The Last Codex
EN

Theme

Language

Drinking Water

Draw, settle, filter, boil

Date
c. 8,400 BCE
Era
Prehistory
Rebuild stage
II Settlement
Difficulty

Plate

Legend
  1. 1Stone-lined well shaft
  2. 2Headwall, cover and windlass
  3. 3Sloping apron and drain
  4. 4Settling jar
  5. 5Fine sand bed with its biological layer
  6. 6Gravel and outlet pipe
  7. 7Pot at a rolling boil

Anatomy

20 µm
Biological layer, the top 1–2 cm of the sand, where most germs are caughtFine sand over small and coarse gravel, in sectionConcrete body, cast in one piece, with its diffuser plateOutlet tube and covered storage bucketA household sand filter of the kind described in the CAWST manual, cast in concrete, 94 cm tall and cut open at its near corner. Water poured onto the diffuser plate sinks through 54 cm of fine sand and two layers of gravel, then leaves by the outlet tube. The tube rises outside and ends 5 cm above the sand, so 5 cm of water always covers the sand and keeps its biological layer wet. The figure shows a run just after the 12-litre reservoir has been filled; the flow then falls as the water level drops, and the readout gives the time for 90 % of the batch to pass. The manual sets 0.4 L/min for this version and its installation checklist calls 0.6 L/min, the rate of earlier versions, an absolute maximum; slower flow gives cleaner water. The 12 litres poured in roughly fill the pores of the sand (porosity 0.40), so each batch pushes out the water of the one before, which has rested in the sand between runs. The loupe shows the top of the sand with the water moving at its real speed; what is caught fades after a few seconds to keep the view clear, whereas in the filter bacteria are grazed over hours and silt builds up until the top is cleaned. In laboratory tests with a mature layer, such filters hold back 99.98 % of protozoa and 90–99 % of bacteria; in homes, 80–98 % of E. coli. Viruses, too small to draw here, get through much more often. The filtered water still has to be boiled or disinfected, and nothing in the filter takes out dissolved chemicals such as nitrate or arsenic.

Safe drinking water starts with a protected source, such as a lined and covered well dug uphill and well away from latrines and animal pens. When the water is doubtful, settling, slow sand filtration and boiling remove or kill most of the germs behind diarrhoea, cholera and typhoid, but none of them takes out dissolved chemicals.

What it is

Most water that makes people ill looks clean. The danger lies in bacteria, viruses, protozoa and worm eggs, nearly all of them from human or animal faeces. Cholera, typhoid, dysentery and the diarrhoea that still kills young children travel this way.

Safe water comes from a chain of habits: a source kept clean, doubtful water cleared by settling or slow sand filtration and then boiled, and careful storage, since a dirty cup or hand can spoil clean water again.

How it works

Rain washes excreta from fields and yards into streams and ponds. Liquid from a pit latrine seeps through the soil towards shallow groundwater, faster through gravel or fissured limestone than through clay.

Groundwater is usually cleaner than surface water because the soil holds back particles and microbes. A well keeps that advantage when its mortared upper lining, headwall, cover and sloping apron stop runoff and spills from entering the shaft.

Many germs cling to particles of silt and clay. When cloudy water stands still for some hours, the particles sink and take part of the load with them. Settling alone does not make water safe.

A slow sand filter is mostly a living system. Water creeps down through fine sand at a few tenths of a metre per hour. Over the first weeks a layer of bacteria, algae and tiny organisms settles in the top centimetres. Particles stick there, the microbes of the layer feed on many of the trapped pathogens, and others die before they get through. Some bacteria and many viruses still pass, so the filter improves water without making it safe on its own.

Boiling kills by heat. According to the WHO, bringing water to a rolling boil is enough to inactivate pathogenic bacteria, viruses and protozoa. Heat does nothing to chemicals.

How to build

Choose and guard the source. Prefer protected groundwater or a captured spring, then rain from a clean roof, and only then rivers and ponds. Site a well uphill of any latrine, at least 30 m away and further in gravel or fissured rock, and fence animals out at about 15 m, as the WHO suggests. Wash clothes and water livestock downhill.

