&EPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600 2-79-169
August 1979
Research and Development
Evaluation of Control
Technology for the
Phosphate Fertilizer
Industry
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-79-169
August 1979
Evaluation of Control Technology
for the Phosphate Fertilizer Industry
by
Vladimir G. Boscak
TRC--The Research Corporation of New England
125 Silas Deane Highway
Wethersfield, Connecticut 06109
Contract No. 68-02-2615
Task No. 10
Program Element No. 1AB604B
EPA Project Officer: Ronald A. Venezia
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
TABLE OF CONTENTS
SECTION PAGE
1.0 EXECUTIVE SUMMARY 1
1.1 Phosphate Fertilizer Processes and Emissions 1
1.2 Control Technologies 3
1.2.1 Air Emission Control Technology 3
1.2.2 Gypsum Pond Emission and Controls 5
1.2.3 Solid Waste Control 6
1.3 Recommended RD & D Projects 6
1.4 References 9
2.0 CONCLUSIONS 10
3.0 RECOMMENDATIONS 12
4.0 PROCESS AND EMISSION DESCRIPTION FOR FIVE
PHOSPHATE FERTILIZER MANUFACTURING
PROCESSES 13
4.1 Wet Process Phosphoric Acid (WPPA) 13
4.1.1 WPPA Process Description 13
4.1.2 Emissions from WPPA 16
4.2 Superphosphoric Acid (SPA) 18
4.2.1 SPA Process Description 18
4.2.2 Emissions from SPA 21
4.3 Diammonium Phosphate (DAP) 21
4.3.1 DAP Process Description 21
4.3.2 Emissions from DAP 24
4.4 Normal Superphosphate (NSP) 24
4.4.1 NSP Process Description 24
4.4.2 Emissions from NSP 26
4.5 Triple Superphosphate (TSP) 26
4.5.1 TSP Process Description 26
4.5.2 Emissions from TSP 29
4.6 Summary of Fluoride Emissions 29
4.7 References 29
5.0 STATE-OF-THE-ART AIR EMISSIONS CONTROL
TECHNOLOGY 34
5.1 Theory and Practice of Wet Scrubbing 34
5.1.1 Gaseous Fluoride Emission Control 34
5.1.2 Particulate Emissions Control 43
5.2 Types of Scrubbers Used in Phosphate Fertilizer Plants. . . 46
5.2.1 Cyclonic Spray Scrubber 46
5.2.2 Venturi Scrubbers 47
5.2.3 Spary-Crossflow Packed Scrubber (SCFS) 47
5.2.4 Coaxial Scrubber 51
11
-------
TABLE OF CONTENTS
(Continued)
SECTION PAGE
5.3 Wet Process Phosphoric Acid (WPPA) 51
5.4 Superphosphoric Acid (SPA) 57
5.5 Diammonium Phosphate (DAP) 63
5.6 Normal Superphosphate (N5P) 72
5.7 Triple Superphosphate (TSP) 72
5.8 References 80
6.0 GYPSUM POND EMISSIONS AND CONTROLS 82
6.1 Location, Description and Role of the Gypsum Pond .... 82
6.2 Gypsum Pond Chemistry 85
6.3 Gypsum Pond Air Emissions 89
6.4 Seepage from the Gypsum Pond 95
6.5 Present Control of Gypsum Pond Emissions 97
6.6 Identification of Control Techniques for Gypsum Ponds. . . 102
6.6.1 Two Pond System 102
6.6.2 Kidde Process 103
6.6.3 Swift Process 103
6.6.4 Liming of the Cooling Pond 105
6.6.5 Dry Conveyance of Gypsum to Stacks 105
6.6.6 Phosphate Rock Calcination 105
6.6.7 Change to Hemi/Dihydrate Process 107
6.6.8. Use of Algae 107
6.6.9 TESI Dry System 107
6.6.10 Comparison of Control Options 108
6.7 References 108
7.0 SOLID WASTE CONTROL Ill
7.1 Sources and Amount of Solid Waste Ill
7.2 Special Waste Regulation Ill
7.3 Phosphate Fertilizer Solid Waste Control 114
7.4 References 117
8.0 REVIEW OF CONTROL EQUIPMENT VENDORS AND
ENGINEERING FIRMS 118
8.1 Environmental Elements Corporation (ENELCO),
Baltimore, MD 118
8.2 The Ducon Company, Inc., Mineola, NY 122
8.3 The Ceilcote Company, Berea, Ohio 128
8.4 Heii Process Equipment Company, Avon, Ohio 132
8.5 Major Engineering Firms Designing Phosphate
Fertilizer Plants 132
8.6 References 135
111
-------
TABLE OF CONTENTS
(Continued)
SECTION PAGE
9.0 RECOMMENDED RD&D PROJECTS 139
9.1 Epidemiological Study of the Phosphate Fertilizer
Industry 139
9.1.1 Background 139
9.1.2 Generalized Methodology for Epidemiological Study . . .
9.2 Studies of Gypsum Pond Emissions and Chemistry
9.2.1 Study of Fluorine Distribution 147
9.2.2 Measurement of Gypsum Pond Emissions 150
9.3 Evaluation and Optimization of Wet Scrubbers 151
9.3.1 Factorial Design Experiments 151
9.3.2 Performance Evaluation of Ionizing Wet Scrubbers. . . . 155
9.4 Ammonia-Sulfuric Acid Mist Interaction Atmospheric
Chemistry and Dispersion Modeling Study 156
9.5 Demonstration of Dry Fluoride Removal System 159
9.6 Evaluation of the Kimre Mist Eliminator 165
9.7 References 166
10.0 BIBLIOGRAPHY 170
10.1 Process Description 170
10.2 Air Emissions and Controls 171
10.3 Wastewater Sources and Treatment 172
10.4 Solid Waste 173
10.5 Miscellaneous 174
10.5.1 Marketing 174
10.5.2 Energy Consumption 174
10.5.3 By-product Recovery 175
IV
-------
LIST OF FIGURES
FIGURE
4-1 Major Phosphate Rock Processing Steps
4-2 Flow Diagram for a Wet-Process Phosphoric
Acid Plant
4-3 Stauffer Evaporator Process .............. 19
4-4 Swenson Evaporator Process .............. 20
4-5 TVA Ammonium Phosphate Process Flow Diagram ..... 22
4-6 Normal Superphosphate Process Flow Diagram ...... 28
4-7 Granular Triple Superphosphate Process Flow Diagram ... 30
5-1 Power Consumed in HF Absorption ........... 39
5-2 Power Consumed in SiF4 Absorption ........... 40
5-3 Inlet Versus Outlet Fluoride Concentration in
Cyclonic Spray Tower ................. 41
5-4 Venturi Scrubber Performance ............. 45
5-5 Cyclonic Spray Tower Scrubber ............. 48
5-6 Gas-Actuated Venturi Scrubber with Cyclonic
Mist Eliminator .................... 49
5-7 Water-Actuated Venturi ................ 49
5_8 Spray-Crossflow Packed Bed Scrubber .......... 50
5-9 Tesi Coaxial Scrubber ................. 52
5-10 Spray-Crossflow Packed Bed Scrubber in Wet Process
Phosphoric Acid Plant ................. 53
5-11 Spray-Crossflow Scrubber System in Superphosphoric
Acid Plant ...................... 59
5-12 Acceptable Scrubber Combinations for DAP
Process Plants .................... 64
-------
LIST OF FIGURES
(Continued)
FIGURE PAGE
5-13 Primary and Secondary Scrubbers in Diammonium
Phosphate Plant 65
5-14 Scrubbing System on Run-of-Pile Normal
Superphosphate Plant 73
5-15 Scrubbing System on Run-of-Pile Triple
Superphosphate Plant 74
6-1 Approximate Locations of Gypsum Ponds in
Central Florida 83
6-2 Fluorine Distribution: Interrelation of Gypsum Pond
System to the Total Process 84
6-3 Major Gypsum Pond Equilibrium 90
6-4 Tatera's Emission Factors for Process Water at
75°, 85°, 95°F 91
6-5 Fluoride Emission Rates for Ponds with Water
Containing 0.628 g moles/liter Fluorides V16 =
Wind Speed at 16 Meters in Meters per Second 93
6-6 The Storage Pond System: Interrelation of Fluorine
Distribution and Water Management 96
6-7 Flow Sheet, "Double Liming" Treatment of Gypsum
Pond Water 99
6-8 Species Predominance Diagram for 0.4 HF Solution .... 100
6-9 Chemical Models of F and P2O5 Removal By Lime .... 101
6-10 Two Pond System for Phosphoric Acid Plant 104
6-11 Proposed Single Liming System 106
7-1 Gypsum Stack and Acid Water System 116
8-1 ENELCO Venturi Gas Scrubber 119
VI
-------
LIST OF FIGURES
(Continued)
FIGURE PAGE
8-2 ENELCO Cyclon-Spray Gas Scrubber 120
8-3 ENELCO Poly Pack Tower Gas Scrubber 121
8-4 Wet Process Phosphoric Acid Flow Sheet 123
8-5 ROP Triple Superphosphate Flow Sheet 124
8-6 Cyclonic and Wet Scrubbers Manufactured by Ducon .... 125
8-7 Oriclone WO Venturi 127
8-8 DAP Process Schematic Including Ducon Scrubbers .... 129
8-9 Simplified Cross-Section and Side-View of the
IWS Scrubber 130
8-10 Three Arrangements of Heil 720 Venturi Scrubbers .... 133
8-11 Typical Phosphoric Acid Process Flow Sheet .134
8-12 Granulation Plant Process Flow Sheet 136
8-13 Tesi Crossflow Scrubber 137
9-1 Uranium Radium Family Mode of Decay 141
9-2 Epidemiological Data for White Males,
State of Florida 142
9-3 Competing Fluoride Chemical Equilibria 149
9-4 Particle Size Distribution Tests A-E, Kearny
Generating Station 157
9-5 Dry Chrornatographic System 160
9-6 Flow Sheet for Pilot Dry Fluoride Removal System .... 164
9-7 Four Stage Scrubber/Mist Eliminator 167
vu
-------
LIST OF TABLES
TABLE PAGE
1-1 Emission Characteristics for Phosphate Fertilizer
Processes at Average Plants 2
1-2 Summary of Fluoride Emission Control Equipment for
Five Phosphate Fertilizer Processes 4
1-3 Information on Five Gypsum Pond Fluoride Control
Alternatives 7
4-1 Fluorine Material Balances for WPPA Manufacture .... 17
4-2 Average Stack Heights and Controlled Emission Factors
for Wet Process Phosphoric Acid 17
4-3 Emission Factors for Diammonium Phosphate Plants .... 25
4-4 Emission Factors for an Average Normal Superphosphate
Plant Based on Controlled Emission Sources 27
4-5 Emission Factors for an Average Run-of-Pile Triple
Superphosphate Plant Based on Controlled Emission
Sources 31
4-6 Emission Factors of Controlled Emissions for an Average
Granular Triple Superphosphate Plant 31
4-7 Phosphate Industries Fluoride Emissions 32
5-1 Advantages and Disadvantages of Wet Air and Gas
Cleaning Devices 35
5-2 Hydrogen Fluoride Absorption Data 38
5-3 Scrubber Performance in Wet-Process Phosphoric Acid
Plants 54
5-4 Operating Conditions for Wet Process Phosphoric Acid
Plant Cross-Flow Packed Scrubber 55
5-5 Capital and Annual Cost for CFPS on WPPA Plants .... 56
5-6 Retrofit Costs for Model WPPA Plant, Case B 58
5-7 Operating Conditions for Wet Scrubbers for SPA Process
Combined Vents Specification 60
via
-------
LIST OF TABLES
(Continued)
TABLE PAGE
5-8 Capital and Annual Control Costs for SPA Plants
Vacuum Evaporation 61
5-9 Retrofit Costs for Model SPA Plant 62
5-10 Operating Conditions for Primary Scrubbers for DAP
Process Dryer Vents 67
5-11 Operating Conditions for Secondary Scrubbers for DAP
Process Dryer Vents 68
5-12 Capital and Annual Control Costs for Diammonium
Phosphate Plants 69
5-13 Spray-Crossflow Packed Bed Scrubber Performance in
Diammonium Phosphate and Granular Triple Superphosphate
Plants 70
5-H Retrofit Costs for Model DAP Plant 71
5-15 Engineering Specifications for Estimating Costs for
Granular Triple Superphosphate Plants 75
5-16 Capital and Annual Control Costs for Granular Triple
Superphosphate Plants 76
5-17 Capital and Annual Control Costs for Run-of-Pile Triple
Superphosphate Production 77
5-18 Retrofit Costs for Model ROP-TSP Plant, Case A 78
5-19 Retrofit Costs for Model GTSP Plant 79
6-1 Fluorine Inventory 700 TPD P2O5 Production
Florida Rock 86
6-2 Major Cation and Anion Concentrations in Gypsum
Pond Water 87
6-3 Possible Modes of F Precipitation in F.G. WAP
Solution 88
IX
-------
LIST OF TABLES
(Continued)
TABLE PAGE
6-4 Comparison of Emission Factors Predicted by King and
Tatera at Various Temperatures ............. 94
6-5 EPA Effluent Limitations Guidelines for Gypsum
Pond Water ..................... 98
6-6 Capital Investment and Operating Costs for Fluoride
Control of 1,000 TPD P2O5 Plant ............ 109
7-1 Special Waste .................... 113
9-1 Cancer Mortality Rate, State of Florida .........
9-2 Parameters of F Distribution in Digester Stage .....
9-3 Statistical Design for Scrubber Testing .......... 153
9-4 Confounding Patterns for Design ............ 154
9-5 Optimization Design for Scrubber Testing ........ 154
9-6 Emission Factors of Some Industrial Activities ...... 158
9-7 Performance of Chromatographic Systems ........ 162
9-8 Performance of TESI Equipment ............ 163
-------
1.0 EXECUTIVE SUMMARY
The Industrial Environmental Research Laboratory of the U.S. Environmental
Protection Agency has the responsibility for ensuring that pollution control technol-
ogy is available for stationary sources to meet air, water and solid waste legislation.
The purpose of this study was to evaluate control technology for the phosphate
fertilizer industry. The evaluation covered five fertilizer manufacturing processes:
wet process phosphoric acid, superphosphoric acid, diammonium phosphate, normal
superphosphate and triple superphosphate.
1.1 Phosphate Fertilizer Processes and Emissions
Wet process phosphoric acid is an intermediate product in the manufacture of
phosphate fertilizer. Most current processes use the dihydrate process in which
phosphate rock is treated with sulfuric acid with gypsum as the major by-product.
Fluosilicic acid can also be recovered as by-product resulting in a 50 percent reduc-
tion in the average emission of fluorides.
Superphosphoric acid is produced by further concentration of the wet process
phosphoric acid using vacuum evaporation or submerged combustion. Seventy-five
percent of the total production of superphosphoric acid is prepared using either the
falling film (Stauffer) or the forced circulation (Swenson) version of the vacuum
evaporation process.
Diammonium phosphate is manufactured through the reaction of anhydrous
ammonia with phosphoric acid. TVA's process is the most frequently used while the
Dorr-Oliver process has limited application. Normal superphosphate is produced
through the reaction between ground phosphate rock and sulfuric acid. Most of its
production has been recently replaced by diammonium phosphate and triple super-
phosphate.
Triple superphosphate is produced through the reaction between ground
phosphate rock and phosphoric acid. Run-of-pile triple superphosphate is manu-
factured by a process similar to normal superphosphate while the TVA or Dorr-
Oliver process is used for the production of granular triple superphosphate.
Table 1-1 summarizes air emissions from all five phosphate fertilizer
processes. The emissions consist primarily of particulate and gaseous fluorides and
some sulfur dioxide. The particulate consists of unreacted rock, final products, and
insoluble salts and can include liquid phosphoric acid aerosols and mists. The major
gaseous compounds are hydrofluoric acid (HF) and silicon tetrafluoride (SiF4). The
major sources of fluoride emission in the phosphate fertilizer industry are uncon-
trolled emissions from normal superphosphate curing buildings and fugitive gypsum
pond emissions consisting primarily of HF.
