Substitutes for Soda Ash in the Treatment of Boiler Feed Water W. A. TANZOL4, R. L. REED, AND J. J. MAGUIRE W. H. & L. D. Betz, Philadelphia Downloaded by CARGILL INC at 13:59:22:740 on June 18, 2019 from https://pubs.acs.org/doi/10.1021/ie50455a009. The current shortage of soda ash is a matter of particuin the treatment lar concern of boiler feed water where lime-soda softening is used to treat make-up water. Various substitutes for soda ash have been worked out on an individual plant basis. Caustic soda can be used in place of lime and soda ash where the carbonate and noncarbonate hardness of the raw water are in the proper proportions. In some cases alteration of the internal chemical treatof the boiler water permits lime softening without ment soda ash, and in others recirculation of boiler water to hotprocess lime-soda softeners is effective. Substitution of zeolite mineral in a filter bed can serve temporarily bv blending the zeolite effluent with lime-softened water. Barium carbonate has been used as a satisfactory substitute for soda ash although the higher cost with this chemical restricts its use to emergency periods. complete industrial plant is dependent on the power or steam generation which, in turn, is directly dependent on the proper softening of the boiler feed water, increased chemical treatment costs of ten, fifty, or a hundred dollars per day are readily justified if continued plant operation can be assured. During the past few years different systems of altered chemical balances and altered chemical treatment have been devised to permit continued and uninterrupted plant operation. Various substitution products and methods have been given consideration either to reduce the quantity of soda ash used or to eliminate it entirely. Some of the plants involved were advised by their suppliers that they would be allotted only about 70% of previous yearly purchases; other plants were faced with early depletion of their supply of soda ash without being able to renew it from The following examples illustrate methods legitimate sources. that have been applied to a number of plants faced with this shortage. a ' nationwide chemical shortage, particularly of caustic soda and soda ash, has resulted in production difficulties in THE many industries. Soda ash is one of the most important chemicals used in the treatment of boiler feed water, and the production of power and steam is dependent on the properly controlled treatment of this feed water. Without properly treated boiler water, unscheduled boiler outages from scale and corrosion cause interruptions in plant operation and shutdowns. One of the most commonly used methods for the external softening of boiler feed water is the lime-soda process, operated both hot and cold. Hydrated lime and soda ash are employed to precipitate the scale-forming calcium and magnesium ions from the raw water; these form insoluble precipitates of calcium carbonate and magnesium hydroxide which are separated from the softened water by sedimentation and filtration. Lime is required for the precipitation of magnesium and the removal of the calcium bicarbonate hardness as follows: + Ca(OH)2 + Mg(OH)2 2CaC03 + Mg(OH)2 + 2H,0 Mg(HC03)2 + 2Ca(OH)2 2CaC03 + 2H20 Ca(HC03)2 + Ca(OH)2 MgSCh = CaS04 = = CAUSTIC SODA While most of the sodium salts were scarce, some plants had available caustic soda which was utilized in place of soda ash (4). Actually, caustic soda is utilized in place of lime and as a result of this substitution, soda ash is formed; Equations 5, 6, and 7 illustrate the softening reactions using caustic soda. Ca(HC03)2 + 2XaOH CaC03 + Xa2C03 + 2H.0 2XaOH Mg(OH), + Xa2S04 MgS04 + Mg(HC03)2 + 4XaOH Mg(OH)2 + 2Xa2C03 + 2H20 = = CaC03 + Xa2S04 (5) (6) (7) Equations 5 and 7 show that soda ash is produced when caustic soda reacts with calcium or magnesium bicarbonate hardness. The soda ash thus becomes available for softening calcium sulfate hardness (Equation 4). Depending on the characteristics of the raw water, complete softening may be brought about by caustic soda alone or by a combination of lime and caustic soda. on waters that consist predominantly of noncarbonate hardness caustic soda can accomplish only partial softening. In general, caustic soda alone can be utilized for softening waters that possess a methyl orange alkalinity that exceeds