TO serve as a basis of discussion, I desire to outline in brief terms the present status of the important subject of water purification. The practical significance of supplying water consumers with pure water is very great, and the importanceof this phase of water works systems, which is receiving much attention in many localities, is so clearly understood by the gentlemen of the association that further comments here are unnecessary.

Starting out with the proposition that water consumers should be supplied with pure water, it is obvious that the ways and means by which this can be accomplished depend largely upon local conditions. In some instances it is practicable for municipalities to draw a practically clear and pure water either from underground sources or from watersheds comparatively near at hand, where the drainageareacan be maintained substantially free from sources of pollution. When local conditions makesucha procedure practicable, there is no room for doubt but that on general principles it is wiser to do so rather than to purify a less distant supplythat is to say, a supply of pure water obtained at a reasonable cost from sources which can be kept free of pollution is unquestionably more desirable than a water supply which must be purified, even though the lattercourse would be somewhat cheaper. Unfortunately for the municipalities, there is a goodly number of them which are unable to obtain a pure water supply within reasonable distances, and, with the steadily increasing pollution of rivers ami lakes due to the increasing population, the day is not far distant when the majority of cities must face the question of water purification.


The problem of water purification is nota new one. But the necessity of considering it from a practical standpoint at various localities possessing widely different local conditions has developed the fact within the |»ast few years that the problem is not a simple matter in many instances. In fact, in some cases the question is a much more difficult one than was generally believed a few years ago.

At the present time more or less definite information is available to serve as a guide, and comes in part from municipal purification works which supply in the neighborhood of 25,000,000 water consumers daily. Most of those works are located abroad where the unpurified water is more or less polluted with sewage, and contains relatively small amounts of silt and clay. With regard to the enormous quantities of very minute clay particles, such as are frequently found in many American rivet wnteis they are practically absent in those waters which are purified with success in Europe.

To define in a practical way those waters which are actually purified abroad, it may be stated in general terms that they represent a type of water which can be filtered satisfactorily through sand-layers about three to five feet thick and at a rate of filtration of about 1.5 to 3 million gallons per acre daily, either with or without preparatory treatment of the water by plain subsidence for one to five days—that is to say, they belong to a type of water which can be purified by what has been culled the English system of purification.

* Paper read at the nineteenth annual convention of the American Water Works association, Columbus, Ohio, June, 1899.

Concerning the merits of the English system of purification, there is very little reason to doubt, so far as my knowledge of the general subject goes, but that for those waters for which it is applicable it is the cheapest and most efficient method of purification, except, perhaps, in a few instances where the local conditions are quite abnormal. But to decide whether or not the English method is applicable for a given water is not an easy problem in some instances for water works engineers to solve, especially those who have to deal with river waters in this general section of the country. Theoretically it is possible to purify any water by the English method by taking sufficient advantage of preliminary subsidence, thick sandlayers, or slow rates of filtration, or, what is more reasonable, by using a combination of these three helpful steps. Practically, however, the situation is not so simple in many cases, because the procedures above stated would become excessively expensive.

Turning from those relatively clear waters which are filtered abroad to the very muddy and turbid rivers, of which the lower Ohio river is typical, it is found that European experience is not a safe and adequate guide for the construction and operation of purification works. The reason of this lies in the behavior of the large quanities of clay particles, not more than 1-100000 in. in diameter, which are present in the water for weeks at a time during the period of heavy freshets. Experience has shown clearly that for this type of water the English system of purification is not suitable, when the period of subsidence, thickness of sand-layers, and rate of filtration are carried to practical limits, due to a number of reasons of which the principal ones are as follows: (1) During prolonged freshets the effluents become very turbid and have a brownish’ appearance; (2) towards the end of periods of markedly turbid effluents the filters show a diminution in bacterial efficiency ; (3) the penetration of fine clay particles into the sand-layers leads to complications in cleaning the filters and makes the cost of operation abnormally large as compared with European data.

