To first consideration, earthenware may not seem to be the best media for water purification. What we know about the material is that it has lots of pores but these are only poorly connected. As conventionally processed and fabricated, red firing clays do not have good permeability. Other materials appear better suited to the task, for example, diatomite. But in considering a widely duplicable production methodology for the poor, those most vulnerable to the water born, gastro intestinal illnesses, dysentery, cholera, typhoid, etc., earthenware clays offer the clear advantage of being available almost everywhere. Sustainability is largely about using local resources. 

For this reason there is an opportunity not only from the perspective of the user, but for a widely duplicable fabrication process, from the perspective of the producer. The model should be a good one with respect to low cost, easiness of use and effectiveness, but it is the low startup cost of production that will encourage the entrepreneur. Expect the factory to lack amenities like electricity and running water. The pavement may be earth. This would be an example for micro-scale production, irrespective of economy of scale. But there may be some question as to maintaining quality at this level. One answer is to apply the silver, required as disinfectant, at the time of purchase, at some central point of sale. 

Materials used for this earthenware purifier include primarily a red firing clay and a very fine grained combustible. In the Bangladeshi scenario, wheat flour is excellent for this purpose, very inexpensive. These two materials make a damp pressed purifier that has proven 99% effective at removal of particles of all sizes down to 1.0 micron, only a little bigger than the size of fecal coliforms. Further, bacteriological challenge testing has proven that a very tiny amount of silver, added to the purifier as a colloid, kills any fecal coliforms not filtered out by the tortuous path they follow. A 1930 patent by the Swiss company Katadyn has indicated that these pathogens are killed by oligodynamic action. This is thought to be the first time that small clusters of silver metal were dispersed within a ceramic filter medium. 

By contrast to clay, sand is a material that offers good permeability but no porosity. A current ‘best practice’ promoted by humanitarian agencies in emergency situations is ‘slow sand filtration.’ Placing sand in a 55 gallon drum, then pouring water through, does remove about 60% of fecal coliforms, but this is certainly no guarantee to the vulnerable. 

Another prospective material for purifiers could be kaolin, white clay, where the smallest pores tend to be around 1.0 micron. At first glance, this would appear to have the clear advantage of filtering out a larger percentage of fecal coliforms, at 0.5 microns each, than might be the case with an earthenware. The red firing clays tend to have their smallest pores between 3.0 and 6.0 microns. But it is thought that it is the tortuosity of the path through the filter that traps so high a percentage of the pathogens as is the case with the earthenware. The membrane filtration used for bacteriological testing has indicated an uncountable number of overlapping bacteria colonies for untreated river water, twenty colonies for a control purifier containing no silver, and zero bacteria for the silver saturated purifier. This might be taken to indicate that at least 80 to 90% of the fecal coliforms were removed by filtration, the remainder by oligodynamic action. 

We refer to the newly developed material that makes a damp pressed earthenware purifier possible as porous or ‘permeable grog.’ Permeable grog is a red clay, combined with fine grained combustible, which is pre-fired, then crushed to a powder, subsequently combined with unfired clay. (Steps necessary in fabrication are listed below.) This grog indicates a permeability not unlike that of sand, along with the porosity of clay. So the silver material applied after the second, and final firing is trapped in the permeable structure, not readily washed away. Where using a colloidal silver, the metal probably gets stuck within the tortuous structure. Where using silver nitrate, some bonding may take place, though this requires a third firing. 

Another resource important to production of earthenware purifiers is intensive labor, at the smallest scale about fifteen minutes total handwork necessary to each purifier. But in places where the purifier is needed, wages tend to be very low, and there are many people in need of the work. Given this kind of economy, the poor can be both the producers and the buyers. Low labor cost input is largely what makes the purifier affordable, about U.S. 30 cents for a candle projected to last three to six months. The required silver costs no more than about 2 cents (U.S.$0.02). Bangladesh has an excellent system of micro credit. This could make the two plastic buckets necessary to the system something that is available to the poorest of the poor on a micro-credit basis. It has been said that in Bangladesh 80% of all illness is caused by drinking unpotable water, fecal coliforms posing the number one problem. In fact, somewhere in the world someone dies of a water related illness every twenty seconds. 

As to fabrication of the purifier, where there is little or no sophisticated machinery, following are the steps necessary to fabrication. Note that in Bangladesh it is quite possible to find skilled and educated workers to carry out the careful record keeping required in this kind of processing. 

1. The composition to begin with is: 37 parts flour, 63 clay and 30 water (excepting for the black clays, described following). However, after all these are mixed, the water should be dried off over two to four hours, final water content to be about 15%. Number 4., below, indicates the time for this step. 

2. The mixing order is important, first combining the 63 parts of clay powder with the 30 parts of water. After hand mixing, this combination should be closed in a bag over night. This pre-mixing of water and clay helps restore some of the plasticity lost in drying. Note that if the clay flour and water were all mixed at the same time, the flour would grab most of the water, almost none going to the clay. 