half of the calcium content by 15 to 30 p.p.m. (Table I). Where methyl orange alkalinity exceeds half the calcium content of the water by more than 30 p.p.m., a combination of lime and caustic soda should be utilized to avoid an excessive alkalinity of the softened Waters that possess a methyl orange alkalinity less than water. half the calcium content can be only partially softened by caustic soda alone. However, (1) (2) (3) remove CaS04 + Xa2C03 = = calcium noncarbonate hardness present in the raw water and also to precipitate the calcium chloride, calcium sulfate, etc., produced in the precipitation of magnesium salts by lime: Soda ash is needed to 24, Pa. (4) When soda ash is unavailable for lime-soda softening, the effluent from the treatment process may be quite high in hardness. Use of unsoftened or partially softened make-up water would be disastrous in many cases of boiler operation under high pressure, despite increased application of internal treatment of boiler To maintain boilers on line during temporary interwater. ruptions in soda ash supply or under continuously curtailed supply conditions constitutes a problem often capable of solution by altering chemical balances throughout the system or by substituting treatment chemicals not usually considered feasible from an economic viewpoint. Where the continued operation of BOILER WATER RECIRCULATION In many cases soda ash requirements have been partially reduced by recirculating boiler water blowdown to the softener, and thus making use of the caustic soda and soda ash content of the boiler water to replace a portion of the lime and soda ash normally used in the softening operation {1, 2, S, 7, 9, 10, 12). Softening reactions follow Equations 4, 5, 6, and 7. Since the major portion of the boiler water alkalinity exists as caustic soda, the additional soda ash produced in Equations 5 and 7 is of bcne1440 INDUSTRIAL November 1947 Table I. Applicability Raw Water Characteristics' . VjCa + (15-30 p.p.m. 3 XI unethyl orange Table II. Hardness CaCOj, as as 180 120 60 20 0 32 60 3d 10 Complete 0 16 4 Boiler Water With Recirculation^ Lime-soda Raw softened Boiler 180 120 60 20 0 120 35 10 32 60 0 16 4 CaCOj, p.p.m. Phenolphthalein Methyl Softening Partial Complete CaCOj. Without Recirculation3 Lime-soda Raw softened Boiler p.p.m. Sulfate as Analyses with and without Recirculation Ca as CaCOj, p.p.m. XIg as CaCOj, p.p.m. Alkalinity alkalinity) and Ca orange as SO4 Chloride 120 3o 10 630 700 o38 154 60 CHEMISTRY 336 374 538 154 60 as Cl 19 Phosphate as PO* 7.3 10.1 11.7 7.3 10.1 11.5 pH a Lime requirement, 147 p.p.m. on raw water; soda ash, 106 p.p.m. on raw water. 6.5 %> boiler water blowdown (69,500 pounds per million pounds BARIUM CARBONATE Barium carbonate was considered an emergency substitute for soda ash. Lime-barium softening for reducing the sulfate concentration is a process of historical interest but is in use, to the authors’ knowledge, at only one plant in this country (-5, 6, 8, It). Barium carbonate was applied during the soda ash shortage, not with the primary intention of reducing the sulfate content of the treated water, but rather to obtain the advantage of the sod urn carbonate produced as a by-product of sulfate removal: + BaCO, CaSO, + BaCOs MgSO, + BaCO, MgCOs + Ca(OH)2 Xa2S04 •steam). Lime requirement, 132 p.p.m, on raw water; soda ash, $4 p.p.m. on water. raw 6.5% system blowdown (139,000 pounds per million pounds steam, of which 69,500 pounds are returned to softener). b tit in the removal of noncarbonate hardness as illustrated by Equation 4. Table II shows the normal balances maintained on a hot process softener and the resulting boiler water characteristics, as well as the balances utilizing boiler water recirculation to the softener to bring about a reduction in soda ash requirements. As a result of recirculation, the reduction in lime requirements was approximately 10% and in soda ash requirements, 21%. A normal balance was maintained on the softener, and all factors of the boiler water balance remained the same with the exception of boiler water alkalinity. The reduction of boiler water alkalinity is a direct result of consuming it in the softening process. In many cases reduction in boiler water alkalinity is a desirable feature to minimize carry-over that may result from excessive boiler water