University of Texas Bulletin No. 1752: September 15, 1917 BUREAU OF ECONOMIC GEOLOGY AND TECHNOLOGY DIVISION OF CHEMISTRY Boiler Waters: Their Chemical Composition, Use, and Treatment By W. T. READ, Chemist With a Preface By E. P. SCHOCH, Head of the Division Pultlilhell by the Unlnrsity l!!ix times a month and entered as seeond-clase matter at the postofflce at AUSTIN, TEXAS B19t-11S.3m. University of Texas Bulletin No. 1752: September 15, 1917 BUREAU OF ECONOMIC GEOLOGY AND TECHNOLOGY DIVISION 01<' CHEMISTRY Boiler Waters: Their Chemical Composition, Use, and Treatment By W. T. READ, Chemist With a Preface By E. P. SCHOCH, Head of. the Division .Published by the University six times a month and entered al secend-class matter at the postomce at AUSTIN, TEXAiil The benefits of education and of useful knowledge, generally diffused through a community, are essential to the preservation of a free govern­ment. Sam' Houston Cultivated mind is the guardian genius of democracy. . • • It is the only dictator that freemen acknowl­e<}ge and the only secmity that free­men desire. Mirabeau B. Lamar TABLE OF CONTENTS Preface. Aid on Water Softening Given by The University of Texas. By E. P. Schoch. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plan of Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Purpose of Thill Bulletin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Service Rendered by the Division of Chemistry. . . . 7 7 7 8 BOILE.R wATERS: THEIR CHEMICAL ·coMPOsrnoN, USE AND 'l'REATMENT By W. T. Read Chapter One. Boilers- Effects of Minerals in Water Used in Steam 1. Common Salts in Boiler Water Supplies. . . . . . . . . . . . 10 CaCO,, Ca(HCO,),, MgCO,, Mg(HCO,),, Caso•. MgSO,, MgCl,, NaCl, Na,SO,, Na,CO,, NaNO,. Potassium salts. 2. What an Analysis Means and What It Does Not Mean 14 Ions, Ion Combinations, Report by Ions, Equivalents (Note: Calculation of Salts from the Amount of Ions), Conversion of Units, Conversion of Salts to Ions. (Note: Table of Factors for Calculating Ions from Salts). 3. Scaling ...... . . .... .... . ........ .... .. . ......... 24 What Makes Scale, Classes of Hardness, Effect of Oil, For­mulas for Scale (Note: Calculation of Scale from Formulas), Effec.ts of Scale (a) Space Loss, (b) Heat L'Oss, (c) Cost. 4. Corrosion ............. .. ........-. . . . . . . . . . . . . . . . 32 Causes of Corrosion: (a) Free Acid, (b) Magnesium Salts. (c) Gypsum Scale, (d) Oil, (e) Gases, (f) Galvanic or Elec­trolytic Action, (g) Oxygen, (h) Alum Coagulants. Formula for Corrosion, Embrittlement of Steel by Caustic Soda, Cor­rosive Effect of Very Pure Water, How to Prevent Corrosion. 5. Foaming ......... . ........... . .... . ............. 40 Foaming and Priming Defined. Causes of Foaming, (a) Sus­pended Matter, (b) Alkalies, (c) Boiler D efects. H-0w to Prevent Foaming, (a) Filtering and Softening, (b) Bl.owing Down, (c) Skimmers and Traps (d) Anti-Foaming Com­pounds, Foaming Formulas. Chapter Two. How to Stop Boiler Troubles Due to Bad Water­ 1. Boiler Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Common Chemical Ingredients, (Note: Some Typical Boiler Compounds), When to Use Compounds, Objections to Com­pounds, (a) Sludge, (b) Foaming, (c) Lack of Definite 'Amx>unt, (d) Bad Boiler Practice, Mechanically Acting Com­pounds, Legitimate Compounds. 2. Preheaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Preheater Not a softener. (Note: Change of Water in Pre­heater.) 3. Water Softeners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Lime Soda Process, (a) Chemical Reactions, (b) Cold Inter· mittent Process, (c) Cold Continuous Process, (d) Hot Con­tinuous Process, (e) Efficiency, Artificial Zeolite or "Per­mutit" Process, (a) Chemical Reactions of "Permutit,'.' (b) Construction of Softener, (c) Method of Operation, (d) Efficiency, (e) Control, (f) Advantages and Disadvantages, Lime Permutit Process, Iron and Manganese Removal with Permutit. 4. Filters .......................................... 70 Mud Removal, Sludge Removal, Type of Filter, Danger of Incorrect Filtration and Coagulants. Chapter Three. Cost, Construction, and Operation of Water Softeners­ 1. Cost of Standard Patented Water Softeners. . . . . . . . . . 73 Variation of Cost with Size, Quotations for Each Individual Case, Municipal Plants. 2. Cost of Home-Constructed Softening Plants. . . . . . . . . . 75 Limited to Small Intermittent Plants, Value and Limitations -Of Home Constructed Plants, General Plan of Home-Con­structed Softener, Construction Details of a Reinforced Con­crete Tank Softener, (a) Size and Capacity of Reaction Tank, (b) Reinforcing (c) Concrete Mix a'r!'d Waterproof­ing, (d) Storage Tank, (e) Cost of Construction of Concrete Softening and Storage Tanks, (f) Pipe and Fittings for the Softener. 3. Cost of Chemicals .......................... ' . . . . . 80 Calculation of Amount of Chemicals Needed from an Analysis of the Water, Table of Factors for Use with Hypothetical Combinations, Variations in Cost of Chemicals, Hot Process Softener as a Preheater. 4. Method of Control of an Intermittent Lime-Soda Softener ...... .. ................ .. ...... . : . . . . 84 Necessity for Chemical Control, How to Fix Feed of Chemicals t-0 Meet Variations in Water, Use of One Constituent to Fix Amount of Chemicals, Preparation of a Table for Amount install a steam boiler is asked to send, by prepaid express, about one gallon of the water used or to be used. The water should be sent in a clean glass bottle, but not in a jug. The bottle should be closed with a well-'6tting stopper, and the. latter se­cured by tying a cloth over it. A letter should be sent with it which gives the source of the water, the state of the boiler-that is, whether it is new or old, and, if old, whether or not trouble from scale formation, corrosion, or foaming has been experienced. After the water has been examined, the applicant will be ad­vised what to do. If he puts in a softening apparatus, the Division of Chemistry stands ready to help him by keeping a chemical check on its ·operation, and if necessary, to send a man to inspect it. All of this is done free of charge except for the traveling expenses of an inspedor sent out on request. • PURPOSE OF Tms BuLLETIN.-The foregoing plan of opera­tion has been followed for about a year. While we have reached comparatively few industrial plants during this time, yet we are more convinced than ever of the public economy of proper boiler water softening; and in order to reach a larger number with less labor of correspondence, and to advise steam users more extensively on this question, we have issued the following Bulletin. It has been written by Mr. W. T. Read, who has had special charge of this work since June, 1916. The Bulletin has been written to give all necessary informa­tion as to the causes of boiler troubles and the remedies for them. It is intended particulary for steam users who operate small plants, to educate them away from slip-shod, wasteful, old­fashioned, careless, and dangerous methods of ''letting things slide,'' pounding and chiselling at boiler scale, dosing boilers with any kind of a compound that a salesman may urge or suggest, worrying along with foaming boilers, and in general, wasting money, time, and labor. It has been written for the general reader who knows nothing about chemistry or the tech­nical handling of a steam plant, and also for the technical man in actual charge of a. plant who wants to have definite informa­tion for securing and operating the proper sort of a softening plant. With this Bulletin, the Division of Chemistry intends to con­tinue its plan of operation stated above and to make its cam­paign for boiler water softening not only more extensive, but more ·effective. We sincerely believe that the perusal of this Bulletin will convince many boiler operators of the economy of using softened waters and thus help us to reduce what is now a great national waste. OTHER SERVICE RENDERED BY THE DIVISION OF CHEMISTRY.­ 'l'his Division has been established for the purpose of giving Chemical aid in the recognition of development of Texas raw products, and to do experimental work in establishing or ad­vancing industries using such raw products. At pr~ent, its investigations are concerned with the :following materials: coal, lignite, petroleum, natural gas, graphite, paper pulp material, clay, cement and plaster material, water used for steam boilers and for irrigating, and other minerals of economic importance. To aid in the exploration for the materials in Texas, the Di­ vision of Chemistry offers to examine free of charge, any sample of these substances collected in such a manner as to secure a sample which is a fair representative of the material or which has a composition that is the average of all the different por­tions of the material represented by the sample. In addition to that, it is necessary that certain information be furnished concerning the material. Details concerning the sampling and the information to be supplied will be sent on request. The extent of the examination made will vary from a critical "look­ing at'' or visual examination of the sample. to a complete chem­ical analysis, as may be considered necessary and profitable by the staff of this Division. Boiier Waters: Their Chemical Composition, ..--Use and Treatment BY w. T. RK\D CHAPTER ON'E EFJ<'ECTS OF MINERALS IN WATER USED IN STEAM BOILERS 1.n general, all waters contain the same substances in solution in all sorts of amounts and proportions. Besides dissolved salts, most waters have certain gases in solution, most commonly car­bon dioxide and air. This latter gas is usually spoken of as dis­solved oxygen, since the nitrogen gas has no effect, and it is only the oxygen that causes trouble. There is also the special class of waters known as ''sulphur waters.'' These contain hydrogen sulphide gas in solution. In addition to the dissolved minerals, most surface waters-especially Texas river waters in a wet season-carry a great deal of mud, fine clay, and vegetable mat­ter in suspension. All of these have their part in causing boiler troubles. While the gases can be gotten rid of in open feed I water heaters, and the mud settled and filtered out, the dissolved solids that cause scaling and corrosion cannot be eliminated or changed so that their harmful effects are neutralized except by chernical treatSnent. Hence before taking up the causes and remedies for boiler ailments, it is well to give a list of all the important salts that are commonly found in boiler waters, with some of their physical properties and chemical reactions. 1. Common Salts in Boile1: Water Supplies·• \ CALCIUM CARBONATE AND BICARBONATE Marble, chalk and limestone, though different in physical structure, are mainly composed of the same substance, calcium carbonate, CaC03. Water charged with carbon dioxide changes calcium carbonate to the more soluble bicarbonate, Ca(HC03 ) 2• Nearly all waters contain carbon dioxide, or carbonic acid gas, dissolved in them, because this gas, coming from decaying vege­table matter, from the smoke of all fires, and from all breathing animals, is readily dissolved by water. When water containing this gas comes in contact with lime-stone, it acts as an acid toward the calcium carbonate, and dissolves it extensively, forming cal~ cium bicarbonate or calcium acid carbonate, according to the equation : Ca.co3+H2o+co2=Ca(HC03)2 As long as there is plenty of carbon dioxide in the water, this salt remains in solution, and the water remains clear. But if this gas is driven out of solution by boiling or simply by long standing, the reverse process begins, the acid carbonate of lime breaks down, giving up its carbon dioxide, and going back to the ·original carbonate of lime or calcium carbonate. Since the latter is not noticeably soluble in water, it separates from the solution as a white, crystalline, solid substance, the form of which depends on the conditions of the solution when the change takes place. The first change, that of solution, is the cause of most of the cutting away of the foundations of dams built on lime rock. The other change, that of precipitation from solution, is responsible for a good part of the scale formed when such water is boiled in any vessel ranging from the stove kettle in the kitchen to the big water tube boiler· in the factory. The change is the reverse of that expressed by the chemical equation given above. Ca(HC03) 2=H2o+co2+CaC03 The carbon dioxide gas escapes, the acid carbonate is changed to the solid carbonate, and scale results. MAGNESIUM CARBONATE AND BICARBONATE In a general way what has been said about calcium carbonate and calcium bicarbonate may be said about magnesium carbonate arn'l magnesium bicarbonate. A great many limestones contain magnesium carbonate. If the amount is considerable, the lime­stone is called a dolomitic limestone, and when the magnesium bicarbonate begins to approach an amount roughly equal to the calcium carbonate, the rock is called dolomite. Water charged with carbon dioxide dissolves magnesium carbonate in the same way as it dissolves calcium carbonate, but the magnesium car­bonate, formed when this gas is gotten rid of, is much more soluble than the calcium carbonate produced under the same conditions. So in order to get rid of the magnesium before the water gets into the boiler, it has to be changed to the hydrate or hydroxide, which compound of magnesium is very soluble. CALCIUM SULPHATE Chemical formula, CaSO,. This salt is familiar to every one in the form of gypsum. In this form it contains a definite amount of water in its crystals, which is called water of cryst~l­lization. When this water is partly driven out, plaster of Paris is formed, and the setting qualities of this substance are due to its taking up water again and forming interlacing crystals. Calcium sulphate is fairly soluble, water at ordinary tempera­tures being capable of carrying nearly 2400 parts per million (i. e. 2400 parts for every million parts of its water solution), or 140 grains per U. S. gallon. When the temperature of the solution rises above the boiling point of water under atmospheric pressure, calcium sulphate becomes appreciably less soluble. At 150" C. or 300° F. it will only go into solution to the extent of 500 to 700 parts per million (30 to 40 grains per U. S. gallon). As a water in . steam boilers is concentrated under heavy pressure and at high temperature, any gypsum crystal­lizing out from it is deposited in the form of a very hard, smooth scale. MAGNESIUM SULPHATE Chemical formula, MgSO,. This is commonly known as "Epsom Salts." It is a very soluble white solid. If present . alone in a boiler water, its only action would be to react with the water itself. This takes place only when the salt separates out on the boiler plates at the water edge. Magnesium sulphate at this point reacts with water to produce magnesium hydroxide and sulphuric acid. The niagnesii:im hydroxide is very insoluble and forms a scale, and the sulphuric acid is very corrosive and will attack the material of the boiler. The chemical reaction is as follows : MAGNESIUM CHLORIDE Chemical formula, MgC12 • This is also a very soluble salt, the only action of which, when present alone, is to react with water at the water edge like magnesium sulphate according to this reaction : Magnesium chloride reacts with water more readily and more extensively than the magnesium sulphate, and through the hydrochloric acid formed corrodes the iron of the boiler. Hence it is a particularly objectionable ingredient of boiler waters. This hydrochloric acid is a gas, and if the solution is concen­trated enough, that is 1 per cent or more of the gas in the solution, it will distill out and corrode the boiler above the water line in the same way that the acid solution attacks it bel~w the water line. SODIUM CHLORIDE Chemical formula, NaCL This substance is what is known as "common salt." It is very soluble, and harmless with reference to scale production. As will be noted in the section on Corro­sion, it simply furnishes the salt solution necessary to battery action, if the other conditions necessary for corrosion are pres­ent. If salt is allowed to accumulate by continual concentration of the boiler contents, and if suspended matter is present, foam­ing is likely to result. SODIUM SULPHATE Chemical formula, Na2S04• What has been said about sodium chloride applies equally well in describing the effect of sodium sulphate, with the ,exception that it is not as active in promoting battery action as the chloride-in other words, its solution is not as good a conductor of electricity as the solution of sodium chloride. SODIUM CARBONATE Chemical formula, Na2C03• This is familiarly known as ''washing soda'' or ''soda ash.'' In waters containing consider­able carbon dioxide, part of it is changed to sodium bicarbonate, NaHC03• When a solution of sodium bicarbonate is boiled, it gives up carbon dioxide rather rapidly and finally passes en­tirely to sodium carbonate. Since this takes place readily, it is not necessary to consider the final effect of anything but the last named salt. This salt is very soluble. It