Managing water in the home: accelerated health gains from improved water supply:
Previous page |
Heat and UV radiation
Overall, the results of both microbiological and epidemiological indicate that solar disinfection of household water has the ability to appreciably improve its microbial quality and to reduce household diarrheal disease of consumers. Additional epidemiological studies to better document the extent of diarrheal disease reduction are recommended because available studies are limited to only one geographic region, Kenya, and study population, Maasai children. Apparently, additional epidemiological studies are now in progress (Mintz et al., 2001). Because of its simplicity, low cost, and the need for only beverage bottles and sunlight, solar disinfection is an appropriate technology for disinfection of household water in the developing world.
The UV radiation technology is simple to use and highly effective for inactivating microbes in drinking water, and it does not introduce chemicals or cause the production of harmful disinfection by-products in the water. While UV lamp disinfection systems have been widely used to disinfect drinking water at the community and household levels, no epidemiological studies of intervention type that document health impacts at the household level have been reported for this technology. There are no reasons to doubt the efficacy of sound UV lamp disinfection technology to adequately disinfect either household or community drinking water when properly applied. However, field studies documenting the ability of this technology to disinfect household drinking water and reduce diarrhea and other waterborne diseases are recommended. Such studies would validate the expected performance of this technology and provide further evidence that the technology is reliable and capable of being used successfully by individuals and communities. Such documentation is needed because UV lamp disinfection has some disadvantages for use as a drinking water disinfectant at the household level. It does not provide a chemical disinfectant residual to protect the water from recontamination or microbial regrowth after treatment. Particulates, turbidity and certain dissolved constituents can interfere with or reduce microbial inactivation efficiency. A reliable and affordable source of electricity is required to power the UV lamps. The UV lamps require periodic cleaning, especially for systems using submerged lamps, and they have a finite lifespan and must be periodically replaced. The technology is of moderate to high cost when used at the household level. Despite these drawback and limitations, UV irradiation with lamps is a recommended technology for disinfection of household and community water.
1 Boiling or heating with fuel
Boiling or heating of water with fuel has been used to disinfect household water since ancient times. It is effective in destroying all classes of waterborne pathogens (viruses, bacteria and bacterial spores, fungi and protozoans and helminth ova) and can be effectively applied to all waters, including those high in turbidity or dissolved constituents. Although some authorities recommend that water be brought to a rolling boil for to 1 to 5 minutes, the WHO GDWQ recommend bringing the water to a rolling boil as an indication that a high temperature has been achieved. These boiling requirements are likely to be well in excess of the heating conditions needed to dramatically reduce most waterborne pathogens, but observing a rolling boil assures that sufficiently high temperatures have been reached to achieve pathogen destruction.
Although boiling is the preferred thermal treatment for contaminated water, heating to pasteurization temperatures (generally $ 60oC) for periods of minutes to tens of minutes will destroy most waterborne pathogens of concern. Even heating to as little as 55oC for several hours has been shown to dramatically reduce non-sporeforming bacterial pathogens as well as many viruses and parasites, including the waterborne protozoans Cryptosporidium parvum, Giardia lamblia and Entamoeba histolytica (Feachem et al., 1983; Sobsey and Leland, 2001). In many situations, however, it is not possible to monitor the temperature of the water with a thermometer or other temperature sensor such as a melting wax visual indicator system. Unless such temperature monitoring is possible, caution is recommended in attempting to pasteurize waters at non-boiling temperatures.
It is also recommended that the water is stored in the same container in which it has been boiled or heated, preferably one with a lid or other protected opening, in order to reduce opportunities for recontamination. It is further recommended that boiled or heat-treated water be consumed soon after it has cooled and preferably within the same day. This is because of the potential for microbial recontamination during prolonged storage. Introduction of microbes from hands, utensils and other sources is a major concern. A major disadvantage of boiling is its consumption of energy in relation to the availability, cost and sustainability of fuel. It is estimated that 1 kilogram of wood is needed to boil 1 liter of water. In areas of the world where wood, other biomass fuels or fossil fuels are in limited supply and must be purchased, the costs of boiling water are prohibitive. Therefore, boiling household water is unrealistic and inaccessible for many of the world's poorest people due to the scarcity and high cost of fuels and the lack of sustainability of biomass or fossil fuels in the community or region. In some areas of the world the use of wood and wood-derived fuels is also a concern because it contributes to the loss of woodlands and the accompanying ecological damage caused by deforestation. However, where affordable and sustainable sources of fuel are available without causing environmental degradation, heating household water to a rolling boil is an effective and accessible method of treatment for collected household water.