alkalinity; boiler water is often recirculated solely for this purpose. In general, the application of boiler water recirculation to waters that, consist predominantly of carbonate hardness brings about a greater percentage saving in soda ash than is covered by the case cited. Noncarbonate hardness waters involve only a minor percentage savings in the soda ash requirements. ZEOLITE = = CaZ + 2XaHCO MgZ + Xa2S04 = = BaSO, + Na.CO, BaSO, + CaCO, BaSO, + CaSO, (10) CaCO, + Mg(OH)2 (13) (11) (12) Blending of Zeolite Softened Water Prior Lime-Soda Softening (in Parts per Million) Table III. to ZeoliteSoftened Raw? Hardness as CaCOj Ca as CaCOj Xlg as CaCOj Alkalinity CaCOj Phenolphthalein Blended b Lime-Soda Softened 121 20 84 37 16 4 0 160 110 50 4 0 130 30 0 130 30 130 30 32 60 30 12 12 12 12 as Xlethyl orange Sulfate as SO< Chloride as Cl Chemical requirements: lime, 149 p.p.m.; soda ash, 68 p.p.m. b 75% raw water, 25% zeolite-softened. Chemical requirements: 137 p.p.m.; soda ash, 27 p.p.m.; salt, 140 p.p.m. a Table IV. Blending of Lime-Soda Effluent Raw plants faced with the need for reducing soda ash requirements had enough filters and filtering capacity to permit conversion of one of the filters to a zeolite softening unit, operated In the zeolite softening process calcium on the sodium cycle. and magnesium hardness is removed in exchange reactions with sodium ions substituted for the calcium and magnesium: = = As reaction 10 shows, sodium carbonate is formed by the reaction between barium carbonate and sodium sulfate present in the raw water. The sodium carbonate is then available for use in accordance with reaction 4. In addition, as reactions 11, 12, and 13 show, calcium sulfate and magnesium sulfate can be removed without the use of soda ash. SOFTENING A few of the Ca(HCO,)$ + Na»Z MgSOi + Xa2Z 1441 In some of the plants employing batch cold-process lime-soda softeners with gravity filters, consideration was given to converting one of the gravity filters to a gravity zeolite softener and blending the effluents of both softening processes to produce a water as illustrated in Table IV. The quantity of soda ash saved by this procedure ivas in proportion to the amount of water softened by the zeolite unit. Since only 75% of the make-up water was softened by the lime-soda process, a direct saving of 25% in soda ash was achieved. Degree of Caustic soda Caustic soda Caustic soda and lime M > VsCa 4- 30 p.p.m. ENGINEERING Caustic Soda of Treatment < i/2Ca = AND Hardness as CaCO, Ca as CaCOj Xlg as CaCOj Alkalinity as CaCOj Phenolphthalein Xlethyl orange Zeolite Softened Water (in Parts per Million) Lime-Soda Softened 220 160 60 35 25 0 45 80 100 Zeolite Softened lime, with Blended'1 4 28 0 100 35 85 80 30 10 80 80 80 Sulfate as SO4 30 30 30 Chloride as Cl 3 75% cold lime-soda softened, 25% zeolite-softened. 8; (9; Reaction 8 shows that the bicarbonate hardness is converted to sodium bicarbonate. When the zeolite softened water is blended with the raw water, the resulting characteristics are such that the hardness becomes totally carbonate. Table III shows the make-up of the blended water and the resulting softened water, which brought about a saving of 8% in lime and 60% in soda ash. At the same time the external chemical treatment cost was slightly lower with this treatment than was obtained by the normal procedure. By softening increased percentages of the raw water by the zeolite process, the use of soda ash could be totally eliminated. where the natural sodium sulfate content of the insufficient to produce the desired sodium carbonate excess required in the softener effluent, it· was necessary to feed sodium sulfate to the softener and thereby produce soda ash directly in accordance with reaction 10. Theoretical efficiency in the use of barium carbonate to produce soda ash has not been obtained in either laboratory test or full scale plant operation. Barium carbonate can be used in both hot· and cold softening, but its action is more efficient in the hot (Table V). Table V illustrates hot process tests in which normal softening action is obtained with theoretical quantities of lime and soda In raw some cases water was INDUSTRIAL 1442 Table V. Results of Raw AND ENGINEERING Laboratory Tests on Hot with Barium Carbonate and ments 100 Barium carbonate Sodium sulfate Amounts used, p.p.m, 172 Soda ash Barium carbonate Sodium sulfate Analysis, p.p.m. Hardness as CaCOt Ca as CaCOt Mg as CaCOt, p.p.m. Alkalinity as CaCOt Phenolphthalein Methylorange Sulfate as SO4 Chloride as Cl 0 Conditions of test: Conditions of test: 100 100 172 200 100 100 100 100 100 100 150 150 150 ... 