sign in sign up
<<
Home , Ammonium phosphate, Ammonium polyphosphate, History of fertilizer

AP42 Section: 3.2.1 Back roundCh 2 1- 1- 19 J. M. Potts, ed., Fluid , Tennessee Valley Authority, Muscle Shoals, AL, Bulletin X-185, September 1984. -hid iertilizers J. M. Potts, Editor

Tennessee Valley Authority National Development Center CONTENTS

WChapter 1.History. Growth. and Status ...... 5 0 Chapter 7-Suspension Fertilizers-Production Status of Fluid Fertilizers (Early 1980s) ...... 10 anduse ...... 86 Basic Studies ...... 87 J Chapter 2-Agronomic Aspects ...... 14 Production of Suspensions ...... 89 Storage of Suspensions ...... 97 Chapter 3-Physical Properties Evaluation 16 Transportation of Suspensions ...... 99

Chapter 4-Chemical Properties ...... 25 0 Chapter 8-Specialty Grades of Fluid Sources for Fluid Production ...... 26 Fertilizers ...... 103 Solubility Diagram for Macronutrients ...... 27 Influence of Production Methods ...... 29 8 Chapter 9-Application of Liquids and Characteristics of Solid Precipitates ...... 33 Suspensions ...... 110 Micronutrients in Fluids ...... 39 Stabilizing Fluids by Fluorine Chapter 10-Materials of Construction for Sequestration ...... 40 the Fluid Fertilizer Industry ...... 120 Hydrolysis of Polyphosphates in Fluid Fertilizers ...... 42 Complex Formations ...... 44 Summary ...... 46 Goals for Future Research ...... 46

Chapter 5-Production. Handling. and Use of Anhydrous ...... 49 Production Processes ...... 49 History ...... 49 ...... 50 Raw Materials for Hydrogen Production ...... 51 Process and Design Details ...... 52 Frank P. Achorn .... 76.86. 120 Ammonia Production Economics ...... 60 Gary W . Akin ...... 14 Waste Disposal Problems ...... 60 H . L. 8alay ...... 16 Annual Production of Anhydrous G . M . Blouin ...... 49 Ammonia ...... 62 Michael F . Broder ...... 110 Anhydrous Ammonia Handling ...... 63 Lawrence C. Faulkner ...... 76 Anhydrous Ammonia Retail Operation ...... 63 A . W . Frazier ...... 25 Use of Anhydrous Ammonia in Agriculture ...... 65 J . G. Getsinger ...... 86 History and Evolution ...... 65 Norman L . Hargett ...... 5 Role of Ammonia as Fertilizer ...... 66 George Hoffmeister ...... 86 Applying Ammonia ...... 68 H . R . Horsman ...... 120 Ammonia Retention and Adsorption T . M . Jones ...... 16 in Soils ...... 72 H. L. Kimbrough ...... 49 Eugene C . Sample ...... 14 Chapter 6-Liquid (Solution) Fertilizers ...... 76 W . J . Sharratt...... 49 Aqua Ammonia ...... 77 Eugene 8 .Wright. Jr ...... 103 Liquid Solution Fertilizers ...... 77 Ronald D . Young ...... 5 FIGURES

Poge Page 1 Growth in production of fluid mixed 27 Liquid fertilizer mix plant ...... 81 fertilizers. U.S...... 6 28 Basic parts of TVA gelometer ...... 87 2 Diagram showing structural formation 29 Production of 13-38-0 of polyphosphates ...... 8 orthophosphate suspension fertilizer ...... 89 3 Viscosity measuring device ...... 19 30 Production of 9-32.0 4 Mix of MAP and DAP crystals ...... 20 suspension fertilizer from merchant-grade 5 Effect of NH, :H3PO4 mole ratio on wet-process ...... 90 solubility of ...... 22 31 Production of urea. 6 Effect of pH and polyphosphate level on >uspension fertilizer (31% N) ...... 91 N.P, OS weight ratio of ammoniated ortho 32 Production of urea-ammonium nitrate and superphosphoric acid soiutions ...... 23 susper~sio,I Iri tiiizei {36% N) ...... 92 7 Relation between aluminum and residual 33 Production of urea- fluorine contents of ammoniated superphosphoric suspension ('29-0-0.5s) ...... 93 acids prepared from wet-process acids ...... 28 34 Mix plant for fluid limestone ...... 95 8 System: ammonia-superphosphoric acid-urea- 35 Micronutrient and clay injector for fluids ...... 98 ammonium nitrate-potassium chloride-water 36 Pipe sparging system for suspensions ...... 99 at32'F ...... 28 37 Cone bottom suspension storage tank 9 System: ammonia-superphosphoric acid-urea- with air sparger system ...... 99 potassium chloride-water at 32°F ...... 28 38 Transport truck with pipe sparger for 10 The system NH3.Ki0.HaPz0,. HzOat 25°C suspension fertilizers ...... 100 projected on the (NH,),O.K, O.(PzOs + H, 0) 39 Cone bottom transport ...... 100 face ...... 29 40 Spherical transport tank ...... 101 11 Solubility in the system MgO.NH,.H,PO,. 41 Tilting-type transport for suspensions ...... 101 H4P,0,.H,0at25'C ...... 34 42 Solubility in the system KzO-PzOs-HzO 12 MgO saturation levels vs polyphosphate content at32'F ...... 103 and pH in liquid fertilizers ...... 34 43 High flotation applicator ...... 111 13 Solid phases in ammonium ortho-pyrophosphate 44 Flow diagram for fluid fertilizer applicator liquid fertilizers saturated with ZnO ...... 35 that uses pump bypass to control pressure 14 Coal-gasification and gas-purification unit and at nozzle ...... 111 gas reforming ammonia piant ...... 53 45 Nozzles used to broadcast and apply 15 TVA ammonia from coal project ...... 54 fluid fertilizers ...... 112 16 This photo shows how anhydrous ammonia was 46 Spray patterns ...... 112 first used at the Mississippi Delta Branch 47 Liquid fertilizer-piston type metering pump .... 113 Experiment Station ...... 67 48 Squeeze pump with round hoses and back plate . . 114 17 fertilizer consumption ...... 69 49 Squeeze pump with flat hoses and hydraulic 18 Consumption of nitrogen fertilizer materials .... 69 drive ...... 114 19 Front swept knife for ammonia application ..... 69 50 Anti-pollution device on fertigation system .....114 20 Various type knives for ammonia application .... 69 57 Nozzles used for dribble or surface band 21 Tractor drawn ammonia applisation equipment . . 70 application ...... 115 22 Effect of soil moisture content and applicator 52 Flow divider for suspension injection ...... 116 spacing on ammonia retention in the soil ...... 71 53 John Blue flow divider for row application 23 V-blade ammonia applicator ...... 71 of liquids ...... 116 24 Effect of polyphosphate level and N:P205 54 Constant-head gravityflow fluid fertilizer weight ratio on solubility of ammoniated applicator and orifice plates ...... 117 phosphoric acids at 32°F ...... 76 55 V-blade equipped with ammonia sparger 25 TVA continuous urea-ammonium nitrate and liquid phosphate injection assembly ...... 118 solution plant ...... 78 56 Dual application knife with ammonia 26 Typical pipe-reactor plants for production and phosphate tubes separated ...... 118 of ammonium polyphosphate solution ...... 80 57 Ammonia knives equipped for dual application . . 119 Page Page 58 Support for PVC plastic pipe ...... 123 62 PVC lined storage pit with floating PVC 59 Neoprene-lined nylon bag (capacity: 500,000 plastic cover ...... 129 gallons) ...... 124 63 Evaporative-type cooler (from refrigeration 60 Transport trucks equipped with plastic tanks ....124 plant) ...... ,129 61 PVC lined earthen phosphoric acid storage pit ...128 64 Evaporative cooler (with plastic packing) ...... 129

.

TABLES

Page Page 1 Chronology of fluid fertilizers ...... 8 19 Summary of micronutrient content of various 2 Selected data on U.S. fertilizer use ...... 11 inorganic micronutrient sources ...... 82 3 Typical physical properties of various 20 Micronutrient content of some organic solution fertilizers ...... 23 micronutrient sources ...... 83 4 Typical properties of various suspension 21 Solubility of micronutrients, 24-hour agitation ... 84 fertilizers ...... 24 22 Viscosities of 5-15-30-XZn mixtures ...... 97 5 Stable compounds in the ammonium potassium 23 Addition of mixture of micronutrients to polyphosphate systems ...... 30 15.15-15 suspension ...... 97 6 Impurities in 11-33-0 prepared from 24 Concentration of plant nutrients in the system wet-acid MAP ...... 32 K,0-P,0s-H20at O°C ...... 104 7 Significant components affecting the chemical 25 Low-chlorine fertilizers made from and physical properties of ammonium superphosphoric acid and potassium hydroxide . . 105 polyphosphate fluid fertilizers...... 33 26 Effect of source of potash on maximum grades . 8 Reaction products of iron with ammonium of liquid fertilizer made with superphosphoric phosphate fluid fertilizers ...... 36 acid ...... 107 9 Reaction products of aluminum with ammonium 27 Production of 0.26-25 potassium polyphosphate phosphate fluid fertilizers ...... 37 liquid fertilizer ...... 108 10 Reaction products of calcium with ammonium 28 Percentage of plant foods in saturated solutions at phosphate fluid fertilizers ...... 38 different temperatures of potash salts used or 11 Reaction products of magnesium with formed in the formation of fluid fertilizers .....109 ammonium phosphate fluid fertilizer ...... 38 29 Suitability of AIS1 type 304 stainless steel and 12 Micronutrient precipitates in fluid fertilizers . . , , 39 mild steel for use in nitrogen solutions ...... 121 13 Comparative energy requirements for 30 Thermal and physical properties of PVC processes using hydrogen ...... 58 and CPVC compared to steel ...... 122 14 Estimated ammonia plant prices (1982) 31 Suitability of various alloys for use in wet-process using four basic feedstocks ...... 61 orthophosphoric acids ...... 125 15 Anhydrous ammonia production and capacity 32 Suitability of various alloys for use in wet-process intheU.S...... 62 16 Consumption of selected nitrogen direct superphosphoric acid ...... 126 application materials-United States ...... 68 33 Suitability of nonmetallic sheet materials. 17 Nitrate-nitrogen content of cotton petioles liners, and coatings for use in wet-process after N applied in irrigation water ...... 72 orthophosphoric and superphosphoric acids , . . , . 127 18 Physical and chemical characteristics of three 34 Mechanical properties of steels used in nonpressure solutions ...... 78 stress corrosion cracking tests ...... 129 Chapter 1 History, Growth, and Status By Ronald D. Young and Norman L. Hargett

Liquid (fluid) fertilizer production and use is generally the solution specified bears some resemblance to the liquid regarded as a relatively new (mid;20th century) practice, mixes made todav (1). which is true if we are thinking ofcominercial significance. But like many other developments popularly regarded as Early Commercial Production new departures, the beginnings of liquid fertilizer go far back in history. Possibly the first U.S. plant for producing a liquid fertil- izer was the G&M Liquid Fertilizer Company, built in Early History Oakland, California, in 1923.This was followed by the work of the Prizer Brothers who in 1928-30 developed the Possibly tlle earliest recorded use of a fertilizer material “Prizer Applicator,” a machine for dissolving a solid fertil- in liquid form, if we broaden the definition of fertilizer to izer and metering the resulting solution into irrigation water include the nimures, was the practice in ancient Athens of through either a gravity-flow or pressure system (2). fertilizing truck gardens and orange groves by city sewage Ammia nroduction began in this period in the United carried in canals. Various other mentions of liquid manure, States. Plants were built in Syracuse, New York, in 1921 by and of other liquids used to promote plant growth, are a subsidiary of Allied Chemical Corporation; at Belle, West found in ancient histories. Virginia, in 1926 by E. I. du Pont de Nemours and Company; Much later, but still quite early in the history of fertilizer and at Hopewell, Virginia, in 1928 by Allied. With a technology, Sir James Murray in 1843 referred to the domestic supply of ammonia available, the liquid fertilizer “new” practice in England of supplying fertilizers in solid industry was in position to move forward at a faster rate, In fomi. The new type, he said, was “exempt from loss by 1932, Shell Chemical Company began research on intro- leakage, more safe, portable, etc.,” as compared with the duction of ammonia into irrigation water and followed previous liquid products, normally sold in 30-gallon casks. this by injection into the soil in 1942 (2). In about the In the same year, Bishop in Scotland reported that aqua same period, ammoniating solutions were developed for ammonia gave improved yields of grass. Subsequently, gas use in making solid fertilizers. For example, in 1932 huuse liquors and other wastes were sold in liquid form. du Pont . first introduced urea-ammonia-water solution, One of the first references to use of a fertilizer in liquid composed of urea, ammonia, , and fmn in the United States was by Joseph H. Webs1er.A water. Such solutions, containing ammonia with urea or ammonium nitrate, were likely candidates for direct appli- “resident of Cumberland, in the County of Webster and cation, and the first application of a nitrogen solution took State of Mississippi,” who in 1899 p&e.nted a solution place on a cane field in Louisiana in 1942. “consisting of water, saltpeter, sal-soda, blue-stone, nitrate

~~ ~ ~ There were problems in handling and application of ammonia, and potash.” He did not use this solution directly solutions that were under pressure and subject to emission as a fertilizer but sprinkled it over a mass of manure, of ammonia fumes. This led to research aimed at the “ashes, , lime, phosphate, kainit, and cotton-seed development of nitrogen solutions without any vapor -_meats.” However, his proposal deserves some note because pressure of ammonia. In 1942, Allied Chemical Corporation 5 began testing a solution of urea and atnmoniuni nitrate, solution and solid urea were used to give such products materials that have suchsutual solubility effect 3s to as 9-9-9, 12-6-6, and 12-8-4 (1). allow nitrogen cuneentrations as high as 32%. This work was interrupted by the war and was resumed in 1950, whell Two Dramatic Decades the first farni application took place. Prior to this, ammonium nitrate solutions were used in California as early The preceding discussion brings us to 1955, at which as 1Y44. time botlimliquids and liquid mixed (NPK) fertilizers Liquid mixed fertilizers, after the start in 1923, grew were strurrlina._- industries. They were striviw to brso very slowly. The practice was of necessity based on dis- an established, cunserwrive. industry with a drastically solution of solid materials because little or no phosplioric difrercnt product forin. The need for a different marketing acid was available. As late as 1943 less than 1_000short system and for providing additional services, including tons ofliquid mixed fertilizer was reported sold in California. custoiii application, were already apparent. The early The high cost of the raw materials restircted liquids mainly pioneers liad both tecliiiical and econimiic probletiis, and to special uses, such as foliar spraying, flower and garden sonie of the coiiipatiies did iiot survive. 111 a few years, fertilization, and correction of unusual soil deficiencies. however, the liquid products began to gain acceptiiuce and In 1937, Shell Chemical Company demonstrated the use increased rapidly. The rate of growth froiii 1955 to feasibility of adding phosphoric acid directly to irrigation about 1975 was greater tliaii for any fertilizer in :iny similar water, and in 1940 the first commercial use of the method period in fertilizer history. took place, Other pioneers in use of phosDlroric acid were Probably.the first review of liquid mixed fertilizer pru- the John Pryor Company (Salinas, Cdifornk) and the duction and use was published in 1955 (4). About 147 Hayward and Wood Company (Oxnard, California) (I). companies reportcdly were oper~tingill tlie field. but In the 1940s, phosphoric acid-mainly the furnace only 72 appeared to bc iniakitig bulk fertilizer for farm use. type-became more available in California and production Only 25 were reported east of the Rockies and most of of liquid mixed fertilizers began to increase. The Leslie- these had gone inti1 production in 1954.55. US. production Agriform Company, later to become a major producer of in the year ended Julie 30, 1954, was only 27,500 toils. liquid mixes, started operation about 1946 in Newark, The dramatic growth in fluid mixed fertilizers duriiig the California. This company originally handled all its product period of 1954-75, both in number of prnducing units in 15- and 30-gallon wooden barrels, and distributed and in productioti of material, is sliiiwii in figure I. solution as far as Phoenix, Arizona. The barrels were The approximately I nil lion totis uf liquid iiiixed icrtil- replaced a few years later with tanks of 1,000- to 2,000- izers cunsunied in 1965 represented ;ibuut 1% of tlie total gallon capacity. Corrosion was a severe problem and raw mixed fertilizers consumed in the United States. Production material costs were high. However, production by Agriform for the approxiniately 1,200 fluid plants averaged less tliiui and others increased, tu the point that by 1953 more than ~. cum JMO 22,000 tons of liquid mix was produced in California: this represented about 9% of all mixed fertilizer made in California. In other areas of the country, Cominco Products, Inc., began making an 8-24-0 solution at Trail, B.C., in 1955. Tlus was a new departure, because Cominco made the 8-24-0 from lower cost wet-process phosphoric acid (in conjunction with ammonium phosphate production), whereas the production in California was based on furnace- type acid. Much of the Cominco product was sold in Washington and Oregon. In the middle and eastern parts of the United States, the only known production of Liquid mixes prior to 1953 was in the form of specialty liquids made from soluble salts and usually packaged rather than sold in bulk. Startup of the Liquilizer Corporation plant at Vincennes, Indiana, in' 1953 marked the beginning uf the new practice- neutralization of phosphoric acid with ammonia and dis- solution of potash in the resulting solutino. The first - hot Figure 1. Growth in production of grade made was a 4-10-10 but the next year ammoniating fluid mixed fertilizers, U.S.

6 1,000 tons per year. The next decade (1965-75) saw even Statcs produced suspensions that accounted for more than more rapid growth in production and use of fluid niixed 40% of the fluid mixed fertilizer market. fertilizers. Suspeiisioii fertilizers, fluids in whicli some In the late 1950s and early 1960s, there were substantial particles aresuspended rather than dissolved, came into the efforts to break through the grade (solubility) ceiling picture alter 1963 when TVA sliie&LcfksLtau~ imposed by liquid (solution) fertilizers of good quality. 12-40-0 base susDensi5offering considcrably higher A main effort was through fluids that contained a large atialvsis than liuuid (s olutiod fertilizers, Production of part of their nutrient content as finely divided solids. In suspensions bccaine substantial by the late 1960s and thc early years such fertilizers were commonly referred their production and use have grown steadily since. Suspen- to as slurries. Later the term suspension came into use sioiis contributed substantially to fluid production by 1970. for higher analysis fluids that were much more dependable The number of iiixed fluid plants in the United States in storage, handling, and application. It may be well here to grew froni 1,200 in 1965 to about 2,800 in 1974. Total differentiate the terms susnension and slurrv.

~~ ~ of nutrient consumption as mixed fluids was only 143,000 wensionSuswension7 Ifertilizersfertilizers ~~ are fluid mixtures solid and tons (N t P,O, + K,O) in 1960. This increased sharply liquid materials in which the solids do not settle rapidly 10 823,000 tons in 1970 and to about 1.3 million tons '%d%d can be redispersed readily with agitatinn to give a_a in 1975. Total nutrient content of mixed fluids (N t uniform mixture. Certain types of clay are added as sus- P-0, + K,O) was estimated to be about 28% in 1960. pendinEagents.-_ The suspension is fluid enoughenoughto- to be pumped.. Gruwtli of tlic liiglier analysis suspensions resulted in an and applied to the soil in application equipment commonly increase to about 34% in 1974, and to about 35.2% total available for liquid fertilizers with little alteration of the plant food content in 1981 (5). eouiometit. Growth in number of plapts slowed, but the total reaclied about 3,000 in 1976 and 3,200 in 1980. The total tonnage 11f mixed (NPK) fluids increased to about 1.8 million tons of nutrients (N t P,O, + K,O) in 1981. This Continuous agitation is required to ensure a uniform represented about 5 milliiin tons of fluid fertilizer materials. mixture that can be pumped and applied to the soil. Com. The 1.8 million tons of nutrients as mixed fluid fertilizers nionly available application equipment usually needs some in 1981 accounted for about 7.5% of total nutricnts applied modification to handle this type of product successfully in the United States, or 17% of total nutrients in mixed (1). fcrtilizers. Beginning dates for suspension fertilizers are difficult to T. P. Hignett in his Sixth Fraiicis'New Memorial Lecture identify, for various innovators have experimented with in 1969 (6) gave a good assessment of the advantages for suspensions from time to time and produced material for liquids. distribution. The use of base suspensions presumably began Liquid fcrtilizcrs arc dcpcndably frcc-flowing. Thcy arc when TVA shipped the first car of 12-40-0in January 1963. wcll adaptcd to handling and application by mcchanized, This pioneering work by TVA, and continuing research, labor-saving Inicthods. Prccisc inetcring and placemcnt and devclopment, and demonstration work with industry during uniform distribution arc casicr with liquids than with solids. the next two decades, led to suspensions reaching a status Liquids are IIOIIIO~C~~OUS,frec frum dustincss or caking, and unaffected by hygroscopicity. Thcy are fully water in the United States approaching that of clear liquids soluble. With liquids, thm~are no sacks to lift, open, and (solutions) by 1980. disposc uttlicrc is littlc delay in rcfiling applicators. Liquid There were many problems and false steps along the way, fcrtilir;itiion inlay be combined with irrigation 01 with appli- and some people became discouraged. But the clear-cut cation uf hcrbicidcs or pesticidcs. advantages of suspensions spurred researchers, technologists, These advantagcs and others, coupled with the high and industry innovators to find ways to deal with the analysis and lower cost of suspensions, have led to the problems. Here are some important main advantages for dcvelopment of a dynamic and progressive segment of the -suspe&QI& ,. fertilizer industry (7). The practical limit in concentration is much higher than for clear liquids (solutions), particularly when History of Suspensions potash is a component. Solubility is not a limiting &or. When the N:&O,:KzO ratio is l:l:l, the The highlight of the decade vf lhe 1970s for fluid fertilizers was the continuing vigorous growth in produc- suspension grade (14-14-14) is twice that of clear tion and use of suspensions. Growth was continuing in liquids. 1982, but in fact, suspensions were the only type of Adequate proportions of secondary and micro- fertilizer showing real growth into the 1980s. By 1981 nutrients can be readily incorporated, as can pesti- a~r170sthalf of tlic 3,200 fluid fertilizer plants in the United cides, herbicides, etc.

7 ic Materials commonly used in formulating suspensions The building block for , based on inoderii generally cost less and are more widely available tlieuries of electronic structure, is the PO, tetrahedron- than those required for clear liquids (8). a atoni surrounded by four oxygen atoms at The growth of suspensions was spurred in the mid- tlie four corners of a tetrahedron. Multiple units (polyniers) 1970s by the develop~nent of equipment and technique of PO4 tetrahedra joined by shared oxygen atoms between for readily incorporating solid intermediates such as niono- the tetrahedra are called condensed pliospliates. The ammonium phosphate (MAP), diarumunium pliusikate simplest cundcnscd pliosphate is tlie pyrophosphate formed (DAF’), and nitrogen materials as major components in by tlie sharing of oxygens between two PO4 groups. Poly- formulations (9). These solid intermediates were available phosphates are linear cliaiii polymers that may contain in unlimited quantities and were substantially lower in cost from two to hundreds of PO4 groups. Branching probably than high-grade wet-process acid and polyphosphates occurs in chains of niore than about 500 PO, groups. required for high analysis liquids (solutions) (IO). Polypliosphates were reported in the literature as early Suspensions are likely to continue to grow, and because as 1816 when Berzelius heated ordinary phosphoric acid of important advantages, may ultimately dominate the fluid and showed that the products were different froin ortho- segment of the US. industry. Slurries, on the other hand, phosphoric acid. They were considered to be iiiixtures have essentially disappeared from the market. of pyru- and metapliosphates and tlie idea UT tripoly- phosphate as a crystalline entity was not accepted until c Highlights of Fluid History 1940. Proof of tlie existence of a Iiomologous series of Table 1 lists a chronology of the history, development, and growth of the fluid fertilizer industry. This listing is primarily for the United States, but developments of lesser scope occurred in France, Belgium, the United Kingdom, and a few other countries,

