Some Tests with Hydrogen Ion Method of Water Treatment
Experiments to Determine Optimum Concentrations for Coagulation of Various Waters—Application to Actual Conditions at Plants
EFFICIENCY and economy in filtration is ever the aim of the efficient operator and his study is always done until a view to obtaining the best results in this respect. The following paper gives some very practical ideas along this line and illustrates the methods of obtaining the optimum of hydrogen ion concentration in order to assist the plant operator. The paper is well worthy of study.
At a conference of engineers and filter plant superintendents of the Southeastern States, held at Atlanta in July, 1922, E. J. Theriault described very comprehensively results obtained by the U. S. Public Health Service in a very exhaustive laboratory study of the value of hydrogen ion concentration in the control of coagulation at water purification plants. This study was made for the most part on artificially prepared turbidities and added buffers, though some results were obtained on natural Potomac River water. As described by Mr. Theriault the indications were that this determination would be of the greatest value to the filter plant man in controlling his dosage, so as to secure economy of chemicals, efficiency in filtration, the elimination of coagulant in filter effluent and possibly the prevention of the corrosive action of certain low alkalinity filter plant effluents.
As suggested by Mr. Theriault, this laboratory research had reached the stage when the next logical procedure was the application of the same line of investigation to the actual conditions found in the various municipal water plants. With this idea in mind, a series of tests were undertaken by the North. Carolina state department of health, on the various public water supplies in that State. As the types of water supply found in this State vary from the highly colored swamp waters of the Atlantic Seaboard to the clear mountain supplies of the Blue Ridge, it was felt that conditions were very favorable for such studies. The important questions to be determined were:
- —What constitutes the optimum or optima hydrogen ion concentration on the various waters?
- —Does the holding of the water treatment to such optima result in improved efficiency of plant results and economy and is it practicable to secure such control?
- —Can a dependable method be devised for determining hydrogen ion concentration that can be used successfully by the average plant operator?
Method of Making the Determination
The determination of hydrogen ion concentration by electrical methods was not considered as the apparatus is too expensive for the small water plant and the technique too elaborate for the average operator. For the large plants it is entirely suitable and of course, furnishes the most reliable figures. It is also useful for standardizing the color methods.
In order to arrive at a suitable colormetric techn_____que, tests were started at the Wilmington, N. C., plant. Close co-operation and assistance was furnished by George D. Norcom, the superintendent of filtration. The standards supplied by the LaMotte Chemical Co., were first tried, but were found unsatisfactory as regards permanence, the methyl red range especially spoiling very quickly. Pure chemicals were next secured and buffer solution made, from which standards could be made up as needed. Very good results were obtained in this way, but only a well-trained chemist with a good laboratory could be relied upon to make such standards.
In the book, “The Determination of Hydrogen Ion Concentration,” by Dr. W. M. Clark, is a color chartreproducing the colors of the various standards, made in accordance with the method of Clark and Lubs. It was found that using these charts in place of the tube standards gave very satisfactory results, that compared very favorably with tubes made from standard buffer solutions and with fresh LaMotte Chemical Co. standards. In order to check possible variation due to personal equation, determinations were made using tube standards and color chart, and by several observers, some of whom had had no previous experience in this class of work. The results showed very close conformity in results and sufficient accuracy using charts for the purpose of water treatment control.
“Laboratory research had reached the stage when the next logical procedure was the application of the same line of investigation to the actual conditions found in the various municipal water plants. With this idea in mind, a series of tests were undertaken by the North Carolina State Department of Health, on the various public water supplies in that State.”
In order to carry out this technique it is only necessary to have one of the charts, which are sold separate from book, some indicator solutions, tubes graduated to hold 10 cubic centimeters, 1 cc pipette graduated in l/10ths, and a bottle of distilled water. The indicator solutions are made from the dry powder purchased from LaMotte Chemical Co. and in accordance with Clark’s method, except that after rubbing up the dry indicator in a mortar with the specified quantities of N/20 soda, the final dilution was made with 85% ethyl alcohol instead of distilled water. These solutions are twenty times too strong in case of methyl red and ten times too strong in the case of the other indicators.
It is necessary to make these concentrations in order to secure keeping qualities, and for making the tests they must be diluted with distilled water. The concentrated alcoholic solutions seem to keep almost indefinitely, a set having been used over a period of eight months with no deterioration. All of the tests given in Table 1 were made by this method, with sufficient checking by tube standards to insure the uniformity of results. It is a significant fact that tests made on the same water at beginning of the eight months showed the same results as at the end of this time, though the same stock indicator solutions were used for entire period. The colors on the charts do not seem to fade appreciably, and they are cheap enough to make their replacement no obstacle.
