DISCUSSION

Allen Hazen, M. Am. Soc. C. E. (by letter).—This paper contains a most interesting and instructive record of the actual operation of a large filter plant, and also a record of a number of experiments. The author has described some useful arrangements for improving the efficiency or reducing the cost.

The utility of raking, as an intermediate treatment between scrapings, seems to have been clearly demonstrated. Its practical effect is to allow a greater quantity of water to be passed between scrapings, thereby saturating—if the term may be used—the surface layer with clay and other fine matter before removing it, instead of taking it off when only a thin surface layer of it has been thus saturated.

The large proportion of the total purification that takes place in passing through three reservoirs successively, holding in the aggregate a quantity of water equal to about 7 days' use, is very striking. Taking all the records, the percentage remaining after passing through these reservoirs, is as follows:

Sediment for the year, 1909-1910, Table 2....................17% Turbidities, 5-year average, Table 3.........................25% Bacteria, 5-year average, Table 4............................24% Bacteria, selected winter months with high numbers in the raw water...................................20% Bacteria, selected summer months with high numbers in the raw water................................... 2.5%

There is considerable seasonal fluctuation in the results of settling and filtering, as is shown in Table 21.

Table 21 Average Removal of Turbidity and Bacteria by Washington Filters for Whole Period, Arranged by Seasons. ===========================+======+====+====+====+====== Winter. Spring. Summer. Fall. Year. + + + + + + Turbidity, in raw 135 96 144 42 105 parts per settled 33 28 27 15 26 million: filtered 4 3 1 0.5 2 + + + + + + Percentage settling 24 29 19 36 25 left from: filtering 12 10 4 3 8 both 3 1 0.3 1 2 + + + + + + Bacteria per raw 16,600 4,150 4,100 1,960 6,700 cubic settled 6,300 980 160 270 1,940 centimeter: filtered 149 29 18 22 54 + + + + + + Percentage settling 38 24 4 14 29 left from: filtering 2.4 3.0 11.2 8.2 2.8 both 0.90 0.79 0.44 1.12 0.81 ==============+========+========+========+========+========+========

The fluctuation in the efficiency of the plant as a whole by seasons is greater with the turbidity than with the bacteria. During the winter the effluent contains 3% of the turbidity of the raw water, and in summer only 0.3 per cent. Most of this difference is represented by the increased efficiency of the filters in summer, and only a little of it by the increased efficiency of settling. With bacteria, on the other hand, the seasonal fluctuation of the plant as a whole is comparatively small, but the settling and storage processes are much more efficient in summer than in winter, the filters being apparently less efficient. The writer believes that they are only apparently less efficient, and not really so, the explanation being that some bacteria always grow in the under-drains and lower parts of the filter, and are washed away by the effluent. The average number of bacteria in summer in the settled water is 160 per cu. cm. and in the filtered water 18. These are very low numbers. It is the writer's view that nearly all of these 18 represent under-drain bacteria, and practically bear no relation to those in the applied water, and, if this view is correct, the number of bacteria actually passing through the various processes is at all times less than the figures indicate. In the warmer part of the year the difference is a wide one, and the hygienic efficiency of the process is much greater than is indicated by the gross numbers of bacteria.

The reduction of the typhoid death rate has not been as great with the change in water supply as was the case at Lawrence, Albany, and other cities, apparently because the Potomac water before it was filtered was not the cause of a large part of the typhoid fever.

The sewage pollution of the Potomac is much less than that of the Merrimac and the Hudson, and it is perhaps not surprising that this relatively small amount of pollution was less potent in causing typhoid fever than the greater pollution of rivers draining more densely populated areas.

The method of replacing the washed sand hydraulically seems to have worked better than could have been reasonably anticipated, and the writer believes that this was due, in part, to the excellent method of manipulation described in the paper. It is his feeling, however, that part of the success is attributable to the very low uniformity coefficient of the sand. In other words, the sand grains are nearly all of the same size, due to the character of the stock from which the filter sand was prepared; and, therefore, there is much less opportunity for separation of the sand according to grain sizes than there would be with the filter sand which has been available in most other cases. Filter sand with a uniformity coefficient as low as that obtained at Washington has been rarely available for the construction of sand filters, and while the method of hydraulic return should certainly be considered, it will not be safe to assume that equally favorable results may be obtained with it with sands of high uniformity coefficients until actual favorable experience is obtained.

The writer believes that in calculating the cost of the water used in the plant itself the price chosen by the author, covering only the actual operating expenses of pumping and filtering, is too low. The capacity of the whole Washington Aqueduct system is reduced by whatever quantity is used in this way, and, in calculating the cost of sand handling, the value of the water used should be calculated on a basis which will cover the whole cost of the water, including all capital charges, depreciation, operating expenses, and all costs of every description. On this basis the water used in the sand-handling operations would probably be worth five or more times the sum mentioned by the author.

The cost of operation of the plant has come within the estimates made in advance, and has certainly been most reasonable. The cost of filter operations has averaged only about 50 cents per million gallons, and is so low that it is obvious that the savings which may be made by introducing further labor-saving appliances would be relatively small. It will be remembered that ten or fifteen years ago the cost of operating such filters under American conditions was commonly from $2 to $5 per million gallons.

The experiments represented by Tables 17 to 19, inclusive, serve to show that preliminary filtration, or multiple filtration, or any system of mechanical separation is incapable of entirely removing the finer clay particles which cause the residual turbidity in the effluent. They also show that this turbidity may be easily and certainly removed by the application of coagulant to the raw water during the occasional periods when its character is such as to require it.

These general propositions were understood by those responsible for the original design of the plant, as is shown by the author's quotations. These experiments, however, were necessary in order to demonstrate and bring home the conditions to those who thought differently, and who believed that full purification could be obtained by filtration alone, or by double filtration, without recourse to the occasional use of coagulant.

The experiments briefly summarized in Table 20 are of the greatest interest and importance. Six small filters, otherwise alike and like the large filters, all received the same raw water and were operated at different rates to determine the effect of rate on efficiency.

That the experimental results from the filter operating at the same rate as the large filters were on the whole somewhat inferior to those from the large filters for approximately the same period, may be attributed to the fact that the experimental filter was new while the large filters had been in service for some time and had thereby gained in efficiency. The greatest difference was in the coli results in Table 20, where it is shown that 24% of the 10-cu. cm. effluent samples from the experimental filter contained coli, in comparison with only from 1 to 3% of such samples from the main filters.

The results from the experimental filter operating at a rate of 1,000,000 gal. per acre daily may fairly be excluded, as the effluent probably contained more under-drain bacteria in proportion than filters operated at higher rates. The number of bacteria in the filter operating at a 3,000,000-gal. rate were 1.7% of those in the applied water; for the filter operating twice as fast, the percentage was 2.4; and, for the one operating more than ten times as fast, was only 3.0; thus indicating a surprisingly small increase in the number of bacteria with increase in rate.

Further and more detailed study by the writer of the unpublished individual results, briefly summarized in Table 20, confirms the substantial accuracy of the comparison based on the average figures as stated in that table.

It must be kept in mind, in considering these results, that the number of bacteria in each case is made up of two parts, namely, those coming through the filter—which number is presumably greater as the rate is greater—and, second, those coming from harmless growths in the under-drains and lower parts of the filter—the numbers of which per cubic centimeter are presumably less as the rate is greater—and these two parts, varying in opposite directions, may balance each other, as they seem to do in this case, through a considerable range. It may thus be that the number of bacteria really passing the filter varies much more with the rate than is indicated by the gross results.

It is also of interest to note that the sand filter (called a preliminary filter) in Table 18, filled with the same kind of sand, when operated at an average rate of 50,000,000 gal. per acre daily for a year, allowed 18% of the applied bacteria to pass, in comparison with 3% found in Filter No. 6 of Table 20, operated at an average rate of 38,000,000 gal. per acre daily.