+--ORTHO+ Role of Polyphosphates e?, - -PYRO- .\ I TRlPOLY, Polyphosphates are the “backbone” of clear liquid -LONGER CHAINS __ (solution) fertilizers. Polyphosphates are formed by removing (or eliminating during production) chemically Figure 2. Diagram showing structural combined water from orthophosphate molecules (figure 2). formation of polyphosphates

~~~ Table 1. Chronology of fluid fertilizers Fertilization of truck eardens and orange groves in Athens by city sewilge in canals. II 400 B.C. ~ .. - 1843 Sir James Murray referred to “New” practice of using solid fertilizers. 1899 First liquid fertilizer patent obtained in U.S. 1923 First liquid mixed fertilizer plant in California. 1942 Injection of anhydrous ammonia into soil introduced (now a major fertiltzation practice of lowest cost). 1953 First liquid fertilizer production by neutralizing phosphoric acid with ammonia and dissolving potash. 1955 147 liquid fertilizer plants in the U.S. 1956 Superphosphoric acid introduced by TVA (now the “backbone” of liquid fertilizer production). 1963 Suspension fertilizers emerge in higher analysis (grades such as 15-15-15 and 7-21-21 practical). 1966 TVA develops wet-process superphosphoric acid. 1967 About 1,500 liquid and suspension fertilizer plants in U.S. 1968-70 Fluids developing slowly in France, Belgium, and United Kingdom. 1972 TVA pipe reactor process for high-polyphosphate liquids introduced (also a boon to wet-process superphosphoric acid production and use). 1974 2,800 fluid fertilizer plants in U.S. 1975-76 Use of solid MAP, DAP, etc., spurs growth of suspensions. 1976 3,000 fluid plants in US. 1980 3,200 fluid plants in US. 1980 Fluids (including NH3 and N solutions) = 32% of U.S. fertilizers. 1981 More than 2 million tons of UAN solution used in US.

1981 Suspensions reach~ 40% of total fluid mixed fertilizers.

8 chain phosphates (polyphosphates) came only after the phosphate) in the pipe produces a melt of high polyphos- development of paper chromatography in tlre early 1950s. phate content that is dissolved in water and ammoniated Tlie progressive removal of water from orthophosphoric further to produce 10-34-0 with 60% to 75% of the P,Os acid as shown in tlre diagram produces a series of poly- as polyphosphate. phosphoric acids. The polyacids are obtained also when The simple energy-conserving TVA pipe-reactor process P,Os from burning of phosphorus is hydrated with water caught on like wildfire in the industry. By 1980 at least short of the requirement for formation of orthophosphoric 130 commercial plants were using this “revolutionizing” acid. process (16). The pipe.reactor process also revolutionized Mixtures of orthophosphoric and polyphosphoric acids the production of wet-process superphosphoric acid by are called superphosphoric acids (1 1). industry, since acid of only 68% to 69% P, Os content is needed. This acid can be readily produced in a single-stage Sequestration of Impurities stainless steel evaporator that requires only steam for heating. The much lower viscosity of this “low-conversion” Sequestration of iInpl! rities in ammoniated wet-process superphosphoric acid also greatly simplifies its shipment / phnsphoric acid is an important phenomenon in the tech- 2nd handling. It bccamc a commodity in world tiade by _Xoogy of preparation of c-1 Ition) fertilizers of 1979 (17). high quality. This is the process whereby polyphosphate form stable, soluble complexes with metal ions and Higher Solubility With Polyphosphates hold tlrenr in solution. ,Aninroniuni polyphosphatesare good sequestering agents and are added to wet-process phosphoric The polyphosphate system provides gieater acid to hold in solution the @ and -m which solubility, and higher grades can be produced as liquid otherwise would precipitate as orthophosphates when the (solution) fertilizers. In the orthophosphate system, wet-process acid is ammoniated. Ammonium pyiophos- maximum solubility at 32’F is 39% total nutrient content phates are more effective than the higher polyphosphates (N t P,O,). With about half of the Pz05in polyphosphate in sequestering iron and aluminum in fertilizer solutions, form, the maximum solubility increases substantially to but the polyphosphates of longer chain length are better 46% total nutrients (N t P,O,). Wet-process superphos- sequestrants for c-m and maenesium, phoric acid is widely used to produce 10-34-0 ammonium Tlie sequestering power of ammonium polyphosphates polyphosphate solution (60% polyphosphate) that is a also aids greatly in production of liquid fertilizers containing major commercial intermediate in production of finished significant anrounts of copper, iron, manganese, and zinc. liquid (solution) fertilizers. Two-component fertilizers with These important micronutrient elements are almost grades such as 18-18-0, 24-8-0, and 22-1 1-0 can be produced insoluble in liquid fertilizers made from ammonium orthu- using 10-34-0. Lower solubility in the orthophosphate phosphates, system limits N:P,O, grades to 14-14-0, 18-6-0, and TVA has acquired inost of its knowledge of ammonium 18-9-0. When NPK grades are produced the polyphosphate polyphosphates since the introduction of electric furnace system offers no grade advantage because solubility of KCI superphosphoric acid by TVA in 1955 (12). is limiting; the grade for a 1: 1 : 1 ratio is 8-8-8 in either TVA’s pioneering research and development work on system. ammonium polyphosphates soon led to 11-37-0 ammonium polypliosphate solution (made by ammoniation of super- Role of National Fertilizer phosphoric acid) that was provided to the industry for Solutions Association use in formulating high-quality liquids and for sequestra- tion of impurities in anrnroniated products of wet-process The National Fertilizer Solutions Association (NFSA) is orthophosphoric acid. the main trade association for the fluid fertilizer industry, TVA followed with the development of much more with member companies in the United States and 14 other economical wet-process superphosphoric acid in pilot- countries. It is dedicated to advancement of all aspects of, plant work in 1961 (13). The industry followed in the fluid fertilizers. This very active organization was established mid-1960s with preparation of this acid by concentrating in 1954-early in the history of the fluid industry, when wet-process acid to 68% to 72% P, 0, content (14). there were less than 100 plants that produced mixed fluid The real boost for wet-process acid-based polyphosphates fertilizers. Membership grew to about 1,700 companies in came about 10 years later in 1974 with development by 1983, but NFSA remains fully as dynamic as during its TVA of the simple pipe reactor and process for preparation formative years. of 10-34-0 ammonium polyphosphate solution (15). With headquarters in Peoria, Illinois, NFSA provides a Ammoniation of wet-process superphosphoric acid of only range of senices to members. These include assisting with 68% to 70% PzO, content (20% to 30% of P,O, as poly- the transfer of technology and information concerning the

9 production, marketing, application, economics, and agro- Palgrave (20) has published an interesting account of nomic aspects of fluid fertilizers. They also coordinate with the history of fluid fertilizer production and use in the other organizations, including governmental regulatory United Kingdom. agencies and universities. Fluid fertilizers are not expected to reach the status in A general meeting each year is attended by more than other countries that they have in the United States. The ,) 3,000, and the annual fluid fertilizer “Round-Up” draws primary driving forces-convenience in handling and appli- large attendancc and is a forum for timely technical presen- cation and savings in farm labor-are not as important in tations. NFSA publishes a bimonthly technical journal, other countries. Farms are smaller and dispersed, thus not Solutions, the leadiug journal of the fluid fertilizer industry. very adaptable to high rates of custom application. In many They also publishand update handbooks from time to time. countries there still is incentive for increased use of lnw- The NFSA Foundation was established in 1982 to promote cost farm labor. and assist in funding research vital to the industry. The Tennessee Valley Authority (TVA) and NFSA have worked Development of Application Equipment actively together for more than 25 years in promoting the fluid fertilizer industry through technology transfer and Fluid fertilizers were not only new from the standpoint other cooperative efforts. of chemistry, physical form, and production equipment, but also required new types of application equipment. Fluid Fertilizers in Other Countries This was a major problem to early producers, along with their many other problems. Weed sprayers were adapted Fluid fertilizers also have found substantial use in France, in some instances, but they were not entirely suitable. It Belgium, England, Mexico, Colombia, the U.S.S.R., and to was necessary to develop new types of equipment, first by a lesser extent in other countries. the producers themselves to use in custom application In the United Kingdom, clear liquid mixtures have been (since the farmers usually had no suitable equipment) used for several years with excellent results; commercial and later by farm machinery companies. production of suspensions started in England in 1978. Early equipment was homemade or adapted. In one of Since about 1970 liquids have been used extensively in the early types, a slit cut by hacksaw in a pipe provided a France. Ammonium polyphosphate solution of grade nozzle that effectively broadcast the solution, in about the 10-34-0 is produced from wet-process superphosphoric acid same manner as the flooding nozzles used today (I). using TVA pipe-reactor technology. Test marketing of With these makeshift hut effective beginnings, design of suspensions was underway in Belgium and the Netherlands application equipment has progrcssed until today there is a in 1978. Fluids are expected to become popular in those wide range of types available. There aregravitv-Row tyDes countries. Other reports show that small amounts of fluids for use in planting, hose Dumps, ~y varieties of nozzle have been used in Spain, Sweden, China, India, and Japan ----types, special types for suspension av~lication,and several (18). others. Many farmers now have their own equipment, The Soviet Union has conducted research on polyphos- although custoni application is still the major practice of phates and liquid fertilizers since the late 1960s, and there the industry (21). is some production and use of NF’K grades. There is sub- The modern “high-flotation” applicator vehicles, with stantial use of anhydrous ammonia and urea-ammonium huge rubber tires, can operate early on frozen or wet fields. nitrate (UAN) solution. There also is some activity with When equipped with an 85-foot boom they apply fluid fluids in Czechoslovakia and Hungary. The Soviet Union fertilizers at rates as high as 2.5 acres_ger &Ute (22, 23). started massive imports of wet-process superphosphoric (The development of application equipment and techniques acid from Florida in 1970 for general use in fertilizer is covered in depth in chapter 9.) production. The more concentrated superphosphoric acid (70% vs 54% P,O,) is imported mainly to save on shipping costs; substantial amounts are used in production of fluid STATUS OF FLUID FERTILIZERS fertilizers. (EARLY 1980s) Solutions, in a 1980 account of the export of fluid fertilizer manufacturing and application equipment, shows Fluids-A Growth Market US. firms marketing to a surprisingly large number of countries. Deliveries to the United Kingdom, Mexico, Use of fertilizer in the United States has increased Netherlands, France, Belgium, Australia, South Africa, spectacularly in the past 20 years. In 1981 plant nutrient Spain, Malaysia, Colombia, Jamaica, Nicaragua, Venezuela, use (N + P,05 t K,O) totaled 23.5 million short tons- and the U.S.S.R. were reported. One manufacturer reported compared with only 7.5 nullion tons in 1960 (table 2). dealings with 23 countries. Nutrient use doubled from 1960 to 1970 and tripled fiom

10 ~___~Table 2. Selected data onU. 5. fertilizer use ~- ____-Plant nutrient consumption (1,000 tons) (24) Fluid mixtures ___ Fluid nutrients - %Of’ Total Materials (mainly Total Year nutrients Mixtures N solutions) Total Mixtures__~~ fertilizers 1960 7464,- 143 327 470 3 2 1970 16,068 823 1,150 1,973 IO 5 I980 23,083 1,613 2,188 3,801 16 7 1981 23,45 1 1,767 2,329 4,096 17 7.5

Average annual growth rates for fertilizer nutrients (23) %total __ % fluid %anhydrous % dry nutrients Mixtures Materials Total ammonia Mix t u r e s Materiais Year _~~_ - 1960-70 8.0 19.1 13 4 15.4 17.2 4.6 8.9 1960-80 5.8 12.9 10.0 1 1.o 10.8 2.9 7.0 1970.80 3.7 7.0 6.6 6.8 4.7 1.3 5.2 1980-81 I .6 8.8 6.9 7.7 2.4 -0.1 0.2

Number and average production of fluid mix plants Estimated number Average throughput Year of plants~- Liquids Suspensions tons per year 1960 390 - - 1970 2,75 1 - - 1974 2,800 3,433 2,573 1976 3,000 2,606 3,252 1980 __ 3,200 3,840 3,604 2

1960 to 1981. In 1981 fluid nutrient use (mixtures plus materials (28-0-0, 32-0-0, 10-34-0, etc.). Anhydrous nitrogen solutions) totaled 4.1 million tons, more than ammonia is not included as a fluid material and is reported doubling since 1970 and increasing from 6.3% to 17.5%of separately. the total nutrient use since 1960. Fluid mixtures (NPK) Nutrient use as mixed fluid fertilizers grew almost 13% use in 1981 totaled 1.8 million tons of nutrients-about per year from 1960 to 1980. The average nutrient content 17% of total mixed fertilizers or 7.5% of total nutrients of mixed fluids also increased substantially. In 1981 fluid used. mixtures averaged 9.0-16.0-10.2, a total nutrient content The proportion of total fertilizer nutrients applied in of 35.2% compared with the estimated average of 28% for fluid form increases greatly if anhydrous ammonia is fluids in 1960. The increased use of suspension fertilizers included. The 4.6 inillion tons of nitrogen applied as has been responsible for the rapid rise in the analysis of anhydrous ammonia in 1981 increases total fluid nutrients fluid mixtures in the past few years. The average nutrient to 8.1 million tons-34.5% of the total nutrients applied content for all U.S. mixed fertilizers in 1981 was 43.8%. in the United States. Fluid fertilizer use has grown nearly Since their introduction in the late 196Os, suspensions have Lwicc as fast as total fertilizer use, averaging more than gained rapidly in market share. They moved from 25% 15% per year increase between 1960 and 1970, and an of the mixed fluids market in 1974, to 33% in 1976, and 11% increase between 1960 and 1980. A large part of this to more than 40% in 1981. increase occurred during the introductory stages of the Ten States-Iowa, Illinois, Nebraska, California, Texas, new product form and was aided by rapid advances in technology. Although the growth rate has slowed-to Indiana, Kansas, Ohio, Minnesota, and Georgia-account 6.8% per year during 1970-80 and 7.7% from 1980-81- for 65% of the U.S. fluid fertilizer market. In these States, )‘e fluid fertilizer use is still increasing faster than total fertil- about 21% of the fluids used is applied as mixtures. Anhy- izer use (table 2). In this table, fluid fertilizers include drous ammonia consumed in these IO States amounted liquids and suspensions as either NPK mixtures or to 75% of the total consumed in the United States (25).

11 Fluid Dealer Characteristics that will make production of high-grade clear liquids more difficult and expensive. Statistics compiled by TVA’s National Fertilizer Develop- Advantages of fluid fertilizers, including the lower invest. ment Center (NFDC) and the Association of American ment for production facilities; ease of preparation and incor- Plant Food Control Officials (AAPFCO) show the dramatic poration of additives, coiivenieiice of handling, and high growth in the number of fluid mixed fertilizer plants from reliability of fluid application systems will contribute tu 1960 to 1980. Resulls (table 2) indicate that those plants the continued strong gtuwtli of fluid ferlilizers iti ttir US. having only liquid mix facilities have an average annual market. Growth will continue in several other countries, throughput of 3,840 tons. Comparable average throughput but fluids will not appro;lch the status of practice by this for suspension plants is 3,604 tons (5). dynamic segment of the industry in the United States. A typical fluid mix plant is difficult to characterize. It may distributeliquid and/or suspension mixtures, anhydrous References ammonia, nitrogen solutions, liquid, and dry direct applica- tion materials. A frequency distribution of all fluid plants 1. Slack, A. V., “History and Growth of Liquid Fertilizers,” indicates the greatest number of plants in the 1,000- to Liquid Iktilizer.Round-Up, Pmceedings, St. Louis, Missouri, 3,000.ton-per-year range. The mode-l,245 tons-is the July 1967. 2. Leavitt, H., Proceedings of the 33rd Annuol Meeting of the value occurring with greatest frequency, and the median- National Committee on Fertilizer Applicotion, pp 85-95 2,067-represents the middle value for all fluid mix plants. (1957). Ownership data relating to all fluid fertilizer plants show 3. Matthieson, G. C., Form Chemicals 129, 45-6, 48, 90 (Sep that 13% of these plants are sole proprietorships, 7% tember 1966). partnerships, 64% corporations, and 16% cooperatives (24). 4. Slack, A. V., Journal of Agricultural and Food Chemicals 3, 568-74 (July 1955). 5. Hargett, Norman L., and Ralph Pay, Retail Marketing of Prospects for Future Growth Fertilizers in the United States presented at the Fertilizer Industry Round Table, Atlanta, Gearea, October 1980. As the fertilizer market becomes more mature, consump- 6. Hignett, T.P., ProceedingsNo. 10SFerrilizerSoeiety ofLondon tion growth rates will tend to be lower than in the past. (1969). 7. LFDC/UNlDO, FertilirerManual 1979, pp 263-265. Total plant nutrient consumption in the United States is 8. Balay, H. L., “Savings Big Attraction for Using Solids in S,us- expected to grow at just over 3% per year through 1985, pension,” Fertilizer Solutions 21 (2), 86, 88, 90, 92 (March- with total use reaching 26.5 to 27.0 million tons. Fluid April 1977). fertilizers have been projected to increase by 4.5% per year Phosphonrso,,dPotossiu,n No. 95,4142 (May/June 1978). reaching a total of 4.7 million tons-18% of total plant 10. Balay, H. L., and D. G. Salladay, “Producing and Marketing Suspensions from Solids,” Custom Applicator (21, 36-8, 40, nutrient consumption in the United States. These projected 9. 43 (February 1979). growth rates are less than those of the 1970s.~However, 1 ‘ 11. Huffman, E. O., and E. L. Ncwman, “Palyphosphates are they represent a significant gain in the total amount of fluid Revolutionizing Fertilizers,” Part I, Farm Chemicals 133 fertilizers used. 1 (21, 28-32 (February 1970). Development of the suspension segment of the fluid 12. Striplin, M. M., Jr., David McKnight, and G. H. Megar, “Fertil- izer Materials: Phosphoric Acid of High Concentration,” fertilizer industry, which can use lower cost materials such Journal of Agricultural and Food Chemisny 6, 298-303 as solid MAP or DAP and TVA base suspensions 13-38-0 (April 1958). and 9-32-0, has helped make fluids cost competitive with 13. Scott, W. C., G. G. Patterson, and H. W. Elder, “Pilot-Plant dry mixtures. This, plus the demonstrated advantages Production of Highly Concentrated Wet-Process Phosphoric for fluids, gave rise to a sharp increase in use of suspensions Acid,” Industrial and Engineering Chemirny 53, 713-16 (September 1961). in the late 1970s. There will be an increased use of lower r 14. Mitehurst, B. M., Fertilizer Indusm’ Round Table Proceedings cost solids to produce suspensions along with further (19741, pp 86-96. i, improvements in suspension application equipment for row 15. Meline, R. S., R. G. Lee, and W. C. Scott, “Use of a Pipe Reactor Production of Liquid Fertilizers with Very High or starter fertilizer application. The narrowing of the retail in Polyphosphatc Content,” Fertilizer Solutions 16, 3245 price gap between fluids and dry fertilizers coupled with (March-April 1972). Q/advantages to the farmer, such as overall convenience, 16. Achorn, F. P., and H. L. Kimbrough, “Latest Deveiopments relieving critical labor constraints, and better uses of limited in Commcrcial Use of the Pipe Reactor Process,” Fertilizer Solutions 18, 8-9, 12, 14, 16, 20.21 (July-August 1974). investment capital, suggest that use of fluids will continue 17. Chemical Marketing Reporter 215 (6), 5, 33 (February 5, to grow more rapidly than the total market. 1979). Growth of suspensions will continue higher than for 18. Achorn, F. P., “lluid Fertilizers and Their Worldwide Poten- tial,” 2nd International Conference on Fertilizers, London, clear liquid (solutions) mixtures because of their lower cost, England, December 1978. and the increasing impurities content of wet-process acid 19. Solutions 24 (3), 48-60 (May/June 1980).

12 20. p;I1gravc, I). A,. Soliirions. pp 3848 (NovcnibcriDeceinbcr 24. Consumptioh of Commcrcial I'ertilizers in the United States, 1980). Annual Rcports; and Consumption of Fertilizers by Class, 21. A/~piicorio,iOJ I'lirid l.i.rrilirerx (April 1970). (TVA Circular Annual Reports; Economics, Statistics, and Cooperative 2-6). Service; Crop Reporting Board, USDA, Washington, D.C. 22. Solsrioriu, pp 31-36 (Septcinbcr/Octobcr 1981). HwcK Norman L., and Janice Berry. 19x0 Fertilizer 23. ~~1,~~~.,:_p., H.L, Ilal;,y, .cl:ql,il,,,, cnt and ill 'TIK Liquid Rcvalution,' '' l:crti/izw Solurioris 16. 30, 32, Sun~mvy Darn, National Ikrtilizer Development Center, 34, 36, 38.40, 4243 (January-l'cbruary 1972). Tcnncsscc Vellcy Authority, Musclc Shoals, Nabama.

13 Chapter 2 Agronomic Aspects By Eugene C. Sample and Gary W. Akin

In any discussion of fluid fertilizers, questions arise when comparing pliospliate fertilizers. For instance. it would concerning the relative agronomic merits of various physical not he valid to compare a highly water-soluble phosphate and chemical forms of fertilizers. Two common questions in fluids with s solid pliospliate of low water solubility. center on whether fluid fertilizers are agronomically better However, wlien solids such as diammoniuiil pllospliate than their dry counterparts and whether polypliosphates (DAP), tiio~ioa~iitiiiiiiiu~iipliospllatc, or amniotliuiii poly- are agronomically better than orthophosphates, The poly- phospliale were compared with fluids such as IO-340. phosphate question is pertinent to a discussion of fluid 8-24-0. or 11-37-0 under similar conditions, long-tcriii fertilizers because polyphosphates are most often studies have sliowii these to bc essentially equal in nutritive encountered in fluids. The answers to these questions are value. Similarly, long-term studies Iisvc shown solid urea or often confusing because responders teiid to include tioii- ammunium nitrate to be essentisliy equal tci nitrogen agronomic factors such as ease of application, uniformity solutions, such as iirea-atiitiioiiiiiiii nitrate. Essenti;illy the of application, and economics in their agronomic evalua- same conclusions would be reaclicd with dry and lluid tions. The answer should be based solely on crop yields NPK mixes. obtained when materials are compared using similar rates of plant nutrients under similar methods of application. Cautions in Comparing Fertilizers

Agronomic Comparisons of For valid comparisons. studies should be conducted Fluid and Dry Fertilizers for several years at the same location iisiiig the same cxperi- mental design to eiisure that the v:iriahility inherent in field Although numerous field studies have been performed studies does nut lead to faulty interpretations. If data are comparing fluid and dry fertilizers under similar conditions, selected from one study. for one year. at one location. very little of this research is reported in refereed journals. evidence can be cited to prove that solids are better tlian Rather, it is usually found in experiment station animal fluids, or vice versa. or that polypliospliates are better tliaii reports which are not widely available. The comparisons orthophosphates, or vice versii. In the United States it is most often made have been among solid urea, ad not uiiusual to encounter evidence in the niarketplace 111 ammonium nitrate, and urea-ammonium nitrate (UAN) in advertising to support eacli of these claims. fluids; =dry ammonium phosphates and fluid Care must be exercised in comparing any fertilizers ammonium phosphates; and between ammonium ortho- under field conditions, whether they be solids or fluids. phosphates- and ammonium polyphosphates. For example, if one wanted to compare conccntrated super- Experimental data from a wide range of studies over- phosphate (CSP) with either 10-34-0 solution or solid whelmingly support the conclusion that there are essentially monoarnnionium phosphate (MAP) as sources of phos- no differences among the liquid, suspension, and dry phorus, one cannot ignore the Fact that both 10-34.0 and fertilizers when they are compared over the long term MAP also contain tiitrugen. To be a wlid comparison, under conditions of similar nutrient rates, placements, equivalent amounts of nitrogen should be applied with the and chemical forms. The latter is particularly important CSP to ensure that the materials are being compared for

14 their pliuspliorus supplying power, not confounded by tlie Potassium fertilizers are even more uniform than either absence of nitrogen. Conversely, if one were comparing a nitrogen or phosphate fertilizers. The dominant source of solution, sucli as UAN,cwitaining only nitrogen, with MAP, potassium for both fluid and dry fertilizers is potassium which contains both nitrogen and phosphorus, some chloride. Even when other sources, such as potassium source of pliosplioius (such as CSP) should be applied phosphate or potassium nitrate, are used, it is the potassium with tlie UAN to equalize pliosphorus. (K’) which the plant root deals with in the soil solution.

Why are Fluids and Solids Implications of Equality Equal Agronomically? There are a number of logical reasons to expect fluid and The relative equality of fluid and dry fertilizers should dry fertilizers of comparable chemical makeup to be equal not be too surprising in light of the fact that the chemical agronomically. Yet some proponents of either fluid or dry constituents of the two physical forms are usually identical. fertilizers feel that they are at a marketing disadvantage if For example, UAN solution or suspension contains both they cannot claim agronomic superiority for their particular urea and a!ninonium nitrate, both popular dry materials. form. From a farmer’s viewpoint, however, the equality Likewise, 8-24-0 solution is an ammonium orthophosphate of tlie various forms is a powerful management tool. It material, just as MAP and DAP are. Some of the ammonium frees the farmer to choose from a wide variety of materials phosphate fluids may contain ammonium polyphosphates; using a multiplicity of nonagronomic factors as criteria but this does not confuse the issue too much, as discussed for the decision. For example, if he practices some form of below. no-till or minimum tillage agriculture, the farmer’s primary The matter of equality of various physical forms is even concern may be to achieve a sod kill with a herbicide and more predictable when one considers the limited variety of fertilize by broadcast application, all in one operation. chemical forms presented to the plant root. Although a His most efficient and economical option may well be to farmer may apply fertilizer nitrogen as anhydrous ammonia, apply a fluid fertilizer-herbicide tank mix. He can choose urea, aninioniuin nitrate, UAN, , or several this option with the assurance that his fertilizer will perform other forms, the same farmer may be assured that, within a about the same as it would have with a broadcast application fairly short time, the roots of his crops will be confronted of a comparable dry material, with a separate herbicide mainly with nitrogen in the nitrate form (NO,?. Thisis spray. There are numerous other technological or conven- because various soil enzymes rapidly convert urea nitro en ience advantages for fluid fertilizers, many of them dis- to ammonium forms, and then soil micro b+. IO oglcdl proc cussed elsewhere in this manual, which he can take advan- esses fairly rapidlv convert the ammonium forms to tage of without worrying about sacrificing fertilizer quality. nitrate. So, for most of the growing season, plant roots If the farmer is in a region of micronutrient deficiency and “see” mainly nitrates. needs to apply relatively small amounts of these nutrients Despite the fact that farmers are offered a wide array of uniformly over relatively large areas, incorporation of the pliospliorus-containing fertilizers, these farmers are assured micronutrient into a fluid fertilizer is a way of assuring that their crops are really confronted with a very limited uniform distribution of the micronutrient in the fertilizer variety of chemical forms of phosphorus. First, the phos- material and, hence, uniform distribution on the field. In phorus in most fertilizers is present in the orthoplios- fact, uniform distribution in the field, with or without phate form. When an orthophosphate-containing fluid micronutrients, is often cited as one of the advantages fertilizer is applied or an orthophosphate-containing dry of fluid over dry fertilizers. fertilizer dissolves in the soil solution. the olant roots are In summary, fluid and dry fertilizers of comparable confronted mainly with two phosphate species (H,P04- - chemical constituency are essentially equal agronomically -;c_nd~,5. If a fertilizer inaterial containing polyplios- phates zpplied, the polyphosphate is fairly rapidly con- when applied at equivalent nutrient rates under similar verted in must agricultural soils to the orthophosphate placements. This equality gives farmers a large number of form. So..” reeardless of the vhvsical._ or chemical form options in choosing fertilizer materials based on economics, of phosphate fertilizer, after a short while in the soil, convenience, compatibility with other operations, dealer plant roots “see” only two very similar forms of phosphate. services, or regional availability.

15 Chapter 3 Physica I Properties Eva Iuat ion Bv T. M. Jones and H. L. Balay

The term, “fluid fertilizers,” usually applies to all tures abovc about -5°F (-10°C) and their viscosity at fertilizers that are capable of flowing. However, the fluid 0°F (-18°C) cannot exceed 2,000 centipoises (see section fertilizers in which the TVA National Fertilizer Develop- on viscosity for iiietliod of detcrminiiig viscosity) because ment Center has been principally involved with are solu- products of liiglier viscosity caiinut be pumped satisfac- tions and suspensions. Each of these classes of fertilizers torily. These base fluid products oftcn coiitaiil significant has chemical and physical properties that significantly proportions of urea, polypliospliate, or sulfate which act affect its production, transportation, storage, and appli- as solidification temperature depressants. However, ortho- cation to the soil. Therefore, it is necessary that careful phosphate base fluids usually do not contain significant evaluations be made of the chemical and physical properties proportions of these materials, and they. tlierefore, usually of the fluid fert.ilizers to determine their suitability for do not meet the solidification temperature and viscosity commercial use. requirements for satisfactory storage in cold wcather Both solution and suspension fertilizers cover two which restricts their winter storage in the colder regions distinct types of fluid products. One type is mixed fluid of the country. fertilizer blends which -are for dircct application to the The evaluatiuns of solution ;rnd suspciisiuii fertilizeis soil. The other type is base fluid fertilizers which may also overlap in that they are but11 rated for many of the same be applied to the soil, but more often are used in produc- chemical and physical characteristics. In making these tion of fluid fertilizer blends of various N:P,O, :K20 evaluations, much of the same equipment iind procedures ratios and grades which are then applied to the soil. These are used. However, some evaluations that arc made apply mixed fluid products can be produced very rapidly by cold only to solutions and others apply only to suspensions. blending different types of base solutions or suspensions. Since the mixed fluid blends are made for direct appli- Solution Fertilizers cation to the soil, long-term storage is considered to be unnecessary and they are usually subjected to storage The major advantage of solution fertilizers over other tests for only 2 weeks by TVA. Longer storage periods fluid products is their freedom from troublesome precipi- may be chosen by individual users to suit their own tants which settle in storage tanks and distribution equip- requirements. ment and can stop up feed lines, orifice plates, and spray The base fluid products usually are produced, then nozzles during application to the soil. A lesser advantage stored for various lengths of time before they are used to is that solutions usually have lower-viscosities (SO to 150 produce mixed fertilizer products; therefore, they must centipoises) than inost solids-containing fluid fertilizers be capable of withstanding storage for long periods of which makes them easier to pump (11 otherwise trausfcr time under various atmospheric conditions without deterio- from one container to another. The principal disadvantages ration in product quality. Some base fluid fertilizers are of solution fertilizers are that, generally, they are lower in designed for storage in the colder regions of the country plant food content and have higher production and sliippiiig and others are not. Products that are considered to be satis- costs than other fluid products of similar N:P205:K20 factory for cold weather storage do not solidify at tempera- ratios.

16 During production of solution fertilizers, measurements Suspension Fertilizers for PH and density are made to determine whether the products are of about the desired grade. Measurements are In past years, TVA has developed numerous types of also made of the saltingout and saturation temperatures suspension fertilizers. Most of the work has been directed to determine the lowest temperature that solutions can be toward development of base suspensions that can be stored subjected to without crystallization of fertilizer salts. until needed, then cold blended for production of mixed After production and before storage, both base and mixed suspensions of various ratios and grades. During planting grade solution products should be evaluated chemically season, most of the farmers would like to get their fertilizer to obtain accurate determinations of their grade and at about the same time, and production of mixed blends N:P,O, :K,O weight ratio. The solutions should also be from base products is considered to be the most convenient evaluated chemically for polyphosphate and for impurities way of rapidly producing premium-grade products of almost (Al,O,, Fe,03, MgO, and F) to indicate the expected any grade or ratio desired by the farmer. stable impurity precipitating phases, the extent of precipi- The principal base suspensions with which TVA is work- tation. and the length of time that the solution would be ing in laboratory, pilot-plant, and demonstration-scale expected to store ratisfactorily. The solutions are evalriated plant development are urea-ammonium nitrate, urea- physically for pourability, clarity, viscosity, crystal size, ammonium sulfate, ammonium orthophosphate, ammo- and solidification temperature to determine whether or nium polyphosphate, urea-ammonium polyphosphate, and not they are satisfactory initially. Mixed solution products potash base suspensions. are then stored for two weeks at 80°F (27OC) and 32°F In production of satisfactory base suspensions, it is necessary that slurries containing an abundance of small (OOC) and then reevaluated for viscosity, pourability, crystals of the stable solid phase be first produced for pre- large (+20 iiiesli) crystals, totalvolume of solids, and volume vention of excessive crystal growth during slow cooling and of adhering type solids (see specific sections on these during fluctuations in temperature during storage. The subjects for methods of determining these physical proper- abundance of small crystals is produced by quick cooling ties). Base solutions are stored at temperatures of 3Z0, SOo, (quenching) solutions from the reactor to well helow their and IOO'F (0'. 27", 38°C) for periods of time up to saturation or salting-out temperature. Recirculation in the about IS0 days. Replicate samples of each solution are coolingcrystallization stage has been found to further pro- stored under quiescent conditions. Since base solutions may mote production of small crystals. To suspend the ahun- be eitlier put into clean tanks or added to solutions that dance of smJl crystals, a gellingtype clay is added during have been in storage for some time, some of the storage or after the cooling step and incorporated in the slurry samples are seeded with a few small crystals of the expected by mixing it thoroughly with a high shear stirrer or pump. crystallization phases and the other samples are unseeded. Generally, during summer months, economics limit opera- tion of the cooler to a minimum discharge temperature of During storage, the samples are evaluated for pourability, about 100°F (38°C). In suspensions that are rated as crystal size, and volume of both adhering and nonddhering satisfactory, the crystals remain suspended during transit types of crystals. During the first month of storage, the and storage for indefinite periods of time. samples are evaluated weekly; during the next 3 months, After production, the suspensions usually are allowed they are evaluated biweekly; and after 4 months they are overnight storage at room temperature. (about 75'F evaluated monthly. [24OC]) for establishment of equilibrium. They are then Solutions that are considered to be satisfactory are com- evaluated chemically for primary nutrient content (N, pletely pourable at both 80°F (27OC) and 32°F (O"C), have P,05, and K,O) to determine the suspension grade and viscosities that did not exceed 400 centipoises at 80'F nutrient ratio. The suspension products are evaluated

(27OC) and 32°F (OOC), and have clarity of at least 25% petrographically to determine the type and size of the light transmittance as compared with distilled water as crystals. The products also are evaluated physically for having 100%. Satisfactory solutions also contain no crystals pourability, viscosity, solidification temperature, and settling of crystals during overnight storage at room that exceed 20 mesh (850 urn), no more than 3% by temperature. Suspensions which are produced for ship volume of total solids, and no more than 0.1% of adhering ment are also evaluated for vibrational settling such as type solids. Solutions that are to be stored in cold weather occurs during transit by rail. The suspensions are then also must be completely pumpahle at very low tempera- used in application to the soil, in production of mixed ture: to meet this requirement, the primary nutrients must grade products, or are shipped and stored. For evaluation not salt out, the solutions must not solidify at tempera- purposes, base suspension products usually are subjected tures above -5°F (-20°C), and the viscosity at 0°F (-18°C) to storage tests for 30, 60, and 90 days at 80°F (27°C) must not exceed 2,000 centipoises. and 100'F (38°C). After each storage period, the base

17 --

suspensions are evaluated at 80°F (27OC) and 32'F (0°C) would be supplied by air sparging. The contaiuer then is for pourability, viscosity, syneresis, crystal type and size, turned downward at a 45' angle for I minute and tlie and settling of crystals during static storage. content (% by volume) that pours out is determined. Field Base suspensions that are considered to be satisfactory tests have shown that fluid products that are at least 98% are at least 98% pourable when evaluated at 80°F (27OC) pourable by this procedure will flow freely by gravity froin and 32°F (O"C), and have viscosities that do not exceed their storage container and they easily can be transferred 1,000 centipoises at 80°F (27°C) or 1,500 centipoiscs at from one tank or container to mother by pumping. 32°F (O'C). Also, satisfactory base suspensions contain Viscosity--Fluid fertilizers are evaluated in the labora- no large +20 mesh (850 pm) crystals, and no crystals that tory for viscosity to indicate their ability to flow. The vis- settle and pack on the bottom of the storage container. cosity generally is measured with a viscometer (figure 3). Suspensions for shipment also contain less than 2% by Before measuring the viscosity, all fluid fertilizers are agi- volume of crystals that settle during vibration tests. tated for 5 minutes with a stirrer operated at propeller- Syneresis, formation of clear liquid that sometimes collects tip speed of 7 feet (2.1 m) per second. For fluids that are either on top or within the suspensions, is undesirable to be applied directly to the soil, the viscosity must not because it may foster growth of excessively large crystals exceed 800 centipoises at 80°F (27°C) and 900 centi- at the suspension-clear liquid interface. However, suspen- poises at 32'F (0°C). The viscosity of base fluids that are sions containing syneresis are not rated unsatisfactory if primarily for use in production of fluid blends must not they can be remixed by recirculation or air sparging. Base exceed 1,000 centipoises at 80°F (27"C), 1,500 centipoises suspensions for cold weather storage do not solidify at at 32°F (OOC), or 2,000 centipoises at 0°F (-18OC). Field temperatures above -5'F (-20°C) and the viscosity at tests have shown that fluid blends which do not meet the O°F (-18OC) must not exceed 2,000 centipoises. above viscosity specifications usually cannot be uniformly Mixed suspensions are for direct application to the soil; applied to the soil, and base suspensions that do not iiieet therefore, they are stored only for 14 days at 80'F (27°C) these viscosity specifications usually cannot be satisfactorily and are evaluated at 80'F (27°C) and 32'F (0°C). They drained from storage tanks or transferred froiii one are initially evaluated chemically for nitrogen, phosphate, container to another by pumping. and potash, and petrographically for crystal size. The mixed Cq~stalType and Size--Fluid fertilizers are considered suspensions also are evaluated physically for pourability, to be unsatisklory if thcy contain any crystals, either viscosity, settling of crystals during overnight static storage, initially or after storage, that exceed 20 mesh (850 pni) and syneresis. After 14 days of static storage, the suspen- in size, because tlie large crystals can stop up dislribritioii sions are reevaluated for crystal size, pourability, viscosity, pipelines, spray nozzles, orifice meters, and oilier types settling of crystals, and syneresis. of flow metering or flow regulatiiig equipment. In suspen- Mixed suspensions that are rated as satisfactory are at sion fertilizers it is necessary that an abuiidancc of sniall least 98%pourable at 80'F (27°C) and 32'F (OOC), contain crystals of the precipitating pliiise be present for prevention no +20 mesh (850 Mm) crystals, and no crystals that are of growth of crystals to excessively large size. The type of settled and packed on the bottom of the storage container. crystals in fluid fertilizer products is also important because Loose crystals on the bottom of the container or syneresis crystals vary in shape from long needle-like to cubical (clear liquid) are considered to be undesirable, but the blocks and to thin sheets (figure 4). The longer type crystals suspensions are not rated as unsatisfactory if recirculation are avoided because they cause stoppages in equipment more or air sparging will easily remix the suspension. readily than crystals of oilier shapes. The thin rectangular shaped crystals are preferred because they have a higher Laboratory Measurement of Physical surface area-to-volume ratio and are therefore less difficult Properties of Fluid Fertilizer to suspend.The size and shape of the crystals are determined in the petrographic laboratory by microscopic analyses. The laboratory methods used in making the evaluations Seftlirig of Crystuls--It is desirable that all crystals in are described below. suspension fertilizers remain suspended at all times. The Pourability--Fluid fertilizers are evaluated in the specifications require that for a suspension to be rated as laboratory for pourability to determine whether or not satisfactory, no crystals can settle and pack on the bottom they can be drained from their container by gravity or of the container during static storage and a maximuni of pumped or otherwise transferred in the field From one only 2% by volume of crystals can be settled after the container to another. This determination is made by vibrational settling test which simulates vibration wluch rotating a stirring rod twice around the inside surface of the occurs during transit by rail or truck. A suspension is storage container, generally an 8- or 16-ounce sample measured in the laboratory for settled and packed crystals bottle, to apply gentle agitation to the fluid. Rotation of the by pressing a stirring rod to the bottom of tlie storage stirring rod is intended to supply no more agitation than container to determine if any resistance is encountered

18 Figure 3. Viscosity measuring device to reach tlie bottoni of the cimtainer. If settled crystals crystals (exclusive of carbonaceous matter) and less' than are very loose ;md are capable of being redistributed in tlie 0.1% by volume of adhering type crystals. A solution is suspension by recirculation or air sparging, the suspension rated as acceptable if it contains no more than 3% by is rated as satisfactory. If the settled crystals are packed volume of loosely held precipitated solids. A solution is and diflicult tu redistribute, lhe product is rated unsatisfactory if it is less than 100%pourable, contains any unsalisfactory. t20 mesh (850 pm) crystals, or contains more than 3% by Crystals of any type are undesirable in solution fcrtil- volume of precipitated solids. izers. To ensure against nucleation and growth of fertilizer Syneresis--Suspension fertilizers sometimes are affected silt crystals in solution fertilizers. Salt-out temperatures are by syneresis, which is formation of clear liquid which usually determined. and the concentrations of the solutions are occurs on top of the suspension with the remainder of the adjusted to enable production of the higllest possible satis- suspension being uniform throughout. The presence of the factory solutiun grades. However,solutions that are prepared clear liquid layer does not necessarily cause a suspension from wet-process phosphoric acids convain impurities that to be unsatisfactory, but it is undesirable because with precipitate slowly as very fine crystals or sludge. Since decrease or fluctuations in temperature or other conditions precipitation 11f these impurity materials is almost impossible that cause changes in the concentration of the liquid phase, to avoid. some precipitated solids are permitted. Solution growth of excessively large crystals occurs at the clear fertilizers are rated as beingverygood if they are completely liquid layer interface. Provided the crystals do not grow to pourable, contain no t20 mesh (850 pm) crystals and exceed 20 mesh (850 pm) in size, syneresis is not objec- cmtain 110 more tlian 1% by volume of nonadhering type tionable if recirculation or air sparging will remix the clear

19 with a spcctropliotoiiieter. The ubject of the tiieasureiiieiit is tu determine wlietlier or iiiit the fertilizer is clear enougli for fertilizcr producers, farmers. or otlicrs that Iiandle or distribute solutioii fertilizers tu the soil tu see large crystals, or foreign iiiatter. that iiiiglit stop up orifice meters, or other equipment, that is used in Iiaiidling or distributing tlic solutioii to tlie soil. Clarity is expressed as tlie percentage transmittance of light ;it the desiied wavelengtli uf tlic liquid absorption spcctruni. A liquid with 25% light transmittance. as coni. pared with that uf distilled water its being loo%, is considered as acccplable. Solidificafioii Temperature--If a fluid fertilizer is pro- duced for wintcr-time storagc in the colder regions of tlie country. its solidification temperature must not exceed -5°F (-20°C) and ils viscosity at 0°F (-18°C) inust not exceed 2,000 centipoises. Field tests show that lluids wliicli nieet these requirements can be truisferred froin we coli- taiiier to another tlirougliout the winter wcatlier of States located in the colder regions of [lie country. Field tests liave further sliow~itliat some fluid products with solidi- fication temperatures liiglier tliaii -5°F (-20°C) solidified during winter storlige and could iiot be removed froiii storage tanks in time for spring application. Solidification temperatures of fluid fertilizers are iiieasured in the labora- tory by cooling tlieiii to soliditicatioii over a period of 1-1/2 to 2 liours with an acetune 111 etlianol-dry ice bath. During cooling, the fluid fertilizer is agitated with a stirrer operated at a propeller-tip specd of about 7 feet (1.1 111) Figure 4. Mix of MAP and DAP crystals per second. Thc solidification tcmperature is recorded wlieii liquid and suspension. Generally, syneresis is caused by either il temperature rise occurs and with continued cooliiig the insufficient clay concentration or improper gelling of clay. fluid product becomes seiiiisolid or solid. Chi@--Fluid fertilizers that are produced from wet- .~alting-Oictarid Sahrration Teniperatures--Sal tiiig-oul process phosphoric acids derived froin uncalcined phos- and saturation temperatures are determilied by cooling tlie phate ores are black in color due to the presence of finely solution until crystallization occ~irs(salting-out teiiiper:i. divided carbonaceous material. The black color has 110 titre) and then warming it very slowly with continuuus effect on the nutrient value of the fertilizers but it does stirring until the last crystals disappear (saturation tetiipera- restrict the sight identification of solid material in solution lure). Usually the cooling and heating rate is about 2'F fertilizers that could clog the equipment used for appli- (1°C) per hour in tlie critical range. sucli nieasureiiieiits cation. Solution fertilizers are often used for row appli- are usually made in duplicate but occasioiially ilre iiiade in cation, whereas the larger part of suspension fertilizers triplicate and the average is taken as the true value. is used for broadcast to the soil. Solution fertilizers often Nutrient Solrrbili@--Animotiiitiii pliospliate sdutiiiiis are gravity fed, and the distribution equipment, orifice exhibit a high degree of supercooling. To obtain nutrient meters, pipelines, and nozzles used for row application solubility in these solutions, 50 milliliters of a series of usually are smaller than those used for broadcast appli- solutions of the ratio being tested differing by sdliticrc- cation of suspension fertilizers. Since the application equip- inelits in water content are stored at constant temper;lture. ment is smaller, it is more likely to get stopped up by large usually 32°F (0°C). Each solution is seeded with about 50 crystals or foreign particles; therefore, the need for being able to see these solid particles in the solution fertilizers milligrams of crystals of a composition previously deter- is much greater than it is in suspension fertilizer. mined tu be the salting-out phase of tlie particular utio atid In production of solution fertilizers, samples of the agitated mildly in a shaking apparatus. The disappearance product are taken and measured for clarity, which is the or growth of crystals it1 the diffeaent saniplcs permits amount of visible light that passes through the sample with. determination of the solubility at 32°F (OOC) witliin out being absorbed. Measurements for clarity are made 0.1% to 0.2% of the plant nutrient content.

20 Specific Gravity--In the production of solution and to vibration such as occurs during shipment by rail, the suspension fertilizers, samples of the product are taken and crystals settled excessively and packed on the bottom of measured for specific gravity to obtain estimates of the the railroad tank cars. Settling of crystals during vibration product grade. Specific gravity is the ratio of the weight of was found to be related to the weight, size, and shape of a given volume of solution or suspension compared with the crystals; the specific gravity of the solution phase of that of the same volume of water: when used together with the suspension; and to the strength of the gel. In order to pH it can give fairly accurate estimates of the grade of the estimate the proportion of crystals that would settle during product. Since specific gravity and pH are both affected transit by rail, a shaker was fabricated and calibrated at by temperature, the measurements are made at 80°F TVA to give approximately the same results during shaking (27OC). Specific gravity measurements are made at TVA by for 2 hours that would be obtained by transit for several either actually weighing il volume ofthe fluid fertilizer and hundred miles by rail. TVA product specifications limit dividing the weight by the weight of an equal volume of settling during vibration to less than 2% by volume of loose water or it is obtained through use of hydrometers. A crystals. Vibration settling tests are made on all experi- hydrometer is a calibrated instrument that will sink to mental materials that are produced in bench. and pilot- ~riousdcpths dcpcndcn:, upon :hc dcnsity of thc fluid. plant-sialc cquipmcnt and on most of the materials thst The specific gravity is rehd directly from the calibrated are shipped by rail. scale. Gel Strength--Suspension fertilizers are composed ofa Field Measurement of Physical gelling agent dispersed in a solution of fertilizer salts- Properties of Fluid Fertilizer usually the solution is saturated and both crystals offertil- izer salts and impurities are present. In production of the Measurements of physical properties made in the field suspension fertilizers, a suspending agent, which is usually are usually not as precise as those made in the laboratory a gelling-type clay, is incorporated in a slurry to form a gel but, with care, useful information can be developed. that is strong enough to suspend the solids in the saturated Density--lt is especially important to measure density solution, but is not too strong to prevent pumping, pouring, because most fluid fertilizer application equipment in the and distribution to the soil. Gel strength, which is the force United States is calibratedingallons per acre and the weight required to break the gel, is measured with a gelometer of a gallon of fertilizer must be known fairly closely to developed at TVA several years ago (see chapter 7). Here- obtain accurate application rates. tofore, there were no attempts to actually measure gel Two methods of measuring density are usually used. strength, The viscometer is used to measure viscosity (the The most common is use of the glass hydrometer calibrated resistance of a fluid to flow) once the gel has been broken. in appropriate units of specific gravity or density. A glass Viscosity is related togel strength, and prior to development hydrometer calibrated at 60°F (16"C), a thermometer, and of the gelometer, the viscometer was considered to be the a suitable container (hydrometer jar) are required. The jar best available instrument for determining the ability of the should be large enough to allow clearance between the sides suspensions to prevent settling of the solid particles. and the bottom of the hydrometer and the jar. To measure In production of suspension fertilizers, gel strength the density, the hydrometer should be gently lowered into measurements are made with the gelometer to determine the fertilizer avoiding wetting the stem above the level of the proportion of gellingtype clay that must be added and the liquid. The sample should be stirred with the ther- the degree of mixing that is required for producing a gel mometer until a constant temperature is reached. The strong enough to keep the solids suspended in the saturated hydrometer should then be given a slight spin and read solution but weak enough to allow the suspension to drain when it comes to rest. When the hydrometer is read, it by gravity or pumped from their containers. Factors that should not be touching the walls or the bottom of the are known to affect the required degree of clay mixing are container. The correct reading is the point at which the pH of the solution, proportion of solids, specific gravity, surface of the liquid touches the scale. The reading is and the presence of polyphosphate and nitrate ions. Once recorded at the temperature of the sample compared to the the proportion of clay that should be added and the degree temperature at which the hydrometer is calibrated. For of mixing required have been established for production of example, if the fluid is at 77'F (25'C) and the specific a satisfactory suspension product, further gel strength gravity reading is 1.439, it should be recorded as 1.439 measurements are only made as occasional spot checks or at 77"/60"F (2S0/160C). If the hydrometer is calibrated in after changes in the process or in feed materials have been specific gravity, the density of the solution can be deter- made. mined by multiplying the specific gravity reading obtained Vibrational Settling--Some suspension fertilizers have from the hydrometer by the density of water at the tempera- been known to store satisfactorily for long periods of ture of the fluid being measured. If the density of water is time under static conditions but, when they were subjected not known at the exact temperature of the fluid, a close

21 approximation can be obtained by multiplying the specific DH gravity obtained times 8.34, the weight of 1 gallon of Water 0 082 164 246 328 411 493 575 657 739 821 903 DEGREE OF AMMONIATION. LB NH,/UNIT at 62OF (17'C). For example, 1.439 times 8.34 equals 12 LL PzO, Yc pounds per gallon. 10 19 29 38 48 58 67 77 86 96 glm, , , I I I I 1 I I I , A more accurate way to measure density, especially with dark colored fluids and suspensions, is to actually weigh a known amount of the mixture. In the laboratory, this can be done by using a triple beam balance reading in grams and a 100 milliliter volumetric flask. The balance is tared to the weight of the flask, or the empty flask is weighed. The flask is then filled to the mark with the fluid to be measured and the flask weighed. The weight of 100 milli- liters of liquid can be read directly from the balance in grams if the flask has been tared, or can be determined by subtracting the weight of the empty flask from the weight of the flask plus the fluid. The density of the liquid at the temperature of the liquid in grams per milliliter can then be determined by dividing the weight of the liquid by 100. The approximate weight of the liquid in pounds per gallon can be determined by multiplying the grams per milliliter obtained times 8.34, as is done with the specific gravity NH,:H,PO, MOLE RATIO determined by glass hydrometer. This can be done because Figure 5. Effect of NH, :H,PO, mole ratio the specific gravity and density in grams per milliliter of on solubility of ammonium phosphate fluids are nearly the same. A variation of this method is to weigh a known amount meter should have a standardization control, temperature of material using a bucket of known volume and any scale compensation control function switch, and an electrode available. Frequently a two-and-ahalf-gallon bucket and holder. The electrode can be the glass-calomel reference dairy scale are used. The percentage error in this kind of or a combination electrode type and, especially in a fertil- measurement is less than might be supposed because of izer plant, spare electides should be kept on hand. The pH the large size of the samples being measured. meter should always be adjusted with a buffer solution Saturation Temperatures--A system, similar to the (pH 6) and washed with deionized or distilled water before .. .. laboratory method, is sometimes used to estimate the and after each use. The electrode should be stored in saturation temperature in a plant workroom. A sample of deionized or distilled watcr and not left dry or in fertilizer the liquid to be tested is put into a test tube and a ther- solution. The operating instructions of tlie pH nicter mometer is inserted into the tube. The test tube is then supplied by tlie manufacturer should be read and followed inserted intb a container of ethanol mixed with dry ice and closely. the fluid is frozen. Methanol-based automobile antifreeze To nieasure the pH of fluid fertilizer. the probe should can be substituted for the ethanol but if this is done, the be inserted into.the fluid and the fluid swirled for about 5 experiment should be performed outside or in a well- seconds. A thermometer should then be inserted into tlie ventilated area to, prevent fumes from the antifreeze from sample, the temperature read, and the temperature compen- collecting in the bork area. The frozen fertilizer is then sation setting adjusted to the temperature of tlie solution. allowed to melt while stirring constantly with the ther- After about 45 seconds the reading should be obtained. mometer. The temperature at which the last crystal dis- For accurate measurements, tlie reading should continue appears (as read from the thermometer) will give a close until a constant value is obtained. approximation of the saturation temperature of the fluid. Diluting the sample 4 to I wid1 distilled or deiunized pH--It is important, when phosphoric acid is being water will sometimes give a more accurate reading. This is ammoniated, to know the ammonia-to-phosphoric acid especially helpful in the case of suspensions where all ratio to obtain the optimum solubility of the fluid being produced. This ratio can, be obtained by measuring the soluble salts can usually be dissolved by a 4 to 1 dilution. pH of the fluid (figure 5). The pH can be obtained either Because calibration and niaintenance of pH meters in with an electronic pH meter or pH paper. fluid fertilizer plants is a problem, pH paper is oftcn used The most accurate method of obtaining pH is with an to control the anumonia:pliosphuric acid ratio. With 3 little electronic pH meter, but a pH meter is a precision instru- practice, pH paper can be used successfully. The pH paper ment and must be used and maintained properly. The pH should be chosen carefully. Some pH papers do not work

