The guy running the snake down our sewer looks matter-of-fact. Our sewage has been backing up. Right next to the pipe connecting our house to the sewer line running down our street stands a 70-year-old willow oak, and I worry the tree's roots have found their way, during the droughty past year, into our line. He shrugs: Maybe it's tree roots, maybe it's a collapsed pipe, maybe it's a yo-yo. The snake went in only a dozen feet or so and found a clog, and now the little claw at the end is spinning. Once he pulls it out we'll know better what's going on. I leave him to his business, though I cast an annoyed glance at the oak. Sewer pipes fit together simply, with a bell joint, and tiny root hairs find their way to the nutrient-rich flow, then grow larger, eventually growing large enough to shatter the vitreous clay pipe that forms so many service lines or dislodge a joint if the pipes are cast iron. Nobody knows what our pipes, 70 years old, are made of, but I fear we're about to find out.

Fifteen minutes later he's winding the snake back up, writing a bill, and exonerating the oak.

"Do you have a baby?" he asks. We do.

"Do you use those flushable wipes?" We do.

"Don't," he says. The entire paper industry in recent years has worked to develop more and more flushable items: baby wipes, moist adult wipes, antibacterial bathroom scrubbers, diaper liners, diapers. He shakes his head: If it doesn't come apart in your hands, don't flush it. All it has to do is hold its form for an hour or so and it can find a place to catch: a joint, a root, a pimple on the inside of the pipe, one of the little mounds of rust buildup called tubercules. Then, like a snag in a river, it starts catching other stuff and you've got yourself a situation, either for you or for your whole neighborhood. We're like a nation of 1-year-olds, throwing everything in the toilet. "Toilet paper and what comes out of you," he says. "That's what should go in the toilet." Take the goldfish outside and bury it; otherwise the best case is it's just going to get caught in a screen at the treatment plant. It won't biodegrade on the way down, and it might cause trouble. And let's not even bring up those garbage disposals--we had had another guy out 6 months before and he excavated enough of a neatly processed carrot that with sufficient patience we could have reconstructed it. The sewer, person after person tells me, is for sewage.

Your favorite pop culture reference to sewage may involve Art Carney, in character as Ed Norton, singing, "Together we stand, with shovel in hand, to keep things rolling along." Or maybe it's one of those scenes from Phantom of the Opera or Les Mis, with all kinds of French high drama occurring amid the atmospheric flow. I prefer Carl Spackler in Caddyshack, cackling while creating plastic explosive animals against a backdrop of sacks of the common golf-course fertilizer Milorganite. You scarcely notice it, but I'll decode that: Milorganite is short for MILwaukee ORGAnic NITrogEn, a soil treatment produced by the city of Milwaukee's wastewater treatment plant since 1925 and now used on lawns all over the country. It's the end result of their sewage treatment, and they ship thousands of tons of it each year.

The point isn't so much that what happens to our sewage reaches into every crevice of our culture. The point is that once you're managing it instead of wishing it away, sewage turns out to be a pretty good thing.

In the North Carolina State Archives in Raleigh, just hanging around atop some cabinets lies an extra set of 4-foot-square planning maps made in 1922--the first planning documents in Raleigh's history. They make great idle-time study: "Locations of Fires in Buildings: One of a Series of Preliminary Zoning Studies," says one. Another shows the water distribution system, a 16-inch and a 14-inch line coming from the pumping station down by Walnut Creek to the city water tower; another shows hard-surface paving; a fourth demonstrates "Barriers to Street Extensions and Residential and Commercial Growth."

By far my favorite is "Sewer Mains and Laterals," with thick, colored-pencil stripes in brown, blue, and yellow showing the locations of different sizes of underground sewer pipes--starting from 6-inch diameters in neighborhoods like mine to the largest mains back then, 24 inches. What I love about the map is the outfalls--at Crabtree Creek north of town and Walnut Creek to the south (safely downstream from the pump that brings drinking water to the city), the colored-pencil stripes simply stop. That's where the sewage goes: into the river.

Those days seem almost absurdly quaint now, but they're not so bygone after all. In 1940, in some of the largest cities in the United States--Boston, Pittsburgh, Cincinnati, St. Louis, Kansas City--every drop of whatever you flushed down the toilet was dumped untreated into a nearby harbor, river, or lake. New York City in 1940 treated approximately one quarter of its sewage, and it reached 100 percent only in 1986. Until then, had you visited your Aunt Louise on the Upper West Side, all your business would have flowed directly into the Hudson.

Historians estimate that before indoor plumbing became widespread, the average person used less than 5 gallons of water per day; nowadays a good round (and low) estimate of American at-home water consumption is 100 gallons per day, per person. Some of that gets sprinkled on lawns and a bit washes cars and pets, but overall we use that water either for cleaning ourselves and our dishes and clothes, in which cases it ends up going down the drain, or for drinking, in which cases it ends up going down the toilet. Every day each one of us turns 100 gallons or so of water into sewage. That's a lot of sewage, requiring a lot of treatment--and very little of it is poop.