-1-
-------
TABLE 1-1. EMISSION CHARACTERISTICS FOR PHOSPHATE FERTILIZER
PROCESSES AT AVERAGE PLANTS (1)
Process
Phosphoric acid
Superphosphoric acid
Ammonium phosphate
Normal superphosphate
Run-of-the-pile triple
superphosphate
Granular triple
superphosphate
Fertilizer complex
Emission point
Rock unloading
Rock transfer and storage
Wet scrubber system
Wet scrubber
Product sizing and material
transfer
Total process emissions
Rock unloading
Rock feeding
Mixer and den
Curing building
Rock unloading
Rock feeding
Cone mixer, den, and
curing building
Rock unloading
Rock feeding
Reactor, granulator dryer,
cooler, and screens
Curing building
Gypsum pond
Emission species
Particulate
Particulate
Particulate
Fluoride
Sulfur oxides
Particulate
Fluoride
Particulate
Particulate
Fluoride
Ammonia
Particulate
Particulate
Particulate
Fluoride
Particulate
Fluoride
Particulate
Particulate
Particulate
Fluoride
Particulate
Particulate
Particulate
Fluoride
Sulfur oxides
Particulate
Fluoride
Fluoride
Controlled
emission factor,
g/kg P205
0.15 + 250%
0.045 + 180%
0.054 + 164%
•0.010+ 47%
0.032 + 200%
0.011 to 0.055
0.0073+ 71%(a)
l.l'b>
0.15 + 120%
0.038 + 30%
0.068 + 75%
0.28(a)
0.055 + 180%
0.26 + 86%
0.10 + 120%
3.~6
1.9 + I20%(d)
0.07(a)
0.014 + 170%
0.10+ 50%
0.10+ 40%
0.09(a)
0.017 + 180%
0.05 + 320%
0.12+ 30%
l.86(e)
0.10 + 240%
0.018 +_ 40%
0.50
(0.025 to 2.5)
-------
1.2 Control Technologies
1.2.1 Air Emission Control Technology
Air emission control technology in the phosphate fertilizer industry consists
almost entirely of wet scrubbers. The most widely used scrubbers are the cyclonic
spray, cross-flow packed and Venturi scrubbers. Wet scrubbers have an advantage
over dry gas cleaning devices such as filters in that they can simultaneously remove
particulate and gaseous emissions. The major gaseous pollutants from phosphate
fertilizer manufacturing are fluorides consisting primarily of HF and SiF4. Most
scrubbers use gypsum pond liquid as a scrubbing liquid. Since fluorides are readily
soluble in the scrubbing liquid, absorption is gas phase controlled. Four to six
transfer units are normally sufficient for over 95% removal efficiency.
The collection of particulate in wet scrubbers is based on the mechanisms of
inertial impaction, interception and diffusion. The major parameters that dictate
particulate removal efficiency are particle size, pressure drop and liquid to gas
ratio.
Table 1-2 shows the combination of wet scrubbers used for removal of
fluorides in the five phosphate fertilizer processes. As a rule, primary scrubbers
are designed for particulate removal while secondary scrubbers remove most of the
fluorides.
Recent findings indicate that fluorides emitted from diammonium phosphate
consist partially of submicron particulate. Conventional wet scrubbers have the
lowest efficiency on particles whose size is between 0.2 to 1.0 microns so that a
scrubber using a removal mechanism other than those mentioned above is needed for
effective control. Two such relatively new devices are used in the phosphate
fertilizer industry: one is based on nucleation while the other uses electrostatic
charging as a mechanism for particulate removal.
Several scrubber vendors supply equipment to the phosphate fertilizer industry
and most of the controlled sources are now in compliance with emission regulations.
Environmental Elements Corporation and The Ducon Company are representatives
of suppliers of large installations while the Ceilcote Company and Heil Process
Equipment Company represent suppliers of smaller installations. Davy Powergas is
one of the major engineering firms designing phosphate fertilizer plants. They also
design their own scrubbers. In the past Davy Powergas has worked closely with
Teller Environmental Systems on the design of air emission control equipment,
specifically cross-flow scrubbers.
The review of air emission control technology carried out for this project
indicates that adequate technology is available for control of gaseous and particu-
late emissions from phosphate fertilizer plants. The only exceptions are normal and
triple superphosphate storage facilities. The use of wet scrubbers at these facilities
would result in significant reduction of fluoride emissions. Further reduction of
emissions is achievable through recovery of fluosilicic acid, for which the recovery
technology is available, but the market for the acid appears to be rather narrow.
-3-
-------
Process
TABLE 1-2. SUMMARY OF FLUORIDE EMISSION CONTROL EQUIPMENT
FOR FIVE PHOSPHATE FERTILIZER PROCESSES
Fluoride Emission
Standard
Grams (F-) per
Kilogram of
Required
Efficiency
Types of Scrubbers Used
Best Control Technology
i
-P-
Wet process
phosphoric acid
Super phosphoric
acid
Diammonium
phosphate
Triple
superphosphate
Normal and triple
superphosphate
storage (granular)
0.01
0.005
0.03
99
50
85
0.1
99.6
0.0025
90
Cross-flow packed,
Water-actuated Venturi,
Centrifugal spray
Water-actuated Venturi,
Conventional Venturi,
Impingement
Primary
Conventional
Venturi,
Cross-flow
packed,
Two-stage
Cyclonic
Primary
Cyclonic,
Venturi,
Spray
Secondary
Cross-flow
packed,
Two-stage
cyclonic,
Two-stage
Venturi
Secondary
Cyclonic,
Cross-flow
packed
Spray,
Cyclonic
Cross-flow packed
Cross-flow packed
Primary
Coaxial or Cross-flow
packed
Secondary
Nucleatioii or Cross-Flow
packed
Primary
Venturi
Secondary
Cross-flow packed
Cross-flow packed
-------
1.2.2 Gypsum Pond Emission and Controls
The gypsum pond is an integral part of the phosphate fertilizer manufacture
wastewater treatment scheme and it serves two functions. It is a settling and
storage area for waste gypsum and fluorides and it provides an area for evaporative
cooling of the pond water. It is also used as a scrubbing liquid in wet scrubbers for
control of air emissions. Gypsum pond water is a complex mixture of various
chemical species. Its pH is about 1.5 and the fluoride concentration ranges between
^,000 and 10,000 ppm. Fluoride concentration in ponds is stabilized due to the
fixation of fluorides in insoluble species like chukhrovite or ralstonite. The
chemistry of fixation is rather complex and only partially understood, but it appears
that the concentration of aluminum and magnesium plays an important role. Two
major environmental concerns regarding gypsum ponds are emissions of fluorides to
the atmosphere and possible leaching of radioactive substances, fluorides and
phosphorus compounds to groundwater supplies and watercourses.
Gypsum pond atmospheric emissions have been calculated and measured and
fluoride concentration above these ponds is in the order of tens of ppb. Based on
gas-liquid vapor pressure equilibria studies and in-field measurements with EPA's
ROSE (Remote Optical Sensing of Emissions) instrumentation, it can be concluded
that HF is the major fluoride compound emanating from the gypsum pond. The
estimated emission rates from gypsum ponds are in the range of 0.2 to 7.1 Ib/acre-
day. Consequently, fugitive fluoride emissions from the gypsum pond at many plants
appear to surpass the quantities from the controlled stationary sources.
There are virtually no data on leaching of fluorides, phosphates and radium
from gypsum ponds and stacks. It appears that contamination of deep aquifers is not
occurring because percolation through clayed sediments under ponds is very slow.
Only some local contamination of the water table aquifer occurs at some sites. A
study currently underway should determine distribution of contaminants around a
gypsum pond.
The pond water is reused for wet scrubbing and processing so that discharge is
needed only when there is a rainfall in excess of evaporation. In order to meet EPA
discharge regulations, two-stage liming combined with settling is needed. The cost
of two-stage liming is relatively high so it cannot be used as a means of reducing
fluoride emission from gypsum ponds. There are several alternatives for control of
these emissions:
1. Two pond system
2. Kidde process
3. Swift process
4. Liming of cooling ponds
5. Dry conveyance of gypsum to stacks
6. Phosphate rock calcining
7. Change to hemi/dihydrate process
8. Use of algae
9. Use of dry systems for air emission control
-5-
-------
A first step in controlling fluoride emissions from gypsum ponds would include
segregation of gypsum and cooling ponds. The gypsum pond is substantially smaller
and pond segregation is a necessary part of applying control options 2 through 5.
Table 1-3 gives information on five gypsum pond fluoride control alternatives. The
economics of a change to the hemi/dihydrate process used overseas is questionable
with low cost and lower quality U.S. rock. Use of algae is at the scientific
feasibility evaluation stage. The dry chromatographic technique has been used in
secondary aluminum plants and appears to be applicable to phosphate fertilizer
plants.
1.2.3 Solid Waste Control
The three sources of solid waste in the phosphate fertilizer industry are by-
product gypsum, sludge and wet scrubber liquor. The solid waste can be disposed
through use of gypsum ponds and stacks, abandoned mine pits, or through sea
disposal. In the U.S. more than 90% of the plants use the gypsum pond-stack
technique.
The planning and building of a gypsum pond-stack system requires several
steps, including: evaluation of climatic and process constraints, sizing, field
evaluation of the subsurface conditions of a tentative site, final site selection,
conceptual and detailed design, and actual building. As a rule two ponds are used;
one is active while the other one is drained. Although several potential solid waste
resource recovery methods are technically feasible, none is economical. The major
environmental concern regarding the gypsum pond-stack system is the groundwater
contamination discussed earlier. Furthermore, gypsum stacks are sources of low
level radioactivity and EPA has classified them as special waste under the Resource
Conservation and Recovery Act (RCRA).
1.3 Recommended RD&D Projects
One of the major objectives of this study was to identify gaps in information
and to recommend RD&D projects regarding evaluation of control technology in the
phosphate fertilizer industry. The following discussion outlines the recommended
projects:
1. Epidemiological Study of the Phosphate Fertilizer Industry
In order to obtain information on the health hazards of the phosphate
fertilizer industry, a pilot epidemiological study is recommended. The
general strategy should consist of evaluation of routinely collected data
on mortality and morbidity. Special emphasis should be on cancer
because of the low level radioactivity of gypsum ponds and stacks, and
possible leaching of radioactive substances. If the pilot study shows
higher mortality or morbidity rates in the "exposed" population, a follow-
up study including evaluation of specifically collected data should be
carried out.
-6-
-------
TABLE 1-3. INFORMATION ON FIVE GYPSUM POND FLUORIDE CONTROL ALTERNATIVES
Process Where Applied
Kidde Barometric
condensers
Swift Absorption Barometric
condensers
Liming All cooling
pond water
Fluoride
Removal
Efficiency
% By-Product
95-98 (NH.LS.F
4 2 j 6
90 H2S.F4
90 None
Total
Capital
Investment
$ MM
2.6
1.3
2.1
Technical Pros
Improved quality of
phosphoric acid.
Increased recovery
of fluoride.
H2S.F6 recovery
may make process
profitable.
Simplest process.
and Cons
No full scale
installation.
Process is
rather complex.
Secondary
environmental
Conveyor
Pre-Calcination
of Rock
Gypsum filter N/A
after
acidulation
Crushed N/A
phosphate rock
None
None
1.1
23.8
problems.
Benefits/disadvantages difficult
to determine
Removes fluorides
before pro-
cessing.
High cost.
-------
2. Studies of Gypsum Pond Emissions and Chemistry
The gypsum pond appears to be a major environmental concern of the
phosphate fertilizer industry, yet the existing information on its emis-
sions and chemistry is scarce.
A study of gypsum pond chemistry is recommended since it could lead to
reduced fluoride emissions. The major objective of such a study would
be an understanding of the factors governing the fixation of fluorine in
the pond and cake.
A study of gypsum pond fluoride liquid-vapor equilibria is also recom-
mended since no reliable data exist. Although some in-field measure-
ment of fluoride emission from gypsum ponds has been carried out, more
data are needed for a reliable estimate of emission rates. A major
purpose of such measurements would be the correlation of emission rates
with atmospheric and plant conditions.
3. Evaluation and Optimization of Wet Scrubbers
There is little reliable information on the performance of wet scrubbers
and two performance evaluations are recommended. The first is
scrubber optimization through fractional factorial design experiments.
The main objective should be to optimize removal of fluoride and
particulate while minimizing energy consumption in wet scrubbing (pre-
ferably for the cross-flow packed type scrubber).
The second study is the performance evaluation of ionized wet scrubbers.
The main purpose of this evaluation would be to measure the particulate
removal efficiency as a function of the particle size distribution. The
verification of the claimed high efficiency of this scrubber in the health-
related submicron particle range is very important since its application
goes beyond the phosphate fertilizer industry.
4. Ammonia-Sulf uric Acid Mist Interaction Atmospheric Chemistry and
Dispersion Modeling Study
In the past some diammonium phosphate plants have experienced a
downwind opacity problem due to formation of ammonium sulfate
particles. An experimental study of the interaction between gaseous
ammonia and sulfuric acid mist in a smog chamber is recommended. A
model describing the interaction of ammonia and the acid mist plume
incorporating an aerosol dynamics model should be developed in this
study.
5. Demonstration of Dry Fluoride Removal System
The demonstration of a dry fluoride removal system is recommended
since its use could reduce the need for gypsum ponds. Several secondary
aluminum plants now use this system in which fluoride and fine
-8-
-------
participates lates are removed simultaneously. A major objective of this
project would be to demonstrate the applicability of the dry system in a
phosphate fertilizer plant and to obtain design parameters for the design
of a full scale system. A pilot model capable of handling 3000 NCMH
(about 1750 SCFM) applied on diammonium phosphate emissions is
recommended.
Optimization of Kimre Mist Eliminator
An efficiency evaluation of the Kimre mist eliminator is recommended
since several scrubber installations in the phosphate fertilizer industry
have been recently retrofitted with this device. The main purpose of the
evaluation program would be to verify the claimed high efficiency in
aerosol removal and to establish optimum operating conditions.
References
1. Nyers, 3.M., et al, Source Assessment: Phosphate Fertilizer Industry,
EPA 600/2-78-004, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, May 1978, 185 pp.
-9-
-------
2.0 CONCLUSIONS
There are two main conclusions from this evaluation of control technology for
the phosphate fertilizer industry. First, there is adequate control technology for
application to gaseous emissions from all five studied processes, and secondly,
gypsum pond emissions appear to be the major environmental problem facing the
industry since there are no simple nor economical means to control such emissions.
The following are more detailed conclusions:
1. Wet scrubbers, represented primarily by cross-flow packed scrubbers,
Venturi scrubbers and cyclonic spray scrubbers, are widely and success-
fully used for control of air emissions consisting primarily of fluorides
and particulate. A review of four representative vendors indicates that
there are several types of scrubbers with adequate efficiency to comply
with existing emission regulations.
2. Cross-flow packed scrubbers appear to be the best control alternative
applicable to all five processes since they demonstrate high efficiency
and relatively low capital, operating and maintenance costs. One of the
new applications in the industry is the use of ionized wet and nucleation
scrubbers which are effective in removal of submicron particulate
matter. The dry chromatographic system used in some other industries
also seems to be applicable to the phosphate fertilizer industry.
3. The only major reduction of fluoride emissions from stationary sources
is achievable through control of emissions from Run of Pile - Triple
Superphosphate (ROP-TSP) storage facilities which are, as a rule, uncon-
trolled, and through recovery of fluosilicic acid which would reduce
fluoride input to gypsum ponds.
^. In spite of the large numbers and size of wet scrubbing installations in
the phosphate fertilizer industry, reliable data on scrubber performance
are scarce. Tests based on factorial experiment design are needed to
develop reliable data and assist in optimization of scrubber performance.
5. The emissions from gypsum ponds appear to be the major environmental
problem facing this industry. The air emission load consisting primarily
of HF seems to be higher per ton of P205 produced than that from the
combined process sources controlled by wet scrubbers.
6. The leaching of radioactive substances, fluorides and phosphorus
compounds from gypsum ponds and stacks could result in aquifer
contamination. A study presently underway sponsored by the State of
Florida and the EPA should determine the distribution of contaminants
around a gypsum pond.
7. There are several potential control techniques for application to gypsum
pond emissions but none appears to be economically attractive. A first
step should include segregation of gypsum and cooling ponds, followed by
-10-
-------
some gypsum pond treatment technique. Single liming is probably the
simplest treatment method while the Swift process may be economical if
all by-product fluosilicic acid can be sold.
8. More research is needed before the environmental impact of gypsum
ponds can be fully assessed. The studies should include gypsum pond
chemistry, vapor pressure equilibria and correlation of fluoride emissions
with plant, pond and atmospheric conditions.
9. Solid waste, consisting primarily of by-product gypsum, appears to be
adequately disposed through stacking although there is some concern due
to its low level radioactivity.
-11-
-------
3.0 RECOMMENDATIONS
In order to further the knowledge of the environmental impact of the
phosphate fertilizer industry and specifically to improve the control technology, the
following studies are recommended to EPA:
1. An epidemiological study should be carried out to assess the impact of
the phosphate fertilizer industry on the population. Special emphasis
should be on the mortality and morbidity rates related to low level
radioactivity, specifically those related to bone cancer. Only a pilot
study is recommended as a first step.
2. Since the gypsum pond appears to be the major environmental concern
related to this industry, pond emisssions and chemistry should be studied.
Special emphasis should be placed on the chemistry of fluoride fixation
in the gypsum pond. An investigation of gypsum pond fluoride liquid-
vapor equilibria should be included. Field measurements of gypsum pond
fluoride emissions should be carried out in order to correlate these with
the plant operating conditions, atmospheric conditions, and gypsum pond
characteristics.
3. In order to improve the efficiency and economics of wet scrubbers, two
evaluation and optimization studies are recommended. A fractional
factorial design experiment is recommended for an evaluation and
optimization of a full scale scrubber in a representative plant. A
performance evaluation of the ionizing wet scrubber is also recom-
mended with special emphasis on submicron aerosol removal efficiency
in a representative DAP plant.
