It was raining hard. The inside of our new house sounded like the percussion section of a high-school band. Thunder, lightning: We got the whole show, a classic of late-summer Washington, D.C. The storm ended abruptly, as though it suddenly got bored. Outside, the sidewalks were steaming, and you could smell electricity in the air.

Our recycling bin and Supercan lay on their sides at the bottom of the driveway, where it slopes down into the garage over a grated trench drain, and there was a big horizontal line of wet dirt two feet up on the garage door. Water—a lot of it—had come and gone and left its mark.

Worried now, we ran to the basement, where our worst fears were confirmed. Like the trench drain, the bathroom shower had experienced a Linda Blair moment, vomiting dirty water all over the tile floor and out the door into the carpeted playroom. The wet filth was full of debris, like the stuff that runs in gutters. I felt invaded. There is something wrong with this house; I knew it all along. We never should have bought it.

“Calm down,” said my husband. “It was an unusually heavy rainstorm, and we’ll plug up the shower drain so nothing more can come out.” Right.

A couple of weeks later, while we were away on vacation, another gargantuan rainstorm swept through D.C. With the shower drain blocked, the water found a different outlet: the toilet. This time the destruction was much worse. The basement carpet was heavily soaked with water that had again come from some unknown and unsavory source. Leaves, dirt, and small sticks were strewn around as though borne in on the tide. Smelly, disintegrating cardboard boxes full of our children’s baby photos, winter woolens, books—everything still packed from the move—sat in the wet. Mold was already growing in the cupboards and on the inside of framed pictures.

I called the insurance company and a cleanup company. Wearing face masks and gloves to protect them from God-knew-what, the crew hauled out the carpet and the rotting stuff and set up huge electric fans in the basement to dry out the rooms. By now, I knew that the house was possessed by an unseen foe that I had to figure out how to conquer: the plumbing.

I had never thought much about what happened after the toilet flushed. The contents simply went away. But that was before I moved to Cleveland Park—and before my toilet started flushing up, rather than down, during rainstorms. And before I became obsessed with Washington’s sewer system.

Those summer storms launched me on a sleuthing mission that started at my toilet and ended at the Chesapeake Bay. It uncovered a long story, and my basement situation was just the opening line. The plot quickly expanded to include the neighborhood’s and then the city’s sewer pipes, all the way to the Blue Plains Sewage Treatment Plant, and beyond.

It turned out that our basement toilet—which is no longer in use—was plumbed into the storm sewer, which drained, eventually, into Rock Creek. As awful as that sounds, it was merely a drop in the billions of gallons of combined raw sewage and storm water that flow every year into Washington’s principal waterways: the Anacostia River, the Potomac River, and Rock Creek. The Anacostia, which is tidal and sluggish and therefore the least capable of processing the bacterial onslaught, is routinely so contaminated after a rainstorm that it is unsafe to swim or even wade in, and it has the highest rate of fish cancers of any river in the United States. And the Chesapeake Bay, which received a D on its most recent annual report card from the environmental-watchdog Chesapeake Bay Foundation, is the ultimate downstream recipient of the excess nutrients and toxic chemicals.

Karen DeWitt, director of communications for the D.C. Water and Sewer Authority (WASA), points out that an ability to properly dispose of waste is “one of the marks of civilization.” With its outdated sewer system and zero support from the Bush White House for desperately needed improvements, Washington has a long way to go to get civilized.

When my husband and I bought a house in Cleveland Park, it was a move up, literally, from Georgetown. Just north of the National Cathedral, our new neighborhood was sited on some of the highest ground in the city—300 feet above the White House, 360 feet above sea level. Known for its stately houses fronted by broad, inviting porches, most of Cleveland Park was developed around the turn of the century and marketed on the cachet of fresh air and healthy living. A 1904 sales brochure extolled the neighborhood’s proximity to the Tenleytown Road (now Wisconsin Avenue), which ran along a ridge, and cited the Weather Bureau’s declaration that the neighborhood was “one of the most healthful regions in the district.” In the summertime, the copy oozed, “the Park is a cool and pleasant resort. The breeze from the hills makes life one grand sweet song, and the music of the birds stirs the soul.”