While it is true that with such waters as that of the lower Ohio river the English system would give per fectly satisfactory results, except during heavy freshets, and on such occasions produces an effluent which is vastly superior to the river water, yet the process cannot properly be regarded as a success, and in my opinion it is not justifiable on the strength of present evidence to spend large sums of public funds for works which cannot fill the requirements for which they are to be built, at times when they are needed most.

In the course of some investigations conducted by the water works commissioners of Cincinnati, a modification in the English system was introduced, with the view of assisting the process at times of heavy freshets. This modification consisted in the use of a coagulating chemical to assist subsidence when the river water wits very turbid, to afford the water an adequate preparatory treatment prior to filtration through English filtei s By the use during freshets of a coagulating chemical, and by allowing the coagulated particles of clay to subside before the water passed on to the filters, a very satisfactory effluent was obtained. But the process was somewhat difficult of management,owing to the possible passage of coagulated masses of clay on to the sand-layer in amounts sufficient to clay the surface of the sand quickly, and it was somewhat more expensive than other methods under consideration.

This brings to our attention the question of the use of chemicals in connection with water purification. That the use of chemicals for this purpose is undesirable there can be no doubt, and there are probably very few water works engineers who fail to recognize that chemicals should be avoided in water purification, whereever it may be practicable to do so. Unquestionably a large number of water consumers, water works officials, and health officers, are prejudiced against the use of chemicals. In a large measure such views are based upon preconceived and erroneous impressions in the absence of a thorough understanding of the action of the applied chemical. Coagulating chemicals as applied to the type of water in question have the power of forming the very small bacteria and the still smaller clay particles into aggregates of such size that they can be handled to a much greater advantage than in the original form In a practical way the effect of coagulation is illustrated well by the familiar action of the white of an egg upon turbid coffee. There are a number of cheap commercial chemicals which are capable of being used for coagulation purposes. Just what chemical is best to use for this work depends somewhat upon natural local conditions and market prices, and in my opinion is not a matter which at the present time can be fin illy settled for all cases. During the Cincinnati experiments already referred to, sulphate of alumina was used as the most suitable for the purpose as judged from the evidence available at that time, although it wus not considered to be necessarily the cheapest. As is probably well known to you all, the clay-hearing water of the Ohio river contains considerable carbonate of lime, and it is largely due to this fact that sulphate of alumina can be used safely and satisfactorily in the treatment of this general type of water.

Directly, sulphate of alumina does not produce the required coagulation, and the direct passage of this chemical in its original form through a purification system is inadvisable for many reasons. The practicable benefit coming from the use of this chemical depends upon its decomposition by the carbonate of lime present in the water. This action has been explained in detail in a number of recent reports. Here it is only necessary to point out that, as a result of the decomposition,there is found aluminum hydrate— a white gelatinous solid which affects the coagulation of the bacteria and clay; and for each grain of applied sulphate of alumina there is approximately half a grain of carbonate of lime converted into sulphate of lime, with the liberation of about 1-5 of a grain of carbonic acid. Thealuminuin hydrate in the form of flakes is removed by subsidence or filtration, while the lime in its modified form and the carbonic acid pass into the filtered water. When sulphate of alumina is applied in large quantities, the conversion of the lime into the form of sulphate and the liberation of carbonic acid become matters for serious consideration. This is due to the fact that the sulphate of lime forms a more objectionable scale in steam boilers than does the carbonate of lime, and that carbonic acid facilitates the corrosion of uncoated iron. When sulphate of alumina is applied at the rate ordinarily of about one to two grains of chemical per gallon of water, and rarely over three grains, the effect of the application, practically speaking, is very small indeed along the two lines mentioned, and, at times, when the quantity applied would be greatest, the filtered water would not, as a rule, be less desirable for boiler use than the river water during stages of low water. Concerning the effect upon the human system of the compounds resulting from the decomposition of the stated quantities of sulphate of alumina, the evidence shows that it is practically nil. With undecomposed chemical—that is, quantities in excess of that which can be broken up by the lime in the water—the situation is different. But, as already stated, the presence of undecomposed chemit in the filtered water is absolutely inadmissible.

