O lll“ 'I,\. .4 . a .. ..70 O -t' ‘ t ' ".11 "8-"1 zaéfiu fight; ~ 5 g 0 In A . .- ""0‘\"‘OO"IV|'OI--u. . (n... f".'f . ov \’.. ~t’. ' .‘J- " ti. .u--' 0:; . La 0 — O O o-IolouFld'b-“oxa '1' I-.- . 1111 -mwm PLANT HEAT REJECTION: REDUCTION-OF ‘ COOLING “CYCLE BLOWDOWN BY SIDE. STREAM ' TREATMENT r' Thesis for theDegreé-Of m, .s; i MICHIGANL'STATE-UNIVERSITY ; ' RALPH. HARTUNEJi-M “‘L —> — m— “‘- ”fig— _—-— J “£15m _:.-.._—. ABSTRACT POWER PLANT HEAT REJECTION: REDUCTION OF COOLING CYCLE BLOWDOWN BY SIDE STREAM TREATMENT BY Ralph Hartline New federal regulations impose strict controls on the discharges from steam-electric power plants. The researcher investigated and evaluated chemical and mechanical methods to reduce the blowdown from a closed cooling cycle. A side stream treatment using lime-soda softening and reverse osmosis was chosen as the most promising process for further evaluation. To aid in computation, a computer program was designed to model the system. The generality of the program makes it readily aadaptable to perform calculations for other water treat- ment processes. Flow charts and a complete listing are included for the program. The proposed side stream treatment process was found to be technically feasible. For a test case, over 99.5 percent of the blowdown was recovered and the makeup water flow rate was reduced by almost 50 percent. POWER PLANT HEAT REJECTION: REDUCTION OF COOLING CYCLE BLOWDOWN BY SIDE STREAM TREATMENT BY Ralph Hartline A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemical Engineering 1975 ACKNOWLEDGMENTS I wish to express my appreciation to Dr. B. W. Wilkinson for his guidance and assistance in conducting this research. I am very grateful to Consumers Power Company of Jackson, Michigan, for their financial assist- ance. Thanks are due to Dr. Paul Schierholz for his assistance in proofreading this manuscript. ii TABLE OF CONTENTS LIST OF TABLES . . . . LIST OF FIGURES . . . . I. II. III. IV. INTRODUCTION . . . BACKGROUND . . . The Nature of Water Corrosion and Scaling Lime- -Soda Water Treatment Ion Exchange . . Electrodialysis and Reverse Freezing . Problem Definition and Analysis Process Description Process Control . SYSTEM MODELING . . Material Balances Chemical Equilibrium Electroneutrality Program CTBD . . Convergence . . Subroutine PRECIP Subroutine R0 . . Subroutine ALPHA . Subroutine ELECTED Subroutine PHCHNG Subroutine FACTOR Subroutine STRCOMB Functions RSC and GAMMA Function ERROR . Subroutine AERATOR Subroutines ACTCOR and PHADJ General Program Application RESULTS . . . . iii Osmosis Page vi V. DISCUSSION Conclusions Recommendations Appendices A. COMPUTER PROGRAM LISTING B. ALTERNATE SYSTEMS C. MATHEMATICAL DETAILS D. PROGRAM INPUT/OUTPUT DETAILS BIBLIOGRAPHY iv Page 75 75 76 78 79 109 113 122 132 Table 1. LIST OF TABLES Test Water Analysis . . . . . . . . Lime-Soda Side Stream Treatment: Effects of Calcium Concentrations on Operating Conditions . . . . . . . . . . Lime—Soda Side Stream Treatment: Effects of Operating pH on Operating Conditions . Lime-Soda Side Stream Treatment with Pond Seepage: Effects of Operating pH on Operating Conditions . . . . . Lime-Soda Side Stream Treatment Plus Reverse Osmosis: Effects of Operating pH on Operating Conditions . . . . . Lime-Soda Side Stream Treatment Plus Reverse Osmosis: Variation of Operating Conditions with the Fraction of Side Stream Treated by Reverse Osmosis . . . Lime-Soda Side Stream Treatment Plus Reverse Osmosis: Effects of Brine ‘Recycle on Operating Conditions . . . Lime-Soda Side Stream Treatment Plus Reverse Osmosis: Effects of Sludge Filtration on Operating Conditions . . Ionic Size Parameters . . . . . . . Thermodynamic Constants Used . . . . . Example Data Deck . . . . . . . . . Sample Output . . . . . . . . . . Page 64 66 66 68 70 70 72 72 125 125 127 129 LIST OF FIGURES Figure Page 1. Electrodialysis Stack . . . . . . . l4 2. Side Stream Treatment Scheme . . . . . 24 3. Typical Cooling Cycle . . . . . . . 30 4. Flow Chart of Main Program . . . . . . 42 5. Flow Chart for Subroutine for Precipitation . . . . . . . . . 51 A.l. Symbols Used for Flow Rates in Program CTBD . . . . . . . . . . . . 80 B.1. Lime Recalcining Process . . . . . . llO vi I . INTRODUCT ION The use of large quantities of water in the rejection of heat from steam-electric power plants can result in thermal and/or chemical pollution. Increased awareness of the effects of pollution on the environment and the increasing occurrence of water shortages during times of peak demand, due to lack of acceptable water sources, have focused attention on the need for greater control of pollution and increased water conservation. Regulations for the control of discharges of steam- electric power plants were proposed by the Environmental Protection Agency in March, 1974, and finalized on October 8, 1974. The regulations propose stringent levels of performance with zero discharge of contaminants as its goal for 1983. Since the discharges are charac- teristically large streams with relatively low concentra- tions of contaminants (including heat), new and efficient water treatment technology will be required to meet these standards. The concept of "zero blowdown" has been the sub- ject of much discussion. The exact meaning of "zero blowdown" and what forms of blowdown disposal are accept- able are not rigidly defined. The decision criteria adopted for this thesis is based on the following defi- nition: zero discharge of contaminants means that any cooling cycle blowdown stream will have the same chemical and thermal content as the receiving body of water. Com— monly, the receiving body of water is also the original source which means that the effluent must be indis- tinguishable from the influent. Recycling the treated effluent then becomes practical, resulting in a reduction of the intake flow rate and eliminating the blowdown. Recycle of this form is an effective and desirable form of water conservation, maximizing usage of a minimum quantity of water. The concern and efforts at finding ways to meet the new standards are evidenced in the literature by the increasing number of papers addressing this topic. Con- sideration has been given to chemical, mechanical, and chemical-mechanical systems for blowdown treatment or pretreatment of cooling water. Of the chemical systems, lime-soda softening and various ion exchange schemes are most often suggested (1, 2, 3). Electrodialysis, reverse osmosis, and freezing have been indicated as applicable mechanical separation processes (4, 5, 6). Use of a chemical treatment to increase the efficiency of mechani- cal processes is recommended (4). Distillation, as applied to sea water desalination, does not appear to be economically competitive with the previously mentioned mechanical processes when applied to low salinity cool- ing waters (7). The scope of this research is to assess the alternatives and propose an economically and technically feasible system to reduce cooling cycle blowdown. Feasi- bility studies were aided by mathematical modeling and development of a computer program for the proposed system. Operating conditions were optimized for a test case, minimizing blowdown within operational limitations. Application of this research is not limited to the treat- ment of cooling cycles. Due to the generality of the computer program, it can with minor modification be used to evaluate other water treatment processes and the devel- opment of a total water management plan for a plant. q I II . BACKGROUND The Nature of Water The clean fresh water seen in rivers and streams is not just pure water, but a complex mixture containing dilute amounts of dissolved gases and minerals, organic matter, and suspended solids. These components are both naturally occurring and introduced by the actions of man. The dissolved gases are mainly carbon dioxide and oxygen scrubbed from the atmosphere by rain. The carbon dioxide concentration is often increased by decom- posing organic matter and industrial wastes. The weak carbonic acid thus formed aids in dissolving minerals in the earth's crust. Excess carbon dioxide can be easily removed by aeration. The most abundant ionic species resulting from the solvent action of water on minerals are calcium, magnesium, sodium, bicarbonate, carbonate, sulfate and chloride. Hydrogen and hydroxide ions are also present in minute quantities due to the ionization of water. Industrial wastes may introduce significant concentrations of bicarbonate, sulfate, chloride, phosphate, iron, zinc, aluminum, and sodium to natural waters. In describing the chemical nature of water, the terms hardness and alkalinity are used. The hardness 4 producing ions are calcium and magnesium. Water hardness is broken down further into two types, carbonate (tem- porary) and noncarbonate (permanent). Carbonate hardness consists of calcium and magnesium bicarbonate while the remaining calcium and magnesium salts constitute non- carbonate hardness. Alkalinity is a measure of the ability of a water to neutralize acids and is usually due to bicarbonate, carbonate, and hydroxide. Organic matter in water consists of bacteria, microorganisms, algae and other small aquatic plants, and decomposing wastes. Municipal wastes and sewage increase the organic content of natural waters. Phos- phates from soaps and fertilizers increase organic con- tent by promoting algal growth. High organic matter in cooling water is undesirable since it causes slime buildup and pitting in heat transfer equipment. The use of chlorine provides a sufficient control of organic content without adversely affecting the chemical nature of water. The term suspended solids refers to the hetero- geneous matter which many waters contain. Silica, mud, humus, sewage and industrial wastes are the major con- stituents making up suspended solids. Cooling towers increase the suspended solids of cooling waters by remov- ing dust from the air. In cooling systems, high concen- trations of suspended solids will cause fouling and sludge deposits in heat transfer equipment. Some form of suspended solids removal is highly desirable. Corrosion and Scaling Control of the chemical composition of cooling water is important to prevent scaling or corrosion in the condenser. Corrosion breaks down condenser tubes, shortening their usable life. Conditions conducive to corrosion vary, but the likelihood of its occurrence increases for acidic solutions at elevated temperatures and high salt concentrations. Therefore, the best con- trols to limit corrosion are maintaining pH near neutrality (on the basic side) and limiting total dissolved solids. Scale deposits on condenser walls are excellent insulators, effectively reducing the heat transfer capacity of the condenser. Hard dense scales are formed by a combination of calcium carbonate or calcium sulfate and silica. Since most calcium salts have inverse solu- bilities, the scale forms on the condenser tube walls where the temperature is the greatest. To prevent scale formation, the concentration of calcium salts must be kept below saturation concentration. Langlier's (8, 9, 10) method for predicting whether a water is corrosive or scale forming is based on equilibrium calculations for calcium carbonate and carbonic acid (dissolved carbon dioxide). Although Langlier's work is widely accepted and still used for natural waters, its utility is limited when applied to industrial waters. No allowance is made to compensate for the presence of other acids or high concentrations of calcium sulfate or other scale-forming salts. The work of Grohman (11) in using a computer to predict scaling and corrosive tendencies of natural water, if applied to industrial waters, suffers the same short- comings. The equilibrium interactions involving all the acids and scale-forming salts must be considered and can be easily solved on the computer to predict