2 Thermal treatment with solar radiation and solar cooking
Although boiling with fuel may be a prohibitive option for household treatment of water, heating water, other liquids and other foods to lower temperatures using solar radiation is a more accessible, economical and technologically feasible option than heating with fuel. Treatment of water with solar radiation was practiced in ancient India more than 2000 B.C.E. The ability of solar radiation to disinfect has been recognized in modern times at least since studies at by Acra at al. (1984) at The American University of Beirut, Lebanon. Since then, it has been shown that water can be heated to temperatures of $ 55oC in transparent bottles (e.g., clear plastic beverage bottles) exposed to sunlight for several hours, especially if the bottle is painted black on one side or is lying on a dark surface that collects and radiates heat (Wegelin et al., 1994; Joyce et al., 1996). This method of treatment utilizes both the UV radiation in sunlight as well as the thermal effects of sunlight to inactivate waterborne microbes, and will be discussed in detail in the next section of this report. Alternatively, if the exterior of the vessel is completely black or similarly capable of absorbing heat (e.g., most metal containers), only thermal effects occur and temperatures can reach >60oC. At these temperatures, water and other liquids can be pasteurized because most enteric viruses, bacteria and parasites are rapidly inactivated (Ciochetti and Metcalf, 1984). Furthermore, if a dark, opaque container is more highly exposed to solar radiation using a solar reflector or solar cooker, the water temperature can reach $ 65oC, a pasteurization temperature capable of inactivating nearly all enteric pathogens within several tens of minutes to hours (Safapour and Metcalf, 1999).
In those parts of the world where solar cooking already is available and widely practiced, solar pasteurization of water, other beverages and weanling foods is a practical, accessible and affordable option for household water treatment. Low cost solar reflectors or cookers can be made from materials as simple and economical as cardboard and aluminum foil. This technology for water treatment and food preparation has been field tested in many parts of the world, including Kenya, Tanzania, Ethiopia, Vietnam and some countries in the Americas. A major limitation of solar heating is that only small volumes (# 10 liters) of water can be exposed conveniently at one time per water container and solar reflector. However, using multiple water containers and alternative solar collectors (e.g., metal roofing material), the volume of water treatable with solar heat at one time can be substantially increased. Another important limitation of solar heating is the availability of sunlight, which varies greatly with season, daily weather (meteorological) conditions and geographic location. However, in many regions of the developing world, sunlight conditions are suitable for solar heating of water and for cooking nearly all year long on full sun or part sun days (approximately 200-300 days per year). A third potential limitation of solar heating to disinfect water is the determination of water temperature. Thermometers are relatively expensive and may not be available or affordable in many regions of the developing world. However, several simple, low cost temperature indicators have been devised. One of the simplest and most effective is a reusable water pasteurization indicator (WAPI) based on the melting temperature of soybean wax. The WAPI consists of a clear plastic tube partially filled with a soybean wax that melts at about 70oC and a piece of nylon (e.g., fish) line attached on each end to stainless steel washers. The WAPI is placed in the water to be heated with the wax at the top of the tube. When the wax reaches 70oC, it melts and falls to the bottom of the tube, thereby giving a simple visual indication of when pasteurization conditions have been achieved. Similar wax indicators have been devised for other target melting temperatures, depending on the type of wax.