At least now we do treat it. Though people have been piping sewage for thousands of years, actual sewage treatment is barely a century old. People had to figure out first that human waste was not just noisome but actually unhealthy and then how it was unhealthy before they could begin figuring out what to do with it. Once they did, they got busy in a hurry; you can all but drink most of the water that comes out of Western treatment plants, and most of the biosolids removed during the process are used to fertilize crops and treat soil. The system is not flawless--biosolids sometimes contaminate water; grease clogs cause sewage spills or system failures; heavy metals, pharmaceuticals, and personal care products build up in biosolids--but overall it works splendidly.

Perhaps the first written sanitation instructions come from the Bible, which, written by and for a nomadic people, takes a small-is-beautiful approach: Deuteronomy urges you to dig a hole and "cover that which cometh from thee." By about 3000 BC, inhabitants of the Orkney Islands had invented toilets: Existing stone hut walls from that period have little niches with holes that drained to underground channels. The sewer historian Jon Schladweiler says that by a thousand years later, civilizations throughout the ancient Mediterranean and Middle East were using pipes to transmit both stormwater and human wastewater away from homes and cities and, usually, into waterways. By about 1500 BC the Cretan palace of Knossos had an actual flush toilet--a seat, a pan, and a slave to pour water to sluice what disposable-diaper companies call "the insult" to a drain in the floor. Cretan techniques for channeling water and wastewater spread throughout Greece, and by the 5th century BC, Athenians were piping wastewater and stormwater to a reservoir outside of town and using it to irrigate crops.

The Romans improved on even that: After considering Rome's many accomplishments, Pliny the Elder called the sewers "the greatest accomplishment of all." (The word "sewer" comes from the Latin exaquare, "to carry away water.") The constant flow of water coming into the city from the aqueducts supplied public fountains and baths, and Romans figured out that public bathwater ought to be changed a couple of times a day. "They built public latrine buildings immediately adjacent to the baths," Schladweiler says, and flushed the latrines by routing the used bathwater under them. The majority of human waste, though, was simply thrown into the streets; aqueduct water was used to wash the streets and sweep that waste into the drains. Because Roman sewers lacked ventilation, the only egress for sewer gas was those same drains and latrines. On the plus side, Romans also invented portable toilets, setting urns by the side of the road near the entrances to the city (vendors would rent you what Schladweiler calls "a modesty cape"). Further, the 1st-century emperor Vespasian had workers collect the contents of urinals, which he then taxed and sold to fullers, tradesmen who cleaned and dyed the Romans' clothing--they had figured out that the ammonia in urine had cleaning powers.

After the fall of the empire, Romans kept throwing filth in the streets, but nobody was washing them. In Rome many sewer pipes fell into disrepair. Everywhere else people got along without them as they always had: at best using latrines (unlined pits) or cesspits (pits lined with perforated masonry that let liquids drain away into the soil while solids piled up for eventual removal) and at worst throwing their waste into the streets and leaving it there. In the 13th century the French king Philip II paved the streets of Paris to reduce the stench, with the result that afterward the waste sat on the stones instead of percolating into the soil. In the 14th century, one of his successors, Philip VI, ordered Parisians to sweep in front of their houses and take the refuse to a dump; crews of sanitation workers were organized to clean up whatever was left. In a return to the technology of the Roman Empire, in 1370 Paris opened a series of drainage canals that also carried waste--the biggest was lined with masonry and called the Grand Egout, or Great Drain. By the 16th century one British royal castle had to post signs reminding people not to "foul the staircases, corridors, or closets with urine or other filth." When the palace of Versailles opened in the 17th century, it had lovely splashing fountains but no bathrooms or sewers.

The world changed in 1842, when the city of Hamburg, after suffering a terrible fire, decided to lay sewer pipes while rebuilding. The new pipes vented through house drains and had a mechanism for flushing using tidewater. The system was efficient, didn't stink, and became a worldwide model. (Before the introduction of these sewers, typhoid, transmitted through water tainted by sewage, caused 48.5 of every 1,000 deaths in Hamburg; after the sewers came into use the number dropped by half.) Immediately thereafter the Parisians began turning their 14th-century sewer system into a wonder of the world, building hundreds of miles of huge brick tunnels to carry away stormwater and everything else Parisians cared to sluice inside.

When early American cities such as Boston and Philadelphia began paving their streets with cobblestones in the 17th century, gutters--and even some underground sewers--were included among the improvements. Private citizens built Boston's first systems, designed, like the Cloaca Maxima and the Grand Egout, to drain cellars and swamps. Bostonians soon grew weary of the constant repairs those wooden sewer lines required and undertook a sort of public-private partnership by issuing construction permits for sewers; everyone who wished to connect a drain had to share in the cost, and the contracts stipulated requirements about pavement reconstruction. Philadelphia had a system of culverts and some underground sewers by 1750, and New York City started putting a few sewers underground later in the century. Human waste, though, remained mostly a personal matter of cesspits and privies.