This technique seems to meet the needs of the small plant operator, and as described, forms a much simpler test than the determination of alkalinity by titration with acid.
Determination of Optima
Clark and Theriault (Public Health Reports, Feb. 2, 1923) state that as a result of their investigation:
“When other possible factors are left out of consideration, optimum conditions for floe formation will be found with a narrow zone of pH centered for dilute solutions at pH 5.5.” J. R. Baylis (Proc. American Society Civil Engineers, Dec., 1921) gives an optimum for Baltimore city water at 6.5. Wagner and Enslow (Journal American W W. Association. May. 1922) give an optimum zone for “non-organic” waters 6.5 to 7.0 and for “organic” waters a zone as low as 5.7 and never above 6.5. These figures indicate quite a variation in optima.
(Continued on page 15)
Test of Hydrogen Ion Method
(Continued from page 8)
Tables 1.2 and 3 give optima determined on a number of North Carolina waters. Those in Table 1, check the figure set by Clark and Theriault. These waters are practically all located in the east central part of the State and contain a certain amount of color as well as turbidity. Table 2 shows some waters located in the south central part of the State with optima consistent with those found by Baylis, and Wagner and Enslow. These waters are more or less turbid with practically no color.
Table 3 gives the figures on the high colored clear waters. The water used at Hamlet and Lumberton with comparatively low colors, give optima around 5.5, while the three supplies at Wilmington (northeast river), Hertford and Elizabeth City, with an excessive color, show much lower optima. The first two are located in interior of the State, while the latter three are on the coast.
The method used for determining optima on these watersAconsisted in using the ordinary “jar test,” and applying the results obtained with this to the plant itself. A series of 1 liter jars were used, and alum, soda and lime solutions prepared so that each cubic centimeter of alum solution put in a jar of raw water would be the equivalent of one-half grain per gallon, and each cubic centimeter of alkali solution in same way was equivalent to one-eighth grain per gallon. In this way any combination of chemicals could be used. The pH was then determined on each jar test, and the jar furnishing a floc most quickly was considered optimum. In this connection it may be noted that the observation of Clark and Theriault that the quickest forming floc is always eventually the best floc was borne out in these tests.
On waters where there was sufficient natural alkalinity to react with the alum, the determination of the optimum is a simple matter, as it is only necessary to run a series with various amounts of alum. Where there is required additional alkalinity the tests were complicated by the two variables. In the latter case, a series of jar tests must be made for each quantity of alum, with the amount of alum the same in each jar and added alkali varied. In each case, the pH is carried to a point on acid side where no floc forms after hours, and similarly on the alkali side to a point of no floc formation.
Tables 4, 5 and 6 will illustrate a typical set of tests.
The alkalinity of the raw water was 21. which was sufficient for necessary alum. If sufficient alkali had been used to bring pH on No. 5 test to 5.5 an excellent floc was formed. This is the water designated No. 1 in Table 1.
This water is the one listed in Table 3 from Hamlet. It has fairlv high color and no turbidity.
This is the water listed in Fable 2 as Concord. It will be noted that there is no color and the natural alkalinity is 39.
The tables 4, 5 and 6 will give an idea of how each test was carried out. It is not considered necessary to burden this paper with the mass of figures producing the optima tabulated in Tables 1, 2 and 3.
High Optimum Waters
It is rather difficult to offer explanation of the high optimum water. The chemical contents not differing appreciably from those with lower optima. An important fact, however, is that there is quite a difference in the alkali requirements on these waters from that of the 5.5 series. It is generally true that the latter require about 6 parts per million for every grain of alum per gallon. In the case of the waters at Charlotte, Concord, Gastonia, Wadesboro, etc., in Table 2, although there is found more than enough natural alkalinity to react with the alum added, it is necessary to add excessive amounts of alkali to get a good floc. In the case of Charlotte, the raw water alkalinity was 21 p. p. m. and this should be entirely sufficient for 2 grs. per gallon of alum. However, it was necessary to add 30 p. p. m. additional alkalinity to get the best floe. It was not possible to get CO2 determinations on the various waters, but at Charlotte and Wilmington these are recorded daily and do not show appreciable difference.