scaling for industrial waters. Lime-Soda Water Treatment The use of lime and soda ash to soften water is the most established form of water treatment, dating to the mid-18003. Many municipalities partially soften water to a residual hardness of 80-120 ppm expressed as equivalent calcium carbonate. The main reactions involved are: Ca0 + H o —-—r Ca(0H)2 LT. Ca++ + 2 OH- 2 Na Co3 112,2 Na++ C0: 2 — — = HCO3 + OH ——>‘ CO3 + H20 Mg++-+2 OH- 2:; Mg(0H)2+ Ca++ + co: 2:: CaC03+ w Ca++ + so= =2: Caso4+ .b (Presentation of the reactions in this form is a devi- ation from convention but more accurately represents what actually takes place.) Lime is slaked with water to produce calcium and hydroxide ions while soda ash dissociates to sodium and carbonate ions as represented in the first two equations. The increase of hydroxide ion concentration causes a shift in the bicarbonate-carbonate equilibrium toward the carbonate ion and the precipitation of magnesium hydroxide. Suspended solids reduction takes place dur- ing the magnesium hydroxide precipitation as silica is adsorbed on the forming crystals. The calcium ions naturally occurring and those from the lime combine with the carbonate produced by the bicarbonate to carbonate shift and the dissociation of soda ash to form calcium carbonate precipitate. For industrial waters with high sulfate concentrations some calcium sulfate may pre- cipitate. Although the lime-soda water treatment process has been used for over a century, improved plant opera- tion has evolved more from practical experience than from basic scientific understanding. Some recent researchers, however, have applied crystallization tech- nology to better understand the precipitation step in water softening (12, l3, 14, 15, l6, l7). Discrepancies between theoretically obtainable residual hardness and higher values obtained in practice are believed to result from the formation of a relatively stable supersaturated solution of calcium carbonate and magnesium hydroxide (15). Suggested methods to reduce this supersaturation are: a. recycle of previously precipitated sludge to promote crystal growth (16, 17); b. addition of treatment chemicals to the recycled sludge to reduce creation of new crystal nuclei (15); c. operating the precipitator at a specific pH and alkalinity for optimum calcium removal (12). Lime-soda water treatment reduces the total dis- solved solids, removing those ions especially objection- able as scale-forming agents. Unfortunately, no control of highly soluble salts is afforded by this form of treatment. Lime-soda softening is a well established economic method of water treatment where the major operat- ing expense is for chemicals. The chemical demand depends on hardness and the flow rate of the stream being treated and can be quite large. This demand can be reduced by the reclamation of lime through recalcining 10 of sludge. See Appendix B for further discussion of chemical reclamation. Ion Exchange Ion exchange processes have been used for over 30 years for the "zeolite" softening of water and more recently for deionization of boiler feed water using mixed beds of cation and anion exchange resins. The principles of ion exchange are similar to those of pre— cipitation, with the ion exchange resin replacing the insoluble precipitate. The most commonly used resins are large multi-charged organic polymers formed into small beads. The resin beads are retained in the shell of the exchange unit by screening and the water is allowed to flow through the resin bed. Although the exact mechanism is not known, the following equations are the reactions that occur in a unit operating in the sodium cycle. Ca++ + 2 Na+--—-R .__-"* Ca++-—-R + 2 Na+ Mg++ + 2 Na+--—R —‘*._ Mg++-——R + 2 Na+ As shown, the calcium and magnesium replace an equivalent amount of sodium in the resin. It should be noted that there is no reduction in total dissolved solids but rather a preferential replacement of sodium for other ions. The reversibility of the reactions 11 accounts for the ability of ion exchange resins to be regenerated using a brine solution. In normal operation, a unit is Operated until the exchange capacity is reached and calcium ions begin to "leak." The unit is then taken off line and regenerated. Regeneration requires an excess of chemicals and extra water for backwashing. Careful control of concentrations is necessary during regeneration to prevent precipitation and the resultant fouling of the resin. Sometimes the cation exchange unit is operated in the hydrogen cycle for water softening. In the hydrogen cycle, hydrogen ions are the main constituent associated with the resin; operation is the same as before with the additional replacement of hydrogen for sodium in solution. For waters high in bicarbonate an added benefit of Operation in the hydrogen cycle is the reduction of bicarbonate to carbonic acid by the acidic solution. The carbonic acid is easily removed as carbon dioxide by aeration. The removal of sodium and bicar- bonate results in a reduction of total dissolved solids. Further purification of the water can be achieved by use of an anion exchanger. Like cation exchange, the anions in solution are exchanged with an equivalent amount of anions associated with the resin. An anion exchanger, operating in the chloride or hydrox- ide cycle, is particularly useful for carbonate removal. 12 Total demineralization of water can be accomplished by using a cation resin in the hydrogen form and an anion resin in the hydroxide form. Hydrogen and hydroxide ions replace the other ionic forms in solution and neutralize each other, forming a very pure water. Ion exchange can be very useful in treating water, particularly water with less than 1000 ppm total dissolved solids. Higher concentrations adversely affect the equilibrium and increase leakage. Some pretreatment is required to prevent organics and suspended solids from entering the unit. Also, some form of metals removal may be desirable to avoid poisoning the resin and thus reducing its life. The physical limitations on the allowable strength of regenerants, concentration levels to prevent precipitation in the resin, and water requirements for backwashing can cause the waste streams from ion exchange to be relatively large. The chemical demands for ion exchange are large, and fairly accurate control of chemical doses is required to assure proper functioning of the system. Reduction in chemicals needed for regeneration is being realized by the develoPment of new resins requiring weaker regenerants, methods requiring no chemicals for regeneration, and systems to recycle regenerants. 13 Electrodialysis and Reverse Osmosis The principles of electrodialysis have been known for many years but practical development did not occur until the late 19403 when durable ion exchange membranes were develOped. Electrodialysis has gained acceptance for desalination of brackish water; the practicability and economic feasibility of this appli- cation are being demonstrated by the operation of many large scale plants. Figure l diagrams the operation of an electrodialysis stack. Electrodialysis is a mechanical separation process using an applied electric field to cause migra- tion of ions toward the anode and cathode. The presence of alternating cation and anion selective membranes alters the migration of the ions causing concentrations to increase in one compartment formed between two mem- branes and decrease in theneighboring compartments.. Cations in a dilution compartment will migrate toward the cathode. The cation permeable membrane will trans- fer the migrating cation into the concentration compart— ment. Further migration of the cation toward the cathode is restricted by the anion selective membrane. Likewise, the anion migration toward the anode is from the dilution cell through the anion membrane to the concentration cell where the anions are retained by the cation membrane. The use of manifolds to distribute 14 Feed to . . B ‘n t c t t' h dilution chambers f1 e 0 con en ra 1n9 C ambers l __ ._.i;__ .7 Electrode wash 1 l - Electrode wash _ l c: I A. l c l A 1, c A _ é, go <69 /® /<+> 8 m 0 61 Ci 9\ L ' i ' 1 a: v t. __ T" __ .1 l 4 v Concentrated Product brine water Legend (9 -- any cation 9 -- any anion C -- cation membrane A -- anion membrane Figure 1.--Electrodialysis Stack. 15 and collect flow from the cells permits recycle of brine to increase the concentration effect and conserve water, and the production of water of any desired purity. Reverse osmosis is a relatively new separation process, still mainly in the developmental stage but with great promise for application to the treatment of saline and waste water. Consideration of reverse osmosis along with electrodialysis is only natural since both are membrane processes and hence are very similar. The difference between the two is that in reverse osmosis water is transferred, not salt, and the driving force is pressure instead of an electric field. The membrane used for reverse osmosis, synthesized from cellulose acetate, a relatively cheap material, allows the passage of water but not other substances. The prin- ciples of reverse osmosis are basically simple and can best be understood by first considering osmosis. This is the phenomenon that causes dehydration in humans drinking salt water. When two salt solutions of different concentra- tions are separated by a semipermeable membrane, a non- equilibrium condition exists at the junction--the membrane wall. In order to be at thermodynamic equilibrium, the concentrations on both sides of the membrane must be the same. Since the salts cannot pass through the membrane 16 there must be an active transport of water. The driving force for this transport is called the osmotic pressure. By applying pressure in excess of the osmotic pressure to the solution of higher concentration the process can be reversed. Relatively pure water will be formed and a brine solution of high concentration will be retained. The use of membranes gives rise to problems common to both electrodialysis and reverse osmosis. Extending membrane life is important to the economics of both processes. As in ion exchange, it is important to prevent suspended solids and organics from entering the units and clogging the membranes. Concentrations in the brine must be controlled to prevent precipitation which will also lead to clogging of membranes. The use of a chemical softening process prior to electro- dialysis or reverse osmosis has merit in reducing suspended solids and precipitate forming ions, protect- ing the membranes and allowing for increased efficiency. Freezing When a solution freezes, the ice formed is pure water. This fact, along with the relatively small heat of fusion, has made freezing seem attractive as a water treatment method. Research has not proceeded beyond the testing stage and has been hampered by the diffi- culty of separating and handling the ice and the 17 carry-over(xfprecipitates and dirt which act as nuclei for ice crystal formation. For these reasons, freezing is not a feasible alternative. Problem Definition and Analysis In power plants, steam leaving the turbine con- tains low value thermal energy which is impractical to recover. This heat is transferred to a large cooling water stream in the condenser. Presently, this warm stream is handled by two different methods. The first method is an Open system often called "once-through cooling." As the name implies, the cooling stream passes through the condenser only once and the warm stream leaving the condenser is returned to the originating stream. Here the pollution introduced to the body of water is the heat. This mode of operation requires a large quantity of water. The second method is a closed system using a cool- ing tower or pond. The warm stream from the condenser enters the cooling tower or pond where the excess heat