Sewers really took off in 1854, with John Snow's discovery that the London cholera epidemic was caused by sewage-tainted drinking water. With advances in microbiology, people began to understand that human waste carried disease in the form of microbes, and increasingly they wanted to protect themselves from their sewage. What's more, the introduction of reliable water service in the 19th century and the spread of the modern flush toilet (the British Public Health Act of 1848, which required every home to have some kind of sanitary arrangement, listed "water closet" as one of the alternatives to an ash pit or privy) vastly increased the amount of wastewater households generated. Cesspits and privies that had already created offensive nuisances now produced vast, vile-smelling seepages, overwhelmed by the new volume of water. And it wasn't just toilets, either--connections draining sinks and tubs began overwhelming sewer pipes, too; in 1844 Boston tried to slow the tide, literally, by passing a law requiring a doctor's order for every bath.

As cities grew in size and density during the Industrial Revolution, they all had to build more, and better, sewers. The cholera epidemic wasn't motivation enough for London, but the "Great Stink" of 1858, when the Thames smelled so bad that Parliament considered relocating, got the city government's attention; it built new sewers in the 1850s and '60s to carry waste downstream from central London. Brooklyn introduced sewers in 1857, and Chicago not long after. Boston, still largely building sewers privately, had about 100 miles of sewers in 1869; by 1885 that had expanded to 226 miles, and new houses were expected to connect to the system both for pump and washbasin waste and for the human waste now going into flush toilets instead of privies.

Every city had its own problems and its own characteristics. Some of Boston's sewers had outfalls dammed by the tide 12 hours of every 24; others, built by unscrupulous contractors in land reclamation projects like the Back Bay, sagged and lost their downhill slope, causing settling, clogs, and backups. Sylvan Seattle had pipes made of wooden staves--and faced a tide problem so severe that at certain times of day toilets became foul geysers; eventually the city simply rebuilt itself higher than its sewer pipes. In Chicago, the outfalls of the sewers made such a mess of Lake Michigan that during large rainstorms the plume of tainted water flowed all the way out to the intake for the water system. In response, engineers built a series of canals and reversed the flow of the Chicago River, turning it from a drainage into Lake Michigan into a flow from Lake Michigan toward the Mississippi. They also moved the intake farther out into the lake.

All these "solutions" merely moved the problem. As one historian said in describing Boston's covering a brook filled with sewage and routing it to the Charles River rather than directly into Boston Harbor, this "somewhat lessened the nuisance caused by it, or at least transferred it to another locality." Older cities on the coasts built combined systems channeling both sewage and stormwater, whereas newer and smaller cities built separate systems--both storm drains and systems of much smaller pipes that handled only sewage--thus preventing sewage overflows in storms. Lennox, Massachusetts, built the first such system in 1875, and Memphis built one in 1880. Since then, that's what everybody has built.

Raleigh laid its first sewer pipes in 1890. Fayetteville Street, Raleigh's main road, wasn't paved until 1886, at exactly the same time the first water pipes were being laid; where water pipes go, sewer pipes soon follow. The privies of Raleigh's population of barely 10,000 almost certainly had not yet polluted the soil enough to foul its wells, and the new sewer pipes, running north to Crabtree Creek and south to Walnut Creek, would not have discharged more than the streams could absorb. (A stream running at about 6 cubic feet per second can absorb the waste of about 1,000 people, so to support 10,000 people the two creeks together would have had to flow at around 60 cfs. Currently, on a dry day in a dry month, they flow at about 75 cfs.) Now, with 2,300 miles of pipes all heading roughly southeast to Raleigh's wastewater treatment plant, the sewer collection system turns out to be the only infrastructure stream that follows that natural tree pattern that I'd expected to find everywhere. The leaves are houses, connected by 4-inch service lines to 6- or 8-inch mains that run mostly beneath streets, and then to 18-, 24-, or 30-inch collectors that start out along streets but head downhill to creek basins, leading to larger and larger pipes and finally to the plant. I sat down with a friendly GIS expert to check it out.

The GIS map easily showed me the path of my own wastewater: the 4-inch lateral in my yard--the same pipe that "flushable" wipe clogged--runs into an 8-inch main, which heads downhill along my street until it crosses the Pigeon House Branch, down by the pool I like to sit by. It runs along the Pigeon House until it joins a 24-inch PVC east of town (the path is following rivers by then, not roads), and thereafter joins larger and larger pipes--some made of PVC, some of reinforced concrete, some of ductile iron. Eventually this stream hits the dual 72-inch reinforced concrete pipes that head directly to the sewer plant, though those sometimes separate into three or four pipes, for ease of maintenance. It's simple and, especially after the spaghetti tangle of the water lines, rather satisfying. It's much like the stormwater system, if every ravine in every drainage basin remained piped and they all came together in one place before entering the Neuse.