3 Solar treatment by combined UV and thermal effects
Treatment to control waterborne microbial contaminants by exposure to sunlight in clear vessels that allows the combined germicidal effects of both UV radiation and heat also has been developed, evaluated and put into field practice (Acra et al., 1984; Conroy et al., 1996; 1996; 1999; Joyce et al., 1996; McGuigugan et al., 1998; 1999; Sommer et al., 1997; Wegelin and Sommer, 1998; Wegelin et al., 1994). A number of different solar treatment systems have been described, but one of the technically simplest and most practical and economical is the SODIS system developed by scientists at the Swiss Federal Agency for Environmental Science and Technology (EAWAG) and its many collaborators and partners. The SODIS system consists of three basic steps: (1) removing solids from highly turbid (>30 NTU) water by settling or filtration, if necessary, (2) placing low turbidity (<30 NTU) water in clear plastic bottles of 1-2 liter volume (usually discarded beverage bottles and preferably painted black on one side), and (3) aerating (oxygenating) the water by vigorous shaking in contact with air, and (4) exposing the filled, aerated bottles to full sunlight for about 5 hours (or longer if only part sunlight). The water is exposed to UV radiation in sunlight, primarily UV-A and it becomes heated; both effects contribute to the inactivation of waterborne microbes. The system is suitable for treating small volumes of water (<10L), especially if the water has relatively low turbidity (<30 NTU). Clear plastic bottles are considered preferable by some workers over glass because they are lighter, less likely to break, and less costly. Bottles made of polyethylene terephthlate (PET) are preferred to those made of polyvinylchloride (PVC), other plastics and most types of glass because they are less likely to leach harmful constituents into the water. In addition, they are lightweight, relatively unbreakable, chemically stable and not likely to impart tastes and odors to the water. PET bottles require period replacement because they can be scratched and they become deformed if temperatures exceed 65oC. The use of an internal temperature sensor is encouraged as an aid to determining if a minimum target temperature of 50oC and preferably 55oC or higher is reached. The reusable sensor contains paraffin wax attached to a screw weight. When the paraffin melts, the weight drops to indicate that the target temperature has been attained.
The effects of several factors influencing microbial inactivation by solar disinfection are summarized in Table 5 below. Microbial inactivation by the SODIS system is attributed to the combined effects of UV radiation in the UV-A range (320 to 400 nm), which is somewhat germicidal, and heating to temperatures of 50-60oC, which are high enough to extensively ($ 99.9%) inactivate many enteric viruses, bacteria and parasites in about 1 hour to several hours. It has been reported that the combined exposure to UV plus heat in the SODIS process has a synergistic effect on microbial inactivation, producing greater inactivation than predicted by comparable levels of exposure to either one of the two agents alone. During the exposure period, UV dose increases to $ 100 Wh/m2 and the water temperature reaches 55oC or even higher. However, others report that even non-UV transmissible sunlight when used to heat water to 60oC in a commercial solar panel system, will inactivate enteric bacteria, spores and viruses (Rijal and Fujioka, 2001). When treated with heat and no UV or heat with UV for 2-5 hours, fecal coliforms, E. coli, enterococci, HPC bacteria and coliphage MS2 were reduced by >3 log10 and Clostridium perfringens spores were 1-2 and nearly 3 log10, respectively. Under cloudy conditions bacterial and spore reductions were much lower, they were lower with heat-no UV than with heat plus UV and temperatures did not reach 50oC. Therefore, achieving a sufficiently high temperature (preferably 55oC or higher for several hours) is an important factor for microbial inactivation by solar disinfection systems. Overall, studies have shown that various bacteria, such as fecal coliforms, E. coli and enterococci, and viruses, such as coliphage f2, rotavirus and encephalomyocarditis (EMC) virus, in water bottles are reduced extensively (by several orders of magnitude) when exposed to sunlight for periods of several hours and sufficiently high temperatures are achieved.
Studies also show that dissolved oxygen in the water contributes to bacterial inactivation, with much greater reductions of E. coli and enterococci after 3 hours in oxygenated water (~6 log10) than in anaerobic water (<2 and <1 log10, respectively) (Reed, 1996). In subsequent studies total and fecal coliforms were inactivated by >3 log10 in 6 and 4 hours, respectively, in aerated water and by # 1.5 log10 in anoxic water or water kept from sunlight (indoors) (Meyer and Reed, 2001). Therefore, aeration of the water by mechanical mixing or agitation is recommended before solar treatment in bottles. The combined process of oxygenation (aeration) by mixing, followed by solar radiation exposure for several hours in a clear plastic bottle is also referred to as solar photooxidative disinfection of SOLAR. Enteric bacteria inactivated by the SOLAR or SODIS process do not appear to regrow or recover their infectivity.
