swollen lake rose ominously until it eventually overtopped and breached the dam. The breach sent billions of gallons of water and an avalanche of debris thundering down the valley of the Little Conemaugh River. After wiping out miles of railroad tracks, small hamlets, and factories, the wall of water and debris eventually slammed into the prosperous steel hub of Johnstown, killing thousands, injuring thousands more, and obliterating the town. Following the disaster, which was soon nicknamed the Johnstown Flood, survivors demanded that justice be served to the South Fork Fishing and Hunting Club, the elite social group which had owned Lake Conemaugh and the South Fork Dam. Yet the Club and its powerful allies had their thumbs on the scales of justice and unduly influenced how elected officials, judges and juries, and even the American Society of Civil Engineers dealt with the aftermath of the disaster.
Prologue: Westward expansion and the Main Line
The events which would culminate in the destruction of Johnstown in 1889 had begun almost a century earlier. The Appalachian Mountains had long acted as a boundary to colonial settlement in Pennsylvania. This was especially true of the Alleghenies, the highest and westernmost portion of the Appalachians which form the Eastern Continental Divide within the Keystone State. Moreover, the Native Americans controlling the mountains of Pennsylvania were notoriously hostile to European settlers. Following the American Revolution, however, Native American tribes were tragically evicted from the Alleghenies, and European-American settlers soon began to establish towns there. In the 1790s, settler Joseph Schantz and his family founded a village on the western slopes of the Alleghenies at the confluence of the Little Conemaugh and Stony Creek (or Stonycreek) Rivers into the Conemaugh River. The junction of the two valleys consists of a small, flat plain surrounded by near-vertical hills several hundred feet high. Although Schantz left about a decade later, the town remained, still bearing the anglicized form of his name - Johnstown (Hanna 2021).
During the early 19th century, citizens of the newly-formed United States of America continued moving westward, beckoned by the lure of the frontier. As the US frontier steadily advanced past the Alleghenies, the wagon roads, both tolled and free, which had long conveyed travelers and goods over the range became overloaded with westbound personal and commercial traffic. The inadequate roadways clearly needed supplementing, and engineers, entrepreneurs, and politicians alike soon began exploring alternative solutions. In the 1810s, Governor DeWitt Clinton of New York proposed a canal across the Empire State which would connect Albany to Buffalo. As construction on the canal began, critics derided the project as “Clinton’s Folly” and “DeWitt’s Ditch”. Yet the criticisms stopped in 1825 when the Erie Canal was completed and began raking in toll revenue for New York State. As construction began on the Chesapeake and Ohio Canal in Maryland shortly thereafter, the Pennsylvania General Assembly realized that the Keystone State was falling behind. In 1828, the Assembly authorized a network of canals and the novel technology of railroads to connect Philadelphia to Pittsburgh. The system, dubbed the Main Line of Public Works, was finished by 1834 (Burgess and Kennedy 1949, Hanna 2021).
The Main Line was a masterpiece of early civil engineering. Travelers embarking in Philadelphia took a railroad to Columbia on the Susquehanna River, where they transferred to canal boats. From Columbia, the passengers took canals along the Susquehanna and Juniata Rivers to Hollidaysburg on the eastern flanks of the Alleghenies. Yet the crown jewel of the Main Line was the 36-mile Allegheny Portage Railroad, which connected Hollidaysburg and rivers flowing toward the Eastern Seaboard with Johnstown and rivers flowing toward the Gulf of Mexico. The Portage Railroad bridged the crest of the Alleghenies using both long, flat portions of conventional railroads, called “levels”, and inclined plane sections with slopes of 7 to 10 percent. Pulleys powered by stationary steam engines and bolstered with cogged tracks and safety stops towed canal boats mounted on railcars up or down each inclined plane. Finally, at Johnstown, travelers returned to conventional canals for the journey into Pittsburgh. Overall, those taking the Main Line covered the nearly 400 miles between the Delaware and Ohio Rivers in about 4 days (Burgess and Kennedy 1949).
Artist’s rendering of a canal boat being hoisted up a plane on the Allegheny Portage Railroad, with a two-wheeled safety stop behind it to prevent derailing
A replica of the section of the Allegheny Portage Railroad depicted above, reconstructed and preserved by the National Park Service
The need for a reservoir
While the Main Line amazed many sightseers, commercial shipments on the route generated much more revenue for the Commonwealth, which made it essential to keep traffic moving between Philadelphia and Pittsburgh. Yet, even as the Main Line was being built, its civil engineers were noticing that the Juniata and Conemaugh Rivers, which served as both the alignments and the water sources of the canals on the Main Line, often froze over during the frigid Pennsylvania winters. More pressingly, the canals also ran low during the hot, dry summers, especially near Hollidaysburg and Johnstown, since the watersheds upstream of these towns were relatively small (Burgess and Kennedy 1949, Hanna 2021).