One of the advantages of the use of a coagulating chemical is that it prepares the water so that it can be filtered through a sand-layer at a very rapid rate. For this purpose so-called mechanical filters are used, and for such water as that of the Ohio river they form the last step in the process which has been called the American system of purification. For a number of months this system was investigated at Cincinnati. The process consisted in allowing the river water to subside from two to three days; applying sufficient chemical to coagulate the subsiding water; affording the chemically treated water to coagulate and subside for a supplementary period of 0.5 to six hours, according to its character; and then filtering the partially clarified and thoroughly coagulated water through a sand-layer at the rate of about 120,000,000 gallons per acre daily. Satisfactory results were obtained by this method, and the difficulty in management and the cost werefound to be somewhat less than by the modified English system already mentioned. The effluent was clear, low in its contents of organic matter and bacteria, absolutely free of undecomposed sulphate of alumina, and a satisfactory water for use in steam boilers. Further information will soon be available in the forthcoming report in full upon the work.



Very recently there was investigated at Cincinnati a new process of purifying clay-bearing water after the general method of the American system of purification. As this subject has not yet been reported upon, I am not at liberty to speak of the detailed in formation obtained from it, but in brief and general terms the procedures may be outlined as follows: This purification plant, which was constructed and operated by the Ohio Sanitury Engineering company at its own expense, was tested officially by the city of Cincinnati for forty-two days, from March 13 to April 25, 1899. The process in its original form consisted essentially of three principal steps, as follows: (1) The treatment of the original river water with lime water; (2) the precipitation of the coarser particles in the water after the lime treatment; (3) the rapid filtration of the partially clarified and thoroughly coagulated effluent from the precipitation tank.

With regard to the action of the applied lime water upon the particular river water, it may be stated that it produced coagulation, and the bacteria and clay particles were aggregated into relatively large masses, thus facilitating rapid precipitation and rapid filtration. To accomplish this result with the river water in question, it was found to be necessary to apply lime water in sufficient quantities to combine with the free and half-bound carbonic acid of the water, and also to afford a sufficient excess of lime water to produce the coagulation. Whether or not it would be necessary to use an excess of lime water in the treatment of all water by this process was not brought out during these tests, although it is said that this would not be the ease with water of more favorable composition. The precipitation tank was operated on the continuous displacement plan, and the period of subsidence was found to be twelve hours, which was found to be adequate when the river water was in its worst condition. With a suitable quantity of applied lime water, adequate to the character of the river water under treatment, the effluent from the precipitation tank was well clarified in consequence of the thorough coagulation. By virtue of this coagulation, the small percentage of bacteria and suspended matter which passed through the precipitation tank were well prepared for satisfactory removal by rapid filtration through a sand filter The sand filter was equipped with mechanical appliances to assist in cleaning

These devices, aided largely by the cleanliness of the applied water, caused the amount of wash water required to be a very low percentage of the water treated. Under efficient management, which is, of course, necessary for all systems of water purification, the effluent of the sand filter was perfectly clear in appearance and very low in its contents of organic matter and bacteria. Concerning the excess of lime water which was used during the test, for reasons already explained, it may be stated that it passed through the system and appeared in the effluent of the sand filter in a practically undiminished quantity. The presence in the effluent of the sand filter of lime water in excess in such quantities as were used during these tests is inadmissible for a number of reasons; and it is necessary to remove it in order to obtain an efffuent of satisfactory character. On a large scale it is said that this would be accomplished by the application of carbonic acid obtained from limekiln gases from limestone burned at the plant; and the subsequent removal of the carbonate of lime formed therewith. During this test, however, the action of the gas was demonstrated by the use of carbonic acid obtained and applied in a liquefied form. For reasons already mentioned, comparative statements as to its efficiency and economy must be deferred for the present _

No posts to display