Table 5. Factors influencing microbial inactivation by solar disinfection of water
4 Advantages, disadvantages and limitations of solar treatment systems
|Factor||Influences on Microbial Inactivation|
|Type of microbe||Microbes differ in sensitivity to inactivation be heat and by UV radiation. Heat is more effective against vegetative bacteria, |
|Water Vessel||Type, composition, volume, and depth influence water temperature, UV penetration of water, cleanability and portability; PET or |
|Sunlight; ambient temperature||Sunlight intensity, duration, and cloudiness influence water temperature and UV penetration; ambient temperature influences inte|
|Vessel placement and orientation||Exposure to full sun without shade (from trees and other objects); influences water temperature and UV exposure; horizontal inst|
|Mixing or movement of vessel||To provide more uniform water exposure to sunlight and minimize differences in sunlight (UV) dosimetry|
|Solar collection or reflection||Solar collection (on dark surfaces) or reflection (by shiny surfaces of reflector panels or cookers) influence water temperature|
|Water quality||UV exposure (UV scattering by particles and absorption by solutes and particles); microbial protection by solids-association|
|Water Aeration (oxygenation)||Increased oxygen content of the water by agitating (shaking) for several minutes in contact with air prior to sunlight exposure |
|Exposure time||Water temperature and duration of exposure to elevated temperature; cumulative UV dose. Typically several hours with full sunlig|
The advantages of disadvantages of solar treatment systems are summarized in Table 6 below. Potential limitations of this and perhaps other solar disinfection systems are: the availability of suitable water containers and other needed materials, lack of sunlight for disinfection, potential difficulties in treating highly turbid water and the availability of simple methods for reducing the turbidity of water before solar treatment, lack of a residual disinfectant to protect water during handling and storage, potential user objections to the technology due to the length of time to treat the water (several hours or longer) and possible objectionable tastes and odors leached into the water from the plastic bottles. Despite these limitations, solar disinfection in clear plastic bottles is one of the most promising and extensively tested methods for disinfection of household water stored in a container.
Table 6. Advantages and disadvantages of solar treatment systems
5 Epidemiological studies of solar disinfection of Household Water
|Microbial inactivation by pasteurization (temperatures of 55°C or higher for several hours are recommended).||Often requires several hours to disinfect and even longer (2 days) if cloudy weather; more heat-resistant pathogens inactivated ||Time to inactivate varies with system (UV+heat) or (heat only) and sunlight conditions; requires a system to indicate that targe|
|Simple, low cost use of small vessels (PET plastic for SODIS and black or opaque bottles for solar reflection or cooking system)||Limited to volumes of 1-several liters per bottle; using 1.5-L bottles (optimum size), several bottles are needed per household ||Availability of sufficient number of suitable bottles, depending on type of solar treatment (simple sunlight exposure vs. solar |
|Does not change the chemical quality of the water.||Provides no chemical disinfectant residual; water must be consumed within a day or so, or else microbial regrowth can occur.||Leaching of chemicals possible from some plastic bottles, causing objectionable tastes and odors; periodic bottle replacement ne|
|SODIS (heat + UV) system effective in water with low to moderate turbidity (<30 NTU).||High turbidity interferes with microbial inactivation; requires turbidity reduction by sedimentation, filtration or other method||Requires clear bottles allowing penetration of UV radiation (preferred plastic is polyethylene terephtalate or PET; some bottles|
|Apparent synergistic effects of heat and UV in the SODIS (UV+heat) system||Requires low (<30 NTU) turbidity water; requires at least several clear plastic bottles and an opaque or black surface on a side||Evidence of synergistic effects documented for vegetative bacteria but it has not been studied for viruses or parasites|
|Improved bacterial inactivation in aerobic water by SODIS system||Requires pre-aeration (e.g., mechanical mixing) to create aerobic conditions; effect may not occur in water with reducing agents||Inactivation of E. coli >10,000-fold higher in aerobic water (99.9999% reduction) than in anaerobic water (90-99%); effect has n|
|Opaque or black bottle system achieves temperatures high enough to inactivate viruses and is less affected by turbidity or UV-ab||System requires solar collector or cooker to deliver sufficient solar energy; small volume of water vessels; poor inactivation o||Solar cooker system gives virus inactivation of 99.99% in 1.5 hours in a 1.4-L black bottle and 99% inactivation in 3 hours in a|
In addition to the essential technical components, the SODIS system for drinking water disinfection also includes important educational, socio-cultural, behavioral and motivational components, such as education and training, behavior modification and motivational training. SODIS has been field-tested in many different parts of the world and in many countries, including South America (Colombia and Bolivia), Africa (Burkino Faso and Togo), Asia (China) and Southeast Asia (Indonesia and Thailand). It has been introduced and disseminated by both governments and NGOs and subjected to economic analysis based on actual costs (estimated at 3 US$ per year for a household of 5 people) for willingness to pay. Acceptance rates, based on willingness to continue use after its introduction as a demonstration project, is reported at to be >80%. However, when introduction was not adequately supported by community involvement activities to address educational, socio-cultural, behavioral and motivational issues, community support for continued use was lower.