is rejected to the atmosphere by evaporation of part of the cooling water. The water from the tower or pond can be reused to cool the condenser. Since water is removed by evaporation, makeup water is required and the concen- trations of chemicals in the cooling water increase. Chlorine must be added to prevent slime from forming in 18 the condenser. Acid is added to neutralize the water, balancing the corrosion and scaling tendency Of the water. Scale prevention in the condenser is important to main- tain a good heat transfer efficiency. As the concentra- tions Of the scale-forming constituents increase, it becomes necessary to remove a blowdown stream. The flow rate of the blowdown must satisfy the Overall mate- rial balance On the system to keep the concentrations Of scale-forming constituents below saturation concentra- tions. The blowdown for even a medium-sized plant will be significant, thus creating a large disposal problem. Despite the size of the blowdown, it is much smaller than the blowdown from the open system. The closed system holds more promise for the future due to its conservation of water. Open cycle cooling for new plants is not feasible in meeting the proposed federal regulations. In light of this, only closed systems were considered in this research. Scale inhibitors have been frequently used to increase the limits on the concentrations of scale- forming constituents, thus reducing the blowdown flow rate. The strict limitations on concentrations of inhibitors in blowdown established by the new regula- tions, the difficulty of removing them from blowdown, their interference of chemical treatment processes, and the associated high costs of their replacement predicate 19 the discontinuance of the use Of inhibitors in the future. Therefore, the use of inhibitors was not con- sidered for the systems investigated. Blowdown from closed systems can be handled by either an Open or a closed system. Deepwell injection and solar ponds. are the two major open systems; the costs Of these systems are directly dependent on the flow rate of the blowdown stream being disposed. Solar ponds and mechanically assisted solar ponds were evalu- ated by Dr. M. C. Smith (18) and found to be impractical and uneconomical for application to large blowdown streams in cool, humid northern climates like Michigan's. The closed systems are the chemical and mechanical sepa- ration processes previously described applied to either pretreatment of the makeup or treatment of the blowdown. Advantages of the closed systems are (a) water conserva- tion by reducing makeup water requirements and water recovery, (b) significant reduction of the waste stream requiring disposal, and (c) costs are more dependent on the amount of dissolved solids removed than the flow rate of the stream being treated. Advantages and dis- advantages of the various closed systems applications to reduce cooling cycle blowdown are discussed in the fol- lowing paragraphs. Lime-soda treatment can be used for either pre- treatment Of makeup water or side stream treatment Of the 20 circulating water. The circulating water with higher chemical concentrations would be easier to treat than the makeup water. Pretreatment Of makeup water should be used only when the quality of water is near saturation. Otherwise, the amount of water needed to be treated to accomplish the same extent Of calcium and magnesium removal will be less for side stream treatment. The result will be a saving in chemicals and a reduction in the size of equipment required. The equipment size reduction can be significant since a large power plant with a cooling cycle evaporation rate Of 15,000 gal- lons per minute would require a pretreatment plant the size of Lansing's Cedar Street Water Conditioning Plant which processes 24 million gallons per day (19). Lime-soda treatment alone usually will not be sufficient to control the corrosive properties of the circulating water; an additional blowdown or treatment probably will be required tO remove the highly soluble ions. When the only blowdown is the sludge, the mate- rial balance is satisfied only if the total dissolved solids in the cooling water is at a very high level. This high salt concentration can be corrosive and also can interfere with the proper operation of the precipi- tator. However, H. W. Frazier reports that use of a high density solids precipitator yields favorable results even with high total dissolved solids (2). 21 Therefore, additional blowdown or treatment is only needed to the extent of maintaining corrosion limits. Ion exchange like lime-soda treatment can be used either for pretreatment of makeup water or side stream treatment of circulating water; regenerant require- ments are the same for both applications. Use of a side stream ion exchanger results in reduced resin requirements but the higher ion concentrations would adversely affect the equilibrium and increase leakage. Therefore, the most efficient use of ion exchange will be for pre- treatment. Since Operation in the sodium cycle does not change total dissolved solids, a blowdown would be required to control corrosion. The need for a blowdown stream could be eliminated by completely demineralizing the makeup water but operating costs would soar because of the extra demand for expensive acids and bases. Pre- liminary calculations indicate that the wastes from regeneration and backwashing, although smaller than blow- down streams, are still very large streams and would pose a diSposal problem. Ion exchange does not appear to be feasible for the above reasons. The necessity of controlling concentrations below saturation levels to protect the membranes limits the application of electrodialysis or reverse osmosis to cooling water treatment. Since the blowdown is already near saturation, further concentration by electrodialysis 22 or reverse osmosis is impossible. Using these methods for pretreatment would produce a brine stream almost equal in size and composition to the blowdown produced if they were not used; hence, no advantage is gained. An application for electrodialysis or reverse osmosis that does have merit is use in conjunction with a lime-soda side stream treatment process. The electro- dialysis or reverse osmosis treatment removes the soluble ions unaffected by the lime-soda softening. The lime- soda treatment would reduce the slightly soluble ion concentrations producing a water that could be concen- trated many times. The electrodialysis stack or reverse osmosis module could then produce a small, highly con- centrated brine stream. The amount Of salt rejected by the unit would depend on the inlet flow rate and con- centrations. Scaling tendencies would be controlled by the lime-soda treatment while the electrodialysis or reverse osmosis concentrator would maintain total dis- solved solids below corrosion limits. The only blowdown from the system would be the sludge from the precipi- tator and the concentrated brine. The sludge has the potential to be recalcined to recover most Of the chemical value but there is no apparent use for the brine or economical method to recover its chemical content. 23 Process Description A flow sheet of the proposed system, using a side stream lime-soda softener followed by a reverse osmosis unit, is presented in Figure 2. The makeup water from a river or well is pretreated before entering the cooling cycle. For a slightly alkaline makeup water the pretreat- ment consists of acid addition to neutralize the water for scale and corrosion control. Acid addition also improves makeup water quality by reducing bicarbonate to carbonic acid which is removed by aeration. No addi- tional aerator is necessary if a cooling tower is used because aeration will take place in the cooling tower. The side stream to be treated is drawn off the cooling cycle after the condenser but before the cool- ing tower or pond in order to take advantage of the higher temperature of the water. Two advantages accrue from the use of warm lime-soda softening: a. The yield of the precipitation step increases due to the inverse solubility of calcium salts. b. Suspended solids removal improves because Of the increased silica adsorption on mag- nesium hydroxide crystals at elevated temperatures. The size of the side stream is determined by the material balance and equals the flow rate of the blowdown that would be required to prevent scaling. 24 .OEmnOm uawEummHB Emmuum mpflmll.~ musmam I A (twobll msflcwoamomu HO ammommww ou Omvsam kuaam ~—-—_———--—-—-——-— + Hmumx mdwxmfiy .H Moumumd cflom Osom MO Hmzou +. maowowm msmxmfi Hmuoe maflaoou cm mauwms u u m Hmwommap F o» madam memosmo coaumuomm>m Omnm>mm OHOMO mcflaooo .M muoumuamaomum_ c.0¢ nmm 60m ammuum mufim Hmmcmvsou 25 The lime-soda softener used in the process is Operated in two stages. In the first stage only lime is added to precipitate the carbonate hardness and possibly calcium sulfate if the water has a high sulfate concen- tration. The second stage removes the residual excess calcium by precipitation using soda ash. The high alkalinity effluent from the precipitators is reacidi- fied to neutrality; the acidification shifts the carbonate equilibrium to the bicarbonate meaning the calcium car- bonate concentration is well below saturation. The sludge produced by lime-soda softening is usually in the range of 3-5 percent solids by weight but can be con- centrated to about 50-60 percent solids by filtration or centrifuging (20). After the lime-soda softener, part of the treated water is passed through the reverse osmosis unit producing a high quality supernatant stream and a small concentrated brine stream. The flow rate to the reverse osmosis unit is determined by a material balance on the soluble salts in the system and the limits on total dissolved solids. The reverse osmosis supernatant, the remaining lime-soda treated water and the makeup water are all combined to form the total makeup to the cooling system. The higher quality total makeup water can be concentrated more than just river makeup water before treatment becomes necessary. 26 Partial brine recycle to the precipitator may be desirable. By recycling the brine, the ratio of soluble salts to hardness will increase. Thus the brine can be concentrated more and the salt content Of the precipi- tator sludge will increase. A smaller waste stream will then be required to satisfy the material balance. An increase of the total dissolved solids in the precipi- tators is a disadvantage of this mode of operation. The reverse osmosis unit was chosen because of its Operational simplicity and the relative ease of modeling it mathematically. A more precise technical and economic comparison of reverse osmosis and electro- dialysis is necessary. The results obtained are equally applicable to each separation method, the main differ- ence would be in flow rates to the units. Although operating costs are mainly a function of the quantity of salt rejected, minimizing the size Of the unit is desirable to reduce capital investments and membrane replacement costs. Process Control The fluctuations that occur in normal power plant Operation to meet electricity demands pose a prob- lem in controlling the treatment system. The large volume of water involved acts as a cushion for small, rapid changes but also introduces a large lag time that 27 could cause instability. A fairly high level of auto- mation would be required to provide continuous trouble- free Operation. The most critical control problem is preventing scaling and corrosion in the condenser. Locating the sensors for circulating water quality control near the entrance to the condenser provides the best safeguard for the condenser. Accurate control of chemical feeders and stream flow rates is required to insure prOper functioning of the system. Control of the chemical feeders is the same for all of them and uses either pH or ion