is regarded as being more prone than any other of the sodium salts to cause foaming when in concentrated solutions. It also reacts with water to some extent to give caustic soda, or sodium hydroxide. This substance, when concentrated, is regarded as a source of danger \ in leading to embrittlement of the steel and consequent cracking. As will be noted under the head of Corrosion, this is a much de­bated and still unsettled question. SODIUM NITRATE Chemical formula, NaN03• This is commonly called "Chile saltpeter." Not many waters in Texas have any appreciable amounts of this salt, except some deep well waters in parts of North Texas, especially northwest from Mineral Wells. It is extremely soluble, and its only effect is probably that of aiding corrosion, because it gives up a part of its oxygen readily. POTASSIUM SALTS These are almost identical in their behavior with sodium salts, and for all practical purposes, the alkali metals, sodium and potassium, may be determined and reported together as sodium. Analyses of a large number of Texas waters have been made, and these waters found to contain an average of ten to fifteen times as much sodium as potassium. Hence the error in con­sidering everything under the single head of sodium is not a great one. 2. What an A.nalysis Means and What It Does Not Mean. IONS In order to gain a clear idea of the meaning of a chemical analysis of water, it is nece~sary to outline in a general way the most recent and generally accepted theories as to the form which mineral matter has in solution. It is not to be expected that one who has never studied chemistry can under~fand the theory thoroughly, but any one can at least get an idea of the general principles involved. For the past twenty years the ''Ion Theory'' has been regarded by chemists as being so thor­oughly borne out by actual facts that it has been made the basis of all analytical chemistry. Stated in the simplest possible terms, the general idea of this ion theory is this. When a sol­u blc salt goes into solution, it is no longer ,electrically neutral, bnt breaks up into two (or more) parts called ions, one (or more) of which is positively charged, the other (or others) negatively charged. For instance: when the chloride of sodium, or common salt, is dissolved in water, it breaks up into sodium (+)ions, each carrying one unit of positive electrical charge, and chloride(-)ions, each carrying one unit of negative elec­trical charge. These ions are constantly moving about and 'Col­liding with each other. If one sodium (+)ion meets another sodium(+ )ion, both being positively charged, they repel each other. In the same way if one chloride(-)ion meets another chloride (-)ion, both being negatively charged, they will also repel each other. The only other possibility is the meeting of a so­dium (+)ion with a chloride (-)ion. There is an attraction and combination to form sodium chloride, but only for an instant. Because of the foree or tendency in tl:re water to produce ions, they break apart once more. Suppose the water is gotton rid of by evaporation. The ions combine gradually as the amount of the water grows less, and its total ionizing power diminishes, until when a'.11 of the water is gone, solid sodium chloride is left in exactly the same form in which it was before. ION COMBINATIONS Suppose some potassium nitrate were added to this salt solu­ tion. Like the sodium chloride, most of it breaks up into these two ions, namely: positively charged potassium (+)ion and neg­ atively charged nitrate(-)ion. Now this solution contains the ingredients for making four different combinations: sodium chlo­ ride, potassium chloride, sodium nitrate, and potassium nitrate. Not a single one of these four salts is present in the water as an individual salt, but the solution contains four kinds of ions: sodium(+), potassium(+), chloride(-) and nitrate(-). Un­ der speci~l ·~onditions, such as evaporation, the first salt which reaches its limits of solubility-which depends upon both its quantity and its soiubility-will be formed first in solid form i:egardless of whether it was originally present or not. If a third salt, say magnesium sulphate is added, we have the necessary ingredients for making the following nine different salts from the six ions present: sodium chloride, sodium nitrate, so­ dium sulphate, potassium chloride, potassium nitrate, potassium sulphate, magnesium chloride, magnesium nitrate, magnesium sulphate. With a solution containing these six kinds of ions, it would be absolutely impossible to find out which three of the possible nine salts were originally put into solution. · The pos­sible formation of nine different salts is shown in the following diagram. ff) 0 z z REPORT BY IONS All the properties of a solution are due to the ions in solution. They each behave like individual substances. Ions are the only things a chemist determines when he makes an analysis. He identifies each ion by certain tests, and finds out accurately how much of each one is present. The only logical way to report a water is by ions, since nothing but ions are determined or can be determined in a water. This is the method employed by the United States Geological Survey, and is the only one used in this bulletin. It is not only just as easy, but it is much more accu­rate to calculate the scaling, corroding, and foaming tendencies of awater from the ions than from any "hypothetical combina­tions'' that might be devised. The following ions of soluble salts are ordinarily found in ground and surface waters. After the name of ·each ion is ·given its chemical symbol, its equivalent weight, and the re­ ciprocal of its equivalent weight, which is a factor frequently used by chemists in calculating water analyses. Reciprocal of Name of Ion Symbol Equivalent Equivalent Weight Weight as· Factor for Calculations Calcium ..... ... . ca++ 20.04 .0499 Magnesium ... ... . Mg++ 12.16 .0822 Sodium . ...... .. . Na+ 23.00 .0435 Potassium .. . .... . K+ 39.10 .0256 Hydrogen ....... . H+ 1.01 .9920 Car.bonate ....... . CO,-­ 30.00 .0333 Bicarbonate ..... • .HCO,_ 61.01 .0164 Sulphate .. ..... . . so.-­ 48.04 .020S: Chloride .. ... . . . . Cl­ 35.46 .0282: Nitrate . . . . . . . . . . NO,­ 62.01 .0161 Besides the ions of soluble salts, water .may contain iron, aluminum and silicon in colloidal solution as oxides; iron oxide, Fe20 3; alumina, Al2 0 3; and silica, Si02• These three substances do not enter into any combination in waters, and they are elec­trically neutral. They are collected and weighed as such-that is in this chemical form-and hence analyses should present,' and usually do present'the amounts of the three oxides which are in a definite amount of a water. The weight of the iron alone m~y be obtained by multiplying the weight of its oxide by .6995; the ]8 University of Texas Bulletin weight of the aluminium is gotten by multiplying the weight . of its oxide by .5303. Silica is present in practically all waters in small amount. Carbon dioxide gas, or carbonic acid gas(C02 ) is also often present dissolved in the water, and is constantly evolved from waters containing the acid carbonates or bicarbo­ .nates as these break up to form carbonates. Hydrogen sulphide 'gas (H2S) is found in so-called sulphur waters. It escapes if .exposed to the air. Acids which form the hydrogen ion exten­sively are rarely present in Texas waters. When such an acid is present, the water has a sour taste, and thus the acidity can be easily detected. In the course of more than a thousand an­alyses of water made in the University of Texas laboratories, only three waters were found which were sour, and these were from a s:and-clay formation in which there was no limestone. If limestone ·had been present, the acid would have been neu­tralized. EQUIVALENTS The analyses of waters usually present simply the number of parts per million of each ion present; but it is advisable to add a second number, which has the following significance. When multiplied by 60.62 times 10 raised to the 22nd power, or 60, 620,000,000,000,000,000,000,000, which is known as the '' Avo­gadro Number", it gives the "number of individual combining portions" of that particular ion present in 1000 litres, which in­dividual combining portion is of such weight that it will combine one to one with the corresponding portion of any other ion of opposite chemical nature. Thus one individual combining por­tion of a negative ion or acid radical is such an amount as will just combine with one individual combining portion of a posi­tive ion or metal. Nature requires that any definite amount of water must con­tain as many individual combining portions of negative ions as of positive ions. This is shown in the following figures, which give the composition of a water as actually found by analysis: Positive Ions Negative Ions Calcium ion.-. .138X.0499=6.89 Carbonate ion 12X.0333=0.40 Magnesium ion.. 29 X.08 22=2.38 Bicarbonate ton 290X.0164=4.76 Sodium ion ... 69X.0434=2.99 Sulphate ion . . 198X.0208=4.12 Chloride ion.. . 106 X.0282=2.99 Total of gram-equivalents of positive ions ....· . . . ... 12.26 Total of gram-equivalents of negative ions.........12.27 Hence the numbers referred to in the preceding paragraph are obtained by multiplying the amount of an ion actually deter­mined by analysis in terms of parts per million by the reciprocal of its combining weight. This bears thus a fundamental relation to the absolute number of individual combining portions. The slight difference between the values of the sum of the number11 in the left column and the right column is due to undetermined traces of other ions, but these figures are sufficiently nearly equal to satisfy all practical purposes. These factors are employed to test the accuracy of an analysis and to determine the most probable source·of error. It is evident that these numbers so obtained in the foregoing calculations do not give the actual numbers of the individual combining portions of all the metal and acid ions, but that they must be multiplieQ. by the Avogadro Number mentioned above in order to give the actual number of individual combining por­tions. Since we are here concerned merely with a comparison or relation of two such. numbers, it is useless and unnecessarily troublesome to multiply them both by a constant value of such immense size as the Avogadro Number. For this reason we al­ways use only the simpler numbers. These simple numbers are known as the number of Gram-Equivalents of any water constit­uent present in 1000 litres of the water. CALCULATION OF SALTS FROM THE AMOUNT OF IONS By means of these numbers it is very easy (though rarely neces­sary) to calculate the total amount of any one salt present in a water. For example, let us consider the question : "How much calcium carbonate (in parts per million) corresponds to the ions present in the water; the analysis of which is given above?" To answer this question, we proceed as follows: Calcium carbonate is formed by a union of the calcium and of the car1bonate ions. Since there is only 0.4 gram-equivalents per 1000 litres of the carbonate ion (which amounts to 12 parts per million), then only 0.4 gram-equivalents per 1000 litres of the calcium ion can be coupled with it to form calcium carbonate. Since 6.89 gram­equivalents per 1000 litres of calcium ion amount to 138 parts per million, 0.4 gram-equivalents amount to 0.4 divided by 6.89, multi­ 40 plied by 138 or of 138, which equals 8 parts per million. 689 Hence in parts per million the total amount of calcium carbonate is 12+8=20 parts per million. This illustrates how the amount of any salt may be figured out from the analytical report of ions as given in this publication. If it is desired to find out how much calcium bicarbonate can be formed from the remaining calcium ions, continue the calculations as follows: It is evident that onl.y 6.49 gram-equivalents of calcium ion are left to be considered (6.89-0.4) . Since this is greater than the number of gram-equivalents of bicarbonate ion ( 4. 7 6), it follows that the amount of this combined form is limited by the latter number, and only 4.76 gram-equivalents of the calcium ion can be coupled up with the bicarbonate ion. The ·number of parts per million of the bicarbonate ion is 290; that corresponding to 4.76 476 gram equivalents of calcium is --of 138 or 95.3 parts per million 689 of calcium. Hence the total amount of calcium bicarbonate is 290+ 95.3, or 385.3 parts per million. This second calculation shows how to ealculate a combination of ions (or salt) with th-e provision that the ions forming a previously calculated combination are not avail­ able for the salt to be calculated. The process here given may be continued until all the ions are combined as salts. Years ago, before it was known that the salts in very dilute solu­tions, exist in the form of separate or free ions, the ions were arbi­trarily assumed to be combined to form salts, and the reports stated the amounts of the combinations; but since there is absolutely no basis of fact that warrants the assumption that any particular com­binations are formed rather than any other equally possible, and since no particularly useful purpose is served in representing the ions as being combined, this manner of reporting water analyses is now befog gradually disp-ensed with, and has not been employed in this publication. To illustrate how arbitrarily different chemists consider the ions to be combined, and how misleading this procedure is, we present here three different reports of the same water, for which the an­alytical results are given below. These three different reports are made on the basis of three different, but definitely used, assumptions concerning the combination of ions. f.11 quantities in plain figures 11re in parts per million, while those in parentheses are in gram equivalents per 1,000 litr_es. Name of Salt Method 1 Method 2 Method 3 Calcium Sulphate .. 119 (1.75) 140 ( 2.06) 280 (4.11) Calcium Carbonate. 20 (0.40) None 20 (0.40) Calcium Bicarbonate 385 ( 4.7 4) 387 ( 4. 7 5) 195 (2.40) Magnesium Sulphate 143 (2.38) 124 (2 .06) None Magnesium Chloride None 17 (0.38) None Magnesium Bicar­ bonate ... . .. . . None None 172 (2.36) Sodium Carbonate .. None 21 (0.37) None Sodium Chloride ... 175 ( 2. 99) 153 (2.62) 175 (2.99) 842 (12.26) 842 (12.26) 842 (12.26) It will be noted that each method gives different results on more than one constituent. In one method, certain salts are reported as present which are reported as absent in another method, yet all of these reports are based on the same analysis. It is evident that some of them must be misleading; as a matter of fact, all three of these reports are somewhat misleading, and the only report that shows exactly the composition of the water is the report given above in terms of ion§. The common ions which occur in ground and surface waters admit of the following possible combinations: Cal­cium chloride; calcium nitrate; calcium carbonate; calcium bicar­bonate; calcium sulphate; magnesium chloride; magnesium nitrate; magnesium carbonate; magnesium bicarbonate; magnesium sulphat-e; potassium chloride; potassium nitrate; potassium carbonate; potas­sium bicarbonate; potassium sulphate; sodium chloride; sodium ni­trate; sodium carbonate; sodium bicarbonate; sodium sulphate. CONVERSION OF UNITS Sometimes it is desired to have the amounts of the salts in a water expressed in other units. An analysis is easily changed from one basis to another by use of a simple factor. To convert parts per million to grains per U. S. gallon, mul­ tiply by .0583. To convert grains per U. S. gallon to parts per million, mul­ tiply by 17.14.' To convert parts per million to pounds per 1000 gallons, mul­ tiply by .008i33. To convert grains per U. S. gallon to pounds per 1000 gallons, multiply by 7. CONVERSION OF SALTS TO IONS Since many analyses, particularly qld analys€s, are only avail­able in terms of possible salt combinatiO:n.s, it is very often nee­ essary to calculate these reports back to the terms of the ions that were originally determirn~d in the water. This can be done very easily by the use of gram-equivalents or pe.rcentage factors. The following method is simpler and more nearly in conformity with other simple engineering calculations. Multiply the salt by a factor which represents the percentage of the desired ion in the salt. This 'Nill give the amount of the desired ion. The differ­ence between the amount of the ion and the original amount of the salt will give the amount of the other ion which went to make up the salt. Apply this procedure to each salt, collect like ions and add them, and the final result will be in terms of ions. Following is a practical example. Given the following analysis in terms of salts, calculate to terms of ions. Hypothetical Salt Combinations Parts per Million Calcium Sulphate 119 Calcium Carbonate 20 Calcium Bicarbonate 385 Magnesium Sulphate 143 Sodium Chloride 175 To -calculate calcium sulphate, CaS04 , to calcium ion. Ca++ and sulphate. ion, SO-!--, multiply the amount of calcium sul­ phate by the factor, .2944. (29.44 per cent of calcium sulphate is ealcium.) 