Most of us walk around unaware of the century-old infrastructure under our feet. In London, Paris, and New York, the active sewer backbone dates to the 1850s-1860s. Yet cities like Brasília (1960) and Songdo (2003) were designed with brand-new, integrated sewer networks. The greenfield approach outperforms patchwork systems for about 15-20 years, then converges toward the same challenges: population overruns, aging joints, and capacity limits. Songdo's vacuum sewer system uses negative pressure to move waste uphill and cuts water use by 60%, but its fixed capacity means it can't easily adapt to growth. Meanwhile, no city has ever fully replaced its sewer network — the cost and disruption are prohibitive. Chicago's Deep Tunnel project, a 109-mile overflow bypass 300 feet underground, has been under construction since 1975 and still isn't finished. Instead of replacement, cities use trenchless rehabilitation like cured-in-place pipe lining, which inserts a new pipe inside the old one through a single manhole. The real surprise? The people who maintain these systems aren't disgusted — they describe the work as technical, calm, and oddly satisfying.
#3980: Sewers from Scratch vs. Victorian Patchwork
Do brand-new sewer systems outperform 150-year-old networks? The surprising answer has an expiration date.
Episode Details
- Episode ID
- MWP-4159
- Published
- Duration
- 22:26
- Audio
- Direct link
- Pipeline
- V5
- TTS Engine
-
chatterbox-regular - Script Writing Agent
- deepseek-v4-pro
AI-Generated Content: This podcast is created using AI personas. Please verify any important information independently.
Downloads
Transcript (TXT)
Plain text transcript file
Transcript (PDF)
Formatted PDF with styling
You Might Also Like
Never miss an episode
New episodes drop daily — subscribe on your favorite platform
New to the show? Start here#3980: Sewers from Scratch vs. Victorian Patchwork
We walk around every day with no idea what's running under our feet. In London, a lot of those pipes were laid in the eighteen sixties after the Great Stink — and now there are companies bolting ruggedized IoT sensors onto Victorian brick, measuring flow rates and gas levels in tunnels that were dug before the telephone existed.
Which is the strangest kind of time travel. The sensor's running a firmware update while sitting on brickwork that Joseph Bazalgette personally signed off on in eighteen sixty-five.
Daniel sent us this one, and he's asking three things. First, are there any sewer systems actually built from scratch in the modern era — planned cities where the pipes went in before the people? Second, has any city ever ripped out its entire sewer network and started over? And third — the one I think a lot of people wonder but don't ask out loud — are sewers as disgusting as we imagine, and what kind of person goes down there for a living?
That third question is the one that's going to surprise people. The reality is not what most of us picture.
Let's start with the first question. Are there sewer systems that were designed and built whole, from scratch, in the late twentieth century or later? Because most of what we've got in the developed world is patchwork — pipes from the eighteen hundreds, nineteen twenties, nineteen seventies, all stitched together into something that somehow still works.
That patchwork is the norm almost everywhere. When people picture a city sewer, they're usually imagining something built in the last few decades. But the numbers tell a completely different story. Paris started laying its system in the eighteen fifties under Haussmann. New York's got pipes from the eighteen fifties still in service — seven thousand four hundred miles of them. London's main interceptors were finished in eighteen sixty-five. These aren't museum pieces. They're the active backbone.
You've got century-and-a-half-old brick tunnels, and someone's down there with a bracket and a drill, mounting a sensor that beams pH data to a cloud dashboard. The sensor's designed to survive sulfuric acid and methane, and it's bolted to something Bazalgette's crews laid by hand.
That tension is exactly why this matters. We're not maintaining Victorian sewers as heritage — we're running a modern city on them. The IoT retrofit is happening because those old pipes are still the primary system. You can't just swap them out. So instead you make them smart. Flow meters, gas detectors, temperature sensors — all ruggedized for an environment that eats standard electronics in weeks.
Which brings us to the comparison Daniel's really asking about. There are two ways a city ends up with a sewer system. The first is the one we all know — the patchwork. A town grows, someone digs a drain, the town becomes a city, and a hundred and fifty years later you've got pipes from seven different eras lashed together with engineering compromises and hope.
The second is the greenfield approach. A planned city where the sewer network was designed before the first building permit was issued. The pipes go in, the streets are laid out around them, and the whole thing is supposed to work as one coherent system from day one.
The question is whether that actually delivers. Because on paper it sounds like the obvious winner. In practice — well, cities have a way of outgrowing anyone's plans.
Let's look at the actual greenfield examples, because they do exist. Brasília is the classic case. Inaugurated in nineteen sixty, the Pilot Plan designed by Lúcio Costa and Oscar Niemeyer included a sewer network sized for about five hundred thousand people. Everything laid out in advance — separate stormwater and sewage lines, treatment plants positioned downwind, the whole thing.
Then reality happened.
