US2476598A - Treatment of steam boiler water - Google Patents

Description

Patented July 19,. 1949 Ralph E. nau moum Lebanon, ran, assignor a Hall Laboratories, Incorporated, Pittsburgh, Pa., a corporation. of Pennsylvania No Drawing. Applicationl ebruary 16, 1945,

' Serial N- 578,349

This application is a continuation-in-part of my copending application Serial No. 483,127, filed April 15, 1943 which has become abandoned.

This invention relatesto treatment of steam boiler water. Particularly, it-relates to the rem-e edy of certain, difficulties which have been very widely encountered in modern steam generating practice, which have been an obstacle-to the 16 Claims. (01.210-23) given any intimation of benefit or value in establishing a maintained ratio of potassium to sodium 1 in the boiler water. For instance, Hall and Jackson Patent 1,903,041 mentions alkali-metal metaand pyrophosphate, for treatment of the boiler water, and in the same breath considers the simplest method of attaining the maintenance of the necessary sodium sulphate-sodium carattainment of best operating efliciency, and the search for whose solution engages widespread attention.

Withthe useof higher steam pressures and ratings, a new set of scale problems has arisen in which calcium and magnesium do not enter into the composition of deposits as primary factors, but which are characterized rather by sodium salts and silica.

In the chemical treatment of water for steamgenerating purposes, the sodium salts have been universally used. So completely have the sodium salts dominated water-treating processes and philosophy, that boiler-water alkalinities are universally expressed as units of caustic soda or sodium carbonate; and recommendations for protection from embrittlement are given as a ratio of the sodium sulphate in the boiler water to the alkalinity therein expressedas sodium,

carbonate.

While there have beenbroadening suggestions in certain prior patents on water treatment that the alkali-metal 'salts ingeneral may be used for their purposes, no special advantage in the use of any other than the sodium salts has been noted or claimed. In fact, in the treatment of water for boilers sodium salts have been employed as a matter of course, because of their more common availability, because they cost less per pound, and because fewer pounds of sodium salt are required than of potassium salt. Thus the phosphate radical for conditioning the boiler water is more cheaply supplied as sodium phosphate than as the potassium salt. The same is true in the use of soda ash rather than potash for the lime-soda softener, sodium sulphate for prevention of embrittlement, sodium sulphite for deoxidizing the boiler water, sodium aluminate for coagulation, and sodium silicate as used in water treatment. In the regeneration of zeolites, sodium chloride is used, thus providing the necessary basic radical at lowest cost. So far as the treatments heretofore or currently carried out are concerned, potassium and the other alkali metals have been and are considered to be merely more expensive and less available chemical equivalents of s'odium. No one has .bonate ratio for protection from embrittlement,

"However, I have found that in its behavior in high temperature boiler operations potassium behaves ina radically different manner from sodium and'that if a minimum ratio of potassium to sodium be maintained in the boiler water in conjunction with appropriate limits of alkalinity; thediificulties encountered with the formation of silica or silicate scales or deposits in the boiler, superheater and turbine are overcome or minimized. With this ratio maintained, I have found that silica, far from being an unwanted constituent of the boiler water, is one that serves useful purposes; that by the maintenance of potassium and silica or similarly hydrolyzable radical such as borate or metaborate in theboiler water in appropriate relation to the other constituents therein, I can provide in the boiler water the alkalinity required therein for all purposes, and at the same time prevent embrittling action thereby on boiler metal, also the destructive corrosive action of boiler water on the metal surfaces where it concentrates highly due to a steam-blanketing or film-boiling development.

' I have also found that hide-out is prevented or minimized when the minimum ratio of potassium to' sodium is maintained in the boiler water, thereby substituting for the sodium phosphate, sodium sulphate, and sodium silicate in the boiler water which are readily deposited by evaporation to dryness at areas where film boiling or steam blanketing occur, potassium phosphate, potassium sulphate, and potassium silicate which make the water more resistant to evaporation to dryness at areas of film boiling or steam blanketing. Furthermore, I have found that by the use of silica for the maintenance of these conditions, magnesium-silicate, and not the less readily dispersed magnesium-phosphate or magnesium-hydroxide sludge, is caused to form in the boiler water when magnesium is present, the result being cleaner boilers, without accumulation on the surfaces of a magnesium-rich sludge too coherent to be entrained in the circulating boiler water. K

Therefore, any boiler water which contains even a very small amount of silica, and in which unfavorable characteristics in the sodium-containing water, which are obviated in the potassium-containing water because of the diiferent properties of potassium silicate. To realize these favorable conditions, I must build the overalldilute boiler water onthe potassium basis, since it cannot be disassociated from the condition of concentration it assumes at these loci in the boiler where rapid evaporation, steam-blanketing or film-boiling occurs. Every bubble of steam that forms at the evaporative surfaces represents a small area thereon on which concentration of the overall boiler water occurs. Similar concentration occurs in the superheater in which any boiler water carried over in the steam becomes highly concentrated or even completely evaporated at the superheater temperatures, while in the turbine, with decreasing temperatures and pressures there occurs deposition on the turbine parts. So dominantly chemically different are the potassium and sodium silicates in boiler waters, that whereas current practice generally excludes silica so far as possible from the sodium boiler waters, in the potassium boiler waters silica is beneficial, and in consequence its maintenance therein in measured amount by addition thereto of silica in satisfactory form as required may be recommended.

To understand this huge difference between sodium and potassium in the boiler water, especially in relation to silica, it is essential to know the characteristics of the ternary systems what happens to the boiler water and the salts dissolved therein in the boiler, and also what occurs in any part of the boiler water which is transported by the steamto the superheater and thence to the turbine. I shall discuss these in turn. In this specification, when I use a chemical ratio such as K/Na or SiOz/NazO, I mean the ratio of their values expressed as chemical equivalents. For example, I shall use 30 as the equivalent weight of S102. I shall express temperatures in Fahrenheit degrees.

