The early water system of colonial Philadelphia wasn’t much of a system. According to the website PhillyH2O.org — curated by Adam Levine, a historical consultant to the Philadelphia Water Department — public wells were dug in the streets for average citizens, and private wells were dug in the backyards of the wealthy. “Human wastes were disposed of in privy pits, often located in the opposite corner of the backyard,” Levine writes on the site. “Since Philadelphia lots are narrow, that often meant that the water for drinking was coming out of a hole in the ground located 20 feet or less from another hole in the ground where wastes were deposited.”
Not surprisingly, mosquito-borne yellow fever struck the city often; in the summer and autumn of 1793, it killed about 5,000 Philadelphians, or roughly 10 percent of the population. “Close to half the citizenry, including President George Washington (Philadelphia was then the national capital), literally ran for their lives, fleeing the death-haunted metropolis for the surrounding countryside,” wrote historian Carl Smith (City Water, City Life: Water and the Infrastructure of Ideas in Urbanizing Philadelphia, Boston, and Chicago. Chicago and London: The University of Chicago Press, 2013.)
The disease returned over four of the next five summers, Smith wrote, killing roughly as many people in 1798 as in 1793. The ongoing epidemic led city leaders to seriously plan for a coordinated water system, one that proved to be, as Smith put it, “the first truly comprehensive waterworks system in a major American city.”
There had been other water systems in America before Philadelphia’s — Bethlehem, about 70 mi north of Philadelphia, had installed a rudimentary water supply system some 50 years prior. But Philadelphia’s was the first at a large (for that time), municipal scale. And Philadelphians themselves had strong opinions about the need for such a system. According to Smith, favorite son Benjamin Franklin, before he died in 1790, “recommended that Philadelphia construct a gravity-driven aqueduct from the Wissahickon Creek, which flowed into the Schuylkill River a few miles northwest of the city.”
But it took the yellow fever outbreaks for the city to create the Joint Committee on Bringing Water to the City, which issued a report in 1798 calling for a new water supply to mitigate the risk of illness. To oversee the system, the city turned to English-born American architect-engineer Benjamin Henry Latrobe, who was in Philadelphia building the Bank of Pennsylvania, the earliest work of neoclassical architecture in the United States.
According to Smith, Latrobe knew the city’s water was poor; he remarked in his journal in 1798 that water in the city was “not to be drank (sic), and it is worst in the most crouded (sic) neighbourhoods,” where it tasted “as if it contained putrid matter.” His solution? Not Wissahickon Creek, as Franklin had suggested, but the Schuylkill River itself.
Today the Schuylkill forms the western boundary of what is known as Center City, the central business district of Philadelphia, but 200 years ago that river was part of a remote wilderness; the city’s development focused more on the larger Delaware River on the city’s eastern side. According to a 1978 Historic American Engineering Record report on the Fairmount Waterworks written by Jane Mork Gibson, a historian of Philadelphia industry and technology, Latrobe reported that the Schuylkill was the best source of water in the area, but that water was useless unless it could be “raised to an elevated level. To do this, in sufficient quantity, very powerful machinery will be required; and I am very certain that human ingenuity has not hitherto invented anything capable of producing the proposed effect with constancy, certainty, and adequate force, excepting the steam engine.”
Latrobe’s innovative plan actually called for two steam engines, located in two separate pumping stations. Steam engines had never been used before to power a water system; up to this point they had mainly been used in the mining industry in England. The first of the two engines Latrobe proposed, on the east bank of the Schuylkill at Chestnut Street, would raise water from the river, according to Smith. This water would travel through an underground pipe to Centre Square, at what is now the corner of Broad and Market streets — the site of the city’s current city hall and the geographic center of town. “The second engine,” Smith wrote, “located in the square, would lift the water to storage tanks forty feet above the ground, and then gravity would drive it through bored-out pine logs to every part of the city.”