To find out what happens in these pipes, I talked to Raleigh's dean of pipage, sewer collection superintendent Hunter "Gene" Stanley. "We're not like New York City," Stanley says right off. "Some of those you can drive a truck through." New York has a combined sewer system that has to be prepared to move the billions of gallons of water that a major storm could dump on the city, not just the comparative trickle of sewage the city generates daily. Combined systems manage overflows with relatively simple mechanical junctions called regulators: basically weir dams in pipes or junction boxes. A weir is nothing more than a low barrier for steering water. When flow is routine, the dam routes it through pipes to the treatment plant; during large rain events, the flow of mixed stormwater and wastewater rises high, overtops the weirs, and flows directly through outfalls to rivers or lakes. Such an event is called a CSO, or combined sewage overflow. New York dumps about 40 billion gallons of CSOs into its rivers and harbors every year.

But before you draw too much comfort from Raleigh's system having to convey only sewage (the plant treats about 45 million gallons per day that are generated by the 400,000 or so customers connected to the system; it's rated for 60 million gallons, and it's being expanded to 75), consider this: The increase in flow caused by nothing more than rainfall and street flow coming in through manhole vents in low-lying areas can nearly double the flow to the treatment plant. Really? "Oh yeah, oh yeah," Stanley says. "A thousand gallons a day [per manhole] if it rains all day." Add in leaky joints, cracks, and holes made by thirsty tree roots, and you can have significant water infiltration during rainstorms. Though catching and correcting the breaks and overflows are an unavoidable part of his job, Stanley stays focused on preventive maintenance.

Stanley grew up in rural North Carolina and has called his preventive maintenance management "an ol' country boy work system"--he copies pages from the map book of his system and gives them to his crews. When the crew has flushed and inspected every line on the map, it comes back. The department logs its maintenance in feet per day, and it likes to reach 300,000 feet per month if it can, meaning that every pipe in the system gets a look-see once every few years. GIS keeps the maps updated, of course, but Stanley's system has been working since they were using nothing more than blueprints and as-built surveys; finding that what's an 8-inch pipe on the map is really a 6-inch is just part of keeping on top of things. That's why you carry different-size saw blades in your truck.

Stanley says a sewer is a simple thing: The pipe needs to drop about half a foot per 100 feet of length, a slope of 0.5 percent, which is fast enough to keep everything moving, but not so fast that the liquid races away from the solids. Bigger pipes--30 inches or larger--can slope even less. But they all must flow downhill, powered by gravity, which is why sewer pipes so commonly crisscross the stormwater drainages: Raleigh Public Utilities Department director Dale Crisp calls all the sewers that run in a particular drainage a "sewershed," which for a while became my favorite new word.

Of course, if wastewater pipes followed only natural gullies, the mains would eventually have to parallel the river, and for many reasons, from aesthetics to the catastrophic results of a spill, nobody wants that. The system generally moves downhill, but pipes sometimes need to cross rises. So the city has more than 100 lift stations, where the contents of pipes are pumped to join other flows or where wastewater from lowlying areas collects in sumps. When the water gets high enough, it trips a float valve and a pump clicks on and lifts it up a hill--kind of like your toilet, only this float valve starts the flush instead of stopping it. I visited one lift station, a 10- by 20-foot rectangle of electrical boxes that look like a central air-conditioning system behind chain-link fences between two houses, controlling an underground sump; even when it's pumping, if you were more than 10 feet away you wouldn't hear it. The station has a backup pump and a generator to power it, plus a little antenna to send information back and forth to the supervisory control and data acquisition (SCADA) system at the treatment plant; that's plenty of equipment, but just the same, if you weren't looking for it you wouldn't know it was there. A much larger station sits on the trunk line, giving a lift to pretty much all of Raleigh's waste on its way to the plant. It's underneath a highway on-ramp, and though some people suggested I could find it by following my nose, it didn't smell when I went out to visit it.

Stanley hands over a laudatory profile of Raleigh's sewer maintenance department in a recent issue of Municipal Sewer and Water magazine, then hands me off to Robert Smith, a sewer monitoring supervisor and asks him to show me around.

First things first: We walk the yard, checking out trucks. Sewer guys basically do three things: They perform maintenance, they respond to crises, and they "TV" pipes, sending tiny little vehicles with cameras on them up the pipes to check both their condition as part of general maintenance and whether the crews who claim to have recently maintained them have actually done so.

Smith shows off the department's various trucks. Rodder trucks have a spool of linked rods, a sort of long chain that the workers feed into a manhole and then rotate, just like someone cleaning roots or a clog out of your drain at home. Some rodders have cutting blades or spiral grabbing implements to clear roots or debris. Flushing trucks carry enormous water tanks to feed high-pressure hoses with spinning heads on the end: Workers feed the hose into the system, usually past the next manhole, and then turn on a pump. Water pressure starts the head spinning, spraying water at thousands of pounds of pressure per square inch back toward the truck as the truck pulls back the hose, scouring the pipes along the way. Standard now is the combination truck, which carries tanks of water for flushing and a garbage-truck-size tank for postflush water, which the truck vacuums up with a huge tube that hangs from a derrick over the cab like an elephant's trunk. The driver eventually empties that tank onto a pad in the parking area, Smith explains; water drains off into the sewer system and the cleaned-out debris--tampons, bricks, gravel, roots, supposedly flushable materials--gets loaded into a dump truck once a week and sent to the landfill. Smith marshals those vacuum trucks when Raleigh has a sewage overflow, too. Another truck he calls a blockbuster has a water hammer--a pipe that uses water to rhythmically pound and break up large blockages. Finally, he shows me a sort of souped-up golf cart that provides access to the many parts of the system that, because they follow ravines rather than roads, are not easily reached by regular trucks.