The soluble components of a boiler water treated with potassium chemicals are mainly potassium and sodium hydroxide, phosphate, chloride, sulphate and silicate. Borate will be present only if specifically used in the treatment. I determine the separate concentrations of potassium and sodium therein, express them as equivalents per million. and thus obtain the ratio K/Na. I also determine the alkalinity, phosphate and silica, and borate if present, and likewise express them as equivalents per million. The alkalinity may be derived wholly or in part by the hydrolysis of the phosphates and silicates, and borates if present. when a boiler water is considerably concentrated as at steam-blanketed surfaces or in the superheater. the hydrolytic action is reversed, with formation anew of the parent salts, and at the same time with corresponding removal of the hydrolytic alkalinity from the boiler water. ilkalinity which so recombines on concentration of the boiler water I designate captive alkalinity; that which remains uncombined and concentrates up concentration of the boiler water I designate free alkalinity. l subtract the captive'alkw iinity of the phosphate, and borate if present,

' from the alkalinity as determined, thus obtaining the free and captive alkalinity associated with the silica in the boiler water, and designate this Z20. Thus the ratio S1O2/Z2O signifies the relation between silica and both the captive and free alkalinity associated therewith. Further sion z Bio- 2 where 02 0) signifies the silica associated with the potassium alkalinity, and

mi o) that with the sodium alkalinity. The ratio of potassium alkalinity to sodium alkalinity KzO/NaaO is assumed to be the same as K/Na.

The relation between the silica associated with the potassium alkalinity, and the alkalinity is given by the ratio and likewise for NaaO.

The advantage to a boiler water of potassium silicate rather than sodium silicate is inherent in the properties of the ternary system H2OK2SiO:-Si02 Asthe ratio of Si0 /K,O

in the boiler water increases, the solid phase in equilibrium at saturation changes from K2Si2Ot(SiO2/K2O=2) next to and finally to SiOz itself. The solid phase separating is a function or temperature as well as the SiO, /K O ratio, and its composition is not necessarily of the same SiOz/KzO ratio as the solution from which it forms.

So long as the ratio of in the boiler water is held below about 2, and the between the potassium disilicate and silica against by supplying S102 or K20 as r ag-races sw mo ratio are reached. Thus, so long as the potassium salts dominate the bollerwater, and the ratio of Sig 2 /K30 is held as mentioned, there will be no silica prob- 1cm in the boiler water, and silica therein is advantageous and not deleterious.

In theHzO-NazSiMiOz system, both so dium metasilicate (NaaSiOa) and disilicate (NazSizOs) are strongly retrograde in solubility in the higher range of boiler-water temperatures, approaching zero solubility at: critical temperature oi water in common with sodium sulphate, sodium phosphate, and sodium carbonate, which likewise are of retrograde solubility in the higher range .of boiler-water. temperatures. The solid phases indicated in the system as the ratio increases are NazSiOa (SiOa/Na20=1), Naisiios soluble Si02 may also form on the surfaces, and

wherever formed be extremely dimcult of removal. Furthermore, the sodium-silicate salts are not readily redissolved, act as a cement on sludges, and likewise cement themselves very tightly to the surfaces where they form. In a boiler water, therefore, silica in' conjunction with sodium is extremely deleterious", and because of these deleterious relationships with sodium, silica has been condemned but sodium has gone soot-free. To

the best of my knowledge, the finger of suspicion has never before been raised against sodium in this connection. When the boiler water concentrates, sodium and potassium therein, have diametrically opposite characteristics in relation 7 to silicate, and the sodium ion because of the insolubility and instability of its silicate combinations is directlyv the cause of and heavily a contributor tothe deposition or silica and silicate scales in boilers and turbines. Thelvalue of K/Na mustbe so maintainedin the? boiler. water that sio,, K,o g.

and not e ioi yn o dominates all silicate equilibria that may develop. Because the potassium silicates are stable and very soluble, their considerable hydrolysis in-the dilute overall boilerwater can be advantageously used to maintain desired alkalinity therein in the form of captive and not free alkalinity. Hydrolysis of potassium and sodium phosphate, and borate if present, may provide the captive alkalinity in part. The remainder is provided merely v uired to hold the ratio notless than 1 andfas an upper. limit not more than 2), and in concentration to give the desired pH value or alkalinity titration of the boiler .6 water. -As pointed out previously,'the overall boiler water should be alkaline (preferably of a pH of. 9.5-11) to prevent corrosion and to keep sludges dispersed. When any concentration'oif such a boiler water occurs, as on the surfacesin sections where the evaporation rate is high, it

results in retraction of the captive alkalinity and formation of the parent salt, This restriction of alkalinity avoids the highly aggressive action of strong caustic solution which would result if free alkalinity were permitted to concentrate. Similar control with NazO and S102 of a sodium boiler Water containing any appreciable amount of silica, is not permissible because it would open thedoor to the many troubles entailed in silica and silicate scale formation.

The boiler water and the steam generated therefrom comprise an isobaric isothermal region ending with the passage of the steam to the superheater. Thus, a'boiler operating at 1400 p. s. i. pressure, for instance, does not deviate in any considerable degree from that pressure. The boiler water, considered as a whole, and as sampled for testing, is a quite dilute solution, 2500 tion) or less being a common concentration. The temperature of the overall boiler water, and likewise of the saturated steam, corresponds. very closely, therefore, tojthe' temperature of pure water boiling .atithistpre's'sure; for "examine, 587 for-1400 p. s. i. absolute-pressure. rData on the full range of boiler-water temperatures'andpressures are set forth in"stea'm tables,";lfor instance those published by the American Society of Me-" chanical Engineers. The separationin the-boiler from the overall dilute boilerwaterof sludge such as calcium phosphate iorfhydroxyapatite) .or magnesiumv silicate, occurs iin accordance with the isobaric isothermal saturationconstants of the substances. :Hdwevr, departure,from' ithef isothermal, but'n'ot the isobaricconditiomoc'curs where a steam bubble is generated or where a. hot section due to steamblanketingoriilm boiling exists, and locally in such areas a minute portion centration in some degree. In these local areas, zthe steam assumesa pressure which is thevapor,

pressure of the overall boiler water at itstemperature; but the fil'm of boiler water'adhering 5.0 to-t'he tube surface which is overheated in .vary ing amount likewise, rises ,in temperaturejand evaporateseither to dryness or in accordancewith Raoults Law to a concentration enough" greater than that of the dilute boiler water so is equal to the-normal steam pressure.

' Ate in rhyme signifies the'increase in tempera-.

v'55 that its vapor pressure at its higher, temperature 'ture over that of saturated steam, $0,131; pressure p0 which results in saturating a solution atthe 6 same pressure pa with respect to at least'one solid phase. vThus, ifa solution saturated with. respect to a solid phase at temperature t is in equilibrium with its-vapor phase at pressure p0 andat this designate t-to as At without subscript.

- This concentration in the film .results' in the] 7 deposition as scale of boiler-water salts, particularly those of retrograde solubility at the ;tem-

peratures concerned, such as calcium sulphate, magnesium hydroxide, and calcium silicate alsosodium silicate and silica derived from SiO2/Naz0 equilibrium relations. Under severe conditions, it

P. P. M. of dissolved solids therein (0.25% solur ,ofthe overall dilute'boiler water undergoes concauses the formation and deposition or the complex sodium-aluminum-silicate, and of sodium phosphate and sodium sulphate therewith. Saturation in the film with sodium sulphate, sodium sodium in the boiler water solves the problem of silica or silicate scale therein.

