Solar disinfection (SODIS)

Summary

Solar disinfection (SODIS, short for solar water disinfection) consists of filling clear plastic PET bottles of 2 L or less with clear water and leaving them lying in direct sun for a full day (at least 6 hours including midday), or two consecutive days if the sky is more than half covered. UV-A radiation and the heating of the water inactivate most of the bacteria that cause diarrhea and cholera, and some of the viruses and protozoa. It uses no fuel or chemical products and works with recovered bottles, which makes it the starting option when there is no firewood or chlorine. Expected, measurable result: with water of turbidity below 30 NTU and good sun, a reduction of fecal bacteria of around 3 logarithms (99.9 %) in 1 day, at a rate of 2 L per bottle exposed (Meierhofer and Wegelin, 2002). It does not sterilize, does not remove chemicals or turbidity, is weak against Cryptosporidium and many viruses, and does not work with continuous rain or for baby-bottle water.

Prerequisites

• Knowledge: none essential. It helps to know boiling water, which is the mandatory alternative for infants, sick people and rainy days, and filtration (slow sand filter) as a pretreatment for turbid water.
• Materials:
  • Clear, colorless PET bottles (or with a slight blue tint), 0.3 to 2 L, with a screw cap that seals tight. Guideline amount: 4 bottles of 2 L per person (2 in the sun and 2 ready to drink), following Eawag's guidance (Meierhofer and Wegelin, 2002); with frequent cloudy skies more are needed, or boiling as a supplement (see “Calculations”). How to identify PET: see step 1.
  • Soap or ash and a brush or a handful of clean sand for the first wash.
  • A support to lay the bottles flat in the sun: corrugated metal sheet (the best), a roof, aluminum sheet or any dark or reflective surface free of shadows.
  • A newspaper or paper with large print (headlines) for the turbidity test.
  • A tightly woven cotton cloth or a sand filter if the water is turbid.
  • Optional: thermometer, WAPI indicator and an E. coli test kit (see “Verification”).
• Tools: none manufactured. Technology-level decision: set at nivel: 0 because the main procedure does not require manufacturing anything: sun, water and recovered PET bottles suffice, which will be abundant in the first years after a collapse (phase F0). PET cannot be manufactured until N4, so the method is sustainable only for as long as recovered bottles last; its manufacturable substitute is clear glass bottles (N2), with the caveats explained in “Variants”. A purely N0 container (gourd, hide, ceramic) is opaque to UV-A and does not work for SODIS; with it, only boiling or pasteurizing is possible.
• Time and people: one person. It requires daily work filling, placing, retrieving and marking bottles, plus 6 h to 2 days of waiting without attention; the 2016 manual only describes it as a “relatively high daily work demand” and gives no minutes per bottle: time it during the first week to plan (see “Calculations”). Eawag's recommendation: that at least 2 people in each household know how to apply it and one is responsible for it every day (Meierhofer and Wegelin, 2002).

How it works

Two effects of the sun and their synergy

1. UV-A radiation (320-400 nm) and visible violet. The UV-B that reaches the ground kills microorganisms by directly damaging their DNA, but PET absorbs almost all of it. What does pass through the bottle is UV-A, which generates reactive oxygen species in the water (singlet oxygen, hydrogen peroxide, hydroxyl radical) from dissolved oxygen and photosensitive substances; these damage the membranes, enzymes and DNA of pathogens (Luzi et al., 2016, chap. 2.1). This is why disinfection depends on light penetrating (clear water, shallow depth) and on dissolved oxygen being present.
2. Heat (infrared). The water in the bottle heats up. On its own, heat inactivates pathogens in 60 minutes at temperatures between 45 °C (Vibrio cholerae) and 63 °C (enteroviruses) (Luzi et al., 2016, chap. 2.2.7).
3. Synergy above 45-50 °C. Together, the two effects add up to much more than separately. At 30 °C a dose of about 555 Wh/m² is needed (350-450 nm, equivalent to about 5-6 h of midday summer sun at mid latitudes: ~5 h according to Wegelin et al., 1994 and table 5 of Luzi et al., 2016; ~6 h according to Meierhofer and Wegelin, 2002, p. 12) to reduce fecal coliforms by 3 logarithms; at 50 °C about 140 Wh/m² is enough, on the order of 1 h of exposure (Wegelin et al., 1994; Meierhofer and Wegelin, 2002). The 2016 manual summarizes this by saying that at 50 °C the dose or time needed is reduced by up to two thirds (Luzi et al., 2016, chap. 2.2.7). Below about 45 °C the temperature dependence is weak.

