I-•
Production of
Granular NPKs monium Phosphate Plants
Some Important Differences
International Fertilizer Development Center
Production
Granular
NPKs in Ammonium
Phosphate Plants
Some Important Differences James J. Schultz
International Fertilizer Development Center
Preface The plant design and operating parameters required fo producing agglomer ated compound (NPK) fertilizers, especially urea-based and other temperature sensitive and hygroscopic granular products, are often quite different from those required fo the production of conventional ammonium phosphate fertilizers, i.e., diammonium phosphate (DAP) or monoammonium phosphate (MAP). These differences should be carefully considered when adapting DAP/MAP plant for th production of NPKs. The plant design and operating criteria described in this bulletin are intended to serve as "check list" that be used to help guide those involved in th planning and design of ne NP projects or the modification of existing units. The data and recommendations contained in this bulletin re th distillation of broad international base of experiences and results collected and formu lated over number of years. The author is grateful to th valuable insight offered by th many production, engineering, and research/development organizations and individuals ho candidly shared their experiences and knowledge, offered suggestions, and otherwise stimulated and supported th preparation of this publication.
Library of Congress Cataloging-in-Publication Data Schultz, James J., 1936- Production of granular NPKs in ammonium phosphate plants-some important differences. (Technical bulletin International Fertilizer Development Center T-36) Includes bibliographical references. 1. Nitrogen fertilizers. I. Title. I1.Series: Technical bulletin (International Fertilizer Development Center) T-36. TP9634.N5S38 668'.624 89-26790 ISBN 0-88090-084-9
International Fertilizer Development Center P.O. Box 2040 Muscle Shoals, Alabama 35662 Phone o. 205-381-6600 TWX-810-731-3970 IFDEC MCHL Telefax: (205)381-7408 Edited by E. N. Roth Typesetting and Layout by F. Sandlin Cover Design by F. Rudolph Graphics by T. L. McGee IFDC publications re listed in Publications of the international Fertilizer Development Center, General Publication IFDC-G-1, which is available free of charge.
Table
Contents
Introduction ................................................................................................................ Im portance of ranule Form ation echanism ................................................... Agglom eration-Type Processes ........................................................................... Accretion-Type Processes ..................................................................................... Equipment and Operating Parameters Unique to Agglomerated NPKs ....... Ra aterial Particle Size ............................................................................... Liquid Phase ....................................................................................................... Estim ating Liquid Phase Values ..................................................................... Heat of he ical Reaction ............................................................................ Insoluble Binders ................................................................................................ Liquid Phase C ontrol ........................................................................................ Acid/Am onia Neutralization ethods ......................................................... Preparing th Production Form ulation ........................................................... aterial Feed System ....................................................................... Dry Ra Solids Recycle .................................................................................................. ranulator ......................................................................................................... Cocurrent Dryer ............................................................................................... Countercurrent Process Cooler (Second-Stage Dryer) ............................... Screens .............................................................................................................. O versize rushers ............................................................................................. Product ooler .................................................................................................. Conditioning .................................................................................................... Storage and Bagging ...................................................................................... ................................. Process Plant Dehum idific ji r, ....................................... Recommendations Specific to the Production of NPKs Containing Urea ....... Verification of Recommended Plant Design and Operating Parameters ...... Conclusion ........................................................................................................... Appendix-Design and Operating Criteria fo Urea-Based NPK ranulation Plants ........................................................................................... Bibliography .........................................................................................................
1 I 1 1 2 3 3 4 4
5 5 5 7 8 10 10 10 12 14 15 16 16 16 16 17 17 18 20
Introduction Conventional ammonium phosphate granulation plants (DAP and MAP) equipped with atmospheric or
pressure tank-type preneutralizers re often rot ideally suited fo th production of granular NPK fertilizers, especially those NPKs that contain relatively large amounts of urea or other very soluble, hygroscopic, and temperature-sensitive salt mixtures. As result, many combination DAP/NPK plants, although well suit or MAP, often have difficulty in ed fo producing achieving th owner's expectations when producing NPKs. This bulletin describes th fundamental differences between th granulation of DAP/MAP and tiat of most NPKs and discusses th plant design features most often needed for routine production of highquality NPKs. In general, it is less difficult to produce DAP or MAP in plant designed to produce NPKs than to do th opposite, provided, of course, that th NP plant is equipped with systems fo neutralizing large quantities of phosphoric acid and recovering ammonia. The PK fertilizer granulation plant design and operating parameters discussed in this bulletin ar intended to describe, in general terms, th mechanical and process conditions that should be et to achieve an optimum level of plant performance and NP product quality. None of th indicated criteria ar absolute; there may often be much latitude fo modification/ variation, depending upon th specific requirements with respect to (1) product formulations (grades and ratios); (2) ra materia; quality and related physical and chemical properties; (3) skill of th plant operatars; (4) expected operating rate (capacity utilization); (5) expected product quality and related regulatory criteria; (6) method of packaging, storage, and use; (7) length of storage between production and use; and (8) host of site-specific, and often variable, tors including climatic conditions, infrastructure, environmental impaci criteria, cultural practices, and
market requirements. An examination of selected papers indicated in the bibliography with numerous other literature sources cited in these papers will provide useful complement to this bulletin.
Importance
Granule Formation Mechanism
The method of granule formation ha pronounced impact on th design and layout of th process equipment. Therefore, good knowledge of th primary mechanism of granule formation, growth, and consolidation is essential in determining th de sign features of the plant. Although more compre-
hensive description of the machanlsms of granule formation can be found in th literature, th folow ing is a brief description of th tw major granule for mation mechanisms, excluding drop formation frmation eredri ecount (priin processes.
granulation
Agglomeration-Type Processes PK products (excluding th slurry based nitrophosphate-type processes), agglomera tion is th principal mechanism responsible fo initial granule formation and subsequent growth (Figure 1). In most NPK formulations, 50%-75% of the raw materi al are fed as "dry" solids. These solid particles are assembled and joined into agglomerates (granules) combination of mechanical interlocking and by cementing-much as stonemason fashions stone wall by using stones of varying size and shape and mortar as th cementing agent. The cementingmedi um for fertilizer granules is derived from salt solutions, fo example, preneutralized ammonium phosphate slurry and/or th dissolution of salt on th surface of th soluble solid particles. The size, shape, surface tex ture, strength, and solubility of th solid particles vary widely (Figure ?) and have major influence on th granulation characteristics of the mixture.
With most granular
Accretion-Type Processes
The slurry-type granulation processes (DAP, MAP, TSP, and some nitrophosphates) are quite different from the processes used fo most NPKs with respect to the mechanism of granule formation and growth. As result, th required process parameters fo optimum operation of these slurry-type plants ar also often quite different. With slurry-type granulation process, relatively thin film of moist slurry is repeatedly ap plied, dried, and hardened to a. relatively firm sub strate consisting of granules that ar often product size and/or nearly product size. Granule growth is primarily by accretion, th process in which layer upon layer of ne material is applied to particle, giving th final granule an "onion-skin" structure (Figure 1). In the process, of course, some agglomer ation also occurs, but this is no th predominant mechanism. The recycle-to-product ratio required for accretion type granule development isnormally higher than that required for agglomeration-type processes. c cordingly, for given production rate, the material handling capacity of the equipment must be larger fo accretion-type granulation plants than fo most agglomeration-type plants. However, because of oar ticular temperature-related processing requirements fo some agglomerated NPKs, certain equipment,
Photo of Actual Fertilizer Granule (Agglomeration) Scale:
Photo of Actual Fertilizer Granule (Accretion)
1m
Scale:
mm
Salt Bridges Between Larger Particles.
Insoluble Binders Ca
LyTp Layer-Type
Often be Used to Strengthen the Salt Bridges Between Particles.