In the superheater, boiler pressure obtains, but the temperature of the steam passing there,-

carbonate, sodiumsilicate or salts of like soluthrough is raised to a designed degree of super.- bility requires an increase in temperature (Ats) heat. Hence, any boiler-water and the salts disin the film at the tube surface of not more than solved therein transmitted by the steam will evapthat required for calcium sulphate or the orate to dryness or remain partly or wholly discomplex sodium aluminum silicates is much less solved in accordance with their properties as than this. In these limited areas then, scale for- 10 defined by their isobaric polythermal saturation mation occurs in accord with the isobaric polyrelations. Thus, any boiler-water carried over in thermal saturation relations of its constituents, the steam must concentrate in its passage through even though the layer of higher temperature the superheater to an extent such that at the steam at the evaporating surface of the film of temperature of the superheater its vapor pressure boiler water is scarcely finite in thickness. is that of the steam pressure, or else evaporate Contrast of the relations under these condicompletely.

TABLE 1 Conditions in the isobaric olythermal region of the superheater Properties of some boiler-water and other salt solutions under these conditions Al. 400 p. S. 1. At 1400 p. s. l.

Condition at 7 Condition at salt 801mm Saturation gg g gg 700 Saturation 525 1000 Tempera- Al, on Per Tempera- AI. i

turc ture er Cent Cent Sodium Sulphate 454 9 33 Complete Evop- 592 5 l5 Complete Evaporation. oration. Sodium Phosphate 1 Similar to Sodium Sulphate Similar to Sodium Sulphate Do. Sodium Carbonate. Similar to Sodium Sulphate Similar to Sodium Sulphate Do. Sodium Di or Met Similar to Sodium Sulphate Similar to Sodium Sulphate Do. Sodium Chiorido 4&8 43 34 652 42 Do, Sodium Bromide 522 77 64 753 166 74 Do. Potassium Chloride 49S 53 50 691 104 61 Do. Potassium Chloride 597 152 76 874 287 (Est Do,

plus At Sodium l3romide 984 413 Unsaturated.v In Solution Potassium Disilicate 533 88 78 1 413 do Do.

Potassium Metasilieatc 1,193 1 255 Unsaturated. 413 do Do. Sodium Hydroxide 1 255 as r (f 413 93 4 Do. Potassium Hydroxide Similar to Sodium Hydroxide I Similar to sodium Hydroxide I Do.

1 Saturation pressure less than boiler pressure.

1 Vapor pressure date not available.

3 Saturation pressure never attains boiler pressure. i Concentration of the unsaturated solution.

tions of sodium and potassium with silica emphasizes their fundamental difference in boiler water. The sodium metaand di-silicates become of retrograde solubility with increased temperature. They reach saturation and are deposited at Ats of not over 10. They are difilcultly redissolved. In contrast to the behavior of the potassium silicates, the sodium silicates split off S102 at relatively low values of SiOz/NaaO. In connection therewith, they exercise strong cementing action on any sludge or other material deposited simultaneously, and thus prevent its Table I gives the results occurring when solo solutions of different salts are subjected to this condition. These results are calculated from the data published by George W. Morey in Journal American Chemical Society, vol. 39, p. 1173 (1917); by N. B. Keevil in Journal American ChemicalSociety, vol. 64, p. 841 (1942); and by F. C. Kracek, International Critical Tables, III, p. 370. 7

Case A portrays .the results when boiler pressure and temperature are respectively 400 p. s. i. absolute and 445, temperature at superheater outlet is 700, and superheat is 255. With its temperature raised to only 454, a solution containing sodium sulphate becomes saturated (At= 9) and evaporates to dryness with any further elevation of temperature. Solutions of sodium phosphate, carbonate, and dior meta-' silicate behave similarly with similar low value of Ats. Solo solutions of sodium chloride or bromide, potassium chloride or disilicate, with Ate respectively 43, 77, 53 and 88, will likewise evaporate to dryness, as will also the mixed solution of potassium chloride and sodium bromide with At; of 152. Potassium metasilicate attains the full value At=255 without reaching saturation, as does likewise sodium hydroxide, the latter solution being of 86% concentration at equilibrium.

Case B summarizes results at 1400 p. s. i. absolute, final temperature of 1000, and superheat of 413. Under these conditions, evaporation to dryness occurs for a number of salts. Potassium disilicate, however, as well as metasilicate, does not-attain saturation at full superheat value (At=413)', and sodium hydroxide concentrates to. 93%. The mixed solution of potassium chloride sodium bromide illustrates how mixtures may behave difierently from salts in solo solution, since upon attaining a temperature of 874 (Ata=287), it reaches saturation; next evapobut a number of them. usually sodium chloride,. sodium sulphate, sodium phosphate, sodium silicate, and sodium hydroxide. In such mixtures.

at saturation. the total amount of dissolved material is in general somewhat greater than the;

saturation concentration of any one of the components, and the usual result of mixture is that the mostsoluble component deviates least from its solubility in solo -solution, the lesser soluble components having their solubility considerably reduced. Thus, sodium chloride'and sodium hydroxide in equilibrium with both solid phases at 356 F. di-solve respectively to the extent of 8.8% and 76.1%. Their solo solubilities at that temperature are 30.7% and 82.3%, respectively. At 302 F., a solution containing 2.9% of K01 and. 73.6% of KOH is in equilibrium with both solid phases.v At this temperature the solubility of KOI-I is 73.6%, and of KC140.2%. At 373, 31.7% of KCland 16.3% of NaCl saturate a solution with regard to both phases. The solo solubilities are, 43.7% for X01, and 31.1% for NaCl. Thus, in

mixed solution KCl is 72.5%, and NaCl only 52.5% of their respective solo solubilities. No data are available for higher temperatures, but it should be noted that with increasing temperature, the number of units of NaCl or KCl present in solution per unit of their respective hydroxides therein, steadily increases, and their concentration at customary superheater temperatures in by small value of At but its tendency to super saturate highly without crystallization may account for the tact that superheater deposits usually contain only a small amount thereof,

probably enough, however, to exercise cementing action on the larger amounts of sodiumsulphate, sodium carbonate and sodium phosphate.