The SODIS system has not been extensively tested for reduction in waterborne disease in epidemiological studies of the intervention type. However, as shown in Table 7, three reported studies found measurable reductions in diarrheal disease and cholera in Maasai (Kenyan) children drinking solar disinfected water (several hours of full sunlight) compared to children drinking undisinfected water (kept indoors) in the same plastic bottles (Conroy et al., 1996; 1999; 2001). While it is clear from these intervention studies that reductions in diarrheal disease and cholera by solar disinfection in bottles are achieved in children under 6 years of age, more studies of this type are needed. This is because it is important to determine the extent of reduction of diarrheal and other waterborne disease by this system in different geographic locations having different water quality conditions and different populations at risk. The measurement of the microbial quality of the water used by intervention and control groups also would be desirable to further document the efficacy of the solar disinfection system. While there are reliable laboratory and field data from studies documenting the inactivation of waterborne microbes by solar disinfection systems, such documentation has not been included in the epidemiological studies reported to date. Therefore, the extent of pathogen or microbial indicator reduction in waters used by those consuming the water and being monitored for diarrheal and other enteric disease is not known.
6 UV irradiation with lamp systems
Table 7. Epidemiological studies on diarrheal disease reduction by the SODIS solar disinfection system of household water
|Location||Water||Treatment||% Reduction in Disease||Significant Microbial Reduction?||Reference|
|Kenya||Household||Solar Disinfection||86%||Not Reported||Conroy et al., 2001|
|Kenya||Household||Solar Disinfection||16%, diarrhea||Not Reported||Conroy et al., 1999|
|Kenya||Household||Solar Disinfection||9 (a)/26 (b), diarrhea||Not Reported||Conroy et al., 1996|
7 UV Inactivation of microbes in water
(a) Total diarrheal disease (b) Severe diarrheal disease
8 UV disinfection systems using lamps
The germicidal activity of ultraviolet radiation from lamps was recognized in the late 1800s and disinfection of drinking water and other media with UV lamps has been practiced since the early part of the 20th century (Baker 1948, Blatchley and Peel, 2001; 1948; Sobsey, 1989; Ward, 1893). This method of drinking water disinfection has received renewed interest in recent years because of its well-documented ability to extensively (>99.9%) inactivate two waterborne, chlorine-resistant protozoans, Cryptosporidium parvum oocysts and Giardia lamblia cysts, at relatively low doses (<10 mJ/cm2).
UV disinfection is usually accomplished with mercury arc lamps containing elemental mercury and an inert gas, such as argon, in a UV-transmitting tube, usually quartz. Traditionally, most mercury arc UV lamps have been the so-called "low pressure" type, because they operate at relatively low partial pressure of mercury, low overall vapor pressure (about 2 mbar), low external temperature (50-100oC) and low power. These lamps omit nearly monochromatic UV radiation at a wavelength of 254 nm, which is in the optimum range for UV energy absorption by nucleic acids (about 240-280 nm). In recent years medium pressure UV lamps that operate at much higher pressures, temperatures and power levels and emit a broad spectrum of higher UV energy between 200 and 320 nm have become commercially available. However, for UV disinfection of drinking water at the household level, the low-pressure lamps and systems are entirely adequate and even preferred to medium pressure lamps and systems. This is because they operate at lower power, lower temperature, and lower cost while being highly effective in disinfecting more than enough water for daily household use. An essential requirement for UV disinfection with lamp systems is an available and reliable source of electricity. While the power requirements of low-pressure mercury UV lamp disinfection systems are modest, they are essential for lamp operation to disinfect water.