Granule Growth
Granule Formed by Agglomeration
Granule Formed by Accretion
Figure 1. Principal Mechanisms for Granule Formation.
dryer and ocoler, may actually be larger in some agglomeration-type plants to achieve
Equipment an
processes. This is discussed in greater detail later. Granules formed by accretion ar almost always harder, more spherical, an more durable than those formed by agglomeration. Fo example, typical well-formed DA or TS granule produced mainly by accretion-type granulation ma have crushing strength of about 4- kg, whereas the crushing strength of an agglomerated granule ma no only be less (often less than kg) bu also more variable depending upon its raw material composition an number of specific factors related to granule forma tion. Some examples of the variability in granule strength for number of fertilizers ar shown In
discussion of the key plant design an processing parameters required for the production of agglomer ated NP products follows. Specific reference is made to the special features needed for the production of most urea-based NPKs an other NPKs that exhibit simi lar high solubility, temperature sensitivity, an critical relative humidity (CRH) characteristics. Th major devi ations from the recommended equipment an oper ating criteria normally found in conventional DAP/MAP
especially th
Operating Parameters
Unique to Agglomerated NPKs
the same production rate as in the accretion-type
plants ar
discussed.
i. Critical relative humidity Isthe ambient relative humidity below which mGcerlal loses moisture to the atmosphere an above which the material ab sorbs moisture from the atmosphere. The CRH of material varies with temperature.
Table 1.
2
cl
Table 2. Typical Particle Size Data for Commercially Available
Table 3. Solubility of Common Fertilizer Salts
Materials Frequently Used to Produce Granular NPKs
Approximate Concentration of Saturated So~utlon .t
Percent Retained on Indicated Mesha 4
16
48
--
-Indicated
16 28 48 10 (4.75) (2.36) (1.00) (0.60) (0.30) (0.15)
aterial Ammonium suhiute (crystalline) Diammonlum phosphateb (granular] Monoammonlum phosphate
0
22
98
(nongranular)
Potassium
(standard)
Potassium chloride
100
Urea (prilled)b
100
100
65
90
55
90
99
Potassium sulfate
0
98
30
(special standard)
(standard) Single superphosphate (run-of.pile) Triple superphosphate (run.of-pile]
75
35
90
15
75
95
100
11
25
36
52
76
14 15
19 92
26 95
51
80 00
...
a. Tyler mesh, opening size in mm indicated In parentheses. b. Crushing of these materials preferred to obtain more uniform
(homogeneous) NP
..
product.
Fertilizer "alt
Ammonium nitrate Ammonium sulfale Diammonlum phosphate Monoammonlum phosphate Potassium chloride Potassium nitrate Ootassium sulfate Urea
---
Temperature
20°C
100oC
(%] 54 41 30 18
22 12 41
66 43
90 51 58 63
41 27
25 24 10 52
36 71
19 88
a. Values indicated are fcrpure salts.
300 kg/tonne of product is about optimal fo most agglomeration-type NPK formulations. Of course, it should be that phase, while im
appreciated
liquid
portant in granule formation is only one of th criter ia that must be carefully evaluated when estimating the granulation characteristics of particular
formulation.
added to the granulator, for example, an ammoni a d d d ophae foxlutens of um phosphate slurry an~d/or solutions of urea or ammonium nitrate, and (2)dissolution of small portion material on th surface of th soluble ra material and recycle part!cles. This dissolution iscaused th combination of heat and water contained in th above solutions or steam or water Injected into th granulator. SolubiNiy data for some of th more comon fertilizer salts are given in) Table 3. Liquid pl-A.se
control is the key to achieving the desired level of granulation efficiency and product quality. Ideally, after drying, th liquid phase (salt solution) forms
strong crystal bonds (salt bridges) between th
porti-
of the agglomerate that are mechanically well. fitted and interlocked-again, th concept of stone les
wall.
Table 4. Liquid Phase Factors for Selected Materials Freluently Used InNPK Granulation (Agglomeration) Formulas INKGnlo
Material
.kglkg Anhydrous ammonia
Ammonlalammonium nitrate solutions (various compositions) Ammonium nitrate (prills) Wet-process phosphoric acid Sulfuric acid
Experience has shown that there are considerable
differences in th amounts of liquid phase generat ed by th varicus materials normally used in th production of agglomerated NPKs. Several years ago (during th 1960s), th Tennessee Valley ALthority (TVA) examined wide variety of PK production formulations that were known to perform well and devised numerical value to express empirically the "apparent" liquid phase contribution that could be expected from number of materials commonly used to produce NPKs. These "liquid phase factors" re shown in Table 4. Experience ha shown that, when these values are used as guide, liquid phase of about
0.50 1.00 0.3U 1.00
1.00 1.00
Ammonium sulfate (crystalline) Single superphosphate (run-of-plle]
0.10
Triple. superphosphate otassium (run-of-pile) chloride (coarse or granular)
Potassium chloride (standard)
Diammonlum phosphate (granular) Monoammonlum phosphate
Values
Liquid Phase Factor
Superphosphoric acid ater or steam
Urea (prilled)
Estimating Liquid Phase
Water
To obtain the total weight of the liquid phase in
the weight
of
2.00 0.10
0.20
0.30 0.00 0.25 0.20 0.30
forniulatlon, multiply each raw material In the formula (ki) by the appropriate
liquid phase factor. total liquid phase weight value of about 300 kg/tonne is considered optimal in many cases
Heat of Chemcal Reaction The level of liquid phase is also closely allied with another criterion, i.e., the expected amount of heat created by various chemical r9actions In given NPK formulation. The amount of heat generated, particu larly within th granulator, can have marked effect on th amount of liquid phase formed and, therefore, th resulting granulation characteristics of th mix ture. In general, to achieve optimum granulation, the calculated total liquid phase for formulation, using
th data in Table 4, should be lowered if the formulalarge amount of chemical heat of tion produces reaction. However, th optimum relationship between liquid phase and heat of reaction for specific formulation must experience,
be
learned
from
actual
operating
The most important heat-generating chemical reaction in most NP plants is th neutralization of acidic materials with ammonia. The approximate amount of heat released when ammonia reacts with some cornon fertilizer materials is shown in Table 5. As with Ifquid phase, experience has shown that if th amount of heat released in the granulator isequivalent to about 45,000-50,000 kcal/tonne of product, conditions ar generally good for obtaining optimum granulation. Of course, like liquid phase, the proper level of heat is just another ne of th many critical criteria that must be me to obtain optimum granulation efficiency.
Insoluble Bnders In some cases, th mechanical and crystal (salt bridge) bonding of particles can be greatly improved by addin about 5%-15% of finely divided insoluble binder powder, for example, kaolin clay, to th granulating mixture. The binder powder helps to fill th many small voids between th particles nd acts much like satu rated wick in helping to join th particles together (Figure 1). This concept works particularly well with NPKs containing large amounts of crystalline ammonium sulfate, potassium chloride, potassium sulfate, and/or kieserite (magnesium sulfate monohydrate) and relatively low levels of suitable soluble salts or solid binders such as ammonium phosphate o. superphosphate. It is important to note, however, that some inso!uble binders (clay, for example) have th capacity for retaining moisture thus making subsequent drying more difficult.
Liquid Phase Control
formulations, most of th liquid phase is obtained from materials that ar Introduced to the process final product with the fixed rate to achieve at In al NP
desired chemical analysis. The resulting liquid phase can, of course, be adjusted within rather specific limits through th selection of ra materials or by controlling th free water content and/or chemical composition of
mole ratio as the slurry (for example, the NH :H P0 shown in Figure 3). However, once this Isestablished (op-
cons flow rates of th liquids must timized), tant to ensure th correct analysis of the final product. For this reason, al agglomeration-type formulations moderate degree of liquid phase should allow for "tuning" performed by th operator using steam and/or water fe directly to th granulator. The discretionary small amount of steam and/or water by the use of operator helps to compensate for variations in granula tion efficiency caused by changes in th temperature of th materials, quantity and particle size of the recy cled material, and minor (but normal) upsets within the overall processing system. This fine tuning of th process th operator is th basis for th observation: "NPK ar than granulation by agglomeration is more of science." The unique skill of an experienced granu lator operator often greatly overrides th effectiveness of th most costly and sophisticated process design and engineering skill.
Acid/Ammonia Neutralization Methods indicated earlier, th
acid/ammonia neutralization
reactions create heat that contributes to the overall Ii quid phase conditions in th granulator and th effi ciency of th granulation process. Thus, th method used for neutralization (reacting ammonia) can signifi cantly influence th overall performance of th plant. brief discussion of th most common methods used for neutralization in agglomeration-type NP granula follows. tion Direct Neutralization in Granulator-This wa on of th most common methods used for reacting am monia in th many NPK granulation plants that were operated in th United States and elsewhere during th 1960s and early 1970s. Direct neutralization in the granulator is particularly well suited fo NPK grades containing large amounts of superphosphate (SSP or and relatively lo level of nitrogen. With direct
Table 5. Approximate Amount of Heat Released Wh in Ammonia Reacts With Various Materials Commonly Used to Produce Granular NPKs Heat Released
Material Reacted With Ammonia
Reaction Product (solid)
NH3 Gas ............