At temperatures beginning at slightly above 578 F., the retrograde solubility curve of sodium sulphate (less markedly,- sodium phosphate also) is reversed in the presence of sodium hydroxide,

sulphate at 486 is 22%; at 706, is 44%. In a 25% solution of sodium chloride at 706, the

= solubility of sodium sulphate is 17%. No data are available, of which I am aware, to give an insight into what the-solubility of sodium sulphate would be at these temperatures in a solution containing both sodium chloride and caustic soda. This means that in high temperature dominating amount of sodium sulphate present relation to their hydroxides may be very much higher than mdicated in these examples.

Because substances of retrograde solubility in the range of boiler or superheater temperatures, such as sodium sulphate, sodium phosphate, and sodium carbonate, may attain saturation at very small values of Ate '(not more than 10 F.), they begin their precipitation in the concentrating droplet of boiler water very shortly after arriving in the superheater. As the droplets of carryover strike against the walls of the superheater, some of the precipitated material ,probably sticks to the walls in much the same manner that muddy water impinging on a surface leaves some of its mud sticking thereto. Its adherence, and the amount thereof, are favored also by the high temperature of the superheater walls, especially in conjunction with the retrograde solubility of these salts. Thus, deposits form in the superheater. The solubility of sodium silicate also be.- comes retrograde in the range of superheater temperatures, and its saturation is also effected in the great'majority of stationary steam plants, whether there naturally or by the treatment applied, makes the early appearance to some extent of sodium sulphate as solid phase in the superheater very probable even at the higher temperatures, and almost certainly present in much greater proportion at low superheater tempera- I tures'. i

As the droplets concentrate in the superheater, the salts whose solubilities'strongly increase with temperature increase, such as sodium chloride, presumably have little or no tendency to adhere to the walls thereof, since it is characteristic that superheater deposits almost invariably contain low content of this type of solids. Inasmuch as the solubility of potassiumsilicate increases with temperature, the substitution of potassium for sodium in the-boiler water not only removes a potential deposit-forming material from the processes occurring in the superheater, but obviates as well the cementing action characteristic of sodium silicate.

With the exception of caustic soda, caustic potash and potassium silicate (sodium silicatealso because of super-saturation as already noted), a considerable proportion of the material carried from the superheater into the turbine is probably in the form of solid salts. If evaporation has been complete, these salts will be dry; but usually, the droplet of boiler water cannot evaporate to dryness and will have concentrated in the super-- If the boiler water is so treated that it contains only the captive alkalinity derived from the hydrolysis of potassium silicate and phosphate, and if the ratio SiOz/ZaO is held at a value of about 1.6-2.0 (not higher), the magma will approach solid condition as nearly as possible without danger of silica deposition.

Passing from the isobaric polythermal region of the superheater, the steam next enters the turbine where conditions are polybaric, pressure decreasing, and polythermal, the total temperature and degree of superheat likewise decreasing.

Dry salts entering the turbine will remain dry for a considerable portion of the passage therethrough, tending to dissolve only in the later stages where superheat has been largely dissipated. When they begin to dissolve, they are resolved into a magma, of the same physical characteristics as the magma formed by boiler water which does not evaporate to dryness in the superheater.

The droplets of boiler water which concentrate to a magma in the superheater are more or less pasty, therefore tacky in some degree, and almost certainly they adhere to some extent to the turcine blades. In this region, the characteristics of the substances in the magma, as defined by their polybaric polythermal saturation curves determine the course of events in step with the changing conditions in the turbine. With the general drop of temperature, pressure, and degiee of superheat in the turbine, the characteristics of the saturation curves of the difi'erent substances concerned change synchronously therewith, and there is little reason to suppose that, in general, the condition of the magma will change considerably until the region of conden-v sation, i. e., disappearance of superheat, is approached. While the magma probably could not of itself accumulate in thickness on the turbine parts in the maelstrom of flow, the rapidly recurring localized changes in the steam in its passage over blade and through nozzle may result in enough dissolving and redeposition so that the solid suspended salt may well be cemented to the turbine blade in the form of deposit found when the turbine is opened. In. much the same manner, dry snow undergoes packing and hardcning with slight changes of pressure or temperatuie. In accord with this conception, deposits the high-pressure stages of the turbine consist mainly of sodium chloride and sodium silicate, with perhaps some sodium hydroxide admixed, and sodium sulphate also under some of the conditions already discussed. With dropping temperatures and pressures, the incongruent solubility of sodium silicate becomes more pronounced, silica becoming more predominantly the solid phase in equilibrium. Further, if the ratio of SiOz/NazO is greater than 1, the deposition of NazSiOa in the higher pressure stages will cause considerable increase in the SiOa/NazO ratio as the droplets come to the lower pressure stages of the turbine, thus favoring silica as the solid phase in equilibrium. Hence it is not surprising to find that silica is the predominant phase in deposits in the lower temperature and pressure range of turbines, appearing in the form of quartz in the range from about 575 to 350, and below that point as amorphous silica. These facts emphasize that effort must be directed to keep the boiler-water salts as highly water-soluble as possible, so that with minimum variation of steam conditions in the turbine, their full dissolution and removal will be accomplished. In this connection, the potassium salts, because of their higher saturation concentration, are advantageous over the sodium salts. Likewise. potassium silicate which is stable in a wide range of the ratio SiOz/KaO, and highly soluble, is tremendously advantageous over sodium silicate.

These advantages of potassium treatment accrue, whether the mechanism of transfer of boiler water salts from the boiler is mechanical or by vaporization. Further, salts with retrograde solubility curve ending in practically zero solubility at the critical temperature and pressure of water such as the sodium silicates are more liable to transport by vaporization than salts like the potassium silicate whose solubilities continuously increase with increasing temperatures until'their melting points are reached.

If deposition in the turbine is composed of highly water-soluble salts, the minor haphazard variations of temperature and pressure therein exercise a continual dissolving and washing action. If this action does not sufllce, and planned steam washing becomes advisable, it should be noted that it does not necessitate entirely taking away the load from the turbine and treating with saturated steam, but rather, dropping off the load enough so that the superheat does not exceed his of salts like sodium chloride or potassium.

chloride, whereupon they dissolve and the blades are cleansed.

Sodium aluminum silicate is a zeolitic material, which if brought in contact with water eontainlng potassium will exchange its sodium for potassium when the boiler-water is changed from the sodium to the potassium basis. Such exchange occurring in sodium aluminum silicate deposited as scale in a boiler gradually changes the scale to a potassium aluminum silicate which disappears from the surfaces leaving them clean. A lower ratio of K/Na will effect some exchange, but for practical purposes, the ratio of K/Na should be 3 or higher. To provide against the formation of silica scale it is essential that the boiler water at points where it may concentrate. as in steam-bound or highly evaporative sections, shall contain sufficient potassium in relation to sodium so that the characteristics of the potassium silicate system, and not those of the sodium silicate system, willdominate the equilibria with the silicates or silica. The same condition is essential to prevent the cementing of sludges in the boiler. The same condition is essential both in the superheater and the turbine.