22 in fertilizer solutions. The paper should cover a pH range that has tlie expected pH near tlie center. For example,, if tlie expected pH is about 6.0, the pH paper range should vary from about 4.5 to 7.5. If readings are difficult to obtain, satisfactory pH readings usually can be obtained with pH paper by diluting the sample 4 to 1 with distilled or deionized water. The relationship between pH and N:P,Os ratio in the pure systems at various polyphospliate levels is given in figure 6. Typical physical properties of some solution and suspension fertilizers are given in tables 3 and 4, respectively.

References

1. Sohiogcn Nitrogen Solutions for Liquid fertilizer Manufacture, OlFFERENCES IN IMPURlT" CCNTENT OF YCUTIONS Sohio Chcmical Co.. Lima, Ohio. WEFROM WET PROCESS ACID 2. SI:tck~~ A. V.. J. hl. Potts. and H. B. Shaffer. "Solubilitv Rela- ~~~~~. ,' tionships in Liquid Fertilizer Systems Bascd on Superphor- I ,' phoric Acid." Ayriarlfirrol old Food Clzemisny. Val. 12,

Table 3. Typical physical properties of various solution fertilizersa Specific Density, Salting-out gravity Ib/gal Viscosity,b cps temperature Grade at ____~75°F at 75'F at 75°F OF("C) pH 8-24-0 1.26 10.5 30 18 (-8) 6.6 1O-34-Oc 1.37 11.4 50

23 _____-~Table 4. Typicala properties of various suspension fertilizers Specific Density Solidification gravity. lb/gal Viscosity, temperature, Grade at 7S°F at 75’F cps at 75’F______’F CC’’C pH 12-40-Ob 1.44 12.0 700 0(-18) 6.0 13-38-0 1.44 12.0 800 1s’(-9j 6.4 9-32-Ob 1.40 11.7 200 -7 (-22) 6.7 10-30-0 1.37 11.4 250 NA~ 6.3 3 1 -0-oc 1.32 11.0 150 NA~ 8.0 18.0-18 1.37 11.4 160 NA 7.4 13-13-13 1.40 11.7 600 NA~ 6.3 4- 12-24 1.41 11.8 400 NA~ 6.3 0-0-31 e 1.30 __ 10.8 360 NA~-~_____ 6.8 aBased on wet-praccssorthophosphoricand polyphosphoric acids. Properties may vary with impurity content of TBW iiiaterials used and amount of suspending clay added. bCantains polyphosphate. ‘Urea-ammonium nitrate suspension. dNot available. ePotash suspension.

24 Chapter 4 Chemical Properties By A. W. Frazier

The chemical knowledge of fluid fertilizers has increased use of these highly impure multicomponent acids. The from a few pure system studies priur to 1960 to tlie present precipitation and gelling problems presented by the redis- development of multicomponent -systems including micro- tribution and interaction of the iron, aluminum, magnesium, nutrients and impurity coinpunenis in ammoniated wet- and fluorine needed study. Considerable progress bas now process polyphosplioric acids. Initially. relatively pure been made in defining the large number of compounds soluble materials were available fur the production of clear (3, 5, 6, 7, 8, 9, IO) that can be present in experimental liquid fertilizers. Even though their physical properties, ammonium phosphate fluids. However, a fundamental such as ease of transport and distribution, lack uf dustiness, application of chemical methods on the complex multi- and compatibility with herbicides and pesticides, were very component phase systems is needed before long-term desirable, the low solubility of these products (1) limited storage of stable fluids can be guaranteed. An applied their fertilizer value. This set the stage fur high-analysis research approach is being directed toward such problems liquids which later became available from ammoniation of as acid purification, solid phase identification, impurity electric-furnace superphosphoric acid. The chemistry of sequestration, and others which have resulted due to the these four-to-five component ammoniuin polyphosphate use of wet-process phosphoric acids. Since wet-process systems was not extremely complicated and the saturation acids are now being supplied with increasingly higher isothernis of tlie pure systems were developed (2). The impurity concentrations, a major effort is now directed additiun of micronutrients complicated the chemistry tu toward the production of stable suspensions; the precipitated some extent and led tu a significant quantity ofknowledge impurities will be more acceptable among other suspended about the sulubilities and characteristics of inany new material than in a clear liquid. compounds. These included both the saturating micro- A primary problemin the study of suspensions is isolating nutrient compounds (3) and the compounds of ammonium and determining the chemical and physical properties of the polyp1iosph;ite (4) dissolved in these solutions. This infor- impurity precipitates. These characteristics such as citrate mation described the limits of nutrient levels that could be solubility, crystallite size, water solubility, and gel-forming obtained from ilic pure systems. The development ofchro- properties will be reflected in the overall properties of the niatugraphic analysis for the polyphosphate species pruvided suspension. Liquid and suspension fertilizers, in essence, a method by which the chemistry of the hydrolysis of are very similar-a liquid approximates a suspension with polyphosphates and their ultimate return to orthophosphate solids removed whereas a suspension is a liquid that contains could be monitored. solids. Both have a liquid base that is almost saturated with Advances made in the polyphosphate chemistry of liquids fertilizer salts. Thus, knowledge gained from one type of prepared froni highly pure electric-furnace acid were in- fluid can be applied to the other. In order to compete with pressive arid provided a background for understanding solids, fluids must have a high plant food content and similar products prepared from the less expensive wet- provide a guaranteed stable storage life over a given period process superphosphoric acid (WPSA). However, there was of time. These are the goals of the chemical research being a g;lp in needed fundamental knowledge presented by the conducted today.

‘J PHOSPHATE SOURCES 000 FOR FLUID PRODUCTION I1 II II 13.0.1'. 0 - P-0 - P - 0. H The initial phosphate source for the production (if fluid I I 1 phosphate fertilizers was orthophosphoric acid prepared 000 from elemental phosphorus in excess of that required by I I I U.S. industry for use in food products and detergents. H H H This excess elemental phosphorus was burned, hydrolyzed, and converted to phosphoric acid for amnioniation to Phosphate Rock Sources 8-24-0 base solution. The basic NH, -H,PO.-HIO isotherms (11) provided data for selection of a satisfactory base Tlie impurity lo;id in WPSAs reflects tlie conceiitratioiis solution. Many pure systems ofmixed fertilizer compunents foulid in tlie phosphate rock source and varies widely among have been reported (1, 12, 13, 14, 15). The iidditional the phospliate liiw materials of North Carolina. Western fertilizer materials included in these studies were ammonium United States. and the various fields in Florid:!. The prim;lry nitrate, urea, potassium chloride, and animoniuni sulfate. rock source ill tlie United States uscd lor research 011 fluid The plant nutrient values for these systems at 32OF (0°C) fertilizer production is central Florida; must fluid prmesses ranged from 15% for saturated KCI to 38% for liquids at have been developed to ;~ccommodate the impurities in the invariant point between NH4H, PO4 and (NH,), HPO, this rock. Merchant-grade xids (54% P, 0,) produced horn Real progress for fluid fertilizer production began with this rock have typically contained about I .5% Al,O,. 0.5% the development of electric-furnace superphosplioric acid. MgO. and black carboiiaceous matlei. all of wliicli interfere The production of this acid follows the stepwise oxidation with tlie productiun of clear, stable liquids. However, of elemental phosphorus with subsequent limited hydroly- suspension processes using Florida xid are well developed sis. Oxidation of phosphorus can occur at three levels: and can result in a stable suspension fertilizer (17). Nnrth Carolina phosphate rock is high in CO, and organic matter. P, to, + 2P, 0 hypophosphite (1) mid is usually calcined prior to acidulation to remove

P,t30, + 2P,03 phosphite (2) organics and prevent fiiaming. Tliis process is beneficial in preventing the formation of black carbonaceous inaterial P4t50, + 2P,05 phosphate (3) in fluid fertilizers. The high magnesium cunteiit ofthis rock The first two oxides arc unstable and oxidize stepwise is being controlled in tlie :acid product to the extent iht it to the stable P, Os form with phosphorus having a valence is the most acceptable pliosph;~tesource in this counry for of t5. The phosphorus oxides are hygroscopic and react successful liquid fertilizer production. A contributor to this readily with water to produce the three acids: success was the developme~~tuf the pipe-reactor process by TVA for the prodoctioil ol' very high polypliosphate (up P,0+3H20 + hypophosphorous acid (41 2H3P0, to 80% uf the P, 0, j products (I 8).. P,03t3H20 + 2H,P03 phosphorous acid (5) Wester11 U.S. phosphate rock SOUIKXS are high in most impurities including organic ni:itter and are less suitable for P2O5t3H,0+ 2H3P04 phosphoric acid (6) fluid production. However. wlien extensive researcli studies These equations provide the basic reactions for the are applied, similar to those on Florida rock. a suitable electric-furnace acid process. For polyphosphoric acid, the suspension fertilizer sliould result. water in equation 6 is limited to a value between 1 and 3 so that condensed phosphoric acids are the final product. Wet-Process Acid These acids have the general formula of Hn+,P,O,n+, I where n for practical purposes varies from 1 to about 15 The advent uf wet-process phosphoric acid (WPA) and (theoretically infinity) and their weight percentage distribu- its condensation product. WPSA. as iiii economical source tion vanes with the final P,O, analysis (16). Acids with of P,Oi for fluid fertilizers significantly increased tlie need P,05 values above 80% are syrupy viscous liquids with for chemical information to control tlie various types of precipitates that are detrimental to the tluid properties uf more of the longer chain length acids whereas acids around these fertilizers. The impurities in WPA play a significant 75% P,O, are more fluid and contain a mixture of poly- role in deterniining the final grade and quality of both lluid phosphates mostly having a chain length from I to 4. These and solid fertilizers. The cations of iron. aluminum. mag- are named orthophosphoric acid (H,PO,), pyrophosphoric nesium, and sodium will reduce tlie possible nitrogeii grade. acid (H4Pz07), tripolyphosphoric acid (H,P,O,,), and whereas sulfate and lluoride lower tlie P205 gradc. It is tetrapolyphosphoric acid (H,P40, ,). The bonding con- possible tn produce standard fluid grades. such 3s 10-34-0 figuration for tripolyphosphoric acid, as an example. is: liquids or 13-38-0 suspensions, at higher impurity co~iceii-

26 tiations but the resultingliquid will have a higher salting-out The primary chemical problem associated with concen- temperature and the suspension will have a higher viscosity. trating WPSAs was directly related to the use of submerged In both cases the increased impurities will be replacing combustion which resulted in localized overheating and some of the water used tu fluidize the fertilizer. caused (Fe, AI)H2P30,0 and (AI, Fe)(P03)3 to precipitate Early in the study of WPSA liquid products, attempts (6). These solids not only interfered with the concentration were made to reach higher and higher levels of polyphos- process but also are citrate insoluble and remained as stable phate (18) in the ammonium polyphosphate liquid using precipitates in ammoniated products. The concentration only a WPA phosphate source. Also. numerous reports and process for these acids effectively eliminated the precipi- patents were issued fur the purpose of defluorinating wet- tation problem caused by fluoride since any unevolved process acids in order to prevent MgAI(NH,),(P, O,), F, fluoride is so strongly complexed with the aluminum that 6H20 from precipitating in the liquids. This cure was it is unavailable for any other reaction. For example, figure 7 only partially effective because the stable precipitate now is a plot of the Al,O, and F content of 11-37-0 high became Mg(NH4)2P, 0,.4H20. Cunsequently, the next polyphosphate fertilizers prepared from several commercial effort to ubtain a stable, clear liquid fertilizer was magnesium WPSAs. This flqoride was not available for the precipitation removal from the original phosphate rock or from the WPA of MgAI(NH4),(PzOIJ2 F, 6Hz0 even though it ranged intermediate. Like the fluorine removal processes, a large up to 0.18%. On the other hand, as little as 0.02 F as number of processes for magnesium removal were proposed NH,HF, would cause the precipitation of this compound and patented but primarily for economic reasons most did in these liquid products. The MgO content of the liquids not prove to be adaptable to fertilizer systems. The two was around 0.6% and Mg(NH,),P,O1.4H, 0 eventually most promising processes appear to be the precipitation of formed in all samples as the higher polyphosphate species MgH,P,O, from 70% P,O, WPSA (6) or the precipitation hydrolyzed to pyrophosphate. of MgSiF6.6H,0 from 55% P,O, acid (19). Considerable The process of increasing the polyphosphate'values for effort is still being applied to the purification of WPA, sequestration purposes in 10-34-0 and 11-37-0 liquids especially since there is increased use of lower-grade rocks reached a practicalslimit with the use of 68%-70% P,05 containing larger quantities of cation impurities notably WPSAs and anhydrous ammonia in a pipe reactor (18). The Mg, AI, and Fe. Probably, some of these purified acid prod- polyphosphate content reached 85% of the total P,05 ucts will be more applicable tu feed- and detergent-grade (I 5% orthophosphate, 35% pyrophosphate, 30% tripoly- products with the by-product sludge fraction going to the phosphate. 10% tetrapolyphosphate, and 10% greater than solid fertilizer industry. tetra) and the storage life was extended to about 3 months. As the cost of electric-furnace superphosphoric acid to However, eventually Mg(NH,),P,01.4H, 0 precipitated produce 11-37-0 increased, a more cost-effective approach in these high poly liquids and it was difficult to increase was developed whereby mixtures of 11-37-0 and ammoni- this 3-month value or to predict any finite storage life for ated merchant-grade wet-process acids were produced. This any particular acid product. provided two desirable effects: (1) the impurities in the WPA were diluted tu help prevent precipitation, and (2) the polyphosphates from the 11-37-0 sequestered the impuri- SOLUBILITY DIAGRAM ties. Later the economics of electric-furnace acid limited FOR MACRONUTRIENTS the use ofeven this process for making clear liquid products. WPSA was developed tu obtain the polyphosphate compo- The maximum fertilizer grade of liquih materials de- nents necessary for sequestration. This new acid form was pends on the saturation composition of the major nutrient components at temperatures down to 32'F (OOC), which produced by continuing the concentration process of filter are likely to be encountered in cooler climates. Likewise, grade (30% P,O,) acid to merchant grade (54% PzO,) in order to obtain maximum grade, it is necessary to under- and on to superphosphoric acid (70% P,O,). An example stand the interaction of components with each other; the of condensation with heat is shown below for an ortho- best known example of which is the use of ammonium phosphoric acid and a tripulyphosphoric acid going to a nitrate and potassium chloride. Both materials have reason- tetrapolyphosphoric acid as they lose a mole of water. able solubilities in fluid fertilizers when used alone. How. Other polyphosphoric acids are formed in a similar manner. ever, attempts to saturate a fluid with either material in the presence of the other, results in salting out of potassium 0 000 nitrate. Numerous studies have been reported (1, 2, 20) II I1 II II describing the solubility relationships in liquid mixed ' HO .P - P - 0 - P-0 - P - OH fertilizer systems with and without polyphosphate. Many I I I I other reports are now available describing the phase systems OH OH OH OH involved in the ammonium polyphosphate mixtures with

27 I .5 I I

1.0 I

I I I 0.05 0.10 0.15 0.20 F. % Figure Relation between aluminum and residual 100.1 7. N fluorine contents of ammoniated superphosphoric acids prepared from wet-process acids Figure 8. System: ammonia-superphosphoric acid-urea- ammonium nitrate-potassium chloride-water at 32OF and without urea amendments (21, 22, 23), and with potassium values (24). The heat of ammoniation of electric- furnace superphosphoric acid from which these fertilizers are obtained is also available (25). All these reports have extensive literature references to earlier reports which describe simpler systems, frequently at 32'F (OOC), 77'F (2SoC), and 122°F (50°C), so that the basic descriptions of the solution and solid phases in fluid fertilizers are avail- able including the optical, X-ray, and infrared properties of the solid components. An examination of these phase systems shows that higher analysis fluids can usually be obtained from multicomponent mixtures and from mixtures of several polyphosphate species. Interaction of the macronutrients is demonstrated by comprehensive diagrams (20) which are reproduced in figures 8 and 9. The total plant food content (YON t 100 %PzOs + %K,O) is shown on lines of equal concentration N and is greatest in solutions of pure ammonium polyphos- Figure 9. System: ammonia-superphosphoric acid phates. Since chloride is a diluent for all three macro- urea-potassium chloride-water at 32'F nutrients and KCI has a low solubility, the use of this component results in a very low grade liquid fertilizer. plant food value of 53%. Systems involving potassium The data in figures 8 and 9 (20) show that plant food ortho. and pyrophosphate (no KCI) with and without urea values up to 45% could be obtained in pure ammonium and ammonia (22, 24, 26) showed that on the high polyphosphate systems. The addition of other nutrients potassium side, nutrient values up to 64% could be obtained such as urea, ",NO,, and especially KCI reduced the at 32-F (OOC). Many new compounds (4, 24. ?6,32) were total nutrient concentration. Because of high shipping isolated, identified, and characterized during the deter- costs and the low analysis of these trinutrient systems, mination of these phase systems. The precipitating solids these solutions are useful at small mixing plants close to that are in equilibrium in the ammonium potassium puly- application sites where one-time applications of special phosphate system are shown in table 5. Some of the ratios are required. conditions under which these compounds are encountered Fundamental solubility diagrams (23,26,27,28, 29.30, in production processes for fluid fertilizers have been 31) are available that give the maximum grade possible at reported (33), and others are described elsewhere in this various pH values and nutrient ratios for pure systems. text. The highest N t P205 value reported was 57.45% at The complexity that can be encountered in a simple pH5.61(27)in the system NH,-H,P04-H,P,0,-HsP,0,n- four component fertilizer system is shown in figure IO H,PIO,,-H,O at 32'F (OOC). Similar systems including where the system K2O-NH,-H4P2O7-H2Oat 77'F (25°C) urea (21) but without the benefit of H,P,O,, reached a is projected on the K20-(NH4)20-H20 face. The figure

28 higher analysis and higher temperatures. Otherwise, these products were trouble free.

Electric-Furnace Pipe-Reactor Products The successof liquid fertilizers produced in a pipe reactor from WPSA precluded extensive studies using the more expensive electric-furnace acid. However, several tests were conducted with this high-purity acid, and variations in the process resulted in sume unusual redistribution of the poly- in 4 phosphate species the ammonium polyphosphate (AFT) melt (4, 35). The parameters causing this redistribution have not been fully researched but limited tests have shown ] , INH&HW, HzO'tc that a preheated, 78.1% P,Os acid with 45% of the phos- " 0 5 b I .-OL p!iatc at P,0,4- rcsu!ts in an M'P mclt having 32% of the IO 15 .\,A20 25 %lNH41 $2 phosphate as pyrophosphate. Likewise, an 83.5% P, 0, acid with 9% of the P,Os above a chain length of 9 was Figure 10. The system NH,-K,O-H,P,O,-H,O at converted. after preheating, to an APP melt having 36% 25OC projected on the (NH4),0-KzO-(Pz0, + H,O) face of the P,O, above nonapolyphosphate. This high-poly shows that each di, tri, and tetra substituted compound amnionium phosphate fraction is water insoluble of aninionia and potassium with pyrophosphoric acid has (2.0.1g/l00 cc H,O) and has the formula, (NH,)n+2- its own field of stability as well as the more acid salt P,O,,+, where n was estimated from the chemical ratio K1H,(P,07)z. Not only is there a lack of isomorphous of NIP to he 51 (4, 36). Studies of pipe-reactot products, substitution between the aninioniuni and putassiuni end- which became polymerized solely by the temperatures members. hut also the mixed ammonia potassium di and of ammoniation (%400°-7000F, 204°-3710C), have shown tri substituted pyrophosph;ites have different crystalline polyphosphate distributions comparable to acids having structures froni these end-niember compounds and conse- the Same polyphosphate level rather than the unusual quently their owti unique stability fields. distributions described above (18).

Tank Reactor WPA INFLUENCE OF PRODUCTION METHODS The use of tank reactors to produce stable liquid grades (10-34-0 or ~11-37-0)from anhydrous ammonia and high Electric-Furnace Acid conversion WPSA was reasonably successful (37). At the low fluoride concentrations found in these products, Tank-type reactors wereoriginally used for the ammonia- all the fluorine is complexed with aluminum so that it tion of 80% P, Os electric-furnace superphosphoric acid cannot precipitate as MgAI(NHa)s(P,0,)2 F,.6Hz0. Thus, tii produce 11-37-0 solution and 12-40-0 suspension grades the stable solid phase is Mg(NH,),P,0,.4H20, which with a very high polyphosphate content (34). The high- precipitates at about 0.2% MgO in these liquids but can analysis acid with its multicomponent polyphosphate remain supersaturated at about 0.3%-0.4% MgO for a few composition was readily obtained by liniiiing the water of weeks. hydration during the process of converting elemental The developmental program for 13-38-0 orthophosphate phosphorus to pulyphosphoric acid. The 12-40-0 grade, suspension in tank reactors has been successful (17) with after 1 to 3 days, would frequently contain a voluminous only small applications of chemical parameters. Fluosilicic precipitate uf (NH4)sP,01 ,-2H20 (34). This did not acid was developed as a crystal modifier for NH,H,PO, restrict the maximum grade of these suspension fertilizers for the purpose of producing thin needles instead of the since these thin plate crystals would not settle. Likewise, thick rods that tend to settle rapidly (38). The major a redistribution uf the polyphosphate species due to emphasis on developing 13-38-0 suspensions was limited hydrolysis would quickly eliminate tripolyphosphate as to central Florida acid. Several exploratory tests using ;I saturating component and the crystals would redissolve after 20 tu 30 days. After about 2 or more years' storage, North Carolina acid having a high magnesium and a low continued hydrolysis of the polyphosphate species in aluminum content required chemical control (39) to solution eventually resulted in the precipitation of prevent the formation of gelatinous precipitates. A simple (NH,), HP04 and naturally this occurred more rapidly at relationship between the cation impurities and fluorine

29 conceiitration in suspensions (17) was developed in order phate (8 Forins during localized uverammoniation (23) <4.8 Comprises the stable crystalline phase when pipe ieactor solids are saturated in water (without additional NH,) (4, 23) S.0-6.1} Piecipitated-froni liigli analysis tluids high (>SO%) in P, 0, c~titeiit } (4. 23) 1 >6.1 }

<3.7 Precipitates in ilnmoiiiu~iitripolyphospllate system at 9% N 2nd 43% P,Os (4. 31) 3.7-5.0 Precipitates in high poly fluids having a t ripu1yphosph;lte content above 38% (4. 3 I) 4.9-5.4 Stable compound in the NH,-H,P,O, o-H20syste111 at 37% P:O, (4, 3 I) >5.4 Forms in. high poly fluids Ixiving ii tripolyphosphate above 35% (4. 3 I)

+6+ Forms when NH4N03 is used tu augnient the N content~ofanimonium polyphosphate fluids high in pyro content (4) 0-6 Precipitates on dissolution of pipe reiictur solids made froin 80%-90% P20, acid (4) 4.6-5.2 Forms in the mixed amiiiuiiioin-potassium pyrophosphate system (24)

4.9-8.6 Forms in the inixed aninioniuni-potassiom pyrophosphate system (14) <1 Forms when KH,PO, is charged to reagent H,PO, (36) 1-9 1 Forms in potassium orthophosphate systems (26) 9-13 <0.43 Stable phase in the potassium pyrophosphate system at 35% P205 and 12%K20(26) 0.43-2.1 Stable phase in the potassium pyrophosphate system at 29% P20, and 12%K20(26) 3 Forms when heated (500°C) K,HPO, is charged to acetic acid (16) 2.1-6.4 The stable potassium pyrophosphate in a potassium polyphosphate mixture (26) 6.4-10.1 The stable potassium pyrophosphate salt in a potassium polyphusphate mixture at 35% Kz 0 and 30% P,O, ~.(26) . >10.1 The stable pyrophosphate is the potassium pulyphosphate system at 38% K20 and 28% PzOs (26) ____- acornpounds most commonly found in fluid fertilizers.

30 111 a single-stage continuous reactor, the acid and contribute to the extended storage life of these samples amlloiiia are added simultaneously to a tank of the fluid as tested. The magnesium concentration in the final prod- product at a pH above 6. Tlie impurities are very insoluble ucts was only 0.3% to 0.4%, the fluoride content was and precipitate rapidly as a colloidal gelatinous form of 0.1% or less, and the P,O,"' concentration was 40% ' FeNH,(HPO,), .nH, 0 and AINHa HP04 F, .nH, 0 which or less of the total P,O,, Since the precipitating phases trap and retain a large amount of water. Recent data of Mg(NH,), P, 0, .4H, 0 and MgAI(NH,h (P, 07hF, ' I (42, 43) indicate that tlie fluoride compound is stable and 6H,O are both pyrophosphate, the lower P2074' in these I does not change. However. the FeNH,(HPO,),.nH, 0 is products, as compared to 50% P,0,4- in tank products, unstable and this amorphous forni undergoes dissolution tends to reduce the rate of precipitation. However, hydroly- with precipitation of the lesser soluble and niore gelatinous sis of the higher polyphosphates to pyrophosphates is FePO,.nH,O which has a strong affinity for water and rapid as compared to the hydrolysis of pyrophosphate foriiis as a very viscous gel. This reaction appears to con- itself (45). Thus, the long-term storage value of these very tinue slowly during storage and depending on the crystal- high polyphosphate solutions is limited by the buildup of linity of tlie FeNH,(HPO,), may require years to reach pyrophosphate. One test sample from a commercial acid stsbi!ity. These snwrplioss precipits!es .'re frequently reached ?. level of 70% pyrophosphate after 3 months and gel-like and tend to form after the viscosity has been contained a solid phase comprising both MgAI(NH,),- adjusted with clay prior to storage. Tlie fine-grained forni (P,07), F2.6H,0 and Mg(NH4)zP,0,.4H20. if AINH,HP04F, and Fe(NH,),(HPO,), F contributes to The "direct pipe-reactor process" in which merchant- the overall surface area of the solids. grade WPA is preheated and reacted with anhydrous aniiiionia in a pipe reactor has been researched (46, 47). Pipe Reactor The heat of ammoniation of this reaction supplies 80% of the heat required for condensing about 30% of the ortho- Products obtained by feeding aninionia, superacid, and phosphate to pyrophosphate. However, when made from water simultaneously to a tank of 10-34-0 product will, acid produced from Florida rock, these liquid products at best, niaintain the siime polyphosphate distribution as have a moderate fluoride concentration, a low polyphos- the original acid (44). The use of tank reactors for ammoni- phate concentration (all P,O,), a very limited storage life, ating phosphoric acid does not take advantage of the heat and result in a voluminous precipitate of MgAI(NH4),- of ammoniation which is removed by heat exchange or (P,0,)2F,.6H,0 after 2-3 days. They are also reported evaporative cooling. to contain citrate-insoluble P,O, (46). Two factors con- The advent of the pipe reaclor (18) greatly enhanced tributed to the failure of the direct process to produce a clear liquid fertilizer growth. Prior to this, batch amnionia- stable clear liquid 10-34-0 product: the low solubility of tion of 70% P,O, WPA to produce 10-34-0 liquids contain- the fluoride compound MgAI(NH,), (P, O,), F, .6H,O and ing approxiniately 50% of the P,O, as pyrophosphate the higher residual fluoride content (-0.6% F) of merchant- resulted in liquids with precipitation problems. While pyro- grade acid as compared to the very low quantity (%0.2% F) phosphate is effective in preventing the precipitation of left in WPSA. Research data (7) show that the quantity of iron, aluminum, and magnesium orthophosphates, the high MgAI(NH,),(P,O,), F2.6H, 0 that can precipitate in these P,0,4- usually caused precipitation of Mg(NH4)2P,07. liquid fertilizers is only limited by the least abundant 4H, 0 after an unpredictable incubation period varying component. Either magnesium, aluminum, or fluoride in from 2 days to 2 months. As discussed above, the total solution can be reduced to 0.01% by precipitation if the magnesium content is an important factor in limiting this other two exceed this value. time period but several other factors are also involved. The worst condition of precipitation occurs when a small Pipe-Reactor Scale Formation and Corrosion residual fluoride concentration (0.7% to 0.9%) is present in the starting acid. In early development of the pipe reactor for the produc- Anhydrous amnionia reacted with 68%-70% P,O,, tion of ];quid fertilizers from high-conversisn WPSA (con- WPSA in a pipe reactor produces liquids having a higher taining about 50% or more polyphosphate), an insoluble polyphosphate content than the original acid, and the precipitate was encountered in the melt within 1 to 2 10.34.0 and 11-37.0 APP liquid products have superior seconds after mixing. Rapid dispersion of the hot melt storage properties as Compared to batchwise, tank in a tank of 10-34-0 at pH 6 prevented the formation of ammoniation of the sanie acids (18). As described else- this compound. Research studies showed that the iron and where, these liquids contain polyphosphate concentrations aluminum crystallized in the melt and on the inside walls up to about 85% and remain free of precipitates for 1 to 3 of the reactor as (Fe, AI)NH,P,O, which is not only months at 100°F (38'C) with one pilot-plant product being citrate insoluble but has a very slow rate of solution in hot satisfactory for 6 months at 8O0F (27OC). Three fdctors concentrated acids. It is readily soluble in hot caustic solu-

31 tions. The parameters controlling the formation of this was beneficial to both pruhlelns by depositing a sinall compound were thoroughly studied (5) and several factors cuntrulled quantity of (Fe, AI)NH4P207 as ii piutcctive were effective in reducing its formation, ix., higher coating. nitrogen to phosphate ratio, lower temperature, higher fluoride content, lower poly levels, and lower retention Fluids from Solids time. None of these variables could he applied to pipe- reactor products so that a solid high polyphosphate APP, Solid aiiinioniuni phosphates froin WPA have hccoine free of citrate-insoluble P,05, could be produced. A lower popular fertilizers because of tlieir ease of inanufacture. temperature resulted in lower polyphospliate, a higher high analysis. and good storage properties. The pruduction nitrogen to phosphate ratio could not be ubtained except of dianiiiioiiiuiii phospliate (DAP) at tlie noniinal grade of under pressurized conditions, tlie retention time was 18-46-0 lias become a standard in the industry and allows already very small, and an econoniic;ll source of fluoride for a total impurity content of 15% (41. 48). This grade was unavailable. The lower temperature parameter was is readily obtained front high-grade rock sources or fruni eventually achieved in the form of a water jacket around clarified WPA products. However. acids from low-grade the pipe reactor to retard the formation of (Fe, AI)- rocks or sludge acids caiin~tiiieet gwde for DAP and arc NH,P,O, on the inside wall of the pipe. This skin-effect niure suitable for a inonstand;ird iiioitoantiitoiiiuiii plios- cooling did not significantly affect the polyphosphate phate (MAP) grade such as 10-50-0which iillows fur 18%- content but did extend the useful life of the pipe reactor 19% total iinpurities. The final forni of tlie ittipurities in before clean-out time was required for the renioval of the these ammoniated prnducts has recently been dctcrntined deposit in the pipe. APP melts from the pipe reactor. by Dillard, et al. (41, 42, 48) u.110 hiwe identified the containing lower polyphosphate concentrations (-25%) impurity conipounds that are in cquilibiiuni with MAP aiid could be granulated successfully without the formation of DAP and tlieir effect on suspension grades prepxed froin insoluble solids. At lower poly levels, iron and iiluriiinuni them (table 6) (43). The productioii of suspension iiiixtures (mainly the iron form) still depusited inside the pipe, frum gwnular aninioniuiii phosphate (43. 49. 50) eliiiiiniites not as the pyrophosphate, hut as (Fe, AI)NH,(HP04), the problems associated with lung-tertii storage of fluids in the form of dense crystalline spheres up to 5 mm across. because they ciin he processed as inceded from the stored This compound also is citrate insoluble and very slowly solids. Recent studies using solid iiniiiioniuiii pllospliates acid soluble hut does not form in the cooled melt. One from different acid sources showed that 11-33.0 suspensioil method for removing these pipe deposits is to heat the grades made froni MAP withvut polypliospliatcs bec:inte removed section of pipe at 600" to 700°C su llial frolhed excessively viscous :~ndcompletely solidified aftel I week glass of iron and aluminum polyphospbate can flow front of storage. These results could iiot bc correlated with the pipe. The stable aluminum fluoride complex in these uperating paraineters mid varied widely depending vn tlie hot melt systems is AI,(NH,),Hz(P04)4F, (5); however, MAP simrce. Two cheniical controls previously applied to the short retention time in a pipe reactor prevents its suspensions produced directly (1 7, 43) were effective ill formation. maintaining their fluid property. A controlled fluorine In contrast to WPSA, electric-furnace superacid reacted concentration solubilized sonie of the irun atid aluminum in a pipe reactor resulted in rapid corrosion of a Type 3 16L and changed the stability region in w,liicli Irydiated gel pipe at the same zone where deposits form with WPA. forination (Fe, AI)PO,.nH,O occurred to one in which Mixing the two acids to obtain an optimum iron content AINH,HPO,F, and Fe(NH,),(HPO,),F iire stable. At

Table 6. Impurities in 11.330 prepared from wet-acid MAP ___ ~ ~~~ Compound in MAP -______Fate in 11-33-0 suspension-~ FeNH,(HPO,), (citrate-insoluble) + FePO, .nH, 0 gel Fe,KH, 4(P0,),.4Hz0 i FePO, .nH, 0 gel Fe(NH,), (HPO,), F (crystalline) + No change AINH,(HPO,), (citrate-insoluble) + AI(",), H(PO,), .4H,O (crystalline) AINH4HP0, F, + No change Mg(NH,)2(HPO,), '4HzO + MgNH4PO4.6H20 CaHPO, } + Ca(NH,), (HPO,),.HZ 0 CaSO, } (Mg,Fe,AI)-NH,-P, 0,-F (gel) + MgAI(NH,),(P20,), F,.6H,O or gel Clear solution (controlled F) (NH4),AIF, (excess F) MgAI("4)z H(P04)z F2 + No change

32 higher concentrations of iron, aluminum, magnesium, and phate, and fluoride are obtained (7, 51). Without this other impurities, the limit of fluoride sequestration is chemical control, the potential for deterioration is great. exceeded and it becomes necessary to reduce the grade to At high pyrophosphate concentration (30-50), the low 10-30-0. It is conceivable that with the use of low-grade residual fluoride is insufficient for the formation of soluble phosphate rocks and sludge acid products, it may be complexes, and MgAI(NH,), (P, 07),F, .6H,0 will precipi- necessary to reduce the grade to 8-25-0 and still need tate as thin needle crystals with a large surface area. At low fluorine in order to produce an acceptable fluid. Similar pyrophosphate levels (< lo%), the R203 (especially the controls may be required for mixed systems such as 8-8-8 iron) will precipitate as (Fe, AI)P04.nH,0 in the form of or 6-6-6 grades with or without micronutrients. an amorphous gel-like material that immobilizes a large Limited unpublished TVA data (4p 43) for several fraction of the water needed to fluidize the product. suspensions prepared with granular ammonium phosphates Experience with these polyphosphate liquids has shown from different acid products has provided a correlation for that redistribution of the impurities can result in semi- controlling the viscosity of these suspensions. These results solidification of the solution to give a nonflowing, clear show, that it is necessary to add fluorine, as HF, NH,HF,, gel. Some of these gel-forming characteristics are known or similar source, until the weight ratio of (Fe,O,+Al,O,+ (7, 17, 39, 431, but most have only been encountered but MgO+CaO/F) is near 3 or less. This ratio produces the not studied. These gels can form in low-grade mixed fluids maximum solubility of impurities and prevents the forma- such as 4-8-4, 6-6-6, and orthophosphate suspensions, and tion of gels which may become stable at other fluoride if detected in the early stages of gelling when they are still values. The highest possible grade can then 6e determined lightly thixotropic, can be fluidized again by air sparging. by the total impurity concentration including weight Long-term storage at summertime temperatures and the use percent Fe,0,+A120,tMgOtCaO+SO~+F,When the total of acids having higher impurity levels will eventually make impurities in the fluid fertilizer sum to 7.5 to 8.0%, an this gelling problem unbearable. 11-33-0 product. stable for at least IO weeks, can be pro- duced. Experience has demonstrated that this relationship is acceptable for MAP suspensions prepared from Florida, CHARACTERISTICS OF Western United States, or North Carolina acids. Other acids, SOLID PRECIPITATES such as spent pickling acids or sludge products abnormally high in Al,O, (>2.S%) or MgO (>1.2%), may require The potential for precipitation of solids in fluid fertilizers further testing to determine the optimum fluoride content. is indicated by the phase rule property of “stable salt The production of acceptable orthophosphate suspen- pairs” which in any chemical system, at any one time, pro- sion fertilizers from granular ammonium phosphate must vides for the possibility of one less solid phase than the begin with chemical control of the granular intermediate total number of ionic components. Table 7 shows the which is usually MAP rather than DAP. Water removal significant components involved in the study of fluids from should be accomplished by concentrating the acid prior wet-process phosphoric acids. Although this group provides to ammoniation because removing water between pH 2 for the presence of 12 solid phases at one time, the fluid and 5 causes the citrate-insoluble compound FeNH4- products, except for DAP in suspensions, are undersaturated (HPO,), to form and results in a serious loss of phosphate with respect to the soluble compounds of ammonium and availability in the MAP and resulting suspension fertilizer. potassium sulfates and phosphates. This eliminates 6 of the Sufficient fluorine will be necessary to provide for the 12 possibilities, and calcium can be removed by proper stoichiometry of AINH,HPO, F, and Fe(NH,),(HPO,), F control of the WPA filtration process. The number of pre- so that both FeNH,(HPO,), and (Fe, AI)PO,.nH,O will cipitating solids is further reduced by the isomorphous be insignificant (42, 48). This would require roughly 1% F substitution of aluminum and ferric iron (also the divalent for each percentage of R,O,(Al,O,+ Fe,O,)-more at high calcium or magnesium contents. Table 7. Significant components affecting Suspensions prepared from granular materials containing the chemical and Dhvsical Droverties of 20% of the total P,O, as pyrophosphate will not require ammonium polyphosphate fluid fertilizers additional fluoride to prevent formation of the citrate- Cations Anions Neutral insoluble compound, FeNH,(HPO,),, since this compo- Ca so, 2 - nent does not form in the melt from the pipe reactor. Mg ~0~3’

Fluids prepared from these solids contain a significant Fez + P,074’ quantity of the impurities as solubilized complexes Fe3+ p30,0s- sequestered by pyrophosphate and fluorides. The ultimate AI F limit of this sequestration is the production of clear stable NH, + liquids in which the proper ratios of impurities, pyrophos- K

33 micronutrients zinc, ferrous iron, and manganese) which will saturate the solution for precipitation of Mg(NH,),- sometimes coprecipitate as the same compounds. This helps (P,0,),.6H20 as cumpared to Mg(NHa)2P20,.4H,0. explain why there are only one or two solids present at In contrast to this solubility variation in APP, the six a time in most APP liquids. hydrate has a relatively high rate of solution and dissolves However, a given impurity can be present in an arninonium in water, whereas the four hydrate is water insoluble. On phosphate fluid in many chemical forms. The large variety the other hand, neither compound will ever be theoreti- of solution complexes, discussed elsewhere, can be in cally in equilibrium with a water solution since pyrophos- equilibrium with solid precipitates which exist as crystalline phate, albeit slowly, will always be hydrolyzing to ortho- compounds having a multitude of compositional arrange- phosphate. Similar data for zinc is shown in figure 13. ments. For example, in a simple ammonium orthophos- The phase relationships shown in figures 11 and I2 will phate system, mono, di, and trivalent substitutes will exist be encountered when WSAs (fluorides vaporized away) are for each metal impurity; also, double salts between these ammoniated to ammonium polyphosphate fluids. The tri- impurities and ammonia will exist-likewise various hydrates polyphosphate generated by the pipe-reactor process has of all these can be encountered at different temperatures. no detectable effect on the solubility of MgO at saturaiion The addition of just one polyphosphate, the tetravalent for the invariant points shown at points A (pH 6.30) and pyrophosphate, can increase the possible crystalline solids B (pH 5.70) in figure II. In these tests, 13% tripolyphos- by a formidable amount. This extensive complexity is the phate gave MgO saturation of 0.03 w 0.02 (pH 5.7) and primary reason’ that so little is known about the chemistry of fluid fertilizers. Even with the expert use of a polarizing microscope, which is mandatory for recognizing and identifying the solid phases, the study of a four component system is difficult and a six component system study is essentially impossible. Consider now that a fluid fertilizer made from WPA effectively contains Fe”, Fez ’, AI”, Mgzt, NH,’, H3P0,, HqP20,, F-, and H,O-nine active components. It is not surprising that “thinking research scientists” have hesitated to attack this problem from a fundamental standpoint. To simplify these types of studies, 10 - the solutions that have been studied were limited to N:P 18 i compositions-mainly from a high analysis point of view. For example, very little research has included fluids with potassium chloride (this introduces two more components) or dilute fluids such as 6-6-6 or 4-8-4 and similar grades. Several major phase diagrams are now available which 28 Y) ,I 34 36 38 a0 a2 44 TC’ThL qo,, wi* identify many of the compounds encountered due to Figure 11. Solubility in the system MgO-NH,. impurity components exceeding saturation limits in H3PO4 -H4P2 0, -Hz0 at 25OC ammonium phosphate fluid fertilizers. Most of these have been reported within the past 20 years. They include CaO-NH3-H3P04-H20(52); CaO-NH,-H, P, 0,-H2 0 (53, 54); CaO-K, O-H,P20,-H, 0 (53, 54); MgO-NH3-H3P0,- H,P20,-H20 (55, 56, 57); and Fe,O3-AI2O3-NH3- H3P04-H20 (58, 59). The magnesium phase diagram is given in figure 11 (55) to demonstrate the ease in which s.5 t the precipitating magnesium compound can change with only slight compositional changes. The solid lines represent equilibrium conditions at 77’F (25OC), and along these isotherms the solid fraction consists of two or three @@@e@@ ammonium phosphates and one magnesium ammonium [email protected] phosphate. The position of the dashed lines can vary slightly depending on the ratio of ortho- to pyrophosphate. ,.o ’@ @ @ 0 @ @ This relationship is shown in figure 12which again shows a 20 Y) LO Ea 60 70 compositional change and demonstrates the effect of pH POLYPWXPHATE. Y. OF TOT& P@, and pyrophosphate content on determining the precipi- Figure 12. MgO saturation levels vs polyphosphate tating solid phase. Note that a much lower MgO content content and pH in liquid fertilizers

34 0.01 vs 0.01 (pH 6.30) demonstrating that amnioiiiuin precipitate as the monohydrate but do not substitute- tripolyphospliate does not sequester MgO-in fact, it calcium forms as a monoclinic crystal while the zinc com- probably lias less effect than nitrate or chloride iuns. pound is orthorhombic. Thus, instead of having one precipi- Other similar pliase systems may be more complicated tlran tate that contains numerous impurities-as in (Mg, Fe, Mn, the niagnesiu~ii one shown here. For example, the zinc Zn)(NH,),(P, O,), .6H, 0-we encounter several different systeiii (8) sliows five zinc compounds instead of three and precipitates when MeZt(N€L)2P2 O,.nHz 0 is the stable tables 8 and 9 sliow that numerous water soluble and phase. insoluble iron and aluniinum compounds can be expected Iron and aluminum do not coprecipitate in fluids except in these systems. This is the primilry reason that these two under limited conditions. The formation of (Fe, AI)",- phase systems have not been studied. Because of the PO4F,.nH2O (48) is a rare occurrence when fresh coniplex clieniical nature of ;mimoniated WPSA. many of orthophosphoric acid is ammoniated before the fluoride the precipitating salts ciinnut be represented by the simple has evolved. Ferric iron dues not substitute in MgAI- pllase systems above; three of the best examples are MgAI- (NH,j, (P, 07)2F, .6H,O even at increasing fluoride (NIL), (P207)2F2-6H20. Mg3Fe(",), z(Pz07)4F2. concentrations. Less is known about the compound Fe,Mg- 12H,O, md the mixed cryrta! (Fe, Mg. Zn)(NH,),- (NH4)i i(P,@7)4Fj .!?H,O bu! chemical :na!ysis diox (P,O,), 6H20. only a trace of aluminum. Similar comparisons exist for Althougli the potential for isomorplious substitution is other iron and aluminum compounds; one major reason great. it is remarkably absent niucli of the time in fluid for the difference is the high degree of complex ion forma- fertilizers. It appears that substitution is greatest in com- tion between aluminum and fluoride which is not true for pounds such as (Mg, Fe. Mn, Zn)(NH4),(P,O,),~6H,O ferric iron and fluoride. In the pure system Fe,03-NH3- which are water soluble like the ammonium pyrophos- H,P,O,-H,Q, there are many highly soluble iron phates indicating that, in thi, compound, crystalline ammonium pyrophosphates (table 8) which precipitate at stritctnre is affected very little by the size of the divalent about 3% to 4% Fe,O,. In contrast, the comparable ciitiuii. On the other hand, substitution is restricted by size aluminum system becomes saturated at about 0.5% Al,O, and affinity for water of the divalent cation in compounds and the stable precipitates are unlike the iron compounds such as Me2t(NH,),P,07, which are wiiter insoluble and (compare tables 8 and 9). Similarly, many aluminum precipitate as various hydrates depending on the divalent ammoniuni orthophosphates have been encountered in cation. The magnesium coinpound is a tetrahydrate with fertilizer research but only a few comparable iron nu significant substitution, ferrous iroii and manganese ammonium orthophosphates. substitute as the dihydrate, and zinc and calciuni each Another solid phase, crystallographic phenomenon that is encountered in fluid fertilizers is that of dimorphic

7 forms. This is a condition where a compound can exist in either of twoor more (polymorphism) crystalline structures. This property is exhibited by Mg(NH,),P,0~.4H20, which forms both as elongated orthorhombic blade crystals (most common form) or as rectangular monoclinic plate 6 crystals both of which have their own fields of stability as shown by figure 11. This phase diagram,shows why the orthorhombic form is more cummon as demonstrated by the more extensive compositional range over which it I a 6. exists. Three other less familiar compounds have been found in fluids as dimorphic forms. These are the ortho- rhombic and hexagonal forms of ZnNH,PO, (hO), the two monoclinic forms of Ca(NH,),(HPO,),~H,O (52): and the orthorhdmbic and monoclinic forms of Zn(NH,)2P,0,. 5. n, 0. The precipitated impurities in fluid fertilizers have been identified and characterized as individual compounds along 07 09 09 09 ( with many of the compatible compounds with which they 5. 0 ,o 0, 0, dan form a stable salt pair in the same liquid (3, 5, 7, 41, 1 50 60 70 61). The calcium, aluminum, magnesium, ferrous iron, PYROPHOSPHATE AS % OF TOTAL P205 (Numbers on tipure rcprcscnt wt % ZnO) ferric iron, and fluoride compounds which have been Figure 13. Solid phases in ammonium ortho encountered in ammoniated phosphoric acids are shown in pyrophosphate liquid fertilizers saturated with ZnO tables 8, 9, 10, and 11. The pure ammonium salts that may

3s !

! m v 6 - v* - mm M *

>aJ 0 -i: 0 $ v 0' +6 f +z nz

36 0. VIVIlnyl

w z % a) cm .-c 0

m 0 n

31 m m W >d c m z m 51 D 4 3 e m 3 P

c * 0 3 W .os P $ a C C .-0 .-0 Iu cu m m wI wI a a d d Qr; Q L3 0 0 v) m

0 N m05

38 form in ammoniated wet-process phosphoric acid (6, 19), and greenhouse studies have demonstrated the value of or partially ammoniated acids, are given in table 5 along these nutrients (62, 63); however, the lack of fundamental with the potassium phosphates that have been identified chemical or physical controls for stabilizing micronutrients from phase system studies. The most common compounds in liquid grades, or providing nongelling components in identified in fluids are noted. Although unusual, as many as suspensions produced from wet-process phosphoric acids, six compounds have been found together in an 11-37-0 seems to have reduced the interest in providing and storing or 10-34-0 polyphosphate liquid at one time. These included fluids with micronutrient additives. This depressed interest ",H2POa, (NH4)2 HPO4, M~(",)ZP,O,~HZO, is most obvious at the production level but is less so at the Mg(NH,),(Pz07),~6H,0, Fe(NH4),P,0,-2H,0, and Mg- distribution site where the mixtures can be applied to the AI(NH,),(P,O,),F,.6Hz0, which are also known as soil before interaction problems arise. Chemical controls "stable salt pairs," i.e., any two or all of them can exist are needed to provide sequestration of these cations both together in the same solution without interacting. during formulation, storage, and distribution of the fluid The possible compounds range from very soluble in products. water to citrate insoluble. Some form as crystalline anhy- Limited progress in the use of micronutrient sources drous compounds and have little effect on the solvent water has demonstrated that essentially all of them are unstable while others are amorphous or even gels (very viscous, when mixed with the components in fertilizers. The inter- clear liquids) which can totally destroy the fluidity of the action with fluid grades occurs rapidly because of the gross fertilizer products by consuming the (so called) free-water. difference between the micronutrient mineral and the Others are intermediate such as the thin needle crystals saturated ammonium phosphate solution. Figures 12 and (0.5 X 0.5 x 10 microns) of MgAI(NH,),(P,0,),Fz~6H,0 13 show the reaction products and saturation composition which not only have a tremendous surface area for thicken- for magnesium and zinc in liquid polyphosphate products. ing the fluids but also remove 6 moles of solvent water for Micronutrient sources such as sulfates or oxides are fre- each mole of magnesium. All of these compounds have an quently added as a powder to solid granular fertilizers effect, some profound, on the chemical and especially where the same reactions occur but at a much slower rate, physical properties of the fertilizer products and they can This may result in caking problems with the reaction prod- be encountered during simple unit processes. ucts serving as the bonding phases between granules. The micronutrient fertilizer reaction products which have been isolated in fertilizer mixtures are shown in table 12. More MICRONUTRIENTS IN FLUIDS compounds have been characterized for zinc than the others, primarily because much more research has been With the exception of boron and molybdenum, which performed in trying to use fertilizers as zinc carriers for do not present solubility problems, the micronutrient corn fertilization. Also, a complete phase system study cations that are desirable in fluid fertilizers are mainly (Zn-NH,-H,P04-H,Pz0,-H,0)(8) is available for the zinc, ferrous iron, manganese, and copper. Crop response region of interest to fluid fertilizer scientists.

-~ Table 12. Micronutrient precipitates in fluid fertilizers (2, 60) Estimated Composition solubility Remarks F;NH,PO~.H~O

39 Most of the solubility data available for micronutrients successful by the fluoride sequestration phenomenon but in fluid fertilizers (table 12) are metastable values obtained other aspects should be studied also. It should be remem- from multicomponent commercial fluids as opposed to the bered that elevated quantities of several micronutrients will equilibrium solubilities that are obtained in pure systems. be accomplished at the expense of the nitrogen grade of These values can vary widely depending on several factors: the fluid. It may be necessary to apply a purification the micronutrient source (Le., ZnO vs ZnSOa), the point process to remove the unwanted ferric iron and aluminum in the fertilizer process at which it is added, the quantity of so that the fertilizer can accommodate the more desirable coprecipitating elements, the quantity of sulfate or fluoride, divalent metallic nutrients. and the temperature environment of the fluid after the micronutrient is added. These saturation concentrations are usually higher than the equilibrium values, most likely STABILIZING FLUIDS because of supersaturation due to elevated temperatures BY FLUORINE SEQUESTRATION or ligand complexing. All of these conditions can change during stofage and cause undesirable precipitation or The saturation compositions for any given compound gelling. in fluid fertilizers depend on the solubility product and The typei of micronutrient source can introduce unsus- activity of the component ions. In fluid fertilizers, these pected pl;oblems which cannot be foreseen. For example, values are impossible to obtain at our present level of ZnSO, is readily soluble,’while ZnS is essentially insoluble knowledge because the precise ionic formations that in fluids, and ZnO has been observed in fluids with a coating are present at any given impurity concentrations in these of reaction product, the thickness of which controls the dis- systems are not known. ‘Also in order to determine the solution rate of the residual zinc oxide. Also, some metallic activity of a complex ion, its equilibrium constant relative oxides have a varying rate of solution depending on the to its ligand and other complexes having the same ligand degree to which they have been “dead burned” during are necessary. Theoretical approaches to defining the preparation. Another problem can result from isomorphous chemistry of fluid fertilizers have largely been ignored pri- substitution of one element for another. For example, if marily because of the magnitude of these complex ion ferrous iron is substituted for zinc in (Zn,Fe)SO,, a ferrous formations. At the present state-of-the-art, saturation limits iron precipitate will result. Ferrous iron from corrosion of for each element have to be reported with limiting condi- storage tanks or reduction of ferric iron and magnesium tions because chemical principles have not been deter- impurities from WPA sources will reduce the saturation mined to predict precipitation reactions. Many simple phase solubility of zinc (and vice versa) in fluids where (Fe, Mg, systems have been reported but none of them include the Zn)(”,)6(P,0,)2.6H,0 is the stable solid phase. This sensitive element fluorine which is always present in.WPA effect is similar to “mass action” where a high concentra- fluid fertilizers and has a profound effect on the other tion of one constituent will reduce the saturation level of impurities and on the ultimate physical properties of the another. This interaction of divalent cations also results product. Solubility products can be applied to pure fertil- in micronutrient deficiencies in soils even when sufficient izer systems in order to predict precipitation concentrations. quantities of a given element are present. High levels of For example, in systems such as MgO-NH,-H,PO.,-H,O, calcium and magnesium in a well-limed soil will prevent the specific complex formation is controlled mostly by pH zinc upt$ke (62) for this reason. Simple solubility values value which also controls the composition of the precipi- for single micronutrients in fluid fertilizers are only relative tant and its level of saturation. Because of the low level and must be considered along with coprecipitating ions. of magnesium over most of the system, the ratio of The. solubility of a micronutrient in ammonium pyro- magnesium to ammonia is not an effective parameter. How- phosphate is higher than the comparable metal ammonium ever, when a fifth and/or sixth component such as pyro- I orthophosphate at the same pH. However, this gain is phosphate or fluoride is added, the type of saturating usua)ly minimal or influenced greatly by the other factors complex is more controlled by the MgO/F and Pz0,4’/F mentioned above to the extent that high solubility values ratios. Varying ratios provide afferent complex formations for micronutrients cannot be guaranteed in polyphosphate in solution and have a major effect on the saturation fluids. The primary advantage gained by polyphosphates, composition and, more importantly, the actual composition in this respect, is available in high poly, pipe-reactor fluids where a high quantity of polyphosphate above pyrophos- of the equilibrium solid phases, some of which are very phate is present and will not precipitate the cation soluble while others are very insoluble. At certain combina- components at the concentrations found in fluids. tion of ratios, the five component fertilizer system (MgO- The solubilization of the divalent micronutrient elements H3P04-H,P,0,-F-H20) can result in a water-white, clear, ,in fluid fertilizers is an important chemical phenomenon almost solid gel which is extremely difficult to biing to :that needs more research study. This goal may be more a fluid condition. These problems can become impossible

