---
titulo: Solar water distillation
descripcion: Glass box solar still for brackish or seawater: construction, yield per m², sizing, salts and variants.
nivel: 2
estado: borrador
requisitos: []
relacionados:
  - 01-agua-y-saneamiento/potabilizacion/hervido-del-agua
  - 01-agua-y-saneamiento/potabilizacion/desinfeccion-solar-sodis
  - 01-agua-y-saneamiento/potabilizacion/desinfeccion-quimica-del-agua-con-cloro-y-yodo
  - 01-agua-y-saneamiento/potabilizacion/coagulacion-floculacion-y-sedimentacion-con-alumbre
  - 01-agua-y-saneamiento/potabilizacion/filtro-lento-de-arena-y-biofiltro-de-carbon
  - 01-agua-y-saneamiento/potabilizacion/filtros-de-ceramica-y-carbon
  - 01-agua-y-saneamiento/potabilizacion/analisis-basico-de-calidad-del-agua
fuentes:
  - "ITDG (Intermediate Technology Development Group, hoy Practical Action). Solar Distillation. Technical Brief, Rugby (Reino Unido), c. 2002. Apartados «Design types», «Output of a solar still», «Experience» y «Would a solar still suit your needs?»"
  - "McCracken, H.; Gordes, J. Understanding Solar Stills. VITA Technical Paper n.º 37, Volunteers in Technical Assistance, Arlington (Virginia), 1985, parte III (diseño: profundidad, separación vidrio-agua, pendiente)"
  - "McCluney, W.R. Solar Distillation of Water. Energy Note FSEC-EN-3-80, Florida Solar Energy Center, febrero de 1984. Apartados «Basic principles» y «Water quality»"
  - "Khalifa, A.J.N. (2011). «On the effect of cover tilt angle of the simple solar still on its productivity in different seasons and latitudes». Energy Conversion and Management, 52(1), 431-436"
  - "Hanson, A.; Zachritz, W.; Stevens, K.; Mimbela, L.; Polka, R.; Cisneros, L. (2004). «Distillate water quality of a single-basin solar still: laboratory and field studies». Solar Energy, 76(5), 635-645"
  - "SolAqua / El Paso Solar Energy Association. Solar Still Basics. Guía técnica, El Paso (Texas), s. f. (recomendación de purga diaria con 3 veces la producción)"
  - "Delyannis, E. (2003). «Historic background of desalination and renewable energies». Solar Energy, 75(5), 357-366 (planta de Las Salinas, Chile, 1872)"
  - "Jackson, R.D.; van Bavel, C.H.M. (1965). «Solar distillation of water from soil and plant materials: a simple desert survival technique». Science, 149(3690), 1377-1379"
  - "Dvorak, B.I.; Skipton, S.O. Drinking Water Treatment: Distillation. University of Nebraska-Lincoln Extension, NebGuide G1493, revisión de diciembre de 2013"
  - "Kozisek, F. «Health risks from drinking demineralised water». En: OMS, Nutrients in Drinking Water, Ginebra, 2005, cap. 12, pp. 148-163"
  - "OMS. Safe Drinking-water from Desalination: Guidance on risk assessment and risk management procedures. WHO/HSE/WSH/11.03, Ginebra, 2011 (arrastre de volátiles en procesos térmicos)"
  - "OMS. Guidelines for Drinking-water Quality, 4.ª ed., Ginebra, 2011, cap. 10 (umbrales de sabor de cloruro, sodio y sólidos disueltos)"
  - "OMS/WEDC. Technical Notes on Drinking-Water, Sanitation and Hygiene in Emergencies, nota n.º 9: How much water is needed in emergencies, tabla 9.1"
  - "Sphere Association. The Sphere Handbook, 4.ª ed., Ginebra, 2018, norma 2.1 de abastecimiento de agua"
  - "Stumm, W.; Morgan, J.J. Aquatic Chemistry, 3.ª ed., Wiley, 1996, cap. 7 (solubilidad de la calcita en equilibrio con el CO2 atmosférico)"
  - "Ministerio de Vivienda / IETcc-CSIC. Catálogo de Elementos Constructivos del CTE (conductividad térmica de aislantes)"
  - "AEMET. Atlas de radiación solar en España utilizando datos del SAF de Clima de EUMETSAT, 2012; Comisión Europea, JRC, PVGIS (irradiación media mensual por localidad)"
  - "Malik, M.A.S.; Tiwari, G.N.; Kumar, A.; Sodha, M.S. Solar Distillation. Pergamon Press, Oxford, 1982 (texto de referencia citado por ITDG; no consultado directamente)"
idioma: en
traduccion: automatica
---
# Solar water distillation

## Summary

A box-type solar still (or "tray still", *basin still*) is a shallow black tray holding brackish
or seawater, covered by an inclined glass sheet. The sun heats the water, the vapor condenses on
the inner, cooler face of the glass, slides down to a trough and exits into a closed container.
Salts, metals, mud and microorganisms stay in the tray. It is mainly useful for **removing salt**,
something that boiling, filtering, chlorinating or SODIS cannot achieve. If the available water is
fresh but dirty, those other methods are much cheaper.
What it yields, in figures: **2 to 3 L of distilled water per m² of glass per day, on average, in
sunny climates** (ITDG, c. 2002: 2.3 L/m² with 18 MJ/m²; McCluney, 1984: 1.9-3.8 L/m²). In winter or
with clouds it drops to 0.5-1 L/m². That is little: a family of 4 needs about 6 m² of still **just
to drink**, and even so, outside summer, will have to supplement it with rainwater or fire
distillation.