9 Costs of UV disinfection for household water
At sufficiently high doses, all waterborne enteric pathogens are inactivated by UV radiation. The general order of microbial resistance (from least to most) and corresponding UV doses for extensive (>99.9%) inactivation are: vegetative bacteria and the protozoan parasites Cryptosporidium parvum and Giardia lamblia at low doses (1-10 mJ/cm2) and enteric viruses and bacterial spores at high doses (30-150 mJ/cm2). Most low-pressure mercury lamp UV disinfection systems can readily achieve UV radiation doses of 50-150 mJ/cm2 in high quality water, and therefore efficiently disinfect essentially all waterborne pathogens. However, dissolved organic matter, such as natural organic matter, certain inorganic solutes, such as iron, sulfites and nitrites, and suspended matter (particulates or turbidity) will absorb UV radiation or shield microbes from UV radiation, resulting in lower delivered UV doses and reduced microbial disinfection. Another concern about disinfecting microbes with lower doses of UV radiation is the ability of bacteria and other cellular microbes to repair UV-induced damage and restore infectivity, a phenomenon known as reactivation. UV inactivates microbes primarily by chemically altering nucleic acids (pyrimidine dimers and other alterations). However, the UV-induced chemical lesions can be repaired by cellular enzymatic mechanisms, some of which are independent of light (dark repair) and others of which require visible light (photorepair or photoreactivation). Therefore, achieving optimum UV disinfection of water requires delivering a sufficient UV dose to induce greater levels of nucleic acid damage and thereby overcome or overwhelm DNA repair mechanisms.
Two alternative configurations or physical systems are used for UV disinfection of small or household water supplies, submerged lamps or lamps in air and mounted above a thin layer of the water to be irradiated. In the units with submerged lamps, the lamps are covered with a protective, UV-penetrable as protection from the electrical hazards associated with water. Water can be treated on a batch basis by placing the lamp in a container of water for several minutes or longer, or on a flow-through basis in a housing or channel, with the water flowing parallel or perpendicular to the lamp(s). In units having the lamps mounted in the air, the UV lamps are in a metal housing with reflective surfaces that direct the UV radiation downward onto a thin layer of water flowing through a channel or tray below the lamps. The advantage of the submerged systems is intimate lamp contact with the water, water-mediated cooling of the lamps, and the use of housing designs that maximize UV exposure of the water. However, the protective sleeves over the lamps must be mechanically or chemically cleaned on a regular basis to overcome fouling by a physical, chemical or biological film that can forms on the sleeve surface, reducing UV passage into the water. The non-submerged, in-air lamp units have the advantage of no need for lamp cleaning due to lamp fouling, but there is some loss of UV radiation due atmospheric and surface absorption. However, both types of UV disinfection system designs are available for disinfection of household water at point- of-use, point-of-entry or at the community level .
UV disinfection with lamps has the advantages of being effective for inactivating waterborne pathogens, simple to apply at the household and community levels, and relatively low cost, while not requiring the use of chemicals or creating tastes, odors or toxic chemical by-products. The disadvantages of UV disinfection with lamps are the need for a source of lamps, which have to be replaced periodically (typically every year or two), the need for a reliable source of electricity to power the lamps, the need for period cleaning of the lamp sleeve surface to remove deposits and maintain UV transmission, especially for the submerged lamps, and the uncertainty of the magnitude of UV dose delivered to the water, unless a UV sensor is used to monitor the process. In addition, UV provides no residual chemical disinfectant in the water to protect against post-treatment contamination, and therefore care must be taken to protect UV-disinfected water from post-treatment contamination, including bacterial regrowth or reactivation.
Because the energy requirements are relatively low (several tens of watts per unit or about the same as an incandescent lamp), UV disinfection units for water treatment can be powered at relatively low cost using solar panels, wind power generators as well as conventional energy sources. The energy costs of UV disinfection are considerably less than the costs of disinfecting water by boiling it with fuels such as wood or charcoal.
UV units to treat small batches (1 to several liters) or low flows (1 to several liters per minute) of water at the community level are estimated to have costs of 0.02 US$ per 1000 liters of water, including the cost of electricity and consumables and the annualized capital cost of the unit. On this basis, the annual costs of community UV treatment would be less than US$1.00 per household. However, if UV lamp disinfection units were used at the household level, and therefore by far fewer people per unit, annual costs would be considerably higher, probably in the range of $US10-100 per year. Despite the higher costs, UV irradiation with lamps is considered a feasible technology for household water treatment.
Managing water in the home: accelerated health gains from improved water supply:
| Next page