40 when iron, aluminum, and an unknown quantity of sulfate tlie hydrolysis of polyphosphates above pyrophosphate are also added. (45). The half-life formula (presented in the hydrolysis In order to relate to the astrononiical chemical variations section) to determine the degree of hydrolysis of poly- ; niay be encountered in fluid fertilizers prepared Srorii phosphate above pyrophosphate (P,O1 05- and higher) WPA. it is necessary to limit many of the paranieters that shows that unconiplexed iron and aluminum are detri- are involved. Fur example, in fluorine sequestration studies mental to the storage life of the higher polyphosphates. we limit pH to 6.0-6.3. plant nutrient content to 10-34-0, However, fluorine can control part of this effect by com- polyphosphate content to IS%, 30%, or 50% of the total plexing the iron and aluminum into ions that are not P20s, and we eliminate sulfate as a probleni because of effective as catalysts for the reaction. its low concentration level. Under these restrictions and Another chemical study involved the use of fluorine to others such as an aluiiiiiiuni to magnesium mole ratio of change the impurity ratios in suspension grades to maintain about 1.8 (wt % AI:Mg - 2 as in Florida rock), several a stable viscosity value below 1,000 centipoises (17). Again, chemical relationships have been determined to help it was found that the ratio of F/Al + Fe. + Mg on a mole control the physical properties of the fluid products. basis was more important than the total quantity of the Ferrws iron is very inru!ub!e and frequently encountered cz!io" inipurities. Even though a mole ratio of me w?s in these prilyphosphate fluids as Fe(NH,), P20,.2H,0 recommended, the study was limited to Florida acids which can be easily prevented by oxidizing to the ferric and variations in the ratios of A1:Fe:Mg were not tested. state in hot acid prior to ammoniation. Tests of typical For example, it is not known if the fluoride addition solved acids show that 0.5% ",NO, or even !as HNOl is a magnesium, iron, or aluminum problem or what effect sufficient for this reaction. In one study, it was determined other parameters such as pH, temperature, or fertilizer that tlie proper ratio or fluqide to iron, aluminum, and grade have on the fluoride requirement. magnesiuni in solution would maintain the clear liquid All these examples are given to show tbat saturation property of the polyphosphate liquid. This relationship limits required for maintaining a stable fluid fertilizer do was expressed as ii sequestration ratio (S.R.) thusly (7, SI): not provide an explanation for solving the physical problems encountered. Notice that in the sequestration wt 9'0 F S.R. = -~ ratio equation and the suspension stabilization study above wt % MgO + 2.5 Al,O, t 0.33,Fe20, total impurity concentration is not a factor. Solubility where the proper value of S.R. for stabilizing a liquid products for the very slightly soluble components can be fertilizer is a function of the pyrophosphate level, the used to define the extent of the precipitating problem when aluniinutii tu magnesium ratio and the total ionic strength, it occurs; however, the saturating solids and their solubility i.e.. the effective S.R. would be higher at higher pyrophos- products are not known for fluid fertilizer under the most phate concentrations and in dilute fluids such as 4-8-4 as favorable conditions such as a clear liquid fertilizer. For compared tu IO-34.0. This fluoride sequestration required example, in a clear, sequestered liquid fertilizer the stable about 1.2% F for a typical 50% polyphosphate liquid pre- solid phases, with respect to the impurities of iron, pared from Florida phosphate rock. Lower polyphosphate aluminum, and magnesium, are not known. The reported ., concentrations require less fluoride (7, 51). This same phase systems data for these metals in ammonium poly- chemical relationship can be applied to suspension grades phosphate solutions cannot be remotely applied to similar to control viscosity values during storage since fluoride can systems containing fluoride. These data in the presence of control. both precipitation and gel formation which are the fluorine will be necessary before the chemistry of fluid primary problems associated with suspensions. Since fertilizers and the corresponding physical properties can be it is known that very small (sometimes 0.1%) fluorine controlled. increases can cause a heavily precipitated gelled product to Experience has shown that the residual fluoride in fluid be a clear, free-flowing fluid, a solubility product relation- fertilizers is usually insufficient to provide the benefits ship cannot be used to explain this effect. However, an desciibed above. Thus, an inexpensive source is needed to equilibrium shift from a very insoluble magnesium complex augment the 0.5% fluoride that is usually present in to one that is soluble can occur; this is similar to changing commercial fluid products. The use of purified HF or from a stability field of an insoluble compound to one of NH, HF, is economically prohibitive and the inexpensive a soluble salt in a phase system study. Thus, the ratio of the impurities which determined the type of complex H, SiF, is not suitable. However, technology is available present is more important for understanding precipitation (64) for readily converting off-gas scrubber solutions from and gelling mechanisms than is the total impurity composi- phosphate rock dissolution processes to concentrated solu- tion. tions of NH4HF, by simple ammoniation. This not only Impurity ratios were also found to be important in provides pollution abatement but also produces a useful determining the catalytic effect of iron and aluminum on fluoride by-product.

41 HYDROLYSIS OF POLYPHOSPHATES At tlie water content required for fluid fertilizers, tlie IN FLUID FERTILIZERS polypliospliate components are metastable arid eveniually hydrolyze (react with watcr) to the stable vrtliopliospliate A theoretical discussion on the hydrolysis of polyphos- form. Tlie scission ofchain pdyphosphare in fluid fertilizers phate chains and rings with many references to early data usually (ant always) occurs at tlie end-member phospliate is available in Van Wazer's Phosphorus aru' Its Compounds and rcsults in one or~hopIiosp1i;ite axid tlie chain member (65). Considerable updated infolmation (2. 30, 31, 33, 34. next lower down. Siiicc tlic liigher spccics have a faster 45) is available for polyphosphates in special enviroiimeiits hydrolysis rate, tlie initial effect is for each cliain length such as strong ammonium and potassium phosphatc solu- to increase iii quantity at tlie expense of tlie next higlier tions or strong sodium and potassiuni phosphate detergents. species until a balance between the tot;il quantity and The surfactive and peptizing properties of both cliain and hydrolysis rate is attained for each succeeding species. This ring polyphosphate solutions have been developed for use hydrolysis of polypliosphates is measured as a half-life by the detergent industry but the hydrolysis characteris- time period in which one-Iialf of a given species is tics of these products, which are usually available as tlie hydrolyzed to lower chain length polyphosphates at a given sodium salts, will not be discussed here. set of eiiviroiimental conditions. This 1i:tIf-life value can be One of the most desired properties in detergent and calculated from chromatograpliic analysis which gives tlie fertilizer applications is the ability of pulyphosphates to percentage change in concentration of a given species solubilize and prevent the precipitation of di and trivalent during a time period, i.e., metallic orthophosphate compounds. These properties have 0.301t been developed for fluid fertilizers where iron, aluminum, Half-life = I% C(,/Ct and magnesium precipitates and gels at relatively low levels of impurity concentrations are almost impossible to control wliere this half-life equation is a fundaniental rate relatioii- without this polyphosphate sequestration effect. ship based only on tlie initial concentration. Ccl, and tlie final cvriceiitrativii, Ct.after time, t. Tlie hydrolysis reaction Hydrolysis of Polyphosphates rate varies for each pvlypliospha~especies atid is faster at Above Pyrophosphate higher cliaiii lengths. The viilue for the very high cliaiii length animoniuni polyphospliate (4) (NH,),,+,PnO,,,,+, Early in the study of polyphospliate sequestration of (whcie II may be as high as 2.000) has ti01 beeti determilied: iron, aluminum, and magnesium, it was apparent that hctter however. Van Wazer (6s) states that a similar chain length storage properties could be obtained at higher polyplios- found in Graham's salt (NaPO.,)x. hydrolyzes ai il rate that phate concentrations. However, when the polyphosphate is of the same order of magnitude as those for pyro- and tri- was mainly pyrophosphate as obtained in a tank reactor, polyphosphates under equivalent eiiviroiimeiital conditions. precipitation of Mg(NH,),P,O, .4H,O or MgAI(NH,), - Nevertheless, since .its solubility in liquid fertilizers is very (P,O,), F,.6H,O occurred within a few days. On the other low compared to the solubility of tripolyphosphate. the hand, when pipe-reactor products (18) were available total quantity hydrolyzed in a given time period wvuld be with a high proportion of the polyphosphate above pyro- extremely small compared to the quantity vf tripolyphos- phosphate, storage was extended appreciably before pre- phate hydrolyzed during the same time perivd. Tlie rates cipitation started. Maintaining this high polyphosphate at ambient conditions and pH 6 are sufficiently slow that content above pyrophosphate is desirable for long-term the fertilizer and detergent industry can take advantage of storage of fluid fertilizers. Several properties of these longer the complexing and surfxtive properties of these chain phosphates contribute to this. They are produced at interesting ions. the expense of the pyrophosphate which precipitates Every compunent io11 in a polyphosphate solutinii 113s insoluble compounds; they do not react with the impurities an effect on the rate of hydrolysis of the polypliospliate to as insoluble solids; they do not form gelatinous materials a lower chain length species. This effect is negligible for with the impurities, and each one is much less concentrated components that are iiisoluble and iiicapable of attaining than the pyrophosphate and thus less susceptible to precipi- a significant concentr:ition. For example. in ail iiiiiiiioiiium pyrophosphate solution, ferrous iron. calcium. and mag- tation as water-soluble compounds. These higher-chain- nesium precipitates as(Fe.Ca. Mg)(NH4)?P2O7 and reduces length polyphosphates are produced in abundance by the the divalent metal to 0.2% or less: thus their effect on tlie pipe-reactor process using 70% P, 0, WPSA (18). The poly- hydrolysis of pyrophosphate is negligible. Tliese same phosphate content above pyrophosphate- can be as high as cations would be temporarily soluble in an iminioniiiiii tri- 50% of the total P,O, in these 11-37-0 liquid fertilizers polyphosphate solution and can be expected to have a large and storage periods of 4 months are possible at room effect on the hydrolysis of tlie tripolyphospliate~atlex1 temperature with some acids. until enough pyrophosphate was formed to precipitate

42 them (55). For fluid fertilizers produced from wet-process When the fluorine concentration exceeds the value required phosphoric acid, the most significant factors controlling to complex with the iron and aluminum, the excess is pres- the hydrolysis ratesare pH, NH,, K, Fe, AI, F, and tenipera- ent as free fluoride which provides a scission of P-0-p ture. Since the desirable properties of polyphosphates are linkages that is far too fast for a practical half-life value to limited by their hydrolysis, understanding the processes be determined. However, this quantity of fluoride is well by which these parameters affect these reactions is neces- above the amount that may be present or added for seques- sary in order to stabilize this species as long as possible. tration purposes, i.e., in our hypothetical liquid at 1.5% The inverse pH effect (higher pH = lower hydrolysis Fe,O, and 1.5% Al,O, without magnesium, fluoride rates) and direct temperature effect (high temperature = would have to exceed 3% before free fluoride would be higher hydrolysis rates) are well-known and compensated available for scission of the polyphosphates. With mag for whenever possible by the fertilizer industry. During nesiuni present, this value would he higher before the production of fluids, the pH is maintained at 6.0 to 6.2; accelerated hydrolysis started. In an attempt to study above this value ammonia vapor pressures become excessive. this fluoride scission, a slurry of (NH4),HP,0,.H,0 and Hot polyphosphate solutions are provided with evaporative NH,HF, was mixed with water. The precipitation of coolers; however; temperatures conler than anihient iisiially (NH4)2PO;F.H,0 ws a!most immediate 2nd neither are not economically justified during storage to obtain reactant could reach saturation as long as the other was slower hydrolysis. Likewise, the industry cannot take present. The hydrolysis of the P03F2' ion (65) controls advantage ofdecreased hydrolysisdue to decreased ammonia the rate at which fluoride is freed again to cause the scission or potassium concentrations since high analysis fertilizer of another polyphosphate ion. grades are mandatory to reduce shipping cost. The effect The relatively low level of pyrophosphate in very high of the animonium ion lids been determined for polyphos- polyphoSphdte fluids is desirable since the common ion phates above pyrophosphate and for tripolyphosphate effect (lower pyrophosphate) retards the precipitation of itself (45). The results of these studies show wt % N to be the magnesium pyrophosphates. However, this environment about equivalent to wt % K and are presented below. deteriorates rapidly with hydrolysis of the higher chain Hydrolysis rates vs concentration length members which on the average produce one pyro- wt %N or K Half-life, days phosphate and one orthophosphate, thus the wt %of pyro- 21 50 phosphate increases twice as fast as the orthophosphate 11 160 (31). Since the half-life for the higher polyphosphates 5 400 averages 27 days for a typical wet-acid, high-poly, pipe- 0.1 1,000 reactor liquid fertilizer, the major to bulk fraction of these species will be gone after 2 to 3 months. The resulting Effect of N or K on hydrolysis Iatc of tripolyphosphate at pH 6 liquid will contain 60% to 70% of the phosphate as pyro- and 25'C phosphate, and precipitation of magnesium will be exces- These data show that the 160-day half-life for polyphos- sive unless fluorine sequestration has been maintained. phates above pyrophosphate is the upper limit that can be Thus in an impure fluid fertilizer, every soluble ion con- expected for concentrated fluid fertilizers at pH 6. How- tributes toward the rapid hydrolysis of the higher (>pyro- ever, this value is further limited by the positive catalytic phosphate) polyphosphate species to the extent that only effect of iron and aluminum on the hydrolysis of these a few months of their desirable properties can be obtained. higher polyphosphates as shown by the half-life formula Technology is available which provides relief from R20, (45): catalytic hydrolysis. The iron and aluminum can he effec- tively removed from 50% to 54% P,O, acid by adding H.L. =2.I9-0.313(wt%Fe,O3 +wt%Al,O,) potassium and precipitating (Fe, Al),KH1,(P0,),.4Hz0 +O.l19(wt%F'[2 wt%A120j +Fe,0,])1/2 (6). This not only provides a longer life for the polyphos- This relationship shows that without fluorine, a value of phates above pyrophosphate but also removes these 1.5% each of Fe,O, and Al,O, will reduce the 160-day elements as diluents. half-life for these higher polyphosphates to only 18 days. At a typical fluorine content of 0.5% F, the second part of Hydrolysis of Pyrophosphates the formula shows that fluoride complexing with part of the iron and aluminum will bring the hydrolysis rate back Except for its rapid reactions with excess fluoride, up to 27 days. Extending this compensating effect of pyrophosphate is unlike the higher polyphosphates in its fluorine to higher concentrations is limited by the empirical hydrolysis mechanisms. The primary reason for this is the formula : ability of pyrophosphate to form stable complexes with the di and trivalent cations. Pyrophosphate hydrolyzes more wt 7% Fg1.5 wt % Al,O, +OS wt % Fe,03 slowly than the other polyphosphates and this reaction is

43 reduced further by the pyro ions being bonded with the di tion of fluids is the initial wet-process acid where complex and trivalent impurities. A multiple nonlinear regression formation has already occurred before ammoniation begins program was used to evaluate the data from an experiment (19, 66). In orthophosphoric acids, the complexes are designed to test the effects of nitrogen, iron, aluminum, sometimes manifested by the various colors that can be fluorine, and phosphate on the half-life value of pyrophos- observed: i.e., a yellow coloration is believed tu be due tu phate hydrolysis. These data were used to determine an vanadium complexes; green niay be caused by utie of the empirical formula relating the hydrolysis rate to these valence states of chroniium; and ferric iron complexes with factors. orthophosphoric acid to give a purple color. Cheniical analysis has also demonstrated that a complex formation In half-life=1.6620+0.1932 wt% N-0.0696 wt% P,O5 exists between aluminum ,md fluorine in hot coiiiiiiercial +0.2987wt%Fe203+0.2054wt%AAlz0,-0.133? wt% F WPSAs (-7670 P20,). Higher aluiiiinuin concentrations will Although the interaction terms have nut been developed, retain higher fluorine values at any given temperature the equation reveals that iron, aluminum, and nitrogen (figure 7). This relationship is not simply between alu- decrease the hydrolysis rate of pyrophosphate while phos- minuiii and fluorine since tlie other ions likewise have an phate and fluoride increase it, These data were determined affinity for the fluorine. Many stability constants have been at fluoride concentrations (O%, 1.0%, and 2.8% F) which reported (67) for vilrious cations and orthopolypliuspliates were below the level required to coniplex with all the iron in solution but they cannot be applied tu strong acids or and aluminum. Fluorine complexed with iron and alu- fluid fertilizers in such a manner that we can predict the minum is effective in reducing the In value about half tlie conditions under which various ones can be expected tu quantity that iron and aluminum increase it. Under condi- precipitate. Sometimes these data can be used in coiijunc- tions of fluoride sequestration (7), magnesiumand probably tion with the types of solid compounds that ai~ein equilib. other divalent elements (fluoride sequestration uf Fe", rium with the liquid phase tu predict the types and propcr- Zn, Ca, or Mn has not been researched) are solubilized as ties of the ions in solutions. For example. as mentioned fluoride complexes and, like the iron and aluminum fluoride elsewhere, Mg(NH,),P,0,.4H20. Mg(NH,),(P,O,),. complexes, are expected to have a negligible effect on 6H20, and MgAI(NH,), (P, 0,)2F, .6H20 can exist to- hydrolysis. gether in the same ammonium polyphosphate solution: The slow hydrolysis of pyrophosphate as shown by the consequently they must be formed from similar complexes equation does not limit the storage properties of fluid in solution sucli as: fertilizers. In fact, since thc divalent metallic impurities, micronutrients, and aluminum form insoluble components MgP20,2~+2NH,++4H,0+ or gels with pyrophosphate, only 3 limited quantity of this Mg(",)z Pz 0, '4H2 0 (7) phosphate species is desirable for long-term storage of fluid MgP2O7'. t 3NH,' t (NH,),P,O; + 6H, 0 + fertilizers. Ferric iron is the only metallic impurity in wet. process phosphoric acid that forms soluble pyrophosphates Mg(~H4)6(P201)2.6Hz0 (8) in fluorine-free ammoniated products around pH 6 (7). MgP,07"+ 2NH,'+ (NH,),P,O;+ AIF,' + 6H,O-t MgAI(NH,)s(P~0,)2Fi-6H20 (9)

COMPLEX FORMATIONS When tlie sulubility products of these ions are exceeded. these compounds precipitate. This occurs regularly in APP Complex formation is the important mechanism by fluids when the MgO content exceeds 0.2% and free fluorine which impurities are sequestered in fluid fertilizers. Since exceeds 0.01% (7). Notice that other complexes have been PO,", P2074-, and F' are very strong ligands in these proposed for equ:itions 8 and 9 so that advantage can be systems, essentially no free ions of the impurity compo- gained from phase chemistry results (55) and experimental nents are present. Each type of complex acts as an inde- observations. For example. the precipitate in equation 8 pendent component with its own chemical and physical forms at higher pH values and higher pyropliospliate concentrations buth of which are shown for No. 8 as coni- effects on the other components. The low level of solvent pared to 7. In No. 9, an aluminum fluoride (insoluble) water and the types and multivalent nature of the coiiipo- is needed and AIF,' fits nicely relative to equations 7 and nent ions in solution are naturally productive of numerous 8. complex ions. Each additional component, i.e., the conden- The sequestration ratio effect that auinioniuiii fluoride sation of orthophosphate to pyrophosphate, creates has on solubilizing all three compounds can be applied to another distinct group of complex multicomponent ions these reactions. Since aluiiiinuiii requires more iliati twice which must establish an equilibrium with the others already the fluoride that magnesium requires. tlie solution reactions in the system. The simplest systeni involved in the produc- can be visualized as:

44 (7) shown in table 8. Except in the special case of AI(NH,),P,0,0H~2H20 which forms in the absence of Mg(NH,),(P, 07)2+ NH, F + MgF+ f fluorine, a pure aluminum ammonium pyrophosphate has 2[(NH4),P,O7]-t NH,'. (1 1) not been encountered. Although the iron ammonium

MgAI(NH,),(P,O,), F, + 3NH4 F + MgFf + pyrophosphate phase system has not been studied, explora- 2(NH4),P,0,-t 2NHaf +A1F4' (12) tory tests show the salts in table 8 precipitate at about 4% Fe,O,. The exception being Fe(NH,)3HP0,P,01~H,0 Thus we have gone from a solution with insoluble (low which forms slowly in ammonium polyphosphate liquids solubility) complexes, MgP,O,*- and AIF,', to one with at 2%-3% Fe,O,. This compound was still precipitating soluble (high solubility) complexes, MgF+ and AIF;. from commercial liquids high in ferric iron after 2 years The complex AIF; is postulated as the soluble form of and demonstrates a slow shift in equilibrium between a aluminum since both the aluminum fluoride (AIF, ') soluble and an insoluble complex since the same study and aluminum orthophosphate (AlHPO,> complexes are showed that increasing the iron content to 4705% Fe,O, insoluble, indicating that MgP,O,'- is the species which did not accelerate the precipitation. Several complex forma- causes precipitation in fluid fertilizers. Also, as indicated tioris are obviously required for precipitating Mg,Fe- above, the sequestration equation required twice the (NH,),,(P,O,),F4OH~I2H,0. One of them is suffi- fluorine for aluminum as for magnesium. ciently short-lived to the extent that the salt forms in fresh Exploratory studies in the system AI-NH,-H,P, 0,- high polyphosphate 10-34-0 sulution but later (up to one H,O have shown that severaldoublesalts are stable between month) redissolves. pH 3 and 8, and all precipitate at low levels of aluminum. In many cases, complexes are formed that are soluble The saturating Al,O, concentration in pure ammonium but have a strong affinity for water which causes the fluid pyrophosphate solutions is 0.3% to 1.4% whereas with fertilizer to become a clear liquid gel. This can be easily small quantities of fluorinel a saturation level with respect demonstrated (7) by starting with a pure 10-34-0 solution to an aluminum ammonium pyrophosphate was never with 33% of the P,05 as pyrophosphate and adding 0.7% reached. As more fluorine is charged more aluminum MgO, 0.5% Al,03, and 0.3% F (1.5% total impurities). dissolved until saturation is reached with respect to After equilibrating 1-2 days at 140'F (6OoC), the product (NH,),AIF, (7). This occurs at about 1% Al,O, and 1.5% will be a clear gel capable of being sliced into stable cubes. F without forming a double salt between pyrophosphate The types of complexes responsible for this will be similar and fluoride. It is apparent that aluminum ammonium to those that cause precipitation but obviously quite differ- pyrophosphate fluorides are very soluble and cannot be ent since a new physical feature (gelling) has now been expected to precipitate at concentrations encountered encountered. Other gels have also been encountered in in fluid fertilizers. Aluminum in orthophosphate fluids strong phosphoric acids, dilute fertilizers, and in ammonium prepared from WPA (17, 41) is almost always complexed orthophosphate systems, but none have been researched. with fluorine, probably as AIF,', and precipitates as the Indeed the solution chemistry of fluids is more complex insoluble compound AINH.HP04F, which can be easily than the standard phase system chemistry indicates where visualized as a combination of the two complexes AIF, + crystalline solids are in equilibrium with a given liquid and NH,HPO,'. phase. Gel formation in fluid fertilizers is a serious problem A major anomaly exists in fluid fertilizer systems which occurs slowly and without warning, resulting in a between iron and aluminum. Use of the singular term "I clear jello-like product with an ever increasing rigid struc- and A" (iron and aluminum) or R, 0, indicates a similarity ture. These gels consist of soluble complexes which have an which does not exist for these elements in fluid fertilizers. affinity for more water than is available in the fertilizer Ferric iron is totally unlike aluminum except for products product. Limited studies have shown that up to a point, prepared from defluorinated WPSA. The iron fluoride small increments of dilution with water result in gels of complex is relatively weak and can be :decomposed by the even liigher viscosities indicating that simply lowering the heat of ammoniation. When fluorine is in excess of that grade of the fluid will not solve this gel problem, required by the aluminum, thecitrate soluble salt Fe(NH,),- Gelled fluid fertilizers have been encountered from (HPO,), F can form; however, the most commonly found many sources and since little is known about them it became ' orthophosphate is FeNH.,(HPO,), which in the absence common practice to search for empirical solutions to of fluoride can coprecipitate aluminum. Unlike aluminum, each without researching its chemical chiracteristics. It which precipitates readily in ammonium pyrophosphate was determined that some gels could be broken by adding solutions as MgAI(",), (P2 O,)> F, .6H20, iron forms small quantities of fluoride whereas some required ammo- strong soluble complexes with P,Ol4. (67); in fluids it is nium pyrophosphate while others were susceptible to small most probably FeP,O; since this form can be readily additions of aluminum. The undesirable element, mag- applied to the stable iron ammonium pyrophosphates nesium, was never tried but probably would also have

45 worked. Again, as with sequestration of solids, the solution also by added suspending agents, and (3) massive gel to the gel forming problem was the elimination of solution formation will virtually stop any further reaction. complexes that have a strong affinity for water. Both prob- The presence of polypliosphates greatly retards the rate lems seem to be susceptible to variations in impurity at which the systems reach equilibrium since the final stable ratios as well as pyrophosphate content. For example, in forms must wait for the polyphosphales to hydrolyze to one study the aluminum content effective in preventing the ultimate orthophosph;ite form The many different gel formation was 1-2 times the magnesium content (39). cumplex ions that will be encountered with time will ciiiise Likewise, the sequestration ratio equation (7) was effective a wide variation in the chemical and physical properties in gel control; however, it was necessary to make allow- of the fertilizer products. Some complexes will be very ances for the aluminum-to-magnesium effect. Many of the soluble with very little ;iffinity fur water and liave little gels encountered in fluid fertilizers were never broken by effect on tlie fertilizers; some are insuluble and will precipi- single exploratory testing, and quantitative records were tate as complex cliemical compounds having ii wide range not maintained on those that did work since processing uf particle sizes and shapes:and some will be soluble with a data was rarely available for comparison. While gel forma- strong affinity for water :ind reduce the product to a clear. tion can occur in all orthophosphate systems, it seems tu be nonflowing viscous liquid or gel. especially prevalent in polyphosphate mixtures containing Very little is known about the soluble. inonhydrated ions pyrophosphate. Recent exploratory results (42) have shown since their satuliitiun concentrations have not been that many simple mixtures are capable of gelling. For exceeded and they liiive nut represelited a problem tu tlie example, in a factorial study of 9-32-0 fluid fertilizers, the development of fluid fertilizers. These are the imic forms simple systems containing 2% Fe,O, and 2% F at 25% present in the so-called "sequestered" products which are pyrophosphate resulted in a purple colored (pyrophosphate) stabilized for periuds up to 5 to 7 years wilh pruper gel but was yellow colored (orthophosphate) in the absence fluoride control. of fluorine and pyrophosphate indicating two different The role of fluoride in fluid fertilizers 1x1s been well types of complexing, both of which are capable of gelling defined. This element acts its a strung ligand with other the fluid fertilizer. Similar gelling, without coloration. was ions-ainsequently, there is no effective quantity of free encountered with aluminum at 3% Al,O,, 1% F, and 25% fluoride present in fertilizer solutions. pyrophosphate. The complex formations responsible for ' these observations will have to be controlled before a stable slurage life for fluids can be guaranteed. GOALS FOR FUTURE RESEARCH

The growth of fluid fertilizers hds been retarded by the SUMMARY persistent storage problem that continue to exist. The attributes of these fertilizers are desired by the industry aiid Fluid fertilizers made from wet-process phosphoric acid the problems are being solved. Background factorial data are complex, multicomponent chemical systems for which covering the rmge of impurity concentrations that will be the stable matrix component is ammonium orthophosphate. encountered from all types of acids are needed to predict As produced, the ammonium phosphates are in a quasi- the consequences of changing the various chemical parani- equilibrium state with the other components and are eters. The lack of these data prevents development of fluid actively trying to reach an equilibrium condition. The rate fertilizers to their full potential. At the present econoiiiic of approach to equilibrium is partially controlled by a status of the fertilizer industry, WPSA of low impurity large number of stability constants which affect both the content and purified WPA are unavailable as a siilution tu rate of formation and the rate of decomposition of each the precipitation uf impurities in fluid fertilizers. Likewise. complex encountered. The quantity and ratio of the as learned early in the study of fluids, diluting the products impurity components predetermine the equilibrium path with water to obtain less viscous, more fluid iiiaterials is that will be followed. The complex molecules in solution not desirable below 10-34-0 grades. Less expensive :itid less are continuously reacting to form more stable complexes energy consuming chemical controls must provide the since those stable in the original acid must rearrange to answers that are needed to contiiiue tlie upward develop- others that are stable in the hot ammoniated products. ment of fluid fertilizers. Since other industries arc willing Redistribution and new complex ion formation is con- to pay higher prices for purified phosphoric acid. an nppor- trolled by many parameters. and proceeds slowly. Three tunity will be available for the fertilizer industry to find a physical parameters which impede equilibrium are easily use for the low cost reject acids from these purification recognized: (1) insoluble complexes removed by precipi- processes. These reject acids along with settled sludge acids. tation cannot react until redissolved, (2) diffusion of WPAs from low-grade phosphate rocks. and spent pickling dissolved impurities is slowed by precipitated solids and acids from industrysources caii be consumed as a11 eamonii-

46 cal source of phosphate fertilizer-if the technology is 24. I:rilzicr. A. W.. E. I:, Dillard, R. D. Thrashcr,nnd K. R. Waerstad, availahlc fur controlling a wide range of impurity concen- J Ag Food Clle,n., Vol. 21, pp 7004 (1973). tidtions. ran. E. P., and 2. T. Wakefield, J, Ag Food Chern., Vol. 13, pp 99.100 (1965). Basic research, experimentation. computer modeling, 26. lhzicr, A. W.,R. M. Schcih, and J. R. Lehr,J. AgFoodChenz.. and theorizing must continue before the industry can feel Vol. 20, pp 146-50 (1972). confident in providing long-term storage fur fluid fertilizers. 27. i-w, T. U., amd J. W. Williard, J. Cliem. Eng. Data., Vol. 17, In this respect. fertilizer scientists are in a position tu benefit No.3,pp317-19(1972). frim several years (if detailed study performed on nimt all 28. I:arr, T. U., and J. W. Williard, J. Chem. E% Dura., Vol. 14, No. 3, pp 367-68 (1969). zlspects of these fertilizer materials. An iniprcssive level of 29. I:m, T. D.. and J. D. I’leming, J. Chem. Eng. Dara., Val. IO, achievenient wluch descrihes the parameters controlling the NO. 1, pp 20-21 (1965). precipitation uf inipurities in fluids from wet-process phos- 30. FK~,T. U., 1. W. Williard. and J. D. Hatfield, J. Cllern. Eng. phoric acid is available now (see references). Doto., Vol. 17,No.3,pp313-17(1972). 31. Fxr, T. I).,1. D. Fleming, and J. D. Hatfield, J. Chem. Eng. Dara., Vol. 12,No. 1,pp 14142(1967). References 32. I:razicr, A. W., J. I:. McCullough, and R. M. Scheib, J. Ag ~oodClimi.,Vul. 19, pp 213-15 (1971). I. Slack. A. V.. J. V. Hatlicld. H. 13. Shaffcr, and J. C. Ilriskcll. 33. Potts, J. M., H. B. Shaffer, and I:. D. Nix, FertilizerSolurions, J. As /bod CIWI~Z.,Vol. 7. 11)’ 404-8 (I959). Vol. 17, No. 3, pp 36, 38, 40, 44, 4647, 50-51 (May 1973) 2. Slack. A. V.. J. M. Potts. and H. 13. Shalfccr, J. Ap Food Cllerrl., (x-170). Vol. 13.pp 165-71 (1965). 34. Scott, W. C., J. A. Wilbanks, and M. R. Burns, Fertilizer 3. Vrnicr, A. \\I., J. 1’. Smith. ;md J. R. Lchr, J. Ap Food C~IC~II., Solutions. Vol. 9, pp 6-7, 10-11, 14-15 (1965) (CD-416). v0i. 14. 522-29 (1966) (~~440). 35. Hignett, T. P., and J. G. Getsinger, U.S. Patent 3,264,085, 4. I:wzicr. A. \\’., J. P. Scnith. and J. R. Lehr. J. As l.’ood Clic?~ August 2, 1966. Vol. 13.pp316-2?(1965). 36. Irazicr, A. W., U.S. Patcnt 3,342,579, September 19, 1967. 5. Ando. Junipci. A. W. I:razicr. and J. R. Lehr, f. Ap I~bod 37. Scott, w. C., 1. A. Wilhanks, and M. R. $urns, Fertilizer Ciici,i.. Vd16. pp 691-97 (1968) (CD487). Soluriuns, Vol. 12,pp 6-8, 10, 12, 14 (July 1968). 6. Lchr. J. R., A. \+’. I.’razicr. and 1. P. Smith, J. Ap /‘bod Cllm., 38. TVA Bulletin Y-136, “New Ikvelopments in Fertilizer Tcch- Viil. 14, pp 27-33 (1966) (CD-424). nology,” 12th Demonhation, pp 16-17, October 1978. 7. I’uiicr. A. W.. R. M. Schcih. and R. 0. Thrasher. .Ax & Food 39. Joncs, T. M., and A. W. I’razicr, U.S. Patent 3,697,247, Chm~..Vol. 20, No. I,PI’ 13845 (1972). Octobcr 10. 1972. 8. I:~aiicr.A. W.. I;. I,’. Dillaid. and J. K. Lchr, J. As /.bod C~lc~ll.. 40. Kcndrick, Lucian A,, Jr., U.S. Patent 3,813,233, May 28, 1974. VoI.23. pp 330-34 (1975) (X-244). 41. Dillard, E. I:., A. W. Frarier, T. C. Woodis, and F. P. Achorn, 9. IFiilicr. A. W.. U.S. Pztcnt 3,861,897, “Stabilization of Poly- .I. Ag Food Chern., Vol. 30, pp 382-88 (1982) (X-543). pliinphatc 1,’crtilircrSolutions.” January 21, 1975. 42. Dillard, E. I?.,and A. W. Fraricr, “Precipitated Impurities in IO. l:ru,icr. A. W.. US. Patcnt 3.711.268. “Stabilization of Poly- Monoilmmonium Phosphate and Thcir Effect on Chemical and phosphate IPcrtilizcr Solutions.” Januiiiy 16, 1976. Physical Properties of Suspension Fertilizers,” TVA Bulletin 11. Hrushccr. J. C., and J. I:. Andcrsm. .I, Am CI!cm Soc., Vol. Y-183, 1983. 68. I) 902 (1946) (CD95). 43. Achorn, F. P., and H. L. Balay, “Quality Control in the Pro- 12. I:lott. R.. ti. Ilruoislholz, and 0. Blumcr, He/!,.Ciiem. Acra.. duction of Suspensions from Solid Ammonium Phosphates, VoI. 38. pp 3753-69 (1955). MAP, and DAP,” presented at the NFSA Convention, 13. Polosin. V. A,, and M. 1. Shakhparunov, J, i’hy~.Chcni., USSR. December 6-9, 1982, Atlanta, Georgia. Vd21,pp119-23(1947). 44. Jamcson, R. I:., J. Chem SOC., p 752 (1959). 14. Polosin. V. A,, :and A. ti. Ticahchov, Izvcst, Timirgazev, 45. Frazier, A. W., and E. F. Dillard, J. Ag Food Chem, Vol. 29, .Srl’rkokhoi. Akod., No. 2 (3). pp 203-20 (1953). No. 4, pp 695-98 (1981). IS. Lindsay. W. L., A. W. I’razier, and H. I;. Stcphenson, Soil 46. TVA 6th Demonstration, “New Developments in Fertilizer . Sci. Soc. Am I’mc., Vol. 26, No. 5. PP~~ 446-52 (1965). Technology,” 1966, p 67 (Direct Process), p 29 (Micro- 16. Iluhti, Anna-Lhra, and P. A. Gartaganis, Conadion J. Chem.. nutrients in Suspensions). Vol. 34. ..DD 785-97 (1956). 47. Siegel, M. R., J. M. Stinson, H. C. Mann, Ir., and R. G. Lee, 17. Joncs, T. M.. and J. G. Gctsingcr, Fcrrilizer Solutions, Vol. 23, 150th Meeting ACS, Atlantic City, New Jersey, September No. 1. pp 68. 70. 72. 74, 76, 78, 80. 82, 84, 86 (January- 1965. i:chruary 1979). 48. Dillard, E. I:., A. W. Fraricr, T. C. Woodis, and F. P. Achorn, 18. Mclinc, R. S.. K. G. LCC. and W. C. Scott, Fertilizer Solurion “Precipitate Impurities in 1846-0 Fertilizer Prepared from Mopmine (March 19723. Wet-Pmccss Phosphoric Acid,” TVA Bulletin Y-162. 19. I:rni~r. A. W., J. K. Lchr. and I:. I:. Ilillard, E‘nviron. Science 49. TVA Bulletin Y-12, “New Developmcnts in Fertilizer Tech- arid Tcch,ioloyy. Vd. 11, pp 1007-14 (1977). nology,” 8th Demonstration, 1970. (Suspcnsians from Granular 20. Slack, A. V., 1. M. Potts, and H. B. Shaffcr, J. An Food Chem., Materials.) Vol. 12. pp 154-57 (1964). 50. TVA Bullctin Y-81, “New Dcvclopments in Fertilizer Tcch- 21. Williaid. 1. W., li. I:. Dillard. and J. D. Hatficld, J, Chem Eng. nology,” 10th Demonstration. (Suspension from Granular Doro. Val. 27, pp 34-38 (1982) (X-540). Materials X Poly.) 22. Williard. J. W., E. I:. Dillard, and J. D. tlatficld,.l. ClIem. Enx. 51. Frazier, A. W., and R. G. Lee, Ferril1:erSolurions. Vol. 16, No. [hro. VoI. 26. ..OD 89-90 (1981) (X-524). 4, pp 3243 (1972). 23. Williard. J. W., and I:. I:. tXllnrd, J. Clzem. Dota, Vol. 20, 52. Frazier, A. W., J. R. Lchr, and J. P. Smith, J. Ag Food Chem.. pp 436-37 (1975) (X-277). Vol. 12, No. 3, pp 198-201 (1964).

47 53. Brown, E. H., J. R. Lchr. A. W. I'razicr, and J. P. Smith, J. AK FoodCIzem., Vol. 12, No. 1, pp 70-73 (1964). 54. Brown, E. H., J. K. Lchr, J. P. Smith, and A. W. 1;rzziw. J. Ag Food Clrwn. Vol. I I, pp 214-22 (1963). 55. I:mzicr, A. W., and E. I:. Dillnrd, J. Ap Food Clmr., Vul. 29. pp 160-62 (1981). 56. Taylor, A. W, k. W. I'razier, and E. L. Gumcy. li.aii.Yocrio,!sof rlie Forodny Society, Vol. 59, pp 487, 1580-84 (196;). 57. Frazier, A. W., J. R. Lehr, and J. P. Smith. Am. Mu., Vol. 48. pp63541 (1963). 58. Taylor, A. W., E. L.' Gurney, and A. W. I:iilzicr, Soil Sci. Soc. Am. Proc., Val. 29, Nu. 3, pp 317-20 (1965). 59. Fmzier, A. W., and A. W. Taylor, Soil Sei. SOC. Am Proc.. Vol. 29, No. 5,pp 545-47 (1965). 60. Lehr, 1. R., E. H. Brown, A. W. Fiazicr. 1. P. Smith. and R. I). Thrasher, "CryStdlOgraQh~C Properties of fcrtilizer Con- pounds," Chem. Eng. Bollctin No. 6. Tenncsrce V:cIIcy Authority, Muscle Shoals. Alnhnma, 1967. 61. Harernsn, 1. I?.,J. R. Lehr, and J. P. Smith, SoilSci, SOC.AIII. prOc.. v0i. 15, pp 76-84 (19~1). 62. Mortvedt, J. J.. Po M. Ciordano. and W. L. Lindsay. "Miciw nutricntr and Agriculture," Soil Sci, SOC. AH., Iw M;idisan. Wisconsin (1972).

48 / Chapter 5 Production, Handling, and Use of Anhydrous- Ammonia By G. M. Blouin. H. L. Kimbrough. and W. J. Sharratt

Nitrogen is y~,essential element in supplying the atmosphere by a complex biological process occurring in burgeoning world population with adequate food and fiber. certain legumes, elemental nitrogen must be converted in Fortunately, it is in abundant supply in the Earth’s large quantities to reactive chemical compounds that are atmosphere, but, unfortunately. is essentially useless in available to growing plants. This process is often referred to intensive agriculture because of its inherent inertness to as . Because of an equal abundance of most growing organisms. Aside from very small amounts water on this planet, the logical and most economical route available in naturally occurring minerals such as Chilean to fixed nitrogen is through the production of highly “caliche” (NaNO, and KNO,) and the large but still reactive hydrogen from water and its subsequent reaction completely inadequate amounts absorbed directly from the with atmospheric nitrogen to produce anhydrous ammonia.

PRODUCTION PROCESSES

HISTORY nitrogen and hydrogen, at Oppau, Germany; that plant soon reached a production rate of 30 tons per day. In 1917 Today anhydrous ammonia is a basic industrial chemical, a 100-ton-per-dayunit was built at Leuna(1). ranking in annual tonnage production with and During the painful development of the Haber-Bosch oxygen. However, reaching this status has required over 200 process in Germany, the fixation of atmospheric nitrogen years of painstaking development since its discovery in the took other directions as well. By 1905 Berkeland and Eyde laboratory by Priestly in 1754. It was some 30 years later had put into operation in their electric arc.process that the composition of this laboratory curiosity was estab- for direct oxidation of atmospheric nitrogen to form lished by Bertholet. In the meantime, atmospheric nitrogen . Also in 1906 the Frank-Car0 calcium cyanamide was identilied by Rutherford in 1795. However, it was not process, under development since 1898, was commer- until the period 1795 to 1823 that consideration was given cialized, first in Italy in 1906 and in the United States in by Hildebrand and Dobereiner to the possible synthesis by 1918. After World War I, these processes were rapidly the catalytic reaction of atmospheric nitrogen and discarded in view of the successful develop- hydrogen. Others continued this work-possibly given ment in Germany. impetus by the knowledge expounded by the work of Lord Thus the era 1920-30 saw the rapid increase in world Dundonaid in 1795 and that of Sir Humphrey Davy in the early 1800s that nitrogen-bearing materials were essential synthetic ammonia capacity. In the United States alone, to plant growth-but it remained to in the the capacity increased to over 380,000 tons per year by period 1902-09 and later to in the period 1932 and again, during World War 11, to about 1.4 million 1910-13 to develop the first commercial ammonia synthesis tons per year. This increased capacity, for the most part, process, using promoted iron oxide catalyst for reaction of was not due to increasingly larger plants-most were of a

49 200. to 350-ton-per-day capacity, or smaller-but to a 2. Shift conversion-catalytic reaction of carbon doubling in the number of plants (I). monoxide with steam; exothermic. The period 1950-70 saw an amazing series of develop- CO + H, 0 + H, + CO, t 9.8 kcal ments-first came new technology involving large single- stage converters and multistage centrifugal compressors- 3. Synthesis gas purification and then a rapid expansion of numbers of plants and plant a. Acid gas (CO, , H, S, COS, or H, S) removal- capacities. In 1950, U.S. capacity was about 1.7 million scrubbing of raw synthesis gas by preferential tons per year. By 1960 it had risen to about 5.5 million, basic chemical (promoted potassium carbonate, almost a fourfold increase (1, 4). Between 1960 and 1970, ethanolamines, etc.) absorbent solutions or capacity increased to 16.9 million tons, another threefold physical absorbent liquids (methanol or other increase. Current (1982) capacity is about 19.5 million tons organic liquids) in packed or plate towers. Major per year in 144 plants. (About half of these plants were energy requirements are in regeneration of closed or idle in mid-1983 because of the general recession sorbent solutions. or because of their low capacity and efficiency.) b. Methanation-catalytic reaction of residual carbon oxides from acid gas rernoval with hydrogen to form methane (COX are synthesis gas poisons, CHEMISTRY methane is not); exothermic.

CO t 3H2 + CH, + H, 0 + 49.3 kcal The ultimate exothermic, catalytic reaction involved in CO, + 4H2 + CH, 2H, 0 + 39.3 kcal the formation of anhydrous ammonia (NH,) is: + c. Liquid nitrogen wash (in lieu of methanation)- N,(g) + 3H2(g)+ 2NH,(I) + 32 kcal residual COX and CH, are scrubbed out of This seemingly simple reaction is, however, only the final synthesis with liquid nitrogen; generally used step of four major steps involved in the production of more only with partial oxidation processes that require than 95% of the tremendous annual tonnage of anhydrous oxygen; also provides nitrogen required for ammonia throughout the world. Only where relatively pure synthesis. hydrogen is available (certain electrolytic processes, 4. Ammonia synthesis-catalytic reaction of hydrogen including water electrolysis, or petroleum refining processes) and nitrogen; exothermic. can the single process (reaction) above be utilized. N,(g)+3H,(g)+2NH3(1)+32 kcal Four basic unit operations are required in the production of nearly all anhydrous ammonia processes (I). These steps, It must be noted that the above reactions are the together with the chemical reactions involved, are described preferred reactions in each step but that in the raw briefly as follows: synthesis gas production nuiiierous side reactions 1. Raw synthesis gas (hydrogen + carbon monoxide) occur arid niust be suppressed by the optimum adjust- production by: ment of pressure, temperature, air, or oxygen: feed- a. Reforming hydrocarbons with steam; catalytic; stock ratio, space velocities, catalyst composition and endothermic; air added in secondary reformer to age, stean1:carbon (feed) ratio, feedstock H:C ratios, burn out residual methane and to provide the etc. Details of this optimization procedure are required nitrogen for synthesis. beyond the scope of this text and are covered thor- (1) Methane oughly in the literature (2, 4, 6). Computerized operation of ammonia plants is of significant benefit CH, + H,O 3H2 + CO -49.3 kcal because of the high degree of interaction of these (2) Naphtha many variables. One aspect of the chemistry of ainmonia synthe- % 2C7HIs +14H20314CO+ 29H2 -520 kcal sis, the thermodynamic (thermal) efficiency of the various processes, is not widely discussed in the b. Partial oxidation of coal or residual oils in literature. However, Hignett (4) provides basic presence of steam; endothermic (emperical thermodynamic data that permits calculation of fo:mula of a typical coal macromolecule; sio, process thermal efficiency. He developed the overall equivalent to ash content). theoretical reaction in the methane steam reforming process as being: (1) CSOH,o 010 S(Si02)+40H20~ 70 Hz 49 CO +COS +Si0,-2400 kcal + 7 CHq(g) + 10 H,O(I) + 8N2 (9) + 202 (g) + (approximate) 16”,(1) + 7Co,(g)

50 Using standard heats of formation. the reaction is Liquified natural gas (LNG) is a growing world com- found to be exothermic to tlie extent of 1.6 x IO6 modity and, in spite of the huge investments required both Btu/ton NH, (liquid). However, the high heating in the exporting and importing countries, is beginning to vdue (HHV) of the 7 moles of CH, given in the displace naphtha as indicated below. equation is equivalent to 1.489 kcal, or 93.1 kcal per Naphtha-Naphtha (emperical formula C,H, 5) is the gram mole NH,. or 19.7 million Btu per ton liquid most widely used feedstock in those countries that have no NH,. The HHV usage here is believed to be justified indigenous natural gas. However, in recent years, the world in that the efficient alternate use of the gas as a fuel price of naphtha has increased to’ the point where LNG is would. by preheat units, recover the latent heat uf beginning to displace naphtha as a feedstock in existing vaporization of the water formed in combustion reforming plants. It is estimated that about 15%-20% of reaction. This serves as the theoretical base for deter- the world ammonia capacity is based on naphtha (4). mining the thermal efficiency of all ammonia synthe- Naphtha is a straight-run, light petroleum distillate sis processes, that is fraction with a final boiling point of about 265°F (130°C); it is primarily a saturated hydrocarbon (95% by weight). Thermal efficiency, % = Unlike natural gas, which is desulfurized at the production 19.7 2: IO6 field, naphtha may contain as much as 0.15% sulfur (I). 1oo(106 Btu process energy input 1 As a result, it must be desulfurized at the ammonia plant. The theoretical energy content, 19.7 x IO6- On the other hand, like methane, it can be steam-reformed Btu-per-ton liquid NH,. is verified in an approximate to produce the required hydrogen for NH, synthesis. In manner by calculating the heat of reaction (combus- fact, the two feedstocks are interchangeable provided the tion) of liquid NH, with O2 : specially promoted iron oxide synthesis catalyst required for naphtha is used. ?”,(I) t40,(g)+N,05(g)+3H,0(l)i Naphtha plant investment costs are about 10% higher ,172.4 kcal than comparable natural gas plants because of the desulfur- Using standard heats of formation, the heat of izing unit and the somewhat larger acid gas removal system reaction of this equation is 172.4 kcal, or 18.3 x IO6 required for the former. Btu per ton NH,. The higher theoretical value (19.7 x Coke-Oven Gus-The production of coke by the reductive IO6 Btu per ton) is normally used in estimating pyrolysis of coal (coking) yields by-product offgas that process efficiency (I). contains hydrogen, carbon monoxide, methane, and nitrogen as wcll as numerous impurities (H2S, COS, CO,, tars, phenols), and some NH,. The gas can be used in the reform- RAW MATERIALS FOR ing process after purification by acid gas removal or cryo- HYDROGEN PRODUCTION genic separation and liquid nitrogen wash (1). This feed- stock is economically feasible when used in a large metal. Virtually any hydrocarbon or carbonaceous material can lurgical operation where tonnage oxygen is required and be converted to hydrogen (plus CO,) by reaction at high thus liquid N, is also available. Its major drawback is high- temperature with water (steam). Water itself may be con- compression energy requirements since the feedstock is sidered as a “feedstock” through the electrolytic process. normally at atmospheric pressure. Its use is dwindling Methane-Natural gas (methane) is, however, by far the because of the competition from large natural gas-based most economical and widely used feedstock in the world plants. (1. 4). About 75% of the world capacity and about 95% Heavy Oil-Raw petroleum, heavy distillates, and residual of the US. capacity is based on the catalytic reforming of fuel oils (Bunker C) are widely available, but have been methane with steam. Only those countries having no used only to a limited extent-perhaps 5% to 10%of world indigenous gas go to other feedStocks. It is very nearly capacity. The primary reason for its past limited use (future the ideal feedstock in that it provides both the carbon for use may tend to increase as indigenous gas reserves dwindle reduction of the hydrogen ions in water to elemental in some areas) is the high plant capital investment required hydrogen and an equal amount of hydrogen by the simul- for the associated air liquifaction plant, the large feed. taneous oxidation of the carbon in the methane molecule. The overall reaction, in its simplest form, is: stock storage facility, and the much larger acid gas removal system for a given capacity (1,4). The future of the partial CH4 t ?H, 0 --f 4H, t CO, oxidation process required by these heavy feedstocks will In addition, the simple molecular structure requires less depend upon the future complex relationships of natural energy to drive the endothermic reaction to the right than versus synthetic pipeline gas (SNG), petroleum refining any other readily available feedstock. demands (gasoline versus naphtha or other petroleum

51 --

products), and the degree of concentration of world market PROCESS AND DESIGN DETAILS supplies from the "gas-rich" areas such as the Persian Gulf, Indonesia, and the North Sea-as well as on the future Feedstock Purification results of the vastly increased gasloil exploration in the United States, Canada, Mexico, etc. 1. Natural gas reforming-Natural gas from the well-head Coal-Partial oxidation of coal is a technically viable invariably contains unacceptable levels of sulfur compounds, but expensive process for hydrogen (ammonia) production. generally H2S, and carbon oxides, both of which are It is even more costly-investment and operating-than the poisons to tlie nickel oxide reformer catalyst. In the United oil partial oxidation process because of the still larger air States, these impurities are nearly always reduced to a few liquifaction and gas purification systems. The basic advan- tenths of a part per million before introduction to the tage is to those countries with large indigenous coal reserves, pipeline system. In other area?, this must be done at the among which is the United States. ammonia plant site. Since virtually all of the acid-gas There are only a small number of coal-based plants in removal systems can be designed to be equally effective for the world-Turkey, South Africa, and India, to name the H2S and CO,, descriptions of these processes will be most prominent. However, the ammonia plants at covered under the Acid Gas Removal section. Modderfontein and Sasolburg, South Africa, are large In some areas, pipeline gas may contain a few parts per units and, at Sasol, are part of a large complex wherein million of sulfur compounds. In this case, a sulfur guard modified Fischer-Tropsch synthesis processes yield syn- may be placed in the ammonia plant gas inlet line. Either thetic petroleum products of a wide variety. Their opera- alternating beds of activated carbon at ambient temperature tion has been quite successful after an initial period of or zinc oxide beds at about 600'F (316"C)(gas preheated) trials and unit modifications. This approach_Le., the inclu- are effective in reducing the sulfur to undetectable levels sion of an ammonia plant in a very large scale coal-based (1). chemical complex, appears to offer the best chance for 2. Naphtha reforming-Nearly all couimercial naphthas economic viability in the near term. contain significant amounts-up lo 1,500 ppn-of sulfur ils The current or first-generation plants operate at low delivered. Thus the ammonia plant must include naphtha pressure in the gasification section. As a result, compression desulfurizer upstream of the primary reformer. The costs are high (1). TVA, at its National Fertilizer Develop- naphtha is vaporized and mixed with a sufficient amuunt of ment Center, has undertaken on a relatively small scale hydrogen from the synthesis section to ensure reduction of (130 tons per day) the demonstration of the integration of all sulfur compounds to H,S in a catalytic converter. The a second-generation coal-gasification and gas purification H,S is removed by a combination of steam sti~ippingof tlie unit (figure 14, left center) with its existing gas reforming recondensed naphtha and iron or zinc oxide sulfur guards ammonia plant (figure 14, right center) (6). The purpose is for the revaporized naphtha (I, 4). to demonstrate the high-pressure (500 Ib/in2g, 3.4 x IO6 3. Most other hydrogen production for ammonia syn- Pa) Texaco gasification process and the means of retrofitting tliesis iiivolves noncatalytic partial oxidation of the feed- a coal gasification system compatible with existing down- stock with relatively high-purity oxygen from an air separa- stream gas-reforming plant purification and synthesis units. tion plant that also supplies the high-purity nitrogen fur Successful demonstration could lead to the savings of hun- syngas makeup and for the liquid nitrogen syngas wadi in dreds of millions of dollars in the event that depleted gas the final Purification step. reserves or exorbitant gas costs force a change to plentiful Under these conditions, the feedstock cannot be purified; coal reserves in this country as an alternative petrochemical therefore, the feedstock impurities such as ash, sulfur, and feedstock. The TVA unit, which is in the early stages of chlorine compounds, and volatile. condensable tars. hydro- operation (figure 15), has completed the 5-day performance carbons, phenols, etc., are present in the raw gasifier offgas test and has produced ammonia at.full design rates. and must be removed from this stream prior to its under- Other Feedstocks-Among other alternative feedstocks- going shift conversion. This purification process is described all requiring the partial oxidation process-are coke and in more detail later in this section. waste cellulases such as paper and wood dust. Their use is Primary Reforming extremely limited; they are mentioned only to emphasize the wide range of potential feedstocks for ammonia syn- As previously indicated. the most widely used primary thesis. reforming reaction is that of methane with steam: Water, of course, may be disassociated into hydrogen and oxygen, in separate streams, by electrolysis. A few small CH,+H,0z3H2 tC0 plants in remote areas produce ammonia from hydrogen The reaction is endothermic. thus rcquiring a large amount obtained in this manner. of heat energy to drive the reaction to the right. Addition-

52 3 ._M U

53 I

54 ally, a catalyst is required to increase the late of reaction tu refiirming in that no pretreatment of the feedstock is an acceptable level. Further. the typic;tl reformer, and sub- required and no sensitive catalysts are involved. On the sequent unit operations prior to syiigas compression. is other hand, it is a much more complex process mechani- operated at abuut 500 lb/inZg (3.4 x IO6 Pa) by coni- cally and much more costly because of the necessity of pressing the meth;ine. as necessxy, prior tn refnrniing and liquid or solid feedstock storage and handling; of the need utilizing high-pressure steam generated by waste heat for an air separation plant; and of significant problems in recovery. The moderately high pressure favors CO shift gaseous. liquid, and solid waste disposal (1,2,4). conversioii kinetics and acid gns reiiiiiyal. iiot reforniing. Extreme detail in process description is covered in the but ii requires less compressivti energy to compress tlie literature (1, 2, 4, 6) and is beyond the scope of this text. inethane (and secondary ieformcr air) than the larger Briefly, the endothermic conversion of steam to hydrogen volutiic of raw reformer ciffgiis (2). Tlie primary reformer by reaction with carbon (coal) or heavy hydrocarbon catalyst gencldly consists of small (5/8 x 3/X x 1/4 itich, involves the respective reactions: 1.6 x 1 x 0.6 cm) rings of alpha aluiiiina base impregn;ltcd C+H,O~H,+co with activated nickel: the catalyst is packed iii nickel alloy tohes (gas side) tliat ;ire extern;tlly heated (fire side) by gas and flames in tlie reformer furnace. Catalyst temperature is generally about 1473’F (800OC). An excess of steatii- about 3.5 mils per tmo1 CH,-versus tlie theoretical 1 mol- Thus, as in steam reforming, the principal products impor- is used to coiitrul tube temperiiture and to prevent carbon tant to ammonia production are H, and CO, the latter (soot) formation due to incomplete metlime cracking (2). being convertible to additional H, by the shift conversion The primary reforming of vapcirized naphtha is quite process. similar to th;tt of methane. The basic difierence is in the The significant differences between the catalytic catalyst empluyed. Whilc still an activated nickel-un- reforming process and the noncatalytic partial oxidation aluniiiia ring, tlie base is “promoted“ with potash. Kalsilite process are (1) the endothermic heat of reaction of the (K,O-AL,O,.?SiO,) is a typical base, or support. for the carbon or hydrocarbon and steam is supplied by complete nickel catalyst. A somewhat higher steam:carbon ratio combustion of part of the feedstock with oxygen within (4.5: I) is generally required to guard against carbon the gasifier rather than by indirect firing in the catalytic depositiiin (I., 2). reformer, (2) the required nitrogen (and combustion oxygen) is supplied by a costly air separation plant rather Secondary Reforming than directly from the atmosphere as in the seconhry reformer, and (3) feedstock purification prior to gasifi- Tlie gas leaving tlie natural gas or naphtha primary cation (reforming) is not required as in natural gas or reformer contains unreformed methane (methane is an naphtha reforming. However, the raw gas from the gasifier intermediate in naphtha reforiiiing). The same nickel does require extensive cleanup, namely, venturi and/or catalyst (somewhat lower iiickel content) is used, but it packed tower scrubbing and electrostatic precipitation for is deposited in a packed tower bed rather than in tubes; solids removal, condensation and separation of condensable potash addition is not required. Compressed, preheated air vapors, and absorptive removal of sulfur compounds and is added in sufficient quaiitity at the gas inlet to provide the some carbon dioxide. both before and after shift conversion. final I :3 N2 :H, tnd ratio in the syngas. A relatively small aniiiunt of additional high-pressure steam is added to carry Carbon Monoxide (Shift) Conversion out soiiie additioriiil methane reforming. The oxygen in the added air reacts, vicourse, with all nf the constituents of The formation of carbon monoxide (CO) in the first the gas so that a net exothermic series of reactions occur. stage of either the steam reforming or partial oxidation Thus the reformer temperature rises to about 1800’F processes is unavoidable. However, it is relatively simple to (982°C) (I, 2). convert the carbon monoxide to additional hydrogcn by the catalytic reaction: Partial Oxidation (Gasification) CO+H,O-tH, +CO,

In lieu of steam reforming of light hydrocarbons, the This IS generally carried out in two stages. The first is a partial oxidation (if heavy hydrocarbons or coal is generally “high temperature,” packed tower catalytic converter con- the source [if hydrogen arid carbon monoxide for ammonia taining chromium promoted iron oxide catalyst tablets and synthesis. This general process was described briefly in the operating at about 780°F (416’C). The second unit (“low Raw Materials for Hydrogen Production section. On one temperature”) is similar except the catalyst is a zinc copper hand, it is a simpler and more rugged process than steam oxide catalyst, sometimes promoted with aluminum oxide,