## Prerequisites

- **Knowledge:** none essential. It helps to know [boiling water](hervido-del-agua.md)
  (backup on cloudy days), [chemical disinfection with chlorine](desinfeccion-quimica-del-agua-con-cloro-y-yodo.md)
  (for stored distilled water) and [coagulation and sedimentation](coagulacion-floculacion-y-sedimentacion-con-alumbre.md)
  or the [slow sand filter](filtro-lento-de-arena-y-biofiltro-de-carbon.md) if the water is turbid.
- **Materials for a 1 m² tray module** (measurements from step 4):
  - **Transparent flat glass**, 3-4 mm thick: a sheet **1.15 m (on the slope) × 1.11 m**, weighing
    about 13 kg if it is 4 mm thick (1.28 m² × 10 kg/m²). Salvaged window glass works. Tempered glass
    withstands hail better, but cannot be cut: it has to be used at its existing size.
  - **Untreated wood boards** 20-25 mm thick: about 8-10 linear meters of 20 cm wide board for the
    walls (including waste), 1.1 m² of board for the bottom, another thin board (1.0 m²) as a false
    bottom, and 4.5 m of 2 × 4 cm batten to hold the glass. **Do not** use wood treated with creosote
    or CCA (greenish, from posts and fencing) or painted with old paint, which may contain lead.
  - **Black liner:** EPDM or butyl pond liner, 1 mm thick, **2.0 × 1.8 m**, so it rises to the top
    edge of all the walls and the wood never touches the vapor. Alternatives: a stainless-steel tray
    painted matte black or, at N2 level, a mortar tray (see "Variants").
  - **Dry insulation:** polystyrene or cork, 5 cm (conductivity ≈0.035-0.045 W/m·K) or straw, sawdust
    or shavings, 8-10 cm (≈0.05-0.08 W/m·K), per the CTE Building Elements Catalogue. For under and
    around the tray: about 120-130 L with 5 cm, or about 250 L with 10 cm.
  - **Trough**, 1.00 m long and 3-5 cm wide: split tube of stainless steel, glass, food-safe PVC or
    glazed ceramic, **lead-free**.
  - **Four short pipes:** inlet (12-20 mm, with a plug), overflow (12-20 mm), distillate outlet
    (12-20 mm) and drain (20-25 mm, with a plug). Preferably stainless steel, glass or food-grade
    plastic; with wall sleeves, or two rubber washers and a nut, where they pass through the liner.
  - **Sealant:** 100 % silicone, aquarium- or food-grade, **without fungicide** (sanitary silicones
    contain fungicide and do not work), from a fresh cartridge. At N2 level, glazier's putty (whiting
    and linseed oil) **only on the outside**, with a rubber gasket on the inside. Also 4.5 m of 10 mm
    rubber weatherstrip to seat the glass.
  - **Collection container** with a lid, of glass, lead-free glazed ceramic or food-grade plastic
    (PET, HDPE), 5 L per m² of still.
  - **Feed water:** about 3 times the volume expected to be distilled each day (step 10).
- **Tools:** saw, hammer or screwdriver and galvanized or stainless nails or screws (N2); glass
  cutter if the glass needs cutting (N2-N3); drill or auger for the pipes (N2); level or a bottle
  laid on its side with water (N1); protractor or square and ruler (N1).
  **Technology-level decision:** `nivel: 2`, because the procedure requires flat glass and
  carpentry with nails. In the early phase after a collapse (F0), when salvaged glass and liner are
  abundant, it can be built right away. Plastic sheeting and the solar still pit ("Variants") avoid
  glass, but plastic is not manufactured before N4 either and does not last long. With N0 means (no
  glass or plastic) there is no solar distillation: only fire distillation.
- **Time and people:** 1 day for 2 people per module; 3-7 days of "curing" in the sun before
  drinking the distillate; afterward, about 10 minutes a day.

## How it works

**Evaporating does not require boiling.** Hot water gives off vapor at any temperature, more so the
hotter it is. At 60-70 °C its vapor pressure (20-31 kPa) is 6-10 times that at 25 °C (3.2 kPa, steam
tables), so a black tray in the sun evaporates a great deal without boiling. Boiling hard even
worsens the purity, because it throws droplets of salt water into the distillate (McCluney, 1984).

**Only water and volatiles evaporate.** Salts, metals (lead, arsenic), nitrates, fluoride, clay and
microorganisms stay in the tray (Dvorak and Skipton, 2013). What passes into the distillate is
substances that boil near 100 °C or below: benzene, toluene and other solvents and fuel components,
some pesticides, chloroform and part of the ammonia and hydrogen sulfide (rotten-egg smell). **A
still does not clean water contaminated with gasoline, diesel or solvents.**

**Condensing requires a surface colder than the water.** The glass, in contact with the air, sits a
few degrees below the water; the vapor condenses on its inner face and gives up its heat of
evaporation, which is lost outward.