To keep the Main Line toll revenues flowing into the Pennsylvania treasury, the civil engineers of the Commonwealth proposed the construction of reservoirs on both sides of the Alleghenies to boost water levels in the canals. Thus, as the Keystone State finished constructing the primary components of the Main Line in the mid-1830s, the state government sent civil engineer Sylvester Welch to survey potential sites for both reservoirs. Based on Welch’s recommendations, the state chose to locate the reservoir for the canals along the Juniata, aptly named the Eastern Reservoir, on the Frankstown Branch of the Juniata River about 2 miles southeast of the canal terminus at Hollidaysburg. For the Western Reservoir, which would supply the canals along the Conemaugh, the state drew upon Welch’s findings and selected a location on the South Fork of the Little Conemaugh River about 14 miles upstream (10 miles due east) of Johnstown. Welch recognized that the 400-foot drop from the Western Reservoir site to Johnstown would give the waters in the reservoir plenty of elevation head to help them complete their relatively long journey to the canals along the Conemaugh (Coleman 2019, Francis et al. 1891, Hanna 2021).
"...a forerunner of the modern geotechnical investigation."
The plan for the dam and the search for the site
In addition to surveying potential reservoir sites, Welch also put forward preliminary sites for the dams which would impound each one. He noted that, if an earth dam were chosen to impound either the Eastern or Western Reservoir, if not both, water could not be allowed to overtop the structure. Welch’s comment reflected how the prevention of overtopping was, by the 1830s, an established part of the standard of care for the engineering of earth dams. Welch elaborated that water leakage through an earthen dam could be prevented either by building a masonry core within the dam or by constructing the upstream half of the dam using a process known as puddling, in which wet clay was placed in thin layers to make it watertight. Ultimately, the Commonwealth chose to value-engineer the dams at both reservoirs by using Welch’s second, decidedly less expensive design alternative. A few years later, the Commonwealth chose civil engineer William Morris to finalize Welch’s preliminary dam designs. The first step Morris took at the Western Reservoir site - less is known about its eastern counterpart - was to have laborers dig shafts and tunnels at the proposed dam location to confirm both the presence of bedrock and the availability of good-quality material for the dam. Such a practice may well be considered a forerunner of the modern geotechnical investigation (Unrau 1979).
Morris used the findings of his crew’s subsurface exploration to convert Welch’s conceptual design for the dam at the Western Reservoir into a final, construction-ready plan. The plan, in its ultimate form, had three main components. First, Morris agreed with Welch that an earth dam would suffice for the Western Reservoir. Second, Morris noted that a drainage outlet at the base of the dam would be necessary to make the full depth of the reservoir available for keeping the canal full. Lastly, Morris observed that, while floods sometimes happened along the South Fork of the Little Conemaugh, “little danger is to be apprehended if proper channels are constructed for their discharge”. Morris thus explicitly stated something that Welch had only implied about the safe design of earth dams. Clearly, only by providing earth dams with sufficient means for safely and quickly lowering the reservoirs they impounded, especially during floods, could such structures be protected from overtopping (Unrau 1979).
Construction of the Western Reservoir began in the spring of 1840 as contractors, supervised by field engineers, cleared and grubbed the dam site and future reservoir bed. The contractors’ work crews then dug an excavation to bedrock and filled it with puddled clay to make the dam watertight. This feature, named a puddle trench, served the same function as a cutoff trench below a modern dam. Next, the crews built a masonry foundation in the center of the dam location to hold five cast-iron discharge pipes, each 24 inches in interior diameter, to be used for controlling the level of the Western Reservoir. To save money while also ensuring the integrity of the dam, Morris had designed the pipes to run only about 80 feet into the 225-foot-wide dam base, at which point they would discharge water into a brick culvert running to the downstream face of the dam. Once the crew had placed the pipes, they began constructing the culvert (Coleman 2019, Francis et al. 1891).