55 operating at about 500°F (26OOC). The exit gas from the Basically, the impurities in the crude synthesis gas are first stage is cooled somewhat by heat-exchange prior to removed by scrubbing the entire volume ufgas, after it has entry into the second stage (1,2). been cooled by heat exchange to less than 150'F (6G"C). The first-stage catalyst is reasonably tolerant to sulfur; as It is scrubbed with aqueous solutions ofchemical absorbents much as 200 ppni H,S can be tolerated. The low tempera- sucli as nionoetlianolamine or potassium carbonate. or with ture catalyst, however, is very sensitive to sulfur; so often a certain organic liquids such as methanol and tlie diniethyl zinc oxide-sulfur guard may be placed upstream of the unit. ether of polyethylene glycol. lo physically absorb the acid Recently a shift conversion catalyst was developed gases. leaving tlie syngas nearly free (if the impurities. Little specifically for use in partial oxidation processes where hydrogen or nitrogen is absorbed by these media. The feedstock sulfur compounds present certain problems. It scrubbing may be accomplished in either packed or plate is an intermediate temperalure, "sulfur-activated" cobalt- towers; the saturated absorbents are regenerated (desorbed) molybdenum catalyst that is completely toleratit of not by heating andlor flashing tlie scrub loop to lower pressure. only sulfur, but also of other likely impurities in the gasifier In partial oxidation processes, the scrubbing is generally raw gas stream. It is operable over a range of 400' to 800'F dune in two stages, one prior to shift conversion. primarily (204O-427"C) (4, 6). Its primary advantage is described for sulfur removal (but including some CO,), and the other below. stage after shift conversion, primarily to recover relatively pore carbon dioxide for urea synthesis. If tlie sulfur- Acid Gas Removal activated shift catalyst is used, the two stages of acid gas removal may follow shift conversion. but two stages are The natural gas and naphtha reforming processes yield, still required. The first is operated under conditions (if after the carbon monoxide shift conversion step, a higli- temperature and pressure that preferentially remove temperature gaseous mixture consisting of Iiydrogen, virtually ill1 the sulfur as HzS. ahing with some CO,, wliile nitrogen, water vapor, carbon dioxide, residual methane, the second removes virtually iill the remaining C02, agzin carbon monoxide, and certain inerts, principally argon. to be recovered fur urea synthesis. from the air fed at the secondary reformer. In tlie gas arid ~i:iplitlia refiirming processes. it is seldoiii The partial oxidation process yields, after particulate that iiiure than one stage of xcid gas (CO,) removal is removal by water scrubbing and possibly electrostatic required since no other impurities cap:ible of being removed precipitation, but before shift conversion, a higli-tenipera- by these systems are noriii:illy present. ture gaseous mixture of hydrogen, carbon monoxide, The physial absorbenl liquids have hecotne tlie diiiiii- carbon dioxide, methane, hydrogen sulfide, carbonyl nant type in both gas reforming and partial uxidation proc- sulfide, condensable organic vapors, ammonia, and water esses in recent years because (1) they are considerably vapor. The methane and condensable vapors are generally' more flexible. i.e.. preferential absorption of H,S in the present only in a relatively low temperature (<1300°F, first stage before or afler shift c~~iiversioii.leaving pure C02 704OC) coal gasification process such as the Lurgi. Most adequate for urea synthesis ftir second-stage recovery and processes operate at high temperature PI80O0F, 982OC) (2) significantly less energy is required for regenerating the to avoid formation of these impurities. absorbent (I, 2.4,6). It is absolutely necessary that in either process, the sulfur compounds, the carbon oxides, and the condensable Final Synthesis Gas Purification organics be reduced to a few parts per million in the 3:l mol ratio hydrogenmitrogen synthesis gas going to the In spite of the often complex and generally effective catalytic ammonia synthesis converter (Ammonia Synthesis processes for removal of syngas impurities (preceding Conversion section). Methane and argon are not poisons to, section), residual impurities always remain in the syiigas. the synthesis catalyst, but do cause pwge losses that are These must he removed or transformed to compounds not necessary to prevent buildup in the synthesis loop. poisonous to the ammonia synthesis catalyst. In the partial oxidation process, it is necessary that at In the gas reforming process. the only residual poisons least the sulfur compounds be removed before shift con- are the carbon oxides. In the m;ijority of cases. these are version unless the new "sulfur-activated" shift catalyst is transformed to methane in a nietlianatiun catdytic used (Carbon Monoxide Conversion section), in which converter. The cold syngas is reheated by heat exchanges to case the acid gas removal step can follow shift conversion. about 500°F (760'C) and passed through a packed tower It is beyond the scope of this text to cover the innumer- containing an alumina-supported nickel oxide catalyst. able interactions of 'the many variables involved in the Both the carboilmonoxide and -dioxide react with the hydrogen present to form methane and water: numerous commercially availahle prbcesses. Only the basic information is given here. The literature abounds in detailed C0+3H,-'CH4tH,0

descriptions of the individual processes (1,2,4,6). C02 + 4H2 --f CH, t ?H2 0

56 The procedure is simple and effective; however, its disad- Following compression, the syngas is fed to the single vantage is apparent. Valuable hydrogen is consumed and catalytic converter where the reaction is carried out at purging requirements in the synthesis loop are increased. about 800°F (427OC) (1, 2). Although converter design Nevertheless, this is the most economically viable mode of varies considerably, the potassium-alumina promoted iron operation for the gas reforming process (1). oxide catalyst is generally arranged in several baskets in the Partial oxidation processes, which require both oxygen vertical converter and the feed gas is divided into several and high-purity nitrogen for syngas makeup, have the streams, one to the top and the remainder to the spaces advantage in at least final syngas purification. The cold gas between catalyst baskets. This is to maintain a relatively from the acid-gas removal system is simply scrubbed in a even temperature profile along the height of the converter plate tower with liquid nitrogen at such temperature that (2). part of the liquid nitrogen boils uff at the proper rate to The catalytic reaction of the hydrogen and nitrogen yield the desired H, :N, syngas mol ratio while the remain- takes place in the converter, but does not approach comple- ing liquid thoroughly reniovcs essentially all impurities tion. In a compromise between chemical kinetics, thermo- including water vapor. Little or no purging is required in dynamic equilibrium, and metallurgical and design consid- the aiiiiiionia synthesis loop following nitrogen wash (2). erations, the conversion per pass is only about 20%-255%. However, the procedure is not economically viable for gas Obviously, the unconverted syngas must be recycled by reforming plants, unless, fortuitously, an air separation means of the recirculating compressor (which may be one plant is required for oxygen in another part of an industrial or two stages in the main compressor, or may be driven by complex. the main compressor drive turbine or motor) after separa- tion by water cooling or refrigerative cooling and conden- Ammonia Synthesis Conversion sation of the ammonia that is formed in each pass. Thus the volume of recirculated gas is about 4 to 5 times that of There is little doubt that the greatest research and the fresh makeup syngas feed, the latter being equivalent to development effort toward establishing a viable synthetic the design production rate of the plant. The recirculation ammonia process was expended on the synthesis step itself; compressor is designed to compress the large volume of Obviously, had this not been the case, beginning with the recycle gas only a few hundred pounds pressure greater more than IO years and 20,000 tests of Haber and Bosch, than its inlet pressure to overcome the pressure drop in the synthetic ammonia industry may never have risen to its the synthesis section. present status. The presence of the inert impurities (methane, argon, It is beyond the scope of this text to explore the vast perhaps helium) in the syngas, mentioned previously in amount of data developed over the years that deal with the the acid gas and synthesis sections, present a problem in complex thermodynamics and chemical kinetics of the the conversion loop. Since they cannot be condensed as is apparently simple catalytic, exothermic reaction. the ammonia product, they build up in concentration in the loop to the point where they interfere with the syn- N,(g)t3H,(g)’2NH,(1)+32.1 kcal thesis reaction. To avoid this, a continuous stream of the Needless to say, many hundreds of man-years and hundreds recycled syngas is “purged” from the recycle loop at a of millions of dollars have gone into the optimization of the point after separation of the product ammonia. This continuous, single-train process with capacities of up to “purge-gas” flow is generally adjusted to mantain a concen- 1,500 tons per day. tration of 10%-12% inerts in the loop. This may amount Basically, the synthesis loop consists of the main com- to about 0.5 volume percent of the total flow to the pressor section, the catalytic converter, the recirculating converter, or about 2 volume percent of the makeup loop condenser, and the recirculating compressor. The feed (2). The purge gas, consisting of methane, hydrogen, purified synthesis gas pressure is raised in the main com- nitrogen, argon, and ammonia may be used direcjly as a pressor from the 400-500 Ibfin’g (2.8-3.4 x IO6 Pa)in the fuel in the reformer or may be separated into its‘compo- syngas preparation section to the “optimum” synthesis nents by cryogenic and/or molecular sieve procedures for pressure of about 2,200-3,500 Ib/in2g (15.2-24.1 x IO6 Pa), recycle to the process. As previously mentioned, in partial oxidation processes, the availability of liquid nitrogen wash or in some cases up to 5,000 Ib/in2g (34.5 x lo6 Pa). The generally makes deliberate purging unnecessary in that main compressor, generally a multistage, electrically driven, what little impurities do bleed through the nitrogen wash reciprocating piston type in plants of up to 500 ton-per-day are dissolved and removed in the ammonia product (I). capacity, or a multistage centrifugal compressor in larger In the past, the liquid product was stored at autogenous plants, is equipped with interstage water-cooled heat pressure and ambient temperature (i.e., 165 Ib/in2g and exchangers to maintain the gas temperature within the 90°F or 1.1 x IO6 Pa and 32OC) in pressure storage vessels. design limitations. Current practice is to use ammonia in a refrigeration unit to chill the liquid product to about -28'F (-33OC), where capital investment of those plants for energy recovery its vapor pressure is essentially zero pounds per square inch systems. gage, for storage at atmospheric pressure. Continuous bleed- off, recompression, and condensation of a small amount of Recent Developments in Technology the tank's vapor phase is required to ensure the low temperature storage stability (4). The drastically increasing costs of the various conven- tional feedstucks, naphtha, fuel oil, and natural gas, have Energy Requirements spurred a vast amount of activity aimed at reducing the overall unit consumption of feedstock (energy) in ammonia Ammonia synthesis from hydrogen and nitrogen is an synthesis. This activity has had two directions: (1) signifi- energy intensive process; however, most of the energy cant reduction in unit feedstock requirements, primarily consumed is in the gas preparation (hydrogen) section. for natural gas, and (2) the development of new partial As indicated in the Chemistry section, the theoretical oxidation technology for gasification of coal, an alternate energy required in the basic methane steani reforming abundant, lower cost feedstock in the United States and process is 19.7 x lo6 Btu per ton of synthetic liquid many other countries. Since either approach would require ammonia. This serves as a basis for determining the thermal several volumes if detailed, only the basic technological efficiencies of all ammonia (hydrogen) processes. Conipara- principles will be listed here (6, 7,9, 10, 11). tive energy requirements for processes utilizing hydrogen 1. Potential improvements in gas/naphtha reforming produced from the various feedstocks are given in table 13. technology. It should be recognized that the above data are approxi- a. Combustion air preheat in reformer-common mate and, in the cases of the hydrocarbon and coal proc- only in large plants erected in last few years. essing, are relative to the present industry average. Many Retrofit will become common. Estimated savings, of the innovations (recent developments section) that 0.7 x IO6 Btu per ton. permit significant reductions in energy consumption in gas b. Process air and gas feed preheat-again, in use only and naphtha plants could be applied to the partial oxida- in new plants. Estimated savings, 0.6 x IO6 Btu tion processes as well, thus also increasing their thermal per ton. efficiencies appreciably. C Generation of superheated steam by heat exchange. It should also be recognized that the significant increases Estimated savings, 1 .Ox lo6 Btu per ton. in thermal efficiency of the gas reforming plants indicated d Use of physical absorbents in acid gas removal above can be achieved only with a significant increase in system; now in use only innewer plants. Estimated savings, about 2 to 4 x lo6 Btu per ton. e Fine-tune existing acid gas removal systems, using Table 13. Comparative energy requirements latest operating indices devised by developers. for processes using hydrogen (2,4, Estimated savings, about 2 x lo6 Btu per ton. 7.8)~- Energy f Substitute scrubbing for final syngas purification consumption Thermal to minimize methanation losses, or selectively Process (feedstock) Btulton x lo6 efficiency, % oxidize CO to CO1 with a bleed stream of air from Natural gasa 41.2 48 main air compressor. Estimated savings, 0.8 x lo6 Natural gasb 29.0 68 Btu per ton. Natural gasC 26.0 76 g Substitute improved centiifugal compressors for Naphthad 43.2 46 reciprocating compressors in 300-500 ton-per-day Fuel oild 45.9 43 units. Possible savings, 1.0 x IO6 Btu per ton. Cod h Substitute gas turbines for steam turbines as prime Existing low pressure 45.0 44 drivers. Estimated savings, 1.0 x lo6 Btu per ton. Future high pressure 40.5 49 i. Recover values (H, and CH,) from purge gas by Electrolysis of water cryogenic, semipermeable membrane, or absorbent (plus air separation plant) systems. Estimated savings, 0.5 x lo6 Btu per ton. Existinge 97.0 20 j. Redesign of synthesis converter to reduce pressure Potentiale 38.0 52 drop and increase level of indirect steam genera- aU.S. industry average, The Fertilizer Institute, 1979. tion in converter. Estimated savings, 1.6 x lo6 Btu bNew Canadian plant, Pullman-Kellagg design, 1981. per ton. 'Potential third generation design. dRelative to footnote a, industry averagc. In addition to these individual unit process improve- eBased on practical 10,000 Btu per kWh. ments, there are several somewhat radical departures

58 from conventional practices that can lead to signifi- the very large volume of departiculized raw synthesis cant energy savings. In one for which data are avail- gas as practiced in present technology. The trade-off able. a portion of the fuel gas is used to power gas for high-pressure operation is in the mechanical diffi- turbine-driven conipressors, with turbine waste heat culties involved in injecting the coal feed and dis- being used to preheat the gas feedstock and conibus- charging the residual ash or slag. Generally this is tion and proccss air. The acid gas removal system is accomplished through time-cycle feed and discharge operated at up to 1,500 Ih/in2g (10.3 x IO6 Pa) lock hoppers. In at least one process, however, coal (vs conventional 500 Ib/in2g, 3.4 x !06 Pa) which feed is in the form of a heavy aqueous slurry that is results in very high purity syngas, low purge losses, injected by high-pressure reciprocating pumps. The and low acid gas sorbent regeneration energy. The TVA gasifier is of this type and one tToubleSome process is reported to consume only 29 x IO6 Btu per lock-hopper operation is eliminated (6). to11 of ammonia and to have a lower capital invest- Gasification temperature has long been the subject ment than conventional plants (9). of intense study. Low raw gas temperature (!,ZOOOF, Recently, an engineering design firm released some 649OC) requires less oxygen per unit of syngas, but details of a design that incorporates many of the unit produces a considerable amount of by-products iiprrations changes listed above and for which is (phenols, ammonia, methane, oils, and condensable claimed an energy conson~plionof only 26 million organic vapors) that most syngas producers, particu- Btu per ton (IO). larly ammonia producers, prefer to avoid. As a result, 2. Potential improvements in partial oxidation tecli- nearly all second generation processes (as well as nology. many first generation processes) operate at high gas It is obvious that some of the potential improve. temperatures (1 ,800°-2,6000F, 982°-i,4270C) so that nients listed above for the hydrocarbon reforming only the principle syngas constituents, H, and CO processes might be incorporated in the unit operations (after acid gas removal), are produced (6). All other downstream of the gasification unit in the partial gases or vapors are cracked at this temperature. The oxidation process. Many have been in existing partial trade-off is the somewhat higher oxygen requirement oxidation processes, both with residual fuel oil and and the more severe effects on the gasifier components. coal. Another area of intense study has been in the Out of the tremendous atliount of literature that determination of the effects of coal ash-formation, has been generated concerning the production of syn- agglomeration and slagging characteristics on mechan- thesis gas (H, + CO) from coal, the single most ical “ash” discharge and gasifier bed (either fixed, important fact emerging is that most research and stirred, fluidized, or entrained) performance, on degree development effort seems to have been directed of coal conversion, and on overall thermal efficiencies toward medium- to high-pressure gasifier operation as (6). The effect on mechanical operation and therefore opposed to numerous existing processes operating at on plant reliability may be the overriding factor in near-atmospheric pressure (6). Although much of this gasification economics. While discussion of such a work is aimed at fuel usage (low to intermediate Btu complex subject is beyond the scope of this text, it gas or pipeline quality gas by methanation), much can be said that out of the many being studied, those consideration has been given to ammonia production few high-pressure gasification-process designs that can as well. solve-and, in time will solve-the problem of mechan- In the latter case, coal gasification even at minimum ical reliability will certainly bring the cost of coal- animonia synthesis pressures (2,000-2,500 Ib/in2 g, based ammonia down to meet the rising cost of gas- 13.8-17.2 x io6 Pa) has been considered and may based ammonia in the years ahead. It may be unfor- be attempted in second or third generation processes. tunate that the recent concurrent surplus of world Generally. however, most designs for ammonia pro- oil supplies and changes in national policies have duction are limited to about the same pressure range seriously reduced activities in the general field of of current reformer operation, i.e., about 500 Ib/in2g synthetic fuels from coal. It remains to be seen when (3.4 x IO6 Pa). In “synthetic” fuel processes, present operational reliability of some few designs can be mechanical and metallurgical considerations have established. TVA is continuing limited operation of limited most gasifier. development to about 1,000 the demonstration-scale Texaco entrained bed, water- Ib/in2 g (6.9 x !06 Pa) (6). quench gasifier system at MuscIe Shoals, and is stili The energy saving to be gained by gasifier operation / pursuing requests for funding of a much larger unit at elevated pressure is obvious. The energy of com- ,/ for methanol production. pression of only the sinall volume of combustion Some Department of Energy contractual work and oxygen is a fraction of the energy of compression of some private sector work still continues. Also, at least

59 one company, Eastman Chemical Products, recently price of naturiil gas in the United States reiiclies %S7 per constructed a dual Texaco gasifier-based chemical million Btu by 1985 becziuse of continued scarcity and/or complex in Kingsport, Tennessee (12). The unit has decontrol, it is probable that coal, which is likely to escalate a capacity of up to 1,800 tons of coal per day. The far less rapidly. will become an economically viable alter. plant is designed to produce acetic anhydride from native feedstock, James (I I), however, predicts that natural captive CO and methanol and imported acetic acid. gas will have to reach the S7-SlO per inillion Btu range While NH, will not be pmdiiced, such a unit could hefirre the econimic vi:ihility of co:il-hased ;anirnrinia produce up to 1,400 tons per day. However, in actual becomes a re:tlity. On tlie other hand, if tlie second genera- operation one gasifier will be on standby. No doubt. tion gasifiers become less costly, more reliable. and more experience gained by the operation uf TVA Texaco efficient. as is believed by many. the break-even price of gas unit at Muscle Shoals was useful in the final design inay fall in the lower range. Residual fuel oil would appear and operation of the Eastman units. to be a likely candidate dso, but its cost is projected to increase :it least as rapidly as that of gas. so. again. coal emerges as tlie potential aninioiiia feedstock in tlie fore- AMMONIA PRODUCTION ECONOMICS seeable future. Retrofit of coal handling and gasification systems to existing reforming synthesis plants should be In view of the widely varying factors influencing given careful consideration, particularly for those located ammonia process economics, this discussion will he limited in coal mining areas or near navigable rivers whereby coal to estimated plant prices from new (1982) plants in an could be barged in cheaply. industrial area on the United States Gulf Coast for the gas- and oil-based plants and in proximity to a coal mining area for the coal-based plant (4, 6, 7, 11). The primary purpose WASTE DISPOSAL PROBLEMS of the data in table 14 is to give an estimate of the relative 1982 prices of ammonia as produced from the four basic The natural gas and n;iphtha steam reforming processes feedstocks-natural gas, naphtha, fuel oil, and coal (present present few problems in waste disposal. In fact. about the technology). It has been assumed that the necessary planning only requirement for efficient cle;inup is in the condensate and the construction were begun on time to start up the from the syngas cooler prior tu the acid gas removal step. plants in early 1982. The line item cost data in table 14 Small aniounts of aniiiionia are formed in the reformers are given in sufficient detail to permit the reader to substi- because of the elevated pressure at which they opente. tute different values that better suit.his own situation. This ammonia condenses. aiong with water. in tlie syngas The calculated prices have little relation to current cooler. It inay be removed by steam stripping to iilliiw United States or world market prices because of the recycle of the condensate to a waste-heat boiler. The depression of prices due to current overcapacity, the stripped aniinonia may be recovered in an aqua-ammonia general slump in the economy, and the fact that many product (1). larger, older plants or plants with existing low-cost gas Some leakage uccurs in all ammonia synthesis plants- contracts (or subsidized gas as in some foreign areas) can, leakage of syngas (CO, H2. N2), as well :IS amnionia. but at the present time, produce and sell at much lower prices adequate maintenance of compressor and valve packings than can a new United States Gulf Coast plant. Further, it can maintain this leakage at :icceptable envirunnientiil is naturally assumed in these estimates that the plant levels. owners have erected the plants on the predication that The partial oxidation processes, p:irticularly that [if coal. their prices will become competitive in the near term, as present formidable. but controllable, total envirimnieiital the fertilizer demand increases. pollution control problem-air, water, and land (solids). While the absolute accuracy of the estimated prices is A significant portion of the capital illvestment is required not great, especially for coal gasification, the relative prices to provide for tlie control of emissions of all types. Thc should be reasonably accurate. problems are complex and details are beyond tlie scope (if The display of two decimal places in the unit costs does this text. The following items are but brief descriptions not imply an accuracy of this order; it is merely reflective of the salient areas of pollution control (I, 6). of computer programming. The data show that naphtha, which has greater value as a feedstock for more expensive Coal Receiving and Preparation petroleum products such as gasoline, obviously is priced out of competition for ammonia production. This has proven to Fugitive dust and water runoff from handling stockpiles be true in many world locations where natural gas is not must be controlled. Dust can be controlled by sprinkling. available. This, then, leaves the partial oxidation processes Pile runoff water must be directed to holding pool, moni- as the alternatives. If, as predicted in many forecasts, the tored for contaminants, and discharged or directed to tlie

60 IIIII

61 wastewater treatment plant, as necessary. If dry grinding is the plant, excluding the condensables recovery system in a practiced, the entire pulverizer-dust transport system must “low” temperature gasificatiun train (6). be exhausted to fugitive dust recovery equipment for recycle. ANNUAL PRODUCTION OF ANHYDROUS AMMONIA Gasifier The fertilizer use of mhydroiis ammonia is discusscd in The gasifier produces solid, aqueous. and gaseous con- chapter I :ind Iatcr in this chapter. However. it should be taminants. The latter are removed from the syngas stream nientiuned that a significant prop

62 ANHYDROUS AMMONIA HANDLING

As indicated in [lie productiun and use sectioiis, con- 100 cars shipped directly from the producing plants to the tinued inipruvements in niiimoiiia pruductioli and transpur- consuiiiing area. Because of these excellent transportation tation facilities. such as aiiinionia pipeline systems, large facilities, it is possible to deliver anhydrous ammonia to regional atmospheric storage tanks. refrigerated barges and dealers for retail sales at a relatively low cust. sliips. and jumbo tank cars have made anhydrous ammonia the iiiost economical source uf nitrogen delivered lo the ANHYDROUS AMMONIA farm. RETAIL OPERATION There are now two long pipeline systems in the United States for transporting ammonia from the producing plants Usually most anhydrous ammonia retail dealers have to the major consuming area. The Mid American Pipeline pressure-type storage tanks varying in size from 12,000 Cnmpany (MAPCO) system extends from northern Texas gallons (45,425 I) to 50,000 gallons (189,270 I). Liquid into Kansas, Nebraska, Minnesuta. and terminates in Iowa. ammonia is transferred from a railroad tank car or a truck The Gulf'Central Pipeline pumps anltnoni;] from the Missis- transport tank into the dealer storage tank with an ammonia sippi Valley area. extending through Louisiana, Arkansas, compressor or a positive displacement pump. The ammonia Missouri, Iowa. and Nebraska, with a branch line into compressor is the must frequently used method. Transfer llliriuis and Indiana. The MAPCO pipeline is about 850 is accumplished by removing vapor from the vapor space in miles (1,367 km) iii length and has a capacity above 1,300 storage tank and pumping these vapors into the tank tons per day. The Gulf Central Pipeline is about 2,000 being unloaded. The removal of vapors from the storage miles (3,219 kni) long and can pump over 3,000 tons per tank reduces the temperature of the liquid ammonia at day. Anhydrous aniiiiunia is transpurted through the pipe- the liquid-vapor interface and, therefore, the pressure lines as a liquid. Usually the ainmunia going intu a pipeline exerted by the liquid ammonia is reduced. The cbmpressed is warmed to the approximate soil temperature surruunding ammonia vapors are heated by the compressor and the hot the pipeline buried 3 to 4 feet (I meter) below the surface. vapors are injected into the vapor space of the transporting Branch pipelines terminate at facilities that store anhy- tank. /The vapor compressor has produced a pressure drous ammonia in large atmospheric storage tanks with differential between the transporting tank and the storage c;ipacities ranging from 6,000 to 30,000 tons. Liquid tank which allows the liquid ammonia to be easily trans- ainniunia in the pipeline is under pressure, and when the ferred from the transport tank into the storage tank. iiniiiionia flows into ii flash tank, the pressure is reduced Reducing the vapor pressure of the ammonia vapor remain- to atinospheric. A portiun of the ammonia evaporates and ing in the transport tank to about 45 Ib/in2g (31 x IO4 Pa), the negative heat of evaporatiun reduces the temperature as discussed later, will result in a loss of only 0.5% by Sf the liquid ammunia to near -33°F (-36.I'C) before it is weight of the original ammonia load. pumped intu atmospheric refrigcrated storage at branch The vapor cornpressor or the liquid-ammonia pump used terminals. The ammonia vapor is liquified by compression for transfer should be labeled for ammonia service. This then water couled and returned into the flash tank. Similar assures that the pump working pressure and pump seals systems fur refrigerated ammunia storage are used on barges will be suitable. If a positive displacement pump is used for ;md on uceangoing ships (4). It is especially important that, ammonia transfer, the pumping system can produce a high with the advent of atmospheric ammonia storage, ammonia pressure and, therefore, the positive displacement pump be heated prior to loading into transport equipment to must be equipped with a pressure regulated bypass valve avuid thermal shock, which can weaken the nietal in the and a discharge pipe to return liquid ammonia into the tank transport tank. being emptied. One railway tank car equipment company is experi- Railroad tank cars are equipped with dip tubes for menting with a tank car fabricated from metal that will liquid withdrawal and one vapor return line. The connection resist thermal shock. Huwever, test results with this tank for fittings and the valves are located beneath the dome car are not yet available. In addition tu refrigerated storage cover. These tank cars require a vapor transfer pump .for tcrniinals, facilities are available along the pipeline route to unloading. Transport trucks with a bottom unloading load the warm ammonia intu trucks and railcars for valve can unload anhydrous ammonia using either the vapor shipment. transfer pump or a positive displacement pump. Some Must major nitrogen-consuming areas are reached by general information published in numerous articles concern- either pipeline or barge transportation. In addition, railcars ing unloading anhydrous ammonia are as follows: are readily available that have pressure tanks for containing 1. An adequate supply of clean water should be avail- 24 tu 73 tuns of anhydrous ammonia. It is a practice to able, either as an emergency eyewash and shower or have "unit train" shipments of ammonia with as many as an open tank that a person can be submerged in.

63 2. Brakes should be set at the unloading rack and Additional procedures for transport truck unloading wheel chocks used to assure nonmovement of the follow: transport tank. 1. Connect the liquid hose between the storage and 3. The hose should be inspected for defects and any transport tank and then the vapor hose connection. kink in the hose straightened. The hose subject to 2. Open the liquid and vapor valves on the storage tank ammonia transfer pressure should be designed for a and the block valve upstream from the pressure by- minimum working pressure of 350 Ih/in2g (241 x pass valve which is usually near, or built into, the lo4 Pa) and a minimum rupture pressure of 1,750 pump. (Note: This valve should always remain open 1b/in2g (1,206 x IO4 Pa). The hose assembly must unless the pump is being sewiced.) be capable of withstanding a test pressure of 500 3. Slowly open the liquid valve on the transport tank lb/in2g(345 x IO4 Pa). and read gauge pressureson the transport and storage 4. The operator should wear a gas mask equipped with tank. an ammonia canister, protective gloves, hoots, and a. If the transport tank pressure is higher, allow the snit made of rubber or other materials impervious to pressure to force the anhydrous anunonia intu anhydrous ammonia while making or breaking con- the storage tank. nections. The wearing of a gas mask should be con- b. When the pressures are nearly equalized, open sidered necessary, especially when working on top the vapor valve on the transport and start the of a railroad tank car. transfer pump. 5. The liquid ammonia valves should be opened slowly c. If the pressure is higher in the sturage tank than to check for leaks around the fittings, to assure that in- the transport tank, open the vapor valve on the bleed valves are closed, and to keep excess flow the transport and allow the pressure to equalize valves from closing. before the transfer pump is started. 6. The liquid valves should be opened completely and 4. An inline flow indicator will help determine when then the vapor line valve should he opened on the the transport is empty. Some operators look for a transport tank. flexing. of the liquid unloading hose. 7. The valve at the compressor should be set so that During the entire unloading procedure the operator vapors are pumped from storage into the transport should: tank. (Different procedures are followed if liquid 1. Continue to wear protective gear. The gas mask could pumps are used.) be exchanged for tight-fitting goggles and a face 8. The vapor valve on the storage tank should be shield. opened slowly and the system pressure allowed to 2. Never leave the transfer operation unattended. equalize and then the liquid valve opened slowly to 3. Never fill a tank to more than 85% of capacity. fdl the ammonia storage tank. 4. Always open the bleed valve~iorelieve the pressure 9. The compressor should be started and operated between the connection before disconnecting the until the liquid transfer is complete as indicated by hose. an inline flow indicator. 5. Always use hose and fittings designed and approved 10. To remove vapor from the tank, the liquid valves for ammonia and never tamper with pressure relief should3,be closed and the vapor valve position valves or other fittings. reversed so that vapor is pumped from the transport tank. The vapor pressure in the tank should be Loading Nurse Tanks reduced to about 45 lb/in2g(31 x lo4 Pa). Truck transports that haul anhydrous ammonia are The general safety precautions listed for rail tank cars usually equipped for liquid unloading through valves in the and transport trucks should be followed. On seeming level bottom of the tank. The transport is usually unloaded using ground, unchocked nurse tanks have rolled from loading apositivedisplacement pump located at the storage terminal areas. or a truck mounted pump powered by a take-off shaft. To fill a nurse tank using a compressor, the liquid hose The general unloading procedure is essentially the same from the storage tank should be connected tu the liquid fill as for unloading railcars. The terminals that handle anhy- valve and the vapor hose connected and the valves opened drous ammonia truck transports are usually adjacent to so that vapors are removed from the nurse tank and frequently traveled roadways and, for -the safety of the pumped into the storage tank. The compressor continually operator and road traffic, the ammonia mask and other reduces pressure in the nurse tank and the ammonia fills to protective equipment should be worn while making con- the proper level. nections and starting the ammonia flow so that corrective If a liquid pump isused, the pump forces liquid ammonia measures can be taken rapidly if connection leaks or hose into the nurse tank, and the vapors are vented through the rupture occurs during the initial transfer. vapor hose and returned into the storage tank.

64 Anhydrous ammonia should be transported, stored, and Institute’s “Operational Safety Manual for Anhydrous handled in accordance with Federal, State, and local regu- Ammonia,” and American National Standard Institute’s lations. For more detailed instructions see the “Agricultural “Safety Requirements for the Storage and Handling of Anhydrous Ammonia Operators Manual,” The Fertilizer Anhydrous Ammonia.”

USE OF ANHYDROUS AMMONIA IN AGRICULTURE

HISTORY AND EVOLUTION Laboratory to undertake experimental work on such a project. Dr. Waynick had established the Association The use of anhydrous animonia in agriculture began in Laboratory at Anaheim, California, to offer citrus growers California in 1932. In the SO years that have elapsed, a advice on irrigation, fertilization, and pest control. Dr. tremendous effort has been invested in bringing ammonia Rosenstein made the necessary arrangements with Dr. to its present level of acceptance. Despite this acceptance, Waynick in 1932 to do experimental work on the use of questions are frequently asked. “Why was anhydrous anhydrous ammonia in irrigation water (IS). ammonia cunsidered for use as a fertilizer when the popular The first experimental applications of anhydrous ammo- solid nitrogen compound. annnoniuin sulfate, was readily nia were made on citrus groves in southern California. available?” This question can be answered by briefly Although the experiments were designed to measure crop reviewing developments that demonstrated the feasibility response to the fertilizer, they were also planned so that of its use and the development 11f equipment and tech- data on many facets of the application procedure coul niques for application collected. This early work showed that the concen a io To gain an understanding of the development of anhy- of ammonia in water was a critical factor4 in a 01 In drous ammonia fur agricultural use it is necessary to look ammonia losses and having even distribution throughout beyond 1932 to the mid-1920s. At that time commercial the field. The length of irrigation run and the%pc-of so9 fertilizers were just beginning to be used in southern -were found to be critical factors. While furrow irrigation California and mechanical devices for their application was the normal procedure in citrus and row crop areas of were not well developed. In the citrus groves, particularly, California, there were many areas where flood or contour sodium and calcium nitrate or ammonium sulfate were irrigation was practiced, particularly in the deciduous being used at S or more pounds per tree and in most orchard areas of northern California, or on preplant irriga- cases were being spread by hand. This was a tedious and tions for grain and cotton. Studies with flood irrigation time-consuming operation. revealed that ammonia distribution with this method of In 1928 two brothers, Eugene and John Prizer (lS), irrigation was also quite uniform. Concurrent with studies concerned with this labor problem on large citrus groves, on distrib t’on of ammonia, work was also carried out on built an applicator that would dissolve soluble fertilizers epth of ammonia penetration into the soil immediately and introduced them into irrigation water. One of the prin- aroun the furrow. Numerous cross section studies were cipal products used at that time was ammonium sulfate emade of irrigation furrows where inch square samples were produced by reacting ammonia with sulfuric acid. The price drawn from the bottom of the furrow downward as much paid for the acid was extremely high, even during the as 12 in. (30.48 cni) and laterally to the center of the plant depression period. In considering ways to reduce overall beds. These early workers also studied the effectiveness of costs, Shell Chemical Company engineers proposed the various application times. Early experiments showed that direct use ofammonia gas to eliminate the need for the acid. the use of ammonia in sprinkler systems was unsatis- It was thought that it might be possible to introduce factory due to high ammonia losses. ammonia into the water of irrigated areas in California. These early studies by Drs. Rosenstein, Hoagland, and The water, being spread in the normal process of irrigation, Waynick and other members of the University of California would distribute the fertilizer uniformly over the field. staff provided the basic information on which the early The principal problem was thought to be one of proving agricultural anhydrous ammonia industry was based. whether water treated with amniunia would be acceptable One of the first applications of ammonia in a closed for growing crops. system was in early 1934 at the grove of R. M. Simon near Dr. Ludwig Rosenstein, at that time chief chemist of Tustin, California. Results were highly undesirable. The Shell Chemical Company, approached Dr. D. R. Hoagland water for this grove was extremely hard, very high in of the University of California to discuss the possibility of calcium and magnesium bicarbonate. The underground applying anhydrous ammonia in irrigation water. Dr. waterlines at this orchard were small, 10 in. (25.4 cm) or Hoagland seemed optimistic and suggested to Dr. Rosenstein less in diameter. During the first tun it was rediscovered that he contact Dr. D. D. Waynick at the Association that when ammonia is dissolved in hard water, calcium

65 carbonate is precipitated. This precipitate would adhere to Agricultural Experiment Station, began field trials with the sides of the pipeline even in fast moving water and a 10- anhydrous ammonia for corn and cotton at many sites in in. (25.4-cm) line could be plugged, or its carrying capacity Mississippi. This work, which was supported by the TVA, drastically reduced over a short period of time. led to widespreadadoptionoftheuse ofanhydrous aiiiiiiunia Although the injection of ammonia into irrigation water in Mississippi, the South. and the Midwest (figure 16). was proven feasible in open systems, the sale of anhydrous Concurrent with studies to investigate the effects of ammonia expanded slowly during the depression years of aiiliydr,~us ammonia on soil conditions and plant gruwtli the late 1930s. Early in 1939 it became apparent that if a was development of the associated equipment needed for big volume of sales was to be obtained it would be necessary distribution and transport. Metering equipment,' tanks, and to apply ammonia in places where irrigation was not transportation methods all evolved with increasing interest practiced, or was not practiced at the time of fertilizer in anhydrous ammonia fur agricultural uses. As the use application. No one had begun work on the problems of anhydrous ammonia spread frum area to area, cumpitnies inherent in the injection of ammonia until equipment became interested in prbducing trailer-type applicators developed by a telephone company was observed laying designed especially for injection. Frequently, the equip- underground cables in the Sacramento area. The equipment ment took the form of a folding "wingup" design whch consisted of an oversized subsoiler with a tube down the allowed movement over highways from field to field witli- back of the blade through which the lead sheathed cable out a special permit. Liquid phosphoric acid was made was passed. The equipment suggested welding a tube down available to agriculture. and applicators were built to dual the back of an ordinary cultivator shank and releasing the apply ammonia and phosphoric acid. Today refinements of ammonia at the depth lo which the shank ran. these early handling and applications systems continue, To test the practicality of such a device, amnionia meter- and the interest in anhydrous anitnonia for agricultural ing equipment was installed on a killifer furrowing tool and purposes continues to increase. Anhydrous amnionia a tube welded down the back of each shank. The first currently represents the niost economical fortii bf nitroget; applicator, though crudely assembled, worked satisfactorily. availahle to the farmer. Three hours after the first application, a cross section of the injection area was exposed and soil was removed in I-in. cubes (2.54-cm) for analysis. Data showed that ROLE OF AMMONIA AS FERTILIZER, ammonia was retained in an area roughly 3 in. (7.66 cm) laterally from the point of ammonia rclcase, downward Principal factors that have made nitrogen the key element about 2 in. (5.08 cm) and upward along the shank channel, in world and U.S. fertilizer use are dramalic crop responses, almost 5 in. (12.7 cm). This type of equipment became changes in economics of production and distribution, and standard for all early experimental work and the first reasonably good availability of raw material. commercial applications. After commercial applications Before the advent of synthetic ammonia, sources of were begun in 1942 and soils ofall degrees of firmness were nitrogen were few and relatively expensive. The major encountered, a more complete study was initiated on the commercial materials consisted of , Chilean nitrate of effect of shank design on the ammonia absorption pattern. soda, by-product ammonia, and other sources. These Concurrently with the early field studies, laboratory work materials were seldom located in areas of consumption; was begun to study the chemical aspects of anhydrous thus handling and transportation increased the cost to the ammonia injection. Ammonia was applied at various rates point where the farmer could afford to use very little for pH and nitrification studies. This work showed that nitrogen. although the pH was initially elevated at the point of injec- The development of the synthetic ammonia industry tion, it soon returned to normal or slightly below as nitrifi- changed this economic picture. Ammonia was converted to cation began. Nitrification was comparable to that for solid new end products and new methods of handling these forms of ammonium fertilizers. Experimental work showed materials were developed. Improved ammonia production that seed germination was inhibited and seedlings were technology led to cheaper products, and these changes led killed if they were placed in close proximity to the point of to an improved competitive relationship between nitrogen injection. However, 3 to 4 in. (10.2 cm) away, all seeds and other plant nutrients. The preeminence of nitrogen germinated and seedlings grew normally, showing nitrogen stimulation soon after they emerged. These early labora- among the plant nutrients is due to its productiun at a great tory and field studies also showed that growing plants could number of geographic locations in the world. Nitrogen for be successfully sidedressed with anhydrous ammonia pro- ammonia production is readily and cheaply available from vided that the shanks were run just at the edge of the air. In the earlier technology of synthetic ammonia pro- rooting zone and from 6 to 8 in. (15.2 to 20.3 cm) deep duction, feedstocks to supply the required hydrogen were (15). In 1943, Dr. W. B. Andrews, with the Mississippi limited to refinery and coke oven gas. In the past 30 years,

66 Ilowever, equipment has been designed to produce synthetic Over the past 30 years, there have been many changes in ammonia from a nuiiiber ofhydrogen wurces. Raw materials the types of nitrogen fertilizers produced. Specific trends now available fur synthetic ammonia production include in nitrogen use are shown in table 16. In 1950, approxi- the following: natural gas, crude oii, coal. lignite, naphtha, mately equal amounts of N were used in fertilizer mixtures and electrolytic hydrogen (see production section, this and for direct application but since that time nearly three chapter). In recent years, suppliesof natural gas and naphtha times as much nitrogen was used in direct application as in have become limited and mure costly. leading to efforts mixtures (figure 17). to develop the tecluiolugy for the production of aiiinioiiia The major factor leading to the rapid increase in direct from coal and lignite. These efforts have only been moder- .application of nitrogen has been the availability of large ately successful thus far, but the development is continuing aniounts of nitrogen solutions and anhydrous ammonia in an effort to make these products more economically plus convenient methods of using these materials in competitive as sources of hydrogen. cropping systems. The econoniics of production and

Figure 16 This photo shows how anhydrous ammonia was first used at the Mississippi Delta Branch Experiment Station

67 distribution favor the use of these intermediate products reason for this economic advantage is that the product can rather than the higher cost solid and liquid NPK products be used directly and requires no additional chemical proc- which require further processing to produce. Figure 18 essing which may add to the cost. The second reason is shows the extent to which the intermediate products have the excellent facilities for transporting and storage of liquid gained favor in the nitrogen marketplace. The economics ammonia. Ammonia is readily transported by truck, rail, of production, distribution, and application of liquids and and barge and now two aninionia pipeline systems exist ammonia so highly favor this system it can be expected to for transportation from producing areas to major consuming increase in relative importance in the future. areas in the Midwest. Many large atmospheric storage tanks have been constructed rarigiiig from 5,500 to 45,000 tons in capacity. These tanks are refrigerated at atmospheric

APPLYING AMMONIA oresure bv using" the ammonia as a refrigerant to inaintain it in liquid state at -28°F (-33.36C) or slightly below. Certain characteristics of ammonia make it a popular fertilizer and also one that must be handled and applied with considerable care to avoid hazards inherent in the Injecting Ammonia product and to maximize crop yields. Anhydrous has a boiling point of -28'F (-33.3OC), and stprage requires high pressure vessels. At tubes mounted on the hack of knives or tines at which the product exhibits a gauge pressure immediately changes to gas. (135.9 x IO4 Pa). The product's high analysis 82% nitrosen, and the fact closure of the injection the lowest cost source of nitrogen for the 19-21 illustrate various

Table 16. Consumption of selected nitrogen direct application materials-United States Fiscal Anhydrous Aqua Nitrogen Ammoniuni Ammonium __year ammonia ammonia solutions nitrate Urea sul fat e (short Ions of nitrogen) 1960 581,924 85.380 194,740 415,855 64.596 106,959 1961 666,234 86,489 292,566 446,585 92,448 115,715 1962 767,425 99,783 371,379 468,523 132,804 1 16,687 - 1963 1,006.762 116,448 471,392 499,378 165,198 147,121 1964 1.148,071 157.531 526,755 55 5,320 184,327 148,865 1965 1,281,968 163,933 595,859 547,488 193,713 161.848 1966 1,606,872 200,814 712,445 610,705 21 1,615 165,797 I967 1,973,596 177,175 784,392 710,503 227,952 174.834 1968 2,457,261 163,047 803,092 786,346 243,359 167,306 1969 2,576,431 143,715 839.925 863.792 264,755 157.042 1970 2,844,058 143,778 969,882 952 36I 242,758 160.91 1 1971 3,254.1 60 151,321 1,036,498 969,091 273,607 185.386 1972 2,982,274 150,165 1.009,l 19 1.015.548 357,386 192.205 1973 2,700,704 I30,4 16 1,144,407 1.095,65I 432.420 197,088 1974 3,427,672 148,071 1,184,507 1,061,743 467,934 193,025 1975 3,294.794 140,497 I.I 9 I ,69 1 941,881 524.183 170.6 14 1976 4,025258 146,180 1,620.308 985.555 729.250 219,760 1977 4.040.328 132.204 1.685.028 936,863 855,344 220.226 1978 3;721.289 1 15,500 1.591.935 820,356 880.753 I @,?OR 1979 4,G03.X45 98.5G3 I ,?23,860 841.272 967,040 157,546 1980 4,499,947 133.219 I ,908.9 17 884.7 I8 946,359 175,566 -1981 4.6 1 0.005 146.474 2,047.464 986.630 ~______980,054 197,937 Source: USI)A. C'onirncrcial I~crtilix~s.St:iti

68 types of knives and equipment which can be used for the Figure 22 (17) illustrates the effect of soil moisture application. content and spacing of applicator outlets on ammonia Proper closure of the channel requires the right amount retention in soil. An increase in the moisture content had a of soil moisture. Overly -1 does not fill the knife pronounced effect onlowering the gaseous loss of ammonia. opening properly and water may actually compete with A point of practical significance is the reduced loss that ammonia for adsorption sites on colloidal surfaces. Very accompanied the narrower applicator spacing when the same dry soil can retain large amounts of ammonia as long as the rate of ammonia wasapplied.Greater losses at wider spacing soil tilth allows sealing of the knife opening. Cloddy soils undoubtedly resulted from greater localized concentration also present problems in closure. at the point of injection. Depth of placement in most soils ranges frommn. (15.2-22.9 cm) but sandy soils require deeper placement because of their low ammonia retention capacity. Spacing ANHYDROUS AMMONIA FROM between channels is usually &&.in. (30.5-45.7 cm) PRESSURE TANK MOUNTED +DIRECTION OF TRACTOR depending upon the crop. Wider spacings often result in ON TRACTOR Ullevrll growth.

Figure 19. Front swept knife for ammonia application

-t = DIRECTION OF TRACTOR PULL

Figure 17. Nitrogen fertilizer consumption (16)

z 0Y t z

z 0 A d I

BACK SWEPT FRONT SWEPT

Figure 18. Consumptionofnitrogen fertilizer materials (16) Figure 20. Various type knives for ammonia application

69 Initially the ammonia is retained in a localized zone increases the number of acres that can be treated in a day about 1 to 3 in. (2.5-7.6 cm) in diameter around the point which results in lower labor arid equipment costs per acre of injection. The concentration in this limited zune usually of applied nitrogen. Another way to decrease application is such that the ammonia is toxic to micro-organisms and costs is to use equipment that applies ammonia during otlier sterilizes the soil to the point where no nitrates form tillage operations. Figure 23 shows a sketch of a V-blade except around the fringes of the zone. The ammonia or used in wheat growing areas where anhydrous ammonia is ammonium ions gradually mvve to less concentrated areas applied during tillage operations which follow wheal and rapidly change to the nitrate form. A inore complete harvest. It is the practice in those areas to pass tlie V-blade discussion of the factors involved in tlie retention of about 3 to 4 in. (IO.? cm) beneath tlie soil. This helps to ammonia in the soil and the interactions with clreniical kill weeds, tills the soil_ and leaves a protective stubble constituents of the soil follows later in the chapter. mulch vn the surface tu prevent wind and water erosion. A possible disadvantage to the use of anhydrous ainnionia Usually 50 to 80 Ib/acre (56 to 90 kg/ha) of aniiiionia is the power requirement necessary to place the material in can be applied without noticeable losses of aiiiinoiiia. the sofit a depth where losses are avoided. The iiicrease in 3-i popularity of nitrogen solutions a!id aqua anini~nia is Application of Ammonia partially due to the application of these niaterials at shal- in Irrigation Water lower depths, thus reducing energy requircnients and fuel costs. There has been an effort to decrease application Application of plant nut rielits tliiwugh irrigation weter costs of ammonia through the use of large tractors so that has several possibleadvantages. Among them are(l)nutrietits an applicator with wider swatli widths can be used. This may be applied when crop or soil conditions would pro-

Figurc 21. Tractor drawn dmmunia application equipment

70 hibit fertilization by other conventional means, (2) nutrient 50 ppm). Due to the high vapor pressure of ammonia it supply can be regulated to coincide with nutrient deniand should not be used in sprinkler systems. by tlie growing crop. (3) less labor and equipment are Distribution-Where furrow or flood irrigation is prac- required for application, and (4) soil conipaction by heavy ticed, more water percolates into the soil at the head of tlie applicatiun equipment is cliniinated. Like any practice, the furrow or flood basin than at the tail. Likewise, niure application of aiiiiiionia in irrigation water has some liniita- ammonia would be expected to be absorbed at the head of tions which niiist be recugnized in adapting it to practical the run. To offset this, ammonia is sometimes withheld iiperJtions. during the first part of the irrigation period, thus avoiding The objective inust be to place tlie N wliere the crop can part ofthe accumulation at the head uf the furrow UT basin. make use uf it efficiently. The ultim;ite niciisure uf success Soil textures, slop of land, crops to be grown, and many will be crop response when aniiiioiiia application in irriga- other factors help determine the length of water runs. The tion water is combined with other practices in a complete shutter the run tlie more uniform the ammonia application crop production program and is evaluated against alternate althougli it inay be inore economical from a labor stand- choices. Discussiuii that follows considers several impurt;int point to run water 1/4 mile (0.4 km) or further. More Fdctors that iiiust be iiiunitored clusely if tlie application of efficient use of ammonia will result by limiting runs to l/S ammonia through irrigation water is to be successful and mile (0.2 km) or less. Extremely long runs. which require incurpuratcd into othei~ciilturdl prdctices. ii long time to apply water, increase the effects of deep Volatilizatioii-Cliapnian (I 8) Sound llie ciiticeiitration and irregiilar wetting (especially on coarse, porous soils). of aiiiiiionia added to irrigation waters could decrease as Since aiiiinunia is adsorbed by soil surfaces, losses by much as 50% between tlie head and tlie lower end of the leaching on mediuni and heavy iextured soil are nut furrow. This decrease may be due to lixation of tlie signific;int. ammonia by tlie soil or vdatilirJtiun; however. the latter Placeriierit-Bec;iuse ammonia is held in the tup 2 or 3 is believed to be the larger factor. The quantity of anhy- in. (5.1 to 1.6 cm) of soil when applied in irrigation water, drous aiiiiiiiiiiia in water shuuld not exceed I 10 pprii and it is essetitially a surf;ice placement. For row crops, the in some liiglilyalkaliiie water sliould not exceed 40-50 ppni. surface placement inay result in N beiiig placed above tlic It is recoiiitiiended that anitiioni:! be applied in irrigation major root zone and perhaps in soil that is too dry for water when air teiiiperatures are 40'F (IOOC) or less. efficient rout feeding during at least part of the period Aiiiiiionia losses by voliitiliziitiu~iare usually negligible up between irrigations. Between irrigations, niucli of the tu 72°F (22.1OC). Studies shuw that very signific;int losses aiiiiiionia inay be converted tu nitrate which would move can and do iicciir at teniperatures over 100°F (373°C) into the ruot mie with the next irrigation. In this case. any unlcss tlie cuticelitration of aniiiiuiiia is kept low (40 to delay in N reaching the crop could be avoided by applying aiiiiiionia tine irrigation earlier. Surface placement of :iiiiiiioiiia inay also stimulate weed growth. thus cuntributing to the problcrns ufweed control. Curbonate Precipitate-Many waters used fijr irrigation coiitaiii large aiiiounts $if dissolved calcium and swietinies

1

IEPTH 6"DEPTH DEPTH

16" SPACING 40" SPACING

Figure 22. Effect uf soil moisture content and applicatur spacing un aiiimuiiia retciitiun in the soil Figure 23. V-blade diiinionia applicator

71 magnesium. When ammonia is added to these waters the pH but should present no problem where N content of the is increased, bicarbonate is converted to carbonate, and a petidles is being monitored and need can be anticipated precipitate of calcium carbonate forms. In closed systems, a few days in advance. This method of fertilization permits this precipitate can cause serious problems-clogging lines, growers to apply fertilizer to a crop which is at the stage valves, gates, siphons, etc. When water is carried in an open where entering the field with mechanical devices for fertili- ditch this deposit is evident as a white precipitate. Loss of zation niay be impossible. dissolved calcium also can have a delctcrious cffect on water quality if the water contains significant levels of sodium. Sodium compounds wluch are more soluble than AMMONIA RETENTION AND calcium compounds at high pH levels remain in the water. ADSORPTION IN SOILS This results in a higher sodium to calcium ratio in irrigation water which reduces percolation rates and restricts down- Following NH, injection, essentially all the ammonia is ward movement of the water and dissolved ammonia. retained by the soil unless improper application techniques In certain areasof the Southwest, sulfuric acid is available are used. High NH, concentrations and high pH may be to growers and can be used to acidulate the irrigation water found within the retention zone. Immediately after prior to incorporation of ammonia, The result is a lowering ammonia application, concentrations ranging from 1,000 of pH, thus eliminating the precipitation of calcium carbon. to 2,500 ppm Nhave been reported to occur in a horizontal, ate from the water. Sodium buildup in the soil is reduced cylindrically shaped zone with a diameter of 1.2 to 2.4 in. and infiltration rates are improved since calcium remains (3 to 6 cm) (17). Concentrations approaching 1,000 ppm in the irrigation water. may persist in this zone for a period of several weeks. Obviously the use of anhydrous ammonia in irrigation Through gaseous diffusion and diffusion in solution, the water has several important advantages. However, use of NH, distributes radially outwards From the line of injection. this practice must be tempered with caution bascd on a According to available data, the ultimate diameter of the complete knowledge of how ammonia reacts with water IocaliZed zone generally does not exceed 3.9 to 4.1 in. and soil, if its maximum effectiveness is to be obtained. (IO to 12 an) with knife application. This means that at The agronomic effectiveness of ammonia applied in wster usual field application rates and,with a knife spacing of appears to be equal to other forms of soluble nitrogen when 39.4 in. (100 cm), a maximum of 20% of the plow layer precautions are taken as outlined above. will be influenced by NH, (20, 21). Distribution patterns Cotton growers in irrigated areas have found water run may be considerably different with other implements such ammonia is an effective itieansofsupplcmenting thc nitrogen as a V-blade (22). supply during the growing season. The data in table 17 The volume of the retention zone mainly depends on the from Hopkins (19) show a slight delay in nitrogen uptake rate of NH, application, row spacing, moisture content, from aninionia compared to a nitrogen solution containing buffering capacity, texture, and nitrification capacity of the 68% of the N as nitrate. This delay in uptake niay suggest soil. The initial distribution pattern is controlled to a large the need for modifying the timing of ammonia application, extent by the moisture content of the soil because NH, is highly soluble in water and NH, diffuses slowly in the aqueous phase as compared to the gaseous phase. Soil moisture functions largely as a temporary repository of Table 17. Nitrate-nitrogen content of cotton petioles NH, from which it is subject to slow movement by mass after N applied in irrigation water (19)-- flow with water or by gaseous diffusion away front the N03-N in petioles zone of high NH, concentration (23). The ammonia gas Days after N source immediately following injection is converted to ammonium yplication Untreated CANa NH, hydroxide and ammonium ions which in turn are absorbed PPnt by the soil clays and organic matter. 0 750 800 750 Under field conditions, soil moisture content may have 1 400 1.150 480 an indirect effect on the soil’s capacity to retain applied 3 400 1,100 500 ilmnionia by affecting sealing of the injection channel (23). 6 500 I.000 800 Hanawalt reported that a high enough moisture level was 11 500 1,000 l.000 inore important in this respect than a high cation exchange 15 500 900 1.100 22 400 900 1.050 capacity (CEC), exchange acidity, and clay and organic matter contents (74). 30 ~-250 780 900 “Cnlciom-llmmoniiiiii nitr3te s&ution containing 11.6% NOq-N In a study using 76 soil samples from 20 different soil and 5.4:; NH4-N. profiles of the Pacific Northwest, Young concluded that total NH3 retention could be described by the following capacity for ammonia adsorption than those with only relationship (25). external surface area.