Wet-process phosphoric acid (54% P20 Wet-process phosphoric acid (54% P20s) Monoammonium phosphate (MAP) Triple superphosphate (TSP)
Monoommonium phosphate (MAP] Diammonium phosphate (DAP) Diammonium phosphate (DAP)
Single superphosphate (SSP) Sulfuric acid (100°o}
Ammoniated SS Ammonium sulfate
Ammoniated TSP
NH
Liquid
-callkgH, reacted)
1,890 1,510
1,370
1.130
610
1,580 1,460 2.165
1,060 940 1,645
90
Degree of Ammonialion, kg NH kg Pg0 0.1
0.2
0.3
degree of reaction performed In th
0.4
0.5
1,000,
goo-
., 00NH3:H3P4 600.
content of th reacted slurry. The neutralization reac tions that are only partially completed In th preneu tralizer are completed in th granulator where additional ammonia is added beneath th rolling bed of material. When sulfuric acid Is reacted In combination with phosphoric acid in the preneutralizer, special pre
500
400-
E
300-
C 200
cautions must be taken in the selection construc tion materials that will resist th more corrosive nvironment caused by th sulfuric acid. Also, at the
100.
0
0.2
0.4
0.
0.8
1.0
NHj:H
PO Mole Ratio
1.2
1.4
preneutralizer Is number of factors; however, th most Important criteria are th production of fluid slurry that is easy to transport (pump) to th granulator an uniformly distribute onto th rolling bed of material In the granulator. The fluidity of th preneutralized slur ry is maintained through careful control of the ole ratio, temperature, and free water determined by
1.6
1.8
2.0
2.2
Figure 3. Effect of Mole Ratio on Solubility of Ammonium Phosphate.
neutralization, th best operation isusually obtained Ifth aiiount of ammonia reacted in th granulator does ot exceed th equivalent of about 50 kg/tonne of product. In th direct neutralization process, ammonia is distributed beneath th bed of material in th granulatar. If sulfuric acid is used, it to is usually distributed beneath the bed of material while th phosphoric acid, Ifused, is most often sprayed or dribbled on top of the bed. When sulfuric acid is used, precautions should be taken to ensure that th acid is added at particular location with respect to th ammonia to en sure quick and uniform neutralization and thus minimize th unwanted reaction of sulfuric acid with potassium chloride that most often is also present in th formulation. This reaction causes th formation of very corrosive hydrochloric acid, which reacts rapidly with ammonia to form dense fume of ammonium chloride that Is very difficult and costly to collect In th plant's emission control (scrubbing) system. Tank-Type Neutralizers-The use of atmospheric
or pressurized tank-type neutralizer offers maximum flexibility in managing th acid/ammonia reactions and obtaining th critical heat/liquid phase criteria needed fo good granulation when producing wide variety NPK grades. Because th acid/ammonia reactions are only partially completed in these tank-type neutralizers, they are often referred to as "preneutrallzers." Such preneutralIzers re commonly used In the majority of today's DAP plants. When preneutralizer Is used, large amounts of acid can be partially reacted with ammonia. The
higher mole ratio and pH (for example, about 1.5 and 6.8, respectively), the presence of ammonium sulfate
crystals tends to thicken th slurry and make pumping
difficult; thus, operation at lower NH :H P0 mole ratio and pH (about 0.4 and 2.0, respectively) is often
preferred. Pipe-Type Reactors-In th mid-1970s, TVA demon strated th feasibility of replacing nonventionrl tank-type preneutralizer with novel device referred to as pipe reactor. This type of reactor wa radical departure from th conventional tank-type preneu tralizer normally used to react large amounts of am monla with phosphoric acid. The pipe reactor consists basically of length of corrosion-resistant pipe (about 3-6 long) to which phosphoric acid, ammonia, an often water ar simultaneously added to ne en through piping configuration resembling tee, thus th name "tee reactor." The acid and ammonia react quite violently, pressuring th unit and causing th su perheated mixture of ammonium phosphate slurry ("melt") and water to forcefully discharge from th op posite end th pipe that is positioned inside the granulator. Uniform distribution of th "melt" on to of th rolling bed of material In th granulator Is achieved by varying the configuration and orlenta tlion of th discharge opening of th pipe. major ad vantage of th pipe reactor over conventional preneutralizers Is that It more effectively utilizes chemIcal heat of reaction to evaporate unwanted water from th relatively dilute acid. Later, th tee reactor wa modified by TVA to also accept an additional flow of sulfuric acid through another pipe inlet located opposite th phosphoric acid Inlet, giving th unit "cross" configuration an thus th name "pipe-cross reactor" (PCR). Use of th PCR makes it possible to react wide varl ety of phosphoric/sulfuric acid mixtures with ammonia. This capability Is particularly useful In granulation plants and allows greater th selection of
ra materials to Improve granulation and optimize th overall cost of production. In general, th mixture discharged from the PC does no require further reaction with ammonia in th granulator. In some cases, however, th level of reaction in th PC ay be altered (decreased) to minimize the escape of ammonia or to obtain improved granulation characteristics of th "melt" when it is combined with the solids in th granulator. Several variations of pipe-type reactors (and materials of construction) are currently used in NPK, DAP, and MAP plants, sometimes in combination with conventional tank-type preneutralizers. Perhaps on of the greatest advantages offered by th use of pipe reactor technolog in th NP industry is that it provides an opportunity to effectively use greater variety of raw materials including, for example, larger quantities of dilute acids an scrubber liquor. This added flexibility in raw materichoices can often result in more favorable production costs and at th same time provide a method for Table 6. Examples of Typical Commercially Proven NP Formation and Growtha
3-9-18
5-20-20
Preparing the Production Formulation According to th foregoing discussion, large number of ra material and process variables must considered when developing NP production formulations. As with th operation of an NP plant, formulation too requires considerable amount of skill and an element of "art" to ensure that th particular formulation will yield th desired results In given plant. given NP fertilizer can formulated in many ways depending upon the available ra materials an specific equipment system. Table shows some exam pies of NP production formulations that have been
Production Formulas In Which Agglomeration Is the Principal Mechanism for Granule
Direct Ammonlatlon In Granulator
Material as Fed to Processo
disposing of certain "problem materials" such as ex cess scrubber liquor. It should be noted, however, that the technology does no fit all situations equally well. Therefore, its poential should be carefully examined with regard to the particular circumstances.
13.13.13
Pipe-CrossReactor [PCR 13.13.13
8.24-24
Tank-Typo Preneutrallier (PN)
15-15.15
17.17.17
Ursa SolutlonISteam (No Chemical Reaction)
12.12.17.1.2 Mg
14-6.21.2.4 Mg
16-0-30
(kglfonne finished product) Anhydrous ammonia-82%
37
52
66
85
0% to PaRC
Ammonia ammoniurn nitrate solutlon-46%
0100% to PCR}
127
82
(75%80%
(65%.70%
,PN)
to PNI
145
Urea Iptils-46%
105
231
Ulea solution-35%
Ammomum sullae (crystatrhne-21% N. 24% Sulfuric a c i d - v a a b s .
50 63
H,50,
64
(93%HSOJ Wet process phosphoric acid-54% PiO
205
(78%HSO.( 170
63
121
Triple superphosphate (OP)-46% Monoommoncum
N. 50%
508
332
156
40
192
261
108
(78%H.SO.