119 X .2944=35 parts per million calcium ion in calcium sul­ phate. Subtract the amount of calcium ion so obtained from the total amount of calcium sulphate and the result will be the sulphate ion in this salt. 119-35=84 parts per million sulphate ion in CaS04 • Calculate calcium carbonate to calcium ion and carbonate ion. 20X.4004-8 parts per million calcium ion in CaC03 20-8=] 2 parts per million carbonate ion in CaC03 Calculate calcium bicarbonate to calcium ion and bicarbonate ion. 385X.2472=95 parts per million calcium ion in Ca(HC03 ) 2 385-95=290 bicarbonate ion in Ca(HC03 ) 2 Calculate magnesium sulphate to magnesium ion and sulphate ion. 143X.202' KNO, to K . ............ .3866 Mg(N0,) 2 to NO, ...... . .8360 KNO, to NO,. ............6134 MgCO, to Mg ......... . .2884 K,CO, to K ...... .......5658 MgCO, to CO, .. ....... . K,CO, tp CO,. .......... .4342 Mg(HCO,), to Mg..... . .1662: KHCO, to K.............3905 Mg (HC0,) 2 to HCO,. . . . . 8338: KHCO, to HCO,. ... ......6095 MgSO, to Mg ......... . .2020' K,SO, to K ........... .4487 MgSO, to SO, ......... . .7980' K,SO, to SO,....... . . . ..5513 CaCI, to Ca ...... ..... . .361() NaCl to Na ..............3935 CaCI, to Cl ........... . .6390 NaCl to Cl. .............6065 Ca(NO,), to Ca ........ . .2442 NaNO, to Na ............2706 Ca (NO,), to NO,. ...... . .7558 NaNO, to NO, ........ ...7294 CaCO, to Ca........... . .4004 Na,CO, to Na ...........4340 CaCO, to CO, .......... . .5996 Na,CO, to CO,. . . . ". ......5660 Ca(HCO,), to Ca ....... . .2472 NaHCO, to Na ...........2738 Ca(HCO,), to HCO,. .... . .7528 NaHCO, to HCO,. ........7262 Caso, to Ca ........... . .2944 Na,SO, to Na . .. ........3238 Caso, to so, ........ .. . .7056 Na,SO, to SO,...........6762 Fe,O, to Fe.. .......... . .6995 MgCI, to Mg............2554 !\.1,0, to Al . . ......... . .5303 3. Scaling WHAT MAIIBS SCALE The chief producers of boiler scale in filtered water are the salts of calcium and of magnesium. Of course, mud or sus­pended matter that is allowed to get into a water supply will form scale, and silica in dissolved or colloidal form (that is, not in actual solution, but so finely divide.d that its suspension be­haves like a solution and cannot be removed by filtering) will .also go to make up scale. Generally the amount of silica in a water is very small, although one instance has been cited in whicl;l a water carried ne~rly five hundred parts per million (30 grains per U. S. gallon) of .silica, and gave a scale that was like glazed porcelain. Iron and aluminium, though both scale formers, are not important, because they are present only in small amounts in most waters. CLASSES OF HARDNESS Thus it appears that the big scale forming constituents are the lime and the magnesia compounds. These: constitute two ·classes of "hardness." The first class is composed of the bicar­·.bonates of calcium and magnesium, and is called ''temporary hardness." In other words, if the water contained only calcium ·and magnesium ions and bicarbonate ions, this water would have -0nly temporary hardness. As has been pointed out above, if a water containing these ions is boiled, the bicarbonate ions will f dissolved oxygen will be noted in another connectjon. (f) Galvanic or Electrolytic Action.-A big cause of corro­sion is the galvanic action (or electrolytic action) which takes place in boilers. To understand this, let us consider the con­struction of an ordinary battery cell, which is familiar to en­gineers. Any two different metals in a salt solution,_ when connected, will set up galvanic action, even though it may be too feeble to be ordinarily noticeable. Suppose a piece of mill .scale, slag, sulphur or manganese is at the surface of the boiler iron, and in contact with the salt solution, which ordinary ground or surface waters are. The piece of mill scale, slag, etc., will act as one pole; a spot of pure iron on the surface of the boiler. wall will act as the other pole; they are both ''dipped'' into the liquid extending from one to the other (whiich serves as the battery liquid) ; and the iron plate to which they are attached serves as the "connecting wire" through which the electric current flows. One of the above two spots-the posi­tive pole-will act just like a zinc pole in a battery cell, and the material there will be consumed or corroded away gradually. At the other spot-the negative pole-hydrogen will be evolved, or will be oxidized to water by free oxygen present in the water. Any difference in the crystalline structure or chemical composi­tion of the steel in different parts of the boiler or difference in structure due to mechanical strains or deformations (such as occur in punching rivet holes) give the.se different spots the property of acting as opposite poles of a galvanic cell. (g) Oxygen.---'When corrosion takes place, a certain amount of hydrogen gas is formed at the negative pole. When free acid attacks iron, the metal goes into solution and the hydrogen of Boiler ·waters 35 the acid appears as a gas. This may cover the surface of the metal as a protecting film and slow down and stop the action, an effeet well illustrat.P.d iu the polarizing of battery cells. But University of Texas Bitlletin i£ there. is dissolved oxygen in the water, or substances which give up oxygen easily-i. . ~., nitrate&-in the water, then this film of hydrogen is continually removed. by simple oxidation, and thf' corrosive action progresses more rapidly. Hence oxygen may be spoken of as a ''stimulating factor'' in cori:osion. (h) Alum Coagulants.-In this connection, it is well to call attention to the danger that may result and has re.suited in many cases from the use of coagulants, espeGially when used in excess in order to clear muddy water. The principal coag­ulating agent is sulphate of aluminium or ordinary alum. Un­less there is sufficient alkali present to neutralize it, the sul­phuric acid produced by the reaction of alum with hot water will attack and ruin the boiler. Only small amounts 0£ eoag­ulants are necessary with an efficient filter, and if the water is closely watched and always kept slightly alkaline, this danger is prevented. FORMULA FOR CORROSION As has been stated, it is impossible to foretell from the an­alysis of a water alone whether or not corrosion will take place when the effect is due to galvanic or electrolytic action. How­ever, it is possible to tell with a very reasonable degree 0£ prob­ability whether the ions present are of the kind that produce acid effects in a boiler. To calculate the corrosion possible from . this source, the best corrosion formula at present available is that of Stabler, given below: Coefficient of Col'Tosion (0)-H+.1116 Al+.0361 Fe+.0828 Mg-.0336 003-.0165 HC03 By this, the probability and degree of corrosion, due to acid forming constituents, may be c.alculated from the ions present. Of course no account whatever can be taken in such .a formula of corrosion due to galvanic action or to the effects 0£ dis­solved gases. According to the formula of Stabler, the following is the class­ification of waters according to their acid forming tendencies~ 1. Corrosive. If the coefficient be positive the water will cer­tainly corrode the boiler. 2. Non-corrosive. If the coefficient plus the factor .0 5 0 3 mul­tiplied by the amount o( calcium ion present (expressed in parts per million) is negative, no corrosion will take place due to acidity. 3. Semi-corrosive. If the coefficient is negative, but the co­efficient plus .0503 times the calcium is positive, corrosion due to acidity may or may not occur, the probability of corrosion varying directly with the value of the expression: ·coefficient of corrosion plus .0503 times the calcium. Given a typical analysis: Parts ·per Million Calcium .... . ............. . ............... 85 Magnesium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 9 Sodium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • 50 Iron ..•...·........ ...... . . . . . . . . . . . . . • . . • 6 Aluminium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • 4 Carbonafes . . . . • . . . . . . . . . . . . . . . . . . • . • . . . . . . 6 Bicarbonates ........................•..... 359 Chlorides .......................•......... 170 Sulrphat~s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . 57 Employ the formula of Stabler. 4 6 69 (Aluminum) (Iron) . (Magnesium) x .1116 = x .0361= x .0'828 = .45 .22 5.71 ---­ 6.38 6 (Carbonates) x .0336 = .20 359 (Bicarbonates) x .0165 = 5.92 6.12 6.38 6.12 +.26 Since the coefficient of this water is positive, even though small, it will cause some corrosion. In order to show the disadvantage, in this connection, of an analysis_reported with ions combined arbitrarily to certain. salts, calculate the "hypothetical combinations" by one of the gen­eral methods. The following results will be obtained for the above analysis. Parts per Million · Sodium Chloride ... ... .. ........... ....... 127 Magnesium Chloride . . .... .... .... . .......• 125 Magnesium Sulphate . . . . . . . . . . . . . . . . . . . . . . . 72 Magnesium Bicarbonate .............. ....... 138 Calcium Bicarbonate ... . ...... . ...... : .. .. . 324 From this statement of the composition ofthe water, it appears to be decidedly corrosive, due to the presence of magnesium salts other than carbonates, while in reality it is only slightly corrosive. "\\'bile the Stabler formula fails in that it does not take ac­count .of any other cause of corrosion but acidity, it is very valuable as far as it goes. Any water that gives a corrosive or semi-corrosive coefficient by this formula should be treated chemically outside the boiler so as to remove the cause of corro­sion, and in the boiler a slight alkalinity of the water should be rigidly maintained. EMBRITTLEMENT OF STEEL BY CAUSTIC SODA Within the last two or three years, a number of boiler ex­plosions have been charged to the embrittlement of steel from caustic soda, as the result of using artesian waters containing carbonate of soda, or boiler compounds in excess which are made up largely of this salt. While we are not at liberty to publish the details of the experiments made by an Eastern man­ufacturing concern of international reputation, we have gone over their results and give below some conclusions drawn from these results, the publication of which is authorized by the firm making these experiments. The effect of caustic soda seems to manifest itself in small, hair­line cracks in the seams below the water line, always starting from a joint, and in many cases so small that they could not be discovered without removing the rivets. The metal of the boilers so afi'ected seems to lose, in the parts that are so damaged, its toughness, and becomes quite brittle. One plant switched from a creek water containing no sodium carbonate to an artesian water carrying a little short of 12 grains per gallon, and immediately began to have trouble with embrittlement. More than twenty cases were reported in one region, and all the troubles had the common characteristic of being in distinct geographical districts, all were below the water line, and all in boilers using artesian water. In these districts where gypsum occurred in the water, there was no trouble with cra~king reported, and it is to be noted for chemical reasons that where gypsum occurs, sodium carbonate will be absent. It seems · evident that the material of the boiler was not at fault, since the same trouble happened with plates from various sources and of differep.t form. The boilers affected represented several different leading and reputable makers, and it is hardly likely that all of them could have put out unreliable products in the same region or at all. Th!) designs were as different as the makes. All manner of operating conditions obtained, without altering the cracking and embrittling troubles. The characteristics are all the same, the cracks being alike, and differing from cracks from all other causes, that have from time to time been found in steam boilers. All were below the water line and in seams. All the boilers, with one ex­ception, were fed with artesian water, the one ingredient common to all ·being sodium carbonate, which changes in part at least to the hydrate or caustic soda, and, depositing in the seams, sets up corrosion. The hydrogen that is freed is absorbed or occluded by the flt.eel, and brittleness is produced. Water taken from a boiler in service which showed this trouble had nearly 250 grains of sodium carbonate per gallon. Serious trouble occurred in a boiler using a compound made up largely of soda ash and caustic soda. In this case the water was almost pure except for the presence of the solids due to the compound, and in general it may be stated that there is no evidence to show that by the use of soda compound for treating waters where there are substances in the water with which reaction can take place, and where the treatment is not to excess, any trouble of this kind has occurred. In the case just cited, the­cracks were so small that they could not be seen until the iron was cleaned. Magnesium sulphate was use<). in a number of cases in order to get rid of the excess of sodium carbonate. The cracking and embrittlement seem to have ceased as the result of such treat­ment. The whole trouble appears to be due to occluded hydrogen. A great deal of work is being done just now on this subject, the results of which will be watched with much interest by boiler water users. In the meantime, if the alkalinity of the boiler water is continually kept down to a few grains per gallon, it seems quite certain that there is no danger to users of lime soda 11ofteners. CORROSIVE EFFECT OF VERY PURE WATER Absolutely pure water is rather corrosive for another reason. Its solvent power is unusually great, hence it dis8olves iron ap· prooiably. However, the addition of a little alkaline material prevents the dissolution of iron. HOW TO PREVENT CORROSION. All this may sound rather dangerous and terrifying to the steam user, but, as a matter of fact, corrosion can be very easily guarded against and prevented. The free acid can be neutral­ized by just :the right amount of alkali; the magnesium, iron, and aluminium salts can be gotten rid' of by proper softening and treatment; dissolved gases usually escape for the most part from the open feed-water heater; the water can be main_tained in a slightly alkaline condition; boilers can be made of a high ,grade iron, so free from impurities that it will not set up gal­vanic action; and correct boiler design can prevent causes due ·to faulty construction. A properly softened water freed from gases in the right kind of a preheater and fed into a boiler of the right material and construction, will not cause corrosion. The common-sense thing to do is not only to test the feed water to see if it carries corrosive constituents, and then use preventatives, but above all things to test water in the boiler, to ascertain whether or not it is deveioping conditions that will make corrosion possible. Slight alkalinity of the boiler water should be maintained, tendencies to acidity being continually .corrected by use of soda ash. 5. Foaming The common boiler trouble known to engineers as "foaming" is a familiar one. Since there seems to be some confusion be­tween the terms "foaming" and "priming," the liberty is taken bere of quoting from a letter from R. C. Bardwell, Chief Chem­ist of the Missouri Pacific . .FOAMING AND PRL'1ING DEFINED " 'Foarning' consists of the steam bubbles building upon the -surface of the water, instead of breaking and freeing steam. It becomes serious when the bubbles build up so rapidly and bang together long enough to reach the point where dry steam is supposed to be taken off, starting a sort bf siphoning action, and considerable water is carried over with the steam.'' " 'Priming' is the sudden evolution of steam from the heat­ ing surface, similar to 'bumping' in laboratory apparatus, and is due either to poor design of the boiler, and its being worked beyond its capacity, or to the sudden opening of the throttle. This is altogether different from foaming, and can be mechan­ically controlled to a large extent.'' CAUSES OF FOAMING It seems practically certain, however, that the main cause of foaming is not due to the concentration of soluble material in the water. Clean, clear water, free from all suspended matter, vegetable matter, and oil will not cause foaming under any reasonable conditions of concentration. The idea that foaming is due to an increase of surface tension because of increased concentration of soluble salts is not borne out by theory or facts. The surface tension of the solution of any salt nor­mally in boiler feed water at the concentration allowable in ordinary boiler practice (never ove~ 3% to 4%) is so near that ·of pure water that the variation is negligible. (a) Suspended Matter.