AT a meeting of the association at Indianapolis, in May, 1896, I read an article on “Filtra ion” in which I made the following statement: “I know of no reason for the use of a filter bed, whether original cost, expense of maintenance, available space, efficiency of filtration, removal of bacteria, facilities for cleaning, etc., are considered, which reasons will not apply with much greater force in favor of mechanical filters, provided only that the proper form of mechanical filter is used.” My experience of two years since that time has confirmed those views, and I am pleased to note that others, among the n Edmund B. Weston,of Providence, R.I., and Nicholas Simin, of Moscow, Russia, seem to have arrived at similar conclusions.

As probably some of my hearers have never seen my original article, I trust I will be pardoned if I repeat some of the statements then made.

The city of Lexington, Ky., is supplied with water by an impounding retsrvoir of about 120 acres in area dependent upon the rainfall and the drainage area of a1 out 3,000 acres for its supply. During the hot months from June to October inclusive, our reservoir gradually becomes filled with algmand vegetable matter, at times in hot weather emitting an unpleasant odor. We also occasionally have an influx of mud after ⅛ heavy rain; but this last we regard as a trivial matter in comparison with our troubles with algce, and the problem that has confronted us has been not merely the removal of mud, algae, and foreign substances, but especially the removal of odors.

Our experience leads us to believe that deodorizing cannot be done by filtration alone, and that aeration must be relied on for this purpose.

We believe we have succeeded in this object in a simple inexpensive way, and after three years’ experience we cannot think of any change to be made in the theory or plans in which we have worked,except possibly to enlarge and carry the ideas still further.

We use mechanical filters of the type known as the “gravity filter.” We adopted this style because it is open to inspection the same as a filter bed, or even more so, and because the rate of filtration, like that of a filter bed. is dependent on the head of water and the filtering area. It also has the merit of being open, and affording an opportunity for the escape of all gases and odors, and entirely prevents carrying air into the water mains and the formation of air pockets.

♦Paper read at the eighteenth annual convention of the Amer.can Water Works association at Buffalo, N. Y., June, 1898,

The pressure filters (so-called) seam to me to be based on faulty theories.

First.—The resistance must increase as the film on the surface of the sand thickens, and an ever growing and uncertain pressure must be maintained to force the water through. This means not only decreased efficiency of filtration, but also increased coal bills.

Second.—Pressure tanks are usually of iron, and however dilute the coagulant used may be, yet if alum or sulphate of alumina is used, the result must be to corrode the iron of which the tanks are made.

Third.—There is no opportunity for aeration or for any gases or air to escape. Whatever air or gas goes into a pres, sure filter must appear in the mains in the shape of air pockets, or escape at the faucets where delivered to the customers, sometimes much to the annoyance of the customers’ olfactories.

For the above named reasons we have adopted the “gravity” filters at Lexington, and contend that there is no result which it is possible to attain by the use of filter beds that we cannot attain by their use. There is certainly some number (more or less) of gravity filters which will produce the same result as so many acres (more or less) of filter beds, and if the original number of filters is not is an easy matter to add to the number.

The use of a coagulant very much increases the safe rate at which water may be filtered. The film which form? on the surface of the filtering material and which is so effective in thorough filtration forms rapidly when a coagulant is used,and slowly when it is not. There is,therefore,little or no lost time with a mechanical filter as compared with a filter bed.

Still there are limits which cannot be safely exceeded with any filter whatever, and I am inclined to think that in many cases, the same as with filter beds,too much has been expected from too little area, and that manufacturers of mechanical filters, and their customers also, from motives of economy have endeavored to get along with too small a filtering surface.

There is no reason that I can see why sand in a filter tank will not do all it will do in a filter bed, and such additional amount as is due to the use of coagulants, and the shorter time lost in cleaning.

But there are some advantages to be derived from the use of mechanical filters which should be taken inty account. Among these are economy of space and also original cost and the substitution of machinery for manual labor. Why should a week be spent in cleaning where a few hours with suitable machinery would suffice? Again, the mechanical filter offers facilities for sterilization by live 3team and for thorough cleansing of the sand used by washings which do not seem practicable with a filter bed.

The special features of our own plant and which we think the principal cause of such success as we have had in filtration and especially in the removal of odors are the thorough aeration given to the water and the mode in which we add the coagulant.