The city raised money for the project by selling shares of the water itself, an early form of user fees. But public hydrants that people could use for free were also installed. Wooden pipes made from the trunks of white oak, yellow pine, and spruce trees were used to convey the water. Levine noted that “each log was bored through its center with an auger either 3, 4.5, or 6 inches in diameter. The bored logs were joined into pipelines with iron couplings and straps.”
The new system was, in the words of environmental engineer C. Drew Brown, the manager of public education for the Philadelphia Water Department, transformative. “One of the things we were successful in was taking forward the basic outline of a water system,” he says. “That is, pumping from a water supply to an elevated storage facility, then supplying by gravity from the elevated storage to the service area.”
Latrobe bequeathed to the city not only the technical achievement of the system but an architectural marvel. The Centre Square pumping station, three stories high and clad in white marble, was admired for its neoclassical beauty. The building concealed both the steam engine and two storage tanks above it. “It featured Doric columns and a domed roof with an opening at the top, recalling the oculus of the Pantheon in Rome, whose purpose was to allow smoke from the engine to escape,” wrote Smith. (Philadelphians nicknamed the building “the Pepper Pot.”)
In 1802, the committee stated that the system was delivering about 400,000 gal. per day. Yellow fever outbreaks diminished, and the system was useful in putting out fires. But despite this — and despite the beauty of its design — the flaws of the pumping station became apparent very quickly.
The steam engines were unreliable and in need of constant repair. “Since the engines were arranged in series, if either one stopped, whether for planned maintenance or by accident, Philadelphians faced the inconvenience of having no running water, not to mention the peril of being without protection against fire,” Smith wrote. And the engines were not safe: “In April 1801 two men who entered the large boiler in Centre Square in order to repair it were suffocated to death, and it had to be torn apart to retrieve their bodies.”
And while the system could deliver a lot of water, the two pumping stations only stored about 17,000 gal. Even with a small number of customers (around 64 homes, according to Brown), that amount of water could be used up in 20 minutes. “They didn’t have the data to know how much water Philadelphians would use,” says Brown. Additionally, Philadelphians also used the water to wash the streets and sometimes left the hydrants running, which further depleted capacity.
By the fall of 1811, the city had invested more than $500,000 in its water system, and it was money down the drain. “The annual income of the works was $12,163, while the expenditures were $29,702,” Smith wrote. “All of the 28-plus miles of wooden pipe that had been laid would have to be replaced, and there were only 2,127 paying customers in this city of well over 50,000 people.”
Years before the city was ready to fund a newer, better system, Latrobe had already moved on (he later worked on the design of the U.S. Capitol). John Davis had taken over as head of the city’s water system; when he left in 1805, Frederick Graff stepped in. Latrobe had hired Graff as a draftsman in the 1790s; he later served as an assistant engineer on the Centre Square project. Many of the lasting elements of the city’s municipal water system came under Graff’s leadership, and he remained chief engineer of the water system until his death in 1847.
Graff turned to a site just a mile from Centre Square — a significant hill with a flat top, Faire Mount (later condensed to Fairmount, its current name). Towering 100 ft over the Schuylkill, the site had space for a reservoir of at least 1 million gal., says Brown.
As before, Graff’s plan called for two steam engines — but this time they were designed to run independently of each other. Additionally, the engines featured different designs. One used a Boulton and Watt steam engine, the same kind as the earlier waterworks. But the second, designed by inventor Oliver Evans, was unique. It was, according to Gibson, the “largest non-condensing high-pressure Columbian steam engine built up to that time” and produced about 100 hp. It also represented, Brown says, a “tremendous leap in technology.” While the Boulton and Watt engine operated at 2 psi in its main cylinder, Evans’ engine operated at 200 psi.
But the steam engines again proved to be problematic, and Graff and city leaders knew it was time to move on from what Gibson described as a “noble experiment.” For one, they remained dangerous: two explosions at Fairmount, in 1818 and 1821, killed several people, according to Levine. And cost and efficiency remained overriding issues. The steam engines used wood-fired boilers, and as trees around the growing city became scarcer, the cost of firewood grew. Brown says the practice of burning coal in a boiler came into use around this time, but by then plans for a water-powered system were well underway. By 1822, the steam engines had been replaced by three waterwheels.