-It is not at all a question of surface tension, but rather of viscosity of surface films. A, soap solu­tion has only a fraction of the surface tension of pure water, but it is sure to cause foaming. It is known from extensive experience on the part of many engineers that finely divided suspended matter will undoubtedly cause foaming. This is not due, as often wrongly stated, to the action of each small particle as a ''steam point.'' The collodial solid matter gives mechan­ical strength or tenacity to the liquid in the thin films formed over bubbles, and hence the steam bubbles rising to the surface retain their films and collect to produce foam. · (b) Alkalies.-Sodium carbonate is considered as the chief offender among salts thought to cause foaming. The most prob­able explanation of its action is, first, that it loosens old scale, and thus adds to the suspended matter; and second, that because of its alkalinity, it produces from the suspended matter a slightly gelatinous, glue-like, viscous solution, manifesting its effect in the surface-film, from which bubbles of steam will form foam. This opinion, then, considers that both the suspended matter and the means for putting it into viscous or emulsified form (which is done by salts like sodium carbonate) enter into the causing of steam boilers to foam. Very few boilers are ever free from suspended matter or dirt of some sort, and as the concen­tration of the soluble salts increases, the effect of the loose sludge becomes more marked. (c) Boiler Defects.-There are still other causes of foaming besides those already giYen. Boiler construction and boiler operation have a great deal to do with foaming. One boiler will give perfect service even up to abnormally high concentration of solids; and another boiler with the same feed water will foam in a day's si.:rvice. Some boilers are so badly made that they will "prime" with the slightest excuse. Some boilers are so well built that they will stand up and give service with al­most any kind of a water. Pressure has much to do with foam­ing. As a rule, the heavier the pressure carried, the greater will be the likelihood of foaming. The rate of operation and sudden variatioi:is in lo~d are to be considered in accounting for foam­ing. The problem is by no means simple, and each boiler and each water supply must be worked out individually. HOW TO PREVENT FOAMING The main thing which interests steam users is not the cause of foaming, but its prevention. In the first place, muddy water should be kept out of the boiler. (a) Filtering and Softening.-Filters are cheap in first cost, and practically cost nothing to operate. A water that will form sludge or scale in any amount has no business in a boiler. (b) Blowing Down.-Concentrations of dissolved salts can be held below reasonable limits by partially emptying the boiler fre­quently (or to use the engineer's term, "blowing down"). The rate and amount of emptying should be regulated not only by the type of boiler and rate of evaporation, but by the amount of dissolved material in the feed water. It is obvious, all other conditions being equal, that a water carrying a thousand parts per million dissolved solids should call for blowing down twice as often as one with £ve hundred parts per million. Following a practice of lo"wering the water level by blowing down a certain distance at arbitrary time intervals (for instance, so many ga.uges every so many hours) regardless of the nature of the water supply is utter foolishness on the part of any engineer. It is true that blowing down means loss of heat and consequent loss in fuel, and it should only be done when it is necessary. But this loss is not to be compared to the value of the protection -of engines and turbines from water coming over. Better a few dollars more on the coal bill than a wrecked engine or a stripped turbine. An engineer who uses his head will soon learn how much his boilers will stand with a certain water supply, and will stay within limits. ( c) Skimmers and Traps.-Since most of the suspended mat­ter that causes trouble is on the surface of the water, mechanical devices, such as skimmers and surface blow-off, are often suc­cessfully resorted to. As a final protection, adequate water traps in the steam line will save the engine or turbine, if foam­ing is sudden or unexpected, until the cause can be ·gotten at and removed. This is an engineer's problem, and no~ a matter for a chemist to decide, or even to discuss. (d) Anti-Foaming Gompounds.-Unless a boiler is to be washed out reasonably often it is best to rely on such mechanical means just referred to, ·rather than anti-foaming compounds. Most of sµch compounds are not only utterly without value, but are dangerous and destructive to boilers. There are some few legitimate compounds, made up mainly ·of castor oil and tannin, that have some virtue, but they are to be used only sparingly and then only with a chemist's advice and supervision. The practice of putting mineral oil, paraffin, ozokerite, and like sub­stances in a boiler is as dangerous as it is foolish. These things may stop foaming a little while but the man who uses them pays a big price in the way of corrosive oil scale and "bagged" flues or tubes. FOAMING FORMULAS Formulas for determining whether or not a certain water will foam have been devised, but these leave out so many factors that their author admits thp.t they have a very doubtful and extremely limited value. Hence they are not given here. In general it may be said that a clean water, free from scale and sludge-foaming constituents can be successfully handled in a decently constructed and sensibly operated boiler, unless it contains much alkali. In that case the heavier the constituents are, the more the boiler has to be blown down. Only a few waters in Texas, these ~ainly artesian, cannot be handled by a conimon-sense engineer, once he knows that he must have a clean, soft water. CHAP'l'ER TWO HOW TO STOP BOILER TROUBLES DUE TO BAD WATER The various causes of boiler trouble have been discussed in considerable detail. The question that will immediately follow is how t-0 stop boiler troubles or reduce them to their smallest possible extent. Like diseases, each trouble has a specific rem­edy, and it is impossible to lay down general methods anl di­rections that will absolutely fit all cases. Each individual prob­lem is a study in itself. It is poS.Sible, however, to outline the general remedies employed in getting rid of costly and trouble­some boiler ailments. 1. Boi1er Compounds In spite of modern advancement, the notion is still extremely prevalent that if the right kind of chemical or combination of chemicals can be found· and put into a steam boiler, all troubles will be at an end. All manner of substances have been used in steam boilers under the general head -0f Boiler Compounds. Some few of these have had a limited efficiency but in the main they have been ridiculous and dangerous remedies. In most cases, boilers have been ignorantly and indiscriminately ''dosed,'' a practice that cannot be too severely condemned. In general the theory of boiler compounds is wrong. The water should as a rule be treated outside of the boiler . . If a compound is used, it should be under the advice -0f a competent chemist, who has picked it out as the best available method of correcting the trouble, and then only in stated and correct amounts. COMMON CHEMICAL INGREDIENTS In the compounds used to prevent scaling, the most common ingredient is carbonate of soda or ''soda ash." By introduc­ing carbonate ion, the calcium and magnesium ions which cause Boiler Waters scaling are largely removed from solution as insoluble salts in the form of a loose sludge. Sodium phosphate has the same general action, but being more expensive than soda ash, it is rarely us~d. Barium salts, which may be used to remove sul­phates, are also quite costly, and are hence rarely found in boiler compounds. These c.ompri<>e about all of the efficient substances that are used.in boilers to convert scaling constituents to a loose and non-adhering sludge. SOME TYPICAL BOILER COMPOUNDS Mr. Dudley K. French has very kindly furnished the author with copies of numerous analyses of boiler compounds concerning which he says: "Some of the greatest fakes are sold on the basis of the story of some wonderful discovery not understood by science at large, but fallen upon by sheer luck. Most of them depend upon the credulity of the engineers for their success." ' For the benefit of those who have often in the past trusted blindly to boiler compounds without knowing what they contained or their chemical effects, a number of these analyses are given together with comments on their value or utter lack of value is remedies for boiler trouble. All of these compounds have been on the open market during the past ten or fifteen years, and the analyses are made from actual samples taken from the marketed products. No names or brands are quoted. 1. Dark colored liquid, neutral reaction. Water ...... . ....... . .................. 98.35 Vegetable Matter . . . . . . . . . . . . . . . . . . . . . . . . 1.28 Ash ....... .. ..... . ................... · .87 The vegetable matter is too small in amount, even if it were tannin, to be of any value. No comment is necessary on the water. 2. Thin, orange-yellow, very limpid liquid. Water .................................. 90.73 Bichromate of Potash..... . ............... 8.89 Undetermined Matter . . . . . . . . . •. . . . . . . . . . . .3 8 The bichromate of potash is intended to prevent corrosion, but is very expensive. No comment is necessary on the water. To allow a corrosive water to go into a boiler, some chemist has said, is like deliberately poisoning a man and then giving him an antidote. 3. Very thin, reddish-brown solution with slight alkaline reaction. Water . ....... . . ............ ... .. .. ..... 90.37 Carbonate of Soda...........•........... · 7. 0 5 Chloride of Soda......................... .47 Organic Matter ................•.. . . . . . . . . 1.49 Undetermined Matter . . . . . . . . . . . . . . . . . . . . . .62 The organic matter seems to be present largely to give a mys­terious color to the compound, as it contained only traces of tannin. Why buy a dirty and very dilute solution of soda ash, when the pure salt is on the market? 4. Light-yellow, very limpid liquid. Carbonate of Soda........................ 2.65 Caustic Soda . . . . . . . . . . . . . . . . . . . • . . . . . . . . 4.80 Chloride of Soda......................... .35 Water .................................. 91.15 Chromate of Potash...................... .90 Loss, etc. . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . . .15 Caustic soda should be kept out of boilers. Another case of buy~ ing water to get a little alkali. 5. Reddish-brown, very limpid liquid. Caustic Soda . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.57 Carbonate of Soda........................ 2.69 Vegetable Matter (Tannin Extract). . . . . . . . . . 2.89 Water ................................. 86.85 Caustic soda undoubtedly leads to embrittlement, and its indis­criminate use is dangerous. More high-priced water. 6. Very limpid, lemon-yellow, transparent solution. Water ........................... . ...... 97.21 Barium Hydroxide . . . . . . . . . . . . . . . . . . . . . . . .1.61 Barium Chloride . . . . . . . . . . . . .• . . . . . . . . . . . .16 Loss and undetermined Matter...... . . . . . . . .59 Barium salts should be used only with certain kinds of waters, mainly those carrying large amounts of the sulphates of calcium and magnesium. In ·other kinds of waters they are useless. Barium hydroxide and barium carbonate are on the market in solid form, and water is usually cheaper at home than .shipped a thousand miles or so. 7. Exceedingly thin dark-brown liquid, neutral reaction. Water .................................. 86'. 13 Carbonate of Soda........................ 2.29 Chloride of Soda. . . . . . . . . . . . . . . . . . . . . . . . . . 6 8 Sulphate of Soda......................... .16 Organic Matter . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7 5 A vegetable extract in very dilute solution, with soda ash to take care of the acidity. The use of tannin has already been discussed. No comment is necessary on the water. 8. Very viscous dirty-gray mixture having a strong odor sug­ gestive of sea-weed or marine growth. Water ........ ..... .............. ....... 89.12 Carbonate of Soda........................ 3.97 Caustic Soda . . . . . . . . . . . . . . . . . . . . . . . . . . . .60 Salt ...... ............................ : . Trace Organic Matter . . . . . . . . . . . . . . . . . . . . . . . . . . 6.31 Amount of organic matter too small even if it were of any value. The danger of such organic compounds has been pointe'tl out. Other­wise, a very dilute solution of' soda ash. No comment is necessary on the water. 9. White, milk-like emulsion, containing considerable quantity of oil in globules, whiieh separate out on standing. Water ....... . ... .. ......•.......•..... .' 86.16 Castor Oil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. 92 Free Acid (principally sulphuric)........... 2.40 Sulphates (in organic combinations)........ .75 Ash ....... ~ ... ..... . . . . . ...·.......... . .18 Loss and Undetermined Matter. . . . . . . . . . . . . .59 Castor oil compounds that are made right may be used under some conditions, but with extreme care. This material contains a free mineral acid,. which will render it highly dangerous. 10. Solid material. Calcium Carbonate 6.04 Tri-Calcium Phosphate . . . . . . . . . . . . . . . . . . . 8 3. 71 Iron and Alumina . . . . . . . . . . . . . . . . . . . . . . . . Trace Magnesia . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . Trace Organic Matter . . . . . . . . . . . . . . . . . . . . . . . . . . 7.25 Water .. .. . .. . .. . . . . . .. .. ... . .. .. .. . .. .. 2.60 Loss and Undetermined Matter. . . . . . . . . . . . . .40 This compound ls worthless, being mainly a filler with a small amount of organic matter. Calcium carbonate is the same compo­sition as scale. The tri-calcium phosphate is also insoluble, and hence valueless. As well put a load of dirty rocks in the boiler for all the good it would do. A pure fake. 11. Solid Material. Tri-Calcium Phosphate ........... ........ 81.70 Iron Phosphate . . . . . . . . . . . . . . . . . . . . . . . . . . 2.73 Tri-Sodium Phosphate . . . . . . . . . . . . . . . . . . . . 13.39 Moisture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 .18 The sodium phosphate is the only effective constituent. It can be bought pure, instead of with 8 5 per cent filler. The remarks under 10 apply here. 12. Dirty-white stick, easily powdered. Moisture ... •. • ........ .. ...... ..... ... . . 9.52 Potassium Chromate ........ , . . . . . . . . . . . . .82 Soap Stone . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89.66 "An absolute fake. Any other kind of powdered stone would have been as useless. 13. Solid Material. Tri-Calcium Phosphate 91.32 Moisture and insoluble matter........ ..... 8.62 A nice load of dirt from the garden would have served as well. 14. Reddish-brown powder. Moisture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. 3 7 Soluble Organic Matter . . . . . . . . . . . . . . . . . • . 2.20 Phosphate of Soda........................ 1.12 Insoluble Mineral Matter...............•.. 89.13 Insoluble Organic Matter.......... ~ ....... 2.18 More dirt. 15. Coarse, yellowish-white, granular powder. Soda Ash .................•.•............72.66 Powdered Soap .. .. .......... .. .....•.... 27.34 The soda ash is all right in its place, which is generally outside the boiler and in a softening tank. What engineer is fool enough to put soap deliberately into a steam boiler? ·16. Salmon-colored Substance. Moisture ....................•.......... 10.66 Insoluble and coloring matter. . . . . . . . . . . . . . . .53 Carbonate of Soda ............•........... 48.83 Caustic Soda . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.85 Chloride of Soda ............•............ 38.14 Salt does no good and may do harm. Why not buy a good grade of soda ash, at least 95% pure? Of course it may not be salmon­colored, but at a pinch, it might be dyed. 17. Pinkish powder, soluble in water Fluoride of Soda.......................... 81.30 Chloride of Soda.... ; . . . . . . . . . . . . . . . . . . . . . 3.62 Moisture ... .. . ...........•... . .......... 13.56 Undetermined Matter (principally coloring matter) . . . . . . . . . . . . . . . . . • . . . . . . . . . . . . 1.52 The main constituent is four or five times as expensive as soda­ash, and no more efficient. Aside from the expense, the compound is all right as far as compounds go. 18. Solid Material. Moisture ................•......·. • . . . . . . 7.82 Sodium Chloride ......•.•••'. .••.•.•••••.•• 74.76 Sodium Carbonate .............•...•..••• 15.90 Undetermined Coloring Matter . . . . . . . . • . . . • 1.52 Why buy useless salt to get 16 % soda-ash. 19. Solid Material. Carbonate of Soda.......•................ 2.65 Chloride of Soda. . . . . • . . . . • . . . . . • . • . • . . • . 9.11 Oil ....•............................... 16.95 Organic Matter (other than oil)............ 1.85 Moisture . . . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . 11.08 . Loss and undetermined matter. . . . . . • . . . . . . 