| Variable | What works well | Why | Outside the range |
|---|---|---|---|
| Water depth | ≈1.5 cm; 1.5-2 cm in practice (VITA, 1985) | Less mass to heat | Beyond about 5 cm, daytime output drops and shifts to the night (VITA); with less than 1 cm the tray dries out in patches and forms a salt crust |
| Glass-to-water distance at the low edge | 5-8 cm (VITA recommends 5-7 cm) | Less opaque wall casting shade, and less air to heat | A 13 cm front wall above the water at 37-40° N casts a winter shadow of 22-25 cm (20-25 % of the tray) |
| Bottom color | Matte black | Absorbs the light that passes through the water | A light-colored bottom reflects it and output drops sharply |
| Insulation | 5 cm of polystyrene or cork, or 8-10 cm of dry straw or sawdust | Heat lost to the ground does not evaporate water | Wet insulation does not insulate |
| Airtightness | Sealed joints | Every vapor leak is lost distillate | Leaks show up as vapor or drops on the outside, and output clearly below the "Calculations" table |
| Glass tilt | 20° to start; more in winter and at high latitudes, less near the equator (Khalifa, 2011) | Drops must slide down against the glass | With clean glass, water slides even with very little slope (VITA: down to 1°); below about 10°, any grease or dirt makes the drops fall back |
| Glass versus plastic | Glass | Water wets the glass and forms a film; on almost all plastics it forms drops that fall | Of 27 large glass stills, 24 were still working after years; of 9 plastic ones, 1 (ITDG) |
| Glass cleanliness | No dust on the outside | Dust blocks light | Output that declines gradually with no other cause |

**Energy balance.** Evaporating 1 kg of water costs about 2.3-2.4 MJ at 40-60 °C (2.26 MJ/kg at
100 °C). A simple box still makes use of 25-30 % of the radiation (ITDG); the rest is lost through
reflection, through the bottom and walls and, above all, with the heat the vapor releases on
condensing. ITDG's estimate:

```
Q = E × G × A / 2.3
Q = liters of distillate per day
E = overall efficiency (0.25-0.30 in a well-built box still)
G = daily global irradiation on a horizontal plane (MJ/m²; 1 kWh/m² = 3.6 MJ/m²)
A = tray surface area viewed from above (m²)
```

A large part of the distillate comes out **late in the afternoon**, with the water still hot and
the air cooling (ITDG): it is collected the next morning.

**When it is not worth it.** Distilling uses more than 10,000 times the energy of pumping the same
liter up 20 m (2,260 kJ/kg versus 0.2 kJ/kg, ITDG). If fresh water is available, even if dirty, it
is far more worthwhile to filter and disinfect it. The still makes sense with brackish or
seawater, with water high in nitrates, arsenic or fluoride, or as a drinking-water reserve.

## Procedure

Single-slope still: the glass runs down from the tall back wall to the low front wall, which faces
the sun (south in the northern hemisphere). All **heights are measured from the surface of the
tray** (the liner over the false bottom).

```
                               glass 4 mm at 20°, 1.15 × 1.11 m
  back wall             ____----‾‾‾‾                  (battens screwed on top)
  46 cm above ____----‾‾‾‾        vapor ↑   drops ↘
  the tray    |                                        ‾‾‾‾----____ front wall
  inlet ⊃     |                                              [trough]| 10 cm above the tray
              |~~~~~~~~~~~~~~ water 1.5-2 cm ~~~~~~~~~~~~~~~~~~~~~~~~| → U-shaped overflow
              |=========== EPDM liner over false bottom =============| → drain (low corner)
              |░░░░░░░░░░ insulation 5 cm (or 10 cm straw) ░░░░░░░░░░|
              |__________________ board bottom ________________________|
  interior clearance between walls: 100 cm × 100 cm; side insulation on the outside
```

1. **Decide whether a still is the solution.** If there is no suspicion of chemical contamination,
   taste the water with the tip of the tongue without swallowing it. If there is, do not taste it:
   measure the conductivity or send it for analysis ([basic water quality analysis](analisis-basico-de-calidad-del-agua.md)).
   Water tastes salty from about 200-300 mg/L of chloride upward (WHO, 2011, ch. 10); seawater
   carries about 35 g/L of salts. Do not use the still as the main treatment if the water smells of
   gasoline, solvent or chemical product. *Success criterion:* the available water is brackish or
   salty, or has a known non-volatile dissolved contaminant; there is no fresh water within a
   reasonable distance; there is sun on more than half the days of the season of use.

2. **Size it.** Calculate the surface with "Calculations", month by month, starting from **2.5-3 L
   per person per day for drinking** (Sphere, 2018; WHO/WEDC, note 9). Cooking requires 3-6 L more
   per person, which should come, at least in part, from non-distilled fresh water. As a reference,
   ITDG calculates 5 L of distillate per person, i.e., 2 m² per person. *Success criterion:* you
   have a number of m² and modules, and know where the water will come from in the months the still
   does not cover (rain, fire).

3. **Choose the site and orient it.** Flat ground, with no shade from 2 hours before to 3 hours
   after solar noon at any time of year, protected from animals and less than 30 m from the feed
   water. The low wall faces south (north in the southern hemisphere), ±15°. Without a compass:
   solar noon is when the shadow of a vertical stick is shortest. North of the Tropic of Cancer,
   that shadow points north; south of the Tropic of Capricorn, south. Between the two tropics, use
   the rising and setting sun or a compass. Level the base (stones, blocks) with a level or a bottle
   laid on its side with water. *Success criterion:* horizontal base (±0.5 cm over 1 m), oriented
   and with no shade at noon.