There was one point of difference in the manipulation: the preliminary filter was washed by a reversed current of water, as mechanical filters are washed, while Filter No. 6 was cleaned by scraping off the surface layer, as is usual with sand filters. Whether the great difference in bacterial results with a relatively small difference in rate is to be attributed to this difference in manipulation the writer will not undertake to state.

If the experimental results of Table 20 indicate correctly the conditions which obtain in filtering Potomac water, then increasing the rate of filtration so as to double it, or more than double it, would make but little difference in the quality of the effluent as measured by the usual bacterial methods. If the increase in rate were accompanied by the preliminary filtration of the water, then, presumably, there would be little change in the quality of the effluent, and the maintenance of excellent results might be incorrectly attributed to the influence of the preliminary filter.

It would also seem that the apparatus which is sometimes used for determining and controlling the rate with more than the ordinary degree of precision is hardly justified by such experimental results as those presented by the author.

In contrast to these results may be mentioned those obtained by Mr. H. W. Clark,[1] for experimental filters operated with Merrimac River water, at rates ranging from 3,000,000 to 16,000,000 gal. per acre daily. The results are the average of nearly two years of experimental work, the period having been nearly coincident with that covered by the author's experiments, and of many hundreds of bacterial analyses of each effluent, and form, with the author's experiments, the most thorough-going studies of the effect of rate on efficiency that have come to the writer's attention.

Mr. Clark's results are given in Table 22.

Table 22. ==========+========+========+============+==========+============ B. Coli in 1 cu.cm. Effective Rate Bacteria per (percentage size of in gallons cubic Bacterial of positive sand. Filter No. acre daily. centimeter in efficiency. tests). -+ + -+ -+ -+ 0.28 A 3,000,000 48 99.1 5.0 0.25 B 5,000,000 85 98.4 24.0 0.22 C 7,500,000 105 98.1 25.0 0.22 D 10,000,000 110 98.0 25.0 0.22 E 16,000,000 280 95.0 38.0 ==========+==========+==========+============+==========+============

It will be seen that the number of bacteria passing increases rapidly with the rate, and whether the total number of bacteria is considered or the B. coli results, the number passing is approximately in proportion to the rate. In other words, doubling the rate substantially doubles the number of bacteria in the effluent.

This is entirely in harmony with all the Lawrence experimental results extending over a period of 20 years. There have been occasional apparent exceptions, but, on the whole, experience with Merrimac River water has uniformly been that more bacteria pass as the rates are higher.

The theory sometimes advanced, that the efficiency of filtration is controlled to a certain extent by gelatinous films, and that, as far as thus controlled, is less dependent on rate, would not seem to be borne out by these results. The Merrimac River water, carrying large amounts of organic matter, would certainly seem better adapted to the formation of such films than the clay-bearing Potomac water, comparatively free from organic matter; but it is the Potomac water which seems to show the least influence of rate on efficiency.

[Footnote 1: Journal, New England Water-Works Association, Vol. 24, p. 589.]

The experiments show that turbidity passes more freely at the higher rates with the Potomac water, as has also been found to be the case with other clay-bearing waters.

In the last lines of Table 20 are given cost per million gallons for filtering at various rates. There is no discussion of these figures, and as they differ considerably from those which the writer has been accustomed to use, the calculation in Table 23, made three years ago for a particular case, may be of interest.

Table 23 Relative Cost of Filtering at Different Rates. ======================+=================================================== Nominal rate, in millions of gallons per acre daily: + + + + 3 5 10 20 + + + + Percentage which average yield is of nominal rate.......... 85 80 75 65 + + + + Average output per acre, in millions of gallons per day....... 2.55 4.00 7.5 13.0 + + + + Cost of that part of filters per acre dependent on rate..... $12,000 $20,000 $40,000 $80,000 + + + + Cost of that part of filters per acre not dependent on rate..... 50,000 50,000 50,000 50,000 + + + + Total cost of filters per acre.............. 60,000 70,000 90,000 130,000 + + + + Cost per million gallons of capacity... 20,600 14,000 9,000 6,500 + + + + Cost per million gallons of average daily output.......... 24,400 17,500 12,000 10,000 + + + + Capital charges and depreciation at 6% on cost per million gallons............... 4.00 2.87 1.97 1.64 + + + + Operating expenses, the same at all rates................. 1.00 1.00 1.00 1.00 + + + + Total cost of filtering, excluding pumping, storage, and all auxiliaries....... 5.00 3.87 2.97 2.64 + + + + Relative cost......... 1.29 1.00 0.77 0.68 ======================+============+============+============+============

When the costs of pumping, pure-water reservoirs usually necessary, etc., are taken into account (which add equally to the cost at all rates), the cost of filtering will vary less with the rate than is indicated.

The effect of rate on cost, as calculated in Table 23, and also the percentages of the bacteria of the raw water found in the effluents by the author and by Mr. Clark, are shown on Figure 10.

Considering all these results together, and also all the other evidence known to the writer bearing on this point, it seems clear that filters are not as sensitive to changes in rate, within reasonable limits, as has been frequently assumed; but, on the other hand, there is usually a substantial increase in the percentage of bacteria passing through a filter with increased rate.

Filters furnish relative, not absolute, protection against infectious matter in the raw water. The higher the bacterial efficiency, the more complete is this relative protection.

The cost of filtering does not decrease in inverse ratio to the rate, but at a much slower rate. This is especially true with rates of more than 5,000,000 or 6,000,000 gal. per acre daily.

In general, a rate of filtration may rationally be selected at which the value of the possible danger resulting from an increase in rate is equal to the saving that may be made in cost by its use. This point must be a matter of individual judgment. The tendency of the last few years has been to use higher rates, or, in other words, to cheapen the process and to tolerate a larger proportion of bacteria in the effluent. The use of auxiliary processes has been favorable to this, especially the use of chloride of lime, in connection with either the raw water or the effluent.



By the judicious use of this substance, efficiency may be maintained while using higher rates than would otherwise have been desirable.

The writer believes that there will be many cases where the added risk of using too high a rate is not worth the relatively small saving in cost that accompanies it.

George A. Johnson, Assoc. M. Am. Soc. C. E.—This paper contains information of an exceedingly interesting nature. There is comparatively little difficulty in obtaining accurate figures on the cost of construction of water purification works, but, with costs of operation of such works, it is different. The data available in published reports and papers are usually more or less fragmentary, and unexplained local conditions with reference to the character of the raw water, the cost of labor and supplies, and methods of apportioning these costs, introduce variables so wide as frequently to render the published figures almost useless for purposes of comparison.

Mr. Hardy's paper is noteworthy in that it presents certain relatively new features of slow sand filter operation which have been only lightly touched on in water purification literature up to the present time. These refer particularly to means whereby a filter may be continued in service without removing a portion of the surface layer of the filter surface itself when the available head has become exhausted, and to methods whereby washed sand may be expeditiously and more economically restored to the filter than has been the case hitherto.

Sand handling is the most important item of expense in the operation of a slow sand filter. Quite recently a charge of $1.50 per cu. yd. for sand scraping, transportation to sand washers, washing, and restoring to the filter, was not considered exorbitant, but the improved methods developed during recent years at Washington, Philadelphia, Albany, and more recently at Pittsburg (at all of which places hydraulic ejection plays an important part), have shown the feasibility of reducing this figure by nearly, if not quite, two-thirds.