However, the early 1840s were a period of tough economic times in the United States, especially for Pennsylvania, which was still paying off the enormous costs of building the original components of the Main Line. By the spring of 1841, the Keystone State was in such dire financial straits that the Eastern and Western Reservoirs, among many other infrastructure projects, were paused to keep the Commonwealth solvent. The pause at the Western Reservoir lasted a decade, as the state’s weak finances, needed repairs to other infrastructure, and even a cholera epidemic shifted public and political attention elsewhere. When limited funds for the reservoirs became available again in the late 1840s, they were used to finish the Eastern Reservoir. Meanwhile, as political and legal wrangling over the half-finished Western Reservoir continued, the incomplete dam east of Johnstown languished without protection from the elements and sustained several partial washouts, which likely compromised its long-term integrity (Coleman 2019, Francis et al. 1891, Hanna 2021, Unrau 1979).
Eventually, though, the Commonwealth’s finances recovered, and, in early 1851, the General Assembly authorized funding to complete the Western Reservoir. Construction resumed almost immediately, continued once more by the original contractors under the supervision of field engineers. William Morris remained the chief civil engineer for the project, and further value-engineered his design for the Western Reservoir by replacing the planned masonry control tower from which the discharge pipes would be operated with an equally effective and more economical wooden structure. Soon, workers had finished both the culvert for the discharge pipes and the control tower. Next, the crews carefully placed layers of puddled clay on the upstream half of the dam to make it watertight while building the downstream half using weathered rock. Both faces of the dam were covered with riprap to further guard against erosion, with larger riprap used on the downstream face. The specifications governing workmanship on the dam were strict, prohibiting the use of “light, spongy, alluvial, or vegetable material” and requiring that shabby earthwork be removed and corrected (Coleman 2019, Francis et al. 1891).
William Morris’s final design for the Western Reservoir dam
Over the next year, as construction progressed on the dam for the Western Reservoir, the structure steadily rose from the future reservoir bed. By the summer of 1852, the dam was complete and the impounding of the reservoir had begun. Supporting construction was finished the following year. The Western Reservoir was initially filled to a depth of 50 feet, as opposed to its maximum safe capacity of 60 feet, so that the engineers and contractors could address problems with the dam if they arose. While no problems materialized at that time, the practice remains standard for new dams nearly 175 years later (Coleman 2019, Hanna 2021).
Artist’s rendering of the dam at the Western Reservoir as originally completed in 1853. Note the sturdy construction (1), the main spillway (2), the control tower (3), and the discharge pipes (4)
The new dam at the Western Reservoir was an immense structure, and was definitely among the largest dams in the US in 1853. It stood 72 feet high and roughly 860 feet long, and tapered from a base width of about 225 feet to a crest width of just 10 feet. The five discharge pipes and the culvert at the base of the dam allowed the operators of the Main Line both to use the full depth of the Western Reservoir to supply the canals west of Johnstown and to safely and rapidly lower the reservoir if it rose too high. In times of flooding, two spillways, one at each abutment of the dam, could augment the discharge through the pipes and culvert. The main spillway, which the crew had cut through sandstone bedrock, lay at the northeastern abutment. The bottom of the main spillway was about 10 feet below the dam crest, and the spillway was about 70 feet wide at its narrowest point. At the southwestern abutment, an auxiliary spillway lay about 3 feet below the crest and was also roughly 70 feet wide. This spillway had not been excavated to bedrock, perhaps due to budget constraints (Coleman 2019, Kaktins et al. 2013, Unrau 1979).
Some scholars note that the auxiliary spillway was never marked on Morris’s plans for the dam, and thus dispute whether it was ever truly intended to be built. However, later surveys determined that a depression was certainly present at the southwestern abutment. This depression would have functioned as an auxiliary spillway during periods of flooding, regardless of Morris’s design intentions. Moreover, it seems likely that either Morris or the onsite engineers could have directed the contractors to fill in the depression, and thus maximize the storage capacity of the Western Reservoir, had either party not approved of the depression. Intended or not, the auxiliary spillway did not negatively impact the performance of the dam during the early years of operation of the Western Reservoir. Although a few minor leaks were noted at the dam in those initial years, they were readily repaired. In fact, the State Engineer of Pennsylvania declared following an 1856 inspection visit that the Western Reservoir and its dam were in “excellent condition” and could ably serve the Main Line for a long time to come (Coleman 2019, Francis et al. 1891).
To be continued
The opinions and conclusions expressed in this article are those of the author only and do not necessarily represent the views of ASCE or the Geo-Institute.