Civilization101Blueprint

City water and sewerage

Keep a dense settlement from poisoning itself.

Tier 4 · Industrial revolution Water t4-sanitation Blocked by prerequisites

Also listed under Water & Sanitation › Make water safe, Settlements & Planning › Settlement and planning, because that subject depends on it.

Sewer pipe going into a trench. Unglamorous, and it saves more lives than any hospital. Sansewer · 2005-04-21 · en:User:Imprezzive1 · Public domain · Commons ↗

Everything for this step

Field guide
Sanitation and disease → 8 sections Micro-organisms → 6 sections
Builds
None yet.
Reference sheets
None.
Sciences it rests on
Microbiology · Engineering
Track
Sanitation, level 2 of 2

Why this matters

Industrial cities concentrated people faster than they concentrated sanitation, and the result was cholera and typhoid at scale. Piped clean water and separated sewage produced the largest gain in human life expectancy in recorded history — larger than any medical treatment.

The shape of it

  1. Separate supply from disposal, absolutely and physically. Take water from upstream of the settlement; discharge downstream of everyone, including your neighbours. Broad Street in 1854 is the canonical lesson: one pump, contaminated by one cesspit, killed hundreds.
  2. Filter, then disinfect. Slow sand filtration removes turbidity and much of the pathogen load; chlorination finishes the job. Chlorination of public water is, per unit of effort, among the most effective public-health measures ever deployed.
  3. Pipe it under pressure so contamination cannot be drawn in through a leak. An intermittent, unpressurised supply sucks in groundwater at every joint.
  4. Build sewers with self-cleansing velocity — egg-shaped sections keep flow fast at low volume. A sewer that silts becomes a blocked sewer becomes a flooded street.
  5. Treat the sewage; do not merely relocate it. Settlement, then biological treatment — trickling filter or activated sludge — then discharge. Untreated discharge just moves the epidemic downstream to someone else.
  6. Recover the nutrients where you can. Treated sludge returns phosphate and nitrogen to farmland, which matters enormously before synthetic fertiliser.
  7. Make it a public function with the authority to compel connection. This never happens voluntarily — every historical case required law, taxation and enforcement. That is a political achievement, not an engineering one.
  8. Monitor continuously. Test the water; do not assume it. Most failures are silent until there is an outbreak.

Doing it, step by step

The summary above says what has to happen. This says how, with the numbers that decide whether it works.

1 · Put the intake upstream and the outfall downstream of everyone

Physically, absolutely, and with a margin. Take water from above the settlement; discharge below every household, and below your neighbours, not just below your own boundary.

Define a protection zone around the intake in writing: no latrines, no burials, no livestock, no workshops and no waste within it, enforced by the health authority. Broad Street in 1854 is the canonical lesson — one pump, contaminated by one cesspit a metre away, killed hundreds in ten days. The failure is almost never the treatment works; it is a pipe or a pit somebody put in the wrong place.

✓ Check: Walking upstream from the intake for a kilometre finds no latrine, cesspit, midden, burial or animal pen, and the zone is marked on the ground and on paper.

2 · Filter, then disinfect — and measure the disinfection

Slow sand filtration takes out turbidity and most of the pathogen load: 0.15–0.35 mm effective grain size, a bed of at least 60 cm, 0.1–0.4 m³ per m² per hour, ripened 2–3 weeks. See the slow sand filter.

Then chlorinate, which is the step that closes the gap on viruses. Aim for a free chlorine residual of at least 0.5 mg/l after 30 minutes of contact, at pH below 8 where chlorine is far more effective, and at least 0.2 mg/l at the furthest tap as the evidence that nothing has been added along the way. Turbidity must be below 1 NTU first — chlorine spends itself on suspended matter.

✓ Check: Residual chlorine measured at the tap furthest from the works reads above 0.2 mg/l, and turbidity into the chlorine contact tank is below 1 NTU.

3 · Keep the mains under pressure, always

An intermittent, unpressurised supply sucks groundwater in at every joint — and in a settlement with any sanitation problem, that groundwater is sewage. A pipe is only a barrier while the water in it is pushing outward.

Hold a minimum of about 15 m of head (1.5 bar) at the highest tap, from a service reservoir or a tower sized for at least a day's demand so pumps stopping does not mean pressure stopping. Chlorinate to leave a residual through the whole network. Fix leaks as contamination events rather than as water losses, and never let a repaired main back into service without flushing and disinfecting it.

✓ Check: Pressure logged at the worst point in the network stayed above 1.5 bar for a month, including during every pump stoppage.

4 · Build sewers that clean themselves

A sewer that silts becomes a blocked sewer becomes sewage in the street. The design rule is velocity, not size: at least 0.6–0.75 m/s at the daily flow, which for a small foul sewer means a steady fall — typically between 1 in 60 and 1 in 150 for a 150 mm pipe, depending on flow.

Bigger is worse: an oversized pipe runs shallow and slow. Egg-shaped brick sections hold velocity at low flow, which is why the Victorian interceptors were built that way. Put access chambers at every junction, every change of direction and every 50–100 m, and ventilate the system — sewer gas is explosive and, in a confined chamber, lethal.

✓ Check: Every reach meets its velocity at the measured dry-weather flow, and opening any chamber shows flowing sewage with no standing silt.

5 · Treat the sewage instead of relocating it

Untreated discharge exports the epidemic downstream to someone who did nothing to earn it.