The practice observed at Washington of raking over the surface of the sand layer when the available head becomes exhausted, in order to avoid the cost and loss of time necessitated by shutting down the filter and scraping off the surface layer, is unquestionably one of the most striking advances in slow sand filter operation in recent years. In rapid sand filter operation, to prolong the period of service between washings, agitation of the filter surface has been used to advantage for many years. The full value of surface raking may not be generally appreciated, but the results which have followed a trial of this procedure at Washington, Philadelphia, and Pittsburg have shown that the output of filtered water between scrapings may be doubled or trebled thereby, with no injury to the filter itself or to the quality of the filtered water. The cost of raking over the surface of a 1-acre slow sand filter unit is less than $10 at all the above-mentioned places, which fact in itself shows the great saving in money and time effected by periodically substituting surface raking for scraping. Under ordinary conditions it has been found that a filter can be raked to advantage at least twice between scrapings.

In the case of filters thus raked, a deeper penetration of suspended matter into the sand layer is inevitable, but at Pittsburg, as at Washington, such penetration does not extend more than about 2 in. below the filter surface. When the filter is finally scraped, a deeper layer is removed, of course, but it is clearly more economical to remove a deep layer at one operation than to remove separately several thinner layers of an equal total thickness.

The lost-time element is an important one, and at Washington this was the main reason for trying surface raking. It became necessary to increase the output of the filters, and the ordinary scraping consumed so much time that the sand-handling force was increased, working day and night. The raking expedient introduced at this time overcame this, and Mr. Hardy states that it is still followed when the work is at all pressing. The speaker has found at Pittsburg, as Mr. Hardy has found at Washington, that raking is nearly if not quite as effective as scraping in restoring the filter capacity.

Eleven years ago the speaker was connected with the preliminary investigations into the best methods of purifying the Potomac River water for Washington. It then appeared that while for the greater part of the time during an average year the Potomac River could be classed among the clear waters of the East, there were periods when excessive turbidity made it necessary to consider carefully methods of preparatory treatment before this water could be filtered effectively and economically. As Mr. Hardy has said, considerable prejudice existed against the use of a coagulating chemical, and the expedient was therefore adopted of giving the water a long period of sedimentation in order to remove enough of the suspended matter to allow the clarified water to be treated on slow sand filters. The expert commission, consisting of Messrs. Hering, Fuller, and Hazen, recommended the occasional use of a coagulating chemical, but this recommendation was not carried out.

The Potomac River is somewhat peculiar, in that the turbidity of its waters, as shown by the results presented in Mr. Hardy's paper, ranges from 3,000 to practically nothing. The bacterial content also varies widely, and Mr. Hardy's tables show this variation to be from 76,000 to 325 per cu. cm. Such a water as this requires particularly careful preparatory treatment. The Dalecarlia Reservoir has a capacity of something like 2 days' storage, the Georgetown Reservoir the same, and the McMillan Park Reservoir nearly 3 days, making a total sedimentation of more than 7 days. Without the use of a coagulant, it is significant that during a period of five years, even with 7 days' sedimentation, the average maximum turbidity of the water delivered to the filters was 106 parts per million, and the maximum average turbidity in one month was 250 parts per million. The water filtration engineer can readily understand that waters as turbid as this cannot be treated economically and efficiently in slow sand filters. It would appear that coagulating works might advantageously have been installed at the entrance to the Dalecarlia Reservoir. If this had been done, and coagulant had been added to the water at times when it was excessively turbid, a considerably shorter period of subsequent sedimentation than now exists would in all probability have rendered the water at all times amenable to efficient and economical slow sand filter treatment.

The prejudice in Washington against the use of coagulants has also manifested itself in other localities, but the results which have been obtained during the past twenty years from rapid sand filters and from slow sand filters, treating waters previously coagulated with salts of iron or alumina, have shown how thoroughly unreasonable were these objections. In this connection it is interesting to note that there are in the United States more than 350 rapid sand filter plants, and that nearly 12% of the urban population of Continental United States is being supplied with water filtered through rapid sand filters, in connection with all of which a coagulating chemical is used in the preparatory treatment.

Table 24—Typhoid Fever Death Rates in Cities of the United States with Populations in 1910 of 100,000, or More.

Statistics gathered by correspondence and from Reports of the Bureau of the Census, Department of Commerce and Labor, Mortality Statistics.

Note.—Statistics from Birmingham, Ala., Dayton, Ohio, Fall River, Mass., Louisville, Ky., Memphis, Tenn., Oakland, Cal., and Providence, R. I., are not included, as they are incomplete.

Columns: A - Average for six years, 1900-05, inclusive. B - Average for five years, 1906-10, inclusive. C - Average for 11 years, 1900-11, inclusive.

====================+=============================================== ~ Typhoid Fever Death Rate City. per 100,000 Population~. + -+ -+ -+ -+ -+ -+ -+ - 1906 1907 1908 1909 1910 A B C + -+ -+ -+ -+ -+ -+ -+ - Albany, N. Y. 20 20 11 19 15 25 17 21 Atlanta, Ga. 50 64 47 44 43 65 50 58 Baltimore, Md. 34 41 31 23 41 36 34 35 Boston, Mass. 22 10 26 14 11 23 16 20 Bridgeport, Conn. 10 13 13 13 9 15 12 14 Buffalo, N. Y. 24 29 21 23 20 29 23 26 Cambridge, Mass. 18 10 10 9 12 18 12 15 Chicago, Ill. 18 18 15 12 14 27 16 22 Cincinnati, Ohio 71 46 19 13 6 54 31 44 Cleveland, Ohio 20 19 13 12 19 51 17 36 Columbus, Ohio 45 38 110 17 13 61 45 54 Denver, Colo. 68 67 58 24 30 37 49 42 Detroit, Mich. 22 28 22 19 16 17 22 19 Grand Rapids, Mich. 39 30 30 17 27 34 28 31 Indianapolis, Ind. 39 29 26 22 31 76 30 55 Jersey City, N. J. 20 14 10 8 10 19 12 16 Kansas City, Mo. 38 40 35 23 38 48 35 42 Los Angeles, Cal. 18 23 19 18 12 35 18 27 Lowell, Mass. 7 9 24 11 21 19 14 17 Milwaukee, Wis. 31 26 17 21 45 19 28 23 Minneapolis, Minn. 33 26 18 20 58 38 29 34 Nashville, Tenn. 66 85 62 53 48 54 58 56 Newark, N. J. 18 24 12 11 13 17 16 17 New Haven, Conn. 54 30 34 20 17 44 31 38 New York, N. Y. 15 17 12 12 12 19 14 17 New Orleans, La. 30 56 31 25 28 40 34 37 Omaha, Nebr. 28 24 22 31 75 20 36 27 Paterson, N. J. 4 11 10 5 7 25 7 17 Philadelphia, Pa. 74 60 36 22 17 47 42 45 Pittsburg, Pa. 141 135 53[1] 13[1] 12[1] 132 71 104 Richmond, Va. 44 41 50 24 22 66 36 53 Rochester, N. Y. 17 16 12 9 13 15 13 14 St Louis, Mo. 18 16 15 15 14 33 16 25 St Paul, Minn. 21 17 12 20 20 14 18 16 San Francisco, Cal. ... 57 27 17 15 20 29 24 Scranton, Pa. 11 76 11 11 14 18 35 26 Syracuse, N. Y. 10 16 15 12 30 14 17 15 Toledo, Ohio 45 36 40 31 32 36 37 36 Worcester, Mass. 12 14 10 8 16 17 12 15 Washington, D. C. 52 36 39 33 23 59 37 49 ====================+=====+=====+=====+=====+=====+=====+=====+=====

[Footnote 1: Filtered water section. Allegheny District not included.]