2.17 Loading soda-ash down with a cheap filler is bad e:n,ough, but t~ put oil with it is very much worse. Oil is dangerous in a boiler and is separated from return water as much as possible. This compound gets out. of the ridiculous into the dangerous class. 2.0. Black Mineral Oil. Mineral Oil ...•.....................•... 98.97 Oil of Mirbane....... .... . .. .. . . . .. . . . . . . 1.03 Gravity .................,. . . . 28.2° Be Flash Point ...... . .......... 410.0° F Burning Point ............... 470.0° F Viscosity at 212° F ... , ....... 104 (Tagliabue) As a "scale resolvent" this stuff is a failure, but as a source of burned and bagged tubes and pitting, it is a grand success. WHEN TO USE COMPOUNDS As to when a boiler 'compound may be used, or when a water admits of treatment within the boiler, it may be said that whenJ ever a water carries less scaling constituents than the average limit of efficiency of a lime-soda process softening plant (from 50 to 75 parts per million in terms of scale formed) it is ad­ missable to use treatment within the boiler. If a water carries as much. as 100 parts per million scaling constituents, it. is best to. treat outside the boiler. OBJECTIONS TO COMPOUNDS (a) Sludge.-The idea of treating a water carrying twQ or . . ' ' three huncired parts per million (ten to twen_ty grain13 per gallon) of incrusting solid~, or even, more than these amounts, with soda ~50 University of Texas Bulletin ash in the boiler is ridiculous. Suppose a water supply carries ·two hundred parts per million of hardness ·in terms of calcium carbonate, and a boiler uses two thousand gallons of water per ·day. That means about three pounds of sludge formed in the water every day. Consider how often it will be necessary, not only to blow down but actually to shut down -the boiler and wash it out so get rid of this mass of sludge. (b) Foaming.-Not only this, but the boiler is by this very treatment full of suspended matter in the presence of an alkali, and thus i~ in the very best and most favorable condition for continuous foaming. (c) Lack of_ Definite Amount.-In most cases the soda ash or other chemical is not added in any definite amount but is simply dumped in according to guess work. If the amount is small, only part of the scaling solids will come out of solution as ·sludge, and the remainder will form scale as usual, and likely include some of the sludge, and the boiler is little better off, than before. If the amount is too much, foaming is added to other troubles. ( d) Bad Boiler Practice.-The theory of this kind of inside treatment is all wrong. It is only a make-shift where the amount of incrusting solids is too small to make ordinary softening methods practicable. There is . absolutely no sense in Jnaking­a sludge tank out of a steam boiler. It is not a chemical re­action tank. It is an apparatus for making the most, the best and the driest steam with the least possible fuel. Such an ap­paratus deserves a clean and properly softened water supply. Modern boiler practice and soda ash and similar boiler com­pounds can never be made to agree, wh€re the raw water carries any c~:msiderable amount of scaling constituents. Treat the wa­ter -outside the boiler. The worst feature of these compounds is that the engineer in most cases has :rio notion· of their com­position and their effect. They are not sold on analysis, but are recommended generally by an ignorant salesman, who him­self is neither chemist nor engineer, and who is consequently privileged to make all manner of guarantees and promises that -cannot be kept. Boiler compounds to prevent scaling should never be used until their composition is known and until a -chemist advises as to the correct amount on the basis of an­alyses of the water. There should always be continuous con­·trol. Only if the amount of incrusting solids is very low, is there any reason or excuse for boiler compounds of this class. MECHANICALLY ACTING COMPOUNDS In addition to those compounds just diEKiussed whose action is purely chemical, there is another general class of compounds ·that are supposed to act mechanically on the theory that they will either coat the walls of the boiler so that scale will not adhere, or else that they will permeate the scaling solids as they settle and keep them loose. Most of these substances are veg­etable in their composition, such as sea-weed, potato meal, bran and the like, all of them more or less harmful and dangerous. Organic materials of all kinds combine with the sludge, bake into the scale, and .burn, causing corrosion. Graphite is thus used, and is especially objectionable. Quite recently a graphite compound was fed into on~ of the boilers of a local power plant. On cleaning the boiler it was found that instead of loosening the scale, the compound had actually ser".ed as a binder. It occurred in alternating layers, which recorded periods of firing ·and cooling the boiler. Worse than this, on removing the scaler severe pitting was found to have taken pliJ,ce under the scale. Suppose the inside of a boiler could be coated smoothly with graphite. The only practicable way to d-0 this seems to be to mix the graphite with oil and apply it with a brush. Then it would last only an hour or'so and do more damage and waste more heat than a thick scale, The idea of dumping loose graphite into a boiler and expecting it to spread uniformly over the boiler surface and cause the scale seeking a resting place to slide off to the mud valves is like wai~ing for ~ miracle. LEGITIMATE COMPOUNDS · There are a few makers of boiler compounds who are going into the matter scientifically as well as legitimately. In the fust place, an honest manufacturer will not recommend a treatment that is not the correct one for the particular case in hand. Then . the a.mount to be used will be specified and the meth-0d of con­ trolling the dosage will be given. The modern compound busi­ ness is taking account of real scientific principles in treating boiler waters. The object in an anti-scaling compound is to alter the small amount of scaling solids (since only waters light in incrusting constituents should be treated anywhere but outside the boiler) into a loose, non-crystalline, non-adhering sludge. It has been claimed by reliable chemists that certain organic com­ pounds, such as the sugars derived from tannins i;n amounts so . small as to be safe will aid very much the action of such sub­ stances as soda ash. Further, alkaline tannates are regarded by chemists who have conducted extensive experiments NO·,, t t 1 ' The followil!-g description of the Sorge-Cochrane hot-process water softener is reproduced through the courtesy of the Harrison Safety Boner Works: "This apparatus is designed to take advantage of the fact that chemical reactions are more rapid and complete in hot water than in cold water, and that the resulting precipitates are coarS"er and settle out mor~ rapidly. The raw water is first heated by means of exhaust steam from pumps, engines or other apparatus to ap­proximat-ely 210° F., or higher, dependi~g upon the back pressure carried. It is then mingled with the softening reagent ang drops into a reaction and sedimentation tank, where a large part, and in many cases practically all the sludge settles out and is removed from th~ conical bottom of the tank . by periodic blowing off. The chemicals ordinarily employed are hydrated lime and soda ash. The chemicals are mixed with water in a tank at .the ground le~el, and the mixture is fed into the chemical pump in accurate pro­portion to the ti.ow of raw water into the softener by ' means of a differential pre·ssure device similar Jn principle to a ti.ow meter. The treatment is controlled by varying the amount of reagents per gallon of mixture. Due to the more rapid and complete reaction obtained with hot water, there is little or no after-precipitation of sludge in piping, pumps or boilers, and the latter can ordinarily be operated satisfactorily at moderate or even high overloads without requiring the use ·of iiiters. Wihere filters are required, as with certain kinds of water, and high overloads, or rapid variations of load, or with boilers of certain design, they are installed externally to the softener, and it is recommended that they be installed only after the need has actually arisen, and after taking into account first cost, cost of operation, and the benefits to be•gained." and filtering easier. The cost. iS no greater than that of a cold process, and if properly handled is even less. The hot water softener takes the place of a preheater. Another advantage is that less sludge gets into the boiler from a water that has been chemically tre11ted }lot. A wate~ softened cold still carries some small ainourits of incrusting solids, due to incomplete reactions. There is always 'present a little excess of soda a.Sh, and the re­actions b~irlg br_qu$ht to completion in the boiler, there will sep­.arate. cmf 11. .carbonate sludge. This sludge begins to separate . in ..th~ pr~$~,t~r, clogging this apparatus, continues to deposit in th~ie4J~i>iii~s, and produces sludge in tne boiler. This adds to..th~2fMmi11idaJiger. 'To avoid this, in most cases, filters are used. betw.een the preheater and the boiler, a thing which is not necessary,_ at all in the hot process, since one filtration suffices. FILTER WATER OUTL.ET THE VATER CONSTANT TEMPERATURE WATER SOFTENING SYSTEM CaO PRE5El4T AS CaSO+ BEFORE TREATMEHT .56fs.l'Et6.AL 5 Cao PRESSltT AS CaCOJ AFTER 5 HOURS TREATMEHT WITH lfafO, AT 50°f. 1.7 6RS. PER GAL. . CaO.PRESE/fT AS Ca~03 AFTER ~/\Iii. TREATME!iT WITH Na,C03 AT ZOO TO Zl0°F. ' ) I.I 0-R.S. PER 6'..41-. . /1~0 PRE.Sf/ff AS .M,tso+ BEFORE TREATME/U . ·'. AFTER s HOURS TREATMEAT '.WITH Ca(OH! AT S0°F. 1.8 &RS. PER &AL. }\~0 PRESfHT AS 11g(oH)~ AFTER s l'\llt TREATMfHT WITH Ca(o~,, AT zoo TO ZI 0° F. -.4 GR.5. PER GAL. .... TABLE OF E.QVIVALE.NTS l~r. Ca.0 IS EQUIVALENT TO 179 ~r. CaC03 ltr 11~0 IS EQUIVALENT TO l.Hgr: M~(OH}z '$' c&o " " " Z.+J gr CaSOary that it should be maintained in alkaline condition. In general, coagulants are not a safe proposition, and there must be close chemical control in their use. It is not necessary to filter water before treatment by the lime-soda process. The chemicals in the reaction tank will serve to some extent as coagulant.;;, and the remaining suspended mat­ter will be removed by the filter, as the treated water flows from the softener. Filters, however, must be used in conmction with the Permutit system, before · water that has even a ' very small amount of mud or vegetable matter is fed into the softener. If muddy water is used, the grains of the zeolite are coated, and their ·chemical action stops. The Permutit apparatus is not a filter at all, but a softener, and requires filtered water. As has been stated, there is no sludge from a Permutit softener, and only the •filter al}ead of the apparatus is necessary. SJ,UDGE REMOVAL Praetically all water softeners are too small, and do not allow sufficient time or space for complete reaction and settling. For that reason the softened water from such softeners carries some sludge. This sludge must be removed by filtration. In fact most water softening systems are constructed to work in co1me·ction with filters, and their manufacturers will not guar­antee their successful operation without filters. Most of the foaming troubles from softened water are caused by the pres­ence of sludg-e. No process of water softening is satisfactory or, efficient unless the amount of suspended matter is reduced to a minimum. This is most surely and safely accomplished by filtration. In case there is any marked separation of sludge in the preheater, which is often the case where a cold lime-soda process is employed,· a small, well-insulated filter should be m­stalled between the preheater and the boiler. TYPE OF FILTER Engineers are very generally agreed that the most satisfactory type of filter is the sand or quartz filter with a down flow of the water. This is the most common type, and its efficiency is a matter of record. Excelsior or hay as a filtering medium is not so safe or satisfactory. Filters are not expensive either in original cost or maintenance cost, and only the type which is of proven efficiency should be used. Summing up: Muddy water, evel.l if it is soft must be fil­ tered. Coagulants are not entirely safe and should be avoided unless water cannot be cleared without them. Hard waters must be softened, and the sludge that remains in suspension must be removed by a filter. The •filter must be of a type that is of proven efficiency, preferably the sand filter with down-flow. It is false economy to install a softener and leave ouJ the filter. Filters are necessary in handling either muddy water or treated water that carries suspended sludge. DANGER OF INCORRECT FILTRATION AND COAGULANTS A successful maker of water softeners has given these two instances in proof of the necessity for filtration. and the danger in the indiscriminate use of coagulants to help out a poor filter. "A central station, lighting, power and heating plant, capacity of softener 15,000 gallons an hour, cold continuous, type of filter: up-flow through excelsior. Results: trouble from sludge; filled up the flow meters, the water regulating valves, and the superheaters; foaming occurred; blading was stripped out of a 5000 K. W. turbine; cost of reblading $5000; complained to manufacturers who recom·. mended the use of a coagulant costing $275 per annum; sulphuric acid released by coagulant ruined four 350 h. p. water tub·e boilers; boiler inspectors condemned them; company installed new boilers at a cost of $16,000; five other boilers so seriously damaged that they cannot be operated two years more. Lesson: Always filter the. water from a softener, and never use a coagulant in a water used for boiler feed. · "A large chemical plant, capacity of softener 16,000 gallons ~n hour, cold continuous, type of filter: up-flow through excelsior. In three years foaming, burning tubes, stopping up of auxiliaries has. caused an expense of $15,000. Company is now spending $12,500 · to install filters to remove sludge. Steel plate is now very high. Softener originally cost $6,500. Was not the failure to install filters originally a crime?" While we are not prepared to go quite as far in condemning the use of coagulants absolutely, it is imperative that the danger from their use be recognized, and continual care taken to . keep the water alkaline. . CHAPTER THREE COST, CONSTRUCTION AND OPERATION OF WATER SOFTENERS Having been convinced of the cost and danger of bad water, the manufacturer naturally asks these two questions: "What will it cost to install and operate a water softener?''; and ''Will a water softener show a reasonable profit on the investment?" In this chapter, the first question will be discussed, and all information given that it is possible to give in a general way. The next chapter. will take up results. 