^. A study of the atmospheric chemistry of ammonia-sulfuric acid mist
interaction and a dispersion modeling study are recommended in order to
better understand and prevent opacity problems occurring downwind
from DAP plants.
5. A dry fluoride removal system with a capacity for treatment of 3000
NCMH (1750 SCFM) should be built in order to evaluate its performance
in the removal of fluorides and particulate matter.
6. Many scrubbing systems in the phosphate fertilizer plants have been
recently retrofitted with Kimre mist eliminators. An evaluation of this
equipment is recommended. Special emphasis should be placed on
obtaining data on performance and optimization.
-12-
-------
4.0 PROCESS AND EMISSION DESCRIPTION FOR FIVE PHOSPHATE FERTILIZER
MANUFACTURING PROCESSES
The phosphate fertilizer industry uses phosphate rock as its major raw
material. The primary objective of this industry is to convert the fluorapatite
(Ca10(PO4)6F,2) in phosphate rock to soluble P205, a form readily available for plant
use. The usual method of making the P205 content of phosphate rock available to
plants is by treatment with a mineral acid-sulfuric, phosphoric or nitric (1).
Figure 4-1 shows the major phosphate rock processing steps (2).
Phosphate fertilizer manufacturing processes are described in detail in several
major references (1), (3), (4), and (5). For a better understanding of emission control
technology, this section will briefly cover five processes with emphasis on emissions.
Detailed discussions of control technology are presented in Sections 5, 6, and 7.
4.1 Wet Process Phosphoric Acid (WPPA)
4.1.1 WPPA Process Description
Phosphoric acid is an intermediate product in the manufacture of phosphate
fertilizer. It is subsequently consumed in the production of triple superphosphate,
ammonium phosphate, superphosphoric acid, dicalcium phosphate and complex
fertilizer.
Most current processes use the dihydrate process in which phosphate rock is
treated with sulfuric acid under conditions where gypsum (Ca SO4 • H2O) is
precipitated (1).
The overall reaction in the dihydrate process is described by the following
equation:
3 Ca10 (P04)6 F2 + 30 H2SO4 + SiO2 + 58 H2O*30 CaSO4 • 2 H2O +
18 H3PO4 + H2SiF6
Major variations of the dihydrate process are the Prayon, Swenson, Dorr-Oliver and
Singmaster processes, with Prayon (designed by Davy Powergas) being used most
frequently.
Figure 4-2 presents a flow diagram for a typical wet process phosphoric acid
plant.*
Unreferenced figures were prepared by TRC based on the open literature
diagrams and some proprietary information.
-13-
-------
PHOSPHATE
ROCK
DEFLUORINATION'
•GRINDING.
-ACIDULATION (HoS04)
•ACIDULATION (HN03')—
-ACIDULATION (H3P04)
I
ELEMENTAL
PHOSPHORUS
PHOSPHORIC
ACID
ANIMAL FEEDS
FERTILIZERS:
DIRECT APPLICATION
NORMAL SUPERPHOSPHATE
NITRIC PHOSPHATES
TRIPLE SUPERPHOSPHATE
AMMONIUM PHOSPHATES
DIRECT APPLICATION
INDUSTRIAL &
FEED
CHEMICALS
Figure 4-1: Major Phosphate Rock Processing Steps (2)
-------
;3HOSPHATE ROCK
SULFURIC ACID DILUTION WATER
SULFURIC ACID
i ' 1
REACTOR/ATTACK
VESSEL
SCRUBBER [)>-
GASOMETER
PHOSPHORIC ACID
]
EVAPORATOR
r~
STEAM
HEAT EXCHANGER
I
BAROMETERIC
CONDENSER
STEAM
STEAM JET EJECTOR
GYPSUM
GYPSUM POND
CONDENSATE
-CONCENTRATED
PHOSPHORIC ACID
CONTAMINATED WATER
Figure 4-2: Flow Diagram for a Wet-Process Phosphoric Acid Plant
-------
In the WPPA process finely-ground phosphate rock is continuously metered
into the reaction vessel and sulfuric acid is added. The reaction vessel has several
ro npartments needed for completion of the reaction, during which the sulfate ion
• nust be closely controlled for a proper crystal growth required for easy filtration.
The heat of reaction is removed from the reactor by vacuum flash cooling.
The reaction slurry is filtered in the rotary horizontal tilting-pan vacuum
filter. The 32% acid obtained from the filter is concentrated in a two or three stage
vacuum evaporator system to about 54%. The evaporators can be equipped with
scrubbers for the collection of H2SiF6 which can be produced at 20% by weight
concentration. This recovery feature is not necessary to the evaporation and is a
matter of economics; it would, however, substantially reduce fluoride transfer to a
gypsum pond.
Vapors from the vacuum flash cooler are condensed in a barometric condenser
and sent to a hot well while the non-condensibles are removed by a steam ejector
and are also vented to the hot well.
4.1.2 Emissions from WPPA
Fluoride emissions from wet-process phosphoric acid (WPPA) will vary depend-
ing upon the type of rock treated and process used. Gaseous fluorine emissions
consist of silicon tetrafluoride generated in the reaction and evaporation phases.
Hydrogen fluoride formed in the reactor is converted to SiF4 according to the
- = act ion (5):
4 HF + SiO2 SiF4 + 2 H2O
Table 4-1* shows a material balance depicting final distribution of the fluorine from
the rock. Wet scrubbers are used for control of air emissions at phosphoric acid
plants. They will be evaluated in Section 5.
Table 4-2 shows stack height and controlled emission factors for WPPA based
on actual data from 10 plants (3). The average emission factors were calculated for
plants with and without recovery of by-product fluorine. Results show that the
average emission factor for the wet scrubber at plants recovering fluorine is half
the amount in plants without recovery.
The emissions of fluorides from gypsum ponds will be evaluated more in
Section 6. A close look at Table 4-2 reveals that the gypsum pond is by far the
largest source of fluoride emission at the WPPA plant.
*Metric units of measurement will be used throughout this report except
where English units were used in the original citation.
-16-
-------
TABLE 4-l.J-LUORINE MATERIAL BALANCES FOR WPPA MANUFACTURE (3)_
Fluorine, 106 R/day
Material
Balance
Aa
Bb
Phosphate
Rock
48.3
127.0
Product
Acid
36.5
16.3
Gypsum
Slurry
And
Process
H20
11.8
110.7
Air
Emission
0.004
0.009
Plant
Daily
Production,
Metric
tons P2O5
368
907
Fluorine
Emission
Factor
g/kg P2O5
0.011
0.024
Data obtained from the public files at the Florida Department of Environmental
Regulations in Winter Haven, October 1976.
Data from Reference (6).
TABLE 4-2. AVERAGE STACK HEIGHTS AND CONTROLLED EMISSION FACTORS
FOR WET PROCESS PHOSPHORIC ACID (3)
Emission Factor, g/k P205
Emission Point
Wet process phosphoric acid:
Rock unloading
Rock transfer and conveying
Wet scrubber system:-
Overall
With recovery of fluorine
Without recovery of fluorine
Gypsum pond
Stack
Height, Total
m Fluoride
12 0
21 0
0.0 10 + 47%
0.0059 + 61%
0.012 + 60%
0.025 to 2.5
(avg 0.50)
Particulate SOx
0.15 + 250% 0
0.045 + 180% 0
0.054 + 164% 0.032 + 200%
a a
a a
0 0
alnsufficient data to determine the effect of fluorine recovery on other
emission species.
-17-
-------
Particulate emissions generated in the reactor consist of unreacted rock and
insoluble salts. The participate emission factor shown in Table 4-2 is based on
measurements at five plants.
The origin of SOX emissions in WPPA manufacture is not clear. The SOX
emission factor in Table 4-2 is based on data from one plant and a Public Health
Service document (7).
4.2 Superphosphoric Acid (SPA)
4.2.1 SPA Process Description
Phosphoric acid produced by the wet process is in ortho form and it can be
converted to polyphosphoric acid through heating. The chemical reaction occurring
during heating is described by:
The product of this reaction is a mixture of orthophosphoric acid and dehydrated
acid called superphosphoric acid. WPPA is concentrated to SPA primarily to reduce
the cost of transportation, but additional benefits are that SPA is less corrosive and
contains fewer impurities than WPPA.
Two commercial processes are used for production of SPA from WPPA:
Submerged Combustion (about 25% of the total production capacity) and Vacuum
Evaporation (about 75% of the total production capacity).
The Submerged Combustion process accomplishes dehydration by bubbling hot
combustion gas through a pool of the acid. It will not be discussed in this report
since vacuum evaporation is more commonly used.
Two vacuum evaporation processes are in commercial use:
1. The falling film (Stauffer) process; Figure 4-3
2. The forced circulation (Swenson) process; Figure 4-4
Both processes are similar and use high vacuum concentrators with high-pressure
steam to concentrate acid to 70% P2OS. Feed acid is introduced into a large volume
of recycling product acid to reduce corrosion. In both systems, product acid is
pumped to a cooler before being sent to storage or shipped. An appropriate recycle
feed ratio is 150:1 in the Swenson process (compared with 80:1 for the Stauffer
process). Both the hot well and cooling tank are vented to a wet scrubbing system.
-18-
-------
vO
I
STEAM
FALLING-FILM EVAPORATOR
CONDENSATE
TO BOILER
54% P203
CONCENTRATED ACID
FEED TANK
EVAPORATOR
RECYCLE TANK
EJECTORS
BAROMETER CONDENSER
HOT
WELL
SUPERPHOSPHORIC ACID
COOLANT
DISCHARGE
COOLING TANK
Figure 4-3: Stauffer Evaporator Process (8)
SUPERPHOSPHARIC
, ACID
(68-72%
-------
n
DOWTHERM
HEATER
WATER
SURGE
TANK
54%
CLARIFIER
F.C.
EVAPORATOR
AIR EJECTOR
CONDENSER
SiF4, HF
T SiF
I
HOTWELL
, HF
A
'i
"i
ii
f
•
••••
.i
T
i
1
i
t
1
1
1
1
M^
1
|
I
19'
O'O
-i
Tr
t-\
•M^
/
— LUULiNb WMItH
OUT
~1
72% P205
STORAGE
COOLING
TANK
Figure 4-4: Swenson Evaporator Process (8)
-------
4.2.2 Emissions from SPA
Emissions from the SPA plant consist primarily of gaseous fluorides and
particulates. Hydrogen fluoride is the major gaseous pollutant, while particulates
are limited to liquid phosphoric acid aerosols and mists produced by the condensa-
tion process. The falling film evaporator can generate submicron aerosols which are
difficult to remove from the gas stream.
Fluorine emission data for two plants using vacuum evaporation were
evaluated by Monsanto (3). The barometric condenser, hot well and product cooling
tank are vented to a two-stage wet scrubber. Fluorine emission factors from these
plants are 0.0036 and 0.011 g/kg P2O5, with an average value of 0.0073 g/kg P205.
One plant reported participate emissions ranging from 0.011 to 0.055 g/kg P2Os.
4.3 Diarnmonium Phosphate (DAP)
4.3.1 DAP Process Description
The typical ammonium phosphates used as fertilizer contain from 11% to 21%
nitrogen and 20% to 55% P2O5 (9). In 1975, 84% of total production (on a P2O5 basis)
consisted of DAP grade.
Diammonium phosphate is manufactured from phosphoric acid and ammonia.
The process consists of three basic steps, including reaction, granulation, and finish-
ing operations such as drying, cooling, and screening.
In the first step, anhydrous ammonia is reacted with phosphoric acid to form
hot liquid DAP as described by the equation (10):
2NH3 + H3PQ4-»-(NH4)2 H PO4 - 81,500 BTU (20,529,850 calories)
(gas) (liquid) (liquid) (AHR (d 60 F)
The most frequently used process for diammonium phosphate manufacture is
TVA's, shown in Figure 4-5. Phosphoric acid is mixed in an acid surge tank with 93%
sulfuric acid (used for product analysis control) along with recycle and acid from
wet scrubbers. Mixed acids have a P2O5 content of 40% to 45% (11). This analysis is
attained by mixing unconcentrated filtered WPPA, 28.7% P2O5, and concentrated
WPPA, 53.3% P2O5.
Mixed acids are then partially neutralized with liquid or gaseous anhydrous
ammonia in a brick-lined acid reactor. In this agitated atmospheric pressure tank,
the mole ratio of NH3: H3PQ4 is maintained at 1.3:1.0 to 1.5:1.0. All phosphoric acid
and approximately 70% of the ammonia are introduced in this vessel. In this molar
range, ammonium phosphates are most soluble, allowing further concentration of
solution while maintaining adequate flow characteristics. The heat of reaction is
used in this vessel to maintain a temperature of 100°C to 120°C and to evaporate
-21-
-------
EMISSIONS GASES
GYPSUM POND WATER
OUT
FILTERED
PHOSPHORIC ACID
GYPSUM POND WATER
N)
KJ
CONCENTRATED
PHOSPHORIC ACID
SULFURIC
ACID
ANHYDROUS AMMONIA
AMMONIATOR
GRANULATOR
ACID
STORAGE
TANK
PRODUCT
TO
STORAGE
BAGGING, OR
BULK SHIPMENT
L-DUST SUPPESSANT
UNDERSIZE
Figure 4-5: TVA Ammonium Phosphate Process Flow Diagram
-------
excess water. A slurry which is primarily MAP and contains 18% to 22% water is
produced and flows through steam-traced lines to the ammoniator-granulator. To
assure no leakage from the reactor, the vessel is ventilated with outside air. In
theory, the reactor could be designed without ventilation or atmospheric discharge,
but in practice, ventilation rates of 57 to 71 m3/min (standard conditions) are com-
mon. Ventilation rate is determined by reactor mechanical design, not process
requirements. Ammonia-rich offgases from the reactor at 77°C to 82°C are wet-
scrubbed before exhausting to the atmosphere. Primary scrubbers use raw material-
mixed acids as scrubbing liquor, and secondary scrubbers use gypsum pond water as
scrubbing liquor.
The basic rotary-drum ammoniator-granulator consists of an open-end, slightly
inclined rotary cylinder with retaining rings at each end and a scraper or cutter
mounted inside the drum shell. Drums vary in diameter from 2 m to 3 m and in
length from 3 m to 6 m. A rolling bed of recycled solids is maintained in the unit.
Slurry from the reactor is distributed above the bed while the remaining ammonia
(approximately 30%) is sparged underneath to bring the final NH3: H3PO4 mole ratio
from 1.8:1.0 to 2.0:1.0. Granulation by agglomeration and by coating particles with
slurry takes place in the rotating drum and is completed in the dryer. Recycle rates
of 2..5 to 4.0 kg recycle/kg product are typical for this type of unit. As with the
reactor, the granulator theoretically could be designed without ventilation, but to
prevent NH-, leakage, approximately 8.5 x 10~4 m3 (standard conditions) per metric
ton PjOj air leakage is allowed into the granulator around inlet and outlet
connections.
The temperature of granular DAP in the rotary drum reaches 85°C to 105°C,
while the temperature of offgases reaches 38°C to 77°C. Ammonia-rich offgases
pass through a wet scrubber before exhausting to the atmosphere.
Moist DAP granules are transferred to a rotary oil- or gas-fired cocurrent
dryer which reduces product moisture content to below 2%. The product is then
cooled to below 35°C. Cooling minimizes caking and product dissociation during
storage. The temperature of offgases from the dryer ranges from 82°C to 104°C,
and the temperature of offgases from the cooler ranges from
-------
4.3.2 Emissions from DAP
Air emissions sources and their related emission species from DAP are (3):
Reactor; ammonia, fluorides.
Ammoniator-granulator: ammonia, fluorides, particulates.
Dryer; ammonia, fluorides, particulates, combustion gases.
Cooler; ammonia, fluorides, particulates.
Product sizing and material transfer; particulates.
Gypsum pond; fluorides.
Ammonia emissions are due to incomplete chemical reaction and excess free
ammonia. While the ammonia emissions alone do not present an environmental
hazard, there is a potential additional effect. Almost all DAP plants also
manufacture sulfuric acid at the same facility. Under adverse atmospheric
conditions, ammonia and sulfuric acid mist can react in the atmosphere. The
resulting ammonium sulfate could affect visibility downwind from the plant. The
analysis of this potential environmental hazard will be addressed as a proposed R&D
project in a separate section (Section 9).
Fluoride emissions originate from the fluoride content of phosphoric acid and
consist primarily of silicon tetrafluoride (5). Dryer offgases contain combustion
products in negligible quantities. Table 4-3 shows emission factors for ammonium
phosphate manufacture. All emissions listed reach the atmosphere through a stack.
