Washington’s Sewage Apocalypse

On January 19, a six-foot-wide sewer pipe broke beneath the land alongside the Potomac River, nine miles northwest of the Lincoln Memorial. A landslide of dirt and rocks dammed the flow, and the products of a million toilets, showers, sinks, and washing machines in the Washington, D.C., suburbs shot up to the surface and gushed into the watershed. According to researchers at the University of Maryland, it was one of the worst raw-sewage spills in U.S. history.

Two months later, there is no official explanation of what went wrong, and the gist of the DC Water report released on March 5 is, roughly, We did everything right. “This was an unprecedented event,” the CEO, David Gadis, said in a press release. “After evaluating our inspection reports and ratings we do not believe there was any reason to change the timing for our planned rehabilitation, which was to start this summer.”

But something did go wrong. In addition to the damage to the watershed, where the utility has elsewhere spent billions of dollars to keep out sewage, Maryland’s C&O Canal spent nearly two months functioning as a jury-rigged open-air sewer to bypass the collapse and has been coated in human muck several inches deep. If the collapse had occurred a few miles upstream, it would have contaminated the intake of the Washington Aqueduct, cutting off the water supply for about 1 million people.

Two investigations into the cause of the collapse are ongoing. But DC Water officials have begun to hint at a possible culprit: a design flaw that has been lurking above the underground pipe since it was installed in the early 1960s, in the form of large boulders that were used as fill when the pipe was buried. This weakness raises the possibility of other land mines along the pipeline’s 54-mile path. The utility is now turning to archival drawings to identify spots with similar methods of construction, and contemplating drilling to assess the risk. Everyone who was involved in designing the interstate sewer is long gone.

[Read: America is doubling down on sewer surveillance]

Searching through old blueprints to discover where the boulders are buried is an extreme example of an institutional memory lapse. The trend is common in a country whose infrastructural progress and prowess peaked decades ago: We are in command of many aging systems whose original design choices are lost to time. Laying boulders over the trench was an error—not some quirky old way of doing things. But the phenomenon plagues the maintenance of our great public works, whether they’re subways or computer systems. This fragility of highly specialized knowledge is sometimes characterized by the “bus factor”—as in, how many personnel would have to be hit by a bus before we couldn’t make this thing work anymore? The Potomac interceptor is so old that that joke could have been about a trolley.

Sewer pipes in America break constantly, but the biggest failures—of the so-called interceptor sewers that drain entire regions, the highways of the underground—are less common. There are three reasons a major piece of infrastructure can fail: Somebody broke it, somebody failed to fix it, or somebody built it wrong. After an interceptor pipe collapsed in the suburbs of Detroit in 2016, opening up a house-eating sinkhole the size of a football field, a subsequent investigation found that contractors working in the sewer had unleashed a “water hammer”—a pressure buildup that burst the pipe. After a four-foot-wide Los Angeles County trunk sewer collapsed in 2021, an audit concluded that sewer corrosion was to blame—the district simply hadn’t caught the issue in time.

DC Water executives have implied that the problem was a construction defect that defied one of the earliest precepts of sewer design: Fill in the hole with soft, tamped dirt, not big rocks, with the care you would give to shoveling dirt onto a casket. The officials clearly do not want to preempt an independent investigation, which will draw its conclusions in the coming weeks. But they have made no secret of what they think of the boulder pile that sat above the interceptor pipe.

“That was probably not the right thing to do,” Gadis said at a public meeting in Bethesda, Maryland, last month. “The standard calls for rock no larger than three inches to be placed back on there as cover. The rock was big; it was huge. And we still have that rock out at the site as we speak right now. So that is what we’re thinking that part of the cause was.” At a press conference earlier this month, he added that pending the result of the investigation, “we will be able to lock in and say, ‘This is the reason why.’” The DC Water executive John Cassidy noted that those big boulders might have imposed “point loads” on a corroding pipe, concentrating pressure that “exacerbated the risk of collapse.”

If that’s true, then the failure of the Potomac interceptor is not the same old story about aging infrastructure that nobody wants to pay to maintain. In fact, DC Water had allocated hundreds of millions of dollars of repairs for work that would have, among other things, shored up the failed section later this year in response to run-of-the-mill corrosion. Instead, the collapse shows how little we sometimes know about society’s aging vital systems; a contractor’s mistake that might have drawn special attention from maintenance officials over the years turned into a lacuna no one knew about.   

Now DC Water is scrambling to figure out which other sections of pipe might be perched in a nest of heavy, pointy boulders.