Under all practical operating conditions, there will be some Na present in the boiler water. The exact ratio of K/Na which is most advantageous will depend upon conditions of boiler water composition and boiler operation. With boilers using evaporated water for make-up and having minimum condenser leakage, the ratio of K/Na may be 5 or 10 or even larger, depending on condenser leakage and evaporator carryover. However, where a sodium-containing natural water is used as make-up, the ratio of K/Na will, in general, be lower. A substantially larger proportion of Na can be tolerated particularly if the conditions are not too severe. Ratios of K/Na of from 3:1 to 4:1 thus may be advantageously employed, particularly where sodium-containing natural water is used as make-up. In any event, the K/Na ratio should be high enough to prevent the deposition of siliceous scale in the boiler and of insoluble silica in the turbine.

In case the minimum value 3 of K/Na is maintained in the boiler water, then the ratio SiO2/Z2O should not be allowed to approach the value of 2,

which is the advisable maximum even if K/Na is l3 very large, since above that value the possibility increases rapidly that quartz may be the solid phase incongruently in equilibrium with the solution. In fact, unless the minimum value 3 of K/Na is exceeded appreciably, SiOa/ZzO may well be maintained at a value of 1. or as low as 0.75 or even as low as 0.3, the latter values signifying free as well as captive alkalinity in the boiler water. It Sim/Z2021, the alkalinity is completely of captive type.

The substitution of K for Na in the boiler water, in conjunction with the establishment of satisfactory ratio sioz mo in appropriate relation to the other constituents therein as already set forth supplies protection from the embrittling action of caustic alkalinity on the boiler steel. Two cases require consideration.

Case 1. When SiO2/Z2O 1, the boiler water contains both free and captive alkalinity.

. boilers.

brittlement is the result of the action of stron caustic solution on stressed steel. If concentration of a boiler-water, as at any seam, results in strong caustic 'concentratiomthat water is embrittling in character. Just as described hereinbefore in relation to steam-bound areas of the boiler, concentration of the boiler water will continue to the point where its vapor pressure is equal to the superimposed pressure. If the boilerwater salts as, for example, sodium sulphate and sodium silicate, which are of retrograde solubility at the temperatures concerned, are of limited solubility, in comparison with the caustic, they will precipitate as concentration continues, and the caustic will concentrate highly before the vapor pressure of its solution arrives at equilibrium with the superimposed pressure. If there are present in the boiler water highly soluble salts which have a solubility increasing with temperature, for example, sodium or potassium chlorides or borates, the chloride and/or borate continue to concentrate along with the caustic. The vapor pressure of their solution will reach equilibrium with the superimposed pressure before the concentration of the caustic becomes high if the chloride or borate is of suflicient concentration in the boiler water in relation to the caustic. Such a. water is not embrittling.

Now, when the boiler water is put on a potassiumbasis, potassium silicate, being practically as soluble at boiler-water temperature as eitherpotassium or sodium hydroxide, will under all conditions concentrate evenly with the alkali, and

if maintained at sumcient concentration in the boiler water either by itself or in conjunction with a regulated concentration of potassium chloride or borate, thus prevent caustic concentration reaching a point where it will attack the steel.

According to Raoults law, the lowering of vapor pressure of any solution is proportional to the total number of molecules dissolved therein, irrespective of the species of the molecule. There fore, the vapor pressure of the film next to the boiler metal under a steam bubble is lowered in accordance with the total number of molecules of the soluble salts which include not only the hydroxide but also the potassium silicate, and potassium chloride if present, and potassium borate if present. There should be enough potassium silicate and, if necessary potassium chloride or borate, in the boiler water so that the vapor i 14 pressure is lowered. sufllciently so that at the temperature of the film it can not concentrate to the point where the caustic constituent reaches .a concentration sufficient to cause caustic embrittlement. An advantage of the use of the soluble potassium silicate for this purpose over the other soluble salts, such as the chlorides, is that the potassium silicates are useful in supplying alkalinity in the captive form and also in supplying silicate to cause the precipitation of the magnesium as a readily disper'sible magnesium silicate sludge. Potassium phosphate and potassium nitrate are, like the potassium silicate, chloride and borate, readily water-soluble salts which have a solubility increasing with temperature at temperatures encountered in steam In accordance with the Hall system as herein applied, potassium phosphate is introduced either as orthophosphate or as a moleculariy dehydrated phosphate which rehydrates in the boiler, water to orthophosphate and insures the precipitation of the calcium as a nonadherent calcium phosphate sludge. phosphate, which is ordinarily maintained in excess of that required to combine with the calcium. will be generally present in the water and; like the potassium silicate, will supply alkalinity in than 1, the potassium silicate is usually formed by the combination of the silica in the water with added potassium salts. The potassium salts added to the water should be of the class of readily water-soluble salts which increase in solubility with an increase in temperature and comprise potassium phosphate, potassium chloride, potassium nitrate, potassium borate, potassium silicate if it is desired to add more silicate. and potassium hydroxide if it is desired to increase the alkalinity. In boiler waters, which are kept upon the potassium equilibrium as herein described, there will be a plurality of such-salts of increasing solubility with temperature at the temperature of the boiler water and since they remain soluble as the water concentrates at a steam blanketed boiler surface, they will serve to lower the vapor pressure of the film next to the boiler metal in accordance with Raoults law and thus, as above explained, prevent a destructive concentration of caustic. I have. found that if a molar ratio of not more than 0.1 is maintained between (1) the concentration of free alkali and (2) the sum of the concentrations of the free alkali and of the water-soluble potassium salts which increase in solubility with increase in temperature at the temperature of the steam boiler water, caustic embrittlement is prevented or greatly minimized.

Case 2. When SiOz/ZzOzl, all boiler-water alkalinity is in captive form. As concentration occurs, the 'hydrolytic action of the phosphates and potassium silicate, and potassium borate if present, is reversed, and the alkali becomes again a part of the potassium silicate or phosphate or borate' molecule.

No damaging high causticity can result particularly in conjunction with a regulated concentration of potassium chloride, a condition which represents the most complete protection against any embrittling action. In this case, the potassiumsilicate is advantageous, both because of its large solubility, and because of its certain control over captive alkalinity. Similarly, potassium borate can be used to supply captive alkalinity which is available as alkalinity in the overall boiler water but which re-unites with the parent Potassium:

molecule and does not concentrate at surfaces where the boiler water may concentrate.