(78%HSO,(
(96%HSO,)
(96%HSO)
(100% tO PCR
(!00% to PCR
(100% tO PN)
(100% to PN)
292
330
443
250 (85% to PCR
216
phosphate (powderedl-10% tI
Dlammonium phosphatr
18.% 46%
240
204
124
P30,
PO
Phosphate
153
155 300
(90% to PCR)
Singlesupe/phosphOe {POP}-18% POs. 12%
115
crushed granules]
130
PO
rock
87
(gr ound- 32% PO
Potassium chloride slandard)-60% K10 Potassium Sulate lstandard)-50%
KO
18%
342
221
402
223
255
289
247
191
276
353
500
320
Sulfate of potash magnesium (standard)
110
22% K,O 11% Mg. 22% Magnesium sulfatekieserle]-16%
Mg. 22%
IS..O
Magnesium oxide-50% Mg
25
Insoluble binder
10
It
Micronuttrlt mix(various nultrents as required
30
Conditioning agent (coaling on productl
Subtotal Eapororlon
10
10
15
15
15
10
1,055
1.048
1,070
1.054
1.083
1.090
1.055
1.040
1.061
-48
.70
.54
83
90
55
.55
.40
.61
000
1.000
1,000
1.000
1.000
1.000
1.000
1,000
Totalproduct busi$
10
1000
15
11-4 -41
1.000
a. Examples intended to partially illustrate argo number offormulation posibililes depending upon spectla process equipment system, available raw materials, agronomic requirements, product composition guarantees, and other local factors Flows ofacid and ammonia to PC
and PN may vary. ndicaled flows ypical forcommercic'l practrce
The addition clsteam and:or water To he granulator is generally required to ptimize ranulation
etlrciency
All materials ore led directly
granulator unless otherwise noted
Indicatedcomposition is ypical at Industry practrce
Values rounded ftar smplicity Phosphoric acid is irsted to fume scrubbing system and then onward to PN
7
successfully used In commercial practice. Because th performance of these formulations in given plant will depend heavily upon number of factors as described in this bulletin, these examples are offered only to illustrate th variability that should be taken into account In th planning and design of an NPK production facility.
ry
aw Material Feed System
Th chemicalplants ranulation analysis finished product that ofusthesignificant from quantities of solid ra materials in the formulation depends heavily upon careful control of the solid ra material feeds ilyupo n t r lcaefu th soidrawmatria feds
to ensure that they are in th correct proportion an that their flow rate is closely matched with th flows of the fluid materials (for example, preneutralized slur ry and ammonia) fed to th granulator. Some granulation plants us
quickly established.
individual belt-type
weigh feeders to measure th continuous flow of each solid material to th process. Other plants, particularly those in th United States and Brazil, use combination batch weighing/continuous feed (stream-out) system. common problem with using individual belt-type weigh feeders for each material is that it is often difficult to maintain accurate control on continu ou basis because lumps and other variations in th flow properties of the nongranular solid materials. Belt-type weigh feeders can be particularly troublesome in NPK plants where large number of grades are produced and/or when lumpy, damp, or
finely textured ra materials are used. Fo example, inadequately cured superphosphate, some forms of
byproduct ammonium sulfate, and some dry but fine ly textured materials such as potash and kieserite often cause problems because they tend to bridge in the weigh feeder surge hoppers. In many cases, mechanical vibrators that are attached to the surge hoppers and designed to overcome these problems actually increase the tendency of the material to compact and bridge. Variability in th ra material feed characteristics may lead to erratic performance of th continuous belt-type weigh feeders an
gram of constantly changing the feed rates without achieving th desired goal. The lag time also varies with th ra material; for example, change in easy to-granulate ammonium phosphate slurry will show up in th product much quicker than change In difficult-to-granulate material such as sulfate of potash. Furthermore, in most NPK formulations, the recycle material ha very different chemical com position from that of th final product; thus, consider able amount of time (often h) Is required to achieve after making in the eed rate equilibrium of ra material. However, change in slurry-type process, such as DAP granulation, th neutralized feed and recycled materials ar essentially th same in chemical composition an equilibrium Is more
increase the need to fre-
quently change (adjust) th feed rates on th basis of th chemical analysis of th product. Correctly ad justing th ra material feed rates on th basis of product analyses is also very difficult because of th lag time in th granulation plant system. For exampie, several hours are required fo change in th ra material feed rate to become fully evident in th final product (Figure 4). Unless th lag time in th plant Is well known and th sampling program is carefully synchronized with this lag time, th analytical findings will be quite misieading and will thus result in pro-
Recycle-to-Product Ratio, lt
10
1.04.0 60
90 70 0-
60
CU 50
These data are theoretical.
Actual time required to effect 4a
longer an mustbe dti rmIned experimentally.
(.
10 _0
Time Following Change In Feed Composition, hour
Figure 4. Relationship Between Time an Change In Product Composition In Granulation Plant at Various Recycle
to-Product Ratios.
Most of th problems with individual belt-type weigh feeders and their possible adverse effect on the chemical analysis of NPKs can be almost complete ly avoided by using combination batch weigh Ing/continuous stream-out feed system (Figure 5) Th batch weighing system avoids th problem of uncer tain weighing accuracy caused by lumps and other variations In th dry ra material flow characteristics. With this system, precisely weighed batch of ry materials (usually 2- to 4-t batch) Is discharged Into
Figure
Raw Materials
I
which
HOLDING HOPPERS
Typically six in clus, each having cllutnne capacity.
5-to
Also smaller hoppers for minor constituents, if needed. DISCHARGE GATES Controall manuall \.ed
or automatically. rbe
HOPPER WEIGH TypIcaly 2- Io4*onne batch size.
Idcorr
Discharge Gate
SURGE HOPPER
Mas flo
minimizes Mnng-ype dein segreation and
weigh feeder.
--
blenders or rotary drum-type mixers can also be used. If rotary drum-type mixer is used, separate weigh hopper offers more practical and reliable mechan ical design. Of course, th separate weigh-hopper design ulsc works well with any type of mixer and ay preferred because of its mechanical simplicity compared with th combination weighing/mixing units. This more elaborate nd more costly version of th batch-type weighing concept (Figure 6) is usually needed only if many nutrient guarantees re re quired (for example, micronutrients), if th size and sur face characteristics of th materials differ greatly from each other, or when high degree of uniformiiy In
.--
L~~iProcess
mass flow-type surge hop
per supplying the stream-out feeder. Ribbon-type
small quantities of insoluble binders or other materials that may used to promote granulation.
FEEDER
Conlinuous slream-out
RTo
mixer Is used, followed by
each product granule is required. Such system is especially helpful in obtaining uniform distribution of
bridging.
Jndicator
depicts an example of such system In combination weigh-hopper/paddle-type
Ra
Granolation
Materials
HOLDING HOPPERS Typically six in cluster, each having 5-to 1O-fonne capacity. Also smaller hoppers for minor constituents, If needed.
Figure 5. Batch-Type Raw Material Weighing Unit With Continuous Feed System.
surge hopper equipped with a belt-type weigh feed-
er (stream-out feeder). In some plants, a variable speed
screw-type conveyor is very effective, especially if the
DISCHARGE
feed material is very fine and free flowing. The rate of this feeder Is synchronized with th flow rate of slurry from th preneutralizer (or pipe-type reactor) and other
fluids fed to the granulator. This system Is quite simple, and the operator can check the accuracy of the stream-out feeder by simply using stopwatch to check the time required to feed batch of known weight. Th operator can do this as frequently as necessary (usually
Certain formulations
ay require
large number of
COMBINATION WEIGH -HOPFERIMIXER
Paddle- or Ribbon-Type Mixer. c-ou b- a szea. Aiali
_
weigh hopper and drum-type mixer.
Discharge Gate
SURGE HOPPER
dry raw materials, which may vary widely In both quan-
tity and physical properties. In this case, the uniformity of th analysis of the individual granules of final product and th consistency of th granulation step can be Improved by mixing th weighed batch of ra materials before It enters the hopper supplying feeder. In this case It is also important, after mixing, to avoid subsequent segregation caused by th material "coning" in the feeder hopper. relatively tall, slender cylindrical hopper with long tapered conical bottom, described In the literature as "mass flow hopper," will minimize "coning" and ensure uniformity of the material entering the stream-out feeder.
Controlled manually or automatically.
'-'
about twice pe work shift) to ensure accurate feeding and proper matching of the dry feed with th slurry and other fluids fed to the process.
GATES
Mass flow design minimizes coning-type segregation an bridging. FEEDER
Continuous stream-out feeder.
____weigh
indicator
----
I
Process
Figure 6. Batch-Type Raw Material Weighlng/Mixing Unit With Continuous Feed System.