1. Cost of Standard Patented Water Softeners No fair estimate can be made on the cost of installing a water softening system until full information covering the individual needs. and requirements of each plant is obtained. Hence in this discussion only the most general figures are given in order that steam power users may gain some sort of an idea of the cost of a water softening plant. It should be kept in mind that, owing to war conditions, structural materials are very high, and any estimate must take into account these ab~ormal conditions. VARIATION OF COST WITH SIZE It is especially :true in water softening plants that the smaller the installation, the higher the cost. With very small i~1stalla­tions, most water softener manufacturers say frankly that it does not pay to install a patented softener. The increased effi­' ciency over a simple, home-constructed, tank softener does not counterbala11ce the higher cost of installation in the case of the smaller. size softeners. For this reason, planS and estimates for small softeners that can be constructed on the ground are given in a later section. Cost Quotations: Following are 'a few quotations given by leading water softening manufacturers on plants of various ca­ pacities. (Several firms preferred not to make any defip.ite statements, due to abnormal war prices of materials.) Capacity per Hour Total Cost Cost per 1000 gallons 500 gallons $ 1,200 to $2,000 $2,400 to $4,000 1,000 gallons 1,500 to 2,500 1,500 to 2,500 5,000 gallons 3,250 to 6,500 650 to i,300 8,000 galions 6,000 750 (one estimate) 25,000 gallons 10,500 420 (one estimate) From the above figures it is evident that under present con­ditions for any plant using less than 1000 gallons per hour, the cost of a water softening system as manufactured by~ power , plant supply companies seems to be prohibitive. QUOTATIONS FOR EACH INDIVIDUAL CASE The reader is advised, however, not to take these figures as anything more than general approximations. Before installing any kind of system, it is advisable to give the full facts of the plant and water supply to several reliable companies for bids and estimates. An honest manufacturer of water softeners will not sell a softener that does not pay a reasonable return on the investment. :Most manufacturers make this as a definite state­ment: ''Our softeners are not for sale until we can prove to you that th~y will pay.'' It cost~ nothing to investigate the relative cost of various makers. MUNICIPAL PLANTS As the size of the installation increases, the unit cost comes down, until for very large installations and municipal plants, the unit cost is as low as one-tenth the figure given for a small plant. . Where the am01~nt of hardness is sufficiently low to make the cost of chemicals a comparatively small item per unit amount of water, large ·softening plants handling the entire water supply of a city are feasible propositions. A notabl~ example of a small ' municipal plant is that of Owensboro, Ky., where something over 1,000,000 gallons of water are treated daily. The tot.al cost of the entire installation was less than $30,000. At :Mc­Keesport; Pa., 10,000,000 gallons daily are treated for the use of the city, and the relative cost of this plant is still less than the one of one-tenth capacity. In such cases, the cost of chem­icals is the important consideration, the original installation being only a small part of the total cost on a long time basis. Section 1n Ghem1Col Tonk.s Fig. 14. Booth water .softener. Cold, continuous, lime-soda process. Municipal plant, Owensboro, Ky. (Plan.) 2. Cost of Hone-Constructed Softening Plants. LIMITED TO SMAJ,L INTERMITTENT PLANTS For water softening plants of any considerable size, home eon­struction is neither economical in first cost or in up-keep. In­termittent systems are more expensive both to build and to operate than continuous systems. Where space is an item, as in cities, this also counts against the intermittent method, since a continuous softener occupies much less ground space per unit capacity. It is not possible for any one not an expert in the construction of water softenin'g plants to devise a chemical pro­portioner. The various· successful types of chemical feeding devices are completely covered by patents. This is after all the key to a continuous process, a successful and simple chemical proportioner. Heating the contents of an intermittent softener is not convenient or profitable. The hot process is used only by continuous plants, and these can not be successfully con­structed except by an expert water softening engineer. Hence home-construction is confined to small capacity inter­mittent plants. Admitting that such plants are less convenient, less flexible, and more expensive to operate, it seems that under present conditi.ons this is about the only way in which water softening can be done by small power plants. It is also often true that an engineer prefers to try out on a small scale with comparatively inexpensive installation the whole problem of water softening, with a view to purchasing at some later time a water softening plant of standard type. VALUE AND LIMITATIONS OF HOME-CONSTRUCTED PLANTS For these reasons, some suggestions are given as to the con­struction of a small, intermittent, cold process softener with concrete reaction and storage tanks. It is not claimed that a softener of this kind will . give anything like the satisfaction that will be given by a standard make, continuous softener. But from practical experience with a home-constr11cted softener of this general type, it is known absolutely that reasonable results · can be obtained. Scaling can be reduced to a neglible amount, corrosion entirely stopped, and savings in fuel and labor effected that will pay large returns on the investment. GENERAL PLAN OF HOME-CONSTRUCTED SOFTENER The softener proper consists of a tank of cypress or rein­forced concrete. Flush with the bottom of the tank there opens a pipe with a valve, which takes off the sludge when the tank is cleaned. Another pipe projects through the bottom eight or ten inches, and is connected through a valve with a standard quartz or sand filter. The feed pipe for the raw ~ater comes up on the outside of the tank. The stirring device here consists of a branching perforated pipe for compressed air or steam. Other devices, such as paddles driven by a water-wheel using the inflow of the water, may be employed, but the use of com­pressed air has been found to be so cheap, convenient and sat­isfactory, that no other devices have been studied. .For a small plant, sound, heavy barrels make satisfactory tankS in which to prepare the solutions of the chemicals to be used. From the filter the water runs into a storage tank of convenient capacity. With all intermittent systems it is necessary to have a storage tank of some sort. It is possible to operate with two or more tanks, drawing the water from one while the softening and set­tling process goes on in another. A storage tank is cheaper and does away with the4d~nger of having to use raw water in case of a: temporary shut down of the softener. CONSTRUCl'ION DETAILS OF A REINFORCED CONCRETE TANK SOFl'ENER (a) Size and Capacity of Reaction Tank.-These plans call for a tank of 10,000 gallons holding capacity, which is a con­venient unit size for an intermittent softener. The tank should have at the base, an inside diameter of 15 feet if circular, or should measure 13 feet each way, if rectangular. The inside depth in either case should be 9 feet. This will allow about one foot margin above the surface of the water if '10,000 gallons of water are run into the tank. This permits adding as much as 1000 gallons more in order to correct a mistake in treatment as determined by tests on the treated water. (A depth gauge worked out in 100 gallon units is very convenient.) The walls should not be less than eight inches thick and the bottom froµi six to eight inches thick. (b) Reinforcing.-The same reinforcing may be used .for either circular or rectangular tanks. The reinforcing should consist of one-half inch round steel rods of a length sufficient to reach from top to bottom of the tank set vertically five inches from center to center with at least three-fourth inch clearance from the inside surface. These vertical rods should be bent ten inches from the bottom at an angle of 45 degrees, the slanting portion pointing inward. Instead of one-half inch round rods five inches from center to center, three-fourth inch round rods eleven inches from center to center may be used for the ver-· tical reinforcing. The horizontal reinforcing should be of one­fourth inch round steel rods placed 18 inches from center to center, and fastened at each intersection. The bottom can best be reinforced by the use of wire mesh of the size usually em­ployed in this kind of work. The wire should be fastened to the veriical reinforcing rods, and should dip from points four feet from each wall to within one inch of the bottom of the · floor slab. It is assumed that this tank is to rest on the ground. (c) Concrete Mix and Waterproofing.-The concrete should be a 1 :2 :4 mix, at a mushy consistency, 11sing only enough water to handle the mixture easily. It should require some tamping and spading to flush the water to the surface when placed in the forms. When inside forms are removed the inside of the tank should be well grouted and surfaced like a sidewalk. To this surfacing mixture, should be added 10 per cent by weight (of the cement) of well slaked lime or some standard water­proofing material, in amount recommended by the manufacturer. If the inside is not to be surfaced, it will be necessary to add this water proofing material to the concrete of tank walls and base as it is made up. If possible, the whole tank should be poured at one time. In case there is an intermission of as much as a day in pouring different parts of the tank, all joint sur­faces should be thoroughly cleaned with a wire brush and plas­tered with neat cement just before pouring additional portions. Otherwise, there will surely be leaks at the joints. (d) Storage Tank.-This is most conveniently set in the ground so that there may be a gravity flow to it from the soft­ening tank through the filter. The details of construction are the same here as for the reaction tank, with the exception that the vertical reinforcing should be next to the outside rather than the inside wall. Its capacity should be as much as twice that of the softening tank so as to provide for all emergencies. (e) Cost of Construction of Co~crete Softening and Storage Tanks. The materials, reinforcing steel and wire, and the labor for making a 10,000 gallon concrete tank of the above size will cost, at present prices, approximately $275.00. For a storage tank that is set in the ground, the cost of excavation will have to be added. ~If the ·excavation is 'inainly in earth, the cost will be from $30 to $40, if in rock about twice this amount. In the case of rock, however, the extra cost of excavation would be partly counterbalanced by the elimination of the reinforcing in . the base and the outside forms. A plant consisting of one 10,000 gallon softening tank above the ground and one 10,000 gallon storage tank set in the ground, would cost about $650 to construct. The pipes, valves and fittings will amount to approximately · $70, and $30 should care for the labor of installation. A total of $750 should easily' cover a complete softening and storage plant of 30,000 gallons daily capacity. (f) Pipe and Fittings for the Softener. Following is a list of the fittings necessary for the softening tank shown in Figure. 2 ells 4". 1 nipple 4x18". 1 nipple 4x12". 2 pc. 4" black pipe 4'. 2 globe valves 4". 1 pc. 2" black pipe 10'. 1 tee 2". 1 pc. 2" black pipe 2'. 3 globe valv-es 2". 1 bushing 2x %, "· 1 pc. %, " black pipe 3'. 1 hose bib %, ". 1 pc. %, " black pipe 14'. 2 angle valves %, " 2 close nipples %, ". 3 ells %, "· 1 pc. %,"black pipe 2' 1 tee %, "· 1 pc. %, " black pipe 3'. 2 pc. 2" black pipe 2'. 4 lock nuts 2". 2 nipples 2x12". 1 pc. %, " black pipe 1 O '. 2 pc. %, " black pipe 7'. 2 ells %, "· 1 cross %, ". 1 t-ee %, "· 1 close nipple %, ". 4 bushings %, " to % "· 4 pc. %"black pipe 6' (each perforated at 1' inlervals with 1-16" holes) . 4 %"plugs. 3. Cost of Chemicals The amount of lime and soda ash used by any type of softener of standard design is approximately the same for a given water. Experience has shown that in a home-constructed intermittent softener of the type given above, the efficiency of the lime is somewhat less, increasing the cost of this chemical in some cases as high as 10 per cent to 20 per cent. The amount of salt nee-. essary to regenerate a Permutit softener is from three and one­half to four times the amount of soda ash necessary to treat the same water. The cost of lime for a lime-Permutit system is of course the same as for a lime-soda system. CALCULATION OF AMOUNT OF CHEMICALS NEEDED FROM ANALYSIS OF THE WATER The amount of lime and soda necessary to soften a given water can be readily calculated from an analysis of the water. The following formulas of Stabler have been found very convenient and accurate. Lime required=.00931 Fe+.0288 Al+.0214 Mg+.258 H+.00~6 HC03+.0118 C02 (gas) Soda ash required=.0167 lfe+.0515 Al+.0232 Ca+ . . 0382 Mg+.462 H-.0155 C03-.00763 HC03 Given the following analysis, and employing the Stabler form­ulas. Let us calculate the amount of lime and soda ash neces­sary to soften this water. (These formulas assume the lime to be lump lime or calcium oxide and of 90 per cent. efficiency; the soda ash to be 95 per cent efficiency.) Parts per Million Calcium .... , ... , ........ , ................ 138 Magnesium . . . . . . . . . . . . . . . . . . . . . • . . . . . . . . . 29 Sodium . . . . . . . . . . . . . ... . . . ... . . . . . . . . . . . . • 69 Carbonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Bicarbonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 Sulphates ................. · .............. 198 Chlorides .. : . . . . . . . . . . . . . . . . . . . . . . . . . . • . . • 1O6 Lime required: .0214X 29 (Magnesium) = .62 .00426X290 (Bicarbonates) =1.24 1.86 lbs. 90%1ump lime per 1000 gals. (To convert lump lime to hydrated lime-Cao to Ca(OH)., multiply by 1.32. 1.86Xl.32=2.46 pounds 90 per cerit hydrated lime per 1000 gallons.) With lump lime at one-half cent per pound the cost per thou­.sand gallons for the lime would be 0.93 cents. Soda ash required: .0232X138 (Calcium) =3.20 .0382 X 29 (Magnesium) =1.11 4.31 .. 00763 X29 0 (Bicarbonates)=2.21 2 .1 O pounds 9 5 % soda-ash per thou­sand gallons. With soda ash at three cents per pound the cost per thousand gallons for this chemical would be 6.3 cents. The total cost £or both chemicals would be 7.23 cents per thousand gallons. Since the amount of dissolved Rolids in practically any water supply varies according to the season, it would be necessary to allow for a variation in cost of chemicals from 6 cents to 9 cents -per thousand gallons at the p11ices given. When an analysis is given in hypothetical combinations, the .amount of chemicals necessary to change or remove each scaling salt may be calculated from the analysis. Different methods of -calculation give varying results, but these are well within the variation limits of the seasonal changes. From the above analysis calculated in combined form, take all the scaling ingredients as follows: Calcium Sulphate . . . . . . . . . . . . . . . . . . . . . . . . . 119 Calcium Bicarbonate ................•...... 385 Magnesium Sulphate ... .. ............ . ..... 143 Using the fact.ors from the table given below, these results are obtained. '(The amounts of each salt being in parts per million, it will be necessary to convert to pounds per thousand gallons by mul­tiplying each result by .00833. ) Lime required : .385 (Calcium bicarbonate) X .346 X. 00833=1.ll 143 (Mag;nesium sulphate) X .46.6X.00833= .57 1.6 8 pounds . or pure lump lime per 1000 gallons. If only 90 per cent efficient, which is the usual strength cal­culated for . water softening plants, the real amount needed would be 1.68 divided by .90=1.86 pounds of 90 per cent lump lime per 1000 gallons. This value checks with the amount cal­culated above from the analysis by ions. Soda ash required : (Magnesium sulphate requires both lime and soda ash and calcium sulphate soda ash alone.) 143 (Magnesium s.ulphate) X:881X.'00833=1.05 119 (Calcium Suiphate) X .779 X .00833= .77 1.82 pounds per 1000 gallons of pure so­dium carbonate. If 95 percent efficient (the usual value for such calculations) the real amount would be 1.82 divided by .95=1.92 pounds of 95 per cent soda ash per 1000 gallons. This value checks fairly closely with the one obtained by the first method. TABLE OF FACTORS FOR USE WITH HYPOTHETICAL COMBI­NATIONS (Multiply the amount of the salt by the .factor to obtain the chemical necessary to, remove this salt as a scaling constituent. If the analysis is in parts per million, multiply the result by .00833 to obtain pounds per thousand gallons. If it is in grains per U. S. gallon, ~ivide the result by 7 to get pounds per thou­sand gallons.) Lime Factors Lump Lime Hydrated Lime CaO Ca(OH), Sodium Carbonate 0.529 0.699 Magnesium Chloride .. 0.589 0.778 Magnesium Sulphate .. 0.466 0.616 Magnesium Bicarbonate 0. 7 6 7 1.014 Magnesium Carbonate 1.330 1.757 Calcium Bicarbonate.. 0.346 0.457 Calcium Carbonate .. . 0.560 0.740 Soda Ash Factors: Soda Ash (Na,CO,) Calcium Chloride . . . . . . . . . . . . . . . 0. 9 5 5 ' Magnesium Chloride . . . . . . . . . . . . 1.113 Magnesium Sulphate . . . . . . . . . . . . 0.881 Calcium Sulphate . . . . . . . . . . . . . . . 