Such a pastoral vision would have been compelling to the denizens of the smelly, disease-laden, steaming miasma that was 19th-century Washington. Hundreds of acres of the city were covered with marshes that bred malarial mosquitoes. Garbage was simply thrown out into the street. As of 1870, the sewer “system” comprised two open ditches, Tiber Creek and the Washington Canal, which drained everything in the city. The ensuing 30 years saw some improvements, including the filling of the Washington Canal because it was a “nuisance,” according to the 1932 book Planning and Building the City of Washington, published by the Washington Society of Engineers. (The word “nuisance” is used throughout Chapter 8, “Drainage and Sewerage,” as a sort of droll engineer’s euphemism for what must have been unspeakably disgusting.)

A somewhat ad hoc series of arches and culverts also evolved. Though intended to channel storm water off the streets, this network, for lack of an alternative, also served to usher sewage into city streams that ran to either the Potomac or the Anacostia. With their offensive contents, these bits of engineering were also “eventually looked upon as creating a nuisance.” Such a nuisance, in fact, that in 1890 President Benjamin Harrison charged a special, handpicked board with figuring out how to “remove sewage from neighborhoods and dispose of it unobjectionably.” No wonder it was appealing to live on the outskirts of the city at 360 feet above sea level. The 1904 Cleveland Park sales brochure might as well have said, “Hey, folks, it’s all downhill from here!”

Like the intricate workings of unhappy families, each case of dysfunctional plumbing is unique. A main source of confusion at my house was the sticks and leaves in the water that came out the basement toilet. How had these elements entered our home’s sanitary line, which was supposed to carry waste (whatever goes down the drain of the toilet, sinks, showers, washing machine, and dishwasher) out to the sanitary sewer pipe in the street? Was the pipe perforated somewhere? Or did the sanitary pipe merge at some point with the storm-water pipe (which is supposed to carry water away from the gutters and driveway)? If they did merge, then when it rained hard and the storm drain on the corner got full, water full of street debris might flow backward up the line and come out the toilet. It was a plausible theory, with its basis in the knowledge that much—though not all—of D.C. relies on a technology called combined sewers, dating from the 19th century.

“Combined” means that a single pipe carries both storm water (including anything washing off the street, such as rain, leaking auto fuel, lost Frisbees, and dog excrement) and liquid sewage (including all waste produced by households, restaurants, hotels, and other businesses) to wherever they are going. Depending on whether the pipe functions as a lateral, branch, or main line, it might be anywhere from 10 inches to 20 feet in diameter. (Twenty feet is big, by the way. The guys at Blue Plains can tell you about the motorcycle that fell into an open catch basin somewhere in the system and came shooting through the main all the way to the plant, breaking a fair number of iron filter bars when it slammed against them at the end of the line.) Depending upon the era in which it was installed, the pipe might be made of wood, brick, stone, clay, plastic, or metal.

The ideal sewer system—a separated system—would channel 100 percent of sewage through one set of pipes to a modern, state-of-the art treatment plant at the same time that storm water—generally cleaner—flowed through a separate set of pipes to the rivers. On the infrastructure front, however, old cities rarely have an ideal prototype of anything, and places such as Washington and Boston evolved as combined systems, with newer, separated systems in only the more recently built areas.

Back at my house, all I knew was that, as they used to say of female trouble, something down there wasn’t working right. But untangling a plumbing problem in an old system is no easy task. It requires more than just working knowledge of plumbing practices over the past 200 years; it also calls for an active imagination. You have to find someone willing to enter where few dare, someone who presses onward when the way is not clear, someone who thinks like a stream of flowing liquid in the thrall of gravity. This how I came to meet the Merlin of wastewater, plumber William Edgar “Special Ed” Flowers.