100 100 100 172 172 200 24 21 172 172 172 370 223 370 334 555 334 38 23 15 248 168 80 170 172 4 3 26 34 30 164 46 64 30 ... 248 168 176 174 16 14 68 80 2 2 13 52 30 22 0 6 16 16 6 12 12 0 38 128 164 44 96 164 100 30 164 100 30 30 40 30 24 80 164 164 164 164 164 1-hour retention, 20( 0 F. 1.0-liter samples. 4-hour retention, 80° F., 1.0-liter samples. Lime alone is effective in precipitation of magnesium, but obviously additional softening power is required to reduce the calcium content. Using theoretical quantities of barium carbonate and sodium sulfate as substitutes for soda ash yields reasonably satisfactory results, with reduction in calcium to 55 p.p.m. as calcium carbonate. Further improvement in lowering the hardness of the treated water is obtained with additional barium carbonate and sodium sulfate. Table V also shows the results of cold process tests with the same raw water. Application of theoretical quantities of barium carbonate and sodium sulfate does not achieve the same degree of hardness removal as secured at higher temperatures. Increased quantities of barium carbonate and sodium sulfate show steadily decreased calcium contents of the softened water, but these data indicate that this process 15 considerably less efficient at lower temperatures, possibly as a result of lower barium carbonate solubility at lower temperatures. The insoluble barium sulfate precipitate formed in these reactions is removed along with the calcium carbonate and magnesium hydroxide through sedimentation and filtration as in normal lime-soda softener operation. Inasmuch as barium salts are toxic, qualitative barium determinations were conducted on the softener effluent, and the efficiency of filtration was also checked closely to make oertain> of the removal of the finely divided barium sulfate precipitate. Only small traces of barium ash. were The Cold Processes 11 for this condition reason has not been closely investigated but was possibly due to insoluble barium sulfate coating on the barium carbonate particles. In addition, 100 100 100 100 100 100 slightly larger lime charges Í0Ó 10Ó 150 2ÓÓ 250 300 were required, of the order 100 150 150 200 250 300 of approximately 10%. In 172 172 172 172 172 172 general, reasonably good agree370 370 740 925 1110 ment is observed between labo223 334 334 466 670 ratory and field tests on the 112 118 66 96 60 56 hot process use of barium car105 101 42 33 31 17 19 24 25 27 bonate. 26 26 26 36 48 54 The use of barium carbonate 38 40 52 36 74 92 in place of soda ash does not 112 176 152 192 248 296 164 164 164 164 164 164 require alteration of normal softener control procedures, so no disadvantage has been encountered in this phase of operation. Lime control can be established on the basis of maintaining certain residual hydrate alkalinity values, and barium carbonate.can be regulated to develop optimum excess carbonate alkalinity concentrations. Determinations for excess sodium sulfate would not be imperative since experience proved that the feed of this material, if required, could be adjusted in direct ratio to the barium carbonate charges. Some state health authorities consider the barium-containing sludge to be a health hazard and will not permit it to be dumped in waterways. Potential users of barium carbonate process who must discharge sludge to streams and rivers should seek information concerning the attitude of their state health authorities before installing the process. Treated waters % of theoretical requireSoda ash Vol. 39, No. an Raw Treated waters CHEMISTRY Table VI. Results of Field Tests with (in Parts per Million) Raw Hardness as CaCOt Ca as CaCOt Mg as CaCOt Alkalinity as CaCOt Phenolphthalein Methyl orange Sulfate as SO* Chloride as Cl a Theoretical b Theoretical c 10% excess d 10% excess Water No. Ia 188 122 66 50 40 10 0 26 40 24 8 48 152 8 lime and barium lime; 50% excess lime; 25% excess lime; 40% excess -Soft* No. 2» 38 28 the Hot Process Efflu No. 3C No. 4d mer No. 5d 22 16 14 2 14 12 2 34 32 26 40 28 30 50 24 30 52 36 8 8 8 8 10 16 carbonate; no sodium