But we're standing in a parking lot while people are out in the field, rodding sewers. "Hey," he says. "You wanna shoot the line?" You bet.

Our first stop is a highway off-ramp, where two flush trucks and a pickup are parked behind orange cones. Several men wearing hard hats, green mesh vests, and rubber-palmed gloves manage a hose coming off a spool on the back of one of the trucks and running to a manhole 20 feet down a steep ravine. A hundred yards away, two guys stand at another manhole looking out for the spinning head of the water jet, which Smith says is called a Warthog. Once it's past, the guys still at the truck turn on the jet and the spool to start reeling it back in. Over the roar of the truck engine Smith explains that on the way out the head sprays as a sort of presoak; "on the way back, it's like a broom." Water-jet cleaning like this is standard for clearing roots, grit, and, especially, grease: "We run into some lines [where] you think, Where in the world did all this grease come from? It looks like you sprayed foam on that pipe." Because the bell-and-spigot joints in sewer pipes provide places for tiny tree roots to enter, and because trees got very thirsty during the drought, crews commonly run a saw through the line after cleaning just to make sure they've got everything. Where the vacuum trucks can't reach a manhole, the crew flushes debris downstream to one the truck can reach.

That's sewer flushing, and the sanitation department does it all day long. Ever since the Hamburg sewers first captured tidal water and then released it all at once to flush out debris, the basic idea hasn't changed much: You use water to flush, you use rods or hooks to attack clogs, and, as Ed Norton sang, you keep things rolling along.

Smith packs us back in his pickup and we drive to a parking lot and a box truck with a picture of a fish on it. The three guys in the truck are going to TV a pipe: Mike is preparing the camera and the screens in the back of the truck while Wayne and someone who introduces himself only as "the Rev" open the manhole, popping the cover off easily with a metal hook. Wayne and the Rev then retrieve the camera from the truck. With six tiny rubber wheels and an inquisitive single eye, it looks a bit like the Mars rover vehicle, only tiny and dangling at the end of a wire. When they come back to the manhole Wayne and the Rev are shocked to find it suddenly filled with sewage. This kind of backup indicates a block in the 6-inch pipe at the bottom of the manhole, though it drains away as fast as it backed up.

A few moments of observation shows two things: The backup comes and goes rhythmically, meaning there's a pump station upstream that sends a pulse of wastewater every couple of minutes, and the blockage is a bunch of pieces of some solid substance that nobody can identify. Out come spoons--hooked, perforated shovels on the end of 12-foot handles. Wayne, Robert Smith, and Eddie, another supervisor who has arrived, take turns scooping, pushing things back and forth between rushes from the pump and pulling them out with an awkward hand-over-hand motion that keeps the gunk barely balanced on the edge of the spoon unless you knock the handle against an overhanging tree branch. It's like using an iced-tea spoon to fish olive pits out of a bleach jug at the back of a cupboard. "And people think it's Ty-D-Bol that keeps their bathrooms clean," Wayne says.

The stuff turns out to be congealed grease, and pieces of it are sufficiently solid--and sufficiently far up the 6-inch pipe--that they block the progress of the camera every time the Rev dangles it down there and tries to get it running. The vacuum nozzle can clear the manhole but can't pull grease out of the pipe and it resists everything else they've got, so the crew finally gives up on TV-ing that pipe for the day, until they can clean the pipe--possibly by using a bucket truck (which feeds a cable past the debris and drags a bucket from one manhole to the next, pulling before it the kind of grit and large debris flushing just doesn't get) or possibly by sending someone down there in the hope that a simple scoop into the pipe will clear the debris. (Sending someone down a manhole, though it's only about 8 feet deep, requires confined spaces training, extra supervision, and ventilation equipment--sewer gas contains methane and hydrogen sulfide, and it has killed workers as recently as 2008.)

Smith shows me video footage from another TV-ing expedition that shows long traverses down shiny pipes half full of dull gray water. The color makes sense--much more of it comes from your washing machine and shower than from your toilet. "First thing people say is 'Eew,'" Smith says; "they think I'm walking around in feces." But even the wastewater filling up the manhole that day smelled more like runoff than poop.