I, similarly, as discussed above under Cases 1 and 2, control the destructive corrosive action on the metal surfaces of bonded oxygen derived from the boiler water where it concentrates, as at points where steam development is most rapid, or where steam blanketing or film boiling occurs. A slight increase oftemperature of these surfaces even as small as 5 over that of the saturated steam concentrates a solo solution of caustic soda to a strength of nearly 5% as shown by the data of Kracek, International Critical Tables, vol. III, page 370. This concentration of caustic is sufiicient to dissolve protective iron-oxide film, whereupon the bared steel surface combines with bonded oxygen of the boiler water. Repetition of the process countless times results in general attrition of the surfaces, devastating gouging with channeling action where conditions are most severe, and accumulation of heavy magnetic-oxide sludge which, if sodium silicate is present to serve as a binder, may become cemented into a scale. By substituting potassium for sodium in the boiler water, I prevent the'binding action because of the ready solubility and the stability of potassium silicate; and by adjusting the concentration of the potassium silicate, or of potassium silicate and chloride, or of potassium silicate and borate, or of potassium silicate, chloride and borate, in the boiler water in relation to the alkalinity so that as concentration occurs to the extent demended by Raoults law, caustic cannot increase to a damaging concentration, I prevent or minimize the corrosive action. By so doing, I maintain more certain circulation in the boiler tubes by eliminating the resistance to circulation caused by the formation of the iron-oxide sludge on their walls. Control in this manner is of particular value to the boilers of forced circulation typeand flash type. As explained above under Case 1 in which the boiler water contains both free and captive alkalinity, the caustic can be prevented from destructive concentration in a water film next to the boiler metal under a steam blanket by the maintenance in the boiler water of water-soluble potassium salts which increase in solubility with temperature at the temperature of the steam boiler water. As stated above, such salts comprise potassium phosphate, potassium chloride, potassium silicate, potassium nitrate and potassium borate. I have found that if there is maintained in the boiler water a molar ratio of not more than 0.2 and preferably not more than 0.1 between (1) the concentration of the, free alkali therein and (2) the sum of the concentrations of the free alkali and of water-soluble potassium salts which increase in solubility with increase in temperature at the temperature of the steam boiler water, caustic corrosion is prevented or greatly minimized.

When phosphate is added to steam boiler Water in accordance with the Hall system, either in the form of orthophosphate, as described in the Hall Patent 1,613,656, or as molecularly dehydrated phosphate, as described in the Hall & J ackson Patent 1,903,041, both referred to above, the phosphate combines with the calcium to form a precipitate of nonadherent calcium phosphate sludge. If the silica content of the boiler water is insufficient, magnesium precipitates largely as magnesium hydroxide or phosphate. The magnesium hydroxide or phosphate sludge thus formed is not removed by blowdown and accumulates in the boiler as soft scale. If the magnesium is caused to precipitate as the more insoluble magnesium silicate by the maintenance of sufficient silica in the boiler water for this purpose, the sludge so formed remains in suspension in the boiler water, and is removed by blowdown. The maintenance of requisite silica for this purpose in a sodium boiler water is attended by the danger of silica scale in the boilers, and silica deposits in the turbines. Furthermore, sodium silicate may act as a binder, thus cementing the sludge together enough to resist removal by blowdown. By substituting potassium for sodium in the boiler water, Ican, as required, advantageously add sufiicient silica in convenient form to the boiler water to effect precipitation of magnesium as the sillcate, and thus overcome these difliculties.

With boilers operating on sodium boiler water, and with some silica in the water, the surfaces become coated generally with a thin film of silica which resists transmission of heat. Where steam blanketing or film boiling occurs, the siliceous deposits become thicker, and tubes burn out. With a potassium boiler water, not only do these siliceous deposits not form, but those already in the boiler are removed, since a boiler water having an alkalinity which is the result of hydrolytic action at an ratio of 1.6 or less, will dissolve silica readily enough, building the ratio to a little higher value. Thus, whereas a sodium boiler water containing some.silica produces deleterious silica deposits, the boiler water with a sufiicient ratio of K/Na not only does not produce such deposits, but dissolves them as one of its characteristic chemical properties. The very fact of the strong retrograde solubility of the sodium silicates in the boiler water at the higher temperature and their readiness to deposit silica as solid phase demonstrates conclusively that in a boiler water containing any silica, sodium and potassium cannot be considered chemically equivalent, since the potassium-silicate salts do not readily deposit silica as solid phase, are very soluble, and increase strongly in solubility with increasing temperature throughout and far beyond the boiler water temperature range.

With boilers operating on sodium boiler water, hide-out is a common occurrence especially in higher pressure boilers, and particularly at higher ratings of operation, and sodium phosphate, sodium sulphate, and sodium silicate, whose solubilities decrease with increasing temperature at the temperature of the'boiler water, readily evaporate to dryness at any areas of film boiling or steam blanketing even though this be mild, since as shown by the data in Table I, only very slight increase in temperature of the surface is necessary to effect their deposition as solid salts. Potassium phosphate and potassium silicate are salts which increase in solubility with temperature increase at the temperature of the boiler water. Potassium sulphate decreases in solubility with increasing temperature at the temperature of the boiler water, but its solubility expressed in mol fractions is about double that of sodium sulphate. Thus, with boilers operating on potassium boiler water, hide-out is prevented or minimized, because the potassium phosphate and silicate particularly, but the potassium sulphate as well, are not deposited as solid salt under mild conditions of film boiling or steam blanketing, but require very severe conditions to effect this, and besides, will dissolve very readily by feed water that the replacement oi sodium by" potassium will be most economically brought about. As an example, in a plant with some condenser leakage but, with all make-up water supplied by an evaporator, the use of ,potassium pyrophosphate, potassium hydroxide, and potassium sulphite, as chemicals supplied to the boiler water, in lieu of the disodium phosphate, sodium hydroxide, and sodium sulphite, theretofore used,

' changed the relationship of K and Nain the boiler water from P. P. M. of K and 360 P. P. M. of Na to 500 P. P. M. of K and 39 P. P. M. of Na. Thus the problem of producing a predominantly K boiler water is relatively simple in conjunction with an evaporator and evaporated make-up.

The same is true also of untreated make-up, if its content of sodium salts is low; otherwise, pre* treatment of the water in appropriate manner is essential.

If treatment of the natural water is by the base exchange process, whether with natural zeolite or the carbonaceous or resinous exchanger, regeneration of the exchanger with potassium salt, for example potassium chloride, will produce an eflluent water in which not only hardness is removed, but sodium is replaced by potassium.