  1. Screening and grit removal first, or everything downstream fills with rag and sand.
  2. Primary settlement — 2 hours or so of quiet retention removes perhaps a third of the oxygen demand as sludge.
  3. Biological treatment — a trickling filter of 40–60 mm stone with rotating distributors, or activated sludge with air blown in. This is where most of the work happens, and it is alive: it needs feeding steadily and it dies if shock-loaded or poisoned with trade waste.
  4. Final settlement, then discharge to water with enough flow to dilute what remains.

Judge it by measurement: falling oxygen demand and no sewage fungus or dead fish below the outfall.

✓ Check: The effluent is clear and odourless, and the river below the outfall supports the same life as the river above it.

6 · Recover the nutrients, carefully

Treated sludge returns phosphate and nitrogen to farmland, which matters enormously before synthetic nitrogen — and phosphate never stops mattering, because there is no synthetic route to it at all.

But treat it first: digest the sludge anaerobically at about 35 °C for 15–30 days, which cuts pathogens sharply and yields methane worth burning, or compost it hot at 55 °C for several days. Keep industrial waste out of the sewers that feed it — heavy metals from plating and tanning make sludge permanently unusable on land, and that decision is irreversible. Apply to arable, not to salad crops, and observe a waiting period before harvest.

✓ Check: Sludge going to farmland has been digested or hot-composted with the temperature and time recorded, and no trade effluent discharges to that catchment untreated.

7 · Make it a public function with power to compel connection

This never happens voluntarily. Every historical case required law, taxation and enforcement, because the household that must pay to connect gets a benefit that is mostly enjoyed by its neighbours.

So: a duty on the authority to provide, a duty on the householder to connect within a stated period, a charge levied on all rather than on users, and a power to do the work and recover the cost where someone refuses. Fund maintenance in the same law as construction — the commonest failure in the record is a works built by a grant with no revenue behind its operating cost, silted up within a decade. See the first public good.

✓ Check: Every dwelling in the served area is connected, and the maintenance budget is a standing line in the accounts rather than an appeal.

8 · Monitor continuously, because failure is silent

Most failures give no sign until there is an outbreak. A schedule, written and audited:

  • Daily — chlorine residual at several points, turbidity, pressure.
  • Weekly — bacteriological samples from the network. The standard is no E. coli in 100 ml, and a single positive is an incident, not a statistic.
  • Continuously — flow and level alarms at the works.
  • After every repair or burst — flush, disinfect, and sample before restoring supply.

Publish the results, including the failures, and keep them for years so trends are visible. Test the water; never assume it.

✓ Check: A year of daily residual and weekly bacteriological results exists with no unexplained gaps, and every positive sample has a written investigation against it.

The better path

What the record shows was done at this step, what to do instead at this tier, and the evidence. All of them: the better path.

Do not repeat

Lead water pipes

What was done
Lead was the standard pipe for water mains and house connections from Rome to the 20th century: soft, easy to join and long-lasting.
Do instead
Lay mains in cast iron, and house connections in iron or fired clay.
Evidence
Vitruvius, around 25 BC, advised clay pipes over lead because lead workers were visibly poisoned. The UK stopped installing lead water pipes in 1970. In 2023 the US Environmental Protection Agency estimated 9.2 million lead service lines still in use; in Flint, Michigan, a 2014 change of water source dissolved lead from them into the supply.

Needs: Blast furnace and cheap iron, Pottery and the first kiln

Do not repeat

Water supply before sewerage

What was done
Cities piped water into houses first and dealt with the waste later. Water use per person rose several-fold and flowed into cesspits built for a fraction of it, which overflowed into the ground and into the rivers the water was drawn from.
Do instead
Build the drain with the supply: no house connection to the mains without a connection to a sewer or a properly sized soakaway.
Evidence
London's flush toilets drained to cesspits and, after 1815, to sewers emptying into the Thames — the source of much of the city's drinking water. Cholera killed about 14,000 Londoners in 1848–49 and about 10,000 in 1853–54, and the 'Great Stink' of 1858 finally funded Joseph Bazalgette's interceptor sewers.

Needs: Brick and tile, Cement and concrete

Do not repeat

Combined storm and foul sewers

What was done
Most 19th-century cities built one sewer for rainwater and sewage together. In heavy rain it overflows, by design, into rivers.
Do instead
Lay separate pipes for foul water and rainwater. The trench is most of the cost of a sewer; a second pipe in it is a fraction of the cost of retrofitting one later.
Evidence
London's combined system discharged about 39 million tonnes of untreated sewage a year into the Thames until the Thames Tideway Tunnel — 25 km long, costing about £4.5 billion — was completed in 2025 to intercept it.

Needs: Brick and tile, Cement and concrete

Adopt earlier

Chlorinate the supply

What was done
Bleaching powder — chlorine absorbed into slaked lime — was made commercially from 1799, yet cities drank unchlorinated water for another century.
Do instead
After filtration, dose chlorine to leave at least 0.5 mg/l free chlorine after 30 minutes of contact, and at least 0.2 mg/l at the far end of the network.
Evidence
Maidstone in England disinfected its mains with bleaching powder during a typhoid outbreak in 1897. Jersey City began continuous chlorination of its supply in 1908, the first in the United States; a court challenge upheld it as safe and effective, and chlorination spread across American cities within a decade.

Needs: Acids, Lime

How this fails

Sewer outfall upstream of an intake — anyone's intake.
Intermittent pressure, which draws contamination in.
Discharging untreated sewage and exporting the problem.
Building the works and not funding the monitoring.

How you know it worked

VerificationWaterborne disease has effectively disappeared from the settlement, and routine testing confirms it rather than assuming it.