Though most blockages are caused by grease or roots, the talk naturally turns to memorable clogs, and I hear about mops, golf clubs, firewood, riprap, and even a refrigerator that have had to be pulled out of manholes. Once a carpet remnant created a block so nasty it took most of a day to clear out. If you're on call and someone calls in a spill, especially one where the overflow is making its way toward a waterway, then it's showtime. "You go running after it like it's a Russian spy," Wayne says. "You chase the spill, pulling hose, four, five, six miles." First the crew finds the end of the spill in the waterway--where the water is still clean--and sets up a block using hay bales, which both dam the flow and filter any water that might trickle through. A pump immediately starts channeling the polluted water into the nearest downstream manhole. And while a crew works on clearing the clog itself, other crews chase the spill, hosing down the sides and bottom of the stream. You can tell when untreated wastewater has hit a stream, Smith says, by the powdery-looking buildup it leaves: "It looks like gray dust in the water," coating the rocks and sticks. The hoses clear the scum off the bottom and stir up the mud. "That muddy water acts like a glue to that stuff--it piggybacks on the mud." Then you pump it out at the end of the spill. "After we go through, that creek looks like nothing ever hit it. It's pretty neat."

One day I pulled over on a main highway to watch a sewer crew fix a leaky pipe using what's called cured-in-place pipe: A long liner impregnated with resin is pushed into a pipe by water, then they pump steam through the pipe to harden the resin, and presto--the pipe is, though slightly narrower, all but new. The crew runs an auxiliary pipe while they fix the leaky one. Workers can find leaks by stopping up pipes with sandbags, pumping smoke down a manhole, and then seeing where the smoke starts creeping out of the ground. Clearly, people have thought about this stuff a lot.

"It gets in your blood," said a senior engineer and inspector named Dave, who managed a construction site near my house. I had stopped by a sewer truck to watch a couple of guys hose out a pipe, which they followed to a manhole at the backyard boundary of a few properties. "I don't like the looks of that," one had said when, spooning out grit from the hole, he unearthed pieces of clay pipe. Rather than dig on several people's private property, the city decided to abandon the line running between the houses and reroute the flow to the end of the street, where it could join a larger main and head downhill. But since the flow had to go downhill, the engineers had to get wastewater to flow away from the manhole in a different direction. That meant, simply, digging a deeper trench in the direction they wanted it to go. They had to do the same thing on the line that received the new flow, so they were digging on two streets for a few weeks, and I occasionally stopped by.

I watched them carefully lower new green PVC pipes and check the slope. They use a laser to measure, with a sort of bull's-eye target at the end. The red dot right in the center means the pipe has the right slope. "We used to run string lines, length by length," Dave told me. "This makes it easier. A lot easier." They lay pipe on a bed of gravel, supporting it with rocks and pieces of brick to maintain the slope, unlike water pipe, which, under pressure, can just be laid in a trench and buried. Any pipe below 12 feet has to be ductile iron, to support the weight of the earth on top; same with any pipe that crosses a creekbed, hangs under a bridge, or does anything but lie directly on the earth. Even though sewer pipes start out low and have to keep going lower, in a city with no subway system there's not much other infrastructure in the way; 12 feet is fairly deep for Raleigh.

The Neuse River Wastewater Treatment Plant, southeast of Raleigh, discharges most of the 40-plus million gallons of water it treats every day in a state bordering on potable. Spread over 300 acres (and surrounded by 1,200 acres of farm fields for application of biosolids), the plant is big enough that to see it you have to tool around in a van; you can't walk it like you can the water treatment plant miles upstream. Superintendent T. J. Lynch started my tour with a drive to the headworks, where screens and vortex filters remove the floating junk and grit from the outflow of the twin 72-inch trunks flowing in from the sewer system. The floating stuff is more obvious, but Lynch emphasizes the basins that settle out grit: In a process heavy on pumps and pipes, grit is anathema. "Think about sand," he says. "It's abrasive. It will literally wear out your equipment." They fill three Dumpsters a day with grit and the debris caught by screening, all of which goes to the landfill.

Remarkably, though the air has a certain tang, it doesn't stink. "Ninety-nine-point-nine percent of it is just water," Lynch says. "We've just got to get that 0.1 percent out."

Outside the headworks Lynch points to a concrete basin almost the exact size and shape of a baseball field, sloping downward from the outfield to a low point at home plate: a 32-million-gallon equalization tank. The sewer system does not have water towers or storage tanks to accommodate regular fluctuations in wastewater volume (flow peaks at breakfast time and again just after midnight), so it stores water in the equalization tank during periods of high flow and, especially, storms: Lynch says, "It's not uncommon for us during a heavy rain to see our flow double." The equalization tank gives Lynch a place to put that excess flow. When the flow slows down again, he uses four enormous screw pumps to push the water to the top of the slope on which the plant is built--the rest of the way through the plant, gravity does the work. He shows off the pumps proudly. They're called "vertical turbine solids-handling pumps," but he notes that they use technology roughly unchanged since the time of Archimedes: turning screws to lift water.

From the headworks the water flows to primary clarifiers--tanks through which the wastewater flows extremely slowly, not unlike the settlement basins at the water plant, and with the same goal: allowing solids to settle to the bottom, creating a "primary sludge blanket," though here grease and oil also float to the top. Scrapers make a circuit along the surface, scooping grease into a small flume, then cycling to the bottom where they shepherd the sludge to its own pumps. Water cleanliness is measured by biochemical oxygen demand (BOD)--the amount of oxygen the bacteria in the water use to remove its organic impurities. The lower the BOD, the smaller the next basins can be and the lower the plant's operating costs. This is a crucial point, given how much energy the next step takes. That step is what Lynch calls "the absolute heart and soul" of his plant: the activated sludge process.