Flowers and his assistant, Chris Puckett, arrived early on a glorious fall day. By 7:30 a.m., they had assessed the situation from the outside and were down in the basement. I got down there to find the toilet under the pingpong table, and Flowers, with his custom-welded rolling equipment cart, in command. He had settled his large, white-jumpsuited frame on a teeny camp stool next to the hole in the floor where the toilet used to be and was preparing to feed a camera down the pipe. The camera, about the size of a lounge singer’s mike, was attached to a length of plastic-coated fiber-optic cable that Flowers paid out from a great coil. The cable connected to a VCR and 10-inch monitor. Flowers fussed with his equipment for a moment before looking up with a dramatic pause. “OK,” he said, snapping on a pair of latex surgical gloves. “We’re going in.”

I held my breath as the camera picked up an infrared image from inside the 4-inch PVC pipe. “We’re in a three-fixture bathroom group,” intoned Flowers, “going down through the water-closet pipe.”

I couldn’t believe what I was seeing on the monitor, how clean it looked inside the pipe. I expected…stuff. The camera forged ahead through a little bit of running water (Flowers had turned the sink faucet on) which let us know where the bottom of the pipe was; without it, we would have no orientation as the camera on its cable twisted around inside the pipe.

Flowers provided commentary à la Walter Cronkite at an Apollo launch.

“The waste line is now turning in a southerly direction.”

“Continuing down the line. There are signs of stoppage here. Note the discoloration. We are now 10 feet from point of entry.”

We barreled along inside our pipe. At 70 feet from point of entry, we encountered a jagged chunk that clearly shouldn’t have been there. From the camera’s perspective it loomed huge, like the teeth of a steam shovel. It turned out to be a 3-inch fragment of an iron drain cover that had rusted off and fallen in. Could this be our problem?

“We are going to attempt to remove this grate drain from this sewer,” said Flowers. “Audio will be off at this time.” He hauled his camera and wet cable out of the depths, hand over hand, then adjusted his aviator glasses.

After Flowers duct-taped some wire hooks onto the camera head, we went back down to extract the little piece of grate, but it was hard to get purchase on the chunk. Then one of the hooks fell off. Flowers was frustrated. He’s not a guy who likes to put in more than he takes out of a waste line. He had just finished telling us about sloppy plumbers who leave all kinds of tools in the pipes that he gets hired to fish out. With a sigh, he hauled up his line again, strapped a magnet onto the camera head, went down again, and pulled up the hook that had fallen off.

On the next go-round, Flowers used the camera to bump the chunk of iron grate farther down the pipe. He had given up on retrieving it and planned to push it into the main instead. At 110 feet from point of entry, however, we encountered a really big piece of something that looked suspiciously like a piece of driveway. This thing dwarfed the chunk. It was so big that Flowers didn’t want to risk damaging his camera by pushing it, so he dispatched Puckett for the K-1500, which was being used on another job up in Chevy Chase. Puckett hopped to. I got the impression it isn’t every day that he gets to use the K-1500.

While Puckett was gone, Flowers went outside and grabbed his shiny brass divining rods. He then paced deliberately across the street with one held up in each hand, looking like a cross between Jesus and the Man From Glad. Passing traffic actually slowed down for a better view. He insisted that I try it, too. Feeling like an idiot, I grasped one of the thin, L-shaped rods loosely in each fist and walked slowly from one side of the driveway to the other. Unbelievably, as I crossed the line where the outflow pipe is buried, the bent rods distinctly moved in my hands, swiveling toward each other.

“How can this possibly be happening?” I asked. I always had thought divining rods were an old wives’ tale.

“Nope,” said Flowers. “They work. And I can tell you right now you’ve got two lines in the street, which means a separate sewer line and storm line in this neighborhood. And my guess is your problem toilet is connected to the storm line.”

When Puckett returned with the K-1500, which turned out to be an extra-long and extra-large-diameter plumber’s snake, and Flowers ran the camera down again alongside the snake, it turned out that Flowers’ divining-rod analysis was exactly right. At 120 feet from the toilet hookup, the pipe leading from our toilet intersected the big storm-sewer pipe in the street. The snake pushed the iron fragment and the concrete chunk along the last stretch of our pipe; they plopped into the storm line like a couple of Alka-Seltzer tablets and disappeared from view. Flowers allowed the camera to linger at the intersection while we gazed in silent awe at the large, empty storm-sewer pipe on the video monitor. Our pipe had been cleared. The street line looked to be in good shape. It was as if we had reached Mecca.