According to Levine’s history, the massive waterworks at Fairmount featured “an earth-filled masonry wall (that) doubled as the wall of the lock at the entry to the canal on the western side of the river,” a 1,200 ft long spillway built across the river to mitigate floods, and the “hollow masonry structure of the millhouse itself, 238 feet long, which housed the waterwheels.” Behind this was a 419 ft long millrace, which brought water to the millhouse and the three waterwheels. The Fairmount Waterworks greatly increased the city’s water capacity; its reservoir contained more than 3 million gal. in 1815 and added another 0.5 million gal. capacity a few years later.
According to an 1824 report Smith cited, the 16 ft diameter waterwheels proved far more efficient than the steam engines: “While previously it had cost $206 to raise 3,375,000 gal. of water a day, this could now be done for $4, a saving of over 98 percent. By adding more waterwheels, the capacity could be increased to 10 million gal. per day, and daily operating costs would be only $10, which was more than $500 less than it would cost to pump the equivalent volume with steam engines.”
Meanwhile, the initial wooden pipe system had grown to more than 32 mi by 1817, and “leakage had been somewhat lessened from the earlier experience by using short sections of iron pipe called ‘connectors’ to join the logs,” wrote Gibson, “but it was soon evident that it was not possible to utilize the new system to its fullest unless better distribution could be made.”
Additionally, frequent right angles in the pipes produced friction, which reduced water pressure as water traveled farther through the system. In 1818, the committee authorized the use of larger iron pipes to replace the old wooden mains. “Over the next three decades new cast iron pipes were laid and old wooden ones replaced until, by 1858, the last wooden pipes were taken out of service,” Levine wrote. “The cast iron pipes had less friction, leaked less readily, and lasted far longer than the wooden pipes, in many cases a hundred years or more.” Indeed, cast-iron pipelines from 1824 were still in use in 1947 when the city became an early member of the Ductile Iron Pipe Research Association’s Cast Iron Pipe Century Club. (The Centre Square Water Works was demolished in 1829, and the Schuylkill Water Works was demolished in 1838.)
Like Latrobe’s original plans, Fairmount was more than a utilitarian piece of hydraulic engineering. According to Smith, “The project also included the building of the first few of what would be a group of elegant structures along the river, to enclose the machinery, afford workspace for the system’s managers and employees, and provide shaded viewpoints for visitors.” These structures were built in the neoclassical style and became significant tourist attractions in their own right.
When the steam engines were finally removed from Fairmount, the engine house that enclosed them fell into disrepair but was refurbished in 1833 and turned into a restaurant. Graff, in what Brown describes as his “first foray into architecture,” had designed and built the engine house in a “Georgian or Federal style whose exterior resembled the nearby summer mansions of wealthy Philadelphians, mansions that were arrayed along the high ground above the east bank of the Schuylkill River, upstream of Fairmount Water Works.”
According to Brown, many of the neoclassical elements that have made the Fairmount Waterworks a tourist attraction to this day were added in stages. The Waterworks’ final ornamental structure, a 20 ft tall pavilion with a pediment, wasn’t added until 1872.
Fairmount remained in service for years, growing with the city. The reservoir impounded 10 million gal. by the 1840s, and by 1843, Fairmount’s millhouse used eight waterwheels. According to Gibson, the building, open to the public, featured a gallery that ran the length of the building, allowing visitors to view all the wheels at once. The Fairmount Waterworks operated continuously from 1815 to 1909, and since then it has served as an aquarium, children’s theater, and, currently, an environmental education center. The Philadelphia Municipal Water Supply was dedicated by ASCE as a Historic Civil Engineering Landmark in 1975.
For Brown, the legacy of Latrobe and Graff was in their innovative melding of architecture and engineering. “These engineers went out of their way to learn some architecture and then apply that to these structures,” says Brown. “We don’t do that today.”
This article first appeared in the January/February 2021 issue of Civil Engineering as “First of Its Kind: The Philadelphia Municipal Water Supply.”