So after a few hours in the clarification tank, the water flows into aeration basins, six concrete pools of several million gallons apiece, the bottoms of which are crisscrossed by air nozzles. These enormous tanks of what looks like boiling brown sewage are just what you imagine when you think "sewage treatment." But it's not boiling: What's happening is aeration, which provides oxygen, in the presence of which bacteria love to eat poop. "We are bug farmers," Lynch says. Bacteria in the basins multiply rapidly, like the starter for sourdough bread. The wastewater provides the food, the nozzles provide the oxygen, and the bacteria feel like they're on a cruise: nothing but breathing, eating, and reproducing, with free food all day long.

This process removes all the harmful chemicals from the water--except nitrates, which feed algae in rivers. These algae propagate wildly and then die. "And the decomposition sucks up all the oxygen in the river, killing the fish. So now we're going into the same tanks and we put walls in there and we have different zones where we stimulate the bacteria to do different things" that remove nitrogen. After about a day of that, the water spends 2 days flowing very slowly through secondary clarification basins: a dozen large circular tanks, each the size of an aboveground swimming pool. "It's very quiet," Lynch says. "The bacteria settle down and create what we call the sludge blanket," which sinks to the bottom and is pumped out. The water trickles out through V-notch weirs around the top of the tank, by which point it has a barely yellowish tinge and no odor whatsoever.

We'll get to the next stage of water filtration, but what's more interesting at this point is what happens to the sludge. Most of it becomes RAS--return activated sludge--and rejoins new wastewater in the aeration basins, providing the starter for a whole new treatment reaction. Bacteria can go around the cycle half a dozen times, Lynch says, but eventually they become WAS--waste activated sludge. In the WAS stream, the bacteria go to four big covered tanks called aerobic sludge digesters, in which instead of wastewater nutrients they eat each other. We climbed up to the top of one of the basins and looked through a porthole inside: You could almost hear the screaming. Okay, not really; it's just a tank of sludge. "If it's good sludge, it's got that nice brown color--it's oxygenated, and it don't smell bad up here," Lynch says. If it's black, something's out of whack--it probably needs more oxygen.

Aerobic digestion, Lynch says, is actually inefficient--it costs money, since the plant has to pump in oxygen; anaerobic digestion, to which the plant hopes to switch, is the opposite: It generates methane, which the plant can use to generate electricity to operate its pumps and blowers; the waste heat from the generators can even help provide the heat the anaerobic digesters need to keep the bugs at optimum eating-each-other temperature. "There are plants now that generate their entire power needs from that methane," he says. "They pull no energy from the grid."

From the digesters we drive to the dewatering building--just as at the water treatment plant, the wastewater plant uses belt presses to squeeze the water out of the last, bug-eat-bug stage of the sludge. Enormous screw conveyors grind big pieces of the dewatered stuff along like the mixers in an industrial bakery. Lynch hands me a piece of what they call cake. It has no odor at all and feels a lot like inch-thick rubber, the kind used under playground equipment. Polymers are introduced to help coagulation and add to the rubbery feel. "Don't forget to wash your hands," Lynch says.

We leave the sludge and return to the water. Primary treatment is mechanical--settling; secondary treatment is biochemical--activated sludge digestion. Tertiary treatment combines mechanical and photochemical: The water flows to a dozen deep sand filters, which remove the remaining fine particles. The final hurdle the water clears is ultraviolet treatment. We walk over to a small, low concrete building, in front of which Lynch pulls up a flat piece of aluminum to reveal a 4- or 5-foot-wide channel of water flowing smoothly past banks of green-glowing UV lamps. They look much like fluorescent lights, though they have a self-wiping mechanism to keep them clean. "It doesn't kill [pathogens]," he says of the light. "It just scrambles their RNA. Sterilizes 'em." He mentions in passing an open concrete channel between the filters and the UV treatment, in which tilapia swim. "It keeps the duckweed down," he says. Once fish are swimming in your treated sewage, you've got to feel like you're getting the job done.

Out behind the UV building we finally reach the finished product: the effluent, flowing out in strong wide arcs into a canal from a dozen UV channels. The water seems to have a greenish tinge, but when Lynch fills a little water bottle and holds it up it looks utterly clear. Then we drive down toward the end of the line, where the effluent enters the Neuse River. There, in a tiny green glade at the bottom of the hill, the river bends into a cove, near the bank of which a few surface bubbles give away the presence of the underground pipe. Lynch says if you didn't know where the pipe was you could follow shad upstream to find it: "The water's warm and highly oxygenated," so the fish love it.