Product sizing and material transfer process steps produce both stack and fugitive
emissions. An uncontrolled emission factor of 2.2 g/kg P2O3 for this source has been
reported (12). The emission factors in Table 4-3 are based on literature and data on
file at the Florida Department of Environmental Regulations, presently located in
Tampa, Florida. Total plant stack emission factors were also calculated as (3):
Particulate: 1.5 g/kg P2O5
Fluoride (as F): 0.038 g/kg P2O5
Ammonia; 0.068 g/kg P2O5
Information on fluoride emissions from gypsum ponds, reported as 0.025 to 2.5 g/kg
PzOj in Table 4-2 for WPPA, also applies here. One half of forty-eight ammonium
phosphate plants produce WPPA. Plants producing only DAP have smaller ponds
with lower fluoride emissions (3).
4.4 Normal Superphosphate (NSP)
4.4.1 NSP Process Description
Because of low P2O5 content the popularity of NSP has decreased from 68%
(total phosphate fertilizer market share) in 1957 to about 18% where it is now
stabilized. The two raw materials used in the production of NSP are 65% to 75%
sulfuric acid and ground phosphate rock. The basic chemical reaction is shown by
the following equation:
-24-
-------
TABLE tt-3. EMISSION FACTORS FOR DIAMMONIUM PHOSPHATE PLANTS (3)
Controlled Emission Factors
Emission Point
Reactor /ammoniator-granulator:
Fluoride (as F)
Participate
Ammonia
Dryer /cooler:
Fluoride (as F)
Particulate
Ammonia
Product sizing and material transfer:
Fluoride (as F)
ParticuJate
Ammonia
Reported as total plant emissions:
Fluoride (as F)
Particulate
Ammonia
Mean
K/kg P2OS
0.023
0.76
_b
0.015
0.75
_b
0.001
0.03
_b
0.038d
0.1 5e
0.068
95% Confidence
Interval, % of Mean
+80
790
_b
+ 160
+60
_b
_c
c
Jb
+30
120
+75
Fugitive emissions are included in the text.
No information available; although ammonia is emitted from these unit operations,
it is reported as a total plant emission.
Emission factor represents only 1 sample.
A fluoride emission guideline of 0.03 g/kg P2O5 input has been promulgated by EPA.
eBased on limited data from only 2 plants.
-25-
-------
[Ca3(P04)2]3 • CaF2 + 7 H2SO4 + 3 H20^3[CaH4 (PO4)2 • H20]
+ 7 CaSO4 + 2 HF
Figure 4-6 shows an NSP process flow diagram showing a TVA double conical
mixer used by 80% of NSP plants. Ground phosphate rock is fed into the double
conical mixer where it is mixed with diluted sulfuric acid. The heat of sulfuric acid
dilution serves to heat the sulf-iric acid to a proper reaction temperature. The fresh
superphosphate discharges from the conical mixer to a pug mill for additional mixing
of acid and rock before discharge to a den. NSP product is kept in the den for one
hour for solidifying. The cutter slices the solid mass of crude product so that it may
be conveyed to the pile storage for "curing." The product takes about 10-20 days to
reach an acceptable P2O5 content. The hardened run-of-pile (ROP) product obtained
in the above process may be granulated. In order to granulate ROP product, it is
pulverized and sent to a rotary drum granulator where steam and water are added.
The mixture is passed through a rotary dryer and cooler to final storage. Most of
the NSP is used in fertilizer mixing plants where it serves as one of the ingredients
in preparation of nitrogen-phosphorus-potassium (NPK) mixtures.
4.4.2 Emissions from NSP
The major sources of emissions at an NSP plant are the mixer, the den, and the
curing building. Fluoride emissions consisting of hydrogen fluoride and silicon
tetrafluoride are controlled by scrubbing with water. Most NSP plants presently
practice fluorine recovery, eliminating the need for a gypsum pond (3).
Table 4-4 shows emission factors for an NSP plant based on source test
data (3). Data were available for 4 tests for controlled emissions from the curing
building, but most emissions from curing buildings are uncontrolled. The emission
factor was therefore derived using a control efficiency of 97%.
4.5 Triple Superphosphate (TSP)
4.5.1 TSP Process Description
Triple superphosphate with 45 to 49% P2O5 content has replaced most NSP
because of transportation economy. Run-of-Pile (ROP) and Granular Triple Super-
phosphate (GTSP) are two principal types of TSP. Both processes achieve high P2O5
content by replacing sulfuric acid with phosphoric acid. The basic chemical reaction
is shown by the following:
Ca3 (PO4)2 + 4 H3P04 + 3 H2O -»-3 Ca (H2PO4)2 • H2O
The ROP process is essentially identical to the NSP process shown in Figure 4-6 and
will not be discussed here.
-26-
-------
TABLE 4-
-------
N)
00
I
SULFURIC ACID
EMISSIONS
A
H DUST
COLLECTOR
T
ROCK
UNLOADING
EMISSIONS
A
GROUND PHOSPHATE ROCK I
1 !
ROCK
BIN
WEIGHER
DUST COLLECTOR
DUST '
— ^^
RECYCLED
AP T n PHMTDni
ML1U LUN 1 KUL
i
i
i
MIX
TO ROCK BIN
aT*
•^
WET
SCRUBBER
CZ
. I I
'I PUG'MILL I I I
'----*-- —1 j /-CUTTER
I
DEN
-*-RECYCLED WATER
^-PRODUCT
CURING
BUILDING
^T^SEVATOR
p
Figure 4-6: Normal Superphosphate Process Flow Diagram
-------
The GTSP process shown in Figure 4-7 is quite different. A lower
concentration acid than that used in ROP is used for reaction with ground phosphate
rock in an agitated tank. This lower strength acid maintains the resultant slurry in a
fluid state and allows the chemical reaction to proceed further. Two methods,
Dorr-Oliver and TVA, are used in GTSP production with the former (shown in Figure
4-6) accounting for over 30% of total production. A thin siurry obtained in this
process is continuously removed and distributed onto dried granules to build up their
size. Granulation is carried out in pug mills and rotating drums. Slurry-wetted
granules are discharged onto a rotary dryer and are cooled before being sent for
curing on a storage pile.
In the TVA granulation process, acidulation and granulation are carried out
simultaneously in a rotary drum. Preheated acid, rock and recycled fines are mixed
in the drum and granulated with steam. The product is cooled, screened and cured.
4.5.2 Emissions from TSP
The major sources of emission in an ROP-TSP plant are the mixer, den, curing
building and gypsum pond. All sources release fluoride and particulates are released
by all sources except the gypsum pond. Emissions of fluorides are controlled by wet
scrubbers that discharge waste to the ponds. Table 4-5 shows emission factors for
an average ROP-TSP plant based on source test data (3). An estimated 8 gF/kg P2O5
are released during production and curing of ROP-TSP. The emission points in GTSP
plants are the reactor, granulator, dryer, cooler, screens, mills and storage. Again,
most sources release fluorides and particulate. The dryer exhaust also contains SO,
from combustion of fuel oil.
Table 4-6 shows emission factors for a GTSP plant (3). Emissions from the
reactor, granulator, dryer, cooler, screens and mills at an average plant are vented
to a common stack. As a result Table 4-6 does not show individual emission factors.
It is estimated that 7 gF/kg of GTSP are released during production and curing.
4.6 Summary of Fluoride Emissions
Table 4-7 summarizes the phosphate industry's fluoride emissions. It shows
soluble fluoride evolution and emission associated with the 1970 and projected
production levels. It should be noted that the emission factors of the individual
processes include gypsum pond emissions (4).
4.7 References
1. Final Guideline Document: Control of Fluoride Emissions from Existing
Phosphate Fertilizer Plants, EPA-450/2-77-005, U.S. Environmental Pro-
tection Agency, Research Triangle Park, North Carolina, March 1977.
-29-
-------
ROCK BIN
WET PROCESS
PHOSPHATE ROCK
GROUND
PHOSPHATE ROCK
MISSIONS!
i
EMISSIONS
I
ACID CONTROL
O
STACK
SCRUBBER
~> r -«— -HSCRUBBERH*- i
T f CRUSHER-^ t
I OVERSIZED. \ F^DUST CYCLONE
).J SCREEN
"^ ' l "PARTICULATE
SCRUBBER
SCRUBBER
(DUST CYCLONE
. j RECYCLED TO
| I GRANULATOR
1 L_—
GRANULATOR
AIR
FUEL
EjJD^ERT&IJELEVATOR
PRODUCT
SCREEN
CYCLONE
OOLER ]^^"^^ |
STACK
RECYCLED
POND WATER
CURING BUILDING
Figure 4-7: Granular Triple Superphosphate Process Flow Diagram
-------
TABLE 4-5. EMISSION FACTORS FOR AN AVERAGE RUN-OF-PILE TRIPLE
SUPERPHOSPHATE PLANT BASED ON CONTROLLED EMISSION
SOURCES (3)
Emission Source
Rock unloading
Rock feeding
Cone mixer, den, curing building
Emission Factor, g/kg P2O5
Particulates
0.07b
0.014+ 170%
0.16 + 50%
Fluorides0
_c
_c
0.10 + 40%
Fluoride released as a vapor.
Based on two sets of data; therefore, 95% confidence limits could not be
determined.
cNot emitted from this source.
TABLE 4-6. EMISSION FACTORS OF CONTROLLED EMISSIONS
FOR AN AVERAGE GRANULAR TRIPLE SUPER-
PHOSPHATE PLANT (3)
Emission Source
Rock unloading
Rock feeding
Reactor, granulator, dryer
sooler, screens
Curing building
Emission Factor, g/kg P2Os
Particulates Fluorides
0.09b c
0.017 + 180% c
0.05 + 320% 0.12 + 30%
0.10 + 240% 0.018 + 40%
sox
c
c
K
1.86b
_c
Fluoride released as a vapor.
Based on two sets of data; therefore, 95% confidence limits could not be meaningfully
calculated.
°Not emitted from this source.
-31-
-------
TABLE fr-7. PHOSPHATE INDUSTRIES FLUORIDE EMISSIONS
1970 Production
(104 tons/yr)
Projected 2000
Production
(10* tons/yr)
Soluble Fluoride
Evolution Factor
Soluble Fluoride
Emission Factor
with Current
Control
Soluble Fluoride
Emission Factor
1 with 99% Control
Vj")
I
Soluble Fluoride
Evolved in 1970
(10' Tons F/year)
Soluble Fluoride
Evolved in 2000
(103 tons/year)
Soluble Fluoride
Emissions in 1970
Wet Process
Phosphoric Acid
3.8
4.07(C)
Ib F/ton P,OS
in Product
3.36
Ib F/ton P,Oj
in Product
Ib F/ton P,05
in Product
7.73
26.4(C)
6.38(C>
Diammonium
Phosphate
7.0(B)
Ib F/ton P,O5
in Product
0.23
Ib F/ton P,0j
in Product
0.013(C)
Ib F/ton P,Qs
in Product
1.57
4.59
0.28(C)
Triple
Superphosphate
2.7(B>
21.2(C)
Ib F/ton P2Os
in Product
5.4
Ib F/ton P,O5
in Product
0.21(C)
Ib F/ton P,Os
in Product
14.8
28.6
3.78
Normal
Superphosphate
0.7(B)
0.2
Industry
Totals
10.4(B)
263(B)
16to9.3(C'E>
Ib F/ton PjOs
Equiv. in
Products
4.1 to 3.l'C>E*
Ib F/ton P,Os
Equiv. in Products
0.09
-------
2. Teller, A.J., Control of Gaseous Fluoride Emissions, Chemical
Engineering Progress, 63:75-79, March 1967.
3. Nyers, J.M., et al, Source Assessment: Phosphate Fertilizer Industry,
EPA-600/2-78-004, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, May 1978, 184 pp.
4. Robinson, J.M., et al, Engineering and Cost Effectiveness Study of
Fluoride Emissions Control, PB 207-506, prepared by TRW Systems
Group for U.S. Environmental Protection Agency, January 1972.
5. Air Pollution Control Technology and Costs in Seven Selected Areas,
EPA-450/3-73-010, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, March 1973, pp. 11-192.
6. King, W.R., Ferrell, J.K., Fluoride Emissions from Phosphoric Acid Plant
Gypsum Ponds, EPA-650/2-74-021, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, October 1974, 329 pp.
7. Atmospheric Emissions from Wet-Process Phosphoric Acid Manufacture,
AP-57 (PB 192 222), U.S. Department of Health, Education, and Welfare,
Raleigh, North Carolina, April 1970, 86 pp.
8. Weber, W.C., Pratt, C.J., Wet Process Phosphoric Acid Manufacture In:
Chemistry and Technology of Fertilizers, Sauchelli, V. (ed).
New York, Reinhold Publishing Corporation, 1960, p. 224.
9. Riegel's Handbook of Industrial Chemistry, Seventh Edition, T.A. Kent,
(ed). Van Nostrand Reinhold Co., New York, NY, 1974, pp. 551-569.
10. Inorganic Fertilizer Materials and Related Products, M 28B (75)-13, U.S.
Department of Commerce, Washington, D.C., December 1976, 6 pp.
11. Shreve, R.N., Chemical Process Industries, Third Edition, McGraw-Hill
Book Co., New York, NY, 1967, pp. 274-277.
12. Rawlings, G.D., Reznik, R.B., Source Assessment: Fertilizer Mixing
Plants, EPA 600/2-76-032c, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, March 1976, 187 pp.
-33-
-------
5.0 STATE-OF-THE-ART AIR EMISSIONS CONTROL TECHNOLOGY
Phosphate fertilizer plants are rather unique in their air pollution control
requirements. The air pollution control equipment cost for some processes may add
up to over 50% of a plant's total capital investment. (In the words of one plant
manager in central Florida, when building a fertilizer plant, first you erect the
scrubbers and then you plan the rest of the equipment around them.) This section
will first review the theory and practice of the \vet scrubber, by far the most
important control apparatus used in the fertilizer industry, and will then examine
each of the five phosphate fertilizer processes for specific air pollution controls. A
separate review of air pollution control equipment vendors will be presented in
Section 8.
5.1 Theory and Practice of Wet Scrubbing
As can be seen in Section 4.0, the major air emissions from fertilizer plants
consist of gaseous fluoride (primarily HF and SiF4) and particulate matter. Air
pollution control equipment can in general be divided into wet and dry cleaning
devices. Dry gas cleaning devices, primarily represented by baghouses and electro-
static precipitators, are usually used in phosphate rock processing which is beyond
the scope of work for this study. Phosphate fertilizer manufacturing processes start
with phosphate rock acidulation which results in gaseous fluoride and particulate
emission. Since wet scrubbers can simultaneously remove both emissions, their use
in the fertilizer industry is widespread.
Wet scrubbers are air pollution control devices in which gaseous pollutants
and/or aerosols are removed from the gas stream through scrubbing by a liquid. The
wet scrubbers which are likely to perform well in phosphate fertilizer plants include
spray towers, Venturi scrubbers, cross-flow packed scrubbers and impingement
scrubbers. The physical and chemical phenomena by which contaminants are
removed from the gas are called unit mechanisms and are somewhat different
depending on whether gas or aerosol pollutants are involved.
Table 5-1 shows the advantages and disadvantages of wet scrubber devices.
5.1.1 Gaseous Fluoride Emissions Control
The basic unit mechanism for removal of a contaminant gas from the air is
mass transfer, which can involve any or several of the transfer phenomena, such as
molecular diffusion, turbulent diffusion, etc. The gas is transferred from one
homogeneous phase (gas) to another (liquid). The driving force for this operation is
concentration gradient. When mass transfer is in a direction from gas to liquid, the
operation is called absorption; if the mass flows in the opposite direction, the
-Deration is called desorption or stripping.
-34-
-------
TABLE 5-1. ADVANTAGES AND DISADVANTAGES OF
WET AIR AND GAS CLEANING DEVICES
(1)
ADVANTAGES:
DISADVANTAGES:
1. Gases and particles collected together.
2. Soluble materials may be readily collected and
pumped.
3. High temperature gases cooled.
**. Corrosive gases and mists may be neutralized.
5. Fire or explosion hazard eliminated.
1. May require recrystallization for soluble
particles.
2. May require sludge pond.
3. Dissoluble particle recovery requires liquid filter.
'f. Particles IMTI not easily collected.
5. Freezing problems.
6. Liquid entrainment in effluent frequent problem.
7. Cleaned air may not be suitable for recirculation,
high dewpoint causes condensation.
-35-
-------
The gas on its way from bulk gas to bulk liquid and to final elimination must
overcome three resistances (2):
1. Diffusion through a gas phase film
2. Diffusion through a liquid phase film
3. Chemical reaction rate.
Any one or a combination of the three resistances can be the rate controlling step.
The rate of mass transfer N. is:
NA = KG (PG - Pi> = KL (Ci - CL>
where
N. = mass transfer
KG = gas film mass transfer coefficient
K. = liquid film mass transfer coefficient
P.- • = pressure, G-gas, i-interphase
u,i
C- . = concentration, i-interphase, L-liquid
i,L
The experimental determination of the individual coefficients K~ and K, is diffi-
cult. When the equilibrium line is straight, overall coefficients similar to those used
in heat transfer can be used and they can be defined from the standpoint of either
the gas or liquid phase. Each coefficient is based on a calculated overall driving
force.