selective electrodes in the sensors. The acid feeder for the makeup water would be controlled with respect to the pH of the circu- lating water. The pH in the lime precipitator would be the criteria for controlling lime doses. Soda ash dosages would increase with an increase in the calcium concentration leaving the soda ash precipitator as measured by a calcium specific electrode. Control Of the acid feeder on the side stream would be based on the pH entering the reverse osmosis unit. The flow rates are all complexly interrelated but can be controlled independently. The flow rate of the makeup would be controlled to maintain a constant volume of water in the cooling cycle. Control of the side stream flow rate would be based on the scaling 28 tendencies of the cooling water, measured as calcium and sulfate concentrations and alkalinity. A conduc- tivity measurement to determine the total dissolved solids of the cooling water would be used to control the flow rate to the reverse osmosis unit. A simple propor- tioning valve is all that is necessary to control brine recycle and it is unaffected by changes in flow rates. III. SYSTEM MODELING The side stream treatment process, like any other process, must satisfy the physical laws governing the system. Specifically, it must satisfy the laws of mass balance, chemical equilibrium and solution electro- neutrality. Steady state Operation was assumed which requires the evaporation rate to be fixed and the compo- sition of the makeup water to be constant. Material Balances The law of mass balance must be satisfied for each piece of equipment, each internal loop, and the overall system. The important material balances for the side stream treatment process are: a. the material balance for the cooling cycle, b. the material balance for reverse osmosis, c. the material balance around the pre- cipitators, and d. the overall material balance. The mass flows for a simple closed cooling cycle are illustrated in Figure 3. The material balance for water simply states that the makeup flow rate must equal the evaporation rate plus the blowdown flow rate. 29 30 Evaporation Cooling _ _______, tower or r Blowdown pond Condenser *————"———-Makeup Water Figure 3.--Typical Cooling Cycle. Since the concentrations of salt in the evaporating water are negligible the only loss of salt is through the blowdown, and the concentration of a salt in the blowdown is a constant factor larger than the concentration of the same salt in the makeup water. This constant factor, called the cycles of concentration or concentration fac- tor, is the same for all salts. The maximum allowable concentration factor for a makeup water is determined by the solubility of the least soluble salt likely to cause scale formation in the condenser. The material balance for the salt in the cooling cycle requires the salt addi- tion by makeup must equal the salt removal by blowdown; or in terms of concentrations and flow rates, for each 31 salt its concentration in the makeup times the flow rate must equal its concentration in the blowdown times the flow rate. Solution of the material balance for the cooling cycle requires only that the evaporation rate and con- centration factor be known. The concentration factor is derived from the makeup concentrations and the solu- bility data for the salts. Flow rates are defined by the following equations: MU = ExF/(F - 1) (1) BD = E/(F - 1) (2) or BD = MU - E Legend: MU is the makeup flow rate, ED is the blowdown flow rate, E is the evaporation rate, and F is the concentration factor. Comparing Figure 3 with Figure 2, it can be seen that the material balance derived for the simple cooling cycle applies equally well to the cooling loop of the proposed system. The makeup to the cooling cycle is the total makeup stream consisting Of the river makeup water plus the side stream treatment recycle stream. The blowdown is the same as the stream identified as the 32 side stream. For a cooling pond, rain is a reduction in the evaporation rate while pond seepage (or tower drift) is a reduction in the side stream flow rate. The material balance for the reverse osmosis module is basically the same with the exception that the makeup rate or flow rate into the unit is the known flow rate. The evaporation stream compares to the supernatant stream and the brine is equivalent to the blowdown. The equations for the flow rates are: R01 = ROI x (F1-l)/F1 (3) BRINE = ROI/Fl (4) or BRINE = ROI - R01 Legend: R01 is the supernatant flow rate, BRINE is the brine stream flow rate, ROI is the flow rate entering the reverse osmosis unit, and F1 is the concentration factor for the reverse osmosis unit. The material balance for the precipitators is more complex due to the many input and output terms and the chemical reactions taking place. The balance for the salts was combined with the equilibrium calculations for the precipitators; these calculations are described in more detail in the section headed Subroutine PRECIP. 33 The concentrations Of the ions in the precipitator effluent and the rate of dry solids production per gallon_ of flow into the precipitator are thus calculated. Specifying a percentage of water in the sludge and using the dry solids production rate establishes the sludge flow rate. A material balance for water determines the flow rate of the effluent. The flow rates are given by the following equations: w s x TOT/(1 - PERW) . (5) R S - PERW X W/8.34 (6) Legend: W is the pounds per minute of sludge, S is the side stream flow rate, TOT is the pounds of dry solids produced per gallon of side stream flow, PERW is the weight fraction of water in the sludge, R is the flow rate of effluent from the precipitator, and 8.34 is the weight of one gallon Of water. Direct solution of the overall material balance was used for the highly soluble ions only. It was not necessary to do so, but much more accurate results were obtained than by indirect calculations. For the slightly soluble ions proper solution of the independent material balances adequately satisfied the overall material 34 balance. The equations used to solve the overall mate— rial balance for each Of the highly soluble ions are presented in Appendix C. Chemical quilibrium The chemical precipitation by the addition of lime and soda ash is complicated due to the equilibrium relations. The dissociation and association of solids, the acid-base reactions and the ionization of water are all interrelated. In order to completely understand the system modeling it is necessary to understand the rela- tions. Consider the general chemical reaction: nN + mM :23 rR + 88 Legend: N, M, R, S identify chemical species; n, m, r, s are the number of moles of each. The thermodynamic equilibrium constant, K, is defined by: r s a - a _ R S K - an . am (7) N M where a is the activity of the species indicated by the subscript. 35 Values of K can be found directly in the litera- ture or calculated from free energy data using the relation (21): _ _ AF° In K —- RT (8) Legend: AF° = standard free energy change, R = universal gas constant, and #5 ll absolute temperature. Temperature correction can be made by the following equation (21): (9) ln ——— = - AH° is the enthalpy of formation and can be considered constant over a small temperature range. The equilibrium relation is easier to use when concentrations are used instead of activities. The activity of an undissolved salt and undissociated water are both taken to be unity. For ionic species in solu- tion, the activity is directly proportional to the ionic concentration. a. = YiC. (10) 36 Ci is the ionic concentration and is usually expressed in moles per liter or moles per 1000 gms of solvent. At low concentrations in the range present in water the two ways of expressing concentration are equivalent. The activity coefficient, Yi’ can be found by use of an analytical equation such as the Debye-Huckel equation (22): N Azi/fi log y. = - (ll) 1 1 + B a1.- /5 Legend: . . .th . Zi = ionic charge of the 1 spec1es, u = total ionic strength = % 2 CiZi2 I i a. = ionic size parameter (values are presented in Table D-l in Appendix D), A = constant = 1.825 x 106 (DT)-3/2, (DT)’1/2, \O B = constant = 5.029 x 10 D = dielectric constant of water (23) = 78.54 [1 - 4.579 x 10'3 6 (T-298) + 11.9 x 10’ (T-298)2 9 + 28.0 x 10' (T-298)3] The Debye-Huckel equation shows good agreement with experimental results up to ionic strengths of about .01 and for one-to—one electrolytes. Use for solutions with 37 higher ionic strengths or mixed electrolytes as may be found in the proposed system is questionable but better than the use of no activity coefficient calculation. The equation for the thermodynamic equilibrium constant can then be rearranged to define an activity . corrected equilibrium constant, K , in terms of concen- trations as: Ym Yn r s K' = K S 2 = [le ’ [Sln (12) YR vs [M1 . [N1 Use Of concentrations is more convenient since they can be easily measured. Further simplification is possible for the equilibrium constant for the dissociation of a solid since the activity Of the solid phase is unity. For example, consider the dissociation of calcium carbonate. CaCO3(S) —-*._.. c...++ + co; a ° a = Ca++ CO3 K = = a ++ - a = aCaCO3 Ca CO3 or , _ ++ . = Sp - [Ca ] [CO3] (13) This special form Of the equilibrium constant is called the solubility product. For any solution in equilibrium, 38 the product of the concentrations will always be less than or equal to the solubility product. If either one or both of the concentrations are increased until the product of the concentrations exceeds the solubility concentration, solid salts will be formed from the solu- tion until equilibrium is once again attained. The dissociation of water, although an acid- base reaction, is similar to the solid dissociation. Since the activity of the undissociated water is unity the equilibrium constant is defined only in terms of hydrogen and hydroxide concentrations. K; = [n+1 . [OH-] (14) The product of the concentrations must always equal the water dissociation constant. Therefore, whenever an acid or base is added to an aqueous solution, the hydro- gen and hydroxide concentrations change to reestablish equilibrium. For the example of calcium carbonate and other saltscfifweak acids the equilibrium concentration of the anion also depends on pH. When carbonic acid dissoci- ates in water three forms of the acid result: undisso- ciated acid, bicarbonate ion, and carbonate ion. The fraction of the total carbonate concentration in each form Of the acid depends on pH and can be calculated by the following equations: 39 [H co 1 [11"]2 a - _________2 3 - (15) 1'" C I + 2 + , , . , T [H ] + [H ]Kl + K1 K2 [HCO-l K' - 3 l a = ——————- = ———— - a (16) 2 CT [H+] l [CO=] K' 3 2 01 = = ° 0!. (l7) 3 CT [H+] 2 Legend: C = total carbonate concentration T K1, K2 = first and second acid dissociation constant. In terms Of the total carbonate concentration at a known pH . = ++ KSp [Ca 3 or K" (18) SP , _ ++ . KSp/OL3 — [Ca ] CT Similar solubility products corrected for activity coef- ficients and based on total anion concentration and pH can be calculated for use in calculating equilibrium concentrations. The derivation of these equations and their general form appear in Appendix C. The concentration of ions after the addition of lime and soda ash and the formation of precipitates must 40 satisfy all the equilibrium conditions. The mathe- matical calculations to find these concentrations require solution Of a multivariable nonlinear system of equations. These calculations are further complicated by the fact that for the residence time of most water treatment plants equilibrium is not Obtained. This is usually explained by the formation of a relatively stable supersaturation which increases the amount of time required to reach equilibrium to the order of days. It has been reported (15) that the supersaturation pres- ent in lime—soda softened waters is 2 to 3 when expressed as a ratio Of the concentration of a salt to the the- oretical equilibrium concentration. The supersaturation depends on many things including residence time, total ionic strength, and concentrations of the species present. However, due to lack of models to quantitatively specify these relations, a constant factor of two was used to correct the concentrations Obtained from the equilibrium concentrations. Actually, the solubility products were increased by the proper power of two so that the super- saturation concentrations were calculated directly by the equilibrium calculations. DevelOpment Of a model Of supersaturations would improve the accuracy of these cal- culations. 