Lynch says the river is the natural place for the water anyhow. Every drop of Raleigh water comes from the Neuse watershed and ought to return there. "All we're doing," he says, "is what a river would do." That is, rivers are naturally self-cleaning, and the Neuse would absorb waste, given time and distance. Waste on the ground would be dispersed by rain and the water would be filtered through the earth, entering the groundwater system clean; matter that made it to the stream would become waterlogged and sink. "On the rocky bottom, bacteria live," he says, that feed off the nutrients in waste. "What happens in our plant is the exact same thing that happens in a stream. That's exactly where the process came from. We've just concentrated it," Lynch says. "It might take the river a couple hundred miles to accomplish what we do in a couple days." We've been talking about sewage treatment for 2 hours--about dewatered cake and grit screens and activated sludge and UV filters--and suddenly I understand.

What the river does is exactly what a sewage plant does, and until around World War I most people didn't really think plants were needed at all. It was understood that watercourses were to some degree self-cleaning, that "the solution to pollution is dilution." But as populations increased, especially downstream, dilution stopped offering much of a solution. Chicagoans felt great about their famous flow-reversal of the Chicago River, sending their sewage into the Mississippi instead of into their own water supply of Lake Michigan. Predictably, the people of St. Louis were not as delighted. Their expression of dissatisfaction eventually landed them in front of the Supreme Court. Raleigh itself was sued in the late 1940s by downstream neighbor Smithfield. Smithfield won, and in 1956 Raleigh's first sewage treatment plant opened, treating 12 million gallons per day. The current plant replaced it in 1977; it's expanded since, and now it's expanding again.

Raleigh's sewage treatment will be limited not by space, but by nitrogen. In 1995 the state of North Carolina set limits on nitrogen based on totals for that year. Raleigh's plant released 1.3 million pounds of nitrogen into the Neuse in 1995, and its current limit is 49 percent of that total, or 676,496 pounds. In 2007 it released 233,061 pounds, its lowest total ever. Still, the state-mandated limit will keep dropping, and Raleigh's capacity to treat its sewage will probably finally reach its limit in pounds of nitrogen rather than in millions of gallons per day.

Modern sewage treatment is a boon, of course, but it creates a new problem. If you treat sewage so effectively that by the time water leaves the plant it's almost drinkable, then what do you do with all the sludge you cleaned out of it? People have been fertilizing with their sewage since the ancient Athenians, but modern Americans seem to mistrust the process. As recently as 2006, when water-starved San Diego considered a water-saving measure to reintroduce effluent from its sewage plant into its reservoir--from which, of course, the city's water would still be treated before use--the Union-Tribune opined, "your golden retriever may drink water out of the toilet with no ill effects. But that doesn't mean humans should do the same." The practice would have made the reservoir no different than the Mississippi or the Neuse or any other source of water into which the effluent of cities upstream has been poured (water managers like to spout the true-enough-for-discussion statistic that by the time Londoners drink water from the Thames, it's been through seven sets of kidneys upstream; others repeat the same chestnut about New Orleans and the Mississippi). Two years later San Diego's mayor claimed the reuse measure was a waste of money and vetoed it. The city council overturned his veto, but it's clear: Even in the parched Southwest, people are frightened of wastewater reuse.

(In 2008, when Raleigh introduced a plan to save water by reusing effluent from the Neuse plant--by piping it only to large-scale users, such as universities, business campuses, and country clubs, and only for irrigation--some politicians resisted. They claimed to object to "subsidizing" large-scale users through lower rates for the reuse water. Fortunately, enough people remembered the drought; the system was built, using pipes foresightedly laid almost a decade earlier.)

Given that people feel that resistant to highly treated water, what on earth to do with all this sludge? New York used to dump its sludge far out in the ocean, a practice that didn't represent a huge improvement over dumping raw sewage, until that was outlawed in 1992, after which the city started shipping sludge by train to a Texas landfill. Plenty of cities still landfill their sludge. Raleigh chooses to follow the example of Milwaukee: More than 90 percent of its sludge is beneficially reused as biosolids (nationwide about 50 percent of sludge is reused). Some of Raleigh's muck is thickened on gravity belt presses and applied directly to fields as Class B biosolids (according to the EPA, this use requires a buffer between fields and the public, and they can't be used on crops for human consumption). Some of this was overapplied to the Raleigh farmland surrounding the plant in recent years; a series of poor practices resulted in groundwater contamination near the plant, so the practice was stopped, though the plant currently has the highest certification level of the National Biosolids Partnership (the NBP is a nonprofit comprising three national groups: Two represent the sewage treatment industry, which may not be the most objective source for biosolids practice assessment, and the third is the Environmental Protection Agency, which probably is).

Some primary sludge is dewatered and sold to a composting company, which treats it and sells it to agricultural products companies as safer, more highly treated Class A biosolids. The rest is mixed with lime, which raises the temperature high enough to pasteurize it, and sold to farms, institutions, and the public as Raleigh Plus, Raleigh's own version of Milorganite. Maybe I'll use some on my lawn. Nobody has any statistics on how it affects gophers.

Reprinted from On the Grid: A Plot of Land, an Average Neighborhood, and the Systems That Make Our World Work by Scott Huler. © 2010 by Scott Huler. By permission of Rodale, Inc., Emmaus, PA 18098