One of the most important engineering parameters in the design of a wet
packed scrubber is the height of the packed section. A useful method for computing
the height of a packed tower is based on the transfer unit concept (3). The physical
meaning of a transfer unit is that it is a section of a packed tower that achieves a
change in composition equal to the driving force in that section (4). The advantage
of this method is that it is much simpler than the use of mass transfer coefficients.
In the case of packed towers which are frequently used in phosphate fertilizer
plants, the following mass transfer relationships have been established:
r°'8
G
-36-
-------
where:
K_ = mass transfer coefficient, Kg moles/hr, m3, atm
G = gas mass flow rate, Kg moles/hr m2
Therefore:
Z
NT ~ ,,0.2
1 Wj
NT = number of transfer units
Z = tower height, m
Thus, the number of transfer units obtainable would be controlled by the height of
the tower, but the number of transfer units usually also increases as the liquid mass
flow rate is increased.
Table 5-2 shows HF absorption data for various wet scrubbers (5). These
scrubbers will be described in more detail in Section 5-2.
An important consideration in gaseous pollutant removal is the performance of
equipment. It is sometimes difficult to compare the performance of two basically
different types of equipment in terms of mass flow rates, height of transfer unit, or
mass transfer coefficients. In order to overcome this difficulty, the performance of
equipment has been studied in terms of the number of transfer units. The effect of
liquid and gas flow rates is also expressed in terms of theoretical power consumed
per unit of gas flow rate, as power consumption is usually of more economic concern
than liquid or gas mass flow rate. Such relationships are mainly a matter of conve-
nience and do not necessarily have a theoretical basis.
Figures 5-1 and 5-2 show the relationship between number of transfer units
and power consumption in absorption of HF and SiF4. It has been shown that the
number of transfer units obtainable on grid towers is controlled principally by tower
height and is only slightly affected by power expended on the liquid and gas phases.
The performance of cyclone spray scrubbers is primarily a function'of power
expended in the liquid phase and is essentially independent of the power expenditure
in the gas phase. Performance of Venturi scrubbers, on the other hand, depends
largely on the power expended in the gas phase but is slightly affected by liquid
power expenditure. These results are useful in characterizing the dominant factors
in the performance of equipment used in the absorption of gaseous fluorides.
Nearly all usable data from the absorption of hydrogen fluoride are based upon
application of spray towers. The performance of this equipment appears to be
dominated by the power expended on the liquid phase, as was the case with the
cyclone scrubber. Significant differences in performance among the various spray
towers in use were found. Wet-cell washers require a higher power consumption
than simple spray towers with the same performance.
-37-
-------
TABLE 5-2. HYDROGEN FLUORIDE ABSORPORATION DATA (5)
00
Absorbing
Type of Equipment Liquor
Cross flow spray Water
Cross flow spray Lime water
Counterflow spray Water
Parallel flow spray Lime water
Counterflow spray Water
Venturi Water
Venturi Water
G = Gas mass flow rate
L = Liquid mass flow rate
Pg = Power in gas phase
PI = Power in liquid phase
Kga = Mass transfer coefficient
G, Lb/(Hr)
(Sq. Ft.)
2,110
1,880
2,080
1,830
1,400
2,050
2,000
13,800
2,000
76,000
-70,000
L, Lb/(Hr)
(Sq. Ft.)
72
72
103
84
92
105
800
3,800
380
42,000
-40,000
65,000
Pg, Hp.,
/M Cu.
Feet/Min
-3.0067
-0.0056
-0.0067
-0.0067
-0.0061
-0.006
-0.2
-0.23
0.24
4.7
2.1
2.4
2.9
3.5
4.0
4.5
Pt, Hp.,
/M Cu.
Feet/Min
0.0089
0.0098
0.0067
0.017
0.017
0.013
0.10
0.017
0.02
0.071
0.074
0.11
0.071
0.095
0.12
0.13
Kga Lb. Mole
(Hr) (Cu. Ft.)
(Atm.)
-11
-12
-12
-15
-25
-35
9
51
4
-
-
-
-
_
_
Nr
-0.33
-0.38
-0.25
-0.62
-1.09
-1.50
~5.85
2.58
2.5
2.9
2.0
2.7
2.3
3.0
3.9
2.3
Nt = Number of transfer units
-------
100
00
t—
*— «
z
10
o
CC
LU
00
0.1
0.001
Tmr
0.01
i mill—i i i mi
1 LEGEND
AVENTURI i
O SPRAYS •
DWET CELL-
0.1
1 ' ' "
1.0
POWER INTRODUCED IN LIQUID PHASE
OR IN GAS PHASE (VENTURI ONLY) - Hp/MCFM
10.0
Figure 5-1: Power Consumed in HF Absorption (5)
-------
100
10
OO
•f.
at.
LU
CO
0.1
0.001
: ' ' """1
•
_
•
•
»
; —
1 1 1 I 1 I III
| | 1 I 1 1 II 1 1 1 1 1 > ' HI |
LEGEND
O CO & COUNTERFLOW
D PACKED TOWER
A JET
O ° A
D °
i i i 1 1 1 ul i i i 1 1 1 nl i
, . iii.^
•
SPRAY ]
A :
—
0.01 0.1 1.0
TOTAL POWER INTRODUCED - Hp/MCFM
10.0
Figure 5-2: Power Consumed in SiF4 Absorption (5)
-------
10
LO
o
o
0.1
T 1—I I I I I
T 1 1 I I
HIGH SAT'D TEMPERATURE (60°C)
SAT'D TEMPERATURE (38°C)
J—I—I I I 111
J—I—I I I I 11
10
INLET CONCENTRATION MGM AS F/SCF
FLUORINE SCRUBBING WITH
LOW FLUORINE CONTENT LIQUOR
100
Figure 5-3: Inlet Versus Outlet Fluoride Concentration
in Cyclonic Spray Tower (6)
-------
The performance of spray towers absorbing silicon tetrafluoride is not consis-
tent with simple gas absorption. One possible explanation is that mists are formed
in the tower, which are collected primarily in the entrainment separators just prior
to emergence from the tower. The mist is probably rather coarse, however, because
high-power consuming devices such as jet scrubbers do not exhibit substantially
better performance than the low-power-consuming spray towers.
In general, the extent of the fluorine abatement system required is determined
by the following parameters:
1. the inlet fluorine concentration,
2. the allowable fluorine emissions,
3. the outlet or saturated gas temperature,
4. the composition and temperature of the scrubbing liquid,
5. the scrubber effectiveness and number of transfer units,
6. the fluorine compounds present, and
7. the effectiveness of entrainment separation.
The inlet and allowable outlet fluorine concentrations must first be established
to determine the overall scrubbing requirement. Figure 5-3 shows the relationship
between saturated gas temperatures and the overall removal efficiency of a
cyclonic spray tower scrubber (6). The gas streams leaving the scrubber are
saturated with water vapor. When the scrubber is operated at a relatively low
saturated temperature (gas temperature close to pond water temperature), the
efficiency is high. Since absorption decreases with temperature increase, efficiency
is lower at a higher saturated temperature. An additional advantage of scrubber
operation is that silica is kept in a gelatinous stage which is easily washed from the
scrubbing device. At higher temperatures, the silica is crystallized on the scrubber
and removed with great difficulty.
The scrubber effectiveness, or the number of transfer units, will determine the
overall scrubbing requirements.
Transfer units are defined by the following formula:
Inlet F
NTU = In
Outlet F + a
is the vapor pressure contribution of fluorine from the scrubbing media.
Once the overall transfer unit requirements are determined, the number of
scrubbing stages may then be set based upon the performance of each scrubbing
device employed.
The information included in Table 5-2 and Figures 5-1, 5-2, and 5-3 is typical
the information necessary for scrubber design.
-------
5.1.2 Particulate Emissions Control
Various types of wet scrubbers are used for removal of particulate matter*
emanating from phosphate fertilizer processes. The principle of particle collection
in wet scrubbers is based on bringing the particle close to the collecting bodies
(liquid droplets) whereupon a number of short range mechanisms accomplish the
actual collection. The basic unit mechanisms for aerosol collection are:
1. Inertial impaction
2. Interception
3. Diffusion
In some special circumstances, other mechanisms involving forces such as electro-
static, thermal, magnetic, etc., may play an important role. Fortunately, in most
cases one mechanism predominates so that the problem can be dealt with on a
theoretical basis.
Analysis of the principle of aerosol collection in wet scrubbers should start
with analysis of the flow of air around a spherical droplet. Throughout this analysis,
one should keep in mind that similar collection principles apply to dry filtration with
the exception that the collecting body is not a sphere but a cylinder (filter fiber). In
cases where a packed tower is used for aerosol collection, the collecting body can be
approximated by a flat plate. Aerosol collection can be best illustrated by the
example of the Venturi scrubber. The stream of gas containing particulate matter is
accelerated in the Venturi throat. The scrubbing liquid is injected into the throat
forming droplets which capture aerosol particles via unit mechanisms. The analysis
may be reduced to viewing two elements: one is stationary (liquid droplet) and the
other (aerosol particle) is immersed in gas moving toward the droplet. The gas
streamlines will follow the surface of the liquid droplet, while the aerosol particles,
because of their larger mass (compared with air molecules), will have sufficient
momentum to continue to move toward the liquid droplet, will break through the gas
streamlines, and become trapped in the droplet. The unit mechanism is referred to
as inertial impaction.
Three factors determine the efficiency of inertial impaction:
1. Velocity distribution of the gas flowing by the droplet which varies with
Reynolds Number.
2, Particle trajectory which depends on mass and size and shape of the
aerosol particle and the size of the droplets, and the rate of flow of the
gas stream.
3. Adhesion of particles to droplets (usually assumed to be 100%-this is
questioned by some researchers when the aerosol is hydrophobic).
*The terms "particulate matter" and "aerosol" are used interchangeably in this
discussion.
-------
Using the equation describing the motion of particles and assuming that the
particles obey Stokes1 Law, a parameter of inertia! impaction has been derived (7).
CPVQ °n
where:
\f/ - inertial impaction parameter (dimensionless)
c = Cunningham empirical correction factor (dimensionless)
p = particle density, gram mass per cc
V = initial (undisturbed) velocity of aerosol stream, cm
per second
D = diameter of aerosol particle, cm
D = diameter of liquid drop, cm
M = gas velocity, poise
The physical significance of the inertial impaction parameter is that it is the
ratio of the force necessary to stop a particle initially travelling at velocity Vo in
the distance Dc/2 to the fluid resistance at a relative particle velocity of Vo. It is
also a ratio of stopping distance to collector diameter (droplet) (8).
The importance of the other basic unit mechanisms on particulate removal in
the phosphate fertilizer industry is negligible and will not be discussed here. For a
detailed description of the theory and practice of wet scrubbing, refer to the
Scrubber Handbook (9).
The overall efficiency of aerosol collection in wet scrubbers is the function of
several parameters. Major influences on collection efficiency are:
1. Aerosol size: The larger the size, the higher the aerosol collection
efficiency. The size with a minimum collection is in a range from 0.1 -
0.8 microns.
2. Pressure drop: The higher the pressure drop, the higher the aerosol
collection efficiency as long as inertial impaction with interception is
the mechanism of collection (see Figure 5-4 (10)).
3. Liquid to gas ratio: There appears to be an optimum liquid to gas ratio.
The size, number, and distribution of liquid droplets generated in the
Venturi throat is usually calculated using the Nukiyama-Tanasawa
equation (11).
-------
10
Oi
£ 60
40
o
o
20
Ut = 200 FT/SE
J.
•T= O.b
LEGEND
L = LIQUID/GAS RATIO (GAL/1000 CU FT)"
Ut = THROAT VELOCITY (FT/SEC)
1 '
8 12 16 20
PRESSURE LOSS, INCHES OF WATER
24 28
Fiugre 5-4: Venturi Scrubber Performance (10)
-------
Other factors of importance include: wettability of the aerosol particles, ratio of
aerosol to drop sizes, and physical properties of the scrubbing liquid.
It has also been demonstrated that the efficiency of a Venturi scrubber can be
increased by lowering the surface tensions and viscosity of the scrubbing liquid (12).
In an article on equipment performance in controlling fluoride emissions,
Lund (5) stresses the difficulty of evaluating wet scrubbers for particulate removal:
The evaluation of liquid scrubbing devices with regard to the collection
of particulate fluorides is far more difficult than with regard to absorption of
gaseous fluorides. The predominant factor controlling the rate of collection of
particulate matter is effective particle size. No data have been published on
the particle size or size distribution of particulate fluoride emissions from
industrial processes (5).
Lapple and Kamack (13) have shown that within wide limits power consumption is a
key criterion in particulate emission control efficiency. The power can be
introduced either in a liquid or gas phase so that scrubber design specifics play a
minor role. This finding was also confirmed by other investigators, leading to
serious questions as to whether one scrubber configuration is really superior to
another at the same power consumption.
Depending on the size of the particulate fluorides, the power consumption for
efficient collection may be no more than that for the absorption of gaseous
fluorides, or it may be vastly greater. When a scrubber is used for simultaneous
removal of gaseous and particulate fluorides, the power requirements for the latter
frequently become the dominating factor.
Although there are a number of large control installations in the phosphate
fertilizer industry for the removal of gaseous fluoride and particulate, little
information has been published on equipment performance and power consumption.
TRC will recommend two studies to fill in this gap in information. The first
study is on the optimization of scrubbers through factorial design experiments; the
second is the evaluation of the potential for energy conservation in phosphate
fertilizer plants. The reasons, objectives and scope of work for these studies are
presented in Section 9.
The next section will briefly describe the scrubbers most frequently used in
the phosphate fertilizer industry.
5.2 Types of Scrubbers Used in Phosphate Fertilizer Plants
5.2.1 Cyclonic Spray Scrubber
Spray towers provide the contact necessary for gas absorption by dispersing
the scrubbing liquid in the gas phase in the form of a fine spray. Cyclonic spray
-46-
-------
towers eliminate excessive entrainment of scrubbing liquid by utilizing centrifugal
force to remove entrained droplets. Figure 5-5 is a schematic diagram of a typical
spray tower. A tangential inlet is used to impart the spinning motion to the gas
stream which flows parallel to the water sprays. Pressure drops across the scrubber
range from 2-8 inches of water. Although solids handling capacity is high, absorp-
tion capacity is limited to about two transfer units
5.2.2 Venturi Scrubbers
The Venturi scrubber provides a high degree of gas-liquid mixing but the
relatively short contact time limits its absorption capabilities, so it is primarily used
for particulate removal. Two types of Venturi scrubbers, gas-actuated and water-
actuated, shown in Figures 5-6 and 5-7 (15), are in general use. The necessary gas-
liquid contact is obtained from the velocity difference between the two phases and
turbulence in the Venturi throat. Both types also require the use of mist eliminators
for removal of entrained scrubbing liquid.
The gas-actuated Venturi uses the velocity of the gas stream as the energy
required for liquid atomization and consequent gas-liquid contact. The water-
actuated Venturi uses the velocity of the water stream as the source of energy. The
gas-actuated Venturi normally operates at a pressure drop of 8-20 inches of water
while the water-actuated develops up to 8 inches at high liquid flow rates (16). The
water-actuated Venturi is limited to gas flow rates up to approximately 5,000 CFM.
5.2.3 Spray-Crossflow Packed Scrubber (SCFS)
The spray-crossf low packed bed scrubber has been accepted as the most satis-
factory fluoride control device available for wet process phosphoric acid plants.
Recent improvements in spray-crossf low packed scrubber designs have alleviated
the initial problem of plugging and allow a greater solids handling capacity.
As shown in Figure 5-8, the spray-crossflow packed bed scrubber consists of
two sections, a spray chamber and a packed bed, separated by a series of irrigated
baffles. Both sections are equipped with gas inlets. Effluents with high fluoride and
particularly high SiF4 concentrations are treated in the spray chamber before
entering the packing. This reduces the danger of plugging in the bed, reduces the
loading on the packed stage, and provides some solids handling capacity. The cross-
flow design operates with the gas stream moving horizontally through the bed with
the scrubbing liquid flowing vertically through the packing. Solids deposited near
the front of the bed are washed off by a cleaning spray. Pressure losses through the
scrubber range from 1-8 inches of water, the average being about five inches.
Recycled gypsum pond water is normally used as the scrubbing liquid in both
the spray and packed sections.