41 Electroneutrality The principle of electroneutrality Of a solu- tion is quite simple. A solution cannot have an over- all electric charge Or, in other words, the summation of ion concentrations times their charges must be zero. Electroneutrality imposes a restriction on solutions that must be met at all times and identifies the solutions to the system of equations that can be physically possible. Program CTBD Program CTBD is the program combining the mate- rial balance and equilibrium calculations to model the side stream lime-soda and reverse osmosis treatment process. Basically, the program is an iterative solu- tion of the system starting with the assumption of no recycle from the treatment process. This means that the total makeup is only the pretreated river makeup water. A flow chart Of the program is presented in Figure 4 and a complete listing of the program is attached as Appendix A. Inputs to the program consist of operating parameters, necessary constants, and the makeup water analysis and ion data. The operating parameters are the treatment operating temperature and the independent variables that need to be specified to calculate the material balances. The necessary constants are the 42 Figure 4.--Flow Chart of Main Program. NO 43 c 3 Input Operating parameters Equilibrium constants Makeup water analysis Initial calculations Evaluate constants Convert concentrations to molar Is aeration “58am? Calculations for aerator Initialize loop Flow rates for treat- ment + 0 Concentrations in TM'+ concentrations in MU 44 Calculate maximum.number of cycles for TM, concentrations in side stream Material Balance for cooling system (calc. flow rates of S and TM) Has the flow rate of TM changed? Yes Calculations for lime precipitator soda ash treatment needed? NO 44 Calculate maximum number Of cycles for TM, concentrations in side stream Material Balance for cooling system (calc. flow rates Of S and TM) No Has the flow rate of TM changed? Calculations for lime precipitator soda ash treatment needed? 1‘ 45 Calculations for soda ash precipitator NO Material balances for precipitators Is reverse osmosis used? Calculations for reverse osmosis Approximate TM, New concentrations in total makeup 46 Was the loop cal- NO culated? Did the number Yes of cycles change? Calculate chemical usage, and total dissolved solids in the cooling water Output Results Flow rates Chemical dosages Concs. in total makeup and the cooling water l C END D 47 free energy of formation and enthalpy data used by function EQK to calculate solubility products and acid constants according to Equations (8) and (9). (The free energy and enthalpy values used are listed in Table D-2 in Appendix D.) The data for the makeup water is stored in two separate arrays, one for the cation concentrations and the other for anion concentrations. An identification matrix read by the program along with the data for solubility products contains the array locations Of the cation and anion concentrations and the proper exponents for each solubility product. The concentrations of the water analysis were expressed in ppm as the ion but had to be changed to molar by dividing by one thousand times the ionic weight. If the concentrations are in ppm as CaCO3 the ionic weight should be set at one hundred so no change in the calculations will be required to Obtain molar concentrations. Before entering the loop, calculations for the pretreatment are performed. First, though, the sodium concentration is adjusted to eliminate any round-off errors and insure the electroneutrality of the makeup water. The total carbonate concentration after aeration and the amount of acid required to adjust the pH to the desired level are calculated for the makeup water. 48 The flow rates and concentrations of the treatment streams are all initialized. Using the total makeup stream concentrations, the maximum allowable composition of the cooling water is calculated by subroutine FACTOR. Material balances then givethe flow rate of the side stream and total makeup. The lime dosage required to Obtain the operating pH and the theoretical compositions entering the precipi- tator are calculated. Equilibrium calculations performed by subroutine PRECIP give the concentrations of ions in the supernatant stream and the dry weight of precipitate formed. Similar calculations are performed for the soda ash precipitator. The amount of acid required to neutralize the supernatant is calculated next. Sub- routine RO is called to calculate the concentration of the brine stream. Finally, material balances give the flow rates Of the recycle, makeup water and waste streams. The concentrations and flow rates of the recycle and makeup water are used to predict the new concentra- tions in the total makeup. The concentrations of the highly soluble ions in the total makeup stream are adjusted to satisfy the overall material balance. The new total makeup concentrations are then used to calcu- late a new concentration factor and hence new flow rates for the side stream and total makeup. These calculations 49 are repeated until the value of the flow rates used to calculate the concentrations in the total makeup are close to the same as those obtained from the material balance for the cooling cycle. Then the entire loop is recalculated. The iterative procedure terminates when the cycles of concentration reach a steady value. The important flow rates, chemical requirements and compo- sitions of the cooling water are the results that are printed. Convergence The iterative method of successive substitutions is probably the most commonly used method of handling systems with recycle. The simplicity Of application is its major advantage but the rate Of convergence is usually very slow and may not converge at all. For the system investigated, convergence was very slow and made a slow cyclic approach to the answer. In an attempt to improve convergence, Wegstein's method (24) was used on the flow rate of the total makeup. Wegstein's method is a weighted averaging tech- nique and is described in greater detail in Appendix C. The use of Wegstein's method resulted in a significant reduction in computer time necessary to reach the steady state answer. Still, the program demonstrated 50 instability for some test values. The instability was found to result mainly from the highly nonlinear nature of the activity coefficients. Therefore, convergence is poor for Operating conditions which cause high total dissolved solids content in the cooling water or precipi- tator. This problem is most apparent for the case of brine recycle. Subroutine PRECIP Subroutine PRECIP, the main subroutine of the program, does the equilibrium-material balance calcula- tions for the precipitators. An iterative procedure is used to solve the system of equations that results from the equilibrium conditions. A flow chart of the system is presented as Figure 5. For each insoluble ion pair, the concentrations in the influent as calculated in the main program are corrected until the equilibrium concentrations are obtained. The correction to the concentrations is calculated by a first order approximation; the deriva- tion Of the approximation is presented in Appendix C. After the corrections are made for each insoluble ion pair, the hydrogen and hydroxide ion concentrations are corrected for time dissociation of water. The solubility products are checked again and concentrations corrected to compensate for any common ion effects. This procedure 51 Figure 5.--Flow Chart of Subroutine for Precipitation. 52 C j Calling Parameters Stream identifier (J) Concentrations in Stream J Concs. in J+l'+ cone. in J Iflag and x-+ 0 Correct solubility products for activity and super- saturation Main iterative loop Set counter to zero Adjust solubility products for acid effects Do calculations " for each product 53 I Iterate correction I for I each product Calculate deviation from equilibrium Is the salt below satura- tion? Yes Calculate first order | correction to con- | centrations NO Set correction to zero ————‘-—_—— I Correct concentrations ' Increase counter the salt close to satura- tion? Yes Calculate correction for H+ and OH- due to the dissociation of water 55 Is the counter Yes zero? l Calculate the dry weight of the precipitate ‘ Re turn D (Dashed lines are used to indicate returns on do lOOps.) 56 is repeated until none of the concentrations need to be changed. The cumulative change for each ion pair is retained and used to calculate the total dry weight of precipitate formed per gallon of influent. The subroutine works best when the least soluble salts are considered first. This can be accomplished ‘ by ordering the solubility products in the input in a corresponding order. A preliminary run will aid in determining a good order. Subroutine RO Subroutine RO contains the calculations used to model the reverse osmosis unit. The model used for reverse osmosis is basically straightforward and simple. The supernatant is assumed to be pure water and the ion concentrations in the brine are assumed to be a constant safety factor less than saturation to prevent membrane clogging. RO returns the brine stream ion concentrations and the concentration factor for the reverse osmosis unit. In actual practice, salt leakage in reverse osmosis is about one percent, under ideal conditions. The assumption of no salt leakage leads to about a one percent error in the calculated flow rates of the influent and supernatant but doesn't affect the flow rate or composition of the brine or the composition of the recycle. 