If ItIs no practical to modify an existing continuous belt-type weigh feeder system to th batch-type feed method, then It is Important to devise some method fo frequently checking th feed rates from th individual belt-type weigh feeders and using these check results as basis fo adjusting th rates. This method of feed rate adjustment Is more accurate and more rapidly responsive than are adjustments made on the basis of product-sampling because It avoids
the complications and uncertainties caused by system lag, difficulty In collecting representative samples, an
other uncontrollable variables. In NPK plants where th agglomeration of solid materials is
th
principal granulation mechanism, it is extremely important to base any change in feed rate(s) on representative product sample collected (composited) over period of at least h. Adjustments in feed rates made on th basis of frequent "grab" samples ca very misleading even Ifth feed rate(s), sampling technique, an
analytical work are precise.
Solids Recycle Unlike most slurry-type/accretion granulation process-
on single pass; therefore, th need for recirculation of th material from tn crusher back to th screens should be carefully examined. This is discussed in more detail later. Table 7. Relationship Between Particle Size and External Surface
Area
ie
........
ominal o...a.. Tylera Tyler.a
Diameter
Screen Size
(mm)
0.106
External Surface Area (m..kg.
15
43.4
0.250
60
0.500
32
1.00 2.00
9.27
16
4.62
18.5
2.31
4.00
1.16
4.75 6.70 .00 15.00
0.971 .5
a. Theoretical, assuming perfect spheres having
0.688 0.577 0.307
density of 1.3 glcm
Granulator
es, PK processes based on agglomeration seldom recycle significant quantity of product-size material to th granulator; only undersize material is normally recycled. The use of recycle control system fo controlling (optinizing) conditions in th granulator is ne of personal choice and is often of limited value except during periods of startup or when changing grades. In most cases, "floating" recycle system is preferred in agglomeration-type PK plants. However, th relative merits of floating and controlled recycle are often topic of debate among and DAP plant operators. In agglomeration-type granulation, particles will agglomerate only If they can first be attracted to each other. This attraction is then followed by interlocking and bonding. Ra material and recycle fines will often no agglomerate into product-size granules if they ar to large. They will simply repel each other and accumulate, eventually overloading th system. Therefore, facility fo crushing portion of th "large" fines is recommended if coarse-grade ra materials and/or large-size product (and therefore large-size recycle fines) are anticipated. Also, th
longer-than-normal rotary drum granulator is preferred (length-to-diameter ratio of or more) fo granulation agglomeration. The longer granula to is needed to (1) provide sufficient room fo th uni form distribution of slurry, acids, solutions, ammonia, and steam, (2)promote intimate mixing of th solids and fluids, and (3) provide time fo initial agglomer ation (granule formation) and consolidation of the granules to occur in view of the operator before the material disappears into the dryer. The "art" of granu lation implies that th granulator operator should able to se what is happening and make th often subtle adjustments needed to avoid inefficient granu lation and th resulting problems caused by upsets in th recycle equilibrium (mass flow and particle size distribution). Some plants that depend heavily upon steam and recovered scrubber liquor as th source of liquid phase very effectively utilize premixing unit (using, for example, pugmill-type mixer) located im mediately upstream from th rotary drum granulator. This added step ensu-es uniform distribution of th Ii quid phase and initiates agglomeration; thus, it at fords th operator another opportunity to optimize the granulation process.
charged from the crushers be recycled to th screens to ensure that only fine material is returned to th granulator. Many crushers are less than efficient
Cocurrent Dryer General-With agglomeration-type granulation, the moisture (usually about 2%-6%) in th material fe to the dryer is distributed quite evenly throughout the granule structure; this situation is very different from that prevailing in th slurry-type processes, fo exam ple, DAP, in which moist film covers relatively ho
ficient crushing of oversize material is essential to en sure that th particle size distribution, and therefce surface area, of th material recycled to th granulator is reasonably uniform. The relationship between particle size and surface area is shown in Table For this reason, it is recommended that th material dis-
and dry particle core. In e'ther case, cocurrent (parallel flow) dryer is preferred because with such flow configuration the ho inlet air first contacts th moist fertili7or and thus allows operation at higher overall temperature difference between th drying air and th moist fertilizer while minimizing th risk of overheating and melting th fertilizer, which becomes more temperature sensitive as it dries, With th agglomerated granules, the diffusion of moisture from th core of the granule to the outside surface and then to th drying ai must be quite cc refully regulated-loo much heat, applied to quickly, ay melt th surface of th particles and cause "cuse hardening" resulting in granules with soft, moist center caLised by entrapped moisture. Sometimes "melting" ay lead to excessive agglomeration and/or fouling of the dryer internals. In still other cases, to rapid evaporation ay cause the granule to "explode." Furthermore, ifevaporation occurs before proper consolidatlon of th granules is achieved (promoted th rolling and tumbling action in th dryer), th granules ay be to porous and therefore quite weak.
With agglomeration, th inlet portion of th dryer should be viewed as extension of th granulator because considerable amount of final granule formation and consolidation occurs here. Therefore, op timal performance of most NPK dryers requires operator ho is skilled and capable of exercising considerable judgment because the performance of th dryer is grectly influenced by th grade being produced and th characteristics of th individual
ra
materials.
Some experienced operators prefer to omit one or tw rows of lifting flights near th inlet en of th dryer to give section of smooth shell immediately followIn th forward-pitched throw (spiral) flights at th dryer inlet. The length of th smooth section usually equals about one-half to on of th dryer. The lifting flights that follow immediately after th smooth shell section must designed to provide full exposure of th moist particles to th dryIn air, ye ot be to closely spaced to cause plugging or make cleaning difficult. Temperature an Humidity Profile-In general, agglomerated products require relatively gentle drying. The temperature profile within th dryer is dependent upon th temperature sensitivity and th CRH of th material. With urea-based or highly soluble ammonium nirate-based NPKs, th dryer outlet temperature (product) usually should no exceed 740C; if kleserite is also present or th N-to-P ratio is high, then temperature of about 680C is abou maximum. The relative humidity (RH) of th dryer outlet air should be at least 10-15 percentage points below that of th CRH of th material at th dryer outlet temperature. Thus, fo most urea-based NPKs, th RH of th air at its
outlet temperature (typically about 74 C) should no exceed about 20% because the CRH of the urea- or ammonium nitrate-based materials at this elevated temperature is usually about 30%, and sometimes even lower (Figure 7) The moisture-holding capacity of air at typical drying temperature and RH con ditions is shown in Figure 8. The unique RH CRH, an temperature criteria required fo drying many NPKs, especially th urea-based NPKs, translate into larger than-usual process airflows. This results in th need for rotary dryers that ar relatively large in diameter compared with those used to dry products such as DAP that ar less sensitive with respect to temperature, humiaity, and particle entrainment. ir Velocity-The particle size distribution of most agglomerated NPKs entering th dryer Is much wider than that of DAP. PK granules are ailso relatively ir regular in shape compared with well-formed DAP granules; thus, they exhibit greater drag in th air stream. The effect of drag is clearly shown in Figure where th actual entrainment of irregular potash granules is compared with theoretical values for per fect spheres. To avoid excessive entrainment of the smaller and more irregular particles (those less than about mm found in most NP plants, It necessary to operate th dryer at lower air velocity than Iscus tomary for DAP dryers. This lower air velocity criterion fo NPKs (about 2.0 to 2.5 m/s empty-dryer [superficial] basis, compared with about to m/s fo DAP) often limits th drying capacity of DAP plants that ar con verted to th production of NPKs. Retention Time-Because th drying process for ag glomerated NPKs is mainly controlled by th rate of diffusion of moisture from th core of th particles, the point at which longer retention time does no sig nificantly promote further drying is more quickly reached with many NPKn (especially the very temperature-sensitive urea-based NPKs) than with DAP or other accretion-type granular products. The ne result isthat long retention time in th dryer is often less important with NPKs and sometimes quite un desirable. The overriding factors are th temperature and RH profile in th unit (most importantly, th differ ence between th material CRH and th RH of th dry ;ng air). Ifth dryer is to long and th cooling effect of evaporation ceases, th product may begin to overheat and "melting" may occur near th outlet end of th dryer. This can result in ring-like buildup of material on th inside of the dryer near th dis charge end. This buildup will cause bed of material to form in th dryer which, in turn, will cause less effi cient drying, overloading of the unit, and number of other problems, including overgranulatioi-. Condition of Procuct at Dryer Discharge-The physical condition of PK products dischaged from the dryer vary widely depending upon the
80 jMonoammonium Phosphate
70 O
'..... --- '-Diammonium
PhosphateO""
60
50
Calcium Ammonium
Urea-Ammnium Phosphate NP
(17117-17)Nirt
.CJ
40-
Nitrophosphate NPK (17-6-18-2MgO)
..