4. **Build the box.** With bottom insulation **inside** the box and side insulation **outside**:
   - **Bottom:** board 1.05 × 1.05 m. On top, 5 cm of polystyrene or cork (or 8-10 cm of pressed
     straw or sawdust) and, on top of that, the false bottom of 1.00 × 1.00 m, shimmed to give a
     **0.5 cm slope** toward the front corner opposite the distillate outlet (the drain corner). So
     the tray ends up 8 cm above the base of the boards (2 for the bottom + 5 for insulation + 1 for
     the false bottom), or 13 cm with 10 cm of straw.
   - **Front wall:** 10 cm above the tray → **18 cm tall board** (23 cm with straw) × 105 cm.
   - **Back wall:** 10 + 100 × tan 20° = 10 + 36.4 ≈ **46 cm** above the tray → **54 cm board**
     (59 cm with straw) × 105 cm.
   - **Sides:** trapezoids 100 cm long (interior clearance between front and back walls), from
     18 to 54 cm (23 to 59 cm), with a straight top edge.
   - For other slopes, the height of the back wall above the tray is 10 cm + 100 cm × tan(angle):
     15° → 37 cm; 25° → 57 cm; 30° → 68 cm. **Choice:** 20° in general; on the Iberian Peninsula,
     if winter matters most, 30-35° (verificar); below 20° latitude, 10-15° (Khalifa, 2011).
   - Outside the four walls, 5 cm of polystyrene or cork, or 10 cm of straw in sacks or between the
     wall and a board or cane lining, protected from rain.

   *Success criterion:* interior clearance of 100 × 100 cm; when a straightedge is laid across the
   sides, the slope matches the one chosen ±2°; the box does not wobble.

5. **Fit the liner and the bottom pipes.** Spread the 2.0 × 1.8 m EPDM liner over the false bottom
   and pull it up **to the top edge of all the walls**, folding the corners without cutting. Fix it
   at the top with staples or nails that will later be covered by the weatherstrip. Make the
   **drain** in the low corner: a 20-25 mm pipe flush with the bottom, going through the liner with
   a wall sleeve or two rubber washers tightened with a nut (or silicone on both faces), and exiting
   through the front wall, with a plug. *Success criterion:* with the drain plugged, 20 L of water
   (2 cm) does not drop in level over 24 h with the box covered by a plastic sheet; the insulation
   stays dry; when unplugged, the tray empties completely and no puddles deeper than 2-3 mm remain.

6. **Install the trough, the outlet, the overflow and the inlet.**
   - **Trough:** along the inner face of the front wall, with the top edge 1 cm below the glass seat
     (at 9 cm above the tray) and the bottom at about 5-6 cm, i.e., **3-4 cm above the water**.
     Slope of 1-2 cm over 1 m (CTE DB-HS 5 requires a minimum of 0.5 %) toward the outlet side,
     where a pipe goes through the wall to the container. The trough must not touch the liner: salty
     water climbs by capillarity.
   - **Overflow:** on the opposite side, using the same wall-sleeve or washer system. Outside, a
     U-bend: **the outlet mouth of the U sets the level**, and must sit 2 cm above the bottom of the
     tray. The U keeps water inside and stops vapor from escaping.
   - **Inlet:** in the back wall, about 30 cm along, with a plug.

   *Success criterion:* 0.5 L poured at the high end of the trough comes out entirely at the outlet
   in under 1 minute; when filled through the inlet, the water overflows and the level settles at
   2 cm ± 2 mm.

7. **Fit the glass (removable).** Stick the weatherstrip onto the top edges of the four walls, over
   the liner. Wash the glass with water and soap and rinse it: grease stops the water from forming a
   film. Rest it so it overhangs 3 cm at the back and sides and its low edge sits over the outer edge
   of the front wall: on the inside, the condensate reaches the weatherstrip line and drips into the
   trough. Fasten it with battens screwed to the walls, tight only enough to compress the
   weatherstrip (if it presses against the wood, it will break when it expands). Seal with a bead of
   silicone **only on the outside**, between the glass and the batten or wall, easy to cut to
   disassemble. *Success criterion:* the glass does not move when pushed and no gaps are visible
   against the light.

8. **Test for airtightness and "cure" it.** Fill up to the overflow and leave it in the sun for
   **3-7 days**, discarding all the distillate. McCluney (1984) recommends several days so that the
   volatiles from the silicone, the liner and the wood evaporate off. Between 12 h and 14 h on the
   first day, check the perimeter. *Success criterion:* the glass fogs evenly on the inside within
   1-2 h of sun; no visible vapor escapes and no drops appear outside at the joints; the last day's
   distillate has no smell in a clean glass.

9. **Pretreat the feed water.** If it is turbid (text cannot be read through 10 cm of water in a
   glass), let it settle 12-24 h and use the clear part, or coagulate or filter it. Disinfecting it
   is not necessary. *Success criterion:* clear water enters, with no suspended solids.

10. **Daily handling**, each morning before 9 h (solar time):
    1. Cover the full container, swap it for a clean one and **record the liters**.
    2. Pour in through the inlet about **3 times the previous day's output** (SolAqua: if it
       distilled 2 L, pour in 6 L). The surplus exits through the overflow, carrying off the brine.
       Use cold water first thing; pouring it in at midday cuts the afternoon's output.
    3. Plug the inlet. Wipe the dust off the outside of the glass with water or a damp cloth, once a
       week or daily if there is a lot of dust.