So this is what we learned: Our neighborhood had separated sewers, which was good. Our basement toilet was improperly connected to the storm sewer, instead of to the sanitary sewer, which was bad. The connecting line had been clogged by chunks of stuff that, during hard rains, caused a backup of all the water flowing off the roof or into the drain at the base of the driveway. It sought the lowest exit point, which was the shower drain, until we plugged it. Then, it was the toilet. If we had plugged the toilet, it would have come out of the sink. This storm water, coming from outside the house, was full of sticks and leaves, but at least it was not sewage per se, for which I was thankful.

After the initial elation of watching the chunks of debris disappear, the implications of our situation began to sink in. Whenever we flushed the basement toilet, it went into the storm sewer, which drained into Rock Creek. The other toilets in the house were plumbed properly, in a stack connected to a lateral waste pipe that ran to the main sanitary sewer. But this one toilet was different; it was in a different part of the house and clearly some shortcuts had been taken, probably by the same sort of sloppy plumbers who leave their tools in pipes. As an officer of WASA, as are all licensed plumbers in the District, Flowers was now obliged to warn us that this condition had been noted and had to be corrected.

I could tell Flowers and Puckett felt bad about solving one problem only to uncover another. “Stop worrying,” said Flowers. “First of all, there are thousands of illegally connected pipes all over D.C. Trust me, you’re far from the only one. Anyway, the storm and sanitary mains may be separate where they run past your house, but knowing this city, they probably combine right down there at the corner.”

In the 1890s, as they tried to “remove sewage from neighborhoods and dispose of it unobjectionably,” the Harrison team came up with a system of interceptor sewers that ran parallel to the District’s streams and rivers and are still in use today. The concept is that during dry weather, waste flows into these intercepting sewers and is carried off to Blue Plains for treatment. During rainy weather, storm water and street runoff combine with raw waste in the intercepting pipes. Because the combined flow volume is much greater—often too great for the pipes to contain it—it overflows the intercepting sewer at specific points and goes directly into the stream or river. Under these circumstances, the thinking goes, storm water dilutes the sewage, so it is not as detrimental to the waterway—and the waterway itself is running more swiftly and at a higher volume. These specific overflow points, or outfalls, are officially designated by WASA as combined sewer overflows (CSOs), and there are 53 of them in the District: 10 on the Potomac, 15 on the Anacostia, and 28 along Rock Creek and its tributaries. According to CSO Update, a monthly newsletter published by WASA, the CSOs in the Anacostia and Potomac Rivers overflow about 70 to 80 times in a year of average rainfall. In Rock Creek, they overflow about 30 times. The total volume of combined wastewater and raw sewage is 3 billion gallons.

If Flowers was correct that the separate pipes running along my street combined somewhere downstream in the system, then it hardly mattered which drains were connected to which lines. It all came out in the wash—or, more accurately, through the CSOs and into the Potomac.

But I had to know, just to satisfy my own curiosity: Was it really the case that all the toilets in Cleveland Park—even the properly plumbed ones—fed into a combined sewer and into a river during big storms? Where exactly did this happen? There was only one place I could find out, and that was WASA’s Documents and Permits Section at the Blue Plains Sewage Treatment Plant.

At Blue Plains, planning-and-design expert Charles Taylor helped orient me to a map showing all the sewers in the city. The red lines are dedicated sanitary sewers, the blue lines are dedicated storm sewers, and the brown lines are combined. In Cleveland Park, the map clearly shows, separated sanitary and wastewater pipes run along my street, just as Flowers had said, but they don’t combine at the corner. They turn east and run downhill toward Connecticut Avenue. The storm line terminates in Rock Creek’s valley, somewhere in the vicinity of the National Zoo. The sanitary line turns southeast and runs along Rock Creek. By this point, it is a 54-inch pipe, with lots of other sanitary lines feeding into it from the heavily populated slopes on the west side of the creek.

In the vicinity of the P Street Bridge, between Dupont Circle and Georgetown, the line clearly intersects a 6-foot combined pipe, surrounded on the map with little dots representing CSOs. Flowers had been right—it doesn’t happen on the nearest corner, but the sanitary sewer does feed into a combined pipe, engineered to overflow in heavy rain. My finger was touching the spot where much of Northwest D.C.’s excretions flow into Rock Creek.