sulfate, barium carbonate; no sodium sulfate. barium carbonate and sodium sulfate. barium carbonate and sodium sulfate. observed. During initial application of barium carbonate to plant operation, the raw make-up water to the softener was depended upon to supply the amount of sulfate required for barium precipitation. Sulfate removal taking place during this process decreases the dissolved solids of the softener effluent, which is beneficial in numerous applications. Further analyses revealed that 100% of the natural sulfate content of the raw water did not react with the barium carbonate, but rather a sulfate residual of 20-40 p.p.m. as SO, remained in the softener effluent. Consequently, not enough carbonate w-as produced to yield as complete noncarbonate hardness removal as desired. Barium carbonate fed considerably in excess of theoretical requirements failed to produce further sulfate or hardness reduction. Excess barium carbonate feed without the sulfate available for precipitation was simply wasted, since this material remained in an insoluble, inert form. Table VI illustrates the results obtained in field tests with barium carbonate in conjunction with a hot process lime-soda softener, having a capacity of 10,000 gallons per hour and operating at 218° F. Somewhat greater than theoretical quantities of barium carbonate were required, varying from 25-40%. Assuming reaction 10 to proceed at 100% theoretical efficiency, the cost of the soda ash thus produced is 6.2 cents per pound. This cost is based on 3.25 cents per pound for barium carbonate of 98% purity and sufficient natural sulfate content of the raw water to complete the reaction. Where sodium sulfate must be supplied for the reaction, based on a cost of 2.1 cents per pound for this chemical, the cost of thus producing soda ash is 9.0 cents per pound. Where higher than theoretical quantities of barium carbonate and sodium sulfate are necessary, the cost of producing soda ash is correspondingly increased. All applications proved that barium carbonate gave as great a hardness reduction as possible with the conventional use of soda ash. Obviously, operating chemical costs increased as a result of the cost of barium carbonate in comparison with soda ash and also the necessity of employing sodium sulfate in a large number of cases. However, the use of barium carbonate enables normal softening operations to be maintained during emergency periods when the soda ash supply may be inadequate, and thereby permits continuous operation without fear of interrupted pro- duction. INDUSTRIAL November 1947 AND ENGINEERING ACKNOWLEDGMENT The authors wish to thank J. W. Polsky, who conducted the majority of the laboratory tests, and also express appreciation to W. H. & L. D. Betz for permission to present these data. LITERATURE CITED (1) Baker, Combustion, 12, 31-4 (1940). (2) Betz, Handbook of Industrial Water 547 (1945). (3) Flickinger, Power, 85, 115-20 (1941). Conditioning, Chap. IS, p. (4) (5) (6) (7) (8) (9) (10) (11) (12) CHEMISTRY 1443 Heiskell, Ibid., 90, 747-9 (1946). Hundesshagen, Z. offentl. Chem., 24, 159-67, 175-86 (1918). Leick, Vom Wasser, 7, 197-205 (1933). Maguire and Tomlinson, Combustion, 11, 26-32 (1939). Mehring, Chem. & Met. Eng., 21, 629-32 (1919). Paris. Chaleur & industrie, 5, 277-9 (1924). Paris, Chimie & industrie, 4, 722-30 (1920). Rodman, Chem. & Met. Eng., 35, 221-3 (1928). Sperry, Combustion, 10, 27-33 (1939). Received September 6, 1947. Presented before the Division of Water, Sewage, and Sanitation Chemistry at the 112th Meeting of the American Chemical Society, New York, X, Y. Butanol-Acetone Fermentation of Wood Sugar REID H. LEONARD AND W. H. PETERSON GEORGE J. RITTER University of Wisconsin, Madison, Wis. United States Forest Products Laboratory, Madison, Wis. Wood hydrolyzates were fermented with Clostridium The wood butylicum No. 39 to butanol and acetone. species and method of hydrolysis affect the fermentability of the liquors. Very mild or very vigorous conditions of by the Madison wood sugar process as described by Harris and Beglinger (4). These samples were neutralized to pH 4.2 with lime at 138° C. (4, 7). The oak sample represented the first 25% of the hydrolyzate received from the digester, and the Douglas fir liquor was from a normal run. hydrolysis do not produce an easily fermentable