Dig and line the well. Dig a shaft at least a metre across and line it with stone or brick as you go down. The WHO recommends making at least the top 3 m watertight with mortared joints, while the lining below the water level stays open to let groundwater in. Raise a headwall, fit a cover, and build a watertight apron at least 1 m wide, joined to the headwall and sloping to a channel that leads spilt water downhill. Keep one rope and bucket for the well and never set them on the ground. A new or flooded well must be disinfected and, where possible, its water tested; until then, boil it. Digging or cleaning a deep well is work for experienced well diggers, never for one person alone.

Cloudy water poured through a folded cloth stretched over the mouth of a clay pot A cloth folded several times holds back silt and the plankton that cholera bacteria cling to.

Settle and strain. Let cloudy water stand in a covered jar for several hours, then pour off the clear upper part without stirring the sediment. Strain it through clean cotton cloth folded into several layers. In Bangladeshi villages, sari cloth folded to eight layers cut cholera by almost half.

Build a sand filter. These figures give the principle; a tested design such as CAWST’s biosand filter specifies sand grading and flow. Take a watertight container about 0.9 m tall. Put gravel at the bottom around the end of an outlet pipe, then coarser sand, then at least half a metre of fine washed sand. Bring the pipe up the outside so that its end sits about 5 cm above the sand surface; once the flow stops, check that about 5 cm of water still covers the sand. A perforated plate or flat stone on top stops pouring from digging into the sand. Pour in one batch at a time and let the filter rest for at least an hour, and no more than two days, between batches. When the flow becomes too slow, lightly swirl the very top of the sand with the palm of the hand, without mixing it deeper, and scoop out the dirty water.

Covered earthenware jar on a stone base, with a ladle hanging on a peg beside it A lid, a raised base and a ladle keep hands and dirt out of the stored water.

Boil and store. Boil all water from a doubtful source before drinking it, even after settling and filtering. Bring it to a rolling boil; the CDC advises keeping it there for 1 minute, or 3 minutes above about 2,000 m. Let it cool covered, then pour it into a clean jar with a narrow neck and a lid. Serve with a clean ladle or by pouring, never by dipping a cup or a hand.

Safety. Boiling and filtering deal with germs, not chemicals. Nitrate from manure and latrines, arsenic, fluoride, lead, salt, fuel and pesticides stay in the water, and boiling concentrates nitrate as water evaporates. If a source may be chemically polluted, find another. Settling and sand filters do not remove every germ: boil the water as well. Keep every latrine at least 30 m from a well and downhill of it, with the bottom of the pit at least 1.5 m above the water table at its highest, in the rainy season. Boiling water scalds: keep the pot stable and children away. Do not climb down into a well: the air at the bottom can lack oxygen and the walls can collapse, and a rope protects against neither. Cover every well so that no child or animal can fall in.

History and evolution

The oldest securely dated wells are in Cyprus: at Kissonerga-Mylouthkia, a shaft cut through limestone around 8400 BCE reached water at 8 m.

Filtered public supplies grew out of an experimental filter built at a Scottish bleachworks in 1804. In 1829 James Simpson, engineer of the Chelsea Waterworks Company in London, brought into use a filter bed of about an acre of brick, gravel and sand for water taken from the Thames.

A cast-iron street pump on cobbles, its long handle lying on the ground at its foot Broad Street, September 1854: without its handle, the pump could no longer be used.

In 1849 John Snow argued that cholera travelled in water fouled by sewage, while most doctors blamed bad air. When cholera struck Soho on 31 August 1854, he linked the deaths to the Broad Street pump: of 73 deaths he looked into in the streets around it, 61 were of people known to have drunk its water. The parish removed the handle on 8 September, when the outbreak was already waning. A local clergyman, Henry Whitehead, later traced the likely origin to a cesspool at No. 40, into which the water from a sick baby’s nappies had been emptied. Snow also compared two water companies serving the same streets: houses supplied from below London’s sewer outlets had about 315 cholera deaths per 10,000 houses, against 37 for the company drawing upstream.

In 1908 Jersey City became the first American city to chlorinate its supply routinely; the chlorine left in the water keeps protecting it in the pipes.

Type a name, a date, a material, a domain…