On the positive side, lab results indicate no problem for samples tested by atomic absorption, for the presence of silver in the filtered water. Silver is found in only a few candles, in barely detectable amounts.  But could there be other health considerations?  Otherwise we proceed in an interdisciplinary spirit, in the urgency that the vulnerable deserve.

The motivation to study alternatives is recent, because several months ago we began experiencing some unprecedented failure, small percentages of e coli not being removed.  While working to compensate, getting back to 100% effectiveness we have learned that for pottery purifiers there is a need for a lot more silver than the 200 part per million that has been used for slower, other types of filters.  When flow rate per surface area is so high as that of the pottery purifier, a small percentage of e coli may become dynamic, or at least nimble, thus passing into filtered water.  Given the fast flow, these bacteria pass because there is inadequate time to be trapped by the ceramic as well as insufficient exposure time to the silver.

To achieve a greater flow rate we simply increase water column height.  The household system has 28 cms. of column height, from container base to top.  So for a number of candles the column height was increased to 46 cms.  For these candles flow increased, upto double the rate of the 28 cms. system.  Because of this, at the higher column there is increased expectation for the disinfection by silver. 

It could be concluded that candles intended for the household system should pass an overkill test, proving 100% effective in systems of the 46 cms. column height.  This kind of quality control would help in designing household systems intended to meet a demand for an increased supply of purified water.

Application procedures with silver nitrate and with silver chloride meet different, important demands in the practicality of their production, so both forms of silver are presented.  On the one hand, AgNO3 can be mixed in with the water used in the mixing of the clay composition.  Or the silver nitrate solution can be applied to 'already-fired' purifiers.  Either way, these purifiers are subsequently kiln fired to at least 900C, at which point the silver bonds to the pottery.  In the meantime the nitrates have burned off at 500C.

Because of the earlier problem of insufficient silver in the purifier, in ppm, with subsequent, bigger amounts we begin to refer to silver per purifier by weight.  This is done for several forms of silver, the alternatives to the colloidal silver (CS).  In the new, alternative methods, small particle size of silver is also assumed.  Note that it is CS saturation that has had widespread use in the production of other kinds of low-temperature filters.  A widespread practice for saturation has been in the use of highly concentrated CS.  When the CS dries out within the medium, this ends up with silver metal and silver oxide.  But though CS is an excellant choice there is a serious disadvantage.  Since CS is imported it tends to be expensive or hard to acquire.  Thus it is important to be able to produce some other form of silver, a form that also gives tiny particles.

Meanwhile, test series for silver nitrate and silver chloride are just several of the forms of silver, and their processing methods, which are under consideration.  Other methodologies include several different kinds of electrolysis, as well as a mock photo process.  But for the current purifiers of AgNO3 and AgCl some sixty membrane filtration tests have indicated that these alternatives should be highly effective toward e coli removal, for example.  And for perspective as to the cost of the silver per purifier, the market price of the metal itself is only about US$0.03.  This 3 cent cost is prior to any further processing of the silver, such as production of the silver nitrate salts.

From A., with too much silver, at too high a cost, we proceeded with testing, through C.  Thus we begin to arrive at a sufficiently low cost of silver per purifier, which still indicates 100% effective removal. 

Not so clear in the graph is the number of candles per group.  With a bit of math we find that for some of the C candles the silver, by weight, is about one third that of A.

Again, the candles shown in the picture above were saturated with a solution of silver nitrate, then dried, before being inserted into the system.  Then very salty, salt water (sodium chloride) is run through these silver nitrate candles.   Thus soluble silver nitrate is replaced with relatively insoluble silver chloride.  NB. Silver chloride has a solubility of 0.000089 grams per 100 cc in cold water.  By comparison one form of silver oxide is 0.0013. 

For the candles shown in the graph:
A.  For these candles, of 0.5 grs. AgNO3, the resultant silver, eg. metal, per candle is 0.32 grs.
B.  Given 0.44 AgNO3 salt applied, remaining silver per candle is 0.28 grs.
C.  Given 0.25 AgNO3 salt applied, remaining silver per candle is 0.16 grs.  However for one sub-group the silver metal per candle is 0.11 grs.

In conclusion, of two groups of purifiers, containing,  1.  silver nitrate, and, 2.  silver chloride,  these can be 100% removal of ecoli.  And these purifiers are low cost.  To first appearances there is little problem with the use of ceramic water purifiers that have been treated with silver nitrate and others with silver chloride.  But being circumspect considering the aspect of health, there is presently no plan to begin the use of such purifiers.  Because of the urgency for the vulnerable these prospective methodologies, and others, might be undertaken and considered in an appropriate way.