30
20-
Source: IFDC Experimental Determinations.
20
30
40
50
60
70
80
90
Temperature, OC Figure 7. Effect of Temperature on Critical Relative Humidity of Selected Fertilizers.
(1) ra material characteristics, (2) formulation, (3) dryer temperature and humidity profile, (4) operator skill, and (5) th particular crystallization habit of th This last point (crystallization habit) is th least predictable. Although it is quite easy to accurately determine th expected free water content, it is only by experience (trial and error) that th crystallization (hardening) characteristics of particular fertilizer for mulation can be accurately determined. Impurities In th phosphoric acid and other ra materials, the type and source of potash, th temperature and moisture content, and host of other subtle and unpredictable factors often have major influence upon th physical properties of th material dis-
charged from th dryer. At times th product may be crisp, hard, and easily screened and crushed; then again, with all conditions "apparently" th same, It may be "dry" but quite soft and plastlc-a condition that can quickly lead to screen blinding, upsetting of th recycle loop, and plugging of th crushers. Countercurrent Process Cooler
(Se.ond-Stage Dryer) The adverse effect of most of the transient and un predictable operational problems described above can be largely eliminated by routing all of the dryer
22
200 ._
180 160
o
140
0.
,_ 120
300/0
RH 20°/0 RH
100 80
RH
~15%
60
"o
40 40
20 50
60
70
80
90
10
ir Temperature, °C Figure 8. Effect of Temperature an
Relative Humidity on Moisture Content of Air.
discharge material to process cooler, which may be more properly referred to as second-stage dryer. countercurrent alr-to-fertilizer flow configuration is preferred fo maximum cooling in this unit. The transfe of fertilizer material from the dryer to the process cooler should be made by means of inclined con veyor belt; bucket elevator is no recommended because of th problems associated with th buildup of material it; th buckets and discharge chute ifthe material is plastic or sticky or if "slug" of "bad" material forms as resuit of an upset caused by operator error, burner/air heater failure, or number of other routine problems that commonly occur In most NPK granulation plants. The heat transfer capacity of the process cooler should normally be about one-half to two-thirds that of th dryer. The cooler should also be equipped with an air heater to temper (heat) the inlet air if needed to decrease Its RH and/or control th temperature profile In th unit. The addition of an air chilling unit with subsequent reheat to more carefully control th RH
and temperature of the inlet air may be needed In some locations; of course, the need fo this feature would be dependent upon local ambient conditions and th N P K product characteristics. The process cooler serves th following useful functions:
It allows drying to continue; typically, about 15% 25 of total drying occurs in th process cooler. It serves as
"safety valve" to accommodate process upsets. It makes th entire system more "forgiving".
It adds time at slightly reduced (and controllable) temperature profile to allow the material to crystallize and harden before It is subjected to screening and crushing. As result, toe screens are more likely to remain clean (free from blinding), and the crushers will perform more efficiently because the oversize particles are harder and therefore will fracture more easily. Material
6.0
5.0 Particle Specific Gravity-m-4.0
Theoretical
rMinimum orizontal
EE
Carrying (Entrainment)
3.0
Horizontal Entrainment Determined Experimentally in IFDC Pilot Plant
.0
Using 0.92 Diameter Rotary Dryer and Irregularly Shaped Granules of Potassium Chloride (1.7 specific gravity),
1.0-
I
20
30
40
50
70
20
100
300
400 500
700
1.000
2.000 3.000 4,000 5.000
Particle Diameter. microns Figure 9. Effect of Particle Size an
Ga
Velocity on Entrainment of Particles.
plasticity and th resultant plugging of th crushers, screens, and downstream chutes are largely eliminated.
Square-mesh wire screen (stainless steel) should be used to ensure th production of well-rounded (minimum surface area) granules. Well-formed granules ar of ke importance to minimize surface contact points and residual dust and small particles that cause caking during long-term storage. The screening action on square-mesh wire screen is more likely to "scrub" th granules and more ef fectively remove bits and pieces of fines. This is es pecially true ifra horizontal gyratory-type screening machine is used. Stainless steel screen wire is
Screens
key element in PK granulation is precision screening to ensure that th bagged product will have long-term storage properties. This is especiallygood applicable to high-nitrogen grades manufactured from either ammonium nitrate or urea. If th process screening system is well designed, it is usually no sreeing nstll dowstram screening a d i to ton a downstream ecesar ecessary to install additional
recommended because its reslstanceto corrosion eoot ires ac resulsn wire surface that resists build In smoother results up of solids and plugging.
facilities (for example, polishing screens after final product cooling or before bagging). Of course, such additional screening will be helpful, but th relative merits and need fo such polishing-type screening should be determined on th basis of the design and performance of th basic process screening system and th specific product characteristics. The followIn process screening system Is recommended fo mosT NPKs especially those containing high levels of
Product granules should be relatively large, in the range of 2-5 mm or even larger, to minimize contact area and, therefore, caking. The production of such large granules will influence th particle size of the recycle and th resultant granulation properties. Even with 2-mm recycle particle (maximum size), it Is often no necessary to crush the recycle; however, this point must be carefully considered depending upon th overall size distribution of the
ammonium nitrate or urea.
14
recycle fraction. The size distribution may change quite significantly from formula to formula even though the screen size range of th final product remains constant.
Single-deck screens ar
recommended because
they allow easy access for inspection and cleaning without interrupting th process. In addition to miniensuring th production of fertilizers with mu tendency to cake, clean screens ar also of ke importance in maintaining good process
screens and crushers, it ensures that the maximum particle size of the material recycled to th granula to will be smaller than th product and thus improves granulation control. Material From
Dryer or Cooler Material From Crusher
OversizeOversize....
control.
" Horizontal gyratory-type screening machines re preferred over inclined units, especially fo th product screens. The horizontal units ar prefera ble because if they become blinded due to th lack of attention they will overload and stall; thus, they will automatically prevent "fines" from being discharged as product, major cause of caking. If inclined screens become blinded, they continue to operate while discharging fines with product thus causing considerable problems with caking on th one hand and control of granulation on th other. Also, as previously mentioned, th horizontal-type units will more effectively "scrub" the granules to remove residual small particles and dust, thus decreasing th risk of caking. The ratio of the fertilizer material feed rate to th low. For th should be quite surfaceabout area 10-20 screening isrecommendtph/m oversize screen, ed, whereas only about 5-20 tph/m is recommended for the product screen with the smaller openings. Of course, the design of the screening machine an th screen open area (product size) have major influence on th screening capacihavuner given circumstance. However, it is impyunder note c i r m a n Hoen suim-r plants often suffer portant to note that many from having to little screening capacity.
Screen
Crusher Undersize
Screen
Product Undersize Material Recycled to Granulator
Figure 10
size distribution of th solid particles afgtheraticle isrtin in agglomeration-type importance of such majorsize isBecausuth granulaion, extraordinary care should be taken in determining th optimum particle size characteristics required for good granulation. This type of informa tion is learned best through testing in commercial scale equipment. Ifsuch large-scale testing is no pos sible, smaller scale (pilot plant) studies can often effectively used to provide valuable comparative data. These test dala should then be used to deter mine th optimum design of the screening and crush-
Because th
In Oversize Crushers
Crushers are no 100% efficient in crushing oversize material on a single pass; an efficiency of about 50% or less Is more likely. The crushing efficiency actually number of factors, obtained, of course, depends most Importantly th required particle size distribution and th crushing characteristics of th material. To avoid th risk of recycling !arge amount of uncrushed or only partially crushed oversize to th granulator and upsetting th granulation process, which Is very sensitive to particle size distribution, th material from the crushers should be recirculated to 10. Although the oversize screens as shown this procedure Increases th capacity needed for th
Closed-Loop Screening and Crushing System.
system.