0.779 VARIATIONS IN COST OF CHEMICALS The cost of chemicals varies according to the amount ~f hard­ness in the water, and the nature of the hardness. Two or three tenths of a cent per th,ousand gallons will cover the cost for a water that is so low in incrusting solids that it is a matter of doubt whether or not it should have outside treatment. As much as thirty cents per thousand gallons will be required to buy the chemicals for softening water that is so heavy in scaling constituents that it is hardly fit to use even when softened, due to the large amount of alkalies left in the water. It is, of course, true that the more expensive the treatment of a water, the more expensive it is to operate a power plant without treating the water. So where it is impossible to change the source of water, it is not a question of high cost of chemicals as much as it is the much higher cost of repairs, cleaning of boilers, and extra fuel used to heat through scale. HOT PROCESS SOFTENER AS A PREHEATER In calculating costs of chemicals on various types of softeners, it is necessary to take into consideration the fact that one kind of softener, namely the hot, continuous, lime-soda process com­bines in one the chemical treatment and the preheating of the boiler feed water. Thus the original cost of a preheater should be deducted from the. cost of installing such 3: softener, alld the lessened fuel cost due to a preheater should be charged off the cost of chemicals. With a well insulated hot process softener, it is estimated, that in the same way as a preheater, there is a saving of one per cent of fuel for each six degrees Centigrade or eleven degrees Fahrenheit. The saving thus amounts to 10 per cent and sometimes as high as 15 per cent of fuel. 4. Method of Conitrol of an.Intermittent Lime-Soda Softener Every manufacturer of a standard water softening system furnishes along with the apparatus full details for its ,control and operation. Hence no attempt will be made here to go into these questions for any other kind than a home-constructed inter­mittent tank softener. The details of the chemical control of practically all lime soda softeners are practically the same. AB stated before, the only control used with a Permutit system is the soap test for hardness. · NECESSITY FOR CHEMICAL CONTROL In the first place, it must be emphasized that a rigid control is absolutely necessary in the operation of any sort of softening plant, and this is particularly true with as crude an apparatu~ as a home-constructed tank softener must necessarily be: No engineer would think of .operating a steam boiler without ac­curate steam gauges. The immediate consequences of lack of control of a softening system are not so serious as disregard of steam pressure, but they are sufficiently expensive. The fail­ure of most water softening systems to deliver a satisfactory quality of boiler feed water may be laid to loose and careless methods of control. The control is so simple that any ordinary :fireman can operate it if he is faithful and conscientious. About the only thing that can disqualify a man of ordinary intelligence from handling the chemical control on a water softener is color blindness. It certainly requires no knowledge of chemistry. HOW TO FIX FEED OF CHEMICALS TO MEET VARIATIONS IN WATER ' An analysis of the water is necessary in order to fix the initial feed of chemicals when the plant operation is :first begun. De­tails of the method of calculation from the analysis have already been given. Once the analysis has been obtained from a chem­ist, it is possible to provide for fluctuations in the pissolved solids in the water due to seasonal changes. USE OF ONE CONSTIT,UENT TO FIX AMOUNT OF CHEMICALS It is usually ::>afe to assume all of the constituents will vary in about the same amount as one individual constituent. The ·easiest one to determine is the bicarbonate ion, the amount of which is tested for by the addition of an acid solution of known strength to a measured amount of the water. For instance, ac­cording to the analysis given above the amount of the bicar­bonate ion is 290 parts per million. From the whole analysis it was calculated thait this water required 1.86 pounds of 90 per cent lump lime and 2.10 pounds 95 per cent soda ash to every 1000 gallons of raw water. A change in the amount of this ion will call for a proportionate increase or decrease in both lime and soda. As will be shown later, when an acid of definite strength, known as "fiftieth normal" is used, each cubic centimeter (mea­ ·ired in an ordinary burette such as is used in any chemical laboratory) equals about 24 parts per million of bicarbonate ion when a 50 cubic centimeter sample of. the water is used. The burette reading in the case. just cited is about 12 cubic centi­meters of acid. If a water requiring that amount of acid for a fifty cubic centimeter sample calls for 1.86 pounds lump 'lime and 2.10 pounds soda ash per 1000 gallons, it is reasonable to in­fer that if all constituents change as do the bicarbonates: when 13 cubic centimeters are needed, the chemical fted must be in­0reased by one-twelfth. For by simple proportion, 12 is to 13 as 1.86 is to 2.02: and 12 is to 13 as 2.10 is to 2.28. Hence the lime feed will be increased to 2.02 pounds and the soda ash feed to 2.12 pounds per thousand gallons. From this it is easy to see how a table may be constructed that would provide for all changes due to variations in the amount of the dissolved solids in the water. PREPARATION OF A TABLE FOR AMOUNT OF CHEMICALS The rule for making a table like this is relatively simple. Calculate from the analysis the amount of lime and soda ash necessary from the formulas of Stabler, or take the value given by the chemist who has made the analysis. Divide by 24 the amount of bicarbonates reported as parts per million in the an­ 1Ysis of the water to be softened. Put down these three values thus: Bicarbonates divided by 24. Lime Required. Soda Required. Take values varying from 4 above to 4 below the value obtained by the division of the bicarbonates by 24 and calculate by simple proportion the amount of lime and soda corresponding to these values, as was just done in the foregoing. paragraph. In the case of the example already cited, the entire table would be as follows: Number of Cubic Cen­timeters of N/50 Acid Amount of Lump Lime Amount of Soda Ash (Bicarbonates di-(90%) R-equired for (95%) Required for vided by 24) 1000 gallons 1000 gallons 8 1.24 1.40 9 1.39 1.57 10 1.55 1.75 11 1.71 1.92 12 1.86 2.10 13 2.02 2.28 14 2.17 2.45 15 2.32 2.62 16 2.48 2.80 HOW TO DETERMINE BICARBONATES The details of the chemical tests to determine the amount of bicarbonates are these: Measure out 50 cubic C·entimeters of the raw water (filtered if muddy) into a porcelain dish. Add three drops of methyl orange indicator. (This is an organic dye that shows whether a solution is acid or alkaline.) The water will .be colored yellow if there are any bicarbonates present. Run in fiftieth normal acid solution from a burette, stirring the water with a glass rod, until the color changes from yellow to a distinct pink. The difference between the reading of the acid levels in the burette before and after, the test is recorded.and is called the bicarbonate reading. APPARATUS AND SUPPLIES FOR TESTING WATER F-0llowing is the equipment necessary for the testing of water al a softening plant: Acid Solution. Fiftieth normal (N/ 50) acid potassium sul­phate (KHS04 ). Five liter lots. Soap Solution. Alcohol (grain, not denatured) and water, 3 :1, s.Olution of castile soap of such strength that each cubic cen­timeter is equivalent to a milligram of calcium carbonate, made up acMrding to Mason's Examination of Water; two liter lots. Methyl orange indicator, water solution, one gram to one liter. Phenolphthalein indicator, ten grams to one liter of 50 per cent alcohol (grain, not denatured). Two 50 cubic centimeter Shellbach burettes with glass stop cocks, 1/10 cc. divisions. Two porcelain evaporating dishes, shallow form, 6 inches in diameter, inside glaze. Two dropping bottles, 'preferabl~ with glass stopper iand grooved pouring point on stopper. One measuring cylinder of 100 cubic centimeters capacity, 1 cc. divisions. One 250 cubic centimeter glass stoppered bottle for harqness test with soap solution. Half dozen stirring rods of glass with rounded ends, about l1alf the thickness of a lead pencil. Half dozen gallon glass bottles with cork stoppers with handle. (Get at any drug store.) · One burette stand for holding two burettes. Half dozen 250 cubic centimeter, low form, Griffin model.' lipped beakers. Two 3 inch· glass funnels. 5. Operating a Home-Constructed Soitener WEIGHING OUT CORRECT AMOUNT OF CHEMICALS Look up in the table the amount of lime and soda ash called for by this bicarbonate reading, and weigh out the correc~t amount for the number of gallons to be treated. Assume that the burette reading on the water ~iven above was 12 cubic ~enti­meters and the tank has a capacity of 10,000 gallons. The amount of lime would be 18.6 pounds and the soda ash 2LO pounds. PREPARING CHEMICALS FOR USE Slake the required amount of lime, if lump lime is used, with the least amount of water necessary to change it to hydrated lime. (A good grade of lime will slake smoothly without lumps. If lumps of unburned material are found they should be gotten . out and their amount allowed for in calculating the lime dosage.) Put it into the lime barrel and fill this with water. In thP. softener designed above, the lump lime is to be slaked and made into milk-of-lime in the barrel by running steam into the lime Boiler Waters ·barrel after the lime charge has been put in and covered with water. Put the required amount of soda ash into the other barrel and allow it to dissolve. This salt goes into solution very readily. TREATING THE WATER Fill the tank about two-thirds full. Add all of the prepared chemicals at' this time. Turn on the stirrers and finish filling the tank. It is always best to add the entire amount of the chemicals to a part of the water and then complete the filling than to fill first and then add the chemicals. The reactions start more readily with an excess of the chemicals and the sludge is coarser. Keep the lime suspension well stirred and flush out the barrel with a hose so as to get all the lime into the softening tank. ;Let the stirrers run .an hour in the full tank. Cut off the stirrers and allow the contents of the tank to stand three hours unstirred. By this time the sludge will have settled out enough so that the tank may be emptied through a filter into the storage tank. VALUE OF OLD SLUDGE It is not necessary to draw off the sludge after each run. In fact it is a decided advantage always to have some old sludge present. The ;very fine precipitates first formed will grow on the larger particles of the old sludge, and hence settling and clearing is quicker'. SUMMARY OF MOST IMPORTANT FEATURES The points to be observed in the actual operation of the softener are these : Employ the correct amount of chemicals. See that the lime suspension is well stirred and that no lime is left in the lime barrel. Add the . chemicals before the tank is more than two-thirds full. Stir thoroughly and never less than one hour. (Practical experience has shown that a tank stirred only by the swirling motion of the foflowing water requires as much as twenty hours to clear, while the same tank with thorough stirring clears in four hours. Stirring increases the capacity of the softener four 1n· five times, and relieves the filter.) Always leave some old sludge in the tank. Never operate a home-constructed softener without a good sand or quartz filter. 6. Testing the 1.'reated Water REASOXS FOR TESTING Every tank of water should be tested to see if the treatment was correct. In the first place, it is possible from these result.'l in a short time to correct the table of chemicals needed for each bicarbonate reading to conform still more closely to the actuai eonditions of the water and the softener. A control is necessary to correct any defects in treatment due to wrong weights of chemicals or to the use of chemicals not up to standard efficiency. After all it is this final test of the. treated water that is most important. These tests should always be made before the softener is emptied, so that any mistakes may be corrected by adding more chemicals or more raw water according to the indications of the tests. DETAILS OF TESTING TREATED WATER Following are the details of the three tests designated as "P ", "1\1", and "H'\ p To 100 cubic centimeters of the water measured into a clean porcelain dish, add five drop~ of phenolphthalein indicator. This will give the solution a purplish red color. While stirring the mixture, run in N/ 50 acid until the solution has just lost its color. Record this amount of acid as "P ". M To the same water sample after the first test has' be.en made, add three drops of methyl orange solution and while stirring the mixture continue to run in acid until the yellow color . changes to a faint red. Read off the total amount of a'Cid used, which includes the amount recorded as ''P'' and call the total reading "M". H Put 100 cubic ·centimeters of the treated water in a 250 cubi·~ centimeter capacity glass stoppered bottle. Run in standard soap solution a few drops at a time and shaking the bottle hard between each addition. At first the lather will disappear in­stantly, but as more soap solution is added, the bubbles become tougher and persist longer. At a point where the magnesium salts have all been used up to form the insoluble compound be­tween the acids of the soap and the magnesium~ the bubbles, though small, will persist' for as much as a minute. This is not an end point. Continue to add soap solution and shake, and as long as there is any calcium radicle present the lather will dis­appear. When the correct end point is reached, there will be a rich lather, which will stand for several minutes. The least · amount of! soap solution whi•ch gives this result should be recorded as a· measure of the hardness of the water as "H." WHAT THE RESULTS MEAN From these results it can be determined whether or not the water has been correctly treated. 1. Soda Ash. ·In a properly treated water one-half of the M reading should be equal to some value which is not less than th·~ H reading plus one nor more than the H reading plus three. If one-half M is less than H plus one, not enough soda ash has been added. If one-half M is greater than H plus three, there has been too much soda ash put in. 2. Lime. When water has been correctly treated with lime, the value for P should fall between one-half and three-fourths the value of M. If P is less than one-half M, there is not enough lime. If P is greater thari three-fourths M, too much lime was used. Following are practical examples: 1. P-7.9, M-12.5, H-4.9. 1j2 M-6.3, which is between H+1(5.9) and H+3(7.9). Soda ash correct. P-7.9, which lies between 1/z M(6.3) and % M(9.4). Lime correct. 2. P-5.3, M-15.2, H-7.5. 1h M-7.6, which is less than H+1(8.5). Not quite enough soda ash. P-5.3, which is le.ss than 1/2 M(7.6). Not quite enough lime. RECORDING RESULTS OF TEST All tests, the amount of chemicals used and "their cost, the amount of water softened, and the total cost for chemicals and labor should .be recorded and kept. Following is a copy of a week's report furnished .by the Austin Street Railway Company to the University Division of Chemistry. A report on the results of the softener used by this company is given in the next chapter. SAMPLE REPORT . Water Softening Report AUSTIN STREET RAILWAY COMPANY Week Ending March 3, 1917 The following is a copy of ·a Weekly Report furnished by the Austin Street Railway Company to the Division of Chemistry. It is given by way of suggesting a convenient form for recording results. Water treated: 30,000 gallons. Labor 4 hours @ 15c.=60c. Sunday: Raw water(50cc.)M 15.1; lime 38 lbs.; soda ash 28 lbs. . ·H+l=4.2 H+3=6.2 H+1=9.5 H+3=11.5 Tank 1. P 6.2, M 12.6, H 3.2; P 7.5, M 14.2, H 8.5, Labor$ .60 H+1=5.6 H+3=7.6 Chemicals 3.09 Tank 2. P 7.6, M 12.8, H 4.6; P-, M-, H-. Total $3.69 114 lbs. lime @ 1hc=$0.57. 84 lbs. soda ash @ 3=$2.52 Water treated: 50,000 gallons. La:bor 10 hours @ 15=$1.50 Monday: Raw water(50cc.)M 15.1; lime 38 lbs.; soda ash 28 lbs. H+l=7.4 H+3=9.4 Labor '$1.50 Tank 1. P 7.3, M 17.3, H 6.4; P-, M-, H . Chemicals 5.15 H+l=8.8 H+3=10.8 Total$6.65 Tank 2. P 6.9, M 13.7, H 7.8; P-, M-, H-. 190 lbs. lime @ 1hc.=$0.95. 140 lbs. soda ash @· 3c.=$4.20. Water treated: 40,000 gallons. Labor 8 hours @ 15c.=$1.20. Tuesday: Raw water(50cc.)M 15.0; lime 38 lbs. ; soda ash 28 lbs. H+1=5.6 H+3=7.6 H+1=5.9 H+3=7.9. Labor $1.20 Tank 1. P 7.5, M 13.9, H 4.6; P 6.4, M 10.9, H 4.9. Chemicals 4.12 H+1=5.9 H+3=7.9 H+l=7.3 H+3=9.3 Total $5.32 Tank 2. P 7.3, M 11.9, H 4.9; P 6.2, M 11.9, H 6.3 152 lbs. lime @ 1h c.=$0.76. 112 lbs. soda ash @ 3c.=$3.36. Water treated:-40,000 gallons. La·bor 8 hours @ 15c.=$1.20. Wednes~ay: Raw water(50cc.)M 14.7; lime 36 lbs.; soda ash 26 lbs. H+l=l0.8 H+3=12.8 H+l=l0.8 H+3=12.8 Labor $1.20 Chemicals 3.84 · Tank 1. P 6.4, M 19.0, H 9.8; P 5.6, M 15.4, H 9.8 Total $5.04 H+l=l0.5 H+3=12.5 H+l=l0.6 H+3=12.6 Tank 2. P 4.5, M 17.4, H 9.5; P 3.7, M 12.4, H 9.6 144 lbs. lime @ 1hc.=$0.72. 104 lbs. soda ash @ 3c.=$3.12. Water treated: 30,000 gallons. Labor 4 hours @ 15c.=60c. Thursday: Raw water(50cc.)M 15.6; lime 38 lbs.; soda ash 28 lbs. H+1=6.2 H+3=8.2 H+l=8.8 H+3=10.8 . Labor$ .60 Tank 1. P 6.1, M l0.4, H 5.2; P 6.1, M 13.2, H 7.8 . Chemicals 3.09 H+1=6.0 H+3=9.0 Tank 2. P 6.7, M 9.7, H 5.0; P-, M-, H-. Total $3.69 114 lbs. lime @ 1hc.=$0.57. 84 lbs. soda ash @ 3c.=$2.52. ·water treated: 30,000 gallons. Labor 4 hours @ 15c.=60c. Friday: Raw water(50cc.)M 15.0; lime 38 lbs.; soda ash 28 lbs. H+1=4.6 H+3=6.6 H+1=4.1 H+3=6.1. Labor$ .60 Tank 1. P 6.2, M. 11.7, H 3.6; P 7.3, M 11.0, H 3.1 Cheruicals 3.09 H+1=6.5 H+3=8.5 Tank 2. P 6.3, M 10.3, H 5.5; P-, M-, H-. Total $3.69 114 lbs. lime @ 1hc.=$0.57. 84 lbs. i;oda ash @ 3c.=$2.52. Water treated: 30,000 gallons. Labor 4 hours @ 15c.=60c. Saturday: Raw water(50cc.)M 15.0; lime 38 lbs.; soda ash 28 lbs. H+1=3.3 H+3=5.3 H+1=4.6 H+3=6.6. Labor$ .60 Tank 1. P 6.6, M 11.9, H 2.3; P 6.2, M 10.9, H 3.6 Chemicals 3.09 H+1=4.4 H+3=6.4 Tank 2. P 6.1; M 10.0, H 3.4; P-, M-, H-Total $3.69 114 lbs. lime @ 1hc.=$0.57. 84 lbs. soda ash @ 3c.=$2.52. REMARKS: Total cost, chemicals and labor, water softening plant for seven days ....................... . ... .. ..... $31. 77 Cost per K. W. hour............................... .0007 Cost per 1000 gallons water.... . . . . . . . . . . . . . . . . . . . . .127 Water treated ......... . ................ .. , .. 