(Continued on page 28)
(Continued from page 15)
Test of Hydrogen Ion Method
High Colored Water
The high colored waters show low optima. The excessively high colored waters at Wilmington, Elizabeth City and Hertford, require optima as low as 4.5, while partly colored waters at Hamlet and Lumberton may run as high as 5.6. The water at Wilmington has been especially studied, George D. Norcom, in charge of tlie filter plant, having made numerous pH determinations during the last year. This is typical of the high colored waters, and a low alkalinity treatment as worked out for it by the author in 1916, and is still in use. This treatment consists in adding only enough alkali to give 4 p. p. m. for each gr. per gallon of alum, and adding additional alkali in a second basin, the principles involved in the precipitation of colloids prompting this treatment as explained at that time. The Elizabeth City water and that at Hertford act similarly.
A very interesting condition exists at Wilmington. Just below the filtration plant, the Cape Fear River forks forming the main Cape Fear and Northeast branch. The main river usually is a low colored high turbidity water, while the Northeast branch is highly stained from cypress swamps. Up to very recently, water has been taken from the Northeast branch, where the tidal changes furnish a clear colored water at times and others a mixture of this with the main Cape Fear water. The tests on the Wilmington water under Table 3 are made on this colored Northeast water and show an optimum of about 4.5. Recently, the intake was extended across to main river, and the Wilmington tests under Table 1 are made on the new water. Thus we have the main river showing an optimum of about 5.5 and its branch fed from the Cypress swamps showing the lower optimum.
Controlling the pH by Adding Acid
In the case of the Tar River waters, tests 1-7 in Table 1, sulphuric acid was used with the result that as good floe was formed using 0.5 grs. per gallon alum as where 1.5 grains was required without acid. It was necessary to add 1.5 g. p. g. alum before the optimum of 5.6 was reached and by adding just sufficient sulphuric acid to produce a pH of 5.6 with 0.5 g. p. g. alum similar results were accomplished.
All attempts to cut the pH of the high colored waters down with sulphuric acid, thus reducing the amount of alum required, were unsuccessful.
Colloidal Clay and Color
A possible explanation of the variable optimum lies in the colloids in the water. Clay in colloidal state has a negative electric charge and color a positive charge. The fairly heavy clay particles have no influence. The whole process of coagulating water has two stages. This first reaction is entirely a chemical one where definite amounts of alkali combine with definite amounts of aluminum sulphate with the formation of aluminum hydrate. Even with the tests where the pH is too high or too low and no good floe is formed, there is. except when tot) little alum is used on a high colored water, a smoky appearance showing the formation of aluminum hydrate. This aluminum hydrate is held in a colloidal state and must he thrown down by physical means. There seems to he required a definite pH or definite electrical charge to precipitate the positively charged aluminum hydrate and negatively charged coloring matter and clay particles.
Value of pH in Plant Control
Thus far, the practical application of the pH determination indicates valuable assistance to the plant operator. Having determined the optimum pH it is only necessary to take samples of water from the mixing chamber and determine the pH. If below the optimum, more alkali should be added and if above optimum alkali is reduced. If supposed optimum is shown and poor floe is forming, it would suggest a change in optimum, and a redetermination of this made. Our experience so far shows that the optimum does not change very frequently. In the case of the Tar River tests (Table 1, tests 1-5) the first test was made in August after a long dry spell, the second test was in October after heavy rains and the third in February during cold weather conditions. In each case the optimum was the same.
Our experience also indicates that coagulation taking place at the optimum pH, gives a heavy, well defined floe, and that it takes place more rapidly, thus requiring less mixing chamber and coagulation basin retention time. Attention is especially called to the influence of this control of coagulation on mixing chamber design.
- —The majority of waters in North Carolina coagulated most satisfactorily at an optimum pH of approximately 5.5. High colored waters require a much lower optimum, approximately 4.5 pH. Some turbid waters, free from color have much higher optimum varying from 6.5 to 7.5 pH.
- —A technique sufficiently simple for the average small plant operator may he devised for determining hydrogen ion concentration in water purification control.
- —Experience on a number of waters in North Carolina indicate that the pH determinations furnish a very valuable assistance jn securing more effective and economical results.
- —There is considerable variation in the optimum for various waters and it is necessary that large numbers of waters throughout the country he investigated in this respect, before all facts in regard to pi I control can he definitely determined.
(Excerpts from paper read before the Detroit Annual Convention of the American Water Works Association.)