Practical consequence: the synergy explains why SODIS works better in warm climates, but the 2016 manual does not recommend shortening the exposure time: water temperature changes throughout the day, a single reading does not prove how long it stayed above 50 °C, and the 6 h or 2 day rule already carries a built-in safety margin (Luzi et al., 2016, chap. 2.2.7 and box 6).

Variables that matter, and their margin

Suitable latitudes

• Most favorable region: between 15° and 35° latitude (north and south): semi-arid areas with plenty of direct radiation and little cloud.

• Second most favorable: between the equator and 15° latitude: plenty of radiation, but more cloud cover and diffuse radiation (Meierhofer and Wegelin, 2002, chap. 2.3).

• Beyond 30-35° (all of mainland Spain, between 36° and 44° N; the Canary Islands, around 28° N, do fall within the favorable band): Eawag warns that its exposure guidelines do not apply beyond 30° N/S (Luzi et al., 2016, box 6). There is no official guideline for these latitudes. At altitude (the Bolivian Altiplano, the Chinese plateau) SODIS has worked well despite the cold because UV is intense.

• Criterion proposed by this project for high latitudes (extrapolation, not an Eawag guideline): table 5 of the 2016 manual estimates a threshold dose of ≈ 5.7 kWh/m² of full-spectrum radiation accumulated during exposure. From this, the following monthly guideline is derived, according to the mean daily global irradiation on a horizontal surface (monthly values in AEMET's solar radiation atlas or in PVGIS):

  Mean daily global irradiation for the month	Guideline
  > ≈ 6 kWh/m²	1 full day on clear days; 2 days if there is more than 50 % cloud
  ≈ 3 to 6 kWh/m²	2 consecutive full days as the rule, only with mostly clear days
  < ≈ 3 kWh/m²	Do not use SODIS: boil or pasteurize

  The thresholds of 3 and 6 kWh/m² are the project's own reasoning without field validation (verificar). They are not fully conservative: in winter the UV-A fraction of global radiation is even lower than in summer and the water barely heats up, so in months close to the lower limit boiling is advisable, and before adopting SODIS at these latitudes it should be checked with an E. coli test (“Verification”).

What it eliminates and what it does not

Typical reductions after 6 h in a PET bottle in tropical countries, with water below 40 °C (Luzi et al., 2016, tables 1-4). LRV = log reduction value (1 log = 90 %, 2 log = 99 %, 3 log = 99.9 %).

WHO (GDWQ 4th ed., 2011, table 7.8) assigns solar disinfection reference reductions of 3 log for bacteria, 2 for viruses and 2 for protozoa, with maximums above 5, 4 and 4 log. Those figures cover all protozoa together: Cryptosporidium falls well below (0.3 to > 0.4 log in 6 h, table above).

SODIS does not remove: chemical products (pesticides, fuels, solvents, nitrates, arsenic, fluoride, heavy metals), turbidity, color, salt or bad tastes. Nor does it sterilize: harmless environmental algae and bacteria can survive and even grow in the bottle (Meierhofer and Wegelin, 2002).

Regrowth

Some studies have observed E. coli or Salmonella multiplying again in the 18-24 h following insufficient exposure; others see no regrowth. To be safe, Eawag recommends keeping the water in the same bottle, in a cool place, and drinking it within a few hours and at most 1-2 days (Luzi et al., 2016, chaps. 2.2.8 and 2.4).