250,000 gallons (Signed) H. D. ODELL, W. T. READ, Chief Engh!,eer. Chemist. CHAPTER FOUR RESULTS ·"Does water softening pay 7" It is the object of this chapter to prove that water softening pays. It pays not only in the in­creased ease and satisfaction in operating a steam power plant, . but it pays big returns on the investment in dollars and cents, as shown in the annual reports of any company using softened water. The experience of a number of manufacturers are given, and one particula,r case which has been thoroughly studied by the Division of Chemistry is given in deta~l. 1. Savings Effected by Water Softening Plants GENERAL FACTS Before taking up specific instances, a few general facts will be given. ·The cost of turbining scaled water tubes has been given at a minimum of about l/2 c. per lineal foot and often runs as high as 2c per foot for each cleaning. Add to this the cost oi fuel required to bring a cold boiler up to steaming after each shut-down for scale removal. It is estimated that it costs 10 pounds of coal for hand fired and 15 pounds of coal for stoke.r­fired furnaces per horse power to cool down a boiler for cleaning and to fire it up again for service. Thus a 600H.P. boiler turbined once a month would cost from $250 to $500 annually for clean­ing alone. All of this may be obviated by the me of a water I softener. ­Several years ago the American Railroad Association adopted as an official standard value seven cents per pound of incrusting solids removed from a scaling water as the saving effected by softening the water supply. Against this was charged the cost and maintenance of the softening plant, reducing the net saving to four cents per pound of scale removed. With present high prices of fuel and labor; it is safe to consider ten cents per pound of scale removed as a fair estimate of saving due to softening water. This means that if a water supply carried 4 pounds of scaling ingredients per 1000 gallons, the removal of this scale would be worth at least 40 cents to the efficiency of the plant. It has been shown that the chemicals necessary to remove this much scale .will cost a little more than 7 cents. Assume labor and overhead expenses, such as interest on investment and depre­ciation of softening plant, cost 7 cents more, which is a most liberal estimate. This gives a softening cost of 14 cents per 1000 gallons. The softening has thus saved 26 cents per thousand gallons of water evaporated. Even if we take the 'figure given ·by the American Railroad Association, the saving is 14 cents per pound of incrusting solids removed~ or-100 per cent profit on the cost of softening. One practical engineer has summed up the book-keeping on the cost and profit of a water softener thus: ''Charge the softening plant with: (1) Interest and depreciation on softening plant, 10 per cenl. (2) Labor cost annually. (3) Chemical cost annually. (Actual cost.) Credit the softening plant with: (1) Cost of turbining tubes, with untreated water. (2) Cost of tubes lost due to scale .and corrosion, with un­treated water. (3) .Cost of shutting down for cleaning, with untreated wate1:. ( 4) Cost of tube repairs, when untreated water is used. (5) Cost of boiler compounds used, with untreated water. (6) Differ_ence in cost of coal between use of treated and un­treated water. (7) Cost of repairing furnace settings, when untreated watel' was used. (8) Increase in horse power hours without in~reased invest­ment. (9) If hot process softener is used, increased temperature of feed water. SOME ACTUAL SAVINQ EXPERIENCl~ absolute personal knowledge has done more to impress upon the chemists of the University of Texas the advantage of treat­ing water to remove scaling cwnstituents than all of the published rf.cords of distant lllanufacturers. NATURE OF PLANT AND CONDITIONS AT BEGINNING OF EX­PERIMENTS In the summer of 1916, the University Division of Chemistry took up the study of water softening, Rsing the plant of the Austin Street Railway Company for experimentation. The plant was home-constructed, and had all the faults of . such systems. It consists of two 10,000 gallon cypress tanks on a platform, with barrels for chemical solutions and the usual arrangement of sludge and drain pipes. There is a large cooling pond for thr turbine exhaust, which serves as a storage tank. The most glar­ing defect of the plant was the utter lack of stirring devices . .The swirling motion imparted to the raw water as it was fed into the tank and mixed with the chemicals had been considered to be adequate agitation. As a result the plant was operating in a very inefficient and unsatisfactory manner. It required 20 hours to settle and clear each tank of water. Since an averag& of 30,000 gallons of water were needed for a day's run, it was necessary to use as much as 10,000 gallons of raw water. As it '98 University of Texas Bi1lletin happened, incorrect amounts of chemicals were being used, thr: water being badly overtreated with soda ash. So the amount of raw water about comprnsated for the excess of chemicals, and a . fair degree of softening was obtained. There was no chemical a b Fig. 16. Effect of stirring on precipitation. One minute after addition of chemical to water: "a" stirred, "b" unstirred. a b Fig. 17. Effect of stirring on precipitation. Five minutes after addition of ch·emical to water: "a" stirred, "b" unstirred. · Boiler Waters 99 control and the whole treatment was somewhat of the hit-or~miss variety that is so often the case when a chemical process is car­ried on without some preliminary chemical advice and adjust­ment. .The president and engineer of the company were thoroughly open to advice and willing and anxious to cooperate fully in correcting defects. CHANGES MADE The first thing done was to change the feed of chemicals, anrl the Iiext thing was to install adequate stirrers. A small motor­clriven rotary air compresser was available, and an air line was pi;tt in to the tanks, and stirring carried out as described in Chapter III. The immediate result was that the tanks that haJ taken 20 hours to settle, cleared in 4 hours. As much as 50,000 gallons of treated water have been gotten from this softene~ in IJ. day's run, and 60,000 gallons could have been gotten just as easily in the same time. NATURE OF THE RAW AND TREATED WATER When the work was first begun, and for several months after­ward, the water was very hard and high in permanent hardness. An analysis made at the beginning of the experiments gave the following values : Parts per Million Calcium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Magnesium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Sodium ... . ....... . .. . ............... . .. 180 Sulphates . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . 276 Bicarbonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437 Chlorides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 Iron, aluminum and silica were present in negligible amounts. According to the Stabler formula this water will give 4.85 pounds of scale per 1000 gallons, with a hardness coefficient of 77 per cent. Some months later, due to emptying Lake Austin, 'the amount of dissolved solids dropped to nearly one-half its former value, which made softening much less of a problem. The water supply is a shallow well, some four hundred yards from the Colorado River. It was dug with the expectation of getting the water that fills all the sands near the river. For some time, about the same kind of water was obtained as that of the river. When the Austin dam was completed, and Lake Austin, which lies some two miles up the river, was filled, the quality of the water suddenly changed. The dissolved solids more than doubled, and" there was a big increase in the relative sulphate content. It seems that the pressure -0f the water in the lake, there being about a sixty foot head, forced water through veins and strata which were impregnated with gypsum as well as through a limestone formation. During the summer of 1911, in order to furnish the rice growers of South Texas with ade­quate water supply, the lake was gradually drained. Within a few days, the quality of water in the street railway company's well improved and resembled the river water once more. -, Samples of the treated water, taken on various O'ccasions, showed from 6 to 14 parts per million of calcium and an average of 10 parts per million of magnesium. This is about the limit of efficiency -0f a cold-process lime-soda softener, that is from 35 to 70 parts per million of incrusting solids remaining in the water. Since there was always a slight excess of soda ash used, the remaining hardness did not form a scale, but prod?ced a slight sludge in the boiler. METHOD OF OPERATING SOFTENER 'l'he details of che)nical control, calculations of chemicals nec­essary by the method of variation of bi-carbonates, testing of treated water, and the actual operation of the softener were all carried out as described in Chapter III. FOAMING When the water had been softened without chemical control, it was often the case that an abnormal excess of soda ash was used. As a result large amounts of sludge were produced in the boilers, and there was violent foaming. Even after correct treatment a great deal of foaming was experienced. With a view of getting at the cause of this 'trouble, samples of water were taken at the water line from the gauge cocks and the fol­lowing analytical results obtained: 1 2 I Suspended Solids ... . .......... . 1,188 492 2,016 Collodial Matter: (a) Ferric Oxide ............. . . 31 43 7 (b) Alumina ...... . ..... ..... . 5 32 11 Total Dissolved Solids . .. ....... . 21,269 26,447 35,971 Ions Calcium .................. . 68 52 45 Magnesium . .. ............• 10 8 19 Sodium (by difference) ..... . 7,520 9,596 12,627 Carbonates ................ . 104 138 188 Bicarbonates .............. . 29 None None Chlorides ..•............. .. 7,168 9,314 12,018 ·sulphates .•................ 5,987 7,308 9,884 Hydroxyl None 8 18 These results show not only an abnormally high concentration of alkali salts, but an excessive amount of suspended solids. The suspended matter obtained in these analyses proved to be largely silica, ferric oxide, alumina, calcium and magnesium salts, and some adsorbed alkalies. There were no carbonates at all in the suspended material. It was extremely finely divided, in fact was as fine as flour. Without altering the chemical feed or any other factors in the operation of the softener, regular blowing down was advised. By blowing down several inches each night after the turbine was shut off, the foaming was stopped and has not since occurred to any serious extent. While the concentrations of the boiler water in these cases are very high, the amount of sodium carbonate in this concen­trated water is by no means as much as in many artesian waters ' that cause no trouble from foaming. It seems evident that the principal -cause o( foaming here was the suspended matter. COST AND SAVINGS DUE TO WATER SOFTENING Through the courtesy of Mr. W. J. Jones, president of the company,..the following cost figures were made available : Boiler R-0om Expenses: One Year with One Year with Untreated Water Tre~tted Water Total Fuel and Water ......... . . .. ... $23,894.03 $22,017.96 *Labor (aside from regular firemen)... . 2,746.26 2,578.59 **Cost of Chemicals ............... : . 10.72 923.00 Net saving of Second Year over First. .... .... .... ... $1,l31.45 Cost of Softener. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 21. 7 2 Per cent Profit on Investment in one year's operation. . . . 182% AVERAGE COST OF OPERATION Following is a summary of a week's run for the week ending April 28, 1917 : Water treated ................. . .. 210,000 gallons Labor 28 hours at 15c . . .... . ............ . $ 4.20 Lime 756 lbs. at 1hc...................... 3.78 Soda 564 lbs. at 3c ....................... 16.92 Total cost ........................-. ... $24.90 Cost per K. W. hour ..... ......... $0.0005 Cost ·per 10 0 0 gallons. . . . . . . . . . . . . .118 (During the past summer,· the cost has dropped, owing to the change in the water, to about Sc per 1000 gallons.) The most interesting things about the experiment are the results obtained in smooth operation of the plant and greatly reduced 'tlost of maintenance. When the raw water was being used, four laborers were continually employed cutting seal~ from one set of boilers while the other set was steaming. In reply to a question as to how often each boiler was cleaned, one of the firemen of the plant replied ''Just as fast as we couW get to them.'' As soon as one boiler was cleaned and filled for use, another was shut down and opened up for cleaning. One boiler kept in service less than two months, while untreated water was being used, required three full weeks for cleaning, with two turbines running all day, and two men going after the scale with chisels. Four barrow loads of seale came out of the boiler, and it was extremely hard scale. The same boiler was operated a full year with treated water. During this time, there was 'This Includes in the first year the cost of cleaning boilers, In the second not only handling softener, but various changes and improvements in boiler room, for which under old conditions men bad no time. ..First year covers boiler compounds used occasionally, second year lime and 5oda ash. quite an interval when the softener was operated without con­trol, sometimes only partly treated water being fed, and at inter­vals ra.:w water going in. Even under these unsatisfactory con-­ditions, when the boiler was opened up after a year's solid ser­vice, it was cleaned in four hours easy work, and about half a barrow of scale, mostly sludge was gotten out. Put the two up against each other: two months run, three weeks hard work to clean, four barrows of scale; six times the length of run, one­fiftieth the labor, one-eighth the scale. Where tube trouble was a continuous thing under old condi­tions, sometimes as many as seven going out in a day's run, when the load hovered around the peak all day, now tube trouble is unknown. A boiler inspector for this district, who has five hundred boilers under his care, has ranked the boilers of the Austin Street Railway Company as among the best half dozen in point of ·condition of his entire five hundred. SU:lUMARY From the above :figures it will be seen that this company paid for the softener in one year and cleared $450 besides. Charging the softener with 10 per cent to cover depreciation and interest on the investment, a yearly charge of about $65 is balanced against a yearly net saving of $1,100, which is a fair profit.. $1,000 p~r year may look like a high chemical bill, but this firm pays this bill and clears more than $1,000 besides as compared with the cost of operation with untreated water. These results were gotten with the crudest and least efficient and most ex­pensive (in point of operation cost) of all types of softeners. It is true that the plant had the care and attention of a chemist for some time; but the chemist had to learn first the practical details of water softening. This experience is available to the public in the pages of this bulletin. By an intelligent reading of what· is given here, by faithfully following directions, sup­plemented by answers to specific questions suggested by indi­vidual difficulties, any engineer ought to be able to operate a water softener successfully. ACKNOWLEDGEMENTS The writer wishes to thank the following gentlemen fo,r their timely suggestions, frank and helpful criticism, and valuable informatiop given in bulletins, personal letters, and interviews. W. T. Alliger, Chemist, Harrison Safety Boiler Works, Phila­delphia, Pa. S. B. Applebaum, Assistant Technical Manager, · The Permutit Company, New York City. Clarence Bahlman, Chemist, ·Cincinnati Board of He.alth, Cin­cinnati, Ohio. R. C. Bardwell, Chief Chemist, Missouri Pacific R. R. Co., Kan­sas City, Mo. Edward Bartow, Director State Water Survey, University of lllionis, Urbana, Ill. A. A. Boettcher, Sales Engineer, Wm. Graver Tank Works, East Chicago, Ind. W. A. Converse, Secretary DearbQrn Chemical Company, Chi­c.ago, Ill. D. K. French, Directing Chemist, Dearborn Chemical Com­pany, Chicago, Ill. G. L. Fowler, Consulting Mechanical Engineer, New York City. G. H. Gibson, Chemist, Harrison Safety Boiler Works, Phil­adelphia, Pa. G. A. Johnson, Consulting Engineer, New York City. W. P. Mason, Professor of Chemistry, Rensselaer Polytechnic Institute, Troy, N. Y. M. F. Newman, Manager Water Purifying Department, Wm. B. Scaife & Sons, Oakmont, Pa. E. P. Schoch, Professor of Chemistry, University of Texas, Austin, Texas. John H. Smith, Jr., Chief Chemist L. 1\1. Booth Company, Jersey City, N. J. · Herman Stabler, Hydraulic Engineer, U. S. Geological Survey, Washington, D. C. F . F . Vater, President Power Plant, Specialty Company, Chi.­cago, Ill. Wm. H. Walker, Professor of Chemistry, Mass. Inst. of Tech­nology, Cambridge, Mass. The author is also indebted to W. J. Jones, President, and H. D. Odell, Chief Engineer, of the Austin Street Railway Com­pany, for their cooperation and many courtesies during the tests and experiments on a water softening apparatus at their power plant. For the chemical control the method given by Professor J. R. Bailey of the University of Texas has been used.