Attention has repeatedly been called to the fact that the relatively high typhoid death rate in Washington, since the filter plant was installed, was a possible indication that the filters were inefficient. It is true that there has not been the marked reduction in the typhoid death rate in Washington, following the installation of the water filtration works, that has been observed in other cities in America. For the six years prior to the date on which filtered water was supplied to the citizens of Washington, the average typhoid fever death rate was 59 per 100,000 population, as against 37 per 100,000 for the five years following, a reduction of 37 per cent. At Albany, N. Y., where the first modern slow sand filter was built in 1899, the typhoid death rate has been reduced by 75 per cent. At Cincinnati, Ohio, the average death rate from typhoid ranged around 50 per 100,000 for years, but since the installation of the filtration plant it has been reduced to a point which places that city, with respect to freedom from typhoid fever, at the head of all the large cities in America; in 1910 the death rate from typhoid in Cincinnati was 6 per 100,000. Similarly, at Columbus, Ohio, where the typhoid death rate before the installation of the filtration plant in 1906 was even higher than at Cincinnati, it was reduced to less than 13 per 100,000 in 1910, whereas, for the previous five years, it was 61 per 100,000. Philadelphia, before the installation of the filtration works, had a typhoid death rate of 60 or more per 100,000, and in 1910 the death rate from this disease was 17. Pittsburg, at least that part of it now supplied with filtered water, for years had a typhoid death rate of more than 130 per 100,000, but the present rate is about 12 per 100,000.

Table 25—Average Monthly Results for the Period, 1905-1910.

Columns: A - Period of sedimentation in days. B - Turbidity in parts per million. C - Bacteria per cubic centimeter. ================================================== Percentage Removed Reservoirs. A B C Turbidity Bacteria - - - River ... 106 6,400 ... ... Dalecarlia 2.2 50 5,000 53 22 Georgetown 2.2 38 3,400 24 32 McMillan 2.8 26 2,000 31 41 - - - Totals and averages 7.2 ... ... 75 69 =========================================

While it may perhaps seem unreasonable to single out Washington as a particular sufferer in this respect, it is highly probable that a large share of the typhoid is still caused by secondary infection, flies, impure milk, and private and public wells. The speaker remembers distinctly that ten years ago, when he made an investigation into the purity of the water of about 100 public wells in that city, a large number of them showed unmistakable evidence of being polluted with sewagic matter. Conclusive evidence would be secured to dispel any doubt as to the sanitary quality of the filtered product if hypochlorite of lime were added to the filtered water throughout one year or throughout the typhoid months. It seems strange to the speaker, that for this, if for no other reason, this safe and non-injurious germicide has not as yet been used at Washington, in view of the fact that at the present time it is being used continuously or intermittently in the treatment of the water supplies of scores of the most important cities of this country, among which may be mentioned New York, Philadelphia, Cincinnati, Pittsburg, St. Louis, and Minneapolis.

Morris Knowles, M. Am. Soc. C. E. (by letter).—This description of the operation of the Washington Filtration Works is timely and of great interest. It is ten years since the writer, in collaboration with Charles Gilman Hyde, M. Am. Soc. C. E., presented a similar record for the Lawrence, Mass., filter. That paper was the first complete, detailed, and continuous history of the actions and results obtained for a long period of time with such a purification works.[1] Since then, the art of filtration has advanced in many ways, particularly in regard to the methods of cleaning slow sand filters and in the accompanying processes. It is well, therefore, again to take account of stock and see really what progress has been made. Therefore, Mr. Hardy's paper, giving a description of the operations of a system thoughtfully designed, after long consideration of the problem, and of operations carried on under efficient and economical administration, with thorough record of all details, should furnish a groundwork for the careful consideration of the question stated above.

The writer, using as a text some of the ideas given in the paper, but more particularly some of those becoming prevalent elsewhere, desires to discuss methods and costs of operation, especially in relation to sand handling; and to offer suggestions looking toward greater efficiency, as well as economy, in carrying out the standard and well-tried methods.

Theory of Slow Sand Filtration.—First, what is the process of slow sand filtration? The answer to this question involves many factors, some of which are even yet but imperfectly understood. In the early history of filtration, at the time of the construction of the London filters, only the straining capacity of the sand bed, to remove gross particles, was known. Later, when the organic contents of water had become better understood, the chemical or oxidizing powers of the process were recognized as performing an important part. Finally, co-existent with the discovery of the so-called "germ theory of disease," a study of the bacterial action of filters resulted in the recognition of its importance. It is now universally thought that each of these factors performs its useful function; that the size of the sand, the amount of organic matter remaining on the surface of the bed, the turbidity of the applied water, and the bacterial content of the influent, are some of the things on which depends the determination of the relative importance of each.

[Footnote 1: Transactions, Am. Soc. C. E., Vol. XLVI, p. 258.]

Engineers have been taught to believe, by the German school of thought, that the film of organic matter on the surface of the sand plays a very important role in filtration. This Schmutzdecke, as it is called, has been considered so precious that stress has been placed on treating it with great care. It was not to be wholly removed at the time of cleaning, and it was not to be walked on, or indented, or in any other way consolidated or destroyed. In fact, in some cases, the wasting of the first water after cleaning has been advocated, for the reason that not a sufficient amount of this organic film would be left on top of the sand to begin the filtration process properly immediately after the cleaning.

In late years, however, there has been a tendency to depart from this fundamental doctrine of slow sand filtration. Various new processes for cleaning the sand surface have been advocated; some of these partly destroy and others completely exterminate any semblance of a bacterial film on the sand bed. These ideas, advanced without any real and serious discussion of their intrinsic merits, or their effects on the public health, are not founded on long continuous records of such results as are necessary to establish confidence in the final value of any of these methods.

Rapid advances along this line have been made more recently, notwithstanding the occurrence of notable instances of trouble and the resultant need of complete repair of filtration beds. Because of the rough treatment of the sand surface, a penetration of organic matter and filth into the bed had taken place. This caused deep clogging, prevented the usual yield of water, and brought about a lessened bacterial efficiency, due to the attempt to force water through the filters, and because some organic matter and growths in the lower part of the bed had furnished a breeding place for more bacteria.

All these endeavors to reduce the work of cleaning have been commendable, because scraping and sand handling are the items of greatest expense in slow sand filter maintenance. Every one has been desirous of minimizing this cost. However, as the writer will endeavor to show, it seems that attempts along this line should be with the idea of doing more economically, as well as efficiently, the things which one knows will accomplish the proper results, rather than unwisely to adopt new methods which have not been tried for a long enough period to determine their effect on the public health.

Pittsburg Methods.—When first taking up the problem of design in Pittsburg, in 1902, the writer had presented to him for consideration and adoption, a suggestion that a certain method of cleaning sand filters, which would involve the washing of the sand in place (similar to that recently tried at the Jerome Park Experiment Station, New York City), would be advisable and economical. The decision then made has never been regretted. As this plan involved such a complete departure from those principles which had been well tried and had proven successful, it was believed that it was not safe to adopt such a method on the municipal filtration works, from which the people were to derive their drinking water. There is more to be considered in such a problem than mere economy of operation; the economy of human life, the effect on which requires far longer than a few months of trial to determine, is a much more important factor. Believing that no one should depart, until after a long period of conclusive experimentation, from that principle which is known to be safe (viz., to take off a small portion of the clogging surface), the writer studied to determine more efficient and economical methods of accomplishing this end.

A device for scraping the material, in just the same way as with shovels, but more efficiently and more exactly, was developed by George P. Baldwin, M. Am. Soc. C. E., under the general supervision of the Bureau of Filtration, of which the writer was in charge. However, on account of the unfortunate and earlier arrangement of other constructive matters, which the City's Legal Department advised could not be changed without upsetting the contract, the entrance doors to the original forty-six filters were not built large enough to permit the rapid and economical transfer of these machines, and, as this act takes so large a proportion of the total time of operation, it has not been found economical to use them. The additional ten filters, recently constructed, with doors especially designed and large enough to pass the machines, have not yet been placed in operation. This is said to be on account of lack of funds and of employees. Therefore, there has been no opportunity to demonstrate what the scraping machines can do, under the conditions for which they were designed to operate. The restoring machine, a complementary device in mechanical operation, which simply replaces the sand in the same condition that it would be if wheeled back, but, with a small percentage of moisture, has accomplished its purpose well and economically. The sand is placed in the filters so that there is no further settling; with a smooth surface, needing no additional adjustment; with absolutely no possibility of sub-surface clogging; and with the filters starting off exceedingly well in operative results.