Procedure

1. Choose the bottles. Clear, colorless or with a very slight blue tint, 2 L or less, with no deep scratches, no cloudiness or yellowing, with a cap that seals. Identify PET: it usually carries the recycling symbol “1” or “PET”. Discard PVC (bluish sheen at the cut edge; symbol “3”; if in doubt, the flame test from “Safety”), polycarbonate (symbol “7”, rigid water-cooler jugs or old baby bottles: can release bisphenol A) and, in general, any plastic other than PET that is not certified for drinking water (Luzi et al., 2016, chap. 2.2.4); also colored bottles, ones that held fuel, pesticide, solvent or industrial oil, and jugs over 2 L. Success criterion: filled and turned upside down for 1 minute it does not drip; looking through it, text reads without distortion or whitish haze.

2. Wash the bottle the first time. Remove all paper or plastic labels (they cast shade). Wash the inside with water, soap (or ash) and a handful of clean sand, shaking for 30 s; scrub the neck thread and the inside of the cap with a small stick or brush, where the most dirt collects. Rinse 2 times. Success criterion: no smell, no film to the touch inside the cap and no visible residue on the thread.

3. Check the water's turbidity. Fill the bottle to the top, stand it in the shade on the headline of an ordinary newspaper and look at it vertically from the mouth of the full bottle, through the whole column of water (Luzi et al., 2016, chap. 2.2.5). Success criterion: the letters can be read → turbidity below 30 NTU, suitable. With Eawag's original template (15 cm SODIS logo), if the fine rays of the drawing can also be made out, turbidity is below 20 NTU (Meierhofer and Wegelin, 2002, p. 15). If they cannot be read, pretreat the water before continuing:

   • Let it settle for 1 day in a covered container and pour off the clear part without stirring up the sediment; or
   • filter it through a tightly woven cotton cloth folded several times, or through a sand filter; or
   • clarify it by coagulation and flocculation (aluminum sulfate or crushed Moringa oleifera seeds; article planned for this folder). Repeat the test. If it still cannot be read after pretreating, do not use SODIS: boil or pasteurize.

4. Aerate and fill. Fill the bottle to 3/4, close it and shake it hard for about 20 s; then fill it to the top and close it tightly. The 2002 guide always recommended this; the 2016 one considers it unnecessary because the water already gets oxygenated when poured. It is essential with stagnant, oxygen-poor water (ponds, cisterns, wells): there, shake it or pour it several times between two containers from ~30 cm. Success criterion: the bottle is filled to the neck, with at most one small bubble, the cap tight and no dripping when tilted.

5. Set up the exposure site. Choose a spot that gets direct sun from early morning to mid-afternoon with no shade at all throughout the day (trees, eaves, walls, the back of a chair). Place a corrugated metal or aluminum sheet on it, or use a metal roof or a dark ground surface. Prefer a waist-height support over climbing onto roofs. Solar time without a watch: drive a vertical stick into level ground; solar noon is the moment its shadow is shortest. “9 h” and “15 h” solar time are roughly 3 h before and after that moment (a test day with any kind of watch, or counting the sun's passage, serves to calibrate it). Success criterion: checked at 9 h, 12 h and 15 h solar time, the spot is still in the sun.

6. Expose. Lay the bottles flat (horizontal or slightly tilted), with the long side facing the sun, without stacking them or shading one another, in the grooves of the sheet so they do not roll. Start as early as possible in the morning. Not behind a window pane (blocks UV-A). Do not move or shake them during exposure. Success criterion: all the bottles receive sun along their whole length and are lukewarm or hot to the touch at midday.

7. Time it according to the sky. Estimate cloud cover in mid-morning and at midday:

   Sky during exposure	Minimum time
   Clear or with clouds covering less than 50 %	1 full day: at least 6 h including midday (e.g., from 9 h to 15 h solar time)
   Clouds covering more than 50 % or fully overcast sky	2 consecutive full days, without opening the bottle in between
   Continuous rain or very dense fog	Do not use SODIS. Drink stored SODIS water, collect rainwater or boil

   Never shorten the time, even if it is very hot or the water is very hot (Luzi et al., 2016, chap. 2.2.7 and box 6). At latitudes above 30°, also apply the monthly criterion from “Suitable latitudes”. Success criterion: the time in the table has been met and the bottle was never in the shade at any point.