The capacity of the base exchange softener between regenerations must be based on sodium plus hardness and not on hardness alone. Wherever a base exchange softener of any type. is regenerated with potassium salts in order to eflect practical replacement of sodium in the raw water by potassium, regeneration with the potassium salts shall be efiected as a rule not later than the time when the sodium content of the eflluent water is equivalent to the sodium content in the raw water, and preferably shortly before this time. In no event shall regeneration be delayed beyond the point at which the sodium content in ally repeating these determinations as changes may occur in the water to be treated, which would be reflected in the potassium and sodium content of the treated water. If desired, treatment of the water with acid either before or subsequent to the exchanger may be made apart of the process. Another variation is to run a carbonaceous or resinous exchanger on the hydrogen cycle with regeneration by acid, with neutralization of the efiiuent water by a potassium salt such as potassium hydroxide or carbonate. Another variation is a parallel arrangement of base exchange softeners, one on acid regeneration, and the other on regeneration with potassium salt, with mixing of the two waters efiluent from the softeners to provide a neutral potassium water.

Another case is that of passing the water through an acid-regenerated cation exchanger, and then next of passing it through an acid absorber, the efiluent water being devoid at practically all dissolved salts, but largely retaining its original silica. The silica is'most undesirable for sodium-treated boiler waters, and the suggestion has been made that it be removed by a silicaguests or without sludge blanket, then the addition of potassium in place of a sodium coagulant, for instance, potassium aluminate, potassium alum or potassium bentonite in lieu of the corresponding sodium salts builds the ratio of potassium to sodium and helps in the maintenance of the requisite K/Na ratio in the boiler water.

' In the use of the precipitation method of softening water as customarily performed by the addition of lime and soda ash so proportioned as to remove both temporary and permanent hardness of the water, the substitution of potassium carbonate for sodium carbonate (or potassium hydroxide for sodium hydroxide) increases the potassium in proportion to the amount of hardness in the water requiring precipitation by the carbonate radical and by the amount of excess potash used, and may in itself be suiiicient to provide the requisite ratio of K/Na in the boiler water. In the phosphate softeners, either by themselves or in conjunction with the lime-potash softener just described, the .substitution of a potassium phosphate for sodium phosphate will likewise build the potassium content of the efliuent water, or in the lime-potash softener an.excess of potash may be supplied and the following phosphate softener may then use phosphoric acid, producing the potassium phosphate as the process proceeds. I These procedures are useful steps in the substitution of potassium for sodium, in order to provide potassium boiler water in place of sodium boiler water. They represent a new departure in the treatment of water for boiler-feedwater purposes. Because sodium and potassium are not 45 chemically equivalent in the boiler water, neither are they thus equivalent in the feedwater entering the boiler, since a potassium boiler water cannot result without consideration of the relations of potassium to sodium in the feedwater. 5 The establishment of a potassium boiler water water both in preparation of. the feedwater, and in maintaining conditions in the boiler water, shall not be supplied indiscriminately as sodium or as potassium salts, whichever is the cheaper or more conveniently secured, but shall be supplied as the potassium salt regardless of the expense of maintaining the requisite ratio of K/Na in the boiler water. Following the primary treatment or in cases where primary treatment is not employed, and depending on what ratio of K/Na is desired in the boiler water, the chemicals introduced into the boiler water may be potassium orthophosphate, or amolecularly dehydrated potassium phosphate such as pyrophosphate or metaphosphate, potassium phosphate glasses, potassium hydroxide or carbonate, potassium sulphite for removing oxygen, potassium nitrate if this further step is desired to be taken for increased protection from embrittlement, this complete substitution of potassium for sodium salt in the boiler-water tree tme'nt giving an excellently high ratio of K/Na in the boiler water. If the desired amount of silica is not provided to the boiler water from the natural requires that chemicals used for treatment of the water, then silica in satisfactory form may be added along with the other treatment in requisite amounts. For example, the silica may be supplied in the form of a potassium silicate or as a mixture of silicate with phosphate in the form of a phosphate glass. If desired, the potassium may be supplied to the boiler water in the form of boiler compounds, which contain potassium salts, and with or without organic material such as dispersive tannins, lignins and alginates.

To control conditions in the boiler water so that the desired relation of K/Na and Sw /K t) shall be maintained, appropriate tests may be made on samples of boiler water withdrawn from the circulating boiler water with appropriate precautions to assure that the sample will be representative. For best precision, the concentration of dissolved silica may be determined by filtering the boiler-water sample and proceeding according to established methods, as recommended in the Standard Methods of Analysis of the American Public Health Association. With suflicient accuracy, the determination may be made by one of the well-known colorimetric methods, due care being taken to avoid interference by the phosphate radical.

For the determination of K/Na ratio it is sufficient that either K or Na be determined, in conjunction with the tests made on the boiler water ior total alkalinity, phosphate, sulphate, chloride, nitrate, and borate if used as a treating chemical. The potassium concentration may be readily determined by evaporating 2-10 ml. of the filtered sample of boiler water to dryness. adding lithium dipicrylaminate reagent, allowing the sample to stand for two hours at room temperature, and then measuring the transmittance of the yellow-red color of the supernatant liquid with a photoelectric colorimeter. Or, the filtered sample of boiler water may be treated with the reagent magnesium uranyl acetate, thus precipitating the sodium as the triple acetate, filtering off the precipitate, and weighing it. With all results of determinations expressed in equivalents per million, the sum of the equivalents of potassium and sodium is for all practical purposes equal to the sum of the negative radicals in equivalents. Hence, if potassium is determined, then the difierence between the sum of the negative radicals in equivalents and the value of potassium in equivalents gives sodium in equivalents; or if the sodium is determined, then the difierence will give potassium in equivalents.

In the plants in which make-up water is evaporated, and condenser leakage is not excessive, the supplying of chemicals for treatment oi the boiler water in the form of potassium salt will assure a ratio of K to Na above the minimum limit herein recommended. But, in any other case, determination of the ratio as routine procedure is advisable, and particularly essential if any mixture or K and Na salts is used for treatment, or if the Na in the natural water is not largely replaced by K.

While I have specifically described my invention and its preferred embodiment, it is to be understood that the invention is not so limited but may be otherwise embodied and practiced metal compounds concentrating in the boiler water would otherwise be predominantly those of sodium which comprises supplying to the boiler water at least one water-soluble potassium salt which increases in solubility with increase in temperature at the temperature of the steam boiler water in amount suflicient to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and a molar ratio of not more than 0.1 between the concentration of free alkali which may be present therein and the sum of the concentrations of such free alkali and of the water-soluble potassium salt or salts which increase in solubility with increase in temperature at the temperature of the steam boiler water.