57 Since electrodialysis can produce water of vary- ing quality it corresponds to reverse osmosis with vary- ing amounts of salt leakage. The reverse osmosis subroutine can be used to model electrodialysis by assuming there is a bypass stream responsible for the salt leakage. This model would give good results for the production of high quality water. For the produc- tion of a lower quality water application of the model is limited; the different ion mobilities become important in determining the composition of the effluent and brine. In this case the model should be expanded to consider the ion mobilities. Subroutine ALPHA For a stream of a known pH and total ionic strength, subroutine ALPHA calculates for each anion of a weak acid the fraction of the total concentration in each stage Of dissociation. ALPHA uses the general equa- tions develOped in Appendix C under the sub-heading Acid Dissociation. It is important that ALPHA be called with the right parameters before subroutines ELECTRO and AERATOR are called. Subroutine ELECTRO Subroutine ELECTRO checks the electroneutrality Of a stream by simply summing the products of the ion charge and concentration. The deviation from 58 electroneutrality for the stream is returned by ELECTRO. Errors caused by the iterative procedure or computer truncation can be avoided by frequent use of ELECTRO and using the deviation returned to correct a key ion con- centration. Subroutine PHCHNG Subroutine PHCHNG calculates the amount of acid or base required to change the pH Of a stream to the specified value. The amount of acid is calculated by an electroneutrality calculation. Since the solution must be electrically neutral after the addition of acid (or base), the deviation from electroneutrality--if just the hydrogen, hydroxide and alpha values are changed--corre- sponds to the amount of anion (or cation) associated with the acid (or base) that was added. The reverse of the procedure, calculating the pH of a stream after addition of a given amount of an acid or base, requires an iterative procedure. Attempts to develOp such an iterative procedure as outlined by Thompson and Trussel (25) were unsuccessful; the pro- cedure would not converge. Subroutine FACTOR The concentration factor or number of cycles for the cooling cycle and the reverse osmosis unit is cal- culated by subroutine FACTOR. The calculations are 59 based on the model of a closed concentrator; that is, the only thing removed is water. Also, it is assumed that acid is added to maintain the pH at the same level. This corresponds to the acid addition to the makeup to control circulating water pH, although the calcula- tion is artificial in assuming the pH is maintained constant. The calculation is carried out in this manner because of the difficulties encountered in predicting pH that were previously discussed. The closed concentrator model applies well to the reverse osmosis unit and a cooling pond but the good liquid-air contacting that occurs in a cooling tower requires modification of the model to consider carbon dioxide stripping and the resulting effect on the car- bonate concentration. A flag would have to be used to indicate when the subroutine is being used for a cool- ing tower in which case the carbonate concentration would not change from the equilibrium value. Subroutine STRCOMB The ion concentrations in a stream that result from the combination Of two streams of different composi- tions are calculated by subroutine STRCOMB. Use of the subroutine is limited to streams with almost the same pH. The concentrations are found by a simple material balance. The subroutine can be used to initialize a stream 60 equivalent to another stream by identifying the two joining streams by the same stream number and using any non-zero combination of flow rates. Functions RSC and GAMMA Functions RSC and GAMMA are used to calculate activity coefficients. RSC calculates the square root Of the total ionic strength of a solution. GAMMA cal- culates the ionic activity coefficients by using the Debye-Huckell equation. Function ERROR Function ERROR calculates the ratio of the solubility product calculated us ing the ion concentrations to the adjusted equilibrium solubility product for an ion pair. The values calculated by ERROR are used to find the concentration fac- tor and the concentration corrections used in PRECIP. Subroutine AERATOR Subroutine AERATOR performs the calculations for the aerator returning the resulting total carbonate concentration and the amount of acid necessary to main- tain the same pH. The equilibrium carbon dioxide con- centration must be specified. Since the removal of carbon dioxide is an easy separation, it is assumed that the equilibrium carbon dioxide concentration is reached. Nordell (26) reports that this concentration for indus- trial areas is approximately 6 ppm as carbon dioxide. 61 Subroutines ACTCOR and PHADJ These two subroutines adjust the equilibrium solubility product for activity and pH effects. ACTCOR uses Equation (12) to correct for activities and PHADJ uses Equation (18) to allow total anion concen- trations to be used. General Program Application To use the program as it is designed requires a minimum of preparatory work. The desired Operating conditions must be determined for the system. The cation and anion concentrations of the water analysis should be separated and ordered in the proper matrices along with the other ion parameters. (The hydrogen and hydroxide ions should be the first in their respec- tive matrices.) The necessary enthalpies and free energies for the solubility products and the proper locators and exponents can then be determined. The identification matrix entries are located on the same card as the thermodynamic constants for the solubility product to minimize confusion. Once the data deck is established, changing a variable requires only replacing the appropriate card. More detail on establishing the data deck is given in Appendix D. The modular design ofiflmaprogram, using sub- routines for similar important calculations, adds 62 flexibility to the program. Variations in the flow pattern or other water treatment processes can be handled by only modifying the main program. Modification of the program will be aided by using a flow diagram of the process and writing out the steps and what takes place at each step. Care must be taken that the streams are labeled prOperly and that the subroutines are called in the right order with the proper parameters. Users should take time to familiarize themselves with the limitations of the subroutines before using to insure proper application. For systems using recycle, extra care must be taken to make sure the highly soluble ions satisfy the material balances and that the calculations converge. IV. RESULTS The computer program described in the preceding sections was used to perform calculations to evaluate various modes of Operation of the lime-soda and reverse osmosis side stream treatment of cooling water. The data used in the calculations is for the Consumers Power Company nuclear power plant being constructed in Midland, Michigan (27). The two nuclear reactors, rated at 2552 thermal megawatts each, are designed to produce 1300 megawatts electricity and 4 million pounds per hour of high pressure process steam for use in a nearby chemical plant. A cooling pond will be used to reject the excess heat, approximately 7 x 109 BTU's per hour. The results Obtained were based on the follow- ing design criteria: a. A constant makeup water composition corre- sponding to the Titabawassee River composition measured at the Poseyville Bridge. The water analysis is pre— sented in Table l. b. A constant evaporation rate of 14,000 gallons per minute reduced 1,400 gallons per minute by rainfall. c. The circulating cooling water is maintained at 63 64 TABLE l.--Test Water Analysis. Concentrations ppm as ions Ca++ 82.0 Mg++ 12.2 Na+ 20.7 K+ 2.3 8003 235.0 SO4 40.0 01‘ 55.0 TDS 447.2 pH = 7.8 Alkalinity = 193 ppm as CaCO Hardness = 255 ppm as CaCO3 Silica = 8 ppm d. Side stream water temperature is 85°F, based on a condenser inlet temperature of 75°F and a 10°F temperature rise. e. Six parts per million of carbon dioxide remain after aeration. Initial calculations based on the use Of sulfuric acid for pH control demonstrated that the use of sulfuric acid was undesirable because of the rapid buildup of the sulfate concentration to scaling levels. Control of sul- fate concentration was possible only through the use Of a large excess of lime. Subsequent calculations were based 65 on the use Of hydrochloric acid for pH control since the salts of chloride are highly soluble, posing no scaling problem. Results for lime-soda ash side stream treatment, without reverse osmosis, at various Operating pH and effluent calcium levels, are presented in Tables 2 and 3. Operation of the soda ash precipitator to produce a water having 20 parts per million of residual calcium minimizes the size of side stream needed to be treated. The amount of waste produced and chemical demands are near minimum while the total dissolved solids value is the lowest at this calcium level. A further reduction in the waste stream flow rate and chemical demands can be realized by Operating the lime precipitator at a pH of 10.9. Operation at these conditions represents over 99 percent recovery of water from the blowdown and pro- duction of a waste stream 150 times smaller than the minimum blowdown resulting from no treatment. Using a lower pH could reduce the amount of waste produced but would increase costs due to increased acid requirements and the need for a larger treatment plant. The total dissolved solids value of approximately 47 thousand parts per million is totally unacceptable from a corrosion vieWpOint. As predicted, some means of soluble salt removal is necessary. Filtration of the sludge to recover more water is impractical because total dissolved 66 TABLE 2.--Lime-Soda Side Stream Treatment: Effects of Calcium Concentrations on Operating Conditions. Calcium Side Stream Waste TDS In Reagents, 1b/min in Cooling Flow Rate Stream Recycle Water , . . Lime HCl Soda Ash ppm gal/min lb/min ppm 40 5,443 788 47,190 8.1 60.5 24.6 30 5,275 788 47,190 8.1 60.5 24.6 20 5,077 791 47,030 8.2 61.0 24.7 10 5,336 825 47,590 9.1 67.0 26.5 7.5 5,548 850 48,040 9.8 71.5 27.9 Operation at (1) pH of 11.0, (2) 5% solids in precipitator sludge. TABLE 3.--Lime-Soda Side Stream Treatment: Effects of Operating pH on Operating Conditions. Side Stream Waste TDS in Reagents, lb/min pH Flow Rate Stream C;::::g gal/min lb/min ppm Lime HCl Soda Ash 11.5 5,000 1,087 50,640 16.4 113.5 40.3 11.4 5,008 997 49,790 13.9 97.5 35.5 11.3 5,016 926 48,980 11.9 85.0 31.8 11.2 5,027 869 48,240 10.4 75.0 28.8 11.1 5,043 825 47,570 9.1 67.0 26.5 11.0 5,077 791 47,030 8.2 61.0 24.7 10.9 5,222 772 47,040 7.7 58.5 23.8 10.8 5,463 764 48,400 7.7 61.0 23.7 10.7 5,626 751 49,910 7.6 63.0 23.6 10.6 5,741 737 51,270 7.4 64.5 23.3 10.5 5,828 723 52,430 7.2 65.0 23.0 Operation at (l) 20 ppm calcium in recycle, (2) 5% solids in precipitator sludge. 67 solids increases even more. Decreasing the water content of the sludge to 90 percent increases the total dis- solved solids of the cooling water to over 100,000 parts per million. The results of Table 4 are included here to show that lime-soda treatment alone can be feasible if there is a stream that serves as a blowdown for soluble ions. Such a stream is provided by pond seepage which is pre- dicted to be 1,400 gallons per minute for the Midland cooling pond. Sufficient removal of soluble ions is provided by this stream; therefore, no additional mechanical treatment or blowdown is needed. A minimum waste stream containing 56 gallons per minute water is produced by Operating the lime precipitator at a pH of 9.5. Since the bleed of soluble ions is provided by the seepage, the sludge could be filtered to remove most of the water without deleterious effects on the system. An economic evaluation may favor a higher operating pH with higher chemical demands but lower equipment capacity. For the range of pH values presented in the table, the required treatment capacity changes by over two million gallons per day. These were the only calculations for which a seepage rate was considered because it is believed that seepage is a blowdown discharging to higher quality 68 TABLE 4.