solution. Complete utilization of sugar could be obtained up to 3% concentrations. Solvent yields ranged from 24 to 38% of the sugar fermented. fermentation of wood hydrolyzates by butanol-acetone is difficult. Sjolander, Langlykke, and Peterson (10) conducted butanol-acetone fermentations on hydrolyzates prepared by a method similar to the Scholler process and obtained a fermentable medium after precipitation of metals at pH 10, neutralization, and clarification with Norite decolorizing carbon. In the present paper other types of hydrolyzates have been studied, and attempts have been made to simplify the pretreatments for fermentation. Two cultures had previously been selected for wood sugar fermentations: Cl. felsineum Carbone No. 41 by Sjolander et al. (10) for hydrolyzates and Cl. butylicum (Fitz strain) No. 39 by Wiley et al. (11) for sulfite waste liquor fermentation. These two cultures, as well as Cl. butylicum No. 37, Cl. beijerinckii No. 67, and Cl. butylicum No. 69, were compared on wood hydrolyzate, and No. 39 was selected as the most suitable organism. Fermentations were conducted with cultures transferred three to five times from the spore stock. Nutrients were supplied by 1% malt sprouts and 0.1% (NHtRHPO». A trace of reduced iron was added to the media before autoclaving. From 0.1 to 0.3% calcium carbonate was added to the media after inoculation. Inoculum was produced on glucose-malt sprouts medium and used after 12 to 20 hours at 8% of the fermentation volume. Determination of reducing sugars was made by the method of Shaffer and Somogyi (9); furfural by a colorimetric method (1); ethanol and butanol by Johnson’s procedure (6); and acetone by Goodwin’s method (3). Volatile acids were determined by titration of 11 volumes of distillate from 1 volume of sample. Hydrolyzates of maple and spruce were prepared in a rotary digester similar to that described by Plow et al. (8). Maple sawdust was hydrolyzed by 3% sulfuric acid with an acid-wood ratio of 1:1 at 181° C. for 30 minutes. A milder hydrolysis of maple and of spruce was made with 1.8% acid, ratio 1:1 at 173° C. for 5 minutes. Oak and Douglas fir were hydrolyzed THE bacteria FACTORS AFFECTING FERMENTATION It was first desirable to repeat the results of Sjolander et al. In their work the amount of decolorizing carbon was not stated, and it was found that with maple hydrolyzate, following their procedure, 10 to 20 grams of Norite decolorizing carbon per 100 ml. were required to duplicate their fermentation results. Fermentation of high temperature maple hydrolyzate prepared in this manner showed that 92% of the sugar in a 4.04 grams per 100 ml. solution was fermented in 5 days. The quantity of decolorizing carbon was found to be important; the use of 1 gram per 100 ml. resulted in the fermentation of 25% of the sugar, 5 grams gave 64%, 10 grams gave 72%, and 20 grams gave 93%. After the furfural was removed from the sample, the quantity of carbon required was decreased to less than 5 grams per 100 ml. The pH was found to be important, since 1 gram of carbon at pH 2.0 gave 17% fermentation, while 1 gram at pH 6.8 gave 51%. The difficulty found with maple wood hydrolyzates was also found with spruce, Douglas fir, and oak. The fermentations were characterized by a long induction period and a slow sugar utilization. In some samples much of the inhibition could be accounted for by the presence of furfural—for example, the ° maple hydrolyzate made at 181 C. contained from 0.5 to 0.8 gram of furfural per 100 ml. On synthetic medium 0.1% furfural decreased the fermentation by 15%. Hydrolyzates containing more than 0.1% furfural stopped the development of the bacteria completely. Furfural was removed easily by distillation or by passing the liquor through a steam stripping colWhen furfural was added back to the stripped liquor, umn. the fermentations were not inhibited to the same extent as initially. This indicated that substances other than furfural were removed by the distillations. When the concentration of the inhibitory substances was decreased by dilution of the liquor, the extent of fermentation was improved. Complete fermentation of sugars in wood hydrolyzates was usually obtained at about 3% concentration. With glucose-malt sprouts medium 5% concentration was about the maximum quantity which could be completely fermented with