Washington Methods.—In Washington, it is stated that the filters are still cleaned by the old-fashioned method of scraping with shovels, throwing the sand into piles, and afterward removing it with a movable ejector. Between scrapings there is also an occasional mid-period action of raking the unwatered sand surface, for the purpose of stirring up the dirty film. This process does not remove any of the clogging material from the bed, but it is said that no injurious effects are produced, and that it is economical. It is stated that the so-called "Brooklyn method," of stirring the surface of the sand while the water is on the bed, has been tried at Washington, but with unsatisfactory results. It seems to have been advocated with greater fervor in some other places.

The method of dry raking does not remove the dirty material, but loosens up the pores of the surface, and through this porosity permits clogging to penetrate deeper into the filter. The method of raking with water on the bed, although it removes some of the organic dirt, also permits deeper penetration of the remainder. The latest devised system of washing the sand in place, by upward spraying with water, called the "Blaisdell method," thoroughly destroys the Schmutzdecke above, and, at the same time, must permit the formation of a subsidiary one below. In the Nichols method, the material removed by shovel scraping is conveyed by an ejector to a portable separator, where it receives a single washing; the dirty water overflows to the sewer, while the washed sand is discharged through a hose and deposited on the recently scraped surface. As the latter is partly impregnated with impurities, there is, by this process, a tendency toward sub-surface clogging.

All these processes are marked and serious departures from the well-tried method of cleaning slow sand filters, which, it is well known, will operate successfully to purify polluted river waters and make them safe to drink. In all there is the danger that they have not been sufficiently and carefully tried, under scientific observation, as to results and possible effects on the public health, to be sure that the bacterial efficiency can long continue to be satisfactory, with the application of specifically infected waters. It is dangerous, and may even jeopardize the safety of human lives, to experiment on water which is furnished for drinking purposes. There is also the added danger, well known from past experience, that in a few years (it may be more or less, depending on the extent and intensity of the new workings) the filters will need renovation, partly, if not wholly, throughout the entire bed. Thus, considering the total cost during a long term of years, the apparently cheaper method may become the most expensive.

There is also an interesting query in regard to the Washington method of replacing sand in the filters, and it is worthy of most careful thought and attention. If the process described can be carried on with success and safety, it will prove to be a long and progressive step in the methods of operation. The difficulty, however, is in determining from any short-term runs whether such a process can be continued permanently without impairing the efficiency of the sand bed. Apparently good conditions may change, after a few years' trial, and be followed by unsafe results and predicaments. This replacing of sand with whatever dirt and detritus may travel with it in the carrying water is certainly not equivalent to the care with which it has been understood that sand should be deposited in filters. It is not comparable with the care with which it is placed, when wheeled from a washer, where dirty water overflows the lip, or where it is placed by a machine restorer in the filter, where the transporting water also overflows the weir and is carried to the sewer.

These cheap and rapid methods of doing the work, advanced in the interests of economy, and the idea that sand filters, receiving polluting waters, can operate at higher rates than those which we have demonstrated, and, therefore, have been led to believe are safe, is a speeding up of the whole organization and of operating conditions. It is like speeding up a machine for the purpose of getting a greater output, with the usual result that fast running means quicker wearing out of both man and machine. Quicker operations generally mean carelessness in doing the work, especially in municipal service. Carelessness is engendered by the thought that such work can be handled in a rough and rapid way, and, further, by the ridicule of all these things, which we have learned to be careful about, as old-fogyish, out-of-fashion, and archaic. Carelessness in operation breeds contempt for the art. Some of the less efficient filter plants, from the standpoint of effect on the public health, may reflect such ill-considered methods

Economy with Efficiency in Operation.—It is particularly important to find out whether one can secure the desired economy, and, at the same time, the required efficiency. The development of efficiency in every line of human endeavor is receiving much attention at present, and not the least cause for this is the growing recognition of the demand for a high standard of service for the expense caused. One of the first requirements is to have well-defined ideals and standards. When one knows how to secure a good and safe result, it is unwise to depart therefrom for a mere whim, or to secure a supposedly lessened expense, unless other facts be also determined favorably. The desire for economy must be tempered by good sense, which means that one should be willing to change a method only when the wisdom of such has been clearly demonstrated. Efficient service can only be secured by strict discipline, accompanied by fair dealing. This means employing no more men than are actually necessary, paying them on the basis of the standard of service and output produced, taking an interest in the working conditions, and providing for their health and welfare.

About twelve years ago, the writer made some investigations of the efficiency of laboring gangs in scraping and handling sand at filter beds,[1] and found that ten men was the most economical number to use in scraping the surface of the Lawrence filter, as then built and operated. This result was determined by numerous studies of the output per man per minute, with different numbers of men working under different conditions. This same sort of study has been carried further by adepts in the art, in reference to shop and similar management, but one fails to find corresponding development along this line in municipal organization except by a few of the scattered Bureaus of Municipal Research. These results, also, have related to a few of the more common and general factors, such as determining the cost per mile, or per square yard, of street cleaned, or per million gallons of water pumped.

[Footnote 1: Transactions, Am. Soc. C. E., Vol. XLVI, p. 291.]

The cost of the management of water-works, one of the largest factors of public enterprise, has never been investigated extensively and thoroughly. There is much possibility in planning for greater efficiency and in determining what can be accomplished under economical administration. Every one is aware of the multiplicity of men in municipal service. Some of these are entirely incompetent, others partly so; the recent appointees may be more efficient, but the majority of them gradually deteriorate under the subtle influence of the prevailing atmosphere, and each new incoming administration places more and more men on the work, without reason or necessity. All these tendencies have made the cost and maintenance of public work greater and greater, and, at the same time, have resulted in frequently and steadily decreasing the output and efficiency per employee.

The Washington situation, however, presents an admirable contrast to this, because of the methods of administration of the public works of the District of Columbia and their freedom from petty political influence. The limited number of employees has tended toward economy, and rendered this plant the envy of all who have desired to obtain good management. Its cost items have been looked on as a result long hoped for, but seldom obtained. It is to be regretted, therefore, that such an abrupt change in methods of removing clogging material and replacing sand has taken place without years of experimental trial on filters not furnishing drinking water to the public, and without an attempt, under such excellent conditions, to maintain the efficiency by a better labor output and by improved working and machine methods in the performance of the older and established order of doing things.

In preparing water for the use of the people, the realms of the unknown are so much larger than those which have been investigated and developed that there may be many undiscovered factors affecting the public health, and many ways in which it is dangerous to depart from well-known and surely safe methods. Who can say that in some subtle and, at present, unknown manner, the failure in some places, where filtration is practiced, to reduce the death rate from typhoid fever may not be due to the introduction of radical departures from the older, slower, safer, and more efficient methods which have produced such excellent results, both in America and in Europe? Further, in cases where there has been a falling off in the typhoid death rate, the failure to secure an accompanying improvement in general health conditions, which follows so closely in communities supplied by water filtered in accordance with the more conservative principles, may be due to the introduction of some of these not thoroughly tried processes. Some day full information may be available as to the influence of these methods of plant operation on the health of the community. Until that time, is it not a much better policy to follow the principles which have been proven by many years of experience to produce safe results, and to make the foremost object the improvement of the methods of operation in accordance with these established truths?