3. The following day mix the 93 parts of clay/ water with the 37 parts of flour. This involves a lot of intensive handwork, breaking up the lumps of clay in order to combine in the flour. At the beginning the material appears white, the flour adhering to the clay lumps. It darkens a bit as the clay lumps are broken up. The last step of mixing is to push all the material through a 10 mesh screen. This step gets any remaining, bigger lumps of clay broken up to some extent 

4. As in number 1., the material is now ready to spread out on plastic, to dry back to the approximate 15% water content. 

5. The material is now ready for pressing grog, kept closed in the bag until needed. This should be done within about twenty four hours, otherwise the mix ferments and smells terribly. A car jack is used to press 1.0 cm. thick disks, appropriate for stacking in the kiln. 

6. Now having the prepared grog, it is ready for firing with little drying. But the temperature where slow care is needed is not that which is typical of most ceramics, at 100C, rather it is around the ignition of the flour, at about 500C. Going through this temperature too fast causes bloating, the disks crumbling to powder, thereafter receiving an uneven firing. Following burnout of all the flour, indicated by a lot of smoke, the kiln can be turned up quickly. 

7. Next that the grog disks are hand crushed to 10 mesh. All the strength of a fired earthenware is present, but with less ‘spot welding’ (the bonding of adjacent clay particles, brought about by firing) crushing is not so difficult. 

8. Final composition for the purifier depends to some extent on the clay used. Black clays require less flour than those that naturally occur as red, yellow, etc. This is because many black clays already contain between three and five percent organic material. A clay that is not black could have a final composition of 50% grog, 40% clay and 10% flour. For a black clay the composition may be 50% grog and 50% clay. 

9. The purifiers are pressed then fired, colloidal silver applied thereafter. 

Note that for black clays by comparison with clays that are not black there is a substantial increase in the rate of flow for only a small additional amount of flour. For example, 37% flour to 63% black clay gives double the rate of flow of the non black clay of the same composition. This means that to get the same rate of flow, as per the composition of the grog, the black clay should contain 33% flour to 67% clay. 

We should also bear in mind that the earthenware candle approach is appropriate technology, developed in a third world setting, and with little recourse to research as to just why things work the way they do. For example, in early development it was thought that a simple, dry pressed earthenware composition would give a very permeable filter medium, lending itself to control of the flow rate. After all, poor permeability in conven-tionally formed earthenwares is primarily due to high water content, ast about 30 to 35%. On a micro-scale the clay particles are pulled together. By contrast, for ‘dry pressed’ clay compositions, where water content is very low, there is a kind of friction on a micro-scale, which pushes clay particles apart. This could help bring about good permeability. 

But in the last step of the forming process for appropriate technology filter candles, in removing a small tapered, wooden spindle from the inside of the purifier, those candles having the desirable 3 to 4% water content did not have adequate strength to withstand the sudden jolt involved. (See the URL of the web site of Ceramiques d’Afrique, given in references, below, for a link to a tour of steps in production.) So to insure the green strength necessary to forming at this stage, it was important to increase minimum water content to about 8%, too much to support any substantial improvement in flow. When this became understood the approach offered by permeable grog was undertaken, development of a red firing clay that had both porosity and permeability. 

For permeable grog purifiers, looking under the microscope it is clear that with respect to permeability the filter medium is not homogeneous. The micrographs indicate that permeability varies from clump to clump, or particle to particle within the purifier. The most permeable material is the porous grog, the least permeable the plastic clay necessary in facilitating forming. Control over final permeability comes in formulating a mix of materials, an averaging that ends up giving the desired flow, the 1.0 to 1.5 liters per hour necessary to the needs of a small family. 

In large part the control over the rate of flow is aided by the proper mixing of particles of different permeability. However, in future work more homogeneity may be considered desireable. For this purpose, some simple machinery may be developed as an aid in the processing, but no design for a mixing device comes readily to mind. The mixture of a plastic clay material with a glutinous bread like material, all with water, would appear to be unlike any other process in ceramics. At present the best aid to the hand process is an old piece of rubber tire, used in the final mixing to push the material through a 10 mesh screen. 

Indeed, no screens other than 10 and 30 mesh are used, sizes that are commonly available in most places. And the same kind of availability of resources is true of all the inputs needed. Dies are made of concrete and iron, the force for pressing from an old car jack, etc. The aim is a widely applicable model, which can be altered and improved upon, depending on location. 

An additional goal is to help dispel an unfortunate perception among developers that ceramics is esoteric technology. This way of thinking may come in part from the wide variation in materials and resources from place to place. But it is important that the ceramics be simplified to the point where those at risk of water born illness can benefit by the seventy year old invention of silver in ceramic media. The fact is that the oligodynamic action, killing of pathogens within ceramic filtering media may offer among the very best of low cost options for water purification. In production there is no high cost input implicit. 

REFERENCES 

The web site of Ceramiques d’Afrique, URL: http://www.geocities.com/ceramafrique/ 

The web site of the Potters for Peace Filter, URL: http://www.potpaz.org/pfpfilters.htm

The Katadyn patent, 1930 

Results of particle tests, Lance Edling, Westpoint 

Results of bacteriological tests, The Dhaka Water Supply and Sewarage Authority