FoundSF

Historical Essay

by David D. Schmidt, 2026

An excerpt from the book, San Francisco Bay Area: An Environmental History, by David D. Schmidt (Backcountry Press, 2025).

This historic brick culvert, built in the 1860s at Kirby Cove in the Marin Headlands, is similar in size and shape to the brick sewers built in San Francisco in the 1860s to 1890s.

Photo: David D. Schmidt, 2025.

Beneath the Streets of San Francisco: Old Brick Sewers

In the pre-dawn hours of December 11, 1995, high winds and rain lashed San Francisco, and a century-old brick sewer backed up with water and burst, spewing 94 million gallons of street runoff and untreated sewage onto a steep sandy slope in the city's ritzy Sea Cliff neighborhood, a mile southwest of the Golden Gate Bridge. The dirty water flowed downhill into Lobos Creek—the drinking water source for the Presidio, a former Army base that transitioned to national park in 1994. As water rushed from the ruptured sewer, it washed away the sand beneath Howard and Iran Billman's three-story home, which collapsed like a house of cards. By this time, it was daylight, and TV crews were on the scene, filming the spectacle for nationwide news networks.

In 2000, 90 of San Francisco's 900 miles of sewers were still made of brick, which had been recognized as problematic more than a century earlier. The sewers of the late 1800s were oval-shaped brick tunnels five feet high and three feet wide—just large enough for a worker to walk through them, which was occasionally necessary to remove blockages. One hundred and thirty-five miles of the city’s sewer system, or 15% of it, had been built before 1906; some of it dated as far back as 1850. The 1995 incident put city officials on notice that they needed to replace the brick sewers.

In the 1850s, the earliest San Francisco sewers collected only street runoff, because there was no piped water, and no flush toilets. By the 1870s, water pipes and toilets were standard features of new homes and hotels. When the gigantic, opulent Palace Hotel opened on Market Street in 1875, it had 755 flush toilets—one for every room or suite.

Before 1880, toilets were not always connected to San Francisco’s sewers. Many discharged their dirty water into privy vaults—covered brick tanks in backyards. In some areas that lacked sewers, such as along Precita Creek (today’s Cesar Chavez Street in the Mission District) in the early 1870s, toilets discharged directly into creeks.

Bay waters at the east end of Folsom Street in San Francisco, c. 1900, are clouded with sewage and sediment. Vertical structures in the background are coal hoppers, use dto store coal unloaded from ships.

Photo: BANC PIC 1905.17500.6:006--ALB, Roy D. Graves Pictorial Collection, Bancroft Library, University of California, Berkeley.

These disposal methods created offensive overflows and odors and transformed Precita and Islais Creeks into open sewers. By the late 1860s, a city ordinance required privy vaults to discharge into the sewers. This didn’t solve the problems—it just moved them downstream. As San Francisco Public Health Officer Dr. I. Rowell explained in 1869,

The method of disposing of filth in this city is the most wretched of any city in the civilized world. . . . [sewage] is sent on its sluggish course through the sewers to the bay, discharging through the open vents at every street corner a volume of disease-bearing effluvia that would cause any animal but man to rush in wild disorder from the spot, and avoid it thereafter as the burned child avoids fire. . . .

The “disease-bearing effluvia”—sewer gas—is a mixture of hydrogen sulfide and methane generated by bacteria consuming organic matter in the sewage. Both gases can be fatal when inhaled in an enclosed area—like a sewer. Rowell and other public health experts of the late 1800s, however, erroneously believed that the gas was spreading disease. Their remedies for the sewer system focused on preventing sewer gas from getting into homes and buildings.

Fighting the Dreaded Sewer Gas

To accomplish this, they recommended "sink traps" and "toilet traps"—those now-familiar bends in the drainage pipes which are always filled with water. By the mid-1870s street corner catch basins, beneath the heavy metal grates on every streetcorner, were designed with water-filled traps for the same reason.

The traps didn’t always work, because the sewers generated gases that pushed up through sinks and toilets into homes and buildings. Eventually someone thought to make small holes in the cast-iron manhole covers on the streets, to equalize the air pressure inside and outside the sewers—a low-tech solution that still works today.