    The brine can be used for washing, but do not always pour it on the same garden plot: it
    salinizes the soil. *Success criterion:* the level returns to 2 cm; no white crust forms at the
    edges of the tray.

11. **Clean the salts and the trough**, every month or when crusts, sludge or algae appear, first
    thing in the morning with the glass cold: open the drain and empty it. Cut the outer silicone
    bead with a utility knife, unscrew the battens and lift the glass between two people, resting it
    on edge on wood. Scrape the crust with a wooden spatula without scratching the liner and rinse
    twice through the drain. Lime scale (hard, fizzes with vinegar) softens with vinegar (Dvorak and
    Skipton, 2013); rinse well. Clean the trough with boiling water and rinse it with distillate.
    Wash the inside of the glass with water only. Put it back and reseal on the outside, and discard
    the first 24 h of distillate. *Success criterion:* black bottom with no white stains and output
    equal to or greater than before cleaning.

12. **Collect and store the distillate.** The outlet pipe enters through a hole in a plug fitted in
    the container, so dust, insects and hands cannot get in. Containers of glass, lead-free glazed
    ceramic or food-grade plastic; **never** copper, brass, galvanized iron or ones with lead solder:
    distilled water is aggressive and dissolves them (Kozisek, 2005). Serve by pouring or with a
    tap, without dipping cups in. The safest thing is to drink it the same day; if it is going to be
    stored, chlorinate it like rainwater (see [chemical disinfection with chlorine](desinfeccion-quimica-del-agua-con-cloro-y-yodo.md)).
    *Success criterion:* covered container, dated and with no contact with unsuitable metals.

13. **Remineralize it if it is the regular drink.** Distillate has no calcium or magnesium, tastes
    "of nothing" and is corrosive. Kozisek (2005, published by WHO) gives as minimum recommended
    values 20 mg/L of calcium, 10 mg/L of magnesium and 100 mg/L of dissolved solids.
    - **Blending (the best option):** 1 part **hard**, disinfected fresh water (from a well or
      spring: the kind that forms lime scale when boiled) to 1-4 parts distillate. Rainwater does
      not work for this either: it has no minerals (Kozisek, 2005).
    - **Limestone or marble bed:** washed pieces of 1-3 cm; pour the distillate over them and leave
      it 12-24 h, or let it drip slowly through a tube packed with pieces (McCluney, 1984).
      Contributes **up to about 20 mg/L of calcium in the best case** (calcite in equilibrium with
      the CO₂ in the air at 25 °C; Stumm and Morgan, 1996), typically 5-15 mg/L in 12-24 h, and
      **no magnesium**, so on its own it does not reach the minimums (Kozisek, 2005). If there is
      dolomite present (looks like limestone but barely fizzes with cold vinegar), mix pieces of
      both.
    - **Salt, only for flavor:** 0.1-0.2 g/L (a pinch) contributes no calcium or magnesium. Do not
      add wood ash: it is very alkaline.

    *Success criterion:* the water does not taste "flat", and it takes a bit more soap to work up a
    lather than with the plain distillate.

## Verification

- **Output:** measure the liters each morning and compare them with the "Calculations" table for
  the weather that day. Less than half the expected amount on a clear day points to leaks, shade,
  water that is too deep or a dirty glass.
- **Salt, taste:** the distillate should not taste at all salty compared with the feed water.
- **Salt, residue:** evaporate 100 mL of distillate in the sun on a clean glass or stainless-steel
  plate and 100 mL of the feed water on another. The distillate barely leaves a ring; brackish water
  leaves a white crust. If the distillate leaves a crust, water from the tray is getting into the
  trough.
- **Conductivity (meter, N4):** conductivity meters read in µS/cm; dissolved solids (mg/L) ≈ 0.5-0.7 ×
  conductivity. A well-sealed still should give on the order of 20-40 µS/cm or less (≈10-20 mg/L)
  (verificar). A high value points to contamination from the tray, but can also come from the
  container: measure directly at the pipe outlet too.
- **Microbiology (if available):** *E. coli*, 0 CFU in 100 mL; H₂S strip, no blackening at
  24-48 h. Hanson et al. (2004) found that a single-tray still eliminated bacteria **as long as
  cross-contamination was avoided**. If it comes back positive, look for where the raw water is
  touched: hands, funnel, trough or container.
- **Smell:** none. A vinegar smell means uncured silicone (keep curing); a plastic or paint smell
  means unsuitable material; a fuel or solvent smell means the feed water carries volatiles and
  **must not** be used.