--Lime-Soda Side Stream Treatment with Pond Seepage: Effects of Operating pH on Operating Conditions. Side Stream Waste TDS in Reagents, lb/min pH Flow Rate Stream CW::;:g gal/min lb/min ppm Lime HCl Soda Ash 11.5 5,820 903 4,160 13.0 102.0 32.5 11.4 5,867 832 3,860 11.0 89.0 28.8 11.3 5,908 777 3,610 9.4 79.0 25.8 11.2 5,946 731 3,410 8.2 70.5 23.4 11.1 5,985 696 3,250 7.2 64.5 21.6 11.0 6,037 666 3,140 6.4 60.0 20.2 10.9 6,090 638 3,100 5.8 58.5 19.1 10.8 6,124 611 3,080 5.3 58.0 18.2 10.7 6,146 587 3,060 4.8 57.5 17.3 10.6 6,181 569 3,050 4.4 57.0 16.7 10.5 6,223 556 3,030 4.1 56.0 16.3 10.4 6,269 544 3,010 3.8 55.0 15.9 10.3 6,319 535 3,000 3.5 54.0 15.6 10.2 6,371 528 2,980 3.2 53.0 15.3 10.1 6,428 522 2,960 2.9 52.0 15.1 10.0 6,486 517 2,940 2.6 50.5 14.9 9.9 6,544 513 2,920 2.3 49.0 14.7 9.8 6,602 509 2,900 2.1 48.0 14.6 9.7 6,658 507 2,880 1.8 46.5 14.4 9.6 6,710 504 2,860 1.6 45.0 14.3 9.5 6,804 503 2,850 1.4 44.0 14.3 9.4 6,938 504 2,840 1.3 43.0 14.4 9.3 7,094 505 2,830 1.2 42.5 14.4 9.2 7,279 505 2,820 1.1 42.0 14.5 9.1 7,505 507 2,820 1.0 41.5 14.7 Operation at (1) 20 ppm calcium in recycle, (2) 5% solids in precipitator sludge, (3) seepage rate of 1,400 gal/min. 69 ground water and as such will not be allowed by the new regulations. The results for the use of reverse osmosis in conjunction with lime-soda softening for side stream treatment of cooling water are presented in Tables 5, 6, 7, and 8. Table 5 shows that when all the side stream passes through the reverse osmosis unit and all the brine is blown down, a high quality cooling water is produced but the amount Of waste is increased by approximately an order of magnitude, reducing water recovery to less than 95 percent. Operation is optimized with respect to both the flow rate of side stream and amount of waste produced at a pH Of 11.3. The soda ash precipitator was considered to be Operated to produce a lower calcium concentration (10 ppm) in the stream entering the reverse osmosis unit. This allows production of a more concen- trated brine stream and hence a smaller waste stream. The use of reverse osmosis to control total dis- solved solids at any desired level is demonstrated by Table 6. At a pH of 11.3, only 3 percent of the side stream needs to be treated by reverse osmosis to maintain total dissolved solids in the cooling water below 20,000 ppm and to recover 99 percent of the water. Twenty thou- sand parts per million has been mentioned in the literature (5) as an upper limit for total dissolved solids in a cooling cycle. The amount of chemicals 70 TABLE 5.--Lime-Soda Side Stream Treatment Plus Reverse Osmosis: Effects of Operating pH on Operating Conditions. Side.Stream Waste Reagents, 1b/min pH Flow Rate Stream gal/min 1b/min Lime HCl Soda Ash 11.5 6,742 7,721 13.3 95.0 34.2 11.4 6,735 7,605 11.3 82.5 30.5 11.3 6,733 7,580 9.7 72.5 27.6 11.2 6,742 7,720 8.5 65.0 25.3 11.1 6,773 8,192 7.5 60.5 23.6 11.0 6,813 8,814 6. 58.5 22.4 Operation at (l) 10 ppm calcium in precipitator effluent, (2) 5% solids in precipitator sludge, (3) 100% of side stream to reverse osmosis, (4) 100% of brine to blowdown. TABLE 6.—-Lime-Soda Side Stream Treatment Plus Reverse Osmosis: 'Vari- ation of Operating Conditions with the Fraction Of Side Stream Treated by Reverse Osmosis. :EaSEQEn s::::m Flow Waste TDS in Reagents, 1b/min Rate to Cooling Stream Flow R.O. Stream Water Soda to R.O. Rate Lime HCl Ash gal/min gal/min 1b/min ppm 1.00 6,733 6,636 7,580 765 9.7 72.5 27.6 .90 6,760 5,997 6,932 840 9.8 73.0 27.7 .80 6,794 5,357 6,445 930 9.9 73.0 27.9 .70 6,822 4,707 5,945 1,045 10.0 73.5 28.1 .60 6,838 4,044 5,410 1,205 10.0 74.0 28.3 .50 6,839 3,370 4,831 1,430 10.1 74.5 28.6 .40 6,814 2,686 4,204 1,775 10.3 75.0 28.8 .30 6,748 1,994 3,514 2,355 10.4 76.0 29.1 .20 6,603 1,300 2,747 3,525 10.6 77.5 29.5 .10 6,276 617 1,884 6,940 11.0 79.5 30.3 .09 6,223 551 1,792 7,665 11.1 80.0 30.4 .08 6,164 485 1,698 8,545 11.2 80.5 30.5 .07 6,099 419 1,605 9,640 11.3 81.0 30.7 .06 6,024 355 1,510 11,035 11.4 82.0 30.9 .05 5,940 292 1,416 12,860 11.5 82.5 31.1 .04 5,844 230 1,321 15,335 11.6 83.5 31.3 .03 5,732 169 1,226 18,850 11.8 84.5 31.7 .02 5,600 110 1,133 24,180 12.0 86.0 32.1 .01 5,443 53 1,045 32,975 12.4 88.5 32.8 .00 5,252 O 964 49,480 13.0 92.0 33.8 Operation at (1) pH of 11.3, (2) 10 ppm Calcium in precipitator effluent, (3) 5% solids in precipitator sludge, (4) 100% of brine to blowdown. 71 required increases as the fraction of side stream to reverse osmosis decreases. Using brine recycle to reduce the amount of waste produced is illustrated by Table 7. For a rela- tively small increase of the chemicals required and an increase in the capacity of the treatment equipment, a significant reduction in the size Of the waste stream results. Brine recycle produces results comparable to partial treatment by reverse osmosis with the exception that a higher quality water is produced by using brine recycle. It appears that the use of brine recycle along with partial treatment by reverse osmosis is attractive. A limit to the amount of brine recycle allowable is the total dissolved solids in the precipitator, because too high of a value will disrupt the precipitation. Due to the inability to quantify this limit without experimental investigation, further Optimization with respect to this variable was not attempted. Comparable to the case with seepage, the total blowdown Of the brine stream provides the major bleed for soluble salts. Filtration of the sludge from the precipitator is therefore feasible to recover some of the entrained water and reduce the amount of sludge pro- duced. The total dissolved solids content of the circu- lating cooling water will not be adversely affected. TABLE 72 7.--Lime—Soda Side Stream Treatment Plus Reverse Osmosis: Effects of Brine Recycle on Operating Conditions. Percent Side Stream Waste TDS In Reagents, 1b/min Of Brine Cooling to Flow Rate Stream Water Blowdown gal/min lb/min ppm Lime HCl Soda Ash 100 6,733 7,580 767 9.7 72.5 27.6 90 80 6,754 6,459 777 9.9 73.5 28.6 70 6,757 5,960 784 9.9 74.0 29.0 60 6,771 5,437 791 9.9 74.5 29.4 50 6,773 4,797 799 10.0 75.0 29.8 40 6,851 4,285 803 10.1 76.0 30.7 30 6,836 3,630 819 10.2 77.5 31.2 20 6,938 2,894 823 10.4 79.0 32.3 Operation at (1) pH of 11.3, (2) 10 ppm calcium in precipitator effluent, (3) 5% solids in precipitator sludge, (4) 100% side stream to reverse osmosis. TABLE 8.—-Lime-Soda Side Stream Treatment Plus Reverse Osmosis: Effects of Sludge Filtration on Operating Conditions. Percent Solids in Side Stream Flow Rate Waste Stream Precipitator Sludge gal/min lb/min 5 6,733 7,580 10 6,707 7,179 20 6,694 6,980 30 6,690 6,913 40 6,687 6,879 50 6,687 6,860 Operation at (1) pH Of 11.3, (2) 10 ppm calcium in precipitator effluent, (3) 100% side stream to reverse osmosis, (4) 100% brine to blowdown. 73 The reduction in the waste stream size resulting from filtration of the sludge is demonstrated by Table 8. Minimization of the flow rate of the waste stream for no brine recycle occurs at the following conditions: a. Operating pH of the lime precipitator equals 11.3. b. Operation of the soda ash precipitator to produce water with 10 ppm of residual calcium. c. Filtration of the preciptator sludge to 50 percent solids. d. Treatment of 4 percent of the side stream by reverSe osmosis. The resulting waste stream flow rate is 475 pounds per minute consisting of about 100 pounds per minute of filter cake and 45 gallons per minute of brine. The total dissolved solids of the cirulating water is approxi- mately the limit of 20,000 ppm. Chemical requirements are: a. 12 pounds per minute of lime, b. 85 pounds per minute of hydrochloric acid, c. 32 pounds per minute of soda ash. This result was Obtained by choosing the best cases from the tables and independently varying the dif- ferent parameters until no more improvement could be made within the constraints of the system. Although these conditions lead to the production of the smallest waste stream--a physical Optimum--the high chemical costs may preclude this result from being the economic 74 Optimum. An economic Optimization requires the quanti- fication Of a cost for waste disposal, and would yield a result equivalent to the physical optimum only when disposal costs are very high. This could correspond to cases requiring ponding of the waste where land usage is at a premium or if the waste is evaporated to produce a solid waste form. V. DISCUSSION Conclusions The reduction of power plant cooling cycle blowdown by side stream treatment using lime-soda soften- ing and reverse osmosis is technically feasible. For the test case the maximum reduction of the blowdown occurred at the following conditions: a. Lime precipitator Operating pH equals 11.3. b. Operation of the soda ash precipitator to produce water with 10 ppm of residual calcium. c. Filtration of the precipitator sludge to 50 percent solids. d. Treatment Of 4 percent Of the side stream by reverse osmosis. The waste stream produced is less than .5 percent of the blowdown needed if no treatment is used. Almost 50 per- cent less makeup water is required due to recovery of water from the blowdown. The reduction Of the blowdown and water conservation afforded by this process are both significant. Greater reduction of the blowdown, makeup flow rate, and chemical requirements can be realized in some cases by using cooling towers instead of cooling ponds. Carbon dioxide stripping which occurs in cooling towers 75 76 maintains the total carbonate concentration relatively constant. If calcium carbonate is the limiting salt concentration for scaling, the number of cycles for a cooling tower would be larger than the number of cycles for a cooling pond. Less water then needs to be treated by the side stream process; ergo, chemical requirements are reduced. Recommendations Further investigation is required to determine the economics of the side stream treatment process. The data for the cases already developed can be used to find the Operating conditions that provide the economic Optimum. The economics may be enhanced by the use of recalcining as outlined in Appendix B. The feasibility of recalcining is basically an economic consideration. The high total dissolved solids possible for the side stream process is well beyond the range of the Debye Huckell equation and the range found in normal municipal water treatment. Experimental investigation of lime-soda softening for water of high total dissolved solids should be carried out. Based on the results of such experiments, limits for brine recycle could be established, and improvements in the mathematical model for activity coefficients and supersaturation could be made. 77 "Zero blowdown" was obtained in the sense that no waste needs to be returned to the water source. The production of no waste at all is impossible but is approached to a degree by side stream treatment. A better approach can be accomplished by reclaiming and reusing some of the chemicals. The alternatives for recycling chemicals described in Appendix B require further investigation to determine the feasibility of their application totflmaproblem of power plant blowdown. The next logical step to this research is the development of a total water management plan for the power plant, using the computer program to help evaluate the plan. It is possible that some of the waste streams could be used as makeup to the cooling cycle. By inte- grating the boiler feed and cooling cycle water plans to the greatest extent possible, better water conservation and usage and reduced blowdown or waste production could result. APPENDICES 78 APPENDIX A COMPUTER PROGRAM LISTING 79 80 .omao Emumoum cw mmumm 30Hm MOM H EMmHum DZ H. 1' poms mHOQE>m11.H.¢ musmfim m4— N Emmhum em. 5 Emonum m 41 29 mm OH Emouum mszm _ Hem HQm \ mwemmnum AMZHmmI3v N3 H3 am 4‘ IHIUWVJIWJ( .JV,V0(JL|(H. H meamHu Ba 14 4M .mw v Emonum \ m H H m smmuum Q 81 whommmm OH0< mom mkonooma >hHJHm340m whmzfiod Soars .onhdaom Jwycnr m>mmc th QZHm: rm ZOH 24 «on hzmHoHummoo >kH>Hho< m2» mmh<4304o no ammzaz m1: MmedJDUJOmmzm swan osz: wwapqwrazm» no zouhozou a ma mhzhH4Hcaeom owhmafioq Oh owhdqnuqao no loqm mxh mmpcqzugqo moamM >pH324 we «a mzmabmm oza zoahoaom a to >pHsampsuazompuunm or» myomxu cm.Uw3m .Ia zw>Ho a pa mhdhm OHO4 redw 2H ZOHZ< no onkoqou mm» mmk<4204FH>HFO< mom mpusnoaa >eH3Hm230m mFOmnwoo wou.ca .mmq capo 2H 0mm: mrowpozau cza mmszsnumnm .ozoa a mo «MIOP quJoou a mmzhHw .m40>o GZHJOOO w>demoaq>u 24 Zn xmhdz 92Hh<430mmu no hzmthumh 24 worocu u humunm wzmzpqmm» amend onhqvhzmozoo zDHUqu cmmvac v zzHoneo meme: 2H WEpaz p2mooma n zuwo omhqmwao mm 44H: rowapHuHomwa yr» Irmzz.»a :0 n anra hzmxeqmmh Iva «rcusmrHJ you chw7oo 7 non owm: oszm DHUq th uc ZQH2q mlk "UHmHkrmCH mHOdH awhdz a: M¥w4 moo muWu PWNHwMQ HIP wH mcoo szHdrszZ mm Oh MH amhqx m7wkaJDOWHQ HI» IOHI? kc In M In mMZOh d mom hmHmo mo fizcn a «on Widw uodouum n aumm mhcn 20HP¢NCQ¢>U u 0 A n mmmmowc w PZMXPQMQP P< Gth? no harhdtmufmk u k uqu mmmD mOWPmHSqum 02H» mzqmakm 00:10 03» oszHmroc.»u enamou rammem q 2H onedmpzmnzoo hr» umrcnzonao myocapm . thMHuHuumou >hH>Hkpq mlh mec42CJmmo MI» 2H .mm: mH mHIr orammkm a uQu IhOZLahm onoH thOH MI» no boom wnnfinv m1» whdJDQJqu mmw kHz: me02wo mmmm>mm mrn aou mzHcr mrh 7M 2¢wh<0»2wmzoo ZOH mrk cza onquFZmozom mo hzrnSq m1» mwhcqboaau 0w .monmu> oHuom to pzooxa my» :23 .ngkqpuoHOLQQ m1» 2H monwaahzwuzoo on 2H wozdru mt: wwwagronco dHowwa . . qu 2H mQZan owwwmmc wry make c» numnd mm or owommz mmdw do oHo< no #23024 er mwhdgbvgqo errors (11 ll (.J'C. L'( (.