Brasília today has over three million people in its metropolitan area. The original sewer system was overwhelmed within two decades. They've been patching and extending it ever since — adding new trunk lines, building satellite treatment plants, basically doing exactly the patchwork approach the original design was supposed to avoid.
The greenfield advantage had a shelf life of about twenty years.
Less, in some ways. The infiltration rates started climbing almost immediately as joints aged and ground settled. But here's the thing — during those first fifteen to twenty years, Brasília's system genuinely outperformed comparable patchwork cities. Fewer overflows, lower maintenance costs per connection, better treatment compliance. The design wasn't wrong. It just assumed a static city, and cities don't do static.
What about the more recent attempts? Songdo, for instance.
Songdo is fascinating because it's the first city designed from the ground up with digital infrastructure baked in. Construction started in two thousand three on reclaimed land near Incheon. They used a vacuum sewer system instead of gravity — smaller diameter pipes, negative pressure instead of slope, which means you can route sewage horizontally and even uphill.
Wait — uphill sewage? How does that work?
Vacuum stations create negative pressure throughout the pipe network. When a valve opens at a collection point, the pressure differential pulls the wastewater in. It's the same principle as those pneumatic tube systems old banks used for checks, except for sewage. The big win is water savings — about sixty percent less water per flush compared to gravity systems, because vacuum toilets use air to move waste. Songdo's entire network has flow meters, pH sensors, and gas detectors embedded from day one, all feeding into a central control room.
They're not retrofitting IoT onto Victorian brick. They designed the sensor grid alongside the pipe grid.
And that integration matters more than most people realize. When you retrofit a sensor into a hundred-and-fifty-year-old sewer, you're dealing with unpredictable flow patterns, chemical environments that vary wildly depending on what's been discharged upstream, and physical access that was never designed for electronics. Thames Water's smart monitoring network in London — which is impressive — still has to contend with the fact that some of those tunnels have never been fully mapped. The original drawings exist, but a century and a half of repairs, collapses, and ad hoc modifications means the as-built reality often doesn't match.
Retrofitting IoT onto legacy systems is actually harder than building it in from scratch.
The sensor housing alone is an engineering challenge. You need something that survives sulfuric acid — which forms when hydrogen sulfide meets moisture on the tunnel walls — plus methane, plus occasional complete submersion, plus the physical impact of debris. And you need it to transmit reliably through several meters of soil and brick. Songdo's sensors don't have to solve any of that. They were specified before the concrete was poured.
Songdo has its own problem, doesn't it? The city was built for a projected population that hasn't fully materialized.
That's the tradeoff in a nutshell. Greenfield systems lock in design assumptions. Songdo was planned for about three hundred thousand residents and a specific mix of residential and commercial usage. If the population grows differently, or if industrial patterns shift, the system isn't flexible in the way patchwork is. A patchwork city can run a new trunk line to a growing neighborhood and tie it in. A vacuum system has fixed capacity at each station — you can't just add a pipe and hope the pressure holds.
The ugly, stitched-together system actually has a hidden strength. Each patch was a response to a specific local condition.
Each patch taught the engineers something about the ground, the flows, the failure modes. That institutional knowledge accumulates. The flip side, of course, is that patchwork systems have way more combined sewer overflows — when stormwater overwhelms the shared pipes and raw sewage discharges into rivers. Greenfield systems with proper separation don't have that problem, at least not at the same scale.
Until the city outgrows the separation design and someone runs a storm drain into a sanitary line because it was the only option that fit.
Which is exactly what happened in Brasília. So the real answer to Daniel's question — do greenfield systems perform fundamentally better — is yes, but with an expiration date. They outperform for a window, then converge toward the same challenges every other city faces. The sensor integration advantage might change that equation, because you can detect problems earlier. But we won't really know for another decade or two.
Greenfield systems have advantages, but they're not magic. That brings us to the harder question. Daniel asked whether any city has ever just ripped out its entire sewer network and started over. I've got to admit, I've never heard of it happening.
Because it basically hasn't. The closest anyone's come is Chicago, and even that isn't a replacement. In nineteen seventy-five, Chicago started digging what's called the Tunnel and Reservoir Plan — TARP, or the Deep Tunnel. It's a hundred and nine miles of tunnels, two hundred fifty to three hundred feet below the existing sewers.
Three hundred feet down. That's not a sewer anymore, that's a geological feature.
It's not replacing anything. The old combined sewers are still there, still flowing. What the Deep Tunnel does is catch what they can't handle during heavy rain. When the old pipes would normally overflow into the Chicago River, the excess gets diverted down into these massive tunnels and held until the treatment plants can process it.
It's a bypass valve the size of a subway system.
It's still not finished. They've been at it for over fifty years, and the total cost is somewhere north of four billion dollars. Not replacement, not new treatment — just somewhere to park the overflow.