## Common mistakes

| Symptom | Likely cause | Solution |
|---------|----------------|----------|
| The glass fogs up but very little distillate comes out | Drops fall back into the tray: greasy glass, too little slope or a plastic cover | Wash the glass with soap; raise the slope to 20° or more; replace plastic with glass |
| Output is half the expected amount on a clear day | Vapor leaks, wet insulation, shade from the walls or from obstacles | Reseal (look for vapor at noon); dry out or replace the insulation; lower the front wall; check for shade |
| The distillate tastes salty | Splashes into the trough, capillarity through the liner up to the trough, tray too full | Separate the trough from the liner; check the U mouth is at 2 cm |
| White crust in the tray | Too little purging; dry patches | Purge daily with 3 times the output; do not let it drop below 1.5 cm |
| Algae or slime in the tray | Algae and bacteria in mildly salty, warm water | Purge more; clean every month; plug the inlet |
| Vinegar or plastic smell | Acetic or sanitary silicone, uncured liner | Discard the distillate for 3-7 more days; if the smell does not go away, change the material |
| The glass cracks | Pressed against the wood, cold water on hot glass, hail | Rest it on weatherstrip with 3-5 mm of play; do not wet the hot glass; mesh against hail |
| Output drops month by month | Dust, scale, wood warping and opening joints | Clean the glass every week; monthly cleaning; reseal every 6 months |
| Positive for *E. coli* | Contamination in the trough, the container or while serving | Boiling water for the trough; closed container with a tap; clean hands; chlorinate what is stored |

## Safety

- **Materials that poison the distillate** (the most serious risk, because it cannot be seen). Hot,
  mineral-free distilled water dissolves metals and picks up volatiles from the apparatus itself.
  - **Lead:** no old paint, lead solder, brass or lead-glazed ceramic in contact with the distillate.
    Kozisek (2005) cites infants poisoned by lead that poorly mineralized water had dissolved from
    brass fittings and tank solder.
  - **Treated wood (CCA, creosote):** not even in the box.
  - **Plastics, silicones and paints:** only ones rated for drinking water (McCluney, 1984) and
    always cured for 3-7 days.
- **Water with volatiles.** If the source may contain fuel, solvents or pesticides (near gas
  stations, workshops, industry, sprayed fields), the distillate can carry them without smelling of
  anything. Do not use it for drinking. An activated-carbon filter after the still reduces volatiles
  (Dvorak and Skipton, 2013; [ceramic and carbon filters](filtros-de-ceramica-y-carbon.md)), but
  without analysis there is no guarantee.
- **Recontamination.** The distillate comes out free of microbes but **without a chlorine
  residual**: if it gets dirty, bacteria can regrow, especially in warm water (Kozisek, 2005). If
  diarrhea occurs in someone who drinks it, treat it as unsafe water: boil until the fault is found,
  and rehydrate.
- **Glass:** about 13 kg per module, and it cuts. Move it between two people, with leather gloves
  and by the edge. If it breaks, pick up all the pieces, **discard the tray's water and the day's
  distillate**, empty through the drain and check the trough.
- **Burns:** in summer, at midday, the water in the tray can reach 60-75 °C (verificar) and the
  vapor comes out hot when the glass is lifted. Only open and empty it first thing in the morning.
- **Frost:** frozen water can break the trough, the pipes or a rigid tray. Empty through the drain
  if nighttime frost below −2 °C is expected.
- **Dehydration.** In bad weather output collapses. Drinking sea or brackish water without
  distilling it **worsens** dehydration and can kill. Keep a reserve for 2-3 cloudy days and a
  backup plan (fire, rain).
- **Demineralized water.** Drinking distillate now and then is not dangerous for an adult with a
  normal diet. As the sole water source for months, remineralize it (step 13). **Do not prepare
  infant formula or drinks for infants with unremineralized distillate**: Kozisek (2005, pp. 152 and
  154) records cases of hyponatremia and seizures. Cooking with distillate strips out up to 60 % of
  the calcium and magnesium from food (Kozisek, 2005): for cooking, hard water or remineralized
  distillate is better.
- **Brine:** pouring it always in the same spot salinizes the ground. Spread it around or take it
  to the sea or a brackish watercourse.

## Variants

### With fewer resources

- **Box with plastic sheeting (N0-N1 with salvaged sheeting).** The same design with taut
  transparent plastic, for example 150-200 µm greenhouse polyethylene, instead of glass. It needs
  more slope, **30-45°** (verificar), so drops do not fall back; the sheeting very taut, with no
  sagging; and a batten forming the drip edge over the trough. Yields clearly less than glass, and
  ordinary polyethylene turns opaque and brittle in the sun (ITDG: only 1 of 9 large plastic stills
  was still working). It is a temporary solution until glass can be obtained.
- **Solar still pit for survival (needs plastic sheeting and a container).** Dig a bowl-shaped pit
  **1 m in diameter and 0.5-0.6 m deep** (≈0.15-0.26 m³ of soil). In the center, place a 1-2 L
  container and, if available, a tube for drinking without dismantling it. Cover with a transparent
  sheet 1.5 × 1.5 m held down with soil and stones, and place a 0.2-0.5 kg stone in the center so
  the sheet forms a cone with its point 5-8 cm above the container. Jackson and van Bavel (1965)
  obtained **at most 1.5 L/day**, with moist soil or cut plants inside. Real limitations:
  - **In dry soil it can produce less water than is sweated out building it.** Only build it if the
    soil is moist to the touch at 20-30 cm depth or if there are green plants to cut, and dig at
    dawn or dusk. In dry soil, output is on the order of 0.1-0.6 L/day (verificar) and drops each
    day: it has to be moved every 2-3 days.
  - Brackish water, seawater or urine can be poured **into the soil of the pit, never into the
    container**. Some ammonia from the urine passes into the distillate (urea breaks down into
    volatile ammonia): use it only in an emergency.
  - Only non-poisonous plants; never plants with milky latex.
  - It is for a few days' emergency, not a supply.
- **Without glass or plastic (N0-N1): fire distillation only.** Catching the vapor of a boiling pot
  with a cloth and wringing it out gives very small amounts, and the cloth picks up salt from
  splashes; an inverted lid is better.