ppni NH,-N = 147 t 596 (%organic C) + 43.4 (%clay) Mechanisms of Ammonia The equation indicates the importance of the organic Absorption on Minerals fraction on NH, retention by soils. The highly pH depen- dent CEC of soil organic matter was largely responsible for The reaction which results in the greatest stability of this property. ammonia on mineral surfaces is its conversion to the Nitrogen applied to soil as anhydrous NH, is subject tu ammonium form. The formation of the NH,' proceeds adsorption by the organic and mineral fractions of the soil when NH, reacts with protons (H'). Thus an acid soil or and may be dissolved in soil water. When applied to moist minerals having exchangeable hydrogen ions on the CEC soil and at proper depth, sorption is nearly complete and may provide the necessary protons for NH,' formation. rather limited inuvenient occurs at the point of placement. A number of physical and chemical reactions are respon- sible for NH, retention. They range from weak physical These exchangeable hydrogen ions are associated with nega- sorption by H-bonding to irreversible incorporation of NH, tive charges within the clay lattice arising from isomorphous into soil organic matter. Physically adsorbed NH, is in a substitution of trivalent aluminum for tetravalent silicon dynamic equilibriuni with the gaseous phase and is subject in the tetrahedral layers and divalent ions such as MG2+ to movement by diffusion. Physical absorption only occurs and FEZ' for AI3+ in the octahedral layers of 2:l type clay when there is a positive pressure of NH, in the soil minerals. The ammonia may react also with protons arising atmosphere. As soon as NH, pressure decreases, equilib- from hydroxyl groups associated with silicon on the edges rium is shifted and some of the physically adsorbed NH, of clay minerals or in amorphous material in the soil. returns to the gaseous phase. While in the ammonium form, the N cannot be lost to The chemisorbed NH, is strongly bound tu specific the atmosphere nor can it be easily leached out of the soil adsorption sites on soil colloids and little migration will by water. The amnionium form (NH,') has a positive occur within the soil. The basic mechanism for chemical electrical charge and is thus attracted by the negative binding of NH, consists of acquisition of a proton by the electrical charge on soil mineral surfaces. It enters into NH, to form NH,'. which is subsequently bound to ordinary cation exchange reactions in the same manner exchange sites on clay and humus particles. Ammonia may as nietal cations on the exchange complex such as K, Ca, also react irreversibly with organic matter with formation and Mg. Exchangcablc NH,' is available for plant utiliza- of various NH,-organic matter complexes. These complexes tion or microbial nitrification. Thus this reaction converts are characterized by high chemical stability and high resis- a toxic gaseous compound into a nontoxic relatively tance to microbial decomposition. imniobile nutrient. The size and kind of iiiineriils in soils affect the amount In addition to aninionium formation, another adsorption of ;imnionia wliich may be adsorbed. Generally. the amount mechanism plays a role in NH,' retention in the soil. This of adsorption is proportional .tu tlie surface area of the reaction involves coordination between ammonia and adsorbent. thus a sandy soil has a lower specific surface exchangeable metal ions on the mineral surfaces, and has (surface area per unit weight) than a clay soil and its capacity a lower energy of reaction than that of NH,' formation for adsorption of ammonia is consequently lower. A general (16, 27). The NH, molecule hasa pair of unshared electrons obseivatioii has been that losses of ainnioniil after applica- on the N atom which niay be shared with ions capable of tion of anhydrous miiionia are more likely on sandy than accepting these electrons. The resulting bond is called a un heavy soil. Even in saiidysoils the clay fraction is usually coordinate bond of "ion-dipole" interaction. The adsorp- the iiiost important iiiiiieral portion with respect tu tion of imiinonia un the cations in the soil system is adsorbing properties. A relatively small percentage of clay analogous to the hydration of these ions by water. In fact. niay coiiipletely dominate tlie chemical and physical the very similar molecules. animonia and water, coriipete fur coordination positions (28. 29). properties of a sandy soil. 111 addition to the amount of The NH,' ioii will enter into ion exchange reactions clay. tlie kind of clay present is a niajur Eictrrr affecting with other cations :ind be displaced into the soil solution. :idsorption. Clay iniiicrals such as kaolinite have unly Under equilibrium conditions. a certain portion of the external surfiice on the order of 2.4 x to 4.9 x io4 ammoiiium ioii will exist on the CEC of the soil and a ii2/lb (5 tu IO iii2/g). wliile expanding clay minerals such smaller portion in tlie solution phase depending on the total as iiii~~~tmorillo~iiteand vermiculite have internal surface amount of NH,' ion present. tlie kind and amount of as well as total surkice areas ofthe order of 3.9 x IO6 ft2/lb associated exchiingeable cations in tlie system. and the (800 mz /g). The exp:indingclay ]minerals thus have a greater iiature of the ]mineral providing the cation exchange sites.

73 Ammonium is very similar to the potassium ion in its although it is now established (36) that the ability of these chemical reactions, being of similar size and valence; thus iniaterials to retain monovalent catiuns is somewhat lower its chemical behavior in soils is quite similar (30). When than for polyvalent ones. Ammonia reacts with carboxyl NH,+ is on the surface of soil minerals, it is considered to+ and possibly other acidic groups present iii tlie organic be easily available for either uptake by plants as the NH4 fractiun to form salts which might be termed i'aninionium ion or for nitrification by soil micro-organisms. If vermicu- liuniates." In tlie immediate vicinity uf tlie ammonia lite t pe minerals arc prcsent in the soil, it is possible that injection band, all such groupings are prob;lbly saturated NH4Y may move from a position on tlie external sureace of with aninioniuin ions. the mineral to internal exchange sites where it niay be At tlie present time, several plausible mechanisms have rendered relatively unavailable fur immediate plant uptake been proposed for ammonia fixation reaction by organic or nitrification (31,32). matter. These are based 011 the fact that ammonia fix:itiun proceeds tilost favorably at lugh pH values and that autoxidatior occurs simultaneously. Mattson and Kuutler- Ammonia-Organic Matter Interactions Anderssoii (37) suggest that fixation takes place in the Nunieruus factors affect the relative importance of lignin-derived fraction vf soil organic inatter. The presence organic matter and clay in retaining ilinnionia in soil. Organic of iir~niaticrings with two or niore hydroxyl groups was matter can play an equal or greater rule than clay in stabi- believed essential. They postulated the furination of amino lizing amniunia in the soil, depending upon the type of soil phenols. which. upon oxidation to more reactive quinone- under consideratiiin. imino conipounds could then form stabie condensation The inniiediate effect uf injeciiiig anhydrous amniunia pruducts with N incorporated in heterocyclic rings. into the soil is to create an alkaline reaction in the retention The important question with respect tu the anini~iniuni zone. The high alkalinity and the associated higli concentra- fixed in soil organic matter is wlietlier it is available to tion of ammonia cause partial sterilization of tlie soil plants. Research on this subject indicates that the avail- including drastic reduction in tlie number of nitrifying ;ibility is relatively low. In a greenhouse experiment with ~irganisms. Nitrification of injected aninionia coninleiices soil containing fixed ammonia which was tagged. Burge at the periphery of tlie retention zone and proceeds toward and Broadbent (38) observed that. in one cutting of the center. Because Nitrobacter which oxidizes nit,rite to sudangrass. 4.3% uf aiiiiiionia was utilized as coinpared with nitrate is iiiore sensitivc to liigli pH and high aminunia 1.14% of thc untaggcd native soil N. In the second cutting. concentr;ition than iVitrosoriiorius which converts animo- only 1.31% uf the remaining tagged N was taken up by tlie ilium to nitrite, there can be an appreciable accumulation pl:ints. This suggests that the availability of at least part uf nitrite; the nitrite may subsequently react chemically (if the fixed ammonia may be siiniewliat better than that of with the soil organic fraction to form stable organic nitrogen indigenous soil N but less than that (if niiiture crop residues. complexes and nitrogen gases. In summary it appears that tlie availability of anhydrous The re1;ttfve iniport:ince of inorganic and organic soii amnionia to plants and the degree to which lusses occur components in retaining ammonia depends (in tlie nature by vol;itilizatiun depend to a considerable extent oii the of the soil, hut in general the urg:inic fraction is more chemical reactioii~ betweeii ainnionia and soil organic reactive on a weight basis. Mortland (33) investigated the matter. The relative inipoitance of organic and inorganic adsorption [if iininionia by dry clays and inuck and found soil components in retaining aninionia depends on tile both H-muck and C;i-muck adsorbed more antnionia than iiatiire of tlie soil, but tlie organic fraction is more reactive did the clays. Sohn and Peech (34) found that organic on a weiglit basis than is the inorganic fraction Ltibur;ilory soils fixed considerably inure aninionia than mitiera1 soils: investigations indicate that as much as one-fourth of the for the niinerd soils. the quantity of :immonia fixed was anhydrous aniiiioniii added to soil niay be fixed by lignin highest for acid soils containing large iinioiints of organic and ligniri-like substances. Aniniuniacal N from sources iiiatler. Thc results of Sohn and Peecli (34) indicated that uilier than anhydrous :Iniiiionia is subject to the same :it Icast 50% of the :~itiiiioni:i fixed in their experiment was fixatiun processes. due to re:ictiun of aiiinionia with soil organic tii:itter. Burge and Bim~dbent (35) tiieasured fixatiiin of anliydriius ainniunia by dry saiiiples [if urganic soil ranging from 14% 10 43% carbon and fiiund that tlie quantity (if aniiiioiiia fixcd was linearly related to carbon content. They CBICII- kited that under :ierobic conditiotis. I molecule of aintnonia was lixcd for every 29 carbcn atiiiiis. Organic nutier in soils exhibits ciition rctentiun cnpac- ities wliicli are liigli in relation to those (if titinera1 soils.

74 4. Fwtiliier Momd International I:ertiliLcr Development Ccnter Nlfmxeri in Apricullwol Soils. Agronomy No. 22. Amer. and Unitcd Natiuni Industrial Development Organization, SOC. Agr., Crop Sci. Soc., Soil Sci. Soc., Madison, WI, 1982. pp 49-71, 333-340, Decembcr 1979. 22. Murphy, Larry, "Ammonia Agronomics in Perspectivc," 5. Bridges. J. D.. "1:ertilizer Trends 1979." National Fertilizer pp 23-32, In Situation 82, TVA Fertilizer Conferencc, Development Ccntcr, TVA Bullctin Y-150. pp 10-13. January St. Louis, Missouri, August 18-19, 1982. 1980. 23. Papendick, R. I..and J. F. Parr, Retention of Anhydrous 6. "Symposium-Ammonia from Coal," Nation31 Fertilixr Ammonia by Soil. 111. Dispensing Apparatus and Resulting Dcvelopment Center, TVA Bulletin Y-143, July 1979. Ammonia Distribution. SoilSei. 102:193-201 (1966). 7. Czoppon. T. A,, and L. J. Buividar, "Which Feedstock for 24. Hanawalt, R. H., "Soil Propcrties Affccting thc Sorption af Ammonia," Hydrocorhorr Processir,~58 (9), 197-200 Sep- Atmospheric Ammania," Soil Sei. Suc. Am. Proc. 33~725.29 tenibcr 1979. (1969). 8. The I:crtilixi Institotc, "Fertilircr Industry Energy tisc 25. Young. J. L., "Ammonia and Ammonium Reactions with some Survcy." Cnlcndar Year 1979. Wnshin#on. D.C., 1980. Pacific Northwcst Soils." Soil Sci. Sac. Am. Proc. 28:33945 9. Waggoner, 0. R., "Effcct of Energy Optimiratian on Design (1964). and Operation of Ammonia Plants," 1:ertilizcr Association 26. Cloos, P., and M. M. Mortland, "Expansion and Electrical of India, Ncw Dclhi, India, Novembcr 30, 1978. Conductivity of Montmorillonite in Ammonia Atmosphere," 10. Davis, C. H., and G. M. Blouin, "l:ertilizers, Encrgy, and Clayroiid C1a.vMirierols 13:217-26 (1965). Opportunities for Convcrsntion." Southwest I'crtilizer Con- 27. Jamcs. D. W.. and M. E. Harword. "Mechanisms of Ammonia ference, Houston, Texas, July 17, 1980. Adsorption by Montmorillonite and Kaolinitc," Cloys md 11. Jamcs, G. R.. "How Ammonia Proccrres May mange io Suit CloyMinerols 11:301-20 (1962). thc Energy Available." 31st Annual Mceting, The I:wtilizcr 28. Brown. J. M.. and W. V. Bartholomcw. Sorption of Anhydrous Industry Round Tablc, Washington. D.C., Novcmbcr 3-5. 1981. Atiimonia by Dry Clay Systcms. Soil Sci. Soe. Atn. Pn,c. 12. "Chcmicals from Coal." Oil and Cas Joimol 80 (I), 120, 133 26:258-62 (1962). January 4, 1982. 29. James, D. W., and M. E. Harward, "Competition of Ammonia 13. Slack, A. V.. and G. R. James. Eds., "Ammonia, Part IV," and Water four Adsorption Sties on Clay Minerals," Soil Sci. Mxccl Dckker. Inc.. NY, 1979. SOC. Am. Proc. 28:63640 (1964). 14. Brown. J. W.. "Nitrogcn Outlook 1982," 31sl Annual Meeting, 30. Mortlnnd. M. M., "Ammonia Interactions with Soil Mincrnls," The 1:crtilizer Industry Round Tablc. Washinptan. D.C., pp 188-97. In 11. M. McVickar el 01. (eds) Agn'cdnrrol Novcnibcr 3-5. 1981 Aiihydroirs Amrwmio Techr!o/o~yold Use. Proc. Symp. 15. Lcavitt. I;. Haucn. 3 rcvicwofchal,tcr7,"Agricultoral Ammoni;t SI. Louis, MO, September 29-30, 1965. Agric. Ammonia Inst., Equipmcnt Dcvclopmcnt and History," pp 12542. McVickar Mcmphis, TN, Am. SOC. uf Agron. and Soil Sci. SOC.of Amer., cf 01. (eds) Ayricrdt!wol AriI,ydroos A,nr,zor,io, Teclr,doy),o,id Madison. WI. 1966. Urc. Proc. Synip. St. Louis. MO. Scptcmbcr 29-30, 1965. 31. Banhad. I., "Cation Erchangc in Soils: 1. Aiiinioniiiin Ikation Agric. Ammonia Inst.. Mcmphis, TN: Am. SOC. of Agron.. and and Its Relation lo Potassium I lion and to Dctermination Soil Sci. SOC. of Amcr.. Madison. Wl. of Ammonium Exchange Capacity," Sui/ Sci. 72:361-71 16. Hxgctt, Norn~in. Person:il corrcspondcncc. Ilcunrmics and (195 1). M:irkcting Rcscarcli Branch, Tcnncsrcc Vallcy Authority. 32. Hanway. J. J.. and A. I). Scott. "Amnioniom I:ixation and Musclc Shonls. Alnbama. 1982. in Ccrtain Iowa Sails." Soil Sci. 82:379-86 (1956). 17. ~I\lcDowcll.L. L., and G. li. Smith. "'I'hc Rctentiunand Rcactions d. M. M.. "Adsorption of Ammoni:l by Clays and of Anhydrous Aminonia on Differcnt Soil Typcr," Soil Sci. Mucks." Soil Sci. 80:11-18 (I955). SOC. A!,>.Pwc., 22:38(1958). 34. Solin, I. 8.. and M. Pcech, "Rctention and Fixation of 18. Clinpnian. II. ID.. "Thc applic:ition of licitiliicri in Irrigation Aiiiinonin by Soils," Soil Sci. 85:1-9 (1 \\'3tcr:'Hrrrer Cmiu vifh Plmr tbod 40:6-IO. 4347 (1956). 35. Burg". W. I)..:md 1'. li. Broadbent. 19. Hopkins. L.. "USC of CAN 17 snd Aminoni;i 3s Supplcmcntal by Organic Soils." Soil Sci. SOC. Ai,icr /'roc. 25: 199.204 Nitrogen on Cotton." OrfhoPloirf IbodNcu,s.(Chcvrrm Chcm. (1961). Co.) 5 (9). 1-3 (1959). 36. Mortland. M. M., "Rcxtions

The term “liquid fertilizer” used in this chapter refers be relatively low. For example. an 8.240 was the popular to solution type fertilizers sucln as aqua ammonia. non- grade norinnally produced from merchant-grade furnace pressure nitrogen solutions, and liquid-mixed fertilizers in orthophosphoric acid. Later the furnace acid process was which all of the plant nutrients are in solution. About I developed for the production of superphosphuric acid that million tons of nonpressure nitrogen solutions and about 3 contained between 75% and 80% P,05 and had a poly- million tons of liquid-mixed fertilizers were used in 1982 phosphate content between 40% and 85%. With the higher for direct application in the United States. Use of nitrogen pulyplnosphate content furnace phosphoric acid, it was solutions is growing at an average rate of about 10% per possible to produce an 11-37-0 liquid grade (figure 24). year and liquid mixtures at a rate of abuut 3% per year. Re;isirns fur llie popularity of solution type fertilizers are: 1. They are easy to transport and handle and usually can be applied at a much faster rate than solid fertil- izers. - 2. They can be applied tinore uniformly tlliatn solids. 3. They are excellent carriers of pesticides. which sotlie- times can be added at the farm. Because solutiun fertilizes-pesticide iiiixtures can be applied uniformly. excellent pest control can be obtained. and one pass across the field can be eliminated. 4. Unless solubility is exceeded. solutions are excellent carriers of inicrunutrients: it is possible to accurately place the sinall amount of required tnicr~nutr~ent at the desired location with efficient fertilizer appli- cation. 5. Solutions in nnost instances have a nianufacturing and application cost abuut the sainne as solid mixtures. 6. Solutiuns are especially well adapted to new :ind special typesof application such asapplication through drip irrigation or sprinkler type irrigation units. Solutions are excellent for accuriite placement in 7. 0200 0290 0300 0310 0320 0330 0340 the ruw at locations where they cain be used efficiently. Solutions were originally produced fronn mercl~ant-grade N P205 WEIGHT R4TIO furnace orthophosphoric xid which conl8ined 54% P205 Figure 24. Effect of polyphosphate level and and no polyphosphate. A nnajur problem in producing sob N:P20, weight ratio on solubility of tion innixtures was that plant nutrient cunce~ntrationhad tu ammoniated phosphoric acids at 32OF

76 Also it was possible to dissolve significantly larger quantities ammonia about 3 to 5 in. (8 to 13 cm) beneath the surface. of micronutrients in the 11-37-0ammonium polyphosphate Applicators used for applying aqua ammonia are similar solution. In the late 1960s and early 1970s, it became to those for anhydrous ammonia because they have injection ilnpractical frorii the standpoint of cost to produce fertil- knives. Since these knives penetrate about one-half the izers from furnace phosphoric acid fur general agriculture. depth of anhydrous ammonia knives, much less power and Somewhat earlier TVA had developed a process for the energy is required to pull them. Also, aqua ammonia can production of superphosphoric acid from wet-process be applied at a much faster rate. It can be applied during phosphoric acid (I. 2) and at that time also developed such field operations as plowing and disking without processes for the use of wet-process superphosphoric acid high ammonia loss. to produce 10-34-0 or 11-37-0 liquid fertilizers (3, 4). Some companies apply aqua or anhydrous ammonia Processes for the production of the ammonium poly- through ditch irrigation systems. This is done at very low phosphate solutions from wet-process superphosphoric concentrations and during early spring when the tempera- acid will be described later in this chapter. Although many ture of irrigation water is low. Enough ammonia is added of the problems encountered with the use of solution type to produce a slight amnionia odor in the irrigation water. fertilizers have been solved. there continue to be some Possibly this technique may help developing countries with problems which require additional work. Some are: application. 1. It is impossible to produce high analysis potash content in solution-type mixtures from usual raw materials. LIQUID SOLUTION FERTILIZERS 1. In some instances, it is impossible to dissolve the desired aniouiit of micronutrients in tlie solutioii Nitrogen (Fe, Mn, etc.). The solution type fertilizer business is well established One of the lnost popular liquid materials for direct and now represents about 60% of the fluid mixtures con- application in the United States is nonpressure nitrogen sumed in tlie United States. solution (28.0-0 to 31-0-0). Use of these solutions is increasing faster than is anhydrous ammonia. During 1982 abuut 7 million tons of these solutions was used AQUA AMMONIA in the United States. The must frequently mentioned reasons for the increased popularity of nitrogen solutions Aqua ammonia is not nearly as popular in the United are usually the same as for liquid mixtures. In addition to States 3s anliydrous aiiinionia;but,it is safer to use. Because the advantages mentioned for liquid mixtures, the reasons of the safety factor. it is a niore practical material for use most dealers and farmers prefer to handle these solutions in developing countries. instead of anhydrous aninionia are: The niost popular aqua aiiiiiiunia solution c~iitains20% 1. Nitrogen solutiuns are safer tohandle,store,transpurt, N and exerts no significant vapor pressure at teniperatures and :ipply. below 97°F (36OC). Therefore. aqua ammonia of this 2. Storage and transport equipment costs are less for concentration is stored in covered nolipressure storage nonpressure nitrogen solutions. tanks usually built to withstand 5 Ib/in2g (3.5 x IO4 Pa) 3. Applicatioti speed is much higher for nonpressure and equipped with pressure and vacuuiii safety valves. nitrogen ~~Iutiun~which results in lower application These valves are set to open at 1.051 and 0.991 times costs. absdute atmospheric pressure. Nitrogen solutions usually are produced frorn urea. Aqua aninionia usually is produced in a plant in wlucli aiiiiiioniuni nitrate, and water. They contain a corrosion anhydrous a1iiiiioiiia. water. and recycled coul aqua inhibitor and can be stured arid used in mild steel (carbon aliinioni:i are iiiixed ciintinuously in a simple pipe iiiixiiig steel) equipment. Solutions sold in the United States have chimber. Enougli wilier is added to adjust the specific three c~~iceiitrations:28%, 30%. and 3'2% N. Their salt-out gravity of the liquid to tliiit of iiqua aiiiiiiunia containing temperatures vary directly with their plant nutrient concen- 20% nitrogen or other desired concentrdtions. A 50-ton trations. Some of the physical and clieinical characteristics tank car of ;~nIiydr~iusa1ii1iionia can be cuiiverted to 206 of the three nonpressure soIutions are shown in table 18. tuns of aqua mnioni:i in 5 to 8 hours depending upon tlie The inhibiting agent niust often used in these solutions is plant's cooling capacity. a small quantity of anhydrous iiniiiionia: usually abuut Because aqua :iiiiiiioni:i has :I low vapor pressure. it does 10 Ib of ammonia/ton (5 kg of NH,/tun) of plwduct is [not need to be injected as deeply into the soil as does ;inIiy- added to adjust the pH of tlie solutiun tu 7.5. Another drous ammonia. Most operators have fuund they do not effective inhibiting agent is aninionium phusphate. Only il [lave excessive :1111111011i:1 losses if they inject tlie aqua snio11 quilntity of aniiiioiiiitlii phosphate (0.05% to 0.3%

77 P,O,) is required to inhibit the solution. This plrosphate Recently in the United States there has been soiiie material reacts with the mild steel tank to form a corrosion interest in producing nitrogen solution from prilled urea inhibiting iron phosphate film. and amtnoniuni tiitrate. When using a b:itcIi mix tank to produce solution from these solid materials. hot water is Nonpressure Nitrogen Production required to speed dissolution of the materials. Usually enough heat is supplied so that all solids will be dissolved Two types of production processes are used-batch and at the end of the 30-minute mixing time. About 50 Ib/ton continuous. Both are fairly simple; in each process concen- (25 kgiton) of saturated steam is required for a reasonable trated urea and ammonium nitrate solutions are measured, mixing time. This process is more expensive than the solu- mixed, and then cooled. In the batch process, solutions are tion process; however. special pricing policies can make it weighed in a mix tank and the corrosion inhibitor is weighed desirable to produce ii solution froin solid materials. Also, separately and added to the mix tank. Finished product it may be practical to use this prucedure in some developiiig is cooled after it is mixed. The continuous process is similar countries where solid materials arc avnil:tble. except the urea solution, ammonium nitrate solution, Recently, one ammonia pipeline cunipany, which also water. and corrusion inhibitor are metered and fed continu- has ii river termin;tl close to the pipeline. decided that it ously lo a mixing chamber similar to the simple baffled was economical to produce hot aiiiinoniuiii nitriltc solution mixer shown in figure 25. Material from the mixing chamber from ainnionia at the pipeline and mix it with prilled urea is cooled and pumped to storage. received by barge. A similar procedure may be advisable for countries receiving anhydrous atntnoiiia and prilled Table 18. Physical and chemical characteristics urea by ship. At the point of delivery, anhydrous ainnionia of three nonpressure solutions (5) could be converted tu nitric acid and then to hut itnimoniuni Grade, 76 N 28 _____~-30 32 nitrate sdutiun. This aninioniuni initiate siilution could Comvosition. '%. by wcieht then be mixed with the prilled urea to produce the non- Ammonium nitrate 40.1 42.2 44.3 Urea 30.0 32.7 35.4 pressure iirea-aniiiiOniom nitrate sulution.

Solution Fertilizer Mixtures

As mentioned earlier. the use (ifurthopliosplinte base solittion (8-24-0 froni furnace pliosplioric acid) is pl-acti- cally nonexistent at the present time. This is because the cost of pioducing the solutions from furnace-grade phiis- phoiic acid is tuo high as compared to producing thetn from the more economical wet-prucess orthophosphoric acid. In addition to these disadvantages. previous experience shows that iiiixtures produced from ~irthopliusphatebiise solutions usually cuntain considerably less plant nutrient than those produced from arntnoiiiuni polyphiisphate solutiotis. There still remains a slight amount of furnace super- phosphoric acid (76% P2 0, and 50% polyphospl~ate)used in the fluid fertilizer industry. However, most of this is used to produce spcci;ilty typc fertilizers by niixing the furnace pliospliiiric acid with p~it;issiuiii Iiydrvxide. iitnmvniil. and urea sulutioii. When this is dune, mixing is usually iiccutn- plished in a batch mix tank that is equipped with an aqu;t cii'. anliydrmts ainnionia sparger. Material is passed from this tank 111 a pipe-type heat exchanger. Couling is accon- plishcd by passing water across the tubes and liquid through the tubes. Most of the siilutiiiti iiiixturcs are prwluced by rcgiunal fertilizer plants which use wet-process supcrpliosplii,ric acid ;tnd the new TVA pipe reactor process tv priiduce Figure 25. TVA cuntinuous urea-ammonium aiiitiiiiiiiiini p~ilyphosphatebase si~lutioti irf gl-ade 10-34-0 nitrate solution plant or 11-37-0(3).

78 The superphosphoric acid is produced by concentrating mixing of the melt in the liquid and also provides an merchant-grade orthophosphoric acid (54% P20s) to excellent means of evaporating and superheating the anhy- S~~perpliosplioricacid in a vacuum or atmospheric concen- drous ammonia used in the pipe reactor. Data indicate that trator. Tlie superpliosplioric acid usually contains about plants with the separate mix tank usually produce an 6816 to 70% tutal P20, of wliicli 20% to 35% is present ainmoiiiuni polyphosphate liquid of slightly higher poly- as pulypliosphate. Physical and chemic;d characteristics of phosphate content than those with the combination mix a typical superpliosplioiic acid are shown in the following tank-cooler. Probably one reason this occurs is because of tabulation: the high temperature of the ammonia used in the pipe reactor. Animonia not added to the pipe reactor is usually Cliciiiical analysis. % by weight added as liquid ammonia to the niix tank. Usually about Total P, 0, 70.0 60% of the ainiiionia is added to the pipe reactor and 40% Pulypliosphate. % of tutal P, 0, 30.0 to the mix tank.This tank is equipped with a small scrubber Sulfate (SO,) 3.8 in which cooled 10-34-0 grade product is used to scrub Aluniinurn (AI,O,) 1.1 exit gases from the mix tank. Irun (Fe,O,) I .o Both plants have efficient, inexpensive evaporative Magnesium (MgO) 0.4 coolers. Very few contaminants are lost from either of the Fluorine (F) 0.27 plants. There is a considerable loss of water as steam and Solids (insol. iti CH,OH) 0.15 wilier vlrpor. Solids (insol. in H,O) 0.0 I Physical and chemical characteristics of a typical ammo- Specific gravity. 75°F (24OC) I .96 nium pulyphosphate solution of grade 10-34-0 made by this Viscosity. centipoises. 126°F (52°C) 400 process are tabulated below: Newly a11 the polyphosphoiic iicid in wet-process super- Chemical Analysis, % by weight phosphoric acid is in pyrupliosphste form; tlie remaining P20s is present as orthuphusphuric acid. Conversion to Nitrogen 10 superpliosplioric acid incieases the cost, but long shipping p2 os 34 distances genelate freight savings that reduce or eliminate Polyphosphatc, 16

79 Nw

80 component liquids that do not salt out down to 32°F (0°C) are 7-14-7, 5-10.10, and 10-20-5.These grades are made by the addition of KCI and water. The 8-8-8 grade is made by the addition of KCI, water, and urea ammonium nitrate solution.

Micronutrients in Solutions The need for micronutrients in the United States is in.creasing and the reasons for this increase are: I. Most cropped land is managed more intensively. 2. The shift to high-analysis fertilizers has reduced the ainount of micronutrients formerly applied as impurities in low-analysis fertilizers. 3. Better methods for identification of deficiencies are Figure 27. Liquid fertilizer mix plant now available. 4. New crop varieties have a higher nutrient require- ment. Deficiencies of one or more micronutrients are reported desirable. For example, if the superphosphoric acid is in most States. In some areas, there is little or no immediate derived from black acid, the product liquid will be black need for the application of micronutrients, but in other unless specially clarified. Also, if the starting acid has areas yields of some crops.are substantially reduced without relatively high impurity content, the product liquid will the addition of micronutrients, have limited storage stability. In view of these problems, Correction of micronutrient deficiencies is made by soil TVA has for several years been investigating methods for or foliar applications. Soil applications of inorganic iron either partially purifying wet-process acid or using other sources are usually ineffective; one or more foliar spray methods to produce satisfactory liquids from impure acid. applications may be required for severe iron chlorosis in One of these methods consists of using crystalline urea certain crops. Boron. copper, zinc, manganese, and phosphate [(NH2)2 CO.H,PO, ] as an intermediate. molybdenum are usually applied to the soil. However, In the first part uf the process, tlie crystalline urea molybdenuni is required in such small quantities.that it is phosphate intermediate is produced from merchant-grade often dusted on seeds. A survey conducted by TVA field (54% P,Os) wet-process phosphoric acid and urea in a representatives who have fertilizer development programs in twrrstage crystallization process. The crystalline urea 41 States indicates that most of the micronutrients now phosphate is of 17-440 grade and contains only 18% to being used are applied with a starter or broadcast fertilizer. 20% of the objectionable inipuiities (iron, aluminum, Since only small amounts of each micronutrient are magnesium, and fluorine) originally contained in the feed usually required, their application together with primary wet-process acid. Most of tlie impurities leave the process nutrient fertilizers is one of the most practical methods for in a by-product side stream of mother liquor that can be applying micronutrients to crops. It is difficult to obtain utilized satisfactorily for other purposes, such as produc- the uniforni distribution and application of a relatively tion of solid or suspension fertilizers. About 20% of the small quantity of a required micronutrient by means of a phosphate feed accompanies the impurities in this by- solid fertilizer mixture. On the other hand, the application product strcani. of micronutrients with fluid fertilizers appears to be the In the second part of the process, the crystalline urea most convenient method of uniformly applying slnall phosphate is iiielted at 210'F (99°C) in 11ne reactor by a quantities of micronutrients over a large area. TVA has combination of direct amnioniatiiin and external steain been conducting research to determine the reactions of heat. The melt then overflows to 2 second reactcir which various micronutrient sources and evaluate their agronomic operates at about 300'F (149OC) using external steam heat effectiveness. Some sources fol 111insoluble reaction products for conversion (if orthophosphate to polyphosphate to which are unavailable to platits. Other inicronutrient SOUIK~S produce about SO% polyphosphate level nielt. This nielt is then dissolved in water and ammoniated to a pH of 6.2 niay cause caking of dry fertilizers, or they will not dissolve to produce a 50% polypliosphate light amber-colored liquid in liquid fertilizers. While certain snurces (if micronutrients of 14-29-0 analysis. niay be applied with solids or fluids to result in products Potential uses of the 14-29-0 base solution iiiclude its of good physical properties, these sources may not be use in three-component liquids. Examples of these tliree- effective for plant growth. Tests have been conducted at

81 TVA and other institutions to determine which combina- Table 19. Summary of micronutrient content of tions are agronomically effective. Results of these tests have various inorganic micronutrient sources been published and reprints are available (6,7,8,9, 10, 11, Micronutrient 12). content, wt. Questions that uften arise concerning the addition uf Material %of element micronutrients tu fluid fertilizers are: Sources of boron 1. What are some of the sources of micronutrients? Sodium tetraborate pentaliydiate 2. How much of each micronutrient can be added to (Na,B,O,-SH,O) 14.7 liquids without crystals forming? Borax 3. What mixing procedure should be used in dissolving (Naz B407.10H, 0) 11.2 the micronutrient into the fluid mixture? Sodium tetraborate pentaborate 20.5 Sources of copper Sources of Micronutrients Cupric sulfate monohydrate (CuS04.H,0) 35.1 Table 19 shows some of the inorganic sources of micro- Cupric sulfate pentahydrate nutrients that are now available. It shows the micronutrient (CuSO4.5H2O) 25.5 content as percentage by weight of the element in each Cuprous uxide source. 020) 86.1 Table 20 is a tabulationof some of the chelated micro- Cupric oxide nutrient sources. The industiy was surveyed and a list was (CuO) 79.8 made of the chelated sources available; however, these Sources of iron sources have increased in number and at such a rapid rate Ferrous sulfate heptaliydrate that some of the newer sources of chelated micronutrients (FeS0,.7HzO) 20.1 qay liave been missed. The cost uf the micronutrient Ferrie sulfate nonahydrate element in must chelates is several times more than tlie cost [Fe2(SO,), .9H, 01 20.4 of the element in tile inorganic suurces. However, most Sources of manganese manuikturers of the chelates report that those sources are Manganous sulfate monohydrate about 5 to 10 times mure effective than the inorganic (MnS0,.H20) 24.7 sources. University personnel have repurted that chelated Manganuus carbonate micronutiients are as much as five times more effective (MnCOd 46 than inorganic sources of inicronutrients (13. 14. 15, 16). Manganous-manganic uxide The solubility of the inorganic sources of micrunutiients (Afn,O,) 69.0 -. --in-clear-liquid fertilizers varies-considerably.. B01:on :IS .borax Suurccs of.molyhdenum . is soluble in large enough quantities in liquid mixed fertilizer Sodium molybdate to meet most agrononiic needs. However, only a small (NI, Moo,) 46.6 quantity of copper and manganese c~nbe dissulved in Sodium molybdate dihydrate orthophosphate ~duti~n~.Larger quantities of micro- (Na>Moo, .2H, 0) 39.6 nutrients can be dissolved in animoniuni polyphusphatc Sources of zinc solutions such as 11-37-0 and 10-34-0 than ciin be dissolved Zinc carbonate in urthuphospliate ~olutiunssuch as 8-24-0. Laboratory (Zd03) 56 tests were made to determine tlie approximate mounts Calcined zinc cuncentiate of Iiiicronutrient materials th:it c:in be dissolved at ruuni (impure ZnO) 74 temperature in 11-37-0 made from electric-furnace plios- Zinc oxide phoric acid (17. 18). For curiip:~rison. similar tests were (ZnOI j, 80.3 made with 8-34-0 base solution prriduced from orthophos- Zinc sulfate monohydiate phoric acid. The amouiits of micronutrients that dissolved (ZnSO,.H,O) ____ after 24 hours of agitatioii (laboratory wrist-type shaker) are shuwn in ttible 31. Alsu shown are the estimated values that c311 be dissulved in this type of 10-34-0 is 75% of tlie for 10.34-0 sulutiun of average purity that is produced quantity that can be expected since part of the polyphus- frum wet-process superphosphoric acid. pliates in this type of 10-34-0 is required to sequester Most of tlie vdues for 10-34-0are based on aver,ige data (essentially dissolve) tlie impurities (Fe, AI, etc.) intro- frum several different cunipanies that use 10-34.0 mde duced by tlie wet-process phosphoric acid. from different wet-process superpliosplioric acids. Tlie These tests indicate that larger quantities of inorganic avcr~geresult indicates that tlie quantity of iiiicruiiutrients niicronutrients can be dissolved in ammonium polyphos-

82 Table 20. Micronutrient content of some 2. This 11-37-0 will dissolve up to 3% zinc as zinc organic___ micronutrient sources oxide. Micronutrient 3. The amount of zinc oxide that can be dissolved in content, wt. a 7-21-7 grade is then calculated by the following --______Material % of element- formula: Sources of copper 1,135 x0.03 x loo= 1.7%Zn Sequestrene Na, Cu copper chelate 13.0 2,000 HampEne 9% copper chelate 9 .O Danitra copper chelate 7.5 A. similar procedure can be used in determining the Rayplex Cu 6.7 amount of inorganic micronutrients that can be dissolved Sources of iron in mixtures produced from 10-34-0 made from wet-process Sequestrene Na Fe irun chelate 12.0 superphosphoric acid. Sequestrene 138Fe irun chelate 6.0 Some plant operators want to include more than one Sequestrene 330Fe iron chelate 10.0 micronutrient in a liquid mixture. An empirical type of Hamp-Ene 14% iron chelate 14.0 proccdurc has also been developed for the use of several Danitra iron chelate 10.5 micronutrients in mixtures produced from 11-37-0. This Rayplex Fe 9.6 procedure is as follows: Sources of manganese 1. Never exceed the maximum amount of each indi- Sequestrene Na,Mn manganese chelate 12.0 vidual micronutrient which can he dissolved in Hanrp-OI 9% manganese chelate 9.0 11-37-0 or 10-340. Danitra manganese chelate 10.0 2. The sum of all the micronutrients added should not Rayplex manganese 8.5 exceed 3% of the aniuunt of 11-37.0 used in the Sources of zinc mixture. Sequestrene Na,Zn zinc chelate 14.2 In most instances? this prucedure for the addition of Hamp-Ene 14.5% zinc chelate 14.5 several micronutrients tu mixtures produced from 11-37-0 Danitra zinc chelate 10.5 . has worked well. It should also work satisfactorily fur Rayplex zinc 10.0 mixtures produced From 10-34-0. An example of its use is Multiole sources of micronutrients as follows: Chelated nutraniin (Mn. Fe, Zn, Cu, B. and Mo) Assume that an 8-8-8-XZn-XB.XMn is to be produced. It Key-El (Zn. Mg. Cu, and Fe) is assumed the Zn:B ratio is 3, and the Zn:Mn ratio is 10. Multiplex (Zn,- Mn, Fe, Cu, and S) 1 First determine the inaxiinum quantity of the three micrunutrients that can be dissolved in 11-37.0. phate solutions than in orthophosphate sulutions. For Care should be taken to avoid exceeding the maxi- example, 60 times as much zinc as zinc oxide can be niuni amount of each individual micronutrient that dissolved in 11-37-0 arnnioniuni polyphospliate sulution as can be dissolved. Alsu the sum uf Zn, B. and Mn in 8-24-0 orthophosphate sulution. Other tests indicate should not exceed 3%. Calculations show that when that little magnesiuln can be dissolved satisfaJurily in tlie niicriinutrients are dissolved in 11-37-0. the either aninionium polypliosphate scilution or orthophus- resulting grade is a 10-34-0-2.0Zn-0.7B-0.2Mn. phate solution. Therefvre. the additiun of compounds that 2 About 471 pounds of r0-34-0-2.OZn-0.7B-0.2Mn contain magnesium to clear liquid fertilizers should bc would be required to produce tlie 8-8-8 liquid. avoided. Therefme, the Zn content of the 8-8-8 would be Most plant operators want to know how mucli of an calculated as follows: inorganic niicronutrient can be dissolved in niixtui~espro-

duced from 11-37-0 or 10.34-0. 11 would be difficult lo 471 x x 100 ~ 0.02 Zn determine this fur all cases. Plant and small-scale tests 2,000 indicate that in most instances it cui be calculated by the The boron content would be calculated as follows: use of the quantity of 11-37-0 or 10-34-0 in tlie forniula- (. tion and the quantity of the micronutrient compound tllat 471 ~0.007~100=0,17%B can be dissolved in the 11-37-0 or 10-34-0 base solution. 2.000 For example. the aniount of zinc as zinc oxide that can be The ni3nganese cwitcnt would be calculated as dissolved in a 7-21-7 produced froin I1-37-0,32-0-0potash, fdlows: and water is calculated as fiillows: I. The quantity of I 1-37-0 used to Q~V~UCCa ton of 471 x 0.002x 100 = 0.05% Mn 7-21-7 giwde is calculated to be 1.135 pounds. 2,000

83 Table 21. Solubility of micronutrients, 24-hour agitation I %by weight of element (Zn, Cu, Fe-LAP---- Mn B Ma) In 11-37-0 In 10-34@ In 8-24-0 Material added base solution base solution ____base solution 0.05 Zinc oxide 3.0 2.2s 0.05 Zinc sulfate 2.0 1 .30 Zinc carbonate [Zn(co,),l 3 .O 2.25 0.05 Cupric oxide (CuO) 0.7 0.53 0.03 Copper sulfate (CuSO,.SHzO) 1.5 0.14 0.13 Ferric sulfate [Fez(SO,), .9HZ01 1 .o 0.80 0.08 Manganous-manganic oxide (Mn304) 0.2 0.15 0.02 Sodium molybdate (Na,MoO,.2H, 0) 0.5 0.38 osob 0.9 0.90 0.90 Borax (Na? B40, .I OH, 0) ___ 3wadcfmn wct-proccss superphosphoric acid. b~arg~~tmount tested.

3. Therefore. tlne clear liquid 1:) :I ratio with Zn, B, popularity is its ease of handling and application. Also it and Mn would be an 8-8-8-0.47Zn-0.17B.0.05Mn. is safer to use tlne nitrogen solution in preference to anhy- Field storage and laboratory data indicate that this liquid drous or aqua ammonia. would store well for several days. However, when tniore Solution mixtures such as 7-21-7 which is a starter than one micronutrient is used in a liquid mixture, the fertilizer continue to be popular and as the farmer strives micronutrients sometimes react with each other and cause to increase his yields, specialized placement of fluids may an increase in the saltingout temperature with the result become niore important. For example, recently Kdnsas that crystals form in the tiiixture. For example, mixtures State University determined that there was some beneficial of 11.37-0, zinc oxide, and manganous-manganic oxide will effect of injecting solutions and anhydrous ammonia react to form insoluble compounds. Sometimes these siiiiultaneously beneath tlie surface of the ground. With reactions take place over a long period of time. but often this type uperatioii, ii trouble-free solution type fertilizer they occur in :I very short period. Also, crystals iiften form is required. Tlie use of nitrogen solution and solution in liquids containing any micronutrients if the liquid is iiiixtures in special applications such as through drip irriga- stored for a prolonged period. For tliis reason in the TVA tion units and sprinkler irrigation systems is gaining in field programs. the storage of liquid mixtures that contain popularity. In the colder regions of the United States. the - .- . micronutrients is nut.reconimended. . - ~use-~ofii starter- .type. -. . fertilizer.. to promote early growth In some instances, the inorganic micronutrients are !not and crop maturity is probably beneficid tb the XiriiZr, sufficiently soluble in clear liquid mixtures to meet agro- Generally. if fluid fertilizer is to be applied by the farmer. nomic requirements where genuine deficiencies exist. it is preferirble to use solution instead of suspension fertil- The solubility of chelated micronutrients is usually not iz,ers because of their trouble-free quality. For these and a problem. Only one-fiflli to one-tenth :is much niicro- other reasons it is expected that the use of nonpressure nutrient element is required as a chelate compared with an nitrogen solutions and solution mixtures will continue at inorganic source. Therefore, the amount of chelated metal increased levels. elements soluble in liquid mixtures produced from 10-34-0 or 11-37-0 is generally sufficient to meet recommended agronomic requirements.

Future Solutions

In recent years. most of tlie growth in the fluid fertilizer industry has been in the pruduction of suspension fertilizers, However, tlie solution fertilizer market is still considerably greater than that of suspension fertilizers. Tine growth of nitrogen solntiiins has been substantially higlier than the growth rate for :inhydrous and aqua amnii~niii.and it is expected that larger quantities of nitrogen solution will be iiscd in the iut~re.The tixiin reason fix tliis increase in

84

L I 12. Shepherd. I.. K. Lawton, and J. 1:. Davis. "The Effcctivcncss of Various Manganese Materials in Supplying Mangancrc to Crops." Soil Sciencc Socier,y oJA,nerico Proceediqs. Vol. 24, pp 218-21 11960). 13. Uinande, R.. B. Kncrck, J. Davis, E. Dollan, and J. Meltan, "l'icld and Laboratory Studies with Zinc and Iron I:ertilizatiun of I'm. Llcans, Corn, and Potatucs in 1967," Michigan Agri- cultural Experiment Station. Michifan State University, East Lansing, Michigan. 14. A~rommyTips. Extension Scrvice, Univcisity of Nebraska, Lincoln. Nchraska. M;irch 25. 1964. 15. Chcsnin, Lcon, Sircccssfirl l:om!in~(Septcmhcr 1965). 16. Lindsay. Willard L.. and S. Dalc Ramrdal. Ayrichernlcol IVesr (Aiigost 1966). 17. Slack. A. V., J. M. Potts. and H. B. Shaffccr, Jr.. "Effect of Polyphosphate Contcnt on Propertics and Uic of Liquid I'crtilizcn.'' Agriadtwol mid Food Cliernlsrry. 13, (2) 165-71 (March-April 1965). 18. Tcnncsrce Vnllcy Authority, L)eveIop,rimts i,r Fwtiliier 7"dziioIogy. fifth dcmonstration. October 6-7. 1964.

c

85 Chapter 7 Suspension Fertilizers- Production and Use By J. G. Getsinger, F. P. Achorn. and George Hoffmeister

Suspension fertilizers are characterized as fluid fertilizers trations of fertilizer solutions with resultant high cost of containing, a suspending agent such as a gelling-type clay. handling, storage, and application. Suspensions usually contain finely divided solids that are The relative purity or impurity of tlie raw materials is held in suspension by the gel. The formulation and nianu- one of the most important considerations in obtaining a facture of a suspension need to be controlled carefully so satisractory sulution. In addition, the clinice of the anions that the gel is strong enough to support the solids without ;ind cations in the solution will affect solubility treinen- their settling but weak enough to allow the suspension to dously. The coiiiiiioti cations are ammonium and potassium. be easily agitared, poured, or pumped. The usual anions are phosphate. nitrate, chloride. and SUI- Suspension fertilizers apparently had their origin in the fate. The phosphate anion is complicated by the solubility- late 1950s (1, 2, 3). Because of the need for including boosting attribute of the polyphosphate form of phosphate. potassium in fluid fertilizers beyond the limit of the solu- When consideration is given Yo impurities, polyphosphates, hilitf of poCassiUrn, growth of suspensions acceleEted (4): salting-out temperatures. and the nse iiFvariuus raw inate: Fluid fertilizers containing nitrogen and phosphate as rials. solution fertilizers are seen to be quite complex aiiiinotiiutii polyphosphates have respectably high soluble Fertilizer materials. concentrations of these twv plant fwds, but potassium The development and use of suspensions was niutivaied salts are less soluble. Therefore, increased need for ii suspen- by the disadvantages and limitations of solutions outlined sion was identified. Other advantages of suspensions in above. Some advantages of suspensions are: comparison with solutions soon became apparent. 1. Less pure and therefore cheaper raw niaterials can A solution is an ideal form of fluid fertilizer for reliable be used in making suspensions. handling, measuring. and applying to the suil. A solution is 2. Solubility of niaterials is inlit a limiting factor; uniform atid hiimogenous. It has a hiw viscosity and does therefore, the concentrations can be higher and tlie not require mixing. It is usually transparent so that any ratios of plant nutrients arc allnust unlimited. stray foreign material can be easily seen. It does not settle 3. Additional plant nutrients, such as micronutrients, in tanks ur clog pipes or iiozzles, and can be gravity fed can be added to suspensions in the quantities in by nieiisuienient throngh a sitnple orifice. Hiiwevcr. sohi- proportions desired without concern for their tiiins have certain characteristics that endow tlieni with solubility (5). disadvantages. Siilutions have solubility limits which dictate 4. Furtherni(ire, less costly forrnsofthe micronutrients, ihe co~~~entraiiiitiarid pmprirtion of mch iif the plant fried such as oxides rather than the niorc soluble sulfates, coinponcnts. These liiiiits dictiite the grate (if the solutiiins. Furtherliiore. since solubility depends oii tempei~ature.the can be used in suspensions, grade tnlist be low enongli to witlist;ind the lowest expected 5. Pesticides are miire compatible with suspcnsiiins temperature without salting out any (if the soluble tiiaterials. tliiin with sdutiiins (6). These rcquireinenls sometimes tiiake for very low co~~celi- Some of the disndvantages (if suspensions are:

86 They contain solids which can settle, clog nozzles, immersed into the sample at a prescribed depth. The cup and cause stoppages in pipes. is rotated by a drive motor at a constant rate. and initially They are more viscous than solutions. thus viscosity the bob rotates at the same speed as the cup, since there is sometimes beconiesa limiting factor in their prepara- essentially no force resisting the bob rotation; however, as tion. the cup and bob rotate further, the wire, which is fixed at Suspensions require periodic agitation during the top end, is twisted and exerts a torque against the gel. storage. When the torque reaches the point at which the gel is The processes for making suspensions require a broken. the rate of rotation of the bob decreases. The clay addition step with the attendant equipment for angle of deflection (amount of rotation) at that point handling and mixing the clay. indicates the torque being exerted by the wire. Thus, the torque is a measure of the gel strength and is directly proportional to the diameter and inversely proportional BASIC STUDIES to the length of the torsion wire. The wire is calibrated to measure torque in gram-centimeters, and the gel strength Measurement of Gel Strength is expressed in these units. Gel strength vahtes determined in this manner can be related to other observations of the Since a suspension is a gel and the ability to suspend quality of the suspensions, and relative gel strengths give an depends on gel strength. research on suspension fertilizers important, fast determination of the suspending character- required sonie tiieans of measuring gel strength. Therefore, istics of a particular suspensiun. Generally, a gel strength a gelotiieter was developed at TVA (7) (figure 28). of about 1 to 2 gram-centimeters is adequate for most A torsion wire suspends a bob, and a turntable supports suspensions. It is more difficult to define a gel strength that a cup that contains a suspension fertilizer. To make a gel is too high. Usually the viscosity would be the limiting strength nieasurement, the bob is lowered iiitu the cup and Factor on a suspension; the viscosity and the gel strength increase together.

Gelling with Attapulgite Clay Usually, tlie suspending agent used in suspension feitilizers is attapulgite clay obtained in south Georgia and north Florida (8, 9). Attapulgite is a hydrated magnesium aluiiiiiiuni silicate. The clay is composed of needle-shaped crystals that are so small they can be seen unly by great Iiiagnitication (8). 111 the dry state, these crystals bond together in bundles as larger particles. These bundles must be disrupted and the individual crystals freed before they can assunie tlie structural arrangement that is character- istic of a gel. When this disruption of the crystal bundles is carried out in the presence of an electrolyte, such as any of the cmmion fertilizei salts present in fluid fertilizers, the individual crystals immediately assunie the regular structural orientation uf a gel. Wetting of the chy by the fluid fertilizer causes some of the required disruption vf crystal bundles and freeing of individual crystals, but not enough to provide good gelling. Research has shown that vnly vigorous shear-type agitation vf the clay-fertilizer suspension can provide the desired degree of disruption and. hence. the desired gel strength This type of stirring requires high tip speed ofthe agitator. One of the niost effective shear-type mixers is a centrifugal pump. Tip speeds vf impellers of centrifugal pumpsusuallyr~tigefroniabout 50 to 80 feet (15.2.24.4 111)

~~ per second. For this purpose. the liigher the speed, tlie better the results. Tu obtain complete gelling. ilultiple Figure 28. Basic parts of TVA gelometer passes of tlie suspension through the pump are required.

87 Therefore, the pump should be large enough to avoid suspension will be difficult to remove from storage tanks. excessive time for gelling. The number of passes required is To prevent excessive increase in gel strength. the suspension related to the speed of the impeller-the slower the inipelkr must be periodically sparged or otherwise agitated in the speed, the more passes needed for complete gelling. The storage tank, as will he discussed later. practical range is about 10 to 20 passes. The amount of shear required depends on several factors. For example, if the fluid contains nitrates or polyphosphates, the amount Fluid Clay of shear required is increased. Also, if the fluid is a liquid with no solids present, the anioutlt of shear required is higher than it would be for a fluid that contained solids. Since dry clay is difficult to gel it1 some suspensions, The action of the solids in increasing the effectiveness of dispersed clay is marketed to improve the gelling of the clay the mixing and the resulting higher gel strength is not (1 I). This fluid clay also offers the advantage of nondusti- understood. However, the presence of solids, such as crystals ness in handling. TVA has developed a fluid clay containing of urea, ammonium nitrate, or ammonium phosphate, urea (12). The dispersing agent is TSPP. The urea serves evidently overcomes much of tlie resistance to gelling two principal purposes: to lower the freezing point of imparted by the anions of nitrate and polyphosphate. the liquid phase to about 15°F (-9.4OC) and to reduce the The efficiency of gelling dry clay in a fluid can be initial viscosity of the product. The clay content is about increased by increasing the cloy content in a portion of the 25% and the nitrogen content from the urea is about 9%. fluid during the gelling step and theti mixing the concen- The fluid clay has two serious disadvantages. One is that trated gel with the remainder of the fluid to obtain the viscosity increases with storage time, particularly in hut desired clay content. For example. for a product with 2% weather and without agitation, such as the conditions that clay, the entireamount ofclay could be added to one-fourth occur in a tank car in the summei. The other disadvantage the fluid for gelling, giving a clay content during the gelling is the higher ireight cust as compared with shipping dry step uf 8%. After gelling, the concentrated gel then would clay. Practical coutitertneasures to these disadvantages he mixed with the other three-fourths of the fluid tu obtain might be to produce the fluid clay only in cool weather and a final product with 2% clay. Gelling of the clay by this near the point of use. procedure allows more efficient use uf the agitator which Research has indicated that hydrolysis of the urea and works un a smaller volume of material. It increases the TSPP C~US~Sthe increase in viscosity. The formation (if viscosity and, therefore, increases the shearing action of the amtiiotiiuni carbonate and orthophosphate, both gelling agitator during tlie gelling step. It increases the proportion agents, partidly gels the concentrated clay and causes of solids, which increases the gelling action as previously thickening. Leaving urea out uf the formulatioti dui~inghut described. weathei probably would he advisable. - - - .- -Another method to facililntc- gclling- of- clay in fluid...... ~. fertilizers is to clien&illy disperse the clay (IO). either in water or in urea or aniiiionia solutions, before adding the Use of Other Suspending Agents clay to the fluid fertilizer. Clay. however. cannot he dis- persed in the presence of fertilizer. salts, and when the dispersed clay is added to tlie fluid 'fertilizer. the salts Sepiulite clay (13. 14). which is mined in Nevada. is cause immediatc gelling of the clay. Chy conceiiti~ationsin similar 111 attapulgite in crystalline structure but contains . a dispersion caii be as high as 25% or 30% with viscosities otrly about cine-tentli (11. the amount of aluminum typically low enough to allow easy Iiatrdling with good flowability. found iti att:ipulgite. Sepiiilite is a hydrated magnesium One of tlie iiiure effective dispersants is tetrasodium pyrw silicate. It furnis a gel in suspensi~msin the samc manner as phosphate (TSPP). The aniuunt uf this dispersant required attapulgite. Sepiolite is iiioic reactive tlian attapulgite with fur dispersing clay in water is 1.25% of the weight uf clay. the slightly acidic liqiiid phase of fertilizer suspensions ;~nd Clay also can be pregelled iii wiilcr without ii dispersing this reaction weakens the gel with lime. agent: however. a clay c~iiicetitriitiun(if about 10% is the Beiitonitc is a clay fuuird in Wyoming and ncighburing limit because of excessive thickness OI~viscosity of greatcr St;itcs- (15). It is primarily a hydrated aluniinuni silicate. coticentraliuns. Tlic use of ii dispersing agent luwcrs the bur its swelling and gelling characteristics when wetted with viscosity mid. therefore. allows higher clay coiiceiitratiiiiis w:iter cumc from the sinal1 amount of sodium it contains. in clay dispcrsions (see Fluid Clay section). Beiitoiiite c3tiiiot be gelled when it is added dry to a Another characteristic of mist suspension fertilizers ciiiiccntratcd fcrtilim ionic solution. Bcntonitc must be is tliiit when suspensims are storcd witliiiiit agitatirin. the gcllcd iii water or noiiioiiic s~rlutiu~ibefurc it is added to gel strenglli of the suspension increases. The gel cmi becmie the suspension. A clay cuiirent of 10% in water is about the so strong that the suspension will no Ioiiger flow atid the tiiaxiinuiii that can be rc;idily handled. PRODUCTION OF SUSPENSIONS to hold the boiling point at about 230'F (110'C). The degree of ammoniation is controlled by measuring the pH Phosphate Base Suspensions of samples of the contents diluted with about four parts of water to one part of material and holding the pH at One of the earliest phosphate base fluids for suspensions about 5.4. The reactants overflow to the second-stage Was 10-34-0 ammonium polyphosphate solution (16). reactor where more ammonia is added. Also, a slight The solution was used together with potassium chloride amount of water is added in this reactor to adjust the to make NPK suspensions; clay was added as a suspending concentration. The temperature in this second.stage reactor agent. TVA developed the concept of .a phosphate base is held at about 200'F (93OC) by recycling a cooled stream suspension containing its own clay. At first, a suspension of of material back to the second stage as necessary. The pH this type with a grade of 12-40-0 was made from ammonium of the material in this stage (sample diluted 4 to 1 with polyphosphate (APP) solution made from electric-furnace water) is held at about 6.5. The effluent from the second phosphoric acid (17). Later the grade was increased to stage is pumped to an evaporative cooler where the product 13-41-0. and the suspension contained crystals of diamrno- is cooled with air to about 125'F (52'C). Effluent from nium phosphate (DAP) (18). Then a phosphate base suspen- thc cooler flows by gravity to a third-stage leactui where sion imide from 3 mixture of electric-furnace and wet- additional ammonia is added. Cooling is provided in the process phosphoric acid was produced. The grade of this third stage to hold the temperature at about 120'F (49°C). material w~s13.39-0 (19). Material flows from the third stage to a clay-mix tank Amnioriiuni Orthophosphate Base Suspensiotl, 13-38-0- where attapulgite clay is added and gelled with a reciicu- More recently. a process was developed for making a base lation pump. The mixture is further cooled with cooling suspension entirely from wet-process phosphoric acid and water to about 1 15'F (46OC). The pH of the product amnioniii (20). A flow diagram of the process is shown in (measured after 4 to 1 dilution with water) is kept at about figure 29. The feed acid is nlerchant-grade with 54% P2 O,, 7.1. The product is then pumped to storage. but it can be as low in concentration as 40% P20,. The The purpose of the three-stage animoniation is to product has a grade of 13-38-0 and contains 1.5% clay. make a concentrated suspension with satisfactory physical Phosphoric acid and gaseous ammonia are fed to a first- properties. The relatively hig!i temperature and low pH stage reactor. The heat of reaction causes the contents to conditions in the first stage promote the precipitation of boil. Water is added to the first-stage reactor as necessary the impurities. principally iron and aluminum, in a form