To prevent the formation of sewer gas, public health experts argued that sewers should transport wastewater to the bay quickly, to minimize decomposition (which generates the gases) during its journey through the sewers. This was nearly impossible in the flatlands of San Francisco and Oakland, where sewage stagnated. Plus, wherever waterfront outfalls were below the high tide mark, high tides held back or even reversed the flow of sewage. When sewage finally did reach the bay, the stench was, according to an 1876 report by William P. Humphreys, San Francisco's official city surveyor, ". . . offensive to the last degree of endurance."

This was one reason why the city’s well-to-do lived on Nob Hill, while the working class lived in the flatlands. Some public health advocates proposed building a new, separate, narrow-piped drainage system for sewage only (not storm runoff). The small pipes, they contended, would be cheaper than sewers and would move the sewage faster to the bay. Others urged building water tanks atop the city's hills, to periodically flush the sewers. These ideas never caught on because the city government was barely able to construct sewers in new neighborhoods fast enough, much less build an entirely new sewer system.

In 1868, San Francisco had only 38 miles of sewers, but more were being added constantly. By 1880 the city had 128 miles, and by 1893, 227 miles. These sewers conveyed storm water and sewage, but they were often clogged with sand, gravel, and according to Humphreys,

“. . . substances which they were never intended to receive; such as street dirt [horse manure], garbage . . . ashes, shavings, sticks, brickbats, coal, bones, bottles, rotten fruit and vegetables, old clothes, boots, shoes and stockings, broken crockery, etc. In some parts of the city dead dogs, cats, and rats have been found in these [streetcorner] catch-basins. The facility with which their covers can be removed and replaced, particularly at night, invite to this easy way of disposing of all kinds of garbage.”

San Francisco’s Sewer Woes

In outlying areas, sewer construction lagged years behind housing. According to an 1892 lecture by physician and public health advocate Dr. I. H. Stallard, new houses dumping their wastewater directly into Precita Creek

". . . grew to be an intolerable and dangerous nuisance . . . About 1875 an Act of the Legislature was obtained to abate the evil at the cost of the whole community."

This "evil" was abated by building a mile-long, oversized brick sewer 11' 6" wide and 8' 9" high in the creekbed, and building Army Street (now Cesar Chavez Street) on top of it. The sewer emptied into a tidal slough, Islais Creek. But rainy season runoff from unpaved streets and construction sites caused soil erosion, dumping sediment into this giant sewer. According to Stallard,

Immense quantities of sand and gravel are brought down. The scour has washed out the cement between the bricks, and the floor of the sewer is like a cobble-stone pavement, providing a lodgement for filth . . . Meantime the fine sand is washed into [Islais] creek and the coarse gravel left behind. Every year it has to be removed by hand labor. Last year it took five gangs of men six weeks to take it out . . .

During the dry season,

"The flow of sewage . . . rarely runs more than three or four inches deep. . . . deltas of filth collect at the entrances of the lateral sewers . . . stench and dangerous gases are given off . . . One effect, however, is to purify the sewage in a small degree . . . The sewage flow . . . is diverted by the accumulation of islands of deposits . . . into a hundred channels . . . Thus we have formed a sort of subterranean sewage farm; . . . and when the sewage arrives at the outlet near Islais Creek . . . it has become bright and sparkling; it has lost all its odor; [according to] the gentleman who has his tannery at the outlet: 'In the early days I have drank much worse looking water, and if I did not know where it came from I would not hesitate to drink this now. . . .'"

Stallard had unwittingly described sewage treatment by bacteria—the basis for sewage treatment technology that would be developed in the early 1900s.

The stagnating sewers caused more problems: When clogged with sand, soil, and debris, they overflowed and flooded streets and basements. A city ordinance required sewers to be placed 10 feet below the surface, to ensure that new sewers could be connected with existing ones. But this meant that on flat land, especially former bay waters that had been filled in, the sewage would stagnate.

The Channel Street [China Basin] sewer, which drained the Mission District, was a case in point—it needed constant cleaning after it was built in 1872. According to Stallard,

". . Nearly all the lateral sewers are . . . choked by sand brought down from the immense drainage area. The sand is black and stinking. . . . In September last [it was] four feet nine inches [deep], occupying considerably more than half the capacity of the sewer. It contained about 18,000 cubic yards of filth . . . At first five men were employed [to clean it out], then for two weeks, 25; . . . Sand enough has been taken from this sewer to fill it three times over.”