This is not an easy question to answer. Picture yourself in an interceptor sewer. The reinforced-concrete pipe is the height of a person, and the stream along its lower half ebbs and flows with the tides of human activity: up in the morning (showers, toilets), down during the workday, up in the evening (laundry, dishwashers), down overnight. “It’s like human arteries,” Michael Chee, a spokesperson for the L.A. County Sanitary Districts, told me. “Just like a 25-year-old man who jogs every day can drop dead of a heart attack, the same thing can happen to a sewer.”

To try to head off that possibility, water-and-sewer districts are supposed to perform regular inspections. Because pipes are narrow, underground, and flowing with several feet of rushing wastewater, these are usually conducted by crawling sewer “tanks,” or boats mounted with video cameras. They measure the buildup of solids on the bottom of the pipe with sonar and try to estimate the corrosion on the top, where hydrogen sulfide gas (a by-product of sewage recognizable by its rotten-egg smell) is eating away at the concrete.

Assigning grades, which run from one (a new pipe) to five (requires immediate attention) in two categories (operation and structural), is more art than science. A Potomac-interceptor inspection from October 2024 by the firm RedZone Robotics, for example, reported two holes with the highest structural-risk rating within sight of the break—a finding that suggests, if not a cause of the break, then a high state of disrepair in need of immediate attention. DC Water said in its March 5 report that these ratings were “incorrect,” and that what RedZone had perceived as dirt was in fact dark-colored concrete, implying that the pipe had not been visibly at immediate risk of failure.

Similarly, a June 2024 inspection used lasers to measure corrosion in the pipe wall and found two segments not far from the break with nearly six inches of corrosion—in a pipe whose walls might be six to eight inches thick, total. That inspection was “rejected due to poor quality.” The October inspection did not measure that segment, but elsewhere found no corrosion more than four inches deep. Those conflicting interpretations are a regular occurence. RedZone’s CEO, David Petrosky, told me his team tries to identify problems but leaves the final say to the system owner, which knows the pipe best.

In my conversations with sewer engineers, I encountered widespread reticence to comment on the Potomac-interceptor collapse, as if I had asked them to comment on someone else’s failed marriage. This reflects equal parts restraint and calculation. It seems uncouth to speculate. But also, contracts with giant agencies are what pay the bills, and no one wants to develop a reputation for saying bad things about utility management.

More than one sewer expert, however, was not impressed with the quality of the pipe. “If those inspections were uploaded to our platform, they would have been blinking red in terms of risk, as measured by likelihood to fail and consequence of failure,” Billy Gilmartin, a co-founder of Sewer AI, an inspection company, told me.

Which gets to another point of wastewater augury: The need for maintenance depends not just on the character of the corrosion but on the peril of inaction. In Los Angeles, for example, the county sewer district maintains a risk matrix that assesses both the condition of the pipe and what might happen if it gives way.

[Marc J. Dunkelman: American infrastructure is about to get even worse]

In Washington, only after excavating the collapse did DC Water learn that the pipe had been surrounded by boulders as large as three feet across that fell through its corroded vault. Instead of an hourglass stream of sand, dirt, and water that could be swept along by the flow, the Potomac interceptor faced a dam of rocks that blocked the tube completely, forcing hundreds of millions of gallons of sewage up to ground level, over the frozen ground, and into the river. Being near the Potomac River is by design—the big sewer pipe needs to be lower than the smaller sewer pipes, and the riverbank is the lowest ground around—but the boulders, Gadis told the D.C. city council this month, were one reason the collapse was “catastrophic.” The combination of this design, the pipe’s sensitive location, and its heavy use all made it an especially risky piece of infrastructure.

As a squadron of pumps lifted sewage up and over the bottleneck, DC Water dispatched crews for emergency inspections of 22 high-priority sites—places where changes in the pipe’s elevation, little sewage waterfalls, would produce higher concentrations of concrete-eating gas. “None of the data that has come back has presented us with information that we believe is an emergency,” DC Water COO Matthew Brown said in a public meeting last month.

But that finding is little comfort if you believe DC Water’s own assessment of the accident, in which the inspection reports did not warn of the danger. Indeed, not everyone is impressed with DC Water’s performance, shares its sense of an exemplary performance under pressure, or buys its explanation for what went wrong. Candice Miller, the Macomb County Public Works commissioner, who supervised the recovery from the 2016 Detroit-area pipe collapse, told me a pipe that shallow should not take so long to bypass. “Sounds to me like it should have been fixed quickly and it wasn’t,” she said. The Potomac repair was finally completed this week.

As for those boulders? Neil Grigg, a professor of civil and environmental engineering at Colorado State University, told me that the big rocks could possibly explain the collapse and the blocked tunnel. But, he said, “effective utility management detects potential problems like that ahead of time and fixes them before you have a failure.” What DC Water saw inside its pipes was only part of what it needed to know.