There is opportunity for the upbuilding of greater efficiency in the conduct of employees and in securing the maximum output, by establishing more comfortable and healthful conditions than usually exist. The elimination of political influence from municipal service is also a task which challenges the people of to-day, and the operating and managing engineer is in a position to perform an important part in accomplishing this end. The number of employees can be reduced to those actually needed, and the way opened for the employment of men who thoroughly understand the necessities of honesty and efficiency in the conduct of public affairs. It should be remembered that to design and construct well is only half the job; to operate economically and efficiently is even more of a problem than to build, and requires just as good talent, just as keen appreciation of the various problems, and is even more essential to public welfare. It seems to the writer that the logical development of the art of obtaining economy as well as efficiency should be along these lines, rather than to revolutionize methods, without having a long-period test of their value, and at the same time allow political influences to control, to a large extent, the labor item.

Preliminary Treatment.—The decision as to the preliminary treatment of the Potomac River water before filtration is of interest, particularly because various other conclusions have been reached in different sections of the country. However, in the main, these decisions have been due to differences in the character of the waters, but it must be evident that they have sometimes been the result of ill-considered action, or the desire to promote some special interest. The use of preliminary filters, which involves a large investment, is not always to be commended, particularly because at times of reasonably good water the removal of some of the organic matter is really injurious and lessens the effect of the final filters.

For a long time, the writer has believed that, where other things are equal, and where there is no important reason for double or preliminary filtration, long periods of storage, accompanied by the use of coagulant at times of severe and extreme muddiness, as planned at Washington, solves the problem in the most practical and economical way. It is true that the investment for a large storage basin may equal, or even exceed, that required for preliminary filters; but the influence of storage on the quality of raw water is never injurious, and, by ripening the condition of the water, may be greatly beneficial in the process of filtration.

The storage available in such a basin makes it possible to shut off the supply from the river during the worst conditions of the water. The duration of the most troublesome spells ordinarily does not exceed a few days, and it is usually possible to secure sufficient capacity in the basin to tide over these periods. Then again, long periods of storage, in addition to assisting in breaking up organic matter, permit the dying out of bacteria, particularly many of the pathogenic kind, and, therefore, the water is rendered much safer from this standpoint. In other words, there is additional insurance in long storage against the faulty and careless operation of incompetent filter employees. The addition of coagulant, especially the fact that only a very small investment of capital is required for the necessary apparatus for dosing the water, and that the cost of the coagulating materials has to be met only when used, seems to give the process, in a most satisfactory manner, the requirement for economical management and thoroughness in preparing the water for final filtration.

Parking Public Works.—It is disappointing that the author has not mentioned some of the steps contemplated in reference to the landscape treatment of the Washington filtration area. Probably every one has been impressed by the barren aspect of the works as they are approached, and as one looks over them. Recently, however, it is stated that some steps have been taken to lay out the grounds, treat the surface in an attractive manner, and make a park of the area. The writer has a firm opinion that when an investment is made for public works, it costs but little in addition to construct buildings along appropriate architectural lines, to treat the grounds in a pleasing manner, and to make the entire works a credit to the municipality from an artistic standpoint. When treated on broad lines, such areas become public parks, and afford open breathing places for the residents, and, if near centers of population, may well be equipped with playground facilities for the children. When thus developed they should have care, that the planting and equipment should not deteriorate and the last state become worse than the first.

The influence which these ever-present examples of attractiveness have on the community is becoming better recognized by students of social progress, and there seems to be no doubt that spending money on such features is not only desirable from the artistic standpoint, but is justified on practical grounds as well. It is cheaper than to create parks, when necessity and demand can no longer be resisted, by buying property and occasionally tearing down buildings and constructing de novo. That this work is now being done in Washington, even after construction, is certainly a recognition of the advisability of original efforts in this direction.

George C. Whipple, M. Am. Soc. C. E. (by letter).—Mr. Hardy's paper is an excellent presentation of the results of the operation of the Washington water filtration plant from the time of its construction in 1905 until June, 1910. Papers of this character are altogether too infrequent, and the actual results from the filters now in use are not readily accessible in detailed form. Yet it is only by studying the results obtained by filters in actual use that improvements can be made and the art advanced.

Among the many important facts brought out by Mr. Hardy, only a few can be selected for discussion. One of these is the operation of filters under winter conditions. It is well known that the efficiency of sedimentation basins and filters is lower during winter than at other times, yet it is just at this season of the year that there is the greatest danger of typhoid fever and similar water-borne diseases being transmitted by water. Most of the great typhoid epidemics have occurred during cold weather, and the very use of the term "winter cholera" is of significance. Apparently, typhoid bacilli and similar bacteria are capable of living and retaining their vitality longest during that season of the year. Just why this is so, bacteriologists have not satisfactorily explained. Doubtless many factors are involved. Because of the increased viscosity of the water, sedimentation takes place less readily at lower temperatures, and inasmuch as sand filtration is partly dependent on sedimentation, the efficiency tends to fall off in cold weather. During winter some of the external destroying agencies are less potent, such as the sterilizing effect of sunlight, and the presence and activity of some of the larger forms of microscopic organisms which prey on the bacteria. Another factor may be the greater amount of dissolved oxygen normally present in water during cold weather, as experiments have shown that dissolved oxygen favors longevity.

Still another reason for the larger numbers of bacteria that pass through a water filter during cold weather may be the effect that the low temperature has on the size of the bacteria themselves. A few experiments made recently by the writer appear to indicate that at low temperatures the gelatinous membrane which surrounds the bacterial cells tends to become somewhat contracted, thus decreasing the apparent size of the bacteria as seen under the microscope. Either this contraction occurs, or the cells themselves are smaller when they develop in the cold. It is possible also that low temperature affects the flagella of the organisms in the same way. It is not unreasonable to suppose that the effect of low temperature is to form what may be, in effect, a protective coating around the cells, which tends to make them smaller, less sticky, and less subject to outside influences. This would tend to make them pass through a filter more readily. In line with this idea also is the well-known fact that disinfection is less efficient in cold water than in warm water.

Another way of viewing the matter is that cold retards the growth of bacteria on the filter, thus reducing the effect of the Schmutzdecke. Still another view of the greater danger from bacterial contamination in winter is the theory that cold prolongs the life of the bacteria by merely preventing them from living through their life cycle and reaching natural old age and death as rapidly as in warm weather.

Another topic in Mr. Hardy's paper which has interested the writer is that of preliminary filters. The experiments described at length indicate clearly that such devices would prove of little or no benefit under the conditions existing in Washington, and that when the river contains considerable amounts of suspended clay nothing less than chemical coagulation will suffice to treat the water so that the effluent will be perfectly clear. Preliminary filters have been used for a number of years at various places and with varying success. In few instances have they been operated for a sufficient length of time or been studied with sufficient care to determine fully their economy and efficiency as compared with other possible methods of preliminary treatment.

Among other experiments on this matter are those made at Albany, N. Y., and published by Wallace Greenalch, Assoc. M. Am. Soc. C. E., in the Fifty-ninth Annual Report of the Bureau of Water for the year ending September 30th, 1909. The Hudson River water used at Albany is quite different in character from the Potomac River water used at Washington, as it is less turbid and contains rather more organic matter. The results obtained in these experiments showed that during the summer the number of bacteria in the effluent from the experimental sand filter used in connection with a preliminary filter did not differ widely from the number found in the effluent of the city filter where there was no other preliminary treatment than sedimentation. In the winter, however, the numbers of bacteria did not increase in the effluent from the experimental filter as they did in the effluent from the city filter. This is shown by Table 26, taken from the report mentioned.

Apparently, therefore, at Albany the benefits of the preliminary filter, as far as bacterial efficiency is concerned, would be confined to a short period of three or four months in each year. Under such circumstances it may well be questioned whether the advantages of preliminary filtration justify its cost.