It was a dirty, dangerous job for the men and boys hired to do it. Stallard cited an 1885 sewer study, which found that in filled-in flat areas,

" . . . [A]s they approach the bay the sewers are either level or run up hill [!]. All the sewers below [east of] Montgomery Street are tidal. . . . they become elongated cess pools . . . Their size and shape is not uniform. Sometimes there is a 16-inch pipe made to take the sewage of a five-foot sewer. . . . In many there are rotten bricks, sandy mortar, and in one place the bricks were found replaced by empty barrels. . . ."

In 1899, San Francisco voters approved $4.6 million in bonds (equal to more than $100 million today) to improve the sewer system, but legal complications invalidated the bonds, so a second vote was needed, and it was delayed by the 1906 Earthquake until May 1908.

This time, voters approved the bonds by an overwhelming 15 to 1 margin. Mission Bay had already been filled in, and its remaining channel, China Basin, was grossly polluted. According to City Engineer Carl Grunsky’s May 19, 1909 report,

The discharge from these sewers [emptying into China Basin] represents more than half of the sewage and rain-water . . . of the city. During half the year there is no rain to dilute the sewage, and it ebbs and flows . . . creating a filthy nuisance beggaring description.”

Grunsky argued that “Dilution in the waters of the bay and ocean [is] without doubt for San Francisco the natural and proper method of sewage disposal." The city’s sewer planners assumed that the tides would sweep the sewage out to sea twice a day—Nature’s own flushing system.

San Francisco continued building sewers in the southeast quadrant of the city in the 1920s and 1930s, routing that sewage to a new outfall at the Islais Creek Channel—another remnant of a bay inlet that had been filled. The channel soon gained the unflattering moniker, “S**t Creek.” South of the city, the dumping of raw sewage contaminated and forced closure of the bay’s last major oyster growing operation on tidal mudflats that would later be filled and paved to make runways at San Francisco International Airport.

Sewage Farms in San Francisco

In the late 1800s, even San Francisco had farms irrigated by sewage. In the Marina District, farmers used wastewater flowing downhill from the mansions on Pacific Heights. According to an 1893 city engineer's report,

"The sewage of the Steiner Street sewer, and to some extent that of the Pierce Street, is being utilized during the spring and summer months for the irrigation of Chinese vegetable gardens . . . north of Chestnut Street."

John McLaren, San Francisco's legendary parks director from 1893 to 1943, was always looking for ways to get more soil and compost to transform the western half of Golden Gate Park from sand dunes to lush greenery. He had the city's street sweepings, mainly horse manure before 1920, delivered to the park daily. Asked what he wanted for his birthday, he famously replied, "Twenty tons of good manure."

As the city rebuilt from the 1906 Earthquake and Fire, McLaren arranged for the raw sewage from the city's western half, then only partly developed, to empty into two open irrigation ditches in Golden Gate Park. The public outcry over seeing (and smelling) sewage in the park was so intense, however, that McLaren was forced to build a septic tank in 1912 to remove the solid material before using the sewage. Continued public pressure, and the discovery of well water sources in the park, prompted McLaren to stop watering the park with sewage in 1916.

Golden Gate Park’s McQueen Plant Pioneers Secondary Treatment, Re-use

In 1926, San Francisco’s Board of Parks Commissioners found a new method of treating sewage so it could again be used for irrigation in Golden Gate Park, but without offensive odors. They hired Henry E. Elrod, a Houston engineer who held patents on several sewage treatment technologies, to design a treatment plant using the “activated sludge process.”

This technique forces jets of air into concrete pools of primary-treated sewage to speed the growth of bacteria that break down the organic pollutants. Some of the resulting sludge—alive with bacteria, and thus “activated”—is constantly routed back to the aeration pools to help the bacteria multiply quickly. This mimics the natural process of bacteria breaking down the waste, but speeds it up from 90 days to as little as 90 minutes. This method, developed in Chicago by 1915, is now used for secondary sewage treatment in urban areas nationwide, thanks to the federal Clean Water Act of 1972.

Back in 1932, park boss John McLaren hired the McQueen Sewage Disposal Company, owned by Frank McQueen, to build the facility Elrod designed in the center of Golden Gate Park. The plant pumped up to a million gallons per day of treated wastewater into Elk Glen Lake,(1) where it acquired a green tint from microscopic algae. From there, it flowed into nearby Mallard Lake and continued westward through a ditch to the park’s Chain of Lakes. Even if you jumped over the ditch—as the author did dozens of times as a high school cross country runner in the 1970s—no sewage smell was noticeable.