Which tells you everything about why full replacement never happens. If just building overflow storage costs that much and takes half a century, imagine actually digging up every pipe.
Let me put some numbers on that. Replacing a single mile of urban sewer runs between five and twenty million dollars, depending on depth, soil conditions, and how much of the street you have to tear up. A medium-sized city might have two thousand miles of pipe. You do the arithmetic.
We're talking tens of billions before you even deal with the disruption.
The disruption is the part that kills these proposals before they leave the committee room. If you're replacing the entire network, you're digging up every street. Every single one. You'd need to reroute traffic for years, maintain temporary sewage handling for the entire population, and somehow keep businesses open while their front doors open onto an excavation pit.
Plus you'd have to coordinate all of this across election cycles, budget fights, and the people whose basements you'd be digging through. It's a political impossibility before it's even an engineering one.
Here's the key insight Daniel's question gets at — there's supposed to be a tipping point where maintenance costs exceed replacement costs, and then rip-and-replace makes economic sense. But that tipping point is a mirage, because trenchless technology keeps moving it further out.
Instead of digging up the pipe, you rehabilitate it from the inside. The most common method is cured-in-place pipe — you insert a resin-soaked liner into the old pipe, inflate it, and cure it with hot water or UV light. You end up with a brand-new pipe inside the old one, and you did it through a manhole.
You're giving the pipe a new interior lining without excavating the street.
There's also pipe bursting, where you pull a new pipe through the old one, fracturing the original outward, and sliplining, where you just slide a smaller pipe inside. All of these methods cost thirty to sixty percent of full replacement, and they can extend pipe life by fifty-plus years.
The old pipe hasn't actually failed yet, and for half the price you can buy it another half century.
Which means the economic equation almost never tips toward rip-and-replace. By the time the rehabilitated pipe degrades again, there'll probably be even better techniques. The tipping point keeps receding.
London's Tideway project — that's the super-sewer they've been digging — is that a replacement?
No, and it's the perfect illustration. The Tideway is a fifteen-mile tunnel running under the Thames, cost four point two billion pounds, completed just last year. But it's not replacing Bazalgette's Victorian interceptors. It's running parallel to them, capturing the combined sewer overflows that the old system can't handle anymore. Same logic as Chicago — add capacity, don't replace.
The pattern is: the old system stays, and we dig deeper to handle what it can't.
We make the old system smarter with sensors. That's the real "new system" — not new pipes, but old pipes with new intelligence layered on top.
Alright, let's pivot to the human question, because I think this is where Daniel's really curious. What's it actually like down there, and who does this work?
The first thing to understand is that the popular image is mostly wrong. Modern sewers in developed countries are not teeming with rats and cockroaches. The flow rates are too high — rats need dry nesting areas, and the main channels are fast-moving and toxic.
The scene in every movie where someone drops into a sewer and is immediately swarmed by rats — that's not real?
Not in a functioning system. Rats live in the cracks and dry side passages, not in the main flow. And cockroaches don't do well in environments with high hydrogen sulfide concentrations. The real hazards aren't vermin. They're gases.
The big one is hydrogen sulfide — H two S. It's colorless, and at low concentrations it smells like rotten eggs. That's your warning. But here's the terrifying part: at higher concentrations, it paralyzes your olfactory nerve. You stop smelling it. And then, within minutes, respiratory failure and death.
The smell disappearing is the danger signal, not the smell itself.
Sewer workers are trained on this relentlessly. If you smell rotten eggs and then suddenly you don't, you don't think "oh good, it cleared up." You evacuate immediately. There's also methane, which is explosive, and a whole soup of pathogens — hepatitis, leptospirosis, E coli. Workers wear full-body hazmat suits, self-contained breathing apparatus, and carry personal gas monitors that alarm at the first trace of H two S.
This isn't unskilled labor by any stretch.
Far from it. These people are trained in confined space entry, gas detection, emergency rescue procedures, and often specialized trades like welding or electrical work for pump maintenance. New York City's DEP has what they call benchmen — inspectors who physically walk the pipes, some of which date to the eighteen fifties. They're checking for cracks, root intrusion, joint separation. It's methodical, high-stakes work in an environment that can kill you if you lose focus for thirty seconds.
What kind of person signs up for that?
The ones I've read about and talked to — they tend to be methodical, unflappable, comfortable with confined spaces, and they have a dark sense of humor that probably functions as a coping mechanism. They're not reckless. Reckless people don't last. They're careful in a way that becomes second nature.
There's something almost monastic about that. Descending into the dark every day, doing precise work in a place most people will never see, keeping a city alive from underneath.
They're almost entirely invisible. You walk down the street, you don't think about the person who was in the pipe below you at three in the morning, checking a joint with a flashlight and a gas monitor. But if they stopped showing up, you'd know within days.