‘L 1‘ l . L. '(_.L (.L . ( (-3 C.‘ C ' :6 u\ 83 mpzahmzoo oHoq mrp «on «he: >onqxwzm mzHPH4HmDJCm Iodn Non wk7hzcaxm C2d o no mon m:» mmHu omam park yHmhqz zoueqowuHezmow zq mu 9H t . . nuke: o mom FZahmzoo onhazoa mH zom ems; mxh . m mezpH>HPOe m1» mg: m 2:3300 a . wnmxh omuowu uh cede >a3wwozm wmuu c we cupboazH . mpzqhmzoo OHzoozmmr» my» me: a zronoo p mhzqhmzoo >eH4Hmooom m1» mzmapzoo xam L mPZMHoHuumoo >kH>Heoa no mrchcaoonuo a mo buxooxomrmmo «on mmmpmzqmqa ONHm zch mrHqwaq mgaqhaas v mg L mzqmozw: .o.m no ozHooooo hzm>wma oe wepoau >wwuv:. rquwhm ermr A 32.0 o a: Myer 3.woe ..n.u.r4_.. 03 ..... upqwmh wc or rem wet ..n mpcy mxo.o mzsz 24(rtn ozchcnmuonou 024 Mm» t» acu mapcqu zompdwhruozo x» no mNHm mt» uc :onquxcoaaq u zm»m>m mzh 2H eroc no Wmmr3z .mpo3cowa >hw3Hr3nnm no umrzzz mzcmye no awar3z vromeao no Oumr37 :400 can rHar 7H mrqz KMHOJQ "3|! .1... the If .. 1 «2 oz MJm. mMOucro 26H24 andkzoo 4mm muowcic rempcn m2H5H4H4340m QZHDZOQmmmMOQ wrp Ihmz mm: Ch mm244> 4an mequszH yy nHud th wmpcszwmo wmqrbz 300 In zm>Ho 4 r4 20Hk4wocmvHo no woakm rodw 2H 20Hh4dhzwozoo 20H24 44hor no 20Hhu4wu mzp szqhzoo 4u44 f-n . L ‘3; (LC? f,. I (W (.1 (JC‘C. L- L t. ( . L ( LI (3 P’.‘ v4 L.) (\l H o«« 85 x.«yma.«mu.aum heru Noo=.raH04 Jthzu kzarao .qz.auH..H.«~H..H.<7oh..m.fius..n.H.zq...doa can. .uz.«uH..h.oNH..H.¢:oH..m.flu...w.H.xu...«oa ram. A run 2H . mHu>4qz< «wkcx a: wx‘r .zz.fiuu..m.dns..w.H.Iv..oa~ cuzoo fiuvoowow u «memo «utc.mm u wa4_ moxaonm.mn u DHocp .amtxnmos.n u «u Jqoxmmzaoa 0h 4\: you myopcqu chmaw>2oo Mp¢43u4qu PO\0mmNaom H mm Mt1k0\®MhJNooH u «a akfiovkmcm n to antfidhomimo¢N+NttahtOIwoowd+ahnmlmckmo30oavQJmomn u a .mmmlh u «k mbqumzo. 4wyosrum>mmo m»y.«ns Nd on .~.H.Jx u >2 :z.duH «a om MDZH 5200 ..N.H.xuu..a.m.ynm.p.xcw u .«pHvxam x.auH cu CL m.«\..noa¢h.u » mhzahmzoo Qch oza >hH4H¢340m mpaaaoaqc monh<4304m<2H1Hqua «m «a Na ca so“ omd L\ h. 04 and 87 .>rz:r.n.o.o.«.m.«.a.mIOUMhm JJqo mmhqx a: wxdt >420 mH w» mrsmmq « ZOthmmpH «on a: wxdt wao>o uzwaooo Jake» m1» Mn N zqmahm .acauu m a n xomIoH chkdmurh mxh no wmhzaoo ma XQMIUH c.9uhxp0h m o.oukow w c.ou«<»4mo h o.muq»4ua o.cu4440Hmam mH pH mtzmmq .o.w town hzqhqzmmanm me o« rqwmwm MNH4mPHJaykqmszomhom4m gauze maszzoc nomo.«..«.u.za\.«.H.zq u .«.H.4 42.~uH mm co 9") 14 NN (J omm Ln (‘J N omN mew cam mom ooN 88 .«om.dq»4mo.c.adxa.uzzuza Jaqo .ovomm u dam am o» oo ..«.F4.«w . uH . o xqwukm . pH you moqwoo mxHA mxh mp<4304nmm wzhmm mH wmmxh mm .pm.q»4me.m.¢4za.ozxoza .440 .m.omu u um n tamapm zH <4:a o» 0220 .x mquw o» moamoc mzHJ mpdgsoaqo moquHaHUMmo wcu mzCthJ:Qch. J c» om .o.o.n4.m.um N ok ow .Noo. .ko.p» .czc. n.uz.yow:c».um r»\.r»or»».mma u hp .2» mz» 2H mozou m:» mpqqncaccum mmoao poz wad >mx» kw .om»<4204o no quartz raruxaz mzh m»44:u.qu .>rznc.o.o.o.«.m.~.~.mzooahm .Juu 2» 2H mmorh o» ancw wonhthzmozoo MNHaquwzH qummpmmoum mxh om44.mu»<2 ozHaooo qupaqzumHu uz» mH m zdwmpm mathzou m u u uowro Io mozwcdm>on mH yuon In Moroqu MJO>O me ya MDZH»ZOL H QOOJ meZm L. L? mom omm u\ [I \ N omm ucm cam 89 ma-m n ran a ~z.~z u 3 dozumxoa u zomIoH H n HHH .mazupzou “ma.c.c.~a.m.¢.¢.mzouahm 444v . - cm.mm.mm .nH.uH mohchuanumma :mq «no» mrp 23a. pzmaquum mrh mg m zpwuanHuq c .>:z:c.o.o.o.«.¢.m.m.mzopapm 44¢c mm m :qmmpm o» >FH4<20 2H Jancm m rawMFm WNHJqHFHzH u own zqwmpm mszm MI» mH . quIFm u mHmoyuo mmmm>mw ”on monh«J:o.qu 8 mm a» co .u.m.o.mw.n.m.m.mzoumpw 44gb aw mm .» ow .o.o .ko.fiuu .8” man zqwcpm w4c>omw mm» mu 5 zqm.km u m :quhm u mozoo N zqumpm Pwm .ommn poz mH mHmczvo .m.m>mo uH u mazH»zgu nu «NJmc+m4wo u N48: Adm4mo.m.mcuo.mo»nqu .44u can .um.m4wc.m.ruvo2xoxu 4.4. mm o» oo Auo. .p4.moocxamoponmooc..gm .m.:.q..d.«.qu4a u Nouu .cm.zn.¢1aga 44cc .mvomm u om mam onhamuq mnmmz m40>umm w:» an yumzu L 91 . ..z»».az.m.~.~.s.mzooa»m 44Jmhazou 0p ooxbmz muZHMkmmuz um: . wnzuhzou .um.Iaon.h.ImvozroIA 44uwm mo Ia ~m350< 9.0 n Nihh a c.o n thh . mxwznmm u «mm .uMImv\Hm I “mm m\3¢a n 31am m\.mzmmmozaa. n 3m - - Hmomum .~.ou.zomruH.uH zouhme MIH wUZHW OMQZGIQ MdI mw40>u uC WWEZTZ wIk mu yuqu m o» 00 A H.»J.HHH . uH «we I .N.¢.U n AN.¢VU HMO + .N.n.< I “Noqu c.N\~wo I duo .Hw0.Nvoahquw 4440 maraon I “N.mv< I aN.m.< Zozw0\.a¢InJmo+.MZHIntamINN#>>I11VoJmaoaztad.m.4. I AN.MV< UDZHkZOU mm u 11 m Non I AND Amucm.om .HJmova H01 n mm a ~Jw3 u Jmo zozw0\..~.mvucnz. u Amomvu zozuo\aANNtAoqo>>vumvaJmo o .~.¢vuenz. n Amuevu onmo> Admtamlmv\.~mlm.tm I xx '5 arm m« o« If; 93 .n.a«mm.»hmuw4 N00 wwmu ad memwO 0\tzaat.xm “Nicu.tut.x««utwqo>cma 7H Jm>m4 ZDHquOctvhdtmou ham . ..au»qz oszagzomHo zH wzcwpamyzmgzoo 4¢2Hua..»uzmou mom .m.cau.. n mopoqm zoupacpzmozooo.\.mH.. zoupcmmkH «one..»qucu mom .u.¢«u.xo«..qu4 ‘xs.¢au.xad..oHoco.\.sz\moz:ca..xdan ..mooma .xs.:«u.x~«.nzmo.\ou amo ..m.mu.on mwn-q~wasz uzukqumaoO..»qrwou new .m.¢Hu.xwa.mm.xod.»a:~ou Now ...2an 2H. ozou..x¢..mzcz vow ..xp«.nzdvhqrucu “om ..am»qz a: quy 2H monkqnhzwozooc.\NH..u mN noncoma >hH4Hmaaom mo wwmzaz .me..u..xu«.moHoa no mmmxnz .xmn.»u.« .xaa..monzq mo mumzaz .me..u».xnd.»mzompqu no mmaxaz..dzd.pqyaou com mm: ow: U14 o1: 95 .om.«.aoohoq 444p 2,38,“. u um whouuum >»H>H»oa com whozooma >hH4Hmagom pommuoo .>zz:c.a.o.o.«. 55.5.q.mzoompw 49¢“ a zquahm 2H muzoo u «.5 rammhm 2H mozoo uNHAthzH 4-mo.« u > «+suss an: u xx MELH 52.0.9 o.o u .N.H.x I m.o u .H.H.x a u .H.quuH ;.HuH o$ on ommN he OQJuH rzq x mNquwthH naughm pzwnJIum mxh wH «.3 rqwupm ozaoaxoo :o4m no Jawxvozaoo m1» mqupzou «a posooma >~H4Hc240m ozchoomwmuou mm» 0» mac owoquo m>Hhcdzsu mrk mzngzop N 2:3400 zonpomuwoo pzwwasu uxh mqubzoo a 223400 mzoapqahzmozoo on mx» 0» 20Hpommaco m1» mzHapzoo x qz.oz.xr.z .ozroc x<.zo\mN\qon.conxaC\oH.vum\«w\q.o\da\zaxrou .m.=«.z<..m.od.zo..c~.om mahoqm mxk mH hqmunn Ikozmmpm cuon no Imam mxp mH cm xogu zqumpmmoam no zogg.w4.om .nza. 9.7m..2.o<4uH.uH 2H4maou>wma QMFQMumcu zmmm m>q1 monpqwk7mozoo mrh uH cza .o.»o.oa. hoaoomu m:» 2H ommakammmuam mH zoaprom my» uH voaxu am oh om .z.um.z.uH o.Ho.m.sn.z.yomum u om m»zc»mzco >hH4Hmr40m omya.ng.Hpqaww mxh urcADQJqo mm.«u4 mm co .Nw.»aoau u an m .NH.»404u u Hm .:.z.ouu~s m .m.z.oHu«sa .m.z.oHu~H m .z.oHu«H 7.Hu7 «¢ 00 puaooma >hHuHngov roam mo. VZOHkaaodqu on mnszzoo AN.ZVDHII»FH4Hmaqom hmzsoa o u pzDOUH zouhquwhm 24 mom mzonhthzmuZUQ 2H mmuzdxo no «szaou mxh ma hzzcoH cm.uu a4 om on 1004 w>Hbzou hoz .H.ma + kxko» .Jmcxymmmr.m . .H.aa ..~.as.z<..:.H.oH + .N.«H.xo..N.H.nH...~.pvx .n.H.oHu«q‘ e .« .chxmczsou 7H rJMH> rozcc mm 0» an .8.» .r.x+.sw.w.q u .33.“.q M .rvx..sn. o.~\.c-..m.r.ynv.nw..3-~..m _ .wa.d.c szHkqubzmcrog .10 are + Z.Unr szwu .mm a H. d & PTDCHH «o C» 00 . > .m4o:zvxvm.m<.07_c. s. .w pq‘dhmzoo zaHumHJHDcu NIH 0... mm. ._ w». Hw.fiz.x..nq.fiH.u - .ws.aH 3w...?,3x.12..amucu 225.4... wZOthubruorou ‘ r. 1nd 0 m H . r. (s ND vv [1‘ :7 )r .— n u a. .55.HH.0\am.z.x mm c» on .o.o.we vzcwquhzmorco nzu one ..\\.~armou tzthoo hypo» .vaa .H.wa .HvoHuHH zy.«uH m. 0: Ma uO MkIwrmZ 04.. u 5.3:...» war..:w.€.u mrthrcc 4.wz:cuw . uH ave u .fiw.«.c .wwom_ u .z.x + .fifi.«vo u m I HUI» buuumou mfiszzon mazuhzou .om.4.um A7.oqauH n hznoon 4.ow .uH :00 u:» xcwxo .ws.aa.u .aw.aw~u .vu th homuuon u .N.z.x uathzoo .z.oq.uH a u mzx» u .«.z.x .Nthqu aqzuoHno mt» u¢ om mm L. cc“ mw on em as 98 ....o2;.th2LU m n AfiuHuq u Afioqu 0.....NHH w." 0,: mnznwzom U t Afiuhvn u “flung” OTmNHH OH Cr. u mOthu m. Cr >1... a, Idmzkm 2.. MZCthwnkwlmozcu kowuwbo O rmapmm .u.a.u4 m.u~ ywo: kcz pm: at o.« zqzu wwwg mu m uH utxuu .Jmn.u.s.nopuau JJaD x >r puzwoa c2: u azHu «7.07.1xhz zozxcu ow.ynw\«mxrq.ro\-\q.p\"n\zorrou .a.o~.oH..m.om.xom..m.a«.zd.Am.a«.zn..ow.oa.d..o«.ofi.p zommxmer aqmuhm ufiwum ml» VuHuHhrwoh w ozHooOJU hzm>mou o» napuau >pmudu mxh mH x zoupqmahqm ow achoam onpcokzmuzoc m1» mH u awmq wmmhmzcuqc .s.x.u.oa hzwkaonmnm *4 m u‘. 99 NDZH»20L Laszzou zozuo\.w.sw.cu.a u .n.sH.qu44 ay.«ua mm on . mnthzom .«+y.s~.qu.u+rczmo u xozwn .a+x.o\:o\.y.o¢.fi+y.aw.«qu .«+y.sH.NHJq¢N4oozoaNom4m zoau zoNNmo MIN mH mo omxomzo oszo zamem or» muNuHszoN s .qu mmwpquaqn .mo.o.omNom4m mzupooycow mm J #6 .HN VL. t.) (.7 mu ow U‘u lOl 2w mDZthou .wmquVONH\4mc - .m.mmcaH~o u .fiowwcmva wZZH ~20 o onkHCCa mzH4o¥4¢ «on «a o» no .LHoaawdwwxamc o .fiNOHUQHvd u .s.cHochq zoNNquuHoHoq no" oo o» oo .o.o.o4.4woouh .owo.N.oaNom4N Joqo xc u .o.a.q m 1n n .s.o.o .om.1o.qzaoq ooco xaNAom..m.«.zo..«.oNH.NH>HNo< ozq oNoa «on mNnoconn >+H4Hm34om pmoooq mozHNzoo .>zxoo.o.o.o.o.«+o.o.o.mzoonpw Juno nan .oofi u n «-2 n z. qz.oz.zz.x 2ozzoo mmqu.cNoahNHNoH.ynmxdmxa.oNnnNzoxzoo .:.oN.oN..m.omoynm..oo.ofl.q..o«.o«.o zonzNzNo ‘ Jm>m4 monomn m:» #4 zn zHc «morn: mzp mH n owha mn+fin u mn (2902. 17.02 307.2un zc.roNNNNoH.ynmNomNa.oNnnNzozzoo .n.o«.z<..n.odoxo.N:.oN.oN..m.oN.ynm..on.o«.c..od.o«.o onwrmzN. mnqm zonn mNH mH on ozd zamwnm kz4>Hnnwnmnm mnNcw zonn nNmr» umq Nn.«n .omzmozoo oszo mzHnHszoH ~5.«s .mmq mamNmzqnqo .mn.~n.an.ro.No.«womxoomhm mzHNoonoow sz .Io.d .m4ou meDOuk ¢vhqrmou Mrzmhzo a Oh cw m u an em or 00 “do. oba.au\.ulduv .mmdunH Aaaoauolafioavdvouxaoaluv u 4mm 2:334 u 77:.de M 24.30 u 3+...odvu mDZHhZOU «« ca cam mm L! L" L k, a“ o: 104 I) f3 2 C) :13 .o.~Nox.Nnow u mzzwn N..AH.QNN..5.H.<.2: u or «ZNHHH mm 0c moanzoo n...H.oNH..n.H.onox n or oz.ouH «N on o.ou:z Irozmahm cH7CH Joke» mrk mung» mH a: :No2mnkm oNon napo» mpqooooqo 3: noun oz.oz.xx.r zorzoo «NH.oNH\«NN4.oNHnNzozzco .o«.4NH..on.oNH..oo.oo.<..oa.on.n onm2m2Ho .xNorwnpm onoN no Noam m1_ mbaooooqo on ommo mm on zqmmnn «I» no «nmrnz or» mH a nnNmzqnqn .ooowm onNozon LLLU; N 0-! v4 (\1 'u'\ 105 oz. .«x.z.xnmNynmonao u mommm Nfito.fi.«fiuc » NHIIAGNNHVO u xawoaqo .:.z.oHuNs m .m.z.oHu«s m .N.z.oHuoH » az.oNuHN oH.ynmN«qu.oN«nNzoz:co .a.om.oN..m.o~.xnm..o«.oa.a..o«.oaoo zofimxmzu. cmaONm nna nhqumzoo omNomwnoo >oywnomn MIN mnwx: xnm no zzonoc NI» mwNnNszmH on omwmonzoo oanm zqnwnm MIN meanzmoH a ommo oszm mH Noooonn >NHonzoom onIz manHNZMoN z .mmq wwmkmrqnqn .fly.o.z.~oonm zoNanon ozu . «12403:.o4 n 42:40 Aumtoaommoc.~.\N¢#chomc<nH4Hmooow nomnmon r. zorxco «NH.ONHNINNI¢.I0NNNNoH.yanamNzorzoo Noa.wn In wnqm MI» qunquI o» omomwz oHoq no ampHJNmnnoI wIN mH omo owNammo ozHuo Iamnnm II» manHszoH s zoapqmmq «Inn: gm>u4 Non nquoomuy mIN mH Noo .umq wnnnmzqman .ono.s.NooouoN NI» mNIJooJao ..ooNNNoa.«Nn«uo-.nxn u «yn mummowo ooN N4 nqumzou IIHImHJNIcm MIN qu4oo«o«4Inzm nIn m« Iowa mummomo cow Na onNIImon no uoquo >owmzm mmmn MIN mH n«mo . z«>«mx mmmmouo . nyopqamazmn whooommd mIn m« N .mna mmmnmIannn .Inmo.n4mo.».xcm onanon ozm w:z«»zoo .3...4.Js.«nnaoN.«I.I.Inm u .N.I.Inm «m o» oo .o .om.4 .mo. o.om.os . n« .N.I.¥Iu« « ..«.x.yyu4o m .4.y.o«u« .«x.x.ynm u .N.I.Inm I.«ux «m oo II.I zozzoo ¢n44.«¥NI«NIooa.xmenNo«.InmN«mNzoIIoo .I.m.«n4<..N.m.ox..~.owoxy..m.m.zooq..m.omoynm..3.oNoo« onmzmrNo omm: mm on Inm no zIoIoo II» mmnquoIH «I zo«»amvznozoo .I mI» m« moonI «mac mmmpmxanqn .«x.moan.oo Ksp To reach saturation some precipitate will form until equilibrium is reached. Mathematically it can be stated: + - II ([A l-X) ([3 l'X) = KSp Solving for X x2 - x.om oococ:m.md ococcnc.mm .Iaa 7H. 0 II I! H H a '0 O h. was N 0200 n¢~o nobodu >huum4 Nou wmmu ouxumuo m40>omm ZH 4w>m4 23H04m so.mm u mmthmmazwh uzmpquoo MOD JU éom X 42 07. 40 Mde ZOH ywhqz 1: wgqi Z“ mzowhqxhzmuzou whooooma >FH4Hmnuom no xmnzoz onu< no rxnzzz monZq do xmnzsz mzompqo no xmntoz .usmuso mHQEmm11.v.a mnmJOmmHo Jakob mmomcm¢.~=~ moo emmnsmm.sm¢- 4o omcwwoo.¢»~m «am mmnemco. To mmm~mca.>m~ x m46cm~>.c-c <2 mmsflcma.sm a: mnbpsom.oo_ «o mmococc. , I .216 zoo ozoo ozqz 23H Imqu 02HL