### Fire distillation (N1 with ceramic, N2 with metal)

For when there is no sun, to get through winter, or in emergencies with seawater. **It uses a lot
of firewood:** 1.1-1.7 kg per liter over an open fire ("Calculations").

**a) Pot with inverted lid (N1-N2).** Pot of 10-20 L with a domed or conical lid, of metal or clay.
1. Fill the pot up to 1/3 of its height, at most. In the center, on three clean stones, place a
   bowl with its rim above the water and 3-5 cm below the lowest point of the lid. *Success
   criterion:* the bowl is stable and splashes do not reach it.
2. Place the lid **inverted**, with its lowest point over the bowl, and fill it on top with cold
   water. *Success criterion:* the lid does not close tightly (vapor can escape at the edge).
3. **Moderate** fire, with a gentle boil: hard boiling splashes salt into the bowl. Change the water
   in the lid every 10-20 minutes, as soon as it is lukewarm. *Success criterion:* drops fall
   continuously into the bowl. Calculated by balance: 2 L of water in the lid warmed 30 °C in
   15 min absorb ≈1 MJ/h, so at most ≈0.3-0.4 L/h condense.
4. Stop when 1/4 of the starting water remains, so the salt does not burn or splash. Remove the lid
   with gloves, turning your face away.

**b) Pot with cooling tube (N2).** Fitted lid with **a single open outlet tube** 1-2 m long,
preferably of stainless steel or glass. If it is copper, it must have no lead solder, must not be
used to store the distillate, and must be rinsed after each use. The tube passes through a bucket
of **at least 10-20 L** of cold water and ends open over the container.
- Change the bucket's water, or add cold water, as soon as it is lukewarm (every 15-20 minutes).
  *Success criterion:* the distillate comes out lukewarm, not hot. If visible steam comes out of the
  end, lower the fire: there is a scald risk and water is being lost.
- Output limited by the fire: with ≈1 kW useful (3.6 MJ/h ÷ 2.6 MJ/L), at most ≈1.4 L/h comes out.
- **Never close the system** or fit valves or plugs on the tube: if it gets blocked, the pot becomes
  a pressure boiler and can burst or shoot out 100 °C steam. Check every 10 minutes that distillate
  is coming out.
- **Discard the first 0.5 L of each batch**: that is where volatiles concentrate (Dvorak and
  Skipton, 2013).
- No car radiators, brass or tin-and-lead solder.
- Smoke: fire in the open air or under an open porch. Carbon monoxide from a fire in an enclosed
  space kills without warning.

### With more resources

- **Double slope (N2):** two glass panes in an inverted V with a trough on each side; useful at low
  latitudes, where the sun passes overhead.
- **Interior reflector (N2-N3):** aluminum sheet or mirror on the inner face of the back wall,
  which sends back to the tray the light that hits that wall; the gain is greater in winter, with a
  low sun.
- **Wick still (N2-N3):** inclined black cloth down which a thread of salt water trickles. It is
  more efficient than the tray (ITDG), but requires a regulated flow and a resistant fabric.
- **Multiple effect (N3):** stacked trays in which the condensation heat of one evaporates the one
  above it; efficiency of 35 % or more, at greater cost and complexity. ITDG advises starting with
  single-tray units.
- **Rainwater collection on the glass:** an **outer** trough at the low edge collects 1 L per m² and
  mm of rain (ITDG). Discard the first 1-2 mm of each rainfall, which washes off the dust, and
  chlorinate the collected water.
- **Waste heat** from an engine or from a refrigerator's condenser (ITDG).
- **Reverse osmosis (N4):** compares with solar distillation from 1 m³/day upward (ITDG).

### Scaling up

| Scale | Water for drinking | Approximate surface (2 L/m²·day) | Comment |
|---|---|---|---|
| Family (4-6 people) | 10-18 L/day | 5-9 m² | Modules of 1-2 m²; 13 kg of glass per m² of tray; 10 min/day |
| Village (200 people) | 600 L/day | ≈300 m² | Rows of trays with 0.6 m aisles; shared closed and chlorinated tank; 1-2 dedicated people; concrete or waterproof mortar trays (ITDG) |
| City (100,000 people) | 300 m³/day | ≈150,000 m² | Not viable: from 1 m³/day, reverse osmosis or electrodialysis are considered, and from 200 m³/day, flash evaporation or vapor compression (ITDG). The city should be located next to fresh water |

Precedent: the Las Salinas plant (Chile, 1872) had about 4,500 m² in 64 glass compartments and gave
about 22,000 L/day when new. Most of the water was for the draft animals of a nitrate operation, and
it ran for several decades (Delyannis, 2003).