Table 26 Results of Experiments with Preliminary Filter at Albany, N. Y. =========================================== Month, Bacteria Bacteria Bacteria in Bacteria in 1906. in raw in pre- effluent effluent water. liminary from from filter experimental city filter. effluent. sand filter. - March 133,480 36,000 151 706 April 77,420 4,810 72 155 May 15,800 2,250 48 37 June 4,520 358 38 34 July 2,090 163 25 22 August 2,740 121 36 22 September 8,280 445 20 24 October 38,350 4,235 67 227 November 67,910 15,570 337 341 December 645,500 25,440 144 2,783 - 1907. January 127,560 4,660 48 443 February 28,000 1,800 13 116 ===========================================

On the diagram, Figure 11, will be found various data taken from the published records of the Albany filter, from 1899 to 1909. These data include: The numbers of bacteria before and after filtration; the percentage of bacteria remaining in the effluent; the average quantity of water filtered, in millions of gallons per day; the quantities of water filtered between scrapings; the turbidity of the raw water; the cost of filtration, including capital charges and cost of operation; and the typhoid death rates of the city per month. Several points are brought out conspicuously by this diagram. One is the uniformly low death rate from typhoid throughout the entire period. The filter was operated from 1899 until the fall of 1907 with raw water taken from what is known as the "Back Channel." Since then it has been taken from a new intake which extends into the Hudson River itself. Until the fall of 1908 the preliminary treatment consisted merely of sedimentation, but since then the water has received an additional preliminary treatment in mechanical filters operated without coagulant, along the lines of the experiments just mentioned. During this time the average rate of filtration of the sand filter has not changed materially, although it is said that the maximum rate has been increased since the preliminary filters were put in service. The study of the bacteriological analyses shows that the best results were obtained during 1902, 1903, and 1904. Since then the numbers of bacteria in both the raw and filtered water have increased. This was especially noticeable during the winters of 1907 and 1908 when the water was taken from the new intake. It will be interesting to compare the results after the preliminary filters have been operated for a long period to ascertain their normal effect on efficiency and on the increased yield.



Another fact to be drawn from the plotted Albany data is the increase in the cost of filtration, both in capital charges and in operation. From 1899 until 1906 the cost of operation, including the cost of low-lift pumping, was approximately $5 per million gallons of water filtered; and the total cost of filtration, including capital charges, was about $10 per million gallons. During the year ending September 30th, 1909, the cost of operation had increased to $7.63 per million gallons, and the total cost of filtration to $15.92 per million gallons, or approximately 50% in three years.

Table 27—Results of Bacteriological Analyses of Samples of Water at Peekskill, N. Y., Before and After Filtration.

Bacteria per cubic centimeter. ============+====+======+======+====+====+====+====== Raw Clear Effluent Effluent Effluent Effluent Tap in Date. water. reser- No. 1. No. 2. No. 3. No. 4. city. voir. + -+ -+ + + + + 1909. + -+ -+ + + + + December 29th 190 100 ... ... ... ... ... + -+ -+ + + + + 1910. + -+ -+ + + + + February 15th 135 10 10 30 20 ... 265 March 31st 225 50 25 45 60 ... 35 May 18th 300 29 22 26 35 43 36 July 6th 300 44 9 3 41 10 31 August 16th 60 5 0 4 1 13 15 October 3d 550 14 12 14 38 ... ... November 21st 315 22 26 17 6 ... ... + -+ -+ + + + + 1911. + -+ -+ + + + + January 25th 415 7 8 4 6 ... 7 + -+ -+ + + + + Average 277 30 14 16 26 22 65 ==============+======+======+========+========+========+========+========

Table 17 (Continued.) Tests for B. Coli. ==================+====================== Percentage of Samples Quantity of water Containing B. Coli. tested. + + - Raw. Filtered. + + - 0.1 cu. cm. 0 0 1.0 cu. cm. 20 0 10.0 cu. cm. 40 0 ==================+========+=============

As a matter of record, the results of a series of analyses made at Peekskill, N. Y., during 1910 are presented in Table 27. A sand filter was constructed for the water supply of this city in 1909, and put in operation in December. The filter has a capacity of 4,000,000 gal. per day. The supply is taken from Peekskill Creek, and the water receives about one week's nominal storage before flowing to the filters. An aerator is used before filtration during the summer, when algae are likely to develop in the reservoir. The filter was installed after an epidemic of typhoid which was apparently caused by an infection of the water supply. Normally, the water has been little contaminated, but the supply is subject to accidental contamination at any time, among other possible sources of infection being the camps of workmen now engaged in constructing the Catskill Aqueduct for New York City.

Table 28 Average Results of Chemical Analysis at Peekskill, N. Y., Made at Intervals of Six Weeks During 1910. ====================+================+================+================= ~Parts per ~Parts per Million.~ Million.~ + + Raw Filtered Raw Filtered water. water. water. water. -+ + + + + Turbidity 2. 0 Total residue 70. 76.00 Color 25. 20. Loss on ignition 19.00 17.00 Nitrogen as albumi- Fixed residue 50.00 59.00 noid ammonia 0.112 0.076 Iron 0.17 0.13 Nitrogen as free Total hardness 38.70 45.10 ammonia 0.024 0.006 Alkalinity 33.90 42.60 Nitrogen as nitrites 0.001 0.001 Incrustants 4.60 4.50 Nitrogen as nitrates 0.06 0.06 Chlorine 2.60 2.70 =====================+========+========+================+========+========

F. F. Longley, Assoc. M. Am. Soc. C. E. (by letter).—In this paper the author has presented a mass of data which will be welcomed by engineers engaged in water purification work, because complete operating records form a substantial basis for improvement in the art, and are often the inspiration for interesting discussions and the exchange of experiences of different observers whose views are mutually appreciated.

Recent tendencies in filtration engineering have been largely in the direction of reducing the cost of operation. A comparison of the operating costs of the earlier American plants of about a decade ago, with those here presented of the Washington plant, is very gratifying to those who have been intimately connected with the latter work. Through perfection in design and reasonable care in operation, the cost of filter cleaning, which is a very considerable part of the total cost, has been reduced to an unusually low figure, without any sacrifice in efficiency, and in the interests of the public health.

Table 14 shows that, from the first year, there has been a progressive increase in the total cost of operation per million gallons filtered, but this has not meant an increase in the annual total expenditure. The largest percentage of increase in any item has been in "Care of Grounds and Parking," and covers much-desired landscape improvements. Aside from this, the principal factor affecting the table of costs has been the reduction in water consumption in the District of Columbia. Nothing pertaining to this reduction has produced any corresponding reduction in the force required for the maintenance and operation of the filtration plant, office and laboratory, and pumping station, though probably there has been some reduction in filter cleaning. Obviously, then, the total cost per million gallons would increase.

This decrease in consumption has been brought about by the elimination of waste in the distribution system, which is not in the same department as the filtration plant, but with regard to which a word may not be amiss in connection with this discussion.

The Washington Aqueduct was built half a century ago on lines which at that time were considered extraordinarily generous. Until recently, therefore, there has been no occasion for concern over the high rate of consumption. During recent years, however, the use and waste of water have increased, reaching a climax under unusual conditions in the winter of 1904-05. The maximum capacity of the aqueduct system is about 90,000,000 gal. The maximum daily consumption at the time mentioned arose almost to 100,000,000 gal., with the result that, before normal conditions were restored, the reservoirs of the system were almost depleted.

This had a beneficial effect, as provision was made for an active campaign for reducing the waste of water, which was known to be very large. These investigations, using the pitometer, were begun in July, 1906, and have been pursued continuously since that time, with most excellent results. Up to January, 1909, leaks aggregating about 12,000,000 gal. per day were detected and eliminated, and about half the house services had still to be covered by the pitometer bureau.