WATER VAPOR

PHOSPHO ACID

RODUT pH 6.5 TO TORAGE GELLING "&~','" 13-38-0 1.5% CLAY 115-F. PH 7.1

Figure 29. Production of 13-38-0 ammonium orthophosphate suspension fertilizer

89 that is crystalline rather than amorphous so that the final by diluting with water in the storage tank to 36% P,O,. product viscosity will he lower. The temperature is lowered Also, the solidification temperature can be lowered to about in the second stage so that the ammonia loss will be mini- -5'F (-51OC) by adding some ammonium polyphosphate mized. In the third stage, the monoammonium phosphate solution to provide about 12% of the P,05 as polyphos- (MAP) crystals are converted to DAP. The DAP crystals phate. The addition of a small amount of sulfate (4%) also are small enough and have a shape that inhibits their lowers the solidification temperature. settling. About 200,000 tons of satisfactory 13-38-0 materid In preliminary operation of this plant without a third has been shipped by TVA and used mostly in the south ammoniation step, the product (1 1-39-0) contained crystals eastern States. as MAP, and these rodlike crystals settled excessively in Polyphosphate Base Suspension, 9-32-0-A process railway tank cars during transport. Further research and employing a pipe reactor was developed at TVA for pro- development work on the process. including a test developed ducing APP suspension from merchant-grade wet-process to simulate vibration received by the products in a railway phosphoric acid, ammonia, and suspending clay (21). tank car, was necessary to develop the modification The ubjectives in this process were threefold: (I) that involving the third-stage formation of DAP. the product would contain polyphosphate so that it would The product is higher in concentration than 10-34-0 withstand cold-weather storage and handle satisfactorily, base solution and, therefore, can be stored and shipped at (2) that the product would be made from merchant-grade costs helow those for 10-34-0. The cost of production of phosphoric acid rather than superphosphoric acid. and 40% to 54% acid, which is suitable for feed to this process, (3) that existing pipe-reactor plants producing 10-34-0 solution fertilizer from superphosphoric acid could be is less than that of superphosphoric acid required for the modified tu produce 9-32.0 suspension in most of the same production of 10-34-0. Furthermore, less energy is required equipment. The process as developed is shown in figure 30. for production of the acids of lower concentration. The The melt dissolution tank and the evaporative cooler 13-38-0 suspension carries its own clay, which is useful for are usually present in most 10-34-0 solution plants. The blending with other materials for making suspensions with- equipment for heating the acidand ammonia and for adding out having to add clay at the mixer-dealer level. clay would need to be provided. The key feature.of this The primary problem with 13-38-0 suspension is that process is tu provide a suspension containing about 25% of it solidifies at about 15" to 20°F (-9.4O to -6.7"C). the P,05 as polyphospliate without outside heating of the Thickening at low temperatures cm bc reduced satisfactorily feed acid or the aninionia. The feed materials are heated by

VENT I

...... - ......

WATER ATTAPULGITE

ACID HEATER MERCHANT-GRADE WET-PROCESS ACI PRODUCT -_- - TANK TO LlOUl STORAGE AMMON 9-32-0 2% CLAY VAPORIZER M Figure 30. Production of 9-32-0ammonium polyphosphate suspension fertilizer from merchant-grade wet-process phosphoric acid

90 heat exchangers where liot solution from the dissolution The phosphate base made in this manner does not have very tank is tlie heating inedium. All of the heat conies from the good storage characteristics in cold weather. reaction uf pliosphuric acid and ammonia. The melt dis- A polyphosphate content of about 10% of the P,O, in Solution tank is kept at about 200°F (93OC) by recycling the solid increases its solubility, shortens the ammoniation cool streaiii from the evaporative cooler. The acid reaches time, and improves the storage characteristics of the base a temperature slightly below 200'F (93'C), and the suspension greatly (24). Also, the grade of the resulting ammonia a temperature of about 165OF (74°C). Hot solu- base suspension can be increased satisfactorily to about tion is pumped from the iiielt dissolution tank to a packed 11-33-0or 12-36-0, depending upon the amount ofimpuri- tower evaporative cooler, where it is cooled to about tics in the phosphate and the proportion of the P20, IOO'F (38°C). Part of the cool stream is recycled to the present as polyphosphate. A process is under development melt dissolution tank to maintain the temperature at about that allows tlie production of polyphosphate in excess of lOOoF (93OC). This is about tlie highest temperature that 10% in the solid product by use of a pipe reactor dis- can be maintained without significant loss of aninionia charging into a granulating drum. This product is expected and witlinut excessive hydrolysis of polyphosphate during to be superior to MAP for use in making suspensions. the time the material is held iii the dissolution tank. The temperature of the s~lutionis desired to be as high as is Nitrogen Base Suspensions feasible in order tu provide ii iiiaxiiiiuin amount of heat for [lie acid and ;imiiionia. When nitrogen base solution is mixed with a phosphate Cool solution from the evaporative cooler is pumped to base suspension, the clay content is diluted and extra clay the clay niixing tank where dry attapulgite clay is added is needed fur adequate suspending action. The gelling of tu give a clay content of about 2%. A recirculation pimp is clay in UAN solution to produce a nitrogen base suspension used for gelling tlie clay in tlie solution. Tlie product with is a very difficult operation. The strong gelling action of a grade of about 9-32-0 is then pumped to storage. ammonium nitrate interferes with the sliearing action of The primary use oftliis product is as a substitute for the the agitator required for separating the particles of dry clay. more expensive 10-34-0APP solution. The 9-32-0 suspeii- A nitrogen base suspension carrying its own well-gelled sion can be used for direct application to the soil or for clay for blending with phosphate suspensions was needed. blending with other niaterials to nuke niixed suspensions. Therefore. TVA has developed processes for producing A further modification uf this process envisages the use nitrogen suspensions. of sulfuric acid togetlier with tlie phosphoric acid to supply Urea-Animoniuni Nitrate Suspensioti (31% Nitrogen- additional heat in tlie pipe tu increase tlie iiielt temperature A flow diagram of a process for producing a UAN suspen- and increase the polyphosphate formation. Wheir phosphoric sioii containing 31% nitrogen is shown in figure 31 (25). acid done is used. the aiiiouiit of polyphosphate formed is sliarply affected by tlie aiiiount of wilier in the phosphoric acid: therefiire. tlie use of sulfiiric acid in conjunction with phosphoric acid would coinpensate for acids with higher water content. This modification is under pilot-plant study IIUW. Susperisior~s frorii Solid Phosphates-In some areas. freight rates for fluid fertilizers are higher than for solid fertilizers. The distance of shipping of fluid fertilizer is limited by economics (22). Therefore. many mixer-dealers use solid fertilizers. such 3s MAP, fur producing their own pliospliate base suspension (23). The usual procedure is to use a batch fertilizer mixer. The water of formulation is first added to the mixer and then ammonia is started. The MAP is fed to tlie mixer and ammoniation is conducted as rapidly as possible to elevate the temperature of the material. The combin~tionof reactioli with the ammonia and tlie heat of imniioiiiatiiin rapidly dissolves MAP in tlie water. A glade of abwt 10-30-0 is produced by this iiieaiis. Usu;illy. urea-aiiiiii(iiiiiiiii iiitntc (UAN) solution. potassium chloride, :ind clay iirc added to complete tlie desired fertilizer mix- ture. llowevcr. ii plivsp1i:ite basc c:in be m:ide by adding Figure 31. Production of urea-ammonium and gelling tlie clay and sending tlie 10-30-0 to stnr~ge. nitrate suspension fertilizer (31% N) Attapulgite clay is dispersed in urea solution by the use ammonium nitrate solution (88%) as feed for the process. of a dispersing agent, TSPP. This dispersing step is a neces- A product containing about 55% urea, 31% ammonium sary prerequisite to effective gelling of the clay in the UAN nitrate, and 13% water has~aviscosity of about 750 cl' solution. Urea solution is used in the dispersing step, and at 32°F (O'C). The solidification temperature is about the clay content is made as high as possible (30%) to reduce -15°F (-26OC). These characteristics would allow this the water content to the lowest possible level. By minimizing product to he stored and handled at extremely cold tem- the water content, the concentration of UAN solution peratures. (32% nitrogen) will be diluted as little as possible when the , A flow diagram of the process is shown in figure 32. clay is added. The concentrated clay dispersion is fed to The concentrated feed materials are mixed and sent to a a mixing tank together with UAN solution for the gelling water-cooled heat exchanger where the material is cooled step. Gelling is accomplished with a centrifugal pump. to about 135'F (57°C) which is above the silturation About five passes through the pump on the average are temperature of about 110°F (43OC). This temperature limit required to gel this concentrated dispersion in UAN solu- is to prevent fouling of the heat exchanger by crystalline tion. Using this system, the product suspension with 2% urea. The solution is pumped to a clay-mix tank where clay has a nitrogen content of 31%. predispersed clay is added. Then the slurry is pumped to a Higher Grade Urea-Ammonium Nitrote Suspension (36% spray cooling tower where air is used to cool the iiiaterial Nitrogen)-The development of a process for producing a to cause rapid nucleation of urea crystals so that small 36% nitrogen suspension with 1.5% clay from urea and crystals will be furnied. The cooling is by combination of ammonium nitrate is underway. Such a product would evaporation and exchange of sensible heat. Very small lower the freight and handling COS~Sand allow storage of urea crystals of a few hundred micrometers in size are more plant fond in the same storage tanks. The process desired to prevent growth of crystals to excessive size would require concentrated urea solution (87%) and during storage.

88 % AMMONIUM NITRATE SOLUTION-^ 87% UREA SOLUTION

UAN MIXER - - - - - __ - . -_ 25OoF-C~~- WATER

XK Figure 32.~ProdE7ioiqurea-ammonium nitrate suspension fertilizer (36%N) Urea-Ammonium Sulfate Suspension-Some soils are at 32°F (0°C) is about 1,200 centipoises and the solidifi- deficient in sulfur, and fertilizers containing sulfur are cation temperature is about -15'F (-26°C). needed for those soils. Since the application of nitrogen The product can he used for direct application to replace can cause sulfur stress in some crops on soils low in sulfur, nitrogen solutions where sulfur is needed. Also, it can he combining sulfur with nitrogen in fertilizer appears to be used in mixed suspensions to provide nitrogen and sulfur desirable. Therefore, TVA developed a process for produc- where sulfur is desired. ing a urea-ammonium sulfate (UAS) base suspension from urea solution, sulfuric acid, and ammonia (26). A flow Potash Base Suspensions diagram of the process is shown in figure 33. Urea solution, sulfuric acid, and ammonia are fed to.a Since potassium chloride is the most economical form of reactor. The heat of reaction of the sulfuric acid and potassium, it is the material commonly used for furnishing ammonia causes the reactor to boil at about 240°F (1 16OC). potassium in suspension fertilizers. Solution-grade potash is The pH is maintained at about 6.3. Water is added to replace usually used. This material is -20 mesh in particle size and water evaporated during the reaction step. Dilute, sperit, or contains 62% K20. by-product sulfuric acid may be used, or concentratcd acid Potassium chloride is itxeived in the mixing plants as is also satisfactory. The reaction product overflows to a solid material and, in many instances, is batch-mixed with surge tank from which it is pumped to a water-cooled heat the phosphate and nitrogen materials to give a complete exchanger. Here it is cooled to about 135'F (57°C) which three-component mixed fertilizer. However, a potash base is just above its saturation temperature. This temperature suspension with a grade of 4-14-28 or 4-12-24 can be made limit is necessary lo prevent fouling of the heat exchanger. by mixing solution-grade potassium chloride in a phosphate From the cooler, the solution goes to a clay-mixing step base suspension, such as 13-38-0, or a phosphate base soh- where 2% clay is added and then to an evaporative spray tion such as 10-34-0 with the addition of attapulgite clay. cooling tower. The product is 29-0-0-5s. It has excellent The necessary water is added to complete the formulation. cold-weather storage and handling properties. Its viscosity Agitation of the mixture, after addition of the potassium

93% SULFURIC AC

WATER -87% UREA SOLUTION

AMMONIA 1 AIR

,

CC AY REACTOR - - 240.F pH 6.3 -- / - lo, CLAY TANKMIX - -d"* ~ I35'F PRODUCT TO STORAGE -29-0-0-5s 2% CLAY

Figure 33. Production of urea-ammonium sulfate suspension (294-0-55) chloride fur IO to 15 minutes, is required in order to form a I. There is no dust during handling or spreading. stable suspension. The reason for this required agitation 2. Very finely ground limestone can be used which time is not understood but is believed to be associated with promotes rapid reaction in the soil. the chemical reactions that occur between potassium 3. Suspension fertilizer manufacturers with equipment chloride and the other components. fur broadcasting suspension fertilizers can use the The advantages of making a potash base suspensioii are: same equipment to apply limestone thereby (I) the potash can be mixed with the other components elimiirating the need for extra dry spreading during slack times when the necessary time of agitation for equipment. adding the potash can be most conveniently expended; A sketch of a typical plant that is used to produce fluid (2) the potash base suspension can be blended with nitrogen limestone as well as suspension mixtures is shown in figure and phosphate base suspensions with simple, low-powered 34. This plant does an excellent job of producing suspen- mixing equipment to obtain plant nutrients in any desired sions in 5- tu IO-ton batches. Finely ground limestone is ratios without having to add and gel clay;and (3) the saving stored arid added to the mix tank in batches. In other of the time that would be required for mixing each batch plants conventional solids handling equipment sucli as during the busy application season. bucket elevators, conveyor belts, drag chains, and augers A potash base suspension is the only feasible fluid base are used to convey the powdered limestone. None of these for potash since the solubility of potassiurri salts is so low. plants are capable of efficiently handling finely ground limestone; therefore. the companies resorted to the design Fluid Lime shown in figure 34 which uses a pneumatic handling system lor transporting and handling this powdered material. Many crop growing areas have low pH soils tliat must be Sonic companies transfer the limestone from dry storage adjusted for optimum crop response. For years this pI-1 tanks to the mix tank through conveyors as shown in adjustment has been accomplished by spreading dry ground figure 34. Others mount an eductor at the discharge of the limestone. Most agriculturalists recommend that all particles dry storage tank and add the limestone to a recirculating in tlie limestone should be less than 10-mesh in size (0.328 fluid line ufthemix tank. Fluid passing through the eductor in.) and that at least 60% should be smaller than 100mesh causes a negative pressure tu develop in the intake of the (0.0058 in.). This is a relatively small size and there is educlor so that thc powdered limestone can be drawn into significant dust generation as it is applied. Generally, most the tluid. agriculturdists agree that the finer the limestone grind tlie The formula fur producing a IO-ton batch of 50% quicker the pH adjustment can be made in the soil. There- liniesloiie mixture is sliown below. fore, it would be desirable to have the particle sizes less than 100-mesh in size. However, because ofthe fineness of Materialsa Lb/batch - - .-.this grind..and.the.dust loss of.limestrmc. as .it is-~applied,it Water - .~ .~ .~ ~. 9,600 is impractical to apply such finely ground limestone in the Attapulgite clay b 400 dry form. In the 1960s, TVA demonstrated that it was Limestone 10,000 possible to produce a suspension which contained 50% limestone, and that this suspension could be easily applied Total 20,000 through flooding type nozzles (27). The concept was ‘Ustwl order of addition. accepted in tlie 1970s and at this time it significant number bl:ivc to 10 Ib of UAN sulution addcd aids in gelling thc clay. of companies offered fluid limestone to their custoniers. Since then. the use of fluid limestone has been gaining in The usual procedure is to add water tu the mixer. add and popihrity. mix the clay, add the limestone, llicn agitate the mixture Most fluid fertilizer manufacturers are interested in until the limestone is thoroughly suspended in the fluid. fluid limestone as opposed to application (if solid limestone Plant and small-scale tests have shown that attapulgite clay because they have suitable applicators for the application uf must be properly gelled to provide enough suspending this type nf suspension and additional dry type applicators strength to keep the limcstone from settling in storage. are mt required. Also. some farmers are particularly Sometimes the operators add 2 to 5 pounds (0.9 to 2.1 kg) interested in iibtaining very rapid pH adjustriienls of the of ammonium nitrate per ton of limestone after the clay is soils on which they plan til grow ii crop. Usually these are dispersed in water tu help gel tlie clay. Aninionium nitrate fmiiers that ilre using rented land or land which they plan is usually added as solution that contains 18% N or UAN to farm for only one seasmi. TVA application demonstra- solution that contains 31% N. Some of the uperators also tions show tlie following advantages for limestune have fuund it advisable to add about 5 to IO pounds suspension (18): (2.7 to 4.5 kg) of dispersant such iis TSPP to the water

I 94 Figure 34. Mix plant for fluid limestone

prior to adding the gelling clay. With these methods, suspen- use of this material. One problem in using calcium oxide sions of 50% or higher liniestone content can be produced. is that it reacts with the water to create some heat which Most dry agricultural liniestone has too large a particle causes the lime suspension to become thick; therefore, size to produce suspensions that will not settle rapidly. dilute lime suspensions must be produced. However, some This is because in many instances conventional agricultural are experimenting with the addition of wetling agents such limestone contains a significant aiiiount of +10 mesh as lignosulfonate and molasses to suspensions made from material which cannot be satisfactorily suspended in "kiln dust." the fluid. For this reason a standard limestone will not Based on these experiences, niost of the deders have make a good suspension. The limestone must be ground so fiiund it advisable to pruduce limestone suspensions fmni that all oftlie particlesare -35 mesh and most have a particle finely ground limestone. Some are mixing the suspension size that will pass through a 100-mesh screen. Many quarries with a phosphate base suspension and applying it so~inafter now have the grinding capability to supply liniestone of it is made. Tests liiive sliown that when this is done imme- this particle size; however. it is usually at a higher price diate application is necessary to avoid reversion of the than standard agricultural limestone. For this reason. inany P2 O5 in the suspension mixture. Others have found that dealers are investigating the possibility of using by-product they can readily mix the finely ground limestone with UAN limestones and other by-product materials. They have had solution and a gelling clay to produce a suspension that excellent results in using limestone dust from air filters that contains 14%-15% nitrogen and SWh limestone, These have been installed in dust removal systems at the quarries. types of nitrogen-limcst~)ne suspensions are gaining in Others have attempted to use by-product calcium oxide popularity since they can help to eliniiwte sonic passes of from cement kilns and have had moderate success in the the applicator across the field.

9 5 "Fluid lime" generally is applied through conventional There is no problem with solubility such as is inherent in fluid application equipment that is used to apply suspension clear liquids, and the incorporation can be accomplished fertilizers. The only change required is the use of larger usually by merely adding appropriate amounts of micro- nozzles to pass the limestone. Usually flooding type nozzles nutrients to the suspension mixture while it is being made. are used. Experiments show that grades such as 10-34.0-6211 and 5.15-30.3211 can be produced from APP solution (11-37-0 from furnace grade superphosphoric acid), putash, zinc Phosphate Rock Suspension oxide, and clay. This is far more zinc than can be dissolved The techniques used to produce limestone suspensions in a clear liquid and would be enough to meet agronomic can be applied also to making suspensions of finely ground requirements. The viscosities of freshly prepared 10-34-0- phosphate rock. There is abundant evidence (29) that rock 6Zn and 5-15-30-3Zn were 225 and 619 centipoises, phosphates from some deposits, when finely ground, are respectively, at 70°F (21°C). This is well below the very effective sources of phosphorus for low-pH soils. In maximum of about 1,000 centipoises at 70°F (?lac) that past years, there was an appreciable amount of fertilization can be applied satisfactorily. carried out in the Midwest by dry application of finely Viscosity increases as the amount of micronutrient ground rock; however, the practice now has given way to material is increased and as storage time is lengthened. use of soluble phosphates. Likewise, no use of phosphate Table 22 shows the viscosities of three 5-15-30 mixtures rock suspension presently is known. In 1976, TVA reported (contain 1% clay, P,O, source APP solution 11-37-0) (30) a study on preparation of phosphate rock suspensions. made with three different levels of zinc supplied as oxide Advantages recognized for the suspension form of rock of high purity (80.3% Zn). were the absence of dust in handling; the possibility of These data show that as zinc content of the 5-15-30 is using a finer, more agronomically effective grind of rock; increased from 2% to 4%, viscosity of the mixture increases and mure uniform application to the field. The TVA tests from 459 to 835 centipoises. If the zinc content shuuld he were made largely with North Carolina phosphate, a rock increased still further, the viscosity of the mixture probably that is among the more effective fur direct application. would exceed 1,200 centipoises. These data show also that Grinds of 67% and 87% through 325 mesh were tested. viscosities of the suspensions increase w.ith storage time. With simple, high-shear mixing of rock with water, However, in some instances periodic agitation of the allowable viscosity was not exceeded with rock contents up suspension will reduce and prevent this increase in viscosity. to about 70% with the coarser rock and 65% with thc When the 5-15-30-3Zn suspension was agitated by air, its finer one;corrcsponding P,05 contents were 21% and 1Y%. viscosity decreased and did not increase with storage time. Use of about 0.6% TSPP dispersant increased maximum The data indicate that the 5-15-30-3Zn should have excel- rock contents to 78% and 70% (23% and 21% P,O,). lent storage characteristics, provided it is agitated periudi- ~-I-loweve:,-. -. none of these suspe~m~?ad: wit&xJ use.>!- cally. This suspension should be a popular mixture in a __ -- -. - ... - - . .- . suspending clay had good storage properties; in unagitated satellite station as a source of both potashiind7inT IiT storage those made with the coarser grind settled excessively these stations it could he mixed in a nurse tank with in I to 2 hours and those made with the finer grind did so 10-34-0 or 11-37-0 and 32-0-0 liquids tu produce various in 1 to 2 days. NPK' solution-type mixtures that contain zinc. Mixed with Rock suspensions with much better storage properties a base suspension and 32-0-0. it could be used also to were made with use of a small proportion of attapulgite produce other suspensions. suspending clay: however, both proportion of rock and The tabulation below shows the approximate quantities proportion of clay were limited by the imparted viscosity. of each micronutrient mateiial that can be incorporated Suspension of 18% P,O, content (60% rock) with good individually into a 15-15-1 5 suspension mixture (about storage properties was made by the following procedure: 1 .O% clay and 70% polyphosphate) without adversely (I) nik rock (87% - 325 mesh) with water, 0.5% attapulgite affecting the physical properties of the suspension. suspending clay, and 0.3% TSPP dispersant, and (2) add 1% 0.35% B as sodium borate ammonium nitrate as a gelling agent during further mixing. 1.20% Cu as copper sulfate Because of the use of dispersant. only moderate mixing action was required. The resultant suspension stored well, I .20% Fe as iron sulfate without agitation. for at least I month. 0.34% Mn as manganese sulfate 2.50% Zn as zinc sulfate Micronutrients in Suspensions The 15-15-15 grade suspension used was produced from 12-40-0 APP suspension (P,05 from furnace phosphoric Suspensions provide a nieatis of incorporating large acid), UAN solution (32-0-0 solution grade), potash(0-0-62), quantities of micronutrients into fluid fertilizers (31). and attapulgite suspending clay. These quantities of micru-

96 nutrients are sufficient to supply most crop requirements. plants that have excellent agitation seem to be able to Several micronutrients added together as a micronutrient dissolve micronutrients in clear liquids or to disperse them package may or may not cause viscosity problems. The data in suspensions with relative ease. The mixing procedure shown in table 23 are from bench-scale tests. that usually works best is as follows: add water and clay to These data sliow that the additiun of several micro- rnk tank first, disperse the clay in the water by mixing, nutrients :is the source materials shown in table 23 did nut and then add the UAN solution. Start circulating this have adverse effects on the 15-15-15 suspension. However, inixture through the injector. Then add the micronutrients plant data show that when manganous-manganic oxide through the injector intu the recirculating fluid. Next was used in combination with zinc iixide, flowers of sulfur, the phosphate base solution (10-34-0 or 11-37.0, etc.) is borax, APP base suspension, 32-0-0, and potash to produce added. The potash is added last in the mixing sequence. a 16-8.16-1.72S-0.85Zn-0.84Mn-0.09B, the suspension Tlus same mixing procedure works well in plants that do became tuc viscuus to pump after about I hour of storage, not have injectors, except the preweighed micronutrients even though immediately after the suspension was pro- are poured from the open bag over the top of the tank duced. its initial Viscosity was only 310 centipoises at while the water-clay-nitrogen solution mixture is recircu- 70°F (21°C). Therefore, in the production of suspensions lated in the mix tank. Care should he taken to avoid containing several inicronutrients, care should be taken to dumping the micronutrients in large batches. They should select micronutrient suurce materials that will not react be sprinkled into the tank at a reasonable rate. If this is with each other and forin excessively viscous suspensiuns. done, the formation of large balls of micronutrients can be It is advisable not tu store suspensions that contain several prevented and the micronutrients can be easily dissolved niicronutrietits unless storage already has been tested and or dispersed in the fluid. pruven satisfactory. The above information relates to the use of micro- Prubably the niost convenient way to add inicronutrients nutrients with phosphate suspensions that were produced (usual size 100% -20 mesh) to a mix tank is through an from furnace phosphoric acid. Actually, good results can injector similar to the one sliuwn in figure 35. Also those also be obtained when wet-process phosphoric acid is used as a source of phosphate for the manufacture of the Table 22. Viscosities of 5.15-30-XZn mixturesa phosphate suspension. However, when the wet-process Ccntipuiscs~~_____~ at 70% (2lOC) acid source is used, no more than about 80% of the micro- I.reshly Oneliay, One-weck Two-week nutrient values shown above should be added and, in __~___Grade prcparcd sloragc rtar2pc" sturagcc 5-IS-30-2Zn 459 966 838 798 addition, the total plant nutrient cuntent may have to be 5-15~30-3Zn 679 1.436 1.290 1.234 lowered by about 10%. 5-IS-30-4Zn 835 'Too thick to - - tic nic:isured or ;igit;ltcd by STORAGE OF SUSPENSIONS -~~__:iii spuging "5-15-30 madc with 11-370 (iumxc grade P,Os), solution-giadc potash. :md ailiiiioniil :ct;ig content 13 Storage Vessels bMotcrial not ilgitatcd prior tu viscosity nieilsurcnient. 'Viscosity taken niter apitiltim by iiir sparzing. It is recommended that suspensions be stored in vertical, mild steel or plastic tanks equipped for air sparging (32). Table 23. Addition of mixture of A tank sire of abuut 12 feet diameter by about 20 feet micronutrients to- 15-15-15 suspension - tall (16.000 gallons) is very satisfactory. Thickness of the Apparcnl ___viscosity metal should be about l/4 ill. Life of a steel tank ciln be _-__C.eiitipoises extended by coating the inside with a material such as an Fresh One-day epoxy-base paint. Mixture added product aging The tanks may lime flat or cune bottom. Although cune 0.17. B as sodiuiii borate bottum tanks are mire expensive than flat bottom, it is (Na? B, o7 .5H20. 14.3% B) easier to resuspend niaterials that settle in the cone of the 0.2% Cu as copper sulfate tank. When suspensions are not uf good quality, solids (CuSO, .5H20) soiiietiiiies scttle and collect on the buttoin at the outside 0.676 Fe as iruti sulfate wdls of a flat bottutii tank atid caniiot be resuspended with (Fe? SO4.OHi 0) 200 470 air sparging. The problem can be avoided with cone bottom 0.6% Mi1 as nianganese sulfate tiinks. (MllS04.XH20) The foundation for a storJge tank shuuld be abuut 1 0.3% Zn as zinc sulfate feet deep and made uf 1/4-in. crushed rock or gravel. The (ZIISO,, 36% Zll) gravel should be wcll comp:tcted. Concrete m;iy be used

91 ._. CLAY HOPPEI

TO WUP INLET LINE

,QNk,? ...... ASSEMBLY VIEW

4" WELDED FLANGE

NQIc USE RUBBER HOSE. AVOID KINKS OR SH4RP BENDS. RUN STRAIGHT INTO PUMP INLET LINE IF POSSIBLE.

Figure 35. Micronutrient and clay injector for fluids

coaled. on adequate. The air reservoir (300-~alluntank) is pressurized - also. Thc- flct- bs!?~~; r;!' ;12!r&t;ink shoiild~ . .. he-. . .- . .~~. . .- .~-. .. ~. ... the outside with asphalt or epoxy resin paint hefore it is to IO0 to 125 pounds pei square in.'Theii tlie contiol valve placed on the foundatiun. The location should be well for regulating the air discharge is opened wide to iillow ail drained so that the tank bottom does not ever stand in tlie sir to discharge through the tank to give the suspension water. a rolling type nix. The reservoir is recharged then to If a separate centrifugal pump is to he used for triins- sparge the next tank. The reservoir may be connected tu ferring the suspension into and out of storage, tlie pump asinany tanksas desired and one tank at tlie time is sparged. should have a 5-in. inlet and ii 4-in. outlet and he equipped Tlie primary purposes of sparging air are to prevent the with a 30-lip inutor. A inure powerful motor may be gel in the suspensioii from becoming tou strong, to ccintroi required if pumping is against a high head. such as a long syneresis (furinatiun of cleaI liquid layer (in top of the IUII uf pipe or an exceptionally high eleviitiun. Tlie transfer suspeiisi(in), :id to prevent giowtli ut" hrge crystals ill the lines should he 4-in. pipe: the pipe may he black iron or suspension. PVC schedule 80 plastic. The plastic pipes shuuld he Fur the purpnse of controlling gel stiength and syneresis. equipped with expansion joints and sh(iu1d be well spargiiig e:icIi tank once or twice a week is adequate. supported (see chapter IO). "Spaiging" inieaiis discharging the air from the fully cliarged reservoii niice into tlie tank. SPargiW Tu prevent excessive crystal growth, ntiier factors liiivc tu be considered. Crystal growth in suspensions occurs Recoinmended design for iiii spargiiig for flat hottoiii niainly as a result (if cooling tlie suspension. As the tempera- tanks is shown in figure 36. A sparging system for a cone ture dccre:ises. tlie aiiiount of salt in the solution phase bottom tank is s1iow11 in figure 37. Ai1 air cumpressor decreases by crystallizing on tlie surfaces (iftlie existing powered by a 3-lip electiic iiiotor or gasoline engine is crystals. If cuiliiig is unifoiiii througliuut the tank, growth

98 Figure 37. Cone bottom suspension storage tank with air sparger system

During winter, in those regions where temperatures are frequently below OOF (-18OC), a suspension sometimes Figure 36. Pipe sparging system for suspensions becomes frozen to a solid mass. However, when the outside temperature subsequently remains above 40°F (4.4OC) for of all crystals will be uniforin and none of the crystals will a period of about a week the suspension will become grow excessively. However, if the cooling is not uniform, fluid and can be sparged. such as getting colder next to the tank wall, the crystals in the colder part of the suspensiun will grow excessively. They will be fed by the warmer. more concentrated solution TRANSPORTATION OF SUSPENSIONS phase by diffusion of salt through the liquid to the colder more dilute solution. By this means. selective crystal One of the drawbacks of liquid mixed fertilizers is low growth occurs and excessively large crystals may form next analysis. With the new high-speed applicators which apply to the cold tank wall. The remedy is to sparge the tank fluid fertilizer at a rate of inure than. 1 acre per minute, it frequently when the suspension is cooling. When the air is very difficult to keep the applicator charged and moving. temperature drops in the Fall and winter, the suspension Part of this problem can be solved by using high analysis should be sparged once or twice daily until the temperature suspensions so that more plant nutrients per load can be of the suspension is about the same as tlie temperature (if delivered to the applicator. However, there are some the outside air. This frequent sparging will cause tlie problems in transporting suspensions that do not exist with temperature throughout the tank to be unifurni while solutions. cooling and will prevent excessive growth crystals. of The main difficulty is tliat road vibrations within the When tlie daily average air temperature is about the same transport tank cause the gel structure of many suspensions as tlie temperature of the suspensiun and remains about to weaken enough to allow crystals to setlle. This difficulty constant or rises, the routine schedule of sparging once 01 twice a week may be resumed. usually can be circuiiivented by recirculation of suspension The suspension always should be sparged imniediately through a sparger while in transport. Figure 38 is a diagram before it is withdrawn from the tank. Spargiiig will make of a transport truck for suspension fertilizers equipped the suspension more fluid :ind should correct any syneresis with an agitation sparger. This transport is equipped with a (mix in my clear liquid layer) that may have (iccurred. recirculating pump driven by ii hydraulic niotor. The pump

99 is a centrifugal type and usually has a discharge diameter of sparger sweeps the bottom of the tank and causes the fluid 3 in. Piping to the pump and sparger is of 3411. schedule 40, to be agitated. stainless steel 01 plastic pipe. The sparger has 1/4-in. holes Some companies prefer cone-bottom trucks with an air drilled on 4411. centers, The liquid recirculating through the unloading system as shown in figure 39. Air is injected

Figure 38. Transport truck with pipe sparger for suspension fertilizers

TOP AIR

~ AIR FOR UNLOPDING

- -NnTF- . SKETCH SHOWS VALVE SETTING FOR \AIR FOR SPARGING AIR AGITATION CENTER TANK. AIR PRESSURE ALSO USED TO FORCE SUSPENSION FROM TANKS. Figure 39. Cone bottom transport

~ 100 at the base of each hopper and fillers up tbrougli a diffuser They have a hydraulic system which raises and quickly pad to agitate the suspension and make it flow. When the unloads the tank as it is dumped into the applicator. appropriate pressure is built up in the tank (up to 25 psi), Other companies have installed large transfer pumps with the discharge valve is opened so that air blown into the capacities up to 300 gallons per minute. suspension forces it througli the discharge line. Pressure The application of suspensions is discussed in chapter 9. remains constant and the flow rate is controlled by a butterfly discharge valve. Other companies transport fluids in spherical tanks on References Ilal-bed trucks. The spheres shown in figure 40 are made of polyolefin plastic and are excellent for storing and trans- 1. Wdtcrs, H. K., Jr., "Salt Suspension I'ertilircra," Corninercial porting suspensions. Materials that settle to the bottom are f~'cwrilizws,pp 25-6, (Scptcmhcr 1959). circulated to the top of the spheres. Nu disficulties lhave 2. Ncwsom, II'. S., Jr., "Suspension I'crtilizcrs," SolstiorIs, pp 30, 31. 33, 35 (January-l'cbruaiy 1960). been reported with deterioralion of the plastic over a 3. Sawyer, E. W., 1. A. Polon. and H. A. Smith, "The Stabiliza- period ui 10 years. tion of Suspcnsion Fertilizers with Colloidal Attapulgitc," Field experience litis slinwii that ciiiises for delay in .Soliitio,is. pp 36-43 (Januaiy-l:chiuary 1960). keeping materials supplied to the large applicators are: 4. Achorn. 1:. P., "Potwiocn: Predict Urc in Liquids Will Double ID ' The size and number of transports used tu haul in I5vc Ycarr." Ferlilizer Sohttionu, 20 (I). 72. 74, 76, 78, 80, 82. 84-5 (Jmuary-I:ebruaiy 1976). suspensions to the applicator. 5. Ach~m,I. P., "Usc of Micronutrients in Fluid I'crtilizcrs." 2. Time required to pump lhe suspension from the Rcprintcd from National I'crtilirer Sulutiuns Association transport to the applicator. Additives and Iluid I'crtilizerr 1969 Round-Up, pp 5-8. One equipment conipmy has developed a tilting type 6. Achom. I'. P., W. C. Scott, and J. A. Wilbnnks, "Physical transport that can be emptied quickly into the applicators. Ci~mp~tihilityTcsts of fluid Fcrtilizer-Pesticide Mii\tures Conductcd Mnv 1970." Ywtilizcr .Solutio,rs (Novcmhcr- A photograph of this transport is shown in figure 41. Deccmhcr 1970). These tanks are self-supporting and hinged on one side. 7. Dovcnuorl. J. E.. J. G. tictsinccr. D. W. Rindt. and D. E. Niclmls, "Suspension 1:crtiIizcr: Gcl-Strcngth Mearurcmcnt 2nd Gel Characteristics Stodicd," Fertilizer Sohirioris. 22 (1). 88. 90. 92-9 (January-l':cbruary 1978). 8. Vivaldi. 1. L. Martin. ;md K.H.S. Kobcrtion. "Polygwskite ;~ndSepiolitc (the hormites)." The &lcctrorr-Upticnl Iuvesri- yolio~01 Claw pp 255-72 (1971). 9. Htidcn. W. L.. Jr.. "Attapulgitu: ItsPropciticsand Application." Imlirstriul orid E,,~~,,ec~,ryClrerriisrr~v.59 (9). 5469 (Scptcnihcr 1967). 10. Van Wazcr, J. I<.,"Pliorphcm~* and Its Compounds." lriter~ s'ciirwe I'i~hlishrrr/I. pp 1679.481 (1941). 11. Ihbkuwski. 'lcd P.. and Alan I. Solonlon. '' 'Liquid' Clay far Supcnsion l:crtilizars." /;wrilizw .Sohilims. 19 (I), 78-82 (Janu;iiy-I:chruary 1975). I?. "Ncw Ucvclnpmcntr in Fcrtilixr Technology." I2111 Ilemon- stration. TVA Ilulletin pp 77-9 (Octohcr 18-19, 1978). 40. Y-136. Figure Spherical transport tank 13. Kulbicki. Geor~cr."Hixh T~nipcratorc Phascs in Scuiolilc.

Figure 41, Tilting-typc trdnspwt fur suspensions 20. Jones, T. M., and I. G. Getsingcr, “HighCrade Siispfnsions Suspcnsion 1:crtilizcrs: Bcnch-Scale and Pilot-Plant Studics,” Produced from Wet-Process Orthophosphoric Acid,” Fertilizer 184th ,Votior,ol rllrrtiriK of’ fiw Amwicon Cl~e,~icolSocirtj Solutions (January-February 1979). (Scptcmbcr 12-17, 1982). 21. Mann, H. C., K. E. hlcGil1, and 7. M. Jones, I&K Product 21. “New Uevclopmcnts in 1:ertiliicr Technology.” I:ifth Demon- Heseare11 & Developmnt. pp 488-95 (September 1982). stration, Tenncssee Vallcy .Authority. Octobcr 6-7. 1964. 22. Balay, H. L., “Marketing Advantages of Using Solid Materials 28. Baloy. H. L., and David G. Salladay. “Proccss and Product in Suspensions,” proceedings NFSA Round-Up: Aim 200. Considcrationr of Iluid Lime.” August 1980. Pcoria, Illinois, pp 41-5 (1976). 29. Khasawneh, I:. E.. and E. C. Doll. “The UECof Phusphaic Rock 23. Kramer, James C., “Using MAP Solids in I.’luids,” Form for Direct Application to Soils,“ Adi,mccr i!! Apro,imi,’. Clremicd~,141 (6). 42-3, 46, 48, 52 (1978). No. 30, PI’ 159-206 (1978). i4. Parker, B. n., M. M. sorton, and T. R. Stumpe, “Pilot-Plant 30. “New Dcvclopmcnts in 1:ertilizcr Tcchnology.” 1 I th Demon- Studies of the Pipe and Pipe-Cross Reactor in Production of stration, TVA Bullctin Y-107. (Octobcr 5-6. 1976). Granular Polyphosphatc Fertilizers," 32~dAnrzual h’eetirzfi of 31. Acharn. Frank P., “Use of hlicronutricnts in l:leid 1:ertilircrs.” Fertilizer lt2dlIstrJ’ Hound Tohle (Octobcr 26-28, 1982). (July 1969). 25. Scott, W. C., M. R. Burns, and J. R. Watson, “New Dcvclop- ments Strcamlinc Suspension I’crtilizcr Production.” Agri- 32. Achom, Frank P.. and Honicr L. Kimbrougli. “lranrporting. ehemicolAge, pp 10, 12, 14, 334 (March 1978). Storing. and Applying Suspcnsion krtilizcrs,’’ I?wiIizer 26. Boles. J. L., and T. M. Jones, “Urea-Ammonium Sulfate Solrrfioo,is.pp 74-81 (May-Junc 1977). Chapter 8 Specialty Grades of Fluid Fertilizers By Eugene 6. Wright, Jr

Specialty fertilizers are frequently thought of as 0-6-7 ratio; it decreases to about 44% for the 0-3-4 ratio, expensive, nonagricultural fertilizer products used mainly increases to 55% for the 0.5-8, and then decreases again as for feeding house plants or nursery stock. A popular form the K,0:P,05 ratio increases. Because of the steep might be tablets or capsules of plant nutrients in a concen- solubility gradient, small changes in composition may trated form which easily dissolves in water to yield a low cause precipitation in the regions of highest solubility. analysis product. Such a picture is incomplete since it does For example, if the P,05 content of 0-25.4-29.6 were 25% not include those products used for fertilizing row crops by foliar application or feeding high-cost crops requiring special consideration. such as low cliloriiie content in tobacco fertilizers. To include these special use materials. this chapter will include those frequently used chemical compounds which supply one or niore of the primary plant nutrients 2nd are used or applied in a distinctive or unique manner in a liquid form. Tlie decide of tlie 1940s appears to be the beginning period of modern liquid fertilizer development at which time production of liquid mixed fertilizers started to become significant iii tlie Pacific Coastal areas ofthe United States (I). This interest in liquids rapidly moved eastward during tlie next decade and a half. During tliis period. interest was deiiiunstrated in nonchluride foinis of potassium base liquids especially since the supply of potassium hydroxide periodiczlly exceeded the inorinnill demand and a natural outlet for this excess could be liquid iinixed fertilizers. John D. tlatfleld in an internal TVA innem~iraiidumoutlined the solubility relationship of various potasiuiii ortliopliospliates in tlie system K, 0-P, Os-H2 0 at 3?F (OOC) which is sliowii in figure 42. Tlie solid liiics indicate [lie solubility relationship. tlie broken lines sliow constaint fertilizer ratios. and tlie dotted line gives tlie solubility iii the inietastable regijoii of dipotassium phosphate triliydrate. The solubilities for various ratios are given in tJble 34. The system K,O.P,O,-tl,O is cliaracterized by P205, PERCENl two regioils of high sulubility. The plant nutrient content Figure 42. Solubility in the system illcreases froiin 29% for tlie 0-1-1 r:itio ti3 about 55% for the KIO-P,0s-H20at 32°F

I03 instead of 15.4, precipitation would occur at 32'F (0°C) sanie rate until the crystals dissolved to determine saturation and the liquid would analyze 0-23-27.2. temperatures. The maximurn grade with a salting-out ti^^ such as 0.2.1 and 0-3-1 are of little interest in this temperature below 32'F (0°C) was determined fur each because of 10% solubility and high acidity of the type of solution in the ratios studied. The liquids that had solutions. Ratios such as 0-1-2 and 0-5-8 are questiunable saturation temperatures helow 32°F (0OC) also were tested because of their alkalinity. in storage for 1 week at 32°F (OOC). The crystallizing The addition of urea to the 0-1-1 ratio not only !oaks phases for most of the liquids studied were identified by promising from the possibility of high analysis, but also petrographic analysis. An electronic pH meter \vas used 'permits the preparation of grades in ratios such as 1-5-5 and for pH measurements and a hydrometer was used for 1.4-4; these liquid grades are too corrosive when prepared specific gravity determinations. from ammonia, phosphoric acid. and muriate of potash. Superphosphoric acid. potassium hydroxide, urea, When dipotassium phosphate (0-40-52) is mixed with and water were used in the priiduction of I:?:?, 1:2:3, ammonium phosphate liquid, complete liquid grades of 1:3:3, 2:4:5, ]:]:I, 1:2:l, and 1:3:l ratio fertilizers. very high analysis are possible. Two such grades are 4-24-1 2 These products contained no chlorine. Ammonia was and 6-24-6. included in formulations for the ]:?:I and 1:3:1 ratios, A 5-20-20 polyphosphate grade with all the potash otherwise low pH products would result since tlie amount supplied as potassium hydroxideand withall of the nitrogen uf potassiuiu hydroxide was nut sufficient to neutralize supplied as urea had a salting-out temperature of 18°F the acid. Most of the products were essentially neutral (-7.8OC) and urea was the crystallizing phase. (pH 7.0 tu 7.5) except those of 1:2:3 and ?4:5 ratios, Tlie introduction of"electric furnace superpliosphoric which had a liigli pH (8.3 to 11 9)because their ratio of acid by TVA in mid-1950 led to the production of high KiO to P,O, was high. The caustic nature of the latter analysis ammonium polyphosphate liquids on a commercial products would not present any serious corrosion problems scale. Research alsu was directed toward liquid fertilizers with mild steel tanks and equipment: however. aluminum based on superphosphoric acid (76% P,O,) and potassium equipment would not be suitable. hydroxide (2). Salt-out and saturation temperatures, In developing tlie data for table 25, it was found that the crystalline phase, pH, and specific gravity measurements for salting-out temperatures decreased and then increased as several grades in representative ratios of luw chlorine the concentration of the products increased. Water was the containing fluids from this study are shown in table 25. crystallizing phase at the lower concentrations. and These grades were produced batchwise by ;idding in increasing the concentration depressed the freezing point. succession the potassium hydroxide, superphosphoric At the 1:2:2 ratio, salting-out temperature of the 5-10-10 acid. water. and ammonia. if needed, while keeping the grade was 12'F (-I1.I"C). lncreasing tlie concentration tu a reaction mixture temperature below 180'F (82.2'C) 7-14.14 grade depressed the. salting-out temperature .to ----by-cooling;solid raw materials such as-urea werc-added iasr. .5"-F --(-l.5.°C).. Further -concentratiun- caused uiea tu The product was then rapidly cooled below 100°F (37.8OC) crystallize :it a higher solution temperature of 36°F (2.2OC). to further lessen the hydrolysis of the polyphosphate which Although small, the differences between siilting-out is enhanced at low pH and/or high temperature. Salting- and saturation teniperatuim shown in the table indicate out temperatures of the products were determined by that there was some supercooling during salting-out cooling the solution at a rate of 4'F (2.2OC) per hour deterininations or overheating during saturation deter- until crystals formed. Tlie solution was then warmed at the minations. Nevertheless, the accuracy is believed to be sufficient for practical application in pruductiiin uf the grades listed. None of the solution^ with siituratiun Table 24. Concentration of Dlant nutrients temperatures below 32°F (0°C) crystallized during storage in the system K,O-P,O,-H,O at 0°C for 1 week at 32'F (OOC). The superphosphiiric acid was Total plant added to the potassiuiii hydroxide rather than vice versa. Ratio Grade tiutrient, %a This is believed to niinimize hydrolysis and also avoids 0:I:1 0-14.5.14.5 29.0 forming the relatively insuluble nionopotassiuni phosphate. 0:9:10 0-22.6.25.2 41.8 Specimens of mild steel A 283 iind stainless steel types 0:6:7 0-25.4.29.6 55.0 304 and 316 were tested for currusiun by a slightly basic 0:9:11 0.21.4-26.2 47.6 piitassiuni polypliospliate su1utioii (3). A 0-24-26 grade 0:3:4 0-18.8-25 .0 43.8 with mi initial polypliosphate content of 77% and a 0:2:3 0- 18.5-27.8 46.3 temperature uf 130°F (54.4'C) was used. At the end uf 0:5:8 0-21.2-33.9 55.1 2 weeks, tlie currilsim rate was measured and the smiples 0:1:2 0-14.9-39.7 44.6 were subjected to further possible attack in fresh solutiun "Pcrccnr (N + P2US+ K20). fur :in additional 2 weeks. Attack on the stainless steel

104 0z03 00MZ Z>ZZ 00: 000: 00: 0oM ZZkh zz> 2Z2> ZZk ZZ>

0,s' I IZI I 3 = r Y

mN-m -mow *ww zmom Nom mvm No m No m

.. 1.. 5 5 -* 1 1 w ~mm mo - mm

11 I

cC?mc-P *om

Ill1 1111 m-m r-mm Ill1 Ill 20'0 -22

105 . _.

106 Was negligible at the end of 28 days. Samples of mild steel A 0-25.25 grade was the most concentrated 0:l:l placed in unaerated solution corroded at the iate of 10 mils ratio that salted out below 32°F (O'C). Water was the per year based upon tlie 14-day test and 14 mils per year crystallizing phase. A 0-26-26 salted out at room tenipera- when reimniersed for ;in additional 14 days. Samples ture and potassiuni dihydrogen phosphate was the crystal- immersed iii an aerated solution were corroded at nearly lizing phase. With 80% P,05 electric furnace acid and twice tlie unaerated rate (19 and 25 mils per year). A KOH, a 0-30-30 solution containing 80% polyphosphate is partially immersed cold worked sample of niild steel was satisfactory at 32'F (0°C) (4). Liquids of maximum plant attacked ;it the rate of 27 and 40 iiiils per year under static food content can be produced from superphosphoric conditions. Grooving to a depth of 7 mils occurred at the acid and potassium hydroxide when the P,O, :KzO ratio liquid air interface. is slightly less than I, This is illustrated by the 0-29-31 and Table 26 compares the niaxiiiiuiii grade possible which 0-27-36 grades with saltingout temperatures below -15°F would store at 32'F (OOC) for 7 days and 30 days at room (-26.1°C) (table 25). With orthophosphoric acid and temperature without salting out when the source of potash potassium hydroxide, the inaximum grade with a salting- is potassium hydroxide instead of the normally used uut temperature below 32°F (0°C) for a 0:l:l ratin is potassium chluride. The acid used was 76% PzO, furnace 0-14-14(tabIe 24). acid and the polyphosphate level was about 50%. The data The source of P,O, for the specialty grades discussed in general sliows the grades tu he two to three units higher so far has been electric furnace acid. So long as solubility when the potassium hydruxide supplied the potash require- relationships were not exceeded, a water clear sulution ments and supplemental nitrogen was supplied as urea. resulted from the reaction of 76% and 80% P,Os electric Supplying some. of the potash as potmiuiii chloride furnace acid and potassium hydroxide. generally caused the salt-out temperature to increase as A slightly cloudy product resulted when extremely high did supplying the supplemental N as UAN solution rather polyphosphate electric furnace (83% P,O, acid - 95% than urea. A 7-14-14 solution with all the potash supplied poly) was reacted with KOH to produce a 0-25.3-25.9 as potassium hydroxide withstood tlie storage tests at solution having a polyphosphate content of 91%. The pH 32°F (0°C). but 3 6-12-12 was the highest satisfactory and specific gravity at room temperature were 8.1 and 1.56 grade when pot:issium chloride was included to supply respectively. 50% of tlie potash requirements. The maximum grade was TVA produced about 1,100 tons of nominal 0-26-25 further reduced tu 5-IO-IO when UAN solution W:IS the potassium polyphosphate solution in the liquid fertilizer source of supplemental nitrogen or when KCI was the demonstration plant in 1976 to fulfill a crash agronomic only source of potash. study on tlie benefits of foliar feeding various crops- soybeans in particular. This product was made at a rate of about 15 tuns per hour by reacting electric furnace super- Table 26. Effect of source of potash on mdxinlum grader phosph(iric acid containing about 80% P,O, witli 45% of liquid fertilizer made with superphosphoric acid potassium hydroxide solution. The KOH solution, super Maxiilium grade of acid. and a small quantity of water to adjust the pruduct liarlid fertilizer tiiade with c~iiicciitrationwere fed simultaneously to a [nixing vessel. The temperature of the product in the mix vessel was Ratio maintained at about 114'F (45.6OC) by recirculating it 0:l:l in a closed loop through a shell and tube heat exchanger. I:I:I 11-11-11 9.9.9 Product ~iverfl~iwingthe mixing vessel into a surge tank was l:2:l 10-20-10 8-16-8 further cooled to about 72'F (22.2"C) in a secondary I:?:? 7-14-14 5- 10-10 cooling system before being pumped to storage. A tabula- 1:2:3 6-1?-18 3-6-0 tion of representative chemical analysis and pH measure- I :3: I 8-24-8 7-21-7 ~meiits is shown in table 27. The nuininal foliar grade l:3:2 8-14-16 4-1 2.8 produced froni the 0-26-25 was 10-2.4-4 witli 0.6% sulfur. 1:3:3 6-18-18 3-9-9 Agrononiic results (ifthese studies were summarized by 1:4:2 6-74-12 -b Gray in a paper presented to The Fertiliser Society of 1:4:4 5-20.30 - b Londwi (111 October 14. I977 (6). l:6:6 3-18-18 -b Liniiled research Ii:is 31so been conducted using wet- 7:4:5 6-13-1 5 4-8,Ln process acid (if vwying polypliosphate content as the source (if P,O,. A 0-?5-?6 grade WJS produced using wet-process acid produced from both calcined and uncalcined rock. The polyphospl~atelevel varied from 0% to 50%. AS found wllen electric furnace acid is used. the solubility of 0-25-26

107 Table 27. Production of 0-26-25 potassium superphosphoric acid to supply about half the K20 in a polyphosphate liquid fertilizer (8) 5-10-10 specialty grade. Leaching with less than about 3 Perccnt parts water per unit of burr asli resulted in unsatisfactory ofP,Os as P20s:K,0 Tcrnp, extraction ofthe K,O from the ash. Grade polyphasphates ratio pll a Ot: A process for producing fertilizer grade KHzPO, was 7.4 1.527 107 0-25.2-24.6 69 1.024 discussed at the ISMA Technical Conference in Sandefijord. 0-26.2-25.3 73 1.036 ~ 1.554 113 Norway, in September 1970 (8). In tlus proccss, KCI is 0-25.7-25.4 82 1.012 7.4 1.551 1 I4 0-25.8-25.5 81 1.012 7.3 1.554 114 reacted with hot H,S04 [400"F (202.2OC)I to evolve '0-26.3-25.7 76 1.023 7.2 1.553 116 relatively pure HCI and form a residue of KHSO4 in H, SO,. I 0-26-0-25.0 76 1.040 7.2 1.553 116 The latter is reacted with rock phosphate to form a slurry 1.551 114 ~~ 7.5 n-2s-7-2s. q 81 0.992 of KH,P04. gypsum, H,PO,, and iiiipurities from the 0-26.6-26 .0 85 1.023 7.3 1.552 114 0-26.5-25.3 83 1.047 7.1 1.551 113 rock. Gypsum is filtered from the slurry; crystals of 0-26.5-25.0 82 1.060 7.0 1.550 112 KH,PO, are centrifuged from a mixture of the filtrate and 0-26.3-25.1 76 1.048 7.1 1.5S0 120 methanol, after which the methanol is recovered by distilla- tion leaving a still bottom rich in H3P04. About the same grade is more favorable than that of either 0-25-25 or time Penmoil United. Inc., and Coulding Fertilizer, Ltd.. 0.26-25 grades. The acid was added to the potassium Ireland, announced similar processes for producing hydroxide to keep the solution basic which allowed pro- potassium polyphosphate (9). Neither process Iias been duction without going through the minimum solubility fully developed to conitnercial scale. area where the K:P mole ratio is 1 (KH2P04 is the least TVA demonstrated a somewhat similar process producing soluble of the potassium orthophosphate salts). In all a solid product using wet-process acid and KCI at the 5th tests, considerable precipitation occurred which was deter- Demonstration of Fertilizer Technology in 1964. Heating mined to be K2S0, along with some iron and aluminum KCI and wet-prucess acid to about 600°F (31.5.6OC) for salts analogous to those produced by ammoniation of wet- 1 hour produced a melt wliicli was sticky and very Iiygro- process acid. Also present were crystals of KH,PO,. These scopic upon cooling (7). Addition of calcined dolomite to dissolved upon dilution with 26% K20solution (31% KOH) the melt yielded a dry product analyzing about 0-48-26 when the grade reached 0-23-26 (pH about 8.0). Further which liad good physicill characteristics; however. while dilution with water to 0-14-16 grade dissolved the 90% of tlie P,O, was citrate soluble. only about 48% ot potassium sulfate. Upon standing. some of the 0-14-16 the P,O, wiis water soluble. If. instead ofadding dolomite, grade produced with low polyphosphate gelled. The undis- the cooled melt was aninioniated to it pH of6.5, a 4-24-12 solved solids tend to precipitate rapidly from the wet- liquid grirde resulted with a11 of the P20s water soluble process solutions so that relatively clear liquid can be and 60% of it present as polyphosphate. Heating the KCI-

separated-either by decattlaliuii ut. iiltratioii. acid mixture..io .iztiiperatures. between 60OoF-(3 l.5.6°C) ~ . .. ~. Sources of potassium other Llian KOH tu pioduce and 1,400'F (760.0"C) caused tlie reaction products to potassium phosphate sulutioiis have been studied. These solidify. Continued heating to 1,500"F (81 5.6"C) include potassiuni bic;irboiiate, potiissiurn carbonate, refluidized the reactants and produced a melt analyzing cotton burr ash, and reaction of KCI and acid at high 0-56-36 with 98% of tlie P,Os as highly condensed poly- temperatuies (7). The salting-out temperatures of 0-'20-20 pllosphates. The product was unsuitable for production of solutions produced with potassium bicarbonate its the high analysis liquids since gels formed in mixtures of cooled source of K were similar to those made with potassium melts and 1 to 2 parts water. carbonate and with potassium hydroxide. Considerable The addition of sodium 1ctrabur;lte priur to heating foaming occurred when tlie acid was mixed with solutions to 1,500"F (815.6"C) ti) furnish 2% boron reduced conden- of either K,CO, or KHCO, : Iiuwever, tlie fuaniing can be satioii of the phosphates ti1 only pyro and tripolyphos- managed by providing extr;! free board in the mixingvcssel phates tliercby increasing the solubility of the solid imelt and a iiieans of foaiii breaking. A mechanical foaiii breaker in water. A 0-25-16-18 liquid having good storage chiirac- similar to a whisk which is attached 10 the agitator tcristics resulted when equal parts of tmelt and water were woi~kedwell. mixed. Samples uf cotton burr ash wliicli cimt:tined iihoiit 37% Altliiqli reaction of KCI with phusptioric acid hits not KzO a5 carbonate. were leached with about 3 parts water become commercially feasible. reaction of KCI wiili nitric per unit uf asli tu yield a 10% Kz0 solution removing about acid tri make Iitrge tonnages uf KNO, has been accu~ii- 80% of the K20 froni the ash. The iemainiiig 20% K20 plished. Southwest Potash in Vicksburg. Mississippi. reiiiaitied in the residue which was separated frolit tlie succcssfully operated a p:ttented process wliicli also yields liqiiiir. This 10% scilutioii could be neutralized with 76% chlorine gas its a secondary product beginning in 1963