8. Retrieve and mark. Collect the bottles at the end of the afternoon without opening them and place them in a cool, shaded spot, separate from the untreated water bottles. Mark the treated ones (a different-colored cap, a notch, or separate “raw” and “ready” shelves). Success criterion: anyone in the household can tell a treated bottle from an untreated one without hesitation.

9. Drink. Drink straight from the bottle or pour into a clean glass, without transferring it to open jugs or containers. Consume it the same day or at most within 1-2 days. If another container has to be used, it should be clean, narrow-mouthed, covered, with a clean tap or ladle. Success criterion: the treated bottle is not opened until the moment of drinking and no one puts fingers or utensils into it.

10. Maintain the bottles. Wash the cap and thread every week: this is a routine recommended by this project, not an Eawag guideline; the source only documents that a dirty thread recontaminates the treated water (Sacabamba, Bolivia; Meierhofer and Wegelin, 2002, p. 27). Each month, hold every bottle up to the light and replace any that are scratched, cloudy, yellowed, deformed or have a leaking cap: scratches and plastic aging reduce UV-A transmission. Success criterion: every bottle in use passes the step 1 inspection.

Verification

• There is no visible sign that the water is disinfected. Treated water looks the same as raw water. The guarantee comes from having followed the process (turbidity, bottle, time, sun with no shade). Keeping a simple log (date, cloud cover, time put out and time retrieved) helps catch failures.
• Temperature (if a thermometer is available): measure, in the early afternoon, the water of a control bottle exposed alongside the rest. If it exceeds 50 °C, the synergy is working with a good margin, but that is no reason to shorten the time (step 7). In warm climates it usually stays around 40-45 °C (Luzi et al., 2016), which is normal and does not invalidate the 6 h.
• WAPI indicator (a tube with wax that melts at pasteurization temperature, see boiling water): if the wax melts inside the bottle or container, the water reached pasteurization temperature regardless of the UV. It is a test of sufficiency, not of insufficiency: if it does not melt, that does not mean SODIS has failed.
• Microbiological testing (the only one that confirms the result): count of E. coli or fecal coliforms in raw and treated water with a field kit (membrane filtration and incubation at 44 °C for 16-18 h; Meierhofer and Wegelin, 2002). Desirable result: 0 E. coli per 100 ml in the treated water; in the WHO risk classification cited by Eawag, 1-10 per 100 ml is low risk, 10-100 intermediate, 100-1000 high and over 1000 very high. Do this when introducing the method, when changing water source or bottle type, and each season. Not valid as indicators: total coliforms or total bacteria count, because harmless environmental bacteria and algae grow in the bottle in the sun. The absence of E. coli also does not guarantee the absence of Cryptosporidium or viruses, which are more resistant. Field methods without a laboratory will be covered in the planned article “Basic water quality testing”.
• Validation test for new bottles (glass or doubtful plastic): requires an E. coli kit. Expose in parallel, with the same water, a reference PET bottle and the candidate bottle, and compare E. coli after the normal time. The candidate is only adopted if it gives the same result. Without a kit there is no possible validation: the candidate bottle is not adopted for SODIS.