Although this reduction in waste has brought about an apparent increase in the cost of filtration, its economical results have been far-reaching. The causes which brought about this investigation also resulted in securing an appropriation for the study of the question of increased supply. The writer was in charge of these studies, and the most significant conclusion was that, owing to the excellent results of the efforts for waste restriction, the total consumption and waste of water in the district during the next few years would be far enough below the safe working capacity of the existing aqueduct system to make it entirely safe to postpone the construction of new works, involving the expenditure of several million dollars, in spite of the threatening conditions of a few years ago.

There has been so much controversy over typhoid fever in the District of Columbia that the writer hesitates to discuss this subject. Viewing the situation through the perspective of several years, however, it does not seem to be as hopeless as the criticisms of four or five years ago would lead one to believe.

In Table 9, showing the typhoid death rates, out of nine years given prior to 1905-06, when the filters were started in operation, only one shows an annual death rate as low as the highest one since that year. Further than this, the annual average typhoid death rate for the period since that year has been one-third lower than for a corresponding period before the filters were started.

The exhaustive researches of the Public Health and Marine Hospital Service into this whole question, covering a period of about four years, have raised the present filtered water supply of the District of Columbia above any well-founded criticism. There has long been a strong and growing feeling that the water supply, before filtration was introduced, had been blamed for more than its share of the typhoid, and this is borne out by much evidence that has been presented from time to time.

It is not an unreasonable conjecture, therefore, that perhaps the reduction of one-third in the total typhoid death rate may represent a much larger reduction in that part of the total which was due to polluted water alone; and that, as the authorities in the District of Columbia and in certain other cities, particularly in the South, are now recognizing, the fight against much of the remaining typhoid must be in the direction of the improvement of milk supplies, precautions against secondary infection, and attention to a large number of details surrounding the individual, which may effectively protect him against the insidious attack of the disease favored by unknown agencies.

Experiments in Filter Cleaning.

The author refers to the difficulty encountered during the first two summers in keeping the filters cleaned fast enough to maintain the capacity of the plant. The real seriousness of this may be judged from the following facts. The average increase in loss of head on all the filters for the entire year, July 1st, 1906, to July 1st, 1907, was about 0.053 ft. per day. During the 1906 period of low capacity under discussion, the loss of head on twelve of the filters increased for a period of eight days at the average rate of 0.45 ft. per day, or about nine times the normal rate of increase. This difficulty was caused by the presence of large numbers of micro-organisms in the applied water. During the first summer (1906) this fact was not recognized, but the sudden decrease in capacity was supposed to have been caused by the unusually high and long-continued turbidity which prevailed during that summer in the Potomac River, and persisted in the water supplied to the filters even after about four days of sedimentation in the reservoirs. During the second summer (1907) the same phenomenon of suddenly and rapidly increasing losses of head appeared again, but without any unusual turbidity in the applied water. Investigation, however, showed the presence of large quantities of organisms, particularly melosira and synedra, in the applied water, and examinations in subsequent years have shown a periodic recurrence of these forms in quantities sufficient to cause the trouble mentioned. In June, 1907, examination showed repeatedly more than 1,000 and 1,500 standard units of melosira per cu. cm., and one count showed nearly 3,000 standard units.

Several expedients were tried in an effort to restore the rapidly decreasing capacity of the filters. One of the earlier conjectures as to the cause of the trouble was that it might be due to the accumulation of large quantities of air under the surface of the sand, as air had been observed bubbling up through the sand, especially in filters which had been in service for some time. The expedient was tried, therefore, of draining the water out of the sand and then re-filling the filter in the usual manner from below, in the hope of driving out the entrained air. Presumably this treatment got rid of the air, but it did not restore the capacity of the filter, as the point of maximum resistance was in the surface of the sand and not below it.

As the author states, raking the filters was tried and found to give results which were satisfactory enough to meet the emergencies already referred to. When the filters were first put in operation, in the fall of 1905, the method of bringing back the capacity of a filter after the end of a run was to remove all the dirty sand to a depth determined by the marked discoloration caused by the penetration of the clay turbidity. This sometimes necessitated the removal of large quantities of sand at a cleaning, as the turbidity was exceedingly fine, and penetrated at times to a depth of 3 or 4 in.

With the idea of effecting an economy in the cost of cleaning the filters, a schedule of experiments was arranged shortly before July 1st, 1907. The general object of the experiments was to determine, first, the relative costs of all different methods tried; second, whether the removal of only a thin layer of sand, or the mere breaking up of the surface of the sand by thorough raking, would give the filter its proper capacity for the succeeding run; third, whether the filters under these treatments would maintain a high standard of quality in the effluents; fourth, whether the continued application of any less thorough method than the one then in use might materially affect the future capacity of the filters.

To this end the filters were divided into four groups which, during a period of about six months, were subjected to treatments as follows:

Group A.—Filters scraped deep at the end of each run; Group B.—Filters scraped light at the end of each run; Group C.—Filters raked at the end of each run, until raking failed to bring back the proper capacity; then they were scraped light, and at the end of the next run the raking was resumed; Group D.—Light scrapings and rakings alternate at ends of runs.

The term "deep scraping" means the removal of practically all the discolored sand, in accordance with the usual practice prior to the beginning of these experiments; "light scraping" means the removal of only a thin surface layer of sand. This depth has usually averaged about 3/8 in. "Raking" means the thorough breaking up of the clogged surface of the filter by iron-toothed rakes, to a depth of about 1 or 2 in.

Results.—A general summary of the results of these experiments is given in Table 29, which also shows the relative costs of the different methods per million gallons of water filtered. A normal period of 9 months just prior to the beginning of these experiments shows a labor cost (corresponding to that in Table 29) of $0.29-1/4 per million gallons filtered.

Table 29—Average Results.

Columns: A - Group. B - Number of filters. C - Number of days of service. D - Million gallons filtered. E - Cost of labor per treatment. F - Sand removed in cubic yards. G - Sand removed in cubic yards. H - Cost of labor. I - Bacteria per cu. cm. in effluent. J - Turbidity in effluent.

====================================================================== Per Million Per Run: Gallons Filtered I J - - - - - - - - A B C D E F G H - - - - - - - - - A 5 82 221.2 $68.44 215 1.11 $0.309 13 1 B 9 36 101.4 29.25 84 0.83 0.288 16 1 C 5 21 60.0 10.92 24 0.40 0.182 18 1 D 10 32 86.0 20.10 46 0.54 0.234 22 1 ===============================================================

Capacity of Filters.—The capacity of the filters under the different methods of treatment are shown in a general way in Table 29 for days of service and millions of gallons filtered per run. This element by itself is decidedly in favor of the deep scrapings, and least in favor of the repeated rakings.

A clearer conception of the capacities of the filters under these different conditions may be obtained from the four diagrams, Figure 12, showing, for the four different groups, the average number of days of service of the successive runs. The diagram for Group A shows that the variations in the period of service of the filters scraped each time to clean sand follow a more or less definite curve from year to year. For the period covered by this curve, the tendency seems to be toward a slight decrease in capacity from year to year, as shown by the lower average maximum and minimum in the second year than in the first. Group B shows a sudden decrease in capacity following the first light scrapings and, since that time, a low but quite constant capacity. Group C shows a constantly decreasing capacity with successive rakings. The only significance attaching to the curve after the first raking is the prohibitively low capacity indicated, and the ineffectiveness of the measures taken to restore the capacity after the sixth raking. Group D, after the first raking, shows a prohibitively low and constantly decreasing capacity. The diagrams for C and D indicate a dangerous reduction in capacity if long persisted in. The method followed with Group C may be dismissed with the statement that it is entirely insufficient, and would be of use only in the rarest emergencies.