-------
DIRTY
GAS
INLET
CLEAN GAS OUTLET
JET SPRAY BOX
DRAIN
Figure 5-5: Cyclonic Spray Tower Scrubber (14)
-------
WATER INLET
AIR INLET
-P-
v£>
WATER
INLET
VENTURI
CYCLONIC MIST
LIMINATION SECTION
WATER OUTLET
SPRAY NOZZLE
AIR OULET
AIR INLET
SEPARATOR BOX
WATER OUTLET
Figure 5-6:
Gas-actuated Venturi Scrubber With
Cyclonic Mist Eliminator (15)
Figure 5-7: Water-actuated Venturi (15)
-------
PRIMARY GAS INLET
POND WATER
to
•a:
C3
3
SECONDARY
GAS INLET
c=>
GAS FLOW
c=O
K%££f%%3WK&.
fl%«s3ftifefe
R^'^^H
?-#5f*..^':^i ,.'1
^i^yw^lf
'^i^vKHSM 3*?;-
PACKED'
BED
ifeiil
'
CLEAN GAS
c^>
TO STACK
Figure 5-8: Spray-Crossflow Packed Bed Scrubber (14)
-------
5.2.4 Coaxial Scrubber
The major objective of the coaxial scrubber and similar devices is to minimize
deposition of solids. An example of the coaxial design as manufactured by Teller
Environmental Systems, Inc. (TESI) is shown in Figure 5-9. This TESI scrubber is
used primarily for ammonia absorption with phosphoric acid.
The liquid flow rate is linear, requiring no turns. Thus, the liquid is not
subjected to mechanical shock that could result in deposition of solids. Liquid
distribution is primarily achieved by underflow weirs. The Venturi throat is of a
diamond design to minimize scale-up. The design of the coaxial scrubber is said to
be such that the definition of flow patterns is far more precise than that in cyclones
of the same diameter resulting in higher efficiency (17).
The next sections will present the state-of-the-art in air emission control
technology in the phosphate fertilizer industry. Special emphasis will be on the
equipment cost.
5.3 Wet Process Phosphoric Acid (WPPA)
The present emission guideline for wet process phosphoric acid plants limits
fluoride emissions to 0.01 grams of fluoride (as F") per kilogram of P2O5 input to the
process. This emission guideline is equal to the standard of performance for new
plants and requires removal of 99% of emitted fluoride. A spray-crossflow packed
bed scrubber installed in 53% of the WPPA plants (accounting for 74% of the total
production capacity) can achieve this requirement.
Figure 5-10 shows the emission streams and crossflow scrubber in a typical
WPPA plant. The scrubber consists of two sections—a spray chamber and a packed
bed—separated by a series of irrigated baffles. A typical unit treating the effluent
streams from a wet acid plant (20,000 SCFM) is 9 feet wide, 10 feet high and 30 feet
long (18). The packed bed is usually 3 or 4 feet in length, while the pressure drop
through the unit averages 4-6 inches of water. Recycled pond water is normally
used as the scrubbing liquid in both the spray and packed section. The ratio of
scrubbing liquid to gas ranges from 0.02 to 0.07 gpm/acfm.
Table 5-3 shows EPA data on crossflow scrubber performance in WPPA plants.
In 1973 the Industrial Gas Cleaning Institute (IGCI), under contract to EPA,
prepared the specifications for a scrubbing system for WPPA using pond water (20).
Table 5-4 shows the operating condition for a scrubber which was to be manu-
factured from PVC, rubber-lined steel, or FRP. Table 5-5 shows average capital and
annual cost obtained from several equipment vendors. It should be noted that these
costs represent 1973 figures. In order to bring the costs up to date one can use
economic indicators from Chemical Engineering (21). From the latest available CE
plant cost index (based on 1957-1959 = 100) for September 1978, an average
correction index of 179 can be derived. Consequently, today's projected total
-51-
-------
GAS INLET
IRRIGATING
LIQUID
OUTLET
SPIN SECTION
IRRIGATING LIQUID OUTLET
Figure 5-9: Teller Coaxial Scrubber (17)
-52-
-------
POND WATER
VENT FROM
HOTWELLS
VENT FROM
FILTER HOODS
VENT FROM
FILTRATE SEAL TANK
VENT FROM
DILUTE H2S04 TANK
VENT FROM
CLARIFIER TANKS
VENT FROM
ATTACK TANK
STACK
Figure 5-10: Spray-Crossflow Packed Bed Scrubber in Wet Process Phosphoric Acid Plant
-------
TABLE 5-3. SCRUBBER PERFORMANCE IN WET-PROCESS PHOSPHORIC
ACID PLANTS (19)
Fluoride Emissions3
Plant
A
B
C
D
Scrubber Design
Spray-cocurrent packed bed
Spray-crossf low packed bed
Spray-crossflow packed bed
Spray-crossf low packed bed
(Ib F/ton P20S)
0.015
0.006
0.002, 0.012b
0.011
Average of testing results.
Second series of tests.
-------
TABLE 5-4. OPERATING CONDITIONS FOR WET PROCESS PHOSPHORIC
"ACID PLANT CROSS-FLOW PACKED SCRUBBER (20)
SMALL LARGE
Plant Capacity, Ton/Day P2O3 500 900
Acid Strength, Wt. % P2O5:
From Digesters 30 30
From Evaporators 54 54
Fluorine Content, Wt. % F * 1.5 1.5
B D
Gas to Scrubbers:
Flow, SCFM 25,000 36,000
Flow, DSCFM 21,050 30,300
Flow, ACFM 28,300 40,600
Temp.,°F 140 140
Moisture, Vol. % 15.7 15.7
Fluorine, Ib/hr 80 115
Fluorine, ppm 1,050 1,050
Particulate, Ib/hr 10.8 15.5
Particulate, gr/SCF 0.05 0.05
Gas from Scrubbers:
Flow, SCFM 22,600 32,500
Flow, DSCFM 21,050 30,300
Flow, ACFM 23,700 34,400
Temp. °F 100 100
Moisture, Vol. % 6.7 6.7
Fluorine, Ib/hr 0.16 0.29
Fluorine, ppm 2.35 2.95
Fluorine Removal, Wt. % 99.80 99.75
Particulate, Ib/hr 0.07 1.4
Particulate, gr/SCF 0.05 0.005
Particulate Removal, Wt. % 91.1 91.1
Estimated y', ppm 1.4 1.4
Estimated NTU required 7.5 6.50
-55-
-------
TABLE 5-5. CAPITAL AND ANNUAL COST FOR CFPS ON WPPA PLANTS (19)
(1974 COSTS)
Plant Size TPD P2O5
500
900
Gas Flow SCFM
25,000
36,000
Capital Cost ($)
Scrubber
Auxiliary equipment (fan,
pumps, etc.)
Installation cost
Total Capital Cost
Annual Cost ($/yr)
Operating labor
Maintenance (5%)
Utilities
Depreciation (10 yr)
Interest (8%)
Property Tax, Ins. (2%)
Administrative (5%)
Total Annual Cost
17,700
8,500
36.300
62,500
2,000
3,100
2,800
6,250
5,000
1,250
3,100
23,500
21,600
9,400
40,900
71,900
2,000
3,600
4,400
7,200
5,750
1,450
3,600
-56-
-------
control cost for a 25,000 ACFM installation is $112,000 and for a 36,000 SCFM,
$129,000.
As mentioned previously, most plants today use spray-crossflow scrubbers and
can meet current regulations. The rest of the plants use spray scrubbers which will
have to be retrofitted with SCFS to be in compliance. Table 5-6 shows the cost
associated with retrofitting a WPPA plant that uses a spray tower with SCFS.
5.4 Superphosphoric Acid (SPA)
The present emission guideline for SPA plants limits fluoride emissions to
0.005 grams of fluoride (as F~) per kilogram of P2O5 input to the process. This
emission guideline is equal to the standard of performance for new plants (NSPS) and
would require approximately 50% removal. A spray-crossflow scrubber is capable of
providing this performance (16). Plants using the vacuum evaporation process (79%
of the SPA industry) will require no additional control.
Figure 5-11 shows a spray-crossflow scrubber system and emission sources in
an SPA plant. The description of the scrubber system is almost identical to that on
a WPPA plant and will not be repeated here.
The alternative controls for SPA process emissions are water-actuated Venturi
scrubbers and Venturi scrubbers with optional packed section. Since the water-
actuated Venturi is cocurrent and limited to one theoretical stage, it may be neces-
sary to add a packed section for additional gas absorption. An absorption efficiency
of 83% equivalent to 5 transfer units may be required to meet the standards of
performance. The energy requirements for the two types of scrubbers are approxi-
mately equivalent. Water flow rate for the water-induced Venturi is 100 gallons per
1,000 SCFM while for the conventional Venturi, 20 gallons per 1,000 SCFM is
required. The conventional Venturi, however, must overcome 10 inches of water
pressure drop in the fan (19).
IGCI scrubber specifications for 300 T/day SPA plants to operate with gypsum
pond water are shown in Table 5-7. The material of construction was limited to
PVC, rubber-lined steel and FRP. Responding to these specifications several
vendors submitted capital and annual control costs. Table 5-8 shows average costs
on both water-actuated and conventional Venturi scrubbers. To update the capital
costs to 1978 one should use an average correction factor of 1.79 which would give a
total capital cost for 1335, 2620, 1335 and 2620 ACFM of $16,500, $21,100, $42,900
and $51,200, respectively.
Two existing submerged combustion plants would have to be retrofitted with
SCFS to meet emission guidelines. Table 5-9 shows the cost of retrofitting (300
T/day PjOj) an SPA plant. The retrofit cost is based on replacement of the
impingement scrubber with SCFS.
-57-
-------
TABLE 5-6. RETROFIT COSTS FOR MODEL WPPA PLANT, CASE B
(500 TONS/DAY P2O5) NOVEMBER 1974 (16)
(25,000 SCFM SCRUBBER)
Cost ($)
A. Direct Items (installed)
1. Spray-cross flow packed bed scrubber 78,800
2. Ductwork 20,000
3- Piping 3,300
4. Pump and motor 5,300
5. Centrifugal fans and motors 16,000
6. Removal of old equipment 12,500
7. Stack 15,800
8. Performance test 4,000
Total Direct Items 155,700
B. Indirect Items
Engineering construction expense, fee, interest on
loans during construction, sales tax, freight insurance
(50% of A) 77,900
C. Contingency
(25% of A) 38)900
D. Total Capital Investment 272,500
E. Annualized Costs
1. Capital charges 4,
2. Maintenance 7 500
3. Operating labor 2*000
14. Utilities 9,300
5. Taxes, insurance, administrative 10*900
Total Annualized Costs 74
-58-
-------
I
tt
POND WATER FROM
SCRUBBER STORAGE
VENT GASES
FROM REACTORS
VENT GASES FROM
GRANULATOR
VENT GASES FROM
DRYER CYCLONE
VENT GASES FROM
EQUIPMENT CYLCONE
P205 TO REACTOR
REAGENT FROM STEP
REAGENT PUMP
PHOSPHORIC ACID
1
ACID
SCRUBBER PUMPl
WATER
SCRUBBER PUMP
STACK
TAIL GAS ^~?
SCRUBBER FAN
EFFLUENT
RETURN PUMP
Figure 5-11: Spray-Crossflow Scrubber System in Superphosphoric Acid Plant
-------
TABLE 5-7. OPERATING CONDITIONS FOR WET SCRUBBERS FOR
SPA PROCESS COMBINED VENTS SPECIFICATION (20)
Plant Capacity, ton/day 300
Acid Strength, Wt.% P2O5
to Evaporator 54
from Evaporator 72
Fluorine Content, Wt.% F
to Evaporator 1.5
from Evaporator 0.4
SCRUBBER Inlet Barometric Cooling Total To
Streams Source Condenser Hotwell Chamber Scrubber
Flow, ACFM 287 2,120 264 2,671
Temp, °F 147.5 100 100 105
Flow, SCFM 250 2,000 250 2,500
Moisture, Vol% 24 6.7 6.7 7.8
Flow, DSCFM 190 1,880 234 2,304
Fluorine, ppm 10 5 30 8.25
Fluorine, Ib/hr 0.008 0.030 0.024 0.062
Scrubber Outlet
Flow, DSCFM 2,304
Temp, °F IQQ
Moisture, Vol% 6.7
Flow, SCFM 2,470
Fluorine, ppm 1,455
Fluorine, Ib/hr 0.011
Scrubber Efficiency, % §2.5
Estimated y1, ppm 1.^
Estimated NTU requirement 5.0
-60-
-------
TABLE 5-8. CAPITAL AND ANNUAL CONTROL COSTS FOR
SPA PLANTS VACUUM EVAPORATION (13)
(197* COSTS)
Water Induced Venturi
Gas Flow Rate (ACFM) at 105°F 1,335
2,620
Conventional Venturi
1,335
2,620
Capital Costs ($)
Collector
Auxiliary equipment
(fan, pumps, ductwork,
instrumentation)
Installation Cost
TOTAL CAPITAL COST
Annual Cost ($/yr)
Operating Labor
Maintenance
Utilities
Depreciation (10 yr)
Interest (8%)
Property Tax, Ins. (2%)
Administration (5%)
TOTAL ANNUAL COST
6,400
8,600
300
2,500
9,200
200
520
880
920
740
180
460
600
2,600
11,800
200
620
1,430
1,180
940
240
590
6,300
7,700
4,200
13,600
24,100
1,230
1,310
990
2,410
1,930
480
1,200
4,700
16,200
28,600
1,230
1,340
1,770
2,860
2,290
570
1,430
3,900
5,200
9,600
11,500
-61-
-------
TABLE 5-9. RETROFIT COSTS FOR MODEL SPA PLANT (300 TONS/DAY P2O5)
NOVEMBER 1974 (16)
A. Direct Items (installed)
1. Spray-crossflow packed bed scrubber
2. Ductwork
3. Piping
4. Pump and motor
5. Removal of old equipment
6. Performance test
Total Direct Items
B. Indirect Items
Engineering construction expense, fee, interest on
loans during construction, sales tax, freight insurance
(50% of A)
C. Contingency
(25% of A)
D. Total Capital Investment
E. Annualized Costs
1. Capital charges
2. Maintenance
3. Operating labor
4. Utilities
5. Taxes, insurance, administrative
Total Annualized Costs
Cost ($)
37,500
5,000
1,900
4,200
12,500
4,000
64,300
32,600
16,300
114,000
18,600
3,000
2,000
700
MOO
28,700
-62-
-------
5.5 Diammonium Phosphate (DAP)
The present emission guideline, which is equal to NSPS, limits fluoride
emission from a DAP plant to 0.03 grams of fluoride (as F~) per kilogram of P^O^
This guideline requires 85% fluoride removal which could be achieved by SCFS added
to existing Venturis. Acceptable alternative scrubber combinations for a DAP plant
are shown in Figure 5-12 (20). The two-stage scrubber combinations may be termed
primary for particulate control and secondary for gaseous fluoride and ammonia
removal.
The sources of emission and a primary-secondary scrubber combination for a
well-controlled DAP plant are shown in Figure 5-13. The scrubbers in DAP plants
are designed for ammonia recovery, particulate collection, and fluoride removal.
The emission control in a DAP plant will be described in detail for three reasons:
1. DAP is a major phosphate fertilizer product and its importance is still
increasing.
2. The control system is relatively complex.
3. Control equipment has been plagued with plugging and poor efficiency
problems.
The ammonia recovery in Figure 5-13 is achieved in a coaxial scrubber
described in Section 5.2.4. A major problem encountered in ammonia scrubbing by
phosphoric acid is the stripping of fluoride. The degree of stripping depends on the
temperature, acid and sulfate concentration, and the temperature in the scrubber is
increased because of exothermic neutralization. The second problem resulting from
ammonia absorption is formation of submicron aerosols of ammonium fluoride and
ammonium bifluoride (17). These emissions are difficult to collect and the most
economical solution appears to be the use of nucleators. The nucleator design is
similar to that of a spray-crossflow scrubber but it uses four mechanisms of
particulate collection when conducted with hydrophilic materials:
1. Condensation of water on the particulates at or above the dew point.
2. Agglomeration of the particulates by inelastic Brownian interception.
3. Short path inertial impact on small targets.
k. Thermophoretic collection in short paths to small target surfaces.
A critical factor in the efficiency of the nucleation process is the provision for short
path inertial and thermophoretic capture. Commercial operating conditions allow
submicron ammonium fluoride removal at a pressure drop of 2 inches of water (17).
It is important to note that a conventional Venturi would require 20 to 50 inches of
water to achieve the same goal. The SCFS is designed for high mass transfer
efficiency with variable liquid flux so that the liquid can be concentrated in the
zone of maximum collection. For example, maximum deposition in the packed bed
will occur in the first foot of packing. Thus, the liquid irrigation rate can be
-63-
-------
GAS
REACTOR
GAS .