Common mistakes

Safety

• What can kill: a false sense of security. SODIS greatly reduces the risk, but does not eliminate it. Cholera, dysentery and dehydration from diarrhea kill quickly, especially young children. If someone gets diarrhea, start oral rehydration immediately (if there are no ORS sachets, WHO's homemade recipe: 1 L of clean water—boiled and cooled—, 6 level teaspoons of sugar and 1/2 level teaspoon of salt, stirred until dissolved; do not add more salt, prepare fresh daily and discard after 24 h; Rehydration Project), switch the whole household to boiling water, and seek medical care if there is blood in the stool, high fever, vomiting that prevents drinking, or signs of dehydration (dry mouth, little urine, sunken eyes, lethargy).
• Groups that should NOT drink SODIS water (Meierhofer and Wegelin, 2002, chap. 3.4): use only boiled water for
  • infant formula, baby porridges and drinking water for infants and children under ~18 months;
  • people who are severely ill or malnourished;
  • immunocompromised people (advanced HIV/AIDS, chemotherapy): Cryptosporidium, barely affected by SODIS, causes them severe, prolonged diarrhea;
  • people with chronic digestive diseases.
• Chemical contamination: SODIS removes nothing chemical. Do not treat water that smells of fuel, solvent or pesticide, nor water from mining or industrial areas without testing it: find another source. Do not reuse bottles that held toxic substances.
• Substances from the plastic: the available studies find no significant risk from migration of substances from PET into SODIS water: the plasticizers measured (DEHA, DEHP) were at levels similar to bottles stored in the dark and well below WHO guideline values (Schmid et al., 2008, cited in Luzi et al., 2016). Antimony from PET only migrates appreciably above 60 °C or with very long storage times: do not deliberately heat PET above that temperature (very efficient collectors) or store water for weeks in the heat. Never use polycarbonate (symbol 7).
• Flame test to tell PET from PVC: PVC releases irritant, toxic smoke (hydrochloric acid). If this test must be done, burn only a 1 cm² piece outdoors, with tongs, without inhaling the smoke and away from children. PET burns more easily and its smoke smells sweetish; PVC's is acrid (Meierhofer and Wegelin, 2002).
• Falls: carrying full bottles (2 kg each) onto roofs causes falls. Prefer a waist-height frame with sheet metal; if a roof is used, always have the same adult do it, with a stable ladder.
• Contact burns: a metal sheet at midday sun can burn to the touch; pick up bottles by the cap or with a cloth and do not rest hands on it or let children sit on it.
• Glass: glass bottles break easily and cause cuts; do not use them on roofs or within reach of small children.
• Mixing up bottles: a child may drink from an untreated bottle. Always mark the treated ones (step 8) and keep the raw ones separate.

Variants

• With fewer resources (no PET bottles):
  • Without a container transparent to UV-A there is no SODIS. The N0-N1 alternatives are boiling and pasteurization in ceramic pots over a fire: see boiling water.
  • Clear polyethylene bags (recovered): trials showed very high disinfection thanks to their shallow depth and larger surface area, but Eawag does not recommend them because they are hard to handle and break more easily than bottles (Meierhofer and Wegelin, 2002, p. 24). Only as an emergency resort: follow the general rule of water depth ≤ 10 cm (variables table; Sommer et al., 1997), lay them flat without folds that cast shade, and apply the same exposure time.
  • Without a metal sheet: any dark ground or flat stone in the sun; the improvement from sheets and dark backgrounds is under 30 %, so lacking one does not require a longer time (Luzi et al., 2016, chap. 2.2.3).
• N2: glass bottles. Glass does not scratch or warp with heat, but it is heavy, breaks, and its UV-A transmission depends on the iron oxide content: 2 mm window glass barely lets UV-A through, while the colorless commercial beverage bottles tested by Eawag transmitted comparably to PET and gave the same effectiveness (Luzi et al., 2016, box 7). That box only backs commercial beverage bottles and recommends measuring the UV transmittance of locally available glass. N2 artisanal glass is usually greenish from the iron in the sand, and at N2 there is no E. coli kit to validate it with (comparative test in “Verification”): without an analysis kit, artisanal glass is not adopted for SODIS; it is used only for solar pasteurization with temperature control (below). If a kit becomes available later, candidate glass must be colorless when viewed edge-on, as thin-walled as possible, and in bottles of ≤ 2 L.
• Solar pasteurization (with more resources or cold climates). Instead of relying on UV, it relies on heat: bottles or containers with the lower half painted black, inside a box with aluminum foil or a solar cooker, until the water is kept above 65 °C (Ciochetti and Metcalf, 1984; time and criterion in the boiling water article), with a WAPI or thermometer as a check. In a test at 32° N, a half-blackened bottle went from 21 °C to 42.5 °C in the open air and to 66.5 °C inside an aluminum-lined box between 9 h and 14 h (Green, Lifewater, cited in Meierhofer and Wegelin, 2002). Advantage: it also inactivates protozoa and viruses that UV-A barely touches, and it works in opaque containers. Disadvantages: PET warps above ~65 °C (use glass or dark metal), and it requires checking the temperature. The 2016 guide no longer promotes painting bottles black for standard SODIS; at altitude, where the water does not heat up, a fully transparent bottle is preferable. Details on pasteurization and the WAPI are in boiling water.
• With more resources (N3-N4): reflectors or compound parabolic collectors that concentrate light (acceleration typically under 50 %; Luzi et al., 2016, chap. 2.3); electronic or chemical UV indicators that warn of the dose received; E. coli kits to monitor quality; combination with filtration (ceramic and carbon filters) for turbid water, or with chlorine when residual protection or effectiveness against viruses is needed.
• Scaling up to a village: SODIS is a household method: it scales by multiplying bottles and homes, not by enlarging containers. It requires an organized supply of PET bottles, communal exposure points (metal roofs of the school or the health center) and, above all, training and follow-up visits: in the field, without follow-up, use drops sharply once promotion ends (in Nepal and Indonesia, around 21 % use years later; Luzi et al., 2016, chap. 3.1.2). See “Calculations”.
• Scaling up to a city: not suitable. Large-volume solar reactors (a few hundred liters a day) are technically complicated to build, clean and maintain, and cannot compete on cost with centralized chlorination (Luzi et al., 2016, chap. 2.3). In a city, SODIS only makes sense as a household backup during outages or emergencies.