We aerate by letting the water fall in a spray of about 1,600 fine jets through a perforated pipe, a distance of about seven feet to the filter below. After passing through the filters it passes through a charcoal vat with a perforated bottom, and again falls in a spray to the clear-water basin.

This simple method of aerating has been quite effective and has entirely removed all odors. We have an air compressor in reserve which was originally designed to aerate it by blowing sprays of air through the sprays of water; but so far we have had no occasion to use it, the aeration by spraying through the air being sufficient and saving the cost of running the compressor.

The depth of sand in our filters is thirty inches and the annual loss about six inches. We use white sand (Horn Island), which we get from New Orleans at a cost of a little over a dollar a barrel, making our annual expense for sand about one hundred dollars. The upper six inches of sand seems to do all the work,and it is seldom it is stained below that depth. We wash from four to five times in twenty-four hours in hot weather, and use water quite freely in so doing, as none of it is lost, but all gets back to the reservoir after a settling and straining process.

We have experimenced with sand and crushed quartz of various degrees of fineness, but hav” found little or no difference in the results; the film or coating which rapidly forms on the surface of the sand seeming to do the woik about the same in either case.

We have also experimented with settling basins with a view of taking part of the work off the filters, but found it impracticable for the reason that the algae and vegetable matter with which we have to contend, is so slight that it will not settle, but floats or remains in suspension, leaving us no alternative but to let the filters do all the work.

As regards the us of a coagulant, I think’much of its effectiveness depends upon how it is used, and I cannot, I think, do better than quote from my former report.

We have a little simple device somewhat on the principle of an injector worked by the rock shaft of our low service pump and worked automatically, which at every stroke of the piston injects a fixed and invariable amount of the solution. It is only necessary to vary the strength of this solution from day to day, or even from hour to hour, if necessary, and it needs 1.0 further attention. The results are perfectly uniform it requires much less coagulant; and is positive and certain irrespective of variations in pressure or speed.

The sulphate of alumina which we use as a coagulant varies in amount from a minimum of one third of a grain per gallon to a maximum of one grain per gallon, varying with the condition of the water and the season, and will not average more than eight-tenths of a grain per gallon for the entire year. This small amount (small when the character of our water is considered) we attribute to our mode of injecting the coagulant,by which we can vary from one grain per gallon to the thousandth of a grain with the utmost accuracy and regularity.

I estimate the cost of filtration and aeration at $6 per million . a!Ions, exclusive of interest on the cost of the plant, which was $27,000; but this is for an output of only a million and a half gallons per day. Better results should be obtaintd where the filtering is done on a larger scale,and where cheaper coal can be obtained. I have figured the coal at $2 per ton.

The small output of one and one-half million gallons per day in a city the size of Lexington is accounted for by the fact that ninety-five per cent of our taps are metered, and there is therefore no waste. In this connection I may remark that meters and filters properly go together, and should not be separated.

With regard to typhoid fever statistics, to which so much importance seems to be attached lately, we have no reliable data in Lexington; but this much at least we do know—that whatever cases there have been, were attributable to the use of wells and other sources cf supply. In thirteen years there has not been one solitary case of typhoid fever, or any other disease which could be traced to the use of reservoir water. Such a charge has never been made against it.

There are some other matters which we think increase the efficiency of our filtration,to which I will briefly call attention.

First.—All our filtered water is stored in a clear-water basin and is carefully excluded from the light and from any possible contamination from germs being blown into it, We find in our warm climate that any exposure to air and light results very quickly in what almost seems to be spontaneous generation.

Second—We provide at the lowest points in our mains, openings which we call “washouts” and through which we flush and completely empty the pipes at intervals during the hot weather. By opening and closing the proper valves, and putting on pressure at the pumphouse we can scour almost any line at pleasure, refilling, of course, with fresh water. These “washouts” are much more effective than hydrants for flushing purposes, the discharge being directly from the main instead of from a point three or four feet above it.

The results thus far have been quite satisfactory to ourselves and customers We have had almost no complaint whatever of the quality of the water since the installation of our filter plant, while before that time complaints were frequent.

I have brought with me a sketch of our coagulant injector, and samples of our water before and after filtration, which I will be pleased to show to any one or give any information in my power. I will also be glad to welcome at Lexington any members of the association who desire to make a personal inspection of our plant.


No posts to display