FROM ^**
30% ACID
t
REACTOR
VENT
SCRUBBER
t
RECYCLE TO
REACTOR
POND
WATER
t
GRANULATOR
DRIER
. SCRUBBER
t
RETURN
TO POND
}GAS T
ATMOSPH
j
TAIL GAS |
SCRUBBER
•CROSS-FLOW
'VENTURI
LUW—i
-CROSS-FLOW
-CROSS-FLOW
•CROSS-FLOW
TWO-STAGE CYCLONIC
OR
VENTURI
OR
CROSS-FLOW PACKED
_J
CROSS-FLOW PACKED
-TWO-STAGE CYCLONIC-
-VENTURI
-VENTURI
(TWO-STAGE CYCLONIC)
•I OR \
(CROSS-FLOW PACKED )
TWO-STAGE CYCLONIC
OR
VENTURI
OR
CROSS-FLOW PACKED
J
t
Figure 5-12: Acceptable Scrubber Combinations for
DAP Process Plants (20)
-------
ON
\J\
I
POND WATER
VENT GAS FROM
DAP REACTOR
VENT GAS FROM
AMMONIATOR-GRAN
PHOSPHORIC ACID
J
VENT GAS FROM
COOLER CYCLONE
VENT GAS FROM
EQUIPMENT CYLCON
VENT GAS FROM
DRYER CYCLONE
CO-AXIAL
SCRUBBER
CO-AXIAL
SCRUBBER
P20 TO
DAP REACTOR
REAGENT
H2S°4
,, 1
J
CO-AXIAL
SCRUBBER
1
r ^-^
CROSS FLOW
PACKED SCRUBBER
L
CROSS FLOW
PACKED SCRUBBER
1
(717
STACK
TAIL GAS
SCRUBBER FAN
EFFLUENT RETURN PUMP
POND WATER RETURN
SCRUBBER PUMP
Figure 5-13: Primary and Secondary Scrubbers in Dianmonium Phosphate Plant
-------
maintained at 20-30 GPM/ft2 in the first foot of packing, whereas it can be as low as
5 GPM/ft2 at the back portion of the packing (17).
Plugging of scrubbers in DAP plants used to be a serious problem until
recently. In a scrubber system similar to one shown in Figure 5-3 such a problem
was attributed to lack of sufficient attention and less than optimum system
performance and control. A rigorous program of repairs, modifications, mainten-
ance, and operation was instituted to solve the plugging problem. The primary
modifications were (22):
1. Increase of clearance of axial downcomer to the coaxial cyclonic shell.
Since this modification, there was no plugging in the primary scrubbers.
2. Separation of inlet headers on the primary scrubbers for cleanout ease.
3. Fresh water flush on the normal 8-10 days maintenance schedule to wash
out any buildup on the primaries.
4. Increased open area of the grating of the nucleators.
5. Increased open area of the distributors in the nucleators.
Upon implementation of the above modifications and alert operation and
maintenance, the plugging problem was resolved.
In many plants particulate carry-over was suspected or determined at the
outlet of SCFS. Kimre mist eliminators were proven effective in alleviating this
problem and a number of plants have retrofitted their scrubbers with them. The
application of Kimre mist eliminators represent a recent development and a study to
evaluate their efficiency will be outlined in a separate section (Section 9).
The IGCI prepared specifications for primary and secondary scrubbers in DAP
plants, as shown in Tables 5-10 and 5-11 (20). Two stage wet cyclonic and Venturi
scrubbers using pond water were specified as primary scrubbers, while crossflow
packed scrubbers were specified as secondary scrubbers. Because of relatively high
temperature PVC was ruled out as a material of construction, so rubber-lined steel
or FRP was to be used. Table 5-12 shows the corresponding capital and annual
control cost (19). Up-to-date costs can again be derived by using an average
correction index of 1.79.
EPA tested SCFS in development of standards for the DAP process. Scrubber
performance is shown in Table 5-13 (23).
The plant using the TVA process was considered for retrofit involving replace-
ment of the cyclonic spray tower on the reactor-granulator stream with an SCFS
and addition of SCFS as a tail gas unit to the dryer and cooler streams (16). Total
capital and annualized cost estimates are shown in Table 5-14 (16).
-66-
-------
TABLE 5-10. OPERATING CONDITIONS FOR PRIMARY SCRUBBERS FOR
DAP PROCESS DRYER AND COOLER VENTS (20)
Plant Capacity, Ton/Day DAP
Plant Capacity, Ton/Day P2O5
Process Weight, Ton/Hr
Gas to Scrubber:
Flow, DSCFM
Flow, SCFM
Flow, ACFM
Temp.,°F
Moisture, vol. %
Particulate, gr/SCF
Particulate, Lb/Hr
Ammonia, Lb/Ton P2O5
Ammonia, Lb/Hr
Ammonia, ppm
Gas from Scrubber:
Flow, DSCFM
Flow, SCFM
Flow, ACFM
Temp., °F
Mositure, vol. %
Particulate, gr/SCF
Particulate, Lb/Hr
Ammonia, Lb/Ton P2O5
Ammonia, Lb/Hr
Ammonia, ppm
Particulate Efficiency, wt. %
Ammonia Efficiency, %
1,000
500
42
30,000
46,100
55,700
180
35
0.5
200
8.0
167
1,350
30,000
46,800
53,500
164
36
0.01
4.0
0.08
1.67
13.4
98
99
1,600
800
66.7
48,000
74,000
89,000
180
35
0.5
320
8.0
267
1,350
48,000
75,000
85,500
164
36
0.01
6.4
0.08
2.67
13.4
98
99
-67-
-------
TABLE 5-11. OPERATING CONDITIONS FOR SECONDARY SCRUBBERS FOR
DAP PROCESS DRYER AND COOLER VENTS (20)
Plant Capacity, Ton/Day DAP
Plant Capacity, Ton/Day P2O5
Process Weight, Ton/Hr
Gas to Scrubber:
Flow, DSCFM
Flow, SCFM
Flow, ACFM
Temp., °F
Moisture, vol. %
Fluorine, ppm as F-
Fluorine, Lb/Hr
Gas from Scrubber:
Flow, DSCFM
Flow, SCFM
Flow, ACFM
Temp., °F
Mositure, vol. %
1,000
500
30,000
46,800
53,500
36
15
2.1
30,000
31,900
33,700
100
7
1,600
800
66.7
48,000
75,000
85,500
36
15
3.35
48,000
51,000
54,000
100
7
Medium Efficiency Case
Fluorine, ppm as F-
Fluorine, Lb/Hr
Fluorine removal, %
Estimated y', ppm
Estimated NTU
4.1
0.415
80
1.95
1.80
4.1
0.67
80
1.95
1.80
High Efficiency Case
Fluoride, ppm as F-
Fluoride, Lb/Hr
Fluorine removal, %
Estimated y1
Estimated NTU
3.25
0.31
85
1.95
2.8
3.25
0
85
2.8
-68-
-------
TABLE 5-12. CAPITAL AND ANNUAL CONTROL COSTS FOR
DIAMMONIUM PHOSPHATE PLANTS (19)
A.
B.
C.
Model Plant Size (P2O5)
Process Source
Reactor-Granulator
1. Capital ($)
2. Total Annual
Cost ($)
Drier
1. Capital ($)
2. Total Annual
Cost ($)
Cooler & Transfer Points
1. Capital ($)
2. Total Annual
Cost ($)
500
Primary
Collector
155,000(l)
62,500
162, 000^
82,000
118,000(2)
47,000
TPD
Secondary
Collector
60,000
23,000
80,000
30,800
57,000
22,100
800
Primary
Collector
210,000
87,400
225,000
120,000
164,000
83,000
TPD
Secondary
Collector
83,400
32,100
111,000
43,000
79,200
30,800
Summary
Total Capital ($) 632,000 873,000
Total Annual Cost ($) 277,400 396,300
Unit Control Cost ($/ton P2O5) 1.68<3> 1.50
Unit Control Cost with NH3
creditsW($) 0.88 0.70
-Stage cyclonic for both model plant sizes.
Venturi cyclone for both model plant sizes.
'3^330 days per year operation.
^ Value of NH3 taken at $40 per ton. Recovered NH3 is estimated at 80 Ibs
per ton NH3 feed for all sources combined.
-69-
-------
TABLE 5-13. SPRAY-CROSSFLOW PACKED BED SCRUBBER PERFORMANCE IN DIAMMONIUM PHOSPHATE
AND GRANULAR TRIPLE SUPERPHOSPHATE PLANTS (23)
Type Of
Facility
Sources Controlled
Primary Controls
Secondary Controls
Fluoride Emissions0
(Ib F/ton P205)
DAP
DAP
GTSP
GTSP
GTSP
storage
reactor, granulator
drier, and cooler
reactor, granulator,
drier, and cooler
reactor, granulator,
drier, and cooler
reactor, granulator,
drier, and cooler
storage building
3 venturi scrubbers
in parallel13
3 venturi scrubbers
in parallelb
3 venturi scrubbers
in parallel
process gases com-
bined and sent to 2
venturi scrubbers in
parallel followed by
a cyclonic scrubber
3 spray-crossf low
packed bed scrubbers
in parallel
3 spray-crossf low
packed bed scrubbers
in parallel
3 spray-crossf low
packed bed scrubbers
in parallel
spray-crossf low
packed bed scrubber
spray-crossf low
packed bed scrubber
0.03*, 0.029
0.039
0.18, 0.06C
0.21
0.00036C
Average of testing results.
Weak phosphoric acid scrubbing solution.
^
'Second series of tests.
Emission rate is in terms of pounds F per hour per ton of P2O5 in storage.
-------
TABLE 5-1 4. RETROFIT COSTS FOR MODEL DAP PLANT
_ (500 tons/day P2O3) November 1974 (16) _
Costs ($)
A. Direct Items (installed)
1. Spray-crossf low packed bed scrubbers (3) $ 285,000
2. Ductwork 16,700
3. Piping 26,200
4. Pumps and motors 41,500
5. Centrifugal fans and motors 33,000
6. Removal of old equipment 12,500
7. Performance test 4,000
Total Direct Items 418,900
B. Indirect Items
Engineering construction expense, fee, interest
on loans during construction, sales tax, freight
insurance (50% of A) 209,500
C. Contingency (25% of A) 104,700
D. Total Capital Investment 733,100
E. Annualized Costs
1. Capital charges 119,500
2. Maintenance 20,000
3. Operating labor
4.1 Utilities 21,200
5. Taxes, insurance, administrative 29,400
Total Annualized Costs 194,100
-71-
-------
5.6 Normal Superphosphate (NSP)
Sources of emissions at an NSP plant include the mixer, den, and curing
building. Emissions of fluoride and participate from the mixer and den are
controlled by a scrubbing system using recycled water. Figure 5-14 shows a
scrubbing system on an ROP-NSP plant. The emissions from the mixer are
controlled in a two stage centrifugal spray scrubber while the emissions from the
den undergo one stage scrubbing.
Since the process for production of ROP-NSP and emission control is almost
identical to that for an ROP-TSP facility, the control technology evaluation pre-
sented in the next section applies to ROP-NSP.
5.7 Triple Superphosphate (TSP)
The present emission guideline for fluoride emission from GTSP and ROP-TSP
(including storage) calls for 0.1 gram fluoride (as F") per kilogram of P7O5 which is
equal to the NSPS. This requires reduction of 99.6% of fluoride from the production
facility which can be achieved in a two stage system. Figure 5-15 shows a typical
scrubbing system at an ROP-TSP plant. This plant displays an efficient control
system since most emissions undergo three stage scrubbing. The third stage is a
spray chamber called a tail scrubber where combined emissions from upstream
scrubbed sources are treated prior to discharge through the stack. The first stage is
3. Venturi scrubber primarily used for particulate removal followed by a cyclonic
scrubber for fluoride scrubbing. Best Control Technology calls for SCFS as an
alternative for a second stage because it has higher absorption efficiency than the
cyclonic scrubber. Table 5-13 shows SCFS performance in a GTSP plant.
IGCI scrubber specifications for 250 and 400 T/day GTSP plants are shown in
Table 5-15. The material of construction was limited to PVC, rubber-lined steel and
FRP while pond water was to be used as a scrubbing liquid.
Table 5-16 shows the capital and annual costs obtained from the equipment
vendors. The scrubber capital costs, translated to 1978 figures, add up to $1,500,000
for 250 T/day and 2,080,000 for 400 T/day. Table 5-17 shows engineering specifica-
tions, capital and annual costs for ROP-TSP scrubbers. In 1978 dollars, capital
scrubber (installed) costs would be $412,000 for a 250 T/day plant and $589,000 for a
400 T/day plant.
In order to meet 99.6% fluoride emission efficiency, a typical retrofit would
involve the replacement of the spray tower with SCFS. About 40% of ROP-TSP
plants are directly affected by the emission guideline. The costs for retrofit are
more severe than for other sources (because of large gas volumes) but are expected
to be manageable. Table 5-18 shows the cost of replacing a spray tower with SCFS
in a conventional TVA process plant (ROP-TSP). The Dorr-Oliver process is
normally used for the production of GTSP. Existing control technology usually
consists of a Venturi scrubber and cyclonic spray tower. The cost of replacing a
spray tower with SCFS involving major retrofit is shown in Table 5-19.
-72-
-------
FROM ROP
MANUFACTURING
VjJ
ROP BLDG FUME
SCRUBBER
STAGE 1
ROP BLDG FUME
SCRUBBER
STAGE 2
et
I—
1/1
FROM ROP
DEN
ROP DEN FUME
SCRUBBER
ROP BLDG FUME
SCRUBBER FAN
ROP DEN FUME
SCRUBBER FAN
SEAL
WASTE TRENCH
— — '
Figure 5-14: Scrubbing System on Run-of-Pile Normal Superphosphate Plant
-------
FROM COOLER
CYCLONE
FROM DRYER
CYCLONE
FROM PHOSPHORIC
ACID SHIFT TANK
FUME DUCT
FROM
PRENEUTRALIZER
FROM DUST
CYCLONE
FROM GRANULATOR
FUME DUCT
SCRUBBER EFFLUENT TANK
SUMP
Figure 5-15: Scrubbing System on Run-of-Pile Triple Superphosphate Plant
-------
TABLE 5-15. ENGINEERING SPECIFICATIONS FOR ESTIMATING COSTS FOR
GRANULAR TRIPLE SUPERPHOSPHATE PLANTS (19)
Model Plant Size (P2O5) 250 TPD 400 TPD
Primary Secondary Primary Secondary
Collector Collector Collector Collector
A. Reactor- Granulator
Effluent Volume, DSCFM 20,000 20,000 32,000 32,000
Gas to Scrubber, ACFM 25,WO 22,700 40,600 36,300
Moisture Content, % 11 7 11 7
Gas Temperature, °F 140 100 140 100
B. Drier
Effluent Volume, DSCFM 40,000 64,000
Gas to Scrubber, ACFM 54,700 87,000
Moisture Content, % 12 12
Gas Temperature, °F 180 180
C. Cooler-Transfer Points
Effluent Volume, DSCFM 50,000 80,000
Gas to Scrubber, ACFM 58,000 93,000
Moisture Content, % 3 3
Gas Temperature, °F 140 140
D. Storage
Effluent Volume, DSCFM 98,400 157,400
Gas to Scrubber, ACFM 100,000 160,000
Moisture Content, % 2 2
Gas Temperature, °F 80 80
-75-
-------
TABLE 5-16. CAPITAL AND ANNUAL CONTROL COSTS FOR
GRANULAR TRIPLE SUPERPHOSPHATE PLANTS (19)
(1974 COSTS)
Model Plant Size (P2O5)
A. Reactor-Granulator
1. Capital
2. Total Annual Costs
250 TPD
Primary Secondary
Collector Collector
93,300(1) 43,700
44,300 17,000
400 TPD
Primary Secondary
Collector Collector
130,000 60,700
64,000 23,400
B. Drier
1. Capital
2. Total Annual Costs
C. Cooler & Transfer Points
1. Capital
2. Total Annual Costs
D. Storage
1. Capital
2. Total Annual Costs
179, 000 '
106,500
186,00(|2)
111,700
335,00cf3)
140,000
247,000
157,000
259,000
160,000
465,000
202,000
Summary
Total Capital ($)
Total Annual Cost ($/yr)
Unit Control Cost ($/ton P2O5)
837,000
420,000
5.09
1,162,000
606,000
3.67
^Venturi-cyclone on both model plant sizes.
"Venturi-packed type on both model plant sizes.
^Cyclonic scrubber on both models.
NOTE: Pond water scrubbing media for all control devices.
-76-
-------
TABLE 5-17. CAPITAL AND ANNUAL CONTROL COSTS FOR RUN-OF-PILE
TRIPLE SUPERPHOSPHATE PRODUCTION (19)
(1974 COSTS)
Model Plant Size (P2Q5) 230 TPD 400 TPD
A. Engineering Specifications;
C,as to Scrubber, ACFM 60,000 100,000
Collector Venturi Cyclone Venturi Cyclone
Scrubbing Media Pond Water Pond Water
B. Cost Summary
1. Capital Requirements ($) 230,000(l) 329,000(2)
2. Total Annual Cost ($) 102,000 153,000
3. Unit Control Cost-* ($ per
ton I'oO.) 1.24 l.lfe
'''Scrubber portion ol costs if $44,000, auxiliary and installation costs are S18b,000.
.Scrubber portion ot costs is $66,000; auxiliary and installation costs are $263,000.
-77-
-------
TABLE 5-18. RETROFIT COSTS FOR MODEL ROP-TSP PLANT, CASE A
(550 tons/day P2O5) November 1974 (16)
A.
li.
C.
D.
E.