At this point, I wanted to know how much longer Rock Creek would be conveying sewage to the Potomac. I figured I might find some clues on a tour of the treatment plant itself, so I tagged along with a class of Georgetown University law students. We met in the conference room for an orientation session conducted by Patrick Chesley, our guide and driver.

Chesley, who also works as a shift manager at the plant, is a wellspring of detailed information about Blue Plains. The facility, he told us, is the “largest advanced wastewater treatment facility in the world.” “Advanced” means that it provides Stage 3, or tertiary, treatment, which not all facilities do. The plant opened in 1938 and now treats 370 million gallons of sewage a day, serving 2 million people in the District and Fairfax, Loudoun, Prince George’s, and Montgomery Counties. From the first step of primary treatment, in which the untreated mixture of storm and wastewater pours in, through the last step of tertiary treatment, when “98 percent clean water” is released to the Potomac via a 54-inch outflow pipe, there is a 24-hour “detention time” for sewage.

All this was fine and good, but what about the impact of my basement toilet, along with the other sources of the 3 billion gallons of untreated waste (in an average year) that skip Blue Plains altogether because they exit the system through the CSOs? What’s the point of “98 percent clean” if you’re dumping in 100 percent filthy at the same time?

Chesley actually had an answer for this, and it involved a two-decade-long, billion-plus-dollar plan for four enormous underground storage tanks. During wet weather when the intercepting sewers couldn’t handle the volume, he explained, the overflow would pour into the tanks instead of the Potomac, the Anacostia, or Rock Creek. The tanks would be buried at least 100 feet down, and the biggest one would hold nearly 100 million gallons.

For some reason, the law students perked up at this last statement. “One hundred million gallons?” asked one. “How big is that?” Chesley looked nonplussed, as though the student had asked how high 10 feet is. “Big,” Chesley said. “This will be a bigger job than building Metro.”

The whole system is big: WASA serves an area covering approximately 725 square miles and operates some 1,800 miles of wastewater pipes. That’s not counting the flow-metering stations, the storm-water-pumping stations, the treatment plant itself, and the thousands of catch basins, street inlets, and manholes throughout its territory.

The “big-tanks” idea is the centerpiece of the city’s Long-Term Control Plan, a cleanup project that has the full support of leading environmental organizations. It is a system that has been used successfully in other cities, including Atlanta and Chicago, as well as in Milwaukee, although there were some glitches there. Once the plan is implemented, sewage that exceeds the capacity of Blue Plains will flow into the big tanks for temporary storage, rather than into the rivers through the CSOs. Two of the tunnels are in Northeast, one is in Southeast, and the fourth one is in Northwest, under Georgetown.

The entire “big-tanks” plan is scheduled to take 20 years to complete, but the early years are key. Even before the digging, by 2008, the plan calls for reducing CSO spillage by 40 percent through a series of modest steps, such as fixing old pumps, cleaning clogged pipelines, and increasing the storage capacity of existing infrastructure. It is not clear, however, how any of it will get accomplished if the federal government doesn’t start ponying up some of its $1.4 billion cost. In recent years, the feds have contributed to the city’s ongoing overflow-reduction efforts. In its 2006 budget plan, however, the White House proposed no contribution for the first, critical year of D.C. long-term sewage project.

According to Sharon Gang, spokesperson for D.C. Mayor Anthony A. Williams, although the funds weren’t granted through the city’s annual appropriation, D.C. Congressional Delegate Eleanor Holmes Norton is pursuing “other federal sources” that may provide a source of money that is “more consistent over time.” Rather than taking it year-by-year, Gang says, D.C. needs to “find a source that will provide a continuing source of funds for this important project.”

The White House, as it happens, contributes to sewage overflow just like the rest of us.

My own plan for clean rivers is straightforward: keep that basement bathroom off-limits and direct visitors to our first-floor facilities.CP

Art accompanying story in the printed newspaper is not available in this archive: Robert Meganck.