Observed health effectiveness

Results from field trials (Luzi et al., 2016, table 8; CDC, 2008): in children aged 5-16 in Kenya, an odds ratio for diarrhea of 0.66 (Conroy et al., 1996); in children under 6, during a cholera outbreak, 3 cases among 155 who used SODIS versus 20 among 144 who did not (odds ratio 0.12; Conroy et al., 2001; table 8 of Luzi et al., 2016 summarizes it as 81 % less cholera in children under 5, a discrepancy with the title of the original article); in Bolivia, no significant effect with only 32 % compliance (Mäusezahl et al., 2009). The CDC summarizes diarrhea reductions of 9 to 86 % across four controlled trials. The difference is made by consistent and correct use: drinking even 5-10 % untreated water can cancel out almost all the benefit (Brown and Clasen, 2012, cited in Luzi et al., 2016).

Calculations

Bottles for a family

Data: 5 people; minimum drinking water of about 2 L per person per day in mild conditions, up to 4.5 L with heat and physical work (Howard and Bartram, 2003). 3 L/person/day is used.

Daily water = 5 people × 3 L = 15 L
2 L bottles that must finish their exposure each day = 15 L / 2 L = 7.5 → 8 bottles

Sunny regime (1 day of exposure):
  8 bottles in the sun + 8 treated ready to drink at the same time = 16 bottles

Cloudy regime (2 days of exposure):
  for 8 bottles to finish each day, 2 alternating batches must be in the sun at once:
  8 (batch on its 1st day) + 8 (batch on its 2nd day) + 8 treated ready to drink = 24 bottles

With only 16 bottles, in a cloudy spell only 8 bottles come out every 2 days, i.e. 8 L/day against the 15 L/day needed: either 24 bottles are kept, or boiling supplements the shortfall during cloudy spells. Eawag's guideline of 4 bottles per person (20 for this family) does not fully cover the 2-day regime at 3 L/person/day either.

Exposure surface

Measure the available bottles lying flat and add about 2 cm of spacing between them. Worked example with a bottle that, lying flat with its spacing, takes up 12 cm × 32 cm:

Surface per bottle ≈ 0.12 m × 0.32 m ≈ 0.038 m²
Family, sunny regime (8 in the sun)  ≈ 8 × 0.038  ≈ 0.31 m²  → a 1 m × 0.5 m sheet is more than enough
Family, cloudy regime (16 in the sun) ≈ 16 × 0.038 ≈ 0.61 m²  → 1 m × 0.7 m sheet