The whole city would back up. Public health would collapse faster than most people realize. These are the people who prevent cholera outbreaks before they start, and we don't even know their names.
We've established that rip-and-replace is almost never the answer, and that sewers aren't the horror show most people picture. What do we actually take away from all this?
That the greenfield versus patchwork debate is kind of a distraction. The real question isn't which approach is better — it's whether you accept that all infrastructure ages and needs adaptation. Brasília had the perfect plan on paper, and within twenty years it was patching things up like everyone else.
That's not a failure of planning. It's just what happens when a city actually lives. People move in, industries shift, rainfall patterns change. No blueprint survives contact with reality. The best system isn't the one that was designed perfectly — it's the one that can be adapted without breaking.
Which is why the patchwork approach, for all its messiness, has a weird kind of resilience. Every fix was a specific response to a specific problem. The system carries its own history.
The real innovation right now isn't about building new pipes. It's about making old pipes intelligent. The IoT sensors going into Victorian sewers — that's the actual new system. It's just distributed across infrastructure that's older than everyone listening to this.
The future of infrastructure isn't gleaming new tunnels. It's a hundred-and-fifty-year-old brick arch with a sensor on it, quietly reporting pH levels to a dashboard that someone checks on their phone.
That reframes the whole investment question. We keep waiting for the tipping point where replacement makes sense, but trenchless rehab plus smart sensors keeps pushing that point further out. Why spend twenty million a mile to dig everything up when you can reline it for half the cost and monitor it in real time?
Next time you flush, think about this. The pipe under your street might have been laid before electric light. And somewhere on that pipe, there's probably a sensor — or there will be soon — that's watching for the first sign of trouble, sending data to someone whose job is to make sure you never have to think about any of this.
That someone is probably in a hazmat suit, holding a gas monitor, working in the dark while the rest of us sleep. The real new system isn't the technology. It's the combination — old brick, new sensors, and the people who go down there.
There's one question I keep coming back to, though. All this math about trenchless rehab beating replacement — it assumes rainfall patterns stay roughly where they've been. But they're not.
That's the climate wildcard. New York and Miami are already recalculating. When a hundred-year storm hits every five years, combined sewer overflows stop being an occasional failure and become a routine public health problem. At some point, rehabbing the old pipe doesn't help if the pipe's diameter was sized for a world that no longer exists.
The tipping point might not be economic in the maintenance-cost sense. It might be forced by water volume.
That's the scenario where rip-and-replace — or at least massive new capacity — stops being theoretical. Miami's dealing with sea-level rise pushing groundwater into aging pipes from the outside, plus heavier rainfall from above. That's a system being attacked from both directions. The old equation assumed the environment stayed constant. It didn't.
The question Daniel's really asking might get answered in the next twenty years, whether we want it to or not.
Even then, I suspect the answer won't be tear-it-all-out. It'll be something like Chicago's Deep Tunnel writ larger — parallel capacity, smarter diversion, more sensors giving earlier warning. The people going underground won't be replaced by machines either. You still need someone to inspect the joint that the sensor flagged, to make the call about whether that crack is stable or about to fail.
Which brings us back to the human part. The invisible guardians. I keep thinking about the fact that modern public health owes more to the people maintaining sewers than to almost any other profession, and we pay them like it's unskilled manual labor.
The training requirements alone should put that myth to rest. Confined space certification, hazardous atmosphere protocols, emergency rescue — and that's before you get to the actual pipe inspection skills. These are technical professionals working in an environment that can kill them silently. The pay doesn't reflect the stakes.
If every sanitation worker in a major city called in sick for a week, we'd have a public health emergency inside of days. Cholera isn't a historical relic — it's kept at bay by the people who go down there.
If you take one thing from this episode, maybe it's that. Next time you see a maintenance crew opening a manhole, those aren't background characters. They're the reason your water runs clear and your street doesn't smell like the Thames in eighteen fifty-eight.
Now: Hilbert's daily fun fact.
Hilbert: In the early medieval period, Polynesian wayfinders identified a specific deep red pigment made from the sap of the noni tree mixed with iron-rich volcanic clay. When applied to canoe hulls, the color shifted subtly with humidity, serving as a rudimentary weather indicator during long ocean voyages.
Hilbert: In the early medieval period, Polynesian wayfinders identified a specific deep red pigment made from the sap of the noni tree mixed with iron-rich volcanic clay. When applied to canoe hulls, the color shifted subtly with humidity, serving as a rudimentary weather indicator during long ocean voyages.
Humidity-sensitive canoe paint. That's clever.
I have so many questions about the chemistry, but I'm going to sit with it.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. If you enjoyed this one, leave a review and send us a weird prompt of your own — we read every one.
We're at my weird prompts dot com.
This episode was generated with AI assistance. Hosts Herman and Corn are AI personalities.