2. The process of treating water for steam boilers for steam generation in which alkalimetal compounds concentrating in the boiler water would otherwise be predominantly those of sodium which comprises supplying to the water at least one potassium compound of the class consisting of potassium phosphate, potassium chloride, potassium silicate, potassium nitrate, potassium borate and potassium hydroxide in an amount suflicient to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and a molar ratio of not more than 0.2 between the concentration of tree alkali which may be present therein and the sum 01' the concentrations of such tree alkali and of the water-soluble potassium salts which increase in solubility with increase in temperature at the temperatureof the steam boiler water.

3. The process of treating silica-containing water for steam boilers for steam generation in which the alkali-metal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to the water a compound of potassium which increases in solubility with increase in temperature at.the temperature of the steam boiler water in an amount sufilcient to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and an alka line potassium reagent in an amount suiiicient to maintain in such water a ratio of silica to the alkalinity associated with the silica in equivalents between 0.3 and 2.

4. The process of treating water for steam boilers for steam generation in which the alkalimetal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to the water a compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water in an amount sufllcient to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and a compound of silica in an amount sufllcient to maintain in such water a ratio of silica to the alkalinity associated with the silica in equivalents between 0.3 and 2.

5. The process of treating water for steam. boilers for steam generation in which the alkalimetal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to the water a compound oi potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water in an amount suiiiclent to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and a silica-alkalinity adjusting reagent in an amount sufiicient to maintainin such water a ratio or silica to the alkalinity associated with the silica in equiva-- lents between 0.3 and 2.- I

6. The process of treating water for steam boilers for steamgeneration in which the alkalimetal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to the water a compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water in an amount sufficient to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and a silica-alkalinity adjusting reagent in an amount suflicient to maintain in such water a ratio of silica to the alkalinity associated with. the silica'in equivalents between 0.75 and 2.

7. The process of treating steam boiler water for steam generation, which comprises supplying to the boiler water at least one potassium salt of. the class consisting of potassium phosphate, potassium chloride, potassium silicate, potassium nitrate and potassium borate in an amount suflicient to maintain a molar ratio of not more than 0.2 between the concentration of free alkali-which may be present therein and the sum of the concentrations of such free alkali and of said potassium salts. v

8. The process of treating water for steam boilers for steam generation in which alkalimetal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to the water a compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water in an amount sufiicient to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and a ratio of silica to alkalinity associated with the silica in the water in the boiler in equivalents'of less than 2.

" which alkali-metal compounds concentrating in the water a molecularly dehydrated potassium p the water a silicate-containing potassium phos- 9. The process of treating water for steam boilers for steam generation in which alkalimetal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to the water a compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water in an amount sufficient to 'maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 5 and a ratio of silica to alkalinity associated with the silica in the water in the boiler in equivalents of lessthan 2.

10. The process of treating calcium-containing water for steam boilers for steam generation in which alkali-metal compounds concentrating in the boiling water would otherwise be predominantly those of sodium, which comprises supplying to the water a phosphate of potassium which increases in solubility with increase in temperature at the temperature of the steam boilerwater in an amount sullicient to insure the precipitation of calcium as non-adherent calcium phosphate and another water soluble compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water to maintain in the water in the boiler a ratio of total potassium to sodium in equivalents of at least 3 and a ratio of silica to alkalinity associated with the silica in the water in the boiler in equivalents of less than 2.

11. The process of treating calcium-containing water for steam boilers for steam generationin phosphate which increases in solubility with increase in temperature at the temperature of the steam boiler water in an amount suflicient to insure the precipitation of calcium as non-adherent calcium phosphate and another water soluble compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water to maintain in the water in the boiler a ratio of total potassium to sodium in equivalents of at least 3 and a.

ratio of silica to alkalinity associated with the silica in the'water in the boiler in equivalents of less than 2.

12. The process of treating calcium-containingwater for steam boilers for steam generation in which alkali-metal compounds concentrating in the boiler water would otherwise be predominantly those-of sodium, which comprises supplying to the water phosphate, silicate and potassium which increase in solubility with increase intemperature at the temperature of the steam boiler water so proportioned as to insure the precipitation of calcium as non-adherent calcium phosphate and a ratio of potassium to sodium in the Water in the boiler in. equivalents of at least 3 and a ratio of silica to alkalinity associated with the silica in the water in the boiler in equivalents of between 0.3 and 2.

13. The process of treating calcium-containing water for steam boilers for steam generation in which alkali-metal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to phate glass which increases in solubility with increase in temperature at the temperature of the steam boiler water and another water soluble compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boilerwater so proportiOned as to insure the precipitation of calcium as non-adherent calcium phosphate and a ratio of potassium to' sodium in the water in the boiler in equivalents of at least 8 and a ratio of silica to alkalinity associated with the silica in the water in the boiler in equivalents of between 0.3 and 2.

14.. The process oftreating water-for steam boilers for steam generation in which alkalimetal compounds concentrating in the boiler wasodium, which comprises supplying to the boiler water atleast one water soluble potassium compound which increases in solubility with increase in temperature at the temperature of the steam boiler water in amount, sufiicient to maintain in the water in the boiler a ratio of potassium to sodium in equivalents of at least 3 and a molar ratio of not more than 0.1 between the concentration of free alkali which may be present therein f generated cation exchanger and an acid absorber whereby the water is substantially freed from mineral salts with the exception of silica, adding to the water so treated a compound of potassium which increases in solubility with increase in temperature at the temperature of the steam boiler water in such proportion as to maintain the silica in the form of soluble potassium silicate in the water in the boiler and supplying the water so treated to a steam boiler.

16. The process of treating water for steam boilers for steam generation in which alkalimetal compounds concentrating in the boiler water would otherwise be predominantly those of sodium, which comprises supplying to the Water at least one potassium compound of the class consisting of potassium phosphate, potassium chloride, potassium silicate, potassium nitrate, potassium hereto, and potassium hydroxide in an amount sufiicient to maintain in the water in the boiler a. ratio of potassium to sodium in equivalents of at least 3 and a molar ratio of not more than 0.2 between the concentration or free alkali which may be present therein and the sum of the concentrations of such free alkali and of the water soluble potassium salts which increase in solubility with increase in temperature at the temperature of the steam boiler water and a ratio 24 of silica to alkalinity associated with the silica in the water in the boiler in equivalents of less than 2.

RALPH E. HALL.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS OTHER REFERENCES "The Chemistry of Water and Sewage Treatment, by Arthur M. Buswell, p s. 74 and 75.