Village of 200 people

Drinking water = 200 × 3 L = 600 L/day
Bottles in the sun = 600 / 2 = 300 bottles/day
Surface = 300 × 0.038 ≈ 11.5 m²  (≈ 6 corrugated sheets of 2 m × 1 m)
Inventory, sunny regime = 300 in the sun + 300 to drink = 600 bottles
Inventory, cloudy regime = 600 in the sun + 300 to drink = 900 bottles (or boil during cloudy spells)
Surface, cloudy regime = 600 × 0.038 ≈ 23 m²
Labor = 300 bottles/day × (minutes per bottle timed in the first week)

There is no published figure for minutes per bottle (the 2016 manual only speaks of a “relatively high daily work demand”): time it in practice before organizing shifts.

That is why in a village it helps for each household to treat its own water at home (spreading the labor and avoiding transfers) and to reserve communal exposure for schools and health centers.

Sources

• Luzi, S.; Tobler, M.; Suter, F.; Meierhofer, R. SODIS manual: Guidance on solar water disinfection. Eawag/Sandec, Dübendorf, 2016. ISBN 978-3-906484-59-4. Caps. 2.1-2.4 (mecanismo, tablas 1-4 de inactivación, tabla 5 de dosis, recuadros 6-8 sobre tiempo de exposición, vidrio y seguridad del PET, recrecimiento, almacenamiento) y 3.1 (adopción e impacto en salud, tabla 8). https://www.sodis.ch/methode/anwendung/ausbildungsmaterial/dokumente_material/sodismanual_2016_lr.pdf
• Meierhofer, R.; Wegelin, M. Solar Water Disinfection: A Guide for the Application of SODIS. EAWAG/SANDEC, informe n.º 06/02, octubre de 2002. ISBN 3-906484-24-6. Caps. 2.1-2.7 (UV-A y temperatura, clima, prueba de turbidez, oxígeno, materiales, procedimiento) y 3.2-3.4 (eficacia en campo, errores de uso, lactantes y grupos de riesgo). Versión española disponible en sodis.ch (manual_s.pdf).
• CDC. Household Water Treatment Options in Developing Countries: Solar Disinfection (SODIS). Hoja informativa, enero de 2008.
• Wegelin, M. et al. (1994). «Solar water disinfection: scope of the process and analysis of radiation experiments». J Water SRT-Aqua, 43(3), 154-169 (dosis de 555 Wh/m² y sinergia a 50 °C).
• Sommer, B. et al. (1997). «SODIS: an emerging water treatment process». J Water SRT-Aqua, 46(3), 127-137 (atenuación de la UV-A por turbidez y profundidad).
• Conroy, R.M. et al. (1996). «Solar disinfection of drinking water and diarrhoea in Maasai children: a controlled field trial». The Lancet, 348, 1695-1697; Conroy, R.M. et al. (2001). «Solar disinfection of drinking water protects against cholera in children under 6 years of age». Arch Dis Child, 85, 293-295.
• OMS. Guidelines for Drinking-water Quality, 4.ª ed., 2011, cap. 7, tabla 7.8 (reducciones de patógenos esperadas con tratamientos domésticos, p. 146).
• Rehydration Project. Oral Rehydration Solutions: Made at Home (receta casera de SRO de la OMS: 1 L de agua, 6 cucharaditas rasas de azúcar, 1/2 cucharadita rasa de sal).
• AEMET, Atlas de radiación solar en España, y Comisión Europea, PVGIS (valores mensuales de irradiación global para aplicar el criterio de latitudes altas).
• Howard, G.; Bartram, J. Domestic Water Quantity, Service Level and Health. OMS, Ginebra, 2003.
• Ciochetti, D.A.; Metcalf, R.H. (1984). «Pasteurization of Naturally Contaminated Water with Solar Energy». Applied and Environmental Microbiology, 47(2), 223-228.
• Estudios citados de segunda mano a través de Luzi et al. (2016), sin consulta directa: Reed (1997, oxígeno), Schmid et al. (2008, plastificantes), Mäusezahl et al. (2009, Bolivia), Brown y Clasen (2012, consumo de agua sin tratar). </content>