## Calculations

### Output by irradiation

With Q = E × G × A / 2.3 and A = 1 m². Efficiency falls with low radiation, because losses weigh
more; the E values in the extreme rows are estimates (verificar). 0.30 is taken as a realistic
summer ceiling.

| Typical day | G (MJ/m²·day) | G (kWh/m²·day) | E | Distillate (L/m²·day) |
|---|---|---|---|---|
| Winter, or cloudy | 8 | 2.2 | 0.15-0.20 | 0.5-0.7 |
| Autumn or early spring at mid-latitude | 13 | 3.6 | 0.20-0.25 | 1.1-1.4 |
| Annual average, sunny zone (ITDG) | 18 | 5.0 | 0.25-0.30 | 2.0-2.3 |
| Clear summer | 24 | 6.7 | 0.25-0.30 | 2.6-3.1 |

On the Iberian Peninsula, average daily irradiation ranges from about 5-9 MJ/m² in December to
21-28 MJ/m² in July, depending on location (AEMET, 2012) (verificar). **Always calculate with the
average monthly G for your location** (PVGIS or the AEMET Solar Radiation Atlas), not the annual
average.

### Worked example: family of 4 on the Mediterranean coast

```
Demand for drinking: 4 people × 3 L/day = 12 L/day
(cooking: 12-24 L/day more, from non-distilled fresh water if available)

Proposed modules: 6 m²
  July       (G ≈ 24, 2.6-3.1 L/m²):  16-19 L/day  → surplus
  May/Sept.  (G ≈ 20-22, ≈2.2 L/m²):  ≈13 L/day    → just enough
  March      (G ≈ 16, ≈1.8 L/m²):     ≈11 L/day    → slightly short
  October    (G ≈ 13, 1.1-1.4 L/m²):  7-8 L/day    → short
  December   (G ≈ 8,  0.5-0.7 L/m²):  3-4 L/day    → very short

Conclusion: 6 m² cover demand from April/May to September.
Covering December would require 12 / 0.5 = 24 m² → not viable; the rest comes from:
  - rain collected on the 6 m² of glass: 1 mm = 6 L; a storm of 20 mm = 120 L
    (after discarding the first 1-2 mm, ≈110 L: 9 days of water);
  - fire distillation for whatever is missing.

Annual output (summed month by month): ≈0.75 m³/m² × 6 m² ≈ 4,500 L,
against 4,380 L of demand (12 × 365). The summer surplus does not make up for winter
unless it is stored and chlorinated.

Feed water in July: 3 × 18 L ≈ 54 L/day (about 36 L come out as brine).
Glass: 6 × 13 kg ≈ 78 kg.
```

ITDG's "1 m³ per m² per year" corresponds to an average of 18 MJ/m²·day (tropics or very sunny
zones); at mid-latitudes the figure is lower.

### Firewood for fire distillation

```
Energy to take 1 L of water from 20 °C to vapor:
  heating: 4.19 kJ/(kg·°C) × 80 °C = 335 kJ
  evaporating: 2,257 kJ
  total ≈ 2.59 MJ/L

Open three-stone fire, efficiency 10-15 % (verificar; same as for boiling):
  2.59 / 0.15 = 17.3 MJ  →  17.3 / 15 MJ/kg = 1.15 kg of dry firewood per liter
  2.59 / 0.10 = 25.9 MJ  →  25.9 / 15 MJ/kg = 1.73 kg of dry firewood per liter

Covering the family's October shortfall (≈4-5 L/day): 5-9 kg of firewood a day.
Covering the December shortfall (≈8-9 L/day): 10-15 kg of firewood a day.
```

That is why fire distillation is a backup, not a supply: boiling fresh water needs about 7-8 times
less firewood than distilling it (see [boiling water](hervido-del-agua.md)).

## Sources

- ITDG (hoy Practical Action). *Solar Distillation*. Technical Brief, c. 2002. Fórmula de producción,
  eficiencias, experiencia con vidrio y plástico, escalas.
- McCracken, H.; Gordes, J. *Understanding Solar Stills*. VITA, 1985, parte III. Profundidad,
  separación vidrio-agua, pendiente mínima.
- McCluney, W.R. *Solar Distillation of Water*. FSEC-EN-3-80, 1984. No hace falta hervir, horneado,
  materiales, remineralización con mármol.
- Khalifa, A.J.N. (2011). *Energy Conversion and Management*, 52(1), 431-436. Inclinación según
  estación y latitud.
- Hanson, A. et al. (2004). *Solar Energy*, 76(5), 635-645. Calidad del destilado.
- SolAqua / El Paso Solar Energy Association. *Solar Still Basics*, s. f. Purga diaria.
- Delyannis, E. (2003). *Solar Energy*, 75(5), 357-366. Las Salinas.
- Jackson, R.D.; van Bavel, C.H.M. (1965). *Science*, 149(3690), 1377-1379. Pozo solar.
- Dvorak, B.I.; Skipton, S.O. *Drinking Water Treatment: Distillation*. UNL G1493, 2013.
- Kozisek, F. En OMS, *Nutrients in Drinking Water*, 2005, pp. 148-163. Agua desmineralizada.
- OMS. *Safe Drinking-water from Desalination*, 2011; *Guidelines for Drinking-water Quality*, 4.ª ed.,
  2011, cap. 10; OMS/WEDC, nota técnica n.º 9.
- Sphere Association. *The Sphere Handbook*, 4.ª ed., 2018.
- Stumm, W.; Morgan, J.J. *Aquatic Chemistry*, 3.ª ed., 1996.
- Catálogo de Elementos Constructivos del CTE; CTE DB-HS 5 (pendiente de canalones).
- AEMET. *Atlas de radiación solar en España*, 2012; Comisión Europea, JRC, PVGIS.
- Malik, M.A.S. et al. *Solar Distillation*. Pergamon, 1982 (no consultado directamente).
</content>