McQueen Sewage Treatment Plant in Golden Gate Park, 1935.

Photo: Golden Gate Park Sewer Archives

The McQueen facility provided water to irrigate about 800 of the park’s 1,017 acres. During California’s 1976-1977 drought, it was recognized as the first of its kind, providing treated wastewater for reuse and helping conserve the city's drinking water. Its long-time operator, George Mallick, loved to give tours of the place. He predicted that it was the wave of the future in water-short California. But the aging plant was unable to meet stringent state water quality standards that took effect in the late 1970s and closed in 1982.

Mallick was vindicated in 2018, when the city started building a recycled water facility at the Oceanside Water Pollution Control Plant, adjacent to the San Francisco Zoo. When construction was complete in 2023, the project started piping up to 4 million gallons per day of recycled water to a new pumping plant in Golden Gate Park, just a few steps from the old McQueen site. The water, treated to meet state standards using membrane filtration, reverse osmosis and ultraviolet disinfection, will irrigate the park as well as golf courses at the Presidio and Lincoln Park.(2)

San Francisco Finally Cleans Up

In 1971 and 1974, San Francisco adopted plans to comply with the state’s 1969 Porter-Cologne Act and the 1972 federal Clean Water Act by building two major secondary treatment facilities. Over the next 20 years, the city spent $1.6 billion to upgrade its entire system to secondary treatment.

In the late 1970s, the city expanded the Southeast Treatment Plant to increase its capacity and provide secondary treatment. The new facilities started operating in July 1982, but that was only half the battle. With San Francisco’s combined wastewater and stormwater sewer system, every time it rained, the volume of water going to the treatment plants suddenly increased tenfold, or more—too much for the treatment plants to handle, so most of the untreated sewage/stormwater mixture went straight into the bay or ocean.

Next, the city routed wastewater from the North Point primary treatment plant near Fisherman's Wharf to the Southeast Plant for secondary treatment. The North Point facility was still needed during the rainy season, when rain events overwhelmed the capacity of both plants. To solve this problem, the stormwater/sewage surge had to be contained, and fed to the treatment plants gradually after a rainstorm. For this, the city built hidden holding reservoirs for wet weather flows, 40 feet deep beneath the surface of Marina Boulevard and the Embarcadero.

With San Francisco's bayside pollution finally under control—more than 70 years after City Engineer Carl Grunsky said all of the city’s raw sewage should be dumped in the bay—the city’s Department of Public Works (DPW) turned its attention to the ocean side. In 1981-1982, the DPW built a mile-long concrete sewage reservoir beneath the Great Highway at Ocean Beach, to store the stormwater/sewage flows after each rainfall, until they could be treated at the aging Richmond-Sunset [primary] Treatment Plant in Golden Gate Park.

San Francisco's Oceanside Water Pollution Control Plant, completed in 1993, enabled the city to shut down and demolish the 1938 primary sewage treatment facility at the southwest corner of Golden Gate Park.

Photo: David D. Schmidt, 2003

This reduced rainy season beach pollution while the city built the new Oceanside [secondary] Treatment Plant adjacent to the San Francisco Zoo. The new facility, partially underground and hidden from public view, opened in 1993, capping 20 years of planning and construction. The Oceanside plant can treat 2.8 cubic meters of sewage per second—up to 65 MGD. It removes up to 95% of the pollutants in the wastewater before discharging it in deep water 4.5 miles offshore, keeping the beaches clean. The old Richmond-Sunset Plant was demolished in 1995, bringing an unlamented end to more than 60 years of sewage treatment in Golden Gate Park.

For the first time, the city had secondary treatment for all its sewage during the dry season and 2/3 of it during wet weather, when the wastewater was 94% street runoff and only 6% sewage. Only 10% of the stormwater/sewage surges were discharged without any treatment, flowing into the ocean and bay at the city's 36 "overflow structures" (outfalls). Two of them are highly visible: the concrete piers on Ocean Beach at Vicente Street and at Lincoln Way.

Questions? Email the author: davidnaturesf@gmail.com

Notes

1. Pinhey, Nicholas, California Water Environment Association, “Forgotten Facilities: Golden Gate Park’s Recycled Water Plant,” accessed in 2021.

2. San Francisco Public Utilities Commission, “Westside Enhanced Water Recycling Project,” accessed June 2023.

---

Excerpted from David D. Schmidt's San Francisco Bay Area: An Environmental History. Available from Backcountry Press.