Product Cooler conventional rotary- or fluldized bed-type product cooler Is recommended. The per forated tray, countercurrent cascade-type cooler is also good choice. Of course, th temperature and RH of th cooling air used in these units must be ad justed, depending upon ambient conditions, to effect cooling without wetting of th product. final product temperature of about 4300 or slightly less Is recom mended. Excessive cooling is no recommended, es pecially in humid areas, because to much cooling could result In th absorption of atmospheric moisture by th material; this could lead to caking.
After screening,
Conditioning
Many NPKs should be conditioned to add extra protection against caking. kaolin-type clay is recommended Usually about 0.5%-1.0% by weight is required; however, this depends heavily upon the NP product characteristics an the properties of the clay. The benefit of using oi or wa to help bind th clay to the granules is no always clear; the oil or wa does, however, help to settle dust in bulk storage an bagging areas even if t does no always fully adhere th clay to the granule surface. If an oil-type binder is used, al conveyor belts downstream from the binder addition system should be "oil proof" to avoid pl separation an failure. Belting made from neoprene or polyvinyl chloride isnot severely affected by oil-type binders or other organic-type conditioning agents. The application of oi to NPKs, especially thcse produced agglomeration, is often difficult Le cause of the relatively unpredictable absorption characteristics of th granules from on grade to another. For those grades containing ammonium nitrate, especially in combination with potassium chloride, th use of oi should avoided for safety reasons. more detailed discussion of th potential safety hazards of mixing organic materials with nitrate-containing fertilizers can be found in th Fertilizer Manual and th references cited therein. The most effective to add solid conditioning agents to granular fertilizers is by use of specially designed rotary drum application unit. The design parameters for such drums ca also found in th Fertilizer Manual.
Storage and Bagging The recommended, and usually most practical, product storage and bagging system for hygroscopic NPKs consists of (1) relatively small humidity controlled bulk storage area (building) representing about to days' production. This bulk storage area should divided according to th expected numer of grades produced over I- to 2-day period;
minimum of tw
or three drive-in storage bays is
recommended. This small and segregated storage method will also greatly facilitate changing of grades with minimum of lost time. This method of bulk storage before bagging will minimize, and usually eliminate, th accumulation of "grade change" products that ar so far off specification that they canno shipped. After brief period of storage in bulk (overnight or perhaps day or two) to allow for initial "pile set" and to certify th analyses, th product should be bagged in moistureproof bags. Bagging directly from th production unit or bulk storage of freshly made product in overhead bins is no recommended.
The optimum bag construction for long-term storage in humid areas consists of an open-mouth bag fitted with loose polyethylene film liner that is at least 0.1 mm thick. The outer jacket should constructed of woven polypropylene or some other stro and dura ble material. The liner should be tied, and the outer jacket should stitched. The stitching should no penetrate th liner. Heat sealing of th liner is no recommended because if precautions a. no taken such sealing usually entraps air that may result In pillow-shaped bags. Handling, especially stacking, such pillow-shaped bags is difficult. Venting en trapped air from such bags by percing the film liner is common practice bu this should be avoided e cause such venting allows atmospheric moisture to seep into th bag, causing wetting and caking.
Process Plant Dehumidification If hygroscopic NPKs ar produced (especially those containing urea) and th ambient relative humidity at th plant site isin excess of about 50% for more than about consecutive hours on routine basis, it is es sential that th process plant building (and bulk product storage building) be tightly enclosed nd ventilated with low-RH air. Furthermore, if th temper ature inside these enclosed buildings is excessive, then it will necessary to cool and dehumidify the ir to maintain comfortable and safe working condi tions at an acceptable RH that will prevent "wetting" of th plant and equipment. Excessive "wetting" caused by th hygroscopic fertilizer material and dust accumulations will lead to excessive corrosion, elec trical failures, safety problems (slippery floors and walkways), and number of adverse process problems particularly with conveyor belt idlers, dust collection systems, screens, and ir handling systems.
Recommendations Specific to th Production NPKs Containing Urea Urea-based NPKs ar among th granular fertilizers that ar most difficult to produce. th basis of the foregoing general discussion and th particular characteristics of most urea-based NPKs, th criteria shown in th Appendix ar recommended as start in point for th basic design of plant well suited for producing urea-basea NPKs or other NPKs that ex hibit similar characteristics such as excessive plastic ity, lo tolerance to elevated temperatures, and low CRH. These recommended design features are also applicable to th more tolerant NPKs. However, with these more tolerant NPKs, more latitude in th design and operating criteria be allowed.
To Almosphere Phospho,; Acid
SolidMaterials
SlVent
WaWaer
V
"n
Phoslrhoriccid
-... (See Nolet)
RawMaterial Surqe oppers (Cluster
Oust Coct so
Hoppers)
System
Product
Vent PaneutlizerScreens
,See Note21
8andMae adM.e
Oversize Recycle
Gnua0~M
CotnosRotary
reednyem
hoidrea
Venloto Aiiit
Soli
UreaSoluioniMeil Preparationystem
anuta
Crushers
Phosphor.ccidSuituriccid -- Ammonia
Vent
or -SeamlWate,
R tailyDryer
A,, Heiler
teamL
re -.
Scrbber --
Rotay Drer
um~
eli
Snaded equipineni
Conditioninggent & Of Bidder) (l
.
.
. um oSleruberPdc
tIrls usuaiiy not inc uOed in connhiionil
ROijry i tiamitlled SProcessiani buiding may hays io OndehuirRdled dependinA upon CRl
ic entiiiti
aem
andlanicent relatvr nurnidiri
Recommended Process Flow Diagram for Maximum flexibility In Producing Granular NPKs Based on Agglomeration.
process flow diagram of an PK granulation plant embodying th recommended features is shown in Figure 11 Also Indicated in Figure 11 ar the features that ar most often missing In plant designed primarily for th production of DAP/MAP.
Verification of Recommended Plant
Design an
Operating Parameters
design recommended an operating parameters for NPK granulation plants described in this bulletin ar thought to be near optimum on the basis of data obtained from number of IFDC pilot plant-scale trials an experience with several commercial-scale operations. However, because no two formulations (grades) or plants respond in exact ly th same way to given set of conditions, extrapo lation an judgment are required when attempting ofnew"sgrade or plant. to set the design basis fo The
Vent
.~I~rv~m
ammninum phosphate iAPHMAPi plants5 Pipe cross reactor PCRi in ,Additon iO eeuiamoei piovides mnxinium ie, ibuity in toimuiailng 050M rod, cont~rin ng rades
11.
Veni
Notes I
Figure
Screens
__
mO-
.....
Con S
Oversize
Gas Scruping
Wale
Conclusion The fundamentals of granule formation and process ing for agglomerated fertilizers differ markedly from those for DAP/MAP For this reason, many well-designed DAP/MAP plants experience difficulties when called upon to produce NPKs. The difficulties ar
most pronounced when the NPKs contain urea or other temperature-sensitive an hygroscopic Ingre
dients.
plant design and operating criteria
described in this bulletin are intended to guide those involved in planning new NPK plant projects o' modifying existing units. This troubleshooting an bulletin ma also be used as checklist during the various stages of project development to help ensure that the special requirements for NPK plants ar not overlooked.
APPENDIX-Design an
Operating Criteria fo Urea-Based NP
Granulation Plants
Solid Urea FeedstockPrilled urea is preferred because its small physical size helps to effect relatively homogeneous incorpora tion into th granule structure. Very small and broken prills re th most desirable. Urea SolutionIfth urea content of formula is more than about 20%, it is often preferable to dissolve portion of the urea to produce ho (1050C) 75%-80% urea solution. The availability of urea solution adds consider able flexibility to th process. The solution is sprayed on top of th bed of material in th granulator. In most instances, solution of urea is preferred over an anhydrous melt.
Other Solid Ra
Materials
Standard-grade materials ar preferred. Nongranular run-of-pile (powdered) monoammonium phos phate and superphosphate ar recommended. If superphosphate is used in combination with urea,
it should be ammoniated to mini.,iize unwanted reactions that result in th release of water of crystalliza tion contained in th superphosphate. In general, th use of superphosphate in combination with urea is ot recommended and should avoided, even th superphosphate is ammoniated. Standard grade muriate of potash is preferred because of it good flow characteristics and more optimum particle-size distribution. Good flow characteristics of all solid ra materials re desired to facilitate handling and accurate metering. Preneutralizerstandard, atmospheric tank-type preneutralizer ispreferred. An NH:HPO mole ratio of about 0.5 to 0.6 is recommended in the preneutralizer to achieve low free water content while still maintaining fluid and pumpable slurry. This is especially important if significant amount of sulfuric acid is also neu tralized in th unit. The preneutralized slurry at mole ratio of about 0.5 to 0.6 should be about 1270C and contain no more than 15 free water. The preneutralizer and its auxiliary equipment (agitator, pumps, and piping) should constructed of corrosion-resistant materials to facilitate th use of mixture of phosphoric and sulfuric acid, thus adding considerable flexibility to th process. In gener l, if higher mole ratio (for example 1.5) is used, especially if sulfuric acid is also present, th free water content of th slurry will have to higher to facilitate pumping. Granulator (rotary)The length-to-diameter ratio should be at least 3. The greater length, compared with most AP and many North American NP plants, facilitates granule formation and gives th operator more flexibility in con trolling the agglomeration process. The NH :H PO mole ratio of the material discharged from th granu lator ay quite variable depending upon th properties of th ra materials. However, mole ratio between 1. and about 1. is expected to optimum. The lower mole ratio will simplify operation of th scrubbing system and tend to improve the process control. The free moisture of th product discharged from the granulator will usually be in the range of about 2% to 3%. The recycle-to-product ratio is expected to vaiy from about to 6. ratio of is recommended for design. If the process is
based on
solid ammonium phosphate source (a preneutralized slurry is no used), then recycle-to-product ratio of about is sufficient.
Cocurrent Dryer-
design
The length-to-diameter ratio of this rotary unit should no exceed about 6. The superficial velocity of the air (a outlet conditions) should no exceed 2.4 m/sec maximum (2.0 m/sec preferred). The maximum temperature of th outlet ai should no exceed 80°C, and th RH at this temperature should ot ex ceed 15%. An outlet ai temperature of about 750C at 20% RH is recommended as th design basis. The moisture of th material discharged from th dryer should no exceed 1.0%; value of 0.8% is recom mended for design provided second-stage dryer (process cooler) is used. The temperature and hu midity profile in th dryer is extremely important, and th optimum values will vary from product to product.
Countercurrent Process Cooler-(Second-Stage Dryer) The length-to-diameter ratio of this unit should no be less than 6; to 8 is preferred. The inlet air should be tempered (heated) and/or conditioned to ensure that the RH i, 60% or less. The superficial ir ve locity should no exceed 2.0 m/sec. The lo velocity is recommended because of th relatively large removed amount of minus 1.0 mm particles in th mate, ial. It is preferred that these small particles (separated) by the screens rather than by th airflowidust collection system. The unit should be designed at a free moisture con to achieve material (discharge) temperature of no more than about 54 tent of no more thon 0.6%. Further cooling of th product fraction is performed in a separate operation. Screening stainless steel Single-deck horizontal gyratory-type units ar recommended. The screen wire should and of th square-mesh style. If inclined, electrically (or motor) vibrated screens re used, they should be for oversize separation only, no fo product screening. The hourly loading of th oversize screen should no exceed about 20 t/m , and th loading of the horizontal gyratory-type product screen should no exceed 50% of this valie (25%-30% preferred). Oversize CrushersDouble rotor chainmil!-ty'..- '.rushers or double row cage mills are preferred. The discharge assem blies of th mills should no restricted and should be constructed of flexible rubber (conveyor belt ing) panels that ca flexed from th outside by an operator using hammer. flared-type discharge assembly for the crushers is recommended to help avoid th accumulation of solids. The crushed oversize should be recycled to th oversize screen on closed-loop basis to ensure that only th fine material fraction from th screens is returned to th granulator as recycle.
Product CoolingEither rotary-drum or fluidized-bed unit is recommended. The single-pass, counter-current, cascade type unit is also acceptable. The cooling unit should be located immediately ahead of th condi tioning unit. The RH of th cooling air at inlet conditions should no exceed 50%. The temperature all cases th of th product discharged from th cooler should no exceed about 430C. about above th average ambient temperature to avoid ture of th cooled product should absorption of atmospheric moisture on th surface of th material. Conditioning-
standard rotary drum-type conditioning unit is recommended. Screw-type mixing units should be avoided as they tend to grind and break th product granules. Bulk Storage and Process Plant DehumidificationIf ambient conditions normally exceed about 50% RH for extended periods (more than about 8-16 h) provisions should be made to dehumidify th process plant and bulk storage buildings. Cooling necessary for worker comfort and safety. Adequate ventilation ay also and proper ventilation is especially important because al buildings should be tightly constructed and closed to maintain dry inside environment (RH of 40%-50%).
BaggingThe products should be bagged in moistureproof bags shortly after production. Direct bagging from th production unit is no recommended because off-specification product could inadvertently bagged and because flexibility is lost in handling product during grade-change periods. In addi tion, bag set (caking) is minimized if th fresh product is allowed to "pile set" for short period of time.
Bibliography
7. Medbery,
1. Achorn, F.P., and D.G. Salladay. 1976. "TVA's Pipe-Cross Reactor Process for Granular
ew
Ammo-
nium Phosphates," Paper presented at American Chemical Society Meeting, San Francisco,
Nlelsson. 1981. "Fun damentals of Granulation," INProceedings of the 182nd American Chemical Society National L., and
Meeting,Paper No 20, New York, New York, U.S.A.
8. Salladay, D.G., and B.R.Parker. 1980. "Commer cialization of th TVA Pipe-Cross Reactor in
2. Chinal, P., C. DeBayeux, and J. F.Priat. 1986. "Dual Pipe Reactor Process fo DAP, NP, and NPK Production," IN Proceedings of lhu 36th Annual Meeting of th Fertilizer Industry Round Table, pp.
Regional NPK and DAP Granulation Plants In the United States," Paper presented at Annual MeetIn of th Fertlllser Association of India, New Delhi,
India.
30.
9. Schultz, J. J., and E.D. Frederick, (Eds). 1988. NPK Fertilizer Production Alternatives, IFDC Special 3. Fischbein, M., and A. M. Brown. 1988. "High QualPublicto P u c i n t e r n a t i a l Deel Fl Ity Granular Ammonium Sulphate Production," Publication SP-9, International Fertilizer Developm o i u m S u l p a t e P r o d c t i n , " m e n t Center, Muscle Shoals, Alabama 35662, ranlar Paper presented at th International Fertilizer InU.S.A. U.S.A. dustry Association, Limited, Conference, Edmonton, Canada. 0. Schultz, J. J., and J. R.Polo. 1987. "The Role g 4. Hemsley, J.D.C., and Francisco Roig. 1972. "The glomeration in the Fertilizer Industry," Paper Manufacture of Granular Compound Fertilizers presented at th 20th Biennial Conference, Insti Based on Urea as th Principal Source of Nitrotute for Briquetting and Agglomeration, Orlando, Florida, U.S.A. ConTechnical gen," IN Proceedings of th ISMA ference held in Seville, Spain 20th to 24th 11. Sheidrick, W.F. 1970. "The Prediction of Granula November, 1972, International Superphosphate tion Plant Performance From Temperature, and Compound Manufacturers' Association Ltd., Moisture Relationships in th Granulator," IN London, England. Proceedings of th 20th Annual Meeting, The Fer 5. Hoffmeister, G., and C. P. Harrison. 1977. "Physical Properties of Granular Urea-Based NP and PK
tilizer Industry Round Table, pp 96-100. 2. Sherrington, P.J. and Oliver. 1981. Granulation, Heyden Son Inc., Philadelphia, Pennsylvania,
Fertilizers," IN Proceedings of th 27th Annual Fertilizer Industry Round Table, pp.
Meeting 162-171.
U.S.A.
6. International Fertilizer Development Center. 1979. FertilizerManual. IFDC-R-I (also available from the United Nations Industrial Development Organization, Vienna, Austria), Muscle Shoals, Alabama 35662 U.S.A.
13. Tennessee
Valley Authority. 1976. Studies of Granulation of Compound Fertilizers Containing Urea-A Literature Review, Bulletin Y-108, Nation al Fertilizer Development Center, Muscle Shoals, Alabama 35660, U.S.A.
20
