Only now is the Nepal project generating the appropriate 'colloidal silver'*, necessary to purifier saturation, as per the methodology given to date (and following).  Several prior attemps at making the CS failed because the distilled water had total disolved solids (TDS) in excess of the requisite maximum, 10mg./ liter.

*  Recent conversation in the internet, 'CS' discussion groups indicates that this term is a mis- nomer.  We will continue referring to 'colloidal silver,' as the accepted term, but as ionic, Ag+  in solution, this should more appropriately be called 'Electrically Isolated Silver.'

As indicated at the bottom of the Erlen- meyer flask, this batch of concentrated colloidal silver is a transparent, dark red.

Where the two prior types of water for which attempts were made were purchased as 'distilled' for the first try and 'de-ionized' for the second, neither gave good results.  Unlike the water of the first two attempts, for the third, and successful attempt the water was first tested for TDS, and indicated 7.0 mg. per liter.

Tests for the new colloidal silver have indicated this is 255 ppm.  Purifiers subsequently saturated with the concentrated CS that is shown have been inset in sample systems, these given to several humanitarian organizations. 

This widely replicable process could be put in place in a principal city of almost any developing country, all the resources present in most locations.  Mexico is a country where a number of brands of highly con- centrated CS are made and are available, and a similar colloidal silver could be made in most places.

In following this production model it can also be possible to produce concentrated CS for easy shipping within a country.  Then the purifiers can be saturated closer to the point of sale, something that would better enable rural production of purifiers.  Following is a list of materials and resources needed for an appropriate, low voltage, 27volt DC, concentrated CS generator.  After that is a description of the use of the generator, and outputting of the CS. 

Please note that aside from the wage of the worker(s) who operate the generator, the primary material cost is for the water, both deionized and mineral free.  For the CS produced, at ~170 parts per million, the amount of silver given off at the electrodes is very tiny, no more than perhaps 5 cents (U.S.$0.05) per two liters.  So the cost of silver would be negligible.

1.  A two liter erlenmeyer flask, with rubber stopper (having three small holes in a row).
2. A thermometer for the flask, indicating to just above the boiling point of water.  This occupies the center hole of the rubber stopper.
3.   Two pure silver (minimum 99.9% purity) electrodes.  An once of pure silver, ingot or coin, can be formed into a strip, approximate 30 cm. length.  These occupy the two outer holes of the rubber stopper.
4.   An electrical convert that outputs 27 volts DC, most simply from three, good quality nine volt batteries.  Wire leads from the batteries, through a DPDT switch (i.e. 'on- off- on' switch), with alligator clips for connection to electrodes.
5.   A stainless steel cook pot, five to six liter capacity, including a stand-off rack, to hold the flask one cm. or two above the bottom of the pot.  This can be a stainless steel rack, or even a collection of (30 or 40) non-breakable beads.  This standoff allows bubbles forming in the water under the flask to escape without moving the flask.
6.   A hot plate, with, preferrably gas, to maintain this double boiler system.
7.   Two liters de-ionized water is required for the flask.  By quality, total disolved solids of this water should be 10 or less.  In addition, several liters of mineral free water is required per hour, for replacement of the water vapor given off to boiling in the cook pot. For example, maintaining the generator at about 200 degrees fahrenheit (or about 93 degrees centigrade) , over four hours requires about 8 liters of mineral free water.
8.   A timer, preferrably to give alarm every one minute.

1.   Fill the stainless steel cookpot with the two liters of deionized water and bring to a boil.  Note that prior to doing this all surfaces, the inside of the pot and its lid and inside of flask need to be properly washed and rinsed several times with mineral free water. 
2.   In a separate, properly washed pot, on a separate burner, boil about five liters of mineral free water.
3.   When both of these two pots of water are boiling, start by pouring the 2 liters of deionized water into the flask.  Place the flask on the standoff rack, in the stainless steel cook pot.  Pour into this pot the five liters of mineral free water.  Note that use of mineral free water is advised because heavy mineral content otherwise accumalates while more and more water escapes as vapor.  This would lead to mineral encrustasion of the flask, pot and standoff, which would inhibit the boil and the life of the items that are exposed to this. 
4.   Since, after boiling, the temperature diminishes as water is moved, container to container, take a bit of time to restore the temperature just under the boiling point, to 200 to 206F  (93 to 97C).  Prior to closing the rubber stopper pour a bit of starter (see no. 7. below) into the flask.  This could be a small amount of previously made CS, enough to get the contents of the flask upto about 2 ppm.  In using two or three tablespoons of the previously made CS, the flask should be filled just to the neck.
5.   Connect alligator clips to the silver electrodes and flip the DPDT switch to either of the two 'on' positions.  As the timer gives alarm every one minute, flip the switch to the other of the two 'on' positions.  Do this 'polarity switching' over the entire three or four hours of operation.
6.   Operation of the generator over three or four hours should result in ~170ppm CS.  It should be understood that it is not really possible to give a precise indication of ppm per time, in large part because of the absence or presence of electromagnetic fields, whether in large or small amounts.  Presence/ absence of electromagnetic fields cause generator operation to vary, plus or minus about one third of the time. 
7.   Previously made CS starter is an aid to initial conductivity of the water, which increases reaction time of the electrodes, getting ionic silver , Ag+, into quicker production. Otherwise, without the starter an additional hour or so may be required.  Note that the generator tends to produce about 50ppm of CS per hour, increasing in over all ppm (per time) by this amount.  So for example, continuation beyond four hours would give something like 500ppm over ten hours.  As to visual monitoring of the process it should be understood that a gradual yellowing will begin, probably around 30 to 60 minutes into the production.  Over time this color will increase to yellow, orange, amber to deep amber, alway remaining clear to light transmitted through the
solution.  However, as the solution becomes very dark it turns to gray or green, visually as reflected light.  Light transmitted through the CS is still quite clear but is very dark, likely dark red, but when a teaspoon or so is poured into a liter of water, this will appear to be clear yellow.