IO8 (IO). The Israel Mining Industries at Haifa, Israel, reacts References KCI and nitric acid to produce hydrochloric acid and potassium nitrate. Laboratory tests were made at TVA to I. A. V. Slack, "Liquids '65: A Searching Analysis of the Liquid determine the.suitability of potassium nitrate for production Industry," Reprinted from February, March, April, May, 1965 Farm Cilenticols. (TVA Reprint Number CD 394). of 1:l:l and 1:2:2 ratio tobacco fertilizers using 8-24-0 2. I. M. Fotts, et 01.. "Liquid Ikrtilizers from Superphosphoric and 11-37-0 grade base solutions. When needed, supple- Acid oad Potassium Hydroxide," Reprinted from Agriculturol mental nitrogen was supplied as urea-ammonium nitrate oi~dFood Clremisfry, 9 (3) 178 (MayIJune 1961) iTVA solution and supplemental potash as potassium chloride. Reprint Numbcr CD 281). The maximum grades that were satisfactory at 32°F (O'C) 3. lnternal TVA Report, Process Engineering Branch Progress Report, April 1976. were 5.5-5 and 3-6-6, when made with both base solutions. 4. New Develuprnents in I'ertilizor Tcchnology, 71h Demunstra- Comparative tests resulted in the same maximum grades tian, October 1-2, 1968. when using potassium sulfate, the usual source of potash 5. "Production uf 0-26-25 Potassium Palyphoaphatc Liquid Icrtilirer." Process Engineering Branch Progress Report, for tobacco. An 8-25-3 made by blending 10-34-0 liquid May 1976, p 47. ammoniuni polyphosphate and KNO, had a salt-out 6. R. C. Gray, "Foliar Fertilisation With Primary Nutrients temperature below 20°F (-7°C). During Reproductive Stage of Plant Growth," l'ertilisei The solubility relationship of several potash salts with Society of London. Piocecding No. 164, October 1977 (TVA Reprint Numbcr X 388). water at Various temperatures are shown in table 28. 7. New Developments in Ferlillzcr Technology. 5th Demonstra- Very slight interest has been shown in solutions of these tion, Octabor 6-7, 1964. Salts even as specialty products because oftheir low analysis 8. Fertilizcr Abstracts. Number 1317, Nu. 8, August 1976. 9. Ciieiiiicol Week. 110 (15) 14 (April 12, 1972). or high cost. The economics of foliar fertilization with 10. New Fertilizer Matwiols. Chemical Yroccsr Rcvicw, Naycs them is yet to become F;lvorable. Developmenf Corporntion, Chapters 9 and 11, 1968.

Table 28. Percentage of plant foods in saturated solutions at different temperatures of potash salts used or formed in the formation of fluid fertilizers Mono-

~~ Potassium Potassium Potassiunl uotassium DiootasciumI ~ Temperature nitrate chloride sulfate phosphate phosphate "F K, 0 N K2 02, K,O s KzO P,O, KzO P2Os % 32 5.4 1.6 13.8 10.4 3.7 1.3 4.3 6.4 30.8 23.3 35 5.8 1.7 14.0 10.5 40 6.6 2.0 14.3 10.8 4.7 7.0 45 7.4 2.2 14.7 11.0 50 8.1 2.4 14.9 11.3 4.6 1.6 5.2 7.7 32.3 24.1 55 9.0 2.7 15.2 11.5 60 9.9 3.0 15.6 11.7 5.6 1.9 5.9 8.6 32.6 24.5 65 10.8 3.2 15.9 11.9 70 11.7 3.5 16.1 12.1 b.4 9.6 33.3 25.1 75 12.7 3.8 16.4 12.4 6.9 10.2 33.9 25.6 SO 13.4 4.0 16.7 13.6 7.0 10.5 34.2 25.8 - 85 14.5 4.3 17.0 12.8 6.1 2.1 7.4 11.0 34.7 26.1

109 Chapter 9 Amlication of Liquids and Suspensions By Michael F. Broder

The applicability of unanswered questions exist, however, concerning use of has contributed primary nutrients in foliar feeding. these materials in the U.S. Fluids can be broadcast at high (& Band application is a general term referring lo several speeds or can be uniformly distributed to multiple outlets appllcation techniques. Surface banding is the application on tillage and planting equipment. Fluid flow rates can of fertilizer in corLtuli~ass trips on the soil surface. This also be accurately metered Dermitting accurate application. can be done by dribbling the liquid at low pressure and Fluid fertilizers are carriers for agricultural chemicals and is termed ij.drib.hle amlication)' Dribbling is often done on they can also be injected into irrigation water. The decline pastures. Surface banding is used to apply fertilizer between in fertilizer consumption by farmers in 1983 heightened rows of established crops. This is often referred to as competition among fertilizer dealer/applicators. More f- However, s-ing also includes the efficient application methods, weed control, and other injection- of fertilizer into the soil between crop rows&- services were offered by fertilizer dealers in an effort to is the term commonly used to describe the boost fertilizer sales. Liquid fertilizer, well suited for these procedure whereby liquid is applied to the soil surface in services, continues to occupy an important part of the strips. spaced as are crop rows but not necessarily aligned fertilizer market. with crop rows, then lightly disced into the soil prior to planting. Row application. in this discussion, refers to the --rr----Annliration Techniques @application of fertilizer in a strip near the seed

All fluid application techniques can be comhidinb planting.~~ The juxtaposition of the seed an d fertilizer band - ~ ~~~~~ two broad classes, broadcast and band application. Broad- zendent on the amount and type of fertilizer used cast application.~ is usually done with ground driven equip- and the tug- For wheat, the fertilizer is often placed ~ nient. Though most broadcast application is done prior directly with the seed. On corn, care is taken to place the to planting or seedling emergence, some crops typically fertilizer slightly below and to one side of the seed to are topdressed by broadcasting liquid fertilizer unto the prevent germination damage. Starter or pop-up fertilizer established crop. For some crops, aircraft are oftgusgd is similar to row applied fertilizer in that it is applied near Q to broad- fluid fer-. In arid parts of the U.S., or with the seed at planting; however, unlike row fertilizer, liquid fertilizer, particularly nitrogen. is broadcast- with starter fertilizer furnishes only part of the crop's fertilizer irrigation water through sprinkler irrigation systems. requirement. Another technique for applying... - liquid. fertilizers is known Another tvpe of band application is referred to as "root as foliar feeding: low rates of fertilizers are applied directly wzone" url'deep'' bandin$Dsp handing is the aoolication otitc crop leaves either by aircraft or high clearance appli- [Lfertilizer 6 or mo~..(l.L2r1&below the soil suriace_

cators. Foliar feeding is commonly used, to.~ apply micro- The bands may or may not lie below the seed. With nutrients such as iron and zinc on certaih cp(- anhydrous ammonia. the location of fertilizer and seed is

I IO crop and time dependent. In row crops, planting may take rated as high as 300 gallons (1,136 1) per minute at 100 place immediately after application of ammonia. in which pounds per square in. (68.9 x lo4 Pa) fluid pressure. case seeds are planted between strips of ammonia. In Agricultural chemicals and solids are sometimes held in grains, planting is delayed to allow aninionia to be suspension by use of a sparger (a pipe with holes oriented converted to nitrate and the seed and fertilizer rows may be to spray material tangentially around the applicator tank). perpendicular or may coincide. When nmiiionia and Two distinctions between solution and suspension phosphate fertilizer are banded together. the process is applicators are the nozzles and strainers used. Of the three to as "dual application.'' Dual application is nozzles most commonly used to apply fluid fertilizer, flat becoming a common practice in the US.Wheat Belt. fan nozzles (right side of figure 45) are only suitable for solutions. Solids in suspensions tend to clog the small Broadcast Application Equipment orifices in flat fan nozzles. Flat fan nozzles are preferred a when uniform distribution of fine droplets is required. Most fluid broadcast applicators are either trucks which Such is the case when herbicides are used. Sprays can drift, have been adapted to spray liquids or are special-built high however, under windy conditions. When winds are strong flotation vehicles. The high flotation vehicles are often enough to cause spray drift, flooding and hollow cone tricycle-type and are commonly referred to as "floaters" nozzles are preferred. Flooding nozzles (left side of figure (figure 43). Booms on either side of the applicator are 45) enlit drops the sane size as average rain drops; there is hinged at the rear of the applicator and can be folded little or no drift. They also give uniform application across forward for highway travel. Most applicators have booms the swath if proper overlap of adjacent nozzle sprays is that span 50 feet (15.2 ni): however. sonic models have maintained. booms that fold .at twu points and span 80 feet (24.4 ni). In most applicators. fluid is pumped to the nozzles by a centrifugd pump driven by a power takeoff of the applicator engine. Figure 44 shows the flow diagram of liquid in tlie truck. Note that pressure at the nozzle is usually controlled by the amount of material bypassed to the applicator tank. The regulating valve and the pressure gauge itre usually located in the truck cah. In some applica- I tors. a regulating valve is also located in tlie product line permitting nozzle flow control independent of recirculation. k Generally, broadcast applicators handle both liquids and suspensions. Sirice suspensions comprise a large part of the fluid broadcast market and require large plumbing and pumping capacity. nianufacturers Iiiive increased the diameter of applicator main lines to 3 in. (7.6 cm) with Figure 44. Flow diagram for fluid fertilizer applicator 314-k. (1.9 cm) fittings supplying each nozzle. Pumps are that uses pump bypass to control pressure at nozzle

Figure 43. High flotation applicator Figure 45. Nozzles used to broadcast and apply fluid fertilizers

Figure 46 (1) shows the importance of proper overlap. Broadcast application is most uniform when 100% overlap (double overlap) is used. Double overlap is achieved when the width of a nozzle spray striking the ground is double thc nozzle spacing. Since most applicators have a fixed nozzle spacing, the boom height is used to adjust spray overlap. Spray overlap increases with an increase in boom height. Some applicator manufacturers tilt the nozzles to spray at 45', instead of vertically downward, to increase spray overlap. A nozzle designed to reduce drift and produce larger droplets than flood nozzles is rapidly being adopted. This norzle produces a hollow cone pattern and is shown in the center of figure 45. The large droplets are produced in Figure 46. Spray patterns a swirl chamber which alsn decreases fluid velocity. The orifices in these hollow cone nozzles are, therefore, slightly A few companies use air pressure applicators. Pressure larger than orifices in flood nozzles of similar capacity. required to spray liquid is supplied by a sinall air In suspension applicators. strainers are sonietimes nut compressor driven by a power takeoff from the truck used because suspension is screened while filling the engine. Air is added to the applicator tank through a applicator. If strainers are used they are bar type or at least sparger to agitate the mixture as it is applied. Pressure is 6-mesh (3.36 lam) size. For solution application. strainers monitored and controlled during application by adjusting are sized tu allow passage of crystals half as large as the an air escape valve inside the cab of the applicator. diaineter of the smallest iiozzle orifice encountered. All of the niethods used to control liquid application ManuFacturers of fluid broadcast equiplnent use rates have limited accuracy due lo variations in fertilizer different methods for controlling application rate. Though flow properties. Pressure and ground speed sensing errors as a proportional flow valve is generally used to control nozzle well as nozzle wear can cause application error. Nozzle flow rate. the wlve is preset on some iipplicators while on flow rates at a given set of application rate settings are used uthers the valve can be adjusted during application. Some to improve application accuracy. This information. obtained applicators employ a pressure and ground speed sensiir by calibration, is often required to achieve accurate applica- connected tu elec1ric;tl controls which autuni:itically tion. Equipment hbricators usually calibrate nozzles with regulate a valve to maintain a constant application rate. water; a more effective calibration is obtained with the Automatic rate controllers are becoming popular because fluid to be applied. Fertilizer flow measurements take into they chininate the need to inonitor pressure. A few applicci- accuunt flow variations due to fluid density and viscosity. tors have piiinps driven by a ground wheel. Ground driven Fertilizer flow is difficult to predict using water data with punips can be geared to maintain :I fixed application rate correction Factors and equations. Data obtained by calibra- regardless of driving speed. tion can improve the accuracy of rate cantroIs and can ensure that the recommended amount of fertilizer is length of the piston stroke. Piston pumps are more deposited on the entire field. expensive than centrifugal pumps and operate at higher pressures. Piston pumps seldom have high enough output Pumps Used for Fluid for recircu~ationagitation. A separate centrifugal pump is Fertilizer Application often used for suspension agitation. Piston pumps can clog when pumping high potash suspensions at low rates. If the Five types of pumps are used for nuid fertilizer applica- pump is set above 3% of full stroke, however, clogging of tion: centrifugal, piston, hose (squeeze), roller, and potash suspensions should not occur. diaphragm. Centrifugal pumps are iiiost coininon for broad- Roller pumps are used for solution and herbicide cast applic:ltion because they ciln liandle high viilumes spraying. Roller pumps are vane-type pumps in which required for liigh speed application at the low pressures resilient rollers are used instead of sliding vanes. They can [IO to 40 psig (6.9 tu 27.6 x IO4 Pa)] required for bruad- be driven by power-take-off (P.T.O.) shafts and are casting. Centrifugal pumps are durable, inexpensive, and can cuminon on tractor-mounted spray rigs. Roller pumps liandle suspensions. Two disadvantages of centrifugd produce higher pressure than centrifugal pumps but are not pumps are tliat they must be driven at high speed (2.500 positive displacement. A disadvantage of roller pumps is to 3.600 rpm) and that they are nut positive displacement. that they are subject tu abrasion by suspensions. That is. output is not iiecessarily proportional tu pump Squeeze or hose pumps are very popular for band appli- speed. cation of suspensions. Hose pumps are much less expensive Pistoil piiiiips (figure 47) are often used for band appli- than piston pumps but are not as accurate. Since flow is catioii and are used fur broddcasting. particularly when the ;ipplicator is tractur drawn. Tlie slow rutational speed of divided in the pump manifold, flow dividers are not pistuii pumps niakes tlieni ideal liur driving uff a ground required. There are two types of hose punips in use. Figure wheel. Since they are positive displnccment, application 48 shows a pump using round hoses and a spring loaded rate is cutistarit when ground driven. regardlcss of driving backup plate. Figure 49 shows flat hoses and no backup speed. Rates of application are varied by changing the plate.

WTLET TO KNIVES OR SFUAY NOZZLES

Figure 47. Liquid fertilizer-piston type metering pump application of nutrients, "spoon feeding." is an agronomi- cally beneficial practice that is commonly done by fertigation. Sprinkler fertigation tests have shown that an ordinary tap into the irrigation piping system is all that is required to get uniform niixing of fluid fertilizer in irrigation wiiter (2). Tu ensure mixing, fertilizer is injected into tlie systeni just ahead of a tee or an elbow in tlie irrigation line. Turbu- lence in tlie water created by tlie pipe fittings aids in iiiixing. The velocity of water iii the irrigation line at tlie region where tlie fertilizer and water iiieet also affects mixing. A water velocity of 6 feet (I .87 m) per second is recnni- mended (2). A system designed to provide unifurm distribu- tion of irrigation water will also provide unifirrni applic.'I I'io11 of fertilizer if the two are well mixed. Injecting fertilizers into sprinkler irrigatiun system is Figure 48. Squeeze pump with round hoses and back plate easily accomplished but instiillittiiiii must include devices to prevent pollution of ground water. One nietliod tu prevent puniping of fertilizer when an eliginc driveii irrigation pump shuts down is tu drive the injectiun pump froni a belt on tlie irrigation pump. In inslallations having an electric innotor on the irrigation punip. the injector pump citn be powered by an electric niotor interlocked electricdly with the tinotor on tl~eirriga-. tion pump (figure SO). Tlierefure. injection of fertilizer stops automatically if the irrigation pumping p1:tnt stops. Note the vacuuiii breaker and check valve between the irrigatioii pump disc1i:irgc and the point of fertilizer injection into the irrigation pipeline. This pixvents back-siplioning uf fertilizer into the well if tlie irrigation puniping plant fails. Tlie check valve on tlie iiijection line stops flow of water frotii tlie irrigation system into tlie fertilizer tiink in tlie event the injection puiiip stops wliile tlie irrigation pump ciititinues IO run (3). 111 suiiie installatioiis, il inoimially closed solenoid vitlve irtterlucked electrically with the engine or motor driving Figure 49. Squeeze pump with flat hoses and hydraulic drive

Diaphragm pumps are used to inject fluid fertilizers into irrigation water. High pressures produced by diaphragm punips arc required to pimp against tlie pressure in irrigJ- tiuii lines. Recently. diaphragm pumps have been used for deep placenieiit of liquid phosphate in dual applicatioii. Tlie rotational speed rif the pimp is niatclicd to tractu~ P.T.O. shaft suced.

Applying Fluids Through Irrigation Systems Applying fertilizers in irrigiition \v:iter. fertigittioii. is becoming ninre popular and widespread because it is inex- pensive and agrunoniically efficient. Tlie cost rrf fertig~tiwi cat1 be 10 tiiiics less than cuiiventioiial iippliciiticin. I'eriodic Figure 50. Anti-pullution device on fertigation system

I I4 the injection pump is placed in the injection line. With this, circumference of the disk. The outlets are 114 to 318 in. neither fertilizer nor water can flow in either direction (0.63 to 0.95 cm) in diameter. Some of these manifolds through the injection line if the injection pump stops (3). are open. Others have spring loaded needle valves in each Use of fertilizer in drip irrigation systems is practical in outlet. The latter type is for solutions and will clog when areas where water costs are high or where high value crops using suspension fertilizers. are grown. Areas where the topography makes conventional Flow dividers designed especially for suspensions have irrigation impossible are also comnonly irrigated by a drip recently been introduced. One system uses a hollow cone system. nozzle mounted in the top of a pot (figure 52). The pot has One problem in fertigation with drip systems is precipi- from 8 to 15 equally sized compartments each of which tation of solids when aninioniuni phosphate is used in hard drains into a single line. Since the compartments drain water. Precipitates clog drip irrigation emitters making faster than they are filled, equal length lines are not fertigation difficult. Phosphates with high polyphosphate content or acidic solutions help reduce the formation of insoluble precipitates. Both approaches are effective in sprinkler irrigation systems; however, fertigation with hard water in drip systems still presents problems.

Surface Band Application Band application, traditionally done by farmers, is rapidly becoming a service offered by many fertilizer dealer/custom applicators. With current fertilizer prices favoring ammonia and dry fertilizer materials, fluid fertilizer dealers are using band application as a marketing tool. Though the benefits ofband application may be exaggerated by soiiie fluid fertilizer dealers, research has shown that banding is superior to broadcasting on many soil/crop systems. Dribble application is a common practice for pasture fertilization. Row crops are often sidedressed by dribble SOLIDSTREAM NOZZLE application and phosphate is soiiietirnes dribbled and disked I into the soil to enhance its effectiveness. Though dribble application implies that liquid is dribbled from an opening and falls by the force ofgravity, mile dribble applicators have iiorzles which produce a solid strcaiii under pressure (figure 5 1). Farmer operated equipment generally dim not have nozzles. Farmers often attach dribble equipment to a tillage implement. A ground driven squeeze pump delivers liquid through hoses with open ends aimed downward and equally spaced on a tool bar in front of disks or some other tillage devices. Ground driven piston punips are also comiiionly used for farmer uperaled dribble application. Pistoii pumps, unlike squeeze pumps, have a single outlet from which flow must he equally divided into as many tubes as there are dribble outlets. Several devices can accomplish this. On siiiall rigs ' the pump oullet can be divided and subdivided by ordinary plumbing littiny. The lines to each knife must be equal in diameter and length lo ensure uniform application. Larger rigs require more elaborate flow dividers. Several types of Inanifold flow dividers are available. Cylindiical MULTIPLE SOLID STREAM"NOZZLES or disk-like inianifolds are co~iimoii.111 these ~nianifdds tliere is typically a 3/4-in. (1.9 cni) or larger inlet in the Figure 51. Nozzles used for dribble center of the disk with 8 10 16 outlets spaced around the or surface hand application

115 FLOW DIVIDER

Y 45'STREET ELBOW NPT

-. OUICK COUPLER a HOSE BARB

0-100 DSI GAUGE

GAUGE DAMPER

3 INJECTION TUBES THREbD END PVC-- PIPE ST4ND

.IMPLEMENT FRAME

Figure 52. Flow divider for suspension injection

required. Since liquid flows by gravity froni the pot: extremely viscous suspensions may cmse flow stoppages. A secund type of suspension flow divider uses multiple urifices in series to achieve tlie low application rdtes desirable for band application (figure 53). Multiple orifices are made large enough to :illow passage of suspension. The flow mte of suspension through the series uf orifices. however. is much less tlian the flow of suspension through a single orifice of the same size. Broadcast spray equipment can be used for dribble application by replacing bro:idc;ist nozzles with stilid streani nozzles. This is done with many solution ;ipplicators wliicli otherwise use flat fan nozzles at close spacing. Multiple solid strean1 nozzles are also used (figure SI). These iiwzles produce two or three divergent streanis :ind :ire designed for solutions. Multiple solid stream noz%les iiiust be spaced priiperly 111 produce strips t1131 are uiiifornily spaced on tlie Figure 53. John Blue flow divider soil. for row application of liquids Suspensioils present problem because the spacing of nozzles for drihbling 112 to 30 in (30.5 tn 76.2 cm)] is this reason the flow dividers previously mentinned are 2 to 10 titlies narrower t1i:iii the nuzzle spacing for suspeti- coninioii for suspension dribble equipment. Both the pot sim broadmstiiig. Orifices iiiust. likewise. be snialler. For and multiple orifice system divide the flow in a manner

116 which permits application of fertilizer at low rates and (V-blades), hydraulic drives are sometimes used to drive narrow uutlet spiicing. fertilizer pumps (figure 49). One manufacturer has converted a high flotation broad- One subsurface applicator is unique in that it has no cast applicator to a dribble applicator by splitting tlie flow pump and relies on gravity to transfer liquid into the soil from each nuzzle intu two equal flows. The outlet spacing (figure 54). The gravity flow system is used for starter is reduced from 60 to 30 in. (151.4 to 76.2 cm). The strip fertilizer or row application. An air-tight tank behind each kits, as they are called. attach to tlie boom as do nozzles planter unit supplies liquid tu or near the seed furrow at a and produce solid streanis wluch are not affected by wind constant late regardless of the liquid level in the tank. A as are streains that are dribbled. This particular rig was cunstant head is maintained by allowing air wluch replaces ' designed especially for application of liquid phosphate liquid drained from the tank to enter the tank via a fertilizer in the Corn Belt. This practice is lypiciilly called breather tube which extends to the bottom of the tank. strip application. The small orifices used Io control application rates do not allow passage of suspensions. Suspension fertilizers must Subsurface Band Application be pumped in subsurface injection systems. Injection knives or chisels are used to inject liquids Subsurface injection equipment is sitnilar tu dribble or beneath the soil surface. Spargers or nozzles are also placed strip application equipment in tliat similar pumps and flow under V-blades to inject liquids with ammonia in dual dividers are used. Tractor drawn subsurface applicators application systems (figure 55). generally liiive ground driven piston or squeeze pumps. Though must subsurface banding is done by farmers Since ground wheels are not practical on sonle tillage tools with planters or tillage equipmcnt. sunie equipment manu-

AIRTIGHT T4NK LID

DOWEL LIQUID LCViL GkUGE /' PIN

VALVE

SEDIMENT J4R

DEL~YLRYTUBE

MLTLRlNG HE40 (~LCDETAIL)

Figure 54. Constant-head gravity-flow fluid fertilizer applicator and orifice plates

117 facturers have adapted custom. fluid broadcast equipment the U.S. Wheat Belt. Ammonia is generally metered at a for subsurface application. All custom subsurhce injection constatit rate by an adjustable orifice. The liquid ammonium equipment is high flotation and 4-wheel drive is sonietimes phosphate is pumped by centrifugal pumps on custom necessary to develop the traction required for pulling applicators and by pistoii or squeeze pumps 011 tractor injection knives through soil. A typical custom subsurface drawn units. V-blades are comnionly used on tractur drawn applicator has a tool bar which is between 30 and 40 feet units and separate lines for ammonia and liquid phosphate long (9.14 to 12.19 m), depending on tlie knife spacing. are attached to the underside of the V-blades. Ammonia The tool bar is generally attached to the applicator frame and phospliate outlets are in line and spaced I? to 15 in. and the knife or chisel depth is controlled hydraulically. apart (30.48 to 38.10 cni) under the V-blade (figure 55). V-blades are not used on custom subsurface applicators. When injection knives or chisels are used for dual appli- Unlike tractor drawn equipment, centrifugal pumps are cation, the phosphate tube is welded to the aninionia tube used on custom equipment. (figure 56). This arrangement often causes freezing of the liquid phosphate near the expanding ammonia. Conveying Dual Application phosphate through a plastic tube helps eliminate tlie freezing problem (figure 57). Dual application of anhydrous ammonia and liquid A variety of tank arrangements are used for dual appli- ammonium phosphate fertilizer is becoming widespread in cation. The ammonia tank is generally pulled behind the tillage implement on a trailer. The liquid phosphate tank is mounted on the tillage implement. pulled on a trailer. or twin tanks are mounted on either side of the tractor. High flotation custom applicators which itre adapted for dual application use the regular applicator tank for fluid phosphate and tlie ammonia tiink is trailered behind the applicator.

Aerial Application

The use of aircraft in agriculture has increased greatly in the past few ycars. Most aeriiil application involves the

VIEW UNDERNEbTH V-BLADE SHOWING LOCATION C+ TUBES

Figure 55. V.blade equipped with ammonia sparger Figurc 56. Dual application knife with' and liquid phosphate injection &scmbly ammonia and phosphate tubes separated

,I I8 Helicopters are becoming more popular for aerial appli- cation because they can reach areas inaccessible by plane and travel slow enough to supply fertilizer at high rates. Bi-winged aircraft are also common because they are easier to maneuver tlian single winged crafts and travel at slower speeds.

Trends and Recent Innovations The threat of high fertilizer prices has spurred interest in agronomically efficient application techniques. Incor- poration of nitrogen and placement of fertilizer near the roots uf plants are becoming more popular. Farmers and Figure 57. Ammonia knives equipped for dual application dealers are also shifting to energy efficient application methods. Equipment is available which can perform several use of herbicides and pesticides. Fertilizer. however, is field operations at once. Planting. fertilization, tillage, and applied extensively by aircraft on some crops. Rice and weed control can be accomplished in a single pass. citrus producing areas in the sootlierii regions of the US. Aniong the innovations which have recently been made are often fertilized by aircraft. Forest producers use aircr;ift in fluid fertilizer application are custoni applicators which to :tpply slowrelease nitrogen fertilizer (sulfur-coated can broadcast, dribble, or inject fertilizer beneath the soil urea). Since aircraft have limited carrying capacity, applica- surface. One such applicator is used for broadcast applica- tion of dry inaterials is more common tlian fluid applica- tion by day and for subsurface injection at night when the tion Aerial spraying, Iiuwever. is ciiinnion for correcting tillage marks enable the operator to see where he has been. micronutrient deficiencies or primary nutrient deficiencies Custom broadcast application equipment has been built that can be corrected by foliar feeding after a crop is to apply fertilizer at a speed of 2.5 acres per minute (I ha/ established. min). Boom movement is hydraulically dampened to permit Aerial application equipment is largely derived from application at higher speeds. Onboard microprocessors ground equipment. Centrifugal pumps and bypass rccircu- Inionitor applicatipn rate and acres covered and adjust lation are conimuii. Fixed wing aircraft and hclicoptcrs thc application rate automatic:~lly, providing extremely are coninionly fitted with booms and iiozzles at close accurate application. spacing. Flat fan nuzzles are coninion: however. the Pneumatic dry fertilizer and seed planting equipment problem of spray drift 113s turned interest to drift reduction is becuniitig more popular. Pneumatic systems are capable (hollow cone) ~iozzles. of distributing dry materials from a single hopper tu several Tlie difficulty in eliminating spray drift is illustrated by outlels spaced across a large implement. Tillage equipment the distances that spr:ly droplets can be carried from the and liquid fertilizer delivery systems are being coupled with path of aircrdft. A fine spray drifts 48 feet (19.63 in) in these pncuniatic planters. Tlie result is a precision planter/ a 3-mile(4.83 kin)-per-hour wind wliile Falliiig wily 10 feet fertilizer :ipplicatur that covers 60 feet (18.’3 m) in a (3.05 m). An aerusul droplet can travel 21 miles (33.8 kin) single pass. under tlie same conditions. To reduce spray drift, inie aerial sprayer ~iianuFxturer1i:is placed extra nuzzles oii tlie Tlie iiiaiiy innovations in fluid fertilizer applicatioli right side of the fuselage to help cumpensate for prnp indictite liow the science is changing tu incet the demands wash wliicli moves spp”:ty frciiii riglit ti1 Icft. Other mctlnids for higher pruductiun. Tlie threat uf higher energy and invulve wing iiiodifications ciii :iirpl:ines to modify their fertilizer costs will surely bring about further advancements wake arid nozzles wliicli control druplet size. in hid fertilizer application. Chapter 10 Mate r ia Is of Construction for the Fluid Fertilizer industry BY F. P. Achorn and H. R. Horsman

Corrosion costs the fluid fertilizer industry niillions of usually contain a corrusion inhibitor; however, corrnsiun dollars each year. Corrosion and erosion are functions of test results with sulutions that du not contain inhibitors temperature, pH, concentration, velocity. and solids sliow hat such sulutiuns are not currusive to most loading. Together corrosion and erosiun can destroy an aluminum alloys. stainless steels (300 series). or polyvinyl expensive piece of equipment in a short time. The most chloride (PVC) plastics. Corrosion tests by TVA showed susceptible to corrosion-erosion attack is equipment that that solutioii containing 32% nitrogen was less corrosive is expused to fast-muving fluids that contain solids. Liquid to carbon steel than more dilute UAN-solution or fertilizers are excellent electrolytes, and the currusion animuniuni initrate (AN) sulutions containing 19% nitrogen. products that form uti the metal slow down the corrosion Also urea sulutiuiis containing 19% nitrogen were less rate; but erosiun strips away the protection. and the result currusive than either the 19% AN or 19% UAN solution is accelerated metal loss. Carbon steel. the most curniiioii and about the xime as the 32% UAN solutioii (5). Carbon material of construction. is attacked by both nitrogen stcel exposed to these sulutiuns has been effectively solutions and fluid fertilizerniixlurcs. Aluminum is resistant inhibited with each ufthe fullowing inhibitors: to most nitrogen solutions, but is attacked by NPK mix- 1. Ammonium polyphosphate or aiiimuniuiii iirtho- tures. The stainless steels are resistant to both nitrogen phosphate (0.05% to 0.2% P205) solutions and mixtures. They are also resistant tu the fluids 2. Anhydrous aiiiiiiiinia in cunibination with micronutrients, herbicides, and pesti- 3. Ammonium thiocyanate (0.1%) cides. The disadvantage of stainless steel equipment is its 4. Sudiuni arsenite (0.1%) high initial cost. and sum alloys are subject tu pitting 5. Sirdiuni dichroniate (0.1%) attack. Corrosion-resistant plastics applications have beer The aniniuniuni phirsph;ites and ammonia are the most used successfully by the fluid fertilizer industry; I~owever, frequently used inhibitors since they are readily available they are susceptible tu impact damage and damage frum in the fertilizer industry and they are environmentally orgxnic solvents often used in herbicides and pesticides. acceptable. Ammuniuni pl~~ispi~atefornis a coating of iron The purpose (if this chapter is to supply infnrniation phosphate on carbon steel. pruviding a barrier tu further that might be used tu help select the must economical corros!on by thc nitrogen solution. Tests show that about material of construction fur each individual application. 10 pounds (4.5 kg) of aiiiiiionia will treat a 1011 of solution (increase pFI fruni 7.0 tu 7.5) and tlie treated sulution Nitrogen Solutions will !not corrode mild stccl. The data given iii table 29 SII~IW that AIS1 304 stainless slcel is usable at 80' and Much lias been written cwccrning corrosiiiii-resistant 120°F (27' to 49-C) cxpscd to AN (19% N) sdutiiin, and equipnient and procedures for handling noiipressurc at 80' and l?O°F (27" tri 49'C) expused tu UAN (31% N, nitrogen so1utioi;s 128%. 30%. and 32% N Iirea-aiiiiiioniuni pH 6.7). Also annealed carbon steel is superior to cold- nitriite WAN) sulutiuns] (I. ?. 3. and 4). Thesc solutions worked carbon steel.

I20 Table 29. Suitability of AIS1 type 304 stainless steel and mild steel for use in nitrogen solutions Suitability and corrosion rate (mils/yr) at indicated temperaturea AIS1 type 304 SSI Carbon steel at ambient tempD at 80" or 120°F Cold worked Annealedc Usable Rate Usable Rate Usable Rated Nitrogen solutions WITH corrosion inhibitorse AN solution (19.0-0 grade) 54.3%AN.45.7% H,O NR - - -

UAN solulioti (28-0-0grade) 30.5% urea, 39.5% AN, Yes - - - 30.0% H20. initial pH = 6.8

UAN solution (32-0-0 grade) 34.8% urea. 45.1% AN Yes - - 20.1% H20, initial pH = 6.8

Nitrogen solutions WITHOUT corrosion inhibitors Urea solutioti (19-0-0 grade) 41.3%urea,58.7%H20, NR 16' Yes 1 initial pl-l = 8.9. final pH = 9.4

AN solution (19-0-0 grade) 54.3% AN. 45.7% n20, NR 87-117f Yes I, PI11 initial pH = 8.5, final pH = 7.9

UAN sdutim (19-0-0 grade) 20.7% urea, 27.2% AN, 52.1% H20, NK 3SOf NR 22 initial pH = 6.4. final p1-l = 7.5

UAN solutioii (28-0-0 grade) 3?.l%iorea. 37.8%AN. 30.1%H20, NR 99f ? P3 initial pl-l= 8.0. final pH = 7.8

UAN solution (32-0-0 grade) Urea:AN ratio unknown. - - - - initial pH = 6.7. final pH = 6.9

UAN s~ilit~i~in(32-0-0 wade) NR 13-20f Yes PI11 initial pH = 8.1, filial pH = 6.4 ~- Ammonium Polyphosphate use of ball and plug cock type valves for flow control and Base Solutions throttling increases corrosion-erosion on the downstream side of the valve body and in the first foot or so of tlie Amnioniuln polyphosphate (APP) solutions of various downstream pipe. Control and throttling valves that direct grades (10-34-0, 11-37-0, 9-300, 12-44-0) contain from the fluid flow through the center of the valve body (gate, about 20% to 80% polyphosphate. Plant experience and globe, needle, etc.) into the downstream pipe minimize laboratory corrosion tests have shown these liquids can he corrosion-erosion damage to valve and pipe walls. PVC stored in carbon steel tanks without excessive corrosion has worked well in some piping of the plant; however, it resulting. Chemical and microscopic examinations show the was not tested in the pipes where severe corrosion occurred. APP solution reacts with the iron to form ferrous phosphate Corrosion due to high fluid velocities has occurred in film on the walls of the tank, protecting it from further many fluid fertilizer plants which use TVA's pipe reactor corrosion (6, 7). The film will deteriorate somewhat at the process for production of high polyphosphate solutions liquid-air interface on the surface of the liquid which tends (10-34-0 and 11-37-0). Since orthophospliate is believed to cause corrosion rings to form around the tank. Corrosion to he the filmfunning species, the corrosion may he due. rings are usually not evident except when the fluid is kept in part: to the low orthophosphate content (about at the same level for extended periods. This problem can 25%-35%) as compared to the liquids fornierly dominating be avoided by floating a film of oil on the surface of the the fluid industry (minimum orthophosphate 50%). liquid (8). Used niotor oil is excellent for this purpose and Chlorinated polyvinyl chloride (CPVC) and other PVC an oil film thickness of about 118 in. (3 mm) seems to work plastics have been used for years in the fluid fertilizer well. This formula may he used for calculating the gallons industry. They are lightweight and do not require a lot of of oil required for the I IS-in. oil film. equipment or highly skilled labor for installation. The most commonly used PVC is PVC-I. Its biggest disadvantage is Gal oil =(tank diameter. fee? its low impact strength. CPVC costs about twice as much 4 as PVC. PVC Type II is an excellent low cost construction iiiaterial for liquid transfer lines where tlie fluid temperature - tank diameter, meter - (- I .22 + does [not exceed 140'F (60°C). It is. llexible, yet has a Iiigh impact strength. APP solutions pumped through carbon steel pipes will I'VC resists corrosion by the polyphusphate liquids, but not corrode them if tlie fluid velocity is kept reasonably solile formulations lose more strength than others at low. Optiiiium fluid velocity is unknown hut most operators elevated temperdtures. Table 30 shows some tliermal and find that if the flow is kept below 8 feet (2.4 m) per second physical properties of PVC and CPVC compared to steel. corrosion is kept within tolerable limits. Corrosion increases CPVC is recommended where fluid temperatures exceed with temperature: thus. tlie solution should he kept at less 140'F (6OOC). This material resists heat distortion at than 150'F (66OC). At higher fluid velocities and tempera- temperatures up to ?IO°F (99°C) at about 260 Ib/in2g tures. the protective iron phosphdte film deteriorates and (I .8 x IO6 Pa). It does not have the impact strength of PVC corrosion of the Steel increases. Entrained air at high fluid Type 11. but it is an excellent plastic for use in hot-mix transfer rates may contribute to the increased corrosion. flu'id fertilizer plants if properly supported and protected. Cast iron is not subject to severe corrosion by APP solutions The support of all plastic lines is important and a dependable and it is recommended for impellers in transfer punips. niethod is to lay the pipe in an angle iron as shown in figure Solutions apparently do not corrode stainless steel (type 58. 316) until fluid velocities exceed 16 feet (4.9 111) per second. 111 TVA's prototype plan1 for the production of APP suspension, 9-32.0 grade. merchant-grdde phosphoric acid Table 30. Thermal and physical propcrtics of PVC (0-54-0) is reacted with ammonia in a pipe reactor to - and CPVC compared to steel pruduce :I fluid that contains suspended solids. Experience Property PVC-I PVC-II CPVC Steel in this plant has shown that carbon steel pipe and cast iron Gilicgr:ivity 1.40 1.35 1.55 7.80 pumps are nut serviceable. especially in tlie high fluid IZOD iiiipact strength. 0.8 8 3.0 30 velocity areits. Cast iron impellers. pumps. casings. and ft-lb/in. of notcha carbon steel piping have been severely damaged after only Ultilnate tensile 8,000 5,500 8,400 60,000- a few hundred hours service. Stainless steel inipellers strcngth, Ib/in2 90,000 and pump casings (Anierican Casting Institute CF-8M. Upper temperature I40 140 210 >l,OOO CD-4MCu) have shown only sliglit evidence of wear in tlie litnit OF Silllie a111ount of time. This operatioil has shown that the '',\SIX standard lc~ts.D 256 for plastic and li 23 for iiietals. Plastics have good overall balance of properties at least or intermediate coat. Not all coating systems require' an cost. intermediate coat, but one is often used to obtain the desired total film thickness. Also, three coats are required Liquid Mixtures to ensure that there are no pinholes. Using a different color paint for each coating is an excellent inexpensive method Chlorides in liquid mixtures corrode carbon steel but not to ensure complete coverage. Once the paint films are at a rapid rate. The main source of chlorides is from potash. penetrated and corrosion starts, the corrosion will spread Many con~panieshave found that carbon steel mix tanks on the metal beneath the film and the paint will start to with 1/4-in. (6 mm) thick walls have not corroded through peel. To ensure good coverage, some companies apply as after 5 years of operation. Carbon steel nurse and applica- many as five coats of paint. The tutal film thickness should tion tanks have been used for years in the application of be at least 5 mils (0.125 mm) for stationary tank walls NPK mixtures. These tanks will usually last at least 5 years. and 9 mils (0.225 mm) for mobile equipment. A relatively Some companies have found that they can extend the life simple method to ensure the desired dry mil thickness is of these tanks if they are frequently washed with water or to use a wet mil paint thickness gauge. The dry mil liquid fertilizer, such as 10-34-0 or 32-0-0, to lower the thickness is then determined by multiplying the measured chloride level on the surface of the metal. Others have wet mil thickness by the percentage solids supplied by the found that if the tanks are kept full ofliquid and a film of coating manufacturer. companies that plan to use these oil is floated on the surface, there is less likelihood of protective coatings should get specific instructions con- corrosion. Some companies fill their applicator tanks with cerning metal preparation and painting from the paint water or liquid fertilizer when they are not in use, thus suppliers. avoiding the corrosive chloride-air combination. Most 300 series stainless steels are resistant to corrosion Some companies coat the inside walls of their mild steel by liquid mixtures. There have been a few reports that tanks with corrosion-resistant paints. These three types of some 304 and 316 stainless steels have been corroded by paints have been effective. mixtures containing potash; however, others have reported 1, Epoxies and epoxy tars no corrosion of these types of stainless steel by NPK 2. Vinyl esters mixtures. Type 304 stainless steel is more susceptible to 3. Some polyethylene and phenolic resins chloride induced pitting than the niolybdenuin containing Proper preparation of the metal prior to painting is 316 and 317 stainless steels. essential for good service life. Most paint companies One steel company reported crevice corrosion of type recommend that the metal in new or old tanks be thoroughly 304 stainless steel by salts in the areas affected by fused- cleaned by sandblasting to white metal for immersion metal splatter during welding. They recommend that the sewice (Steel Structure Paint Council-SPS, National tanks be flushed with water after each use to prevent Association of Corrosion Engineers No. 1). After carefully buildup of salt deposits on the metal (9). There has been inspecting the metal to ensure that it has been thoroughly only one report of corrosion to type 316 and 317 stainless cleaned, a corrosion-resistant bonding primer should be steel. and these are recommended for use with all the applied. followed by application of a vapor barrier primer conventional fluid fertilizers. These two metals are also the most expensive. PVC plastic is an excellent relatively

,4"DlA.PLASTIC PlPE low cost material for pipes and tank liners exposed to fluid mixtures. Care should be taken to.select the type of PVC that suits the type of plant and process (hot-mix or cold- mix). Soinc companies store liquid mixtures in neoprene-lined nylon bags as shown in figure 59. This is a photograph of a rubber bag with a capacity of 500,000 gallons (I .9 x lo6 I) installed in a pit. If the bag ruptures, the liquid is trapped in the pit and CJII be pumped im~nediatelyinto another container. Some conipanies have used these bags to store pl~osphoricacid. Figure 60 is a photograph of a transport truck 011 which are installed several spherical type tanks made of polyolefin plxtic. These tanks are excellent for stordge of liquids and suspensions. The material tht tends to settle in the botio111 Figure 58. Support for PVC plastic pipe can be easily withdrawn and recirculated to the top ofthe

I23 sphere. No difficulties have been reported with deteriora- To avoid hgh cost, some conipanies store acid in less tion of the plastic. Some have been in use for as long as expensive fiberglass reinforced plastic tanks. Many tanks of 10 years. this type have given excellent service for more than 10 years but some have ruptured after only a few months of Storage and Transfer of Phosphoric Acid service. Many different plastic resins (PVC. polyolefins. polyesters, polyethylene, phenolics, and others) are used The container iiiost frequently used for storing with fiberglass reinforcement and with varied constructiori phosphoric acid is a rubber-lined carbon steel tank. The techniques. The resin provides the corrosion resistance, and types of rubber tilost frequently used are Butyl and natural tlie fiberglass and its application method provides the tank rubber. Type 316 stainless steel is also an excellent with its strength. The strongest vessels are wound with construction material for storage tanks and transfer lines. continuous strands of fiberglass. The buyer of tanks should Table 31 shows the suitability of various alloys for use in wet-process orthophosphoric acid and table 32 shows the investigate the perfornlance history of tlie supplier’s tanks suitability of variuus alloys for use in wet-process to ensure their reliability. Also, liiost wet-process phos- superphosphoric acid. Tanks with the rubber lining or phoric acid contains lluorine compounds which attack stainless steel are still in service after 20 years or more. fiberglass. The tank manufacturer can protect agditist this These require a high initial investinent. damage by applying several coats of the resin (sometimes

Figure 59. Neoprene-lined nylon bag (capacity: 500,000 gallons)

Figtirr 60. Transport trucks equipped with plastic tanks

I24 called a veil) over the fiberglass. Some lining niaterials In the past IO years there have been numerous pit tested by TVA in phosphoric acid are found in table 33. storage containers built for storing phosphoric acid. Occasionally. old tiinks are lined. When this is done. the Considerable care was taken with the construction of the tank walls should first be smoothed, sandblasted. and first pits to avoid puncturing the liner. Laler there were coated with a vapor barrier primer (see seclion on coating sonie con1pl:lints concerning ruptured pit linings and inside of tanks). Some coxting companies reconinlend a investigations showed that in must instances the lining was high build (about 50 mils or I .25 111111 tliick) spray applied not properly constructed or the support surface beneath butyl rubber system. the lining was not properly prepared.

Table 31, Suitability of various alloys for use in wet-process orthophosphoric acids Suitability and corrosion rate (mils/yr) ___ at indicated temperaturea 150°F 200°F 250°F Usableb Rate Usableb Rate Usableb Rate 2O%P,0,. .. acid AIS1 Type 304 sstc ~ - Yes - AIS1 Type 316 sst - - ') - Lead. 10% antinionial - - Yes -

29% P20, acid AIS1 Type 316 sst Yes 3 ') 9 - - AIS1 Typc 3 I7 sst Yes I YCS 4 - - Carpenter 20 Yes I Yes 5 - -

20%-40%P, Os acid Alloy 20 (cast. CN-7M) Yes I Yesd 5 - - Le;id, cliciiiical - - Yes

39%-56% P,Os acid AIS1 Type 304 sst No'

5l%.55% P,05 acid AIS1 Type 317 sst Yes 1 '! 6-8 NR 40-85 AL 6X Yes

60c/o-66% P: Os iicid AIS1 Tviie 3 16 sst - - Yes

I25 Figure 61 is a sketch of an excellent design for an acid Holes can be patched with a suitable solvelit and PVC storage pit. The pit bottom slopes at an angle ofabout 0.2 material. It is usually advisable when repairing the liner tu to 0.3 in. per foot (1.7-2.5 cmlnl) from one end to the place a patch on both sides of tlie hole so that there are other. Usually one end has a depth of 6 feet (1.8 m) and three thicknesses of PVC around the hole. the other end 8 feet (2.4 m). All four walls of the pit The acid is withdrawn from the sump end through an should be sloped at an angle of 45' or less so that no open-end pipe or a perforated pipe sparger. Companies that stress is applied to the liner as fluid is added. use the sparger find that as new acid is added to tlie pit, It is important that a drainage device be installed as the streanis from the sparger resuspend tlie solids that settle shown in figure 61 so that ifa rupture occurs, the acid will in the acid. Most companies cover their pits with an drain into a sump from which it can be recirculated until A-frame type of wooden roof covered with a standard the pit can be emptied and the liner repaired. This sump is composition material. Sonic cover their pits with an often used to detect small leaks by filling it with water and additional PVC liner which floats on top of the fluid. checking the pH of the water daily. A drop in pH indicates Figure 62 shows a pit with a floating cover. a leak. This design shows reinforcing wire on the bottom Some companies have used these pits as long as I2 years of the pit and drainage trench. Over the wire is a layer of to store phosphoric acid, UAN solution, and APP ,base sand Over wluch a layer ofasphalt is sprayed. The perforated solutions. They report no difficulties in the storage of these plastic pipe is laid in the drainage ditch and crushed rock is liquids in a properly installed pit. The estimated installation laid around the pipe. Next, the layer of asphalt and rock is cost (1975) of pit storage fur phosplioric acid was $34 per covered with sand so that there is a smooth surface onto tor of P,Os. This is considerably less than the S98 per ton which the PVC liner is laid. of P,O, for rubber-lined storage tanks (17.000 gallons) Prior lo installing the liner (usually 30 mils or 0.75 inm (6.4 x IO4 I). minimum thickness), it should be inspected for Roles. This can be done by first laying the liner in the pit and then Conduit blowing air between the sandfloor and the PVC liner to lift and buoy the film. A high intensity light directed on the PVC or PVC-coated conduit is the most corrosion- surface will reveal holes to an operator beneath the liner. resistant for use around fluid fertilizer plants. PVC.

Table 32. Suitability of various alloys for use in wet-process superphosphoric acida______Suitability and corrosion rate

AIS1 Type 304 sstd 70-74 Yes

I26 Schedule 40 and 80, is the only nonmetallic conduit at the Ammonia present time approved by the National Electric Code for use above ground. Carbon steel and aluminum need to be coated if exposed to strong free acids. Aluminum is not Carbon steel can be used for piping and tanks. Zinc, resistant to strong alkalies. tin, copper, and copper-base alloys should not he used

Table 33. Suitability of nonmetallic sheet materials, liners, and coatings for use in wet-process orthophosphoric and superphosphoric acids Suitability at indicated temperaturea 80'F 120'F I 5O0F 200°F 300°F Orthopliosphoric acids 3 I %-34%P, 0, acid Butyl rubber - - Usableb ?C ~ EPDM~ - - Usableb Usableb - I-lypalone Usable Usable Usable Usable -

54%.55% P, 0, acid Buta-bond rubber - Usable Us;ible Butyl rubber Usible Usable Usable Natural rubber Usable - ? Neoprene - Usable Usable Penton plastic - Usable Usable Polyethylene Type I - Usable - I-lypalone Usable Usable Usable

57% P, 0, acid Pe ii ton -

Superphosplioric acid 70%-74% P, 0, acid Liners, slieetg Natural rubber Usable - ? Butyl rubber Usable - Usable Neoprene Usable - Usable PVC II Usable-? - Usable-? Tygon Usable - Usable Polyethylene Type 111 - - Usable Polypropylene - - Usable - - Tefloii Usable - Usable Usable Usable coatings. appiiedg tleresite,," p-403 t L-66 Usable - Usable Usable - Neobon' Usable - Usable Nukeiiiite No. 401 Usable - Usable

127 where moisture may be present as they are corroded rapidly stress corrosion crxking of carbon steel. Vessels that are by wet amnionia gas or aqua ammonia. stress relieved after fabrication are less susceptible to Most alloys of aluminum are resistant to corrosion by stress cracking than those in the stressed condition (cold anhydrous ammonia and aqua ammonia except for dilute formed and welded). Installation of elbows iinniediately aqua ammonia (1%-5% NH,); this corrosion can be downstreain of restriction orifices should be avoided for controlled by the addition of nitrates (10). Aluminum anhydrous ammonia. alloys that contain copper should be avoided, especially the 2000 series that contains copper as the principal Fluid Fertilizer Additives alloying element. Anhydrous ammonia does not corrode carbon steel at Fluid fertilizers are an excellent medium for the atmospheric temperature. However, contaminants such as application of micronutrients and pesticides. They are carbon dioxide and air (oxygen) in ammonia cause stress usually added to the mixture in the applicator; therefore, corrosion cracking of carbon steel (1 I). Some mechanical there is some concern about corrosion ti1 the applicator. properties of steels used in tests (1 1) to show the effects If pesticide-fertilizer mixtures are allowed to stand in tanks of strength and hardness on their susceptibility to stress for an extended period of time. they inlay separale (float corrosion cracking are shown in table 34. on top or fall to the bottom). Also they may soften sonic The high-strength steels are more susceptible to cracking plastic tanks. especially those thal luve additives that than medium- and low-strength steels. The addition of 0.2% depend on organic solvents to stay in solution. Also, many minimum of pure water to anhydrous ainmonia inhibits additives contain fluorine and chlorine which become ionic

4'PVC PERFORATED WYP PlPEtYIITO

Y

w Figure 61. PVC lined earthen phosphoric acid storage pit I28 when the additive degrades, and they contribute to pitting Some of the coolers liave been redesigned, as it1 figure 64, attack in nietal equipment. Copper compounds cannot and packed with PVC plastic Intalox saddles or Pall ring be used safely in aluminum tanks because of the unfavor- packing. The plastic will not deteriorate and the saddles or able galvanic relationship between copper and aluniinuni. rings provide a large evaporative surface area per unit Usually ino corrosion problems can be attributed to the volume. micronutrient compounds that are added to fluid fertilizer.

Cooling Equipment

During the past 5 years, several liquid fertilizer producers liave used evaporative type coolers that are normally constructed for the refrigeration industry (figure 63). The coolers are packed with redwoud slat-type packing which lasts for about 5 years in a liquid fertilizer plant.

......

Figure 63. Evaporative-type cooler (from refrigeration plant)

Figure 62. PVC lined storage pit with floating PVC plastic.cover

Table 34. Mechanical properties of steels used in stress corrosion cracking tests Yield Tensile Strellgtll. Stlellgtll, Ib/in2 Ib/iii2 tlardness ASTM A 285 37.000 59.000 RU66 ASMI' c~se1056 48,000 79,000 RB85 ASTM A 2 I? grade B 5 I ,000 83,000 RU84 NOZZLE ASTM A 202 grade 13 57,000 93.000 RB94 * P4CKlNO TO BE IYTbLOX rn LDYL T-l steel 116.000 125.000 RC 35 I AIS1 4130 I19.000 134.000 RC 29 Figure 64. Evaporative cooler (with plastic packing) Summary Steel Founders' Society of America (SFSA) 2061 I Center Ridge Road This chapter by no means covers all of the corrosion- Rocky River, OH 441 16 resistant materials used in the fluid fertilizer industry but Steel Structure Paint Council (SSPC) should serve to highlight important materials and practices 4400 Fifth Avenue for reducing troubles with corrosion and erosion. Those Pittsburgh, PA 15213 personnel responsible for choosing the materials of construction in this industry should weigh all the informa- All sources of information should be utilized since tion that is available to make suitable and economical minimizing corrosion lowers operating costs by minimizing selections of materials. The manufacturers of equipment maintenance, limiting unscheduled downtime. and and materials continue to develop new materials and maximizing equipment life. methods that give each construction material better corrosion and erosion resistance. All carbon steels, aluminunis, stainless steels. coatings. and liners are not References equally resistant even if they have the same generic name. Different manufacturing methods and formulations give I. Wynia. B. H.. and R. H. Wagner. "Aluminum Usc in tlic different characteristics. The one recommended for each I-crtilizcr Industry." Clzer,iicol Eirpiricrririy Pn,xrers. 60 (3). specific application should be selected. Beside textbooks (May 1964). and reference books that provide general informatioti on 2. Sauchelli. Vinccnl. Fcrrilizer Nilrogm ~llo,mppi!,Scrics No. each material, the manufacturers provide specific 161. Rcibald. information about the performance of their products. 3. Honk. \\I, P.. "Inhibiting Mild Sled in NitroScn I~ertili~ers." :vorcrio/ ~rorecrim.vot. 7. pp 35-7 (~Marc111968). This information is provided through the home office 4. Rerr. H. 1.. "Nitrogen Solutions." NFSA Liquid I:ertilircr or their local representative. Other organizations will Roundup. Sf. Louis. hlinsou". pp 35-8. July 11-12, 1967. provide useful material properties and application 5. Crow. G. L.. and H R. Hurmian. "Hot-Rolled Mild Sled information. Some of these org:inizations are: Recumniendcd fc~r Nitrogcn Sulutions Slurage Tank." SO!UI~WI. 22 (3). pp 62-70 (M~Y-JWW1978). Aluminum Association (AA) 6. Crow. G. L., and J. Silverbcrg. "Corrusion of Mild Stccl by 818 Connecticut Ave.. NW Liquid I:crtilizm." Fmilixr So!uriom lpp 8-1 3 (January- Washington. D.C. 20006 Ikbruary 1967). 7. Hatficld. J. D.. A. V. Slick. C. L. Crow. and H. 6. Shiffer. National Association of Corrosion Engineers (NACE) "Corrosion of Melds by Liquid hlixed IFcrlilizcrr." P.O. Box 986 ilyricdfsrolorid FoodC'1~o~i.q~.Vol. 6. p 524 (1958). Katy, TX 77450 8. P:dgravc. D. A,. and C. I). Smith. "Corrosion of Mild Sled National Associatioil of Plastic Distributors (NAPD) by fatilizcr Sdutioas." J. Ayri Luyimcriny Res. Vol. 17. Gilson Road lip 236-45 (1972). 9. Smith. K.. '~Prwcnting Corrosion of Sktinlcss-Steel Liquid Jaffrey. NH 03452 I..L~tilizer .' Tanks." Ayricidrrrro! iVirroRoi i\'en,s. Vd. SIX, National Fertilizer Solutions Association (NFSA) p 41 (Noveinbcr-Uuceiiiber 1969). 8823 North Industrial Road IO. I:lournoy. R. W.. "Aluinimini-Its hporlml Role in 1:crtilizer Peoria. IL 61615 Sulutions Eqoipmcnt." Furilizcr Solurims, 8 (I) pp 6-14 (J:inu3ry-I:cbruary 1965). Society of the Plastic Industry (SPI) 11. Laginow. A. W.. and E. H. Phclps. "Stress-Corrosion Crxking 355 Lexington Avenue of Slccls io Agricultural Ammonia." C~ro~io?!,18 (8) New York. NY 10017 pp 2991-3091 (Augiist 1962).

I io NATIONAL FERTILIZER DEVELOPMENT CENTER Bulletin Y-185 TENNESSEE VALLEY AUTHORITY TVAIOACD-841 MUSCLE SHOALS, ALABAMA 35660 September 1984

TVA is an equal opportunity employer, and is committed to ensuring that the benefits of .-programs receiving- TVf financial assistance are available to all eligible persons regardless of ;ace, color, national origin, handicap, or age