Direct Items (installed)
1. Spray-crossflow packed bed scrubbers
2. Ductwork
3. Piping
4. Pumps and motors
5. Centrifugal fans and motors
6. Removal of old equipment
7. Stack
8. Performance test
Total Direct Items
Indirect Items
Engineering construction expense, fee, interest
on loans during construction, sales tax, freight
insurance (50% of A)
Contingency (25% of A)
Total Capital Investment
Annualized Costs
1. Capital charges
2. Maintenance
3. Operating labor
4. Utilities
5. Taxes, insurance, administrative
Costs ($)
$ 294,000
9,800
33,300
31,900
28,800
12,500
44,000
4,000
458,300
229,200
114,600
802,100
130,700
21,700
4,000
26 , 500
32,000
Total Annualized Costs 214,900
-78-
-------
TABLE 5-19. RETROFIT COSTS FOR MODEL GTSP PLANT
(WO tons/day P2O5) November 1974 (16)
Costs
A. Direct Items (installed)
1. GTSP Production
a. Spray-crossflow packed bed scrubbers (3) $ 261,000
v b. Ductwork 22,800
c. Piping 26,200
d. Pumps and motors 35,900
e. Removal of old equipment 18,000
f. Performance test 4,000
g. Centrifugal fans and motors 14,400
2. GTSP Storage
a. Crossflow packed scrubber 150,000
b. Ductwork 56,600
c. Piping 27,800
d. Pumps and motors 19,400
e. Centrifugal fan and motor 23,000
f. Structural steel supports/building 50,000
g. Sealing of storage building 10,000
h. Performance test 4,000
Total Direct Items 723,100
B. Indirect Items
Engineering construction expense, fee, interest
on loans during construction, sales tax, freight
insurance (50% of A) 361,600
C. Contingency (25% of A) 180,800
D. Total Capital Investment 1,265,500
E. Annualized Costs
1. Capital charges 206,300
2. Maintenance 33,800
3. Operating labor 6,000
4. Utilities 40,600
5. Taxes, insurance, administrative 50,500
Total Annualized Costs 337,200
-79-
-------
Uncontrolled emissions from ROP-TSP and GTSP storage emissions present
major fluoride emissions from stationary sources in the phosphate fertilizer indus-
try. Adoption of the emission guidelines would result in a 50% emission reduction
for GTSP and 90% for ROP-TSP (16). The present emission guideline for a GTSP
storage facility limits fluoride emissions to 2.5 x 10~4 grams per hour per kilogram
of PjOs in storage.
These limitations can be easily reached by applying 5CFS on the storage
facility. Recycled gypsum pond water can be used as a scrubbing liquid.
5.8 References
1. Billings, C.E., Fabric Filter Manual, The Mcllvaine Co., Northbrook,
Illinois, 1975.
2. Boscak, V., Tendon, J., Odor Abatement in Animal Food Manufacturing
Plants, Proceedings of the First Conference on Energy and Environment,
College Corner, Ohio, 1973.
3. Colburn, A.P., Trans. AICHE, 35, 211, 1939.
4. McCabe, W.L., Smith, J.C., Unit Operations of Chemical Engineering, p.
655, McGraw-Hill Book Co., 1959.
5. Lunde, K.E., Performance of Equipment for Control of Fluoride
Emissions, p. 293-298, Ind. Eng. Chem., Vol. 50, No. 3, March 1958.
6. The Mcllvaine Scrubber Manual, The Mcllvaine Co., Northbrook, Illinois,
1974.
7. Ranz, W.E., Wong, 3.B., Impaction of Dust and Smoke Particles, Ind.
Eng. Chem., Vol. M, No. 6, p. 1371, 1952.
8. Johnstone, H.E., et al, Gas Absorption and Aerosol Collection in a
Venturi Atomizer, Ind. Eng. Chem., Vol. 46, No. 8, p. 1601, 1954.
9. Calvert, S.3., et al, Scrubber Handbook, EPA-R2-72-118a, PB 213-061,
1972.
10. Ekman, P.O., 3ohnstone, H.F., Collection of Aerosols in a Venturi Scrub-
ber, Ind. Eng. Chem., Vol. 43, No. 6, p. 1358, 1951.
11. Nukiyama, S., Tanasawa, Y., Trans. Soc. Mech. Engrs (Japan) Vol. 5, No.
18, p. 68, 1939.
12. Boscak, V., Ph.D. Thesis, Mathematical Modeling and Digital Simulation
of Venturi Scrubber, University of Belgrade, Yugoslavia, 1972.
13. Lapple, C.E., Kamack, H.3., Chem. Eng. Progress, 5J_, 110-121, 1955.
-80-
-------
14. Inspection Manual for Enforcement of New Source Performance Stan-
dards: Phosphate Fertilizer Plants, EPA 340/177-009, U.S. Environmental
Protection Agency, Washington, D.C., January 1977, 80 pp.
15. Chatfield, H.E., Ingels, R.M., Gas Absorption Equipment. In: Air Pollu-
tion Engineering Manual, Danielson, J.A. (Ed). U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, 1973.
16. Final Guideline Document: Control of Fluorides Emissions from Existing
Phosphate Fertilizer Plants, EPA 450/2-77-005, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, March 1977.
17. Teller, A.3., Scrubbers in the Fertilizer Industry; Their Success, Near
Future and Eventual Replacement, Presented at Fertilizer Round Table,
Washington, D.C., November 1973.
18. Atmospheric Emissions from Wet-Process Phosphoric Acid Manufacture,
National Air Pollution Control Administration, Raleigh, North Carolina,
Publication Number AP-52. April 1970, p. 25-26.
19. Technical Report: Phosphate Fertilizer Industry. In: An Investigation
of the Best Systems of Emission Reduction for Six Phosphate Fertilizer
Processes. U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, April 1974, p. 25.
20. Air Pollution Control Technology and Costs in Seven Selected Areas,
EPA-450/3-73-010, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, December 1973.
21. Economic Indicators, C.E. Plant Cost Index, Chem. Engineering, January
15, 1979.
22. Teller, A.J., Lombardi, C.E., Scrubbing Fluoride Emissions. Presented at
Fertilizer Institute Environmental Symposium, New Orleans, Louisiana,
January 14-16, 1976.
23. Technical Report: Phosphate Fertilizer Industry. In: Group HI Back-
ground Document. U.S. Environmental Protection Agency, Research
Triangle Park.
-81-
-------
6.0 GYPSUM POND EMISSIONS AND CONTROLS
6.1 Location, Description and Role of the Gypsum Pond
The gypsum pond is an integral part of the wastewater treatment scheme at a
phosphate fertilizer plant and it serves two functions as both a settling and storage
area for waste by-product gypsum and fluorine compounds, and as an area for
atmospheric evaporative cooling of the pond water prior to its return to process
units. The pond system operates in a closed loop mode and releases water for
treatment only during heavy rainfall. The size of the pond at a WPPA plant is
approximately 2.23 x 10~3 Km2/metric ton P2O5/day (1). The surface area ranges
from 0.26 to 2.71 Km2.
The majority of gypsum ponds in the United States are located in the Polk-
Hillsborough County area of Florida. Figure 6-1 shows the location of several
gypsum ponds in that area.
There are three major considerations of environmental significance regarding
gypsum ponds (2):
1. Civil engineering design to ensure mechanical integrity.
2. Process engineering consideration of water management and recycle
uses.
3. Chemical processes that control fluorine behavior.
The chemistry of fluorine behavior is fundamental to virtually all aspects of
environmental control and will be discussed next. Figure 6-2 shows the inter-
relationship of the storage pond system to the total phosphate manufacturing
process (2). Fluoride transport to the gypsum pond system follows two basic
pathways: Fluorides delivered from the acidulation process as by-product wastes
consigned to the gypsum storage section and fluoride recovered in recycled process
water uses and returned to the pond water storage section where effluents are
mixed.
General operating parameters which affect fluoride loading in the gypsum
pond for all phosphoric acid and phosphate fertilizer unit processes are as
follows (1):
a. Product specification
b. Raw material specification
c. Temperature of the reactions
d. Acid concentration
e. Fluoride emissions standards
f. Fluoride recovery (if any).
-82-
-------
S PORT TAMPA
PASCO
HILLSBOROUGH COUNTY
PLANT CITY
**#'
POLK COUNTY
LAKELAND
MULBERRYjV
$ TLE&
LAKE ALFRED
WINTER HAVEN
LAKE HANCOCK
FORT MEADE
PEACE1 RIVER
MANATEE COUNTY ' HARDEE COUNTY
^APPROXIMATE LOACTION OF GYPSUM PONDS
Figure 6-1: Approximate Locations of Gypsum Ponds in Central Florida
-------
00
•P-
TGS/BC*
I i
WPA
REACTOR
1
T
J_
FILTRATION
WASHING
CAKE/PW
SLURRY
I
WPA
EVAPORATOR
GYPSUM POND
COOLING POND
LJ
SEEPAGE DITCH *TGS: SCRUBBERS
' BC: BAROMETRIC
Figure 6-2: Fluorine Distribution: Interrelation of Gypsum CONDENSER
Pond System to the Total Process (2)
-------
A major operational parameter in phosphoric acid production is control of the
crystalline form of calcium sulfate in the acidulation/f iltration process.
6.2 Gypsum Pond Chemistry
One important role of the gypsum pond is to remove fluorides from the ferti-
lizer manufacturing process. Table 6-1 indicates that over 50% of the soluble
fluorides generated in the manufacture of phosphate fertilizers "disappears" in the
pond (4). The major concern in studying gypsum pond chemistry will be establishing
what fluoride species are likely to escape to the atmosphere and in what form
fluoride is fixated in the pond.
Gypsum pond water is a complex mixture of various chemical o species.
Table 6-2 (3) shows some of the major cations and anions present. Fluorine,
designated as F~ in Table 6-1, will occur as the fluoride ion F~, as H2SiF6, HF (aq),
SiF«, HF2~ and complexed with metals such as FeF3, AJFe++, A1F2+, AF3, AF4~, etc.
Since one would like to immobilize as much fluorine as possible in the pond,
the first important question is what modes of phosphate precipitation are possible.
Table 6-3 shows the composition of 12 fluoride precipitates. A laboratory study was
carried out to determine the impurity factors that favor the formation of chukhro-
vite (Ca4AlSiSO4F,3 • 10 H20) over the usual alkali fluorosilicates and possibly other
major fluoride compositions such as ralstonite (NaxMgxAL (F,OH)6 • H20) and
cryolite (Na2SiF6). The major effects are due to aluminum "and magnesium. High
alumina concentrations favor chukhrovite, while a high magnesium content supresses
fluoride precipitation. Increasing the sodium content favors the formation oi
Na25iF6 as a competing precipitation process with chukhrovite. Laboratory test
results indicate that the preferential precipitation of chukhrovite during the rock
digestion stage can remove as much as 85% of the input fluorine. The fluorine can
be immobilized in the by-product cake in a salt that will minimize fluorine
redistribution in gypsum ponds. This is an effective way of controlling fluoride
emissions but it is not fully understood. Because of decreasing phosphate rock
purity, further study is recommended and will be described as a proposed research
project in Section 9.
The transport of fluoride from pond water to the vapor phase is obviously one
of the principal environmental concerns. The chemical equilibrium governing
fluoride emissions involves the complicated interactions of HF, HF2", H2SiF6, SiF4,
and SiO2. These reactions are as follows:
H2SiF6(aq)^H±2HF(aq) + SiF<(aq)
3SiF4(aq) + 2HaO<—^ 2H2SiF6(aq) + SiO2(s)
Si02(s) + 6HF(aq)^=±:H2SiF6(aq) + 2H2O
HF(aq) + F"< *" HF2"
-85-
-------
TABLE 6-1. FLUORINE INVENTORY 700 TPD P2O5 PRODUCTION
FLORIDA ROCK (4)
LB/DAY
INPUT
Fluoride in rock 170,000
OUTPUT
Solid Fluorides
In Gypsum 46,000
In Fertilizer Product [7,000
Total Solid Fluorides 63,000
Soluble Fluorides
Lost to atmosphere from pond 50
Lost to atmosphere from stacks 40
Lost to pond discharge and seepage 1^,000
"Disappeared" 92,910
Total Soluble Fluorides 107,000
-86-
-------
TABLE 6-2. MA3OR CATION AND ANION CONCENTRATIONS IN GYPSUM
POND WATER (3)
CATIONS
Ca++
Na+
AJ3+
Fe3+
Mg++
H+
CONCENTRATION (mg/1)
2000
1600
500
300
240
200
pH = 1.4
CONCENTRATION (M)
0.05
0.07
0.018
0.005
0.01
0.005
0.04
ANIONS
F"
50 =
Cl-
H2PO4~
8000
4800
200
1000
0.42
0.05
0.006
0.02
-87-
-------
TABLE 6-3. POSSIBLE MODES OF F PRECIPITATION IN
F.G. WAP SOLUTION (2)
FLUOSILICA'TES FLUOALUMINATES
Na2SiF6 N
NaKSiF6 N
CaSiF6.2H2O (MgxNa
MgSiF6.6H2O
SIMPLE FLUORIDES COMPLEX FLUORIDES
CaF2
M
F DISTRIBUTED TO CAKE (% OF INPUT)
Na2SiF6 20-30%
(NaMg)xAl2_xF6.2H2) _ 70%
Ca^AlSiSO-.F-a-lOHoO 80-85%
-------
3HF2" + SiO2(s)
HF(aq) ^=^H+ + F
HF(aq)^==HF(v)
The above reactions are overwhelmingly complex and can be considered only
qualitatively. The first reaction equilibrium ^appears to be rate controlling but
others (e.g., SiO2) can play significant roles.
Figure 6-3 shows the reaction pathways considered to be the most significant
in a gypsum pond. This simplified model meets the criteria that all chemical species
entering the pond must either precipitate into the pond sediments or volatilize, and
that waters in the pond are near saturation with respect to many chemical species
due to continuous recycling.
The fluoride species involved and the quantities in which they are emitted to
the atmosphere will be discussed next.
6.3 Gypsum Pond Air Emissions
Three studies have been conducted in an attempt to determine the amounts of
fluoride emitted from gypsum ponds. The first study was performed in 1967 (5). It
consisted of measuring fluoride concentrations emitted from a small A-frame
structure placed on a 0.64 Km2 gypsum pond. The intent of this study was to
practically measure an emission rate. Unfortunately, due to the fundamental errors
caused by using arbitrary areas and flow rate, the results are meaningless (1). A
major study was conducted in 1970 (6). This study consisted of two phases. In the
first phase, saturation vapor pressure of fluorides over a liquid solution of
hydrofluoric acid (HF) and fluosilicic acid (H2SiF6) was measured. The results
showed that a linear relationship exists between fluoride concentrations and fluoride
vapor pressures above the solution. The second phase was concerned with the
determination of a fluoride emission factor and involved experiments in a wind
tunnel. Figure 6-4 shows the study's summary of results on process water.
What is probably the most complete study on fluoride emissions from gypsum
ponds was conducted in 1974 (7). The approach to the development of an emission
factor was:
1. Development of a correlation for predicting the mass transfer coeffi-
cient (K.) from existing data.
.£
2. Measurement of the equilibrium vapor pressure (P,) of fluorides over
samples of pond water.
3. Prediction of fluoride mass transfer rates by the equation:
-89-
-------
ATMOSPHERE
SOLUBLE Fe & Al
COMPLEXES
Figure 6-3: Major Gypsum Pond Equilibrium
-90-
-------
E = 0.0306V
E • 0.0103V
E = 0.00816V
Process Water 9 75°F
Process Water @ 85°F
4- Process Water @ 95°F
1 1
300
V - WIND VELOCITY (FPM)
400
600
Figure 6-4: Emission Factors for Process Water at 75°, 85°, 95°F (6)
-------
where
N, = Fluoride transfer rate per unit surface area from
pond to atmosphere VR-moles f)
(hr-M2)
K, = Overall gas-side fluoride mass transfer coef-
ficient (g-moles F )
(hr-M2-mm Hg)
*
P{ = Partial pressure of fluoride at the gas liquid in-
terface in equilibrium with pond water (mm Hg)
P. = Partial pressure of fluorides in the atmosphere
above pond.
Figure 6-5 shows fluoride emission factors developed as a function of temperature
and wind speed. A simulation model for fluoride concentration prediction was also
developed. The model was verified through field sampling and analysis, but the
correlation between measured and calculated ambient air concentrations is some-
what questionable (1).
Comparisons of emission factors developed in these studies are shown in Table
6-4. These results should be interpreted with caution because a judgment cannot be
made as to the validity of either method of estimation when applied to a particular
pond. The best statement that can be made at this point is that the characteristic
emission factor appears to lie in the range of 0.1 to lOlb/acre-day (1).
In August 1977, TRC - THE RESEARCH CORPORATION of New England and
EPA jointly carried out a field program at the CF Industries-Bartow Phosphate
Complex near Bartow, Florida. The purposes of the program were to evaluate the
ROSE (Remote Optical Sensing of Emission) system for the measurement of fluoride
concentrations above and at the edge of the gypsum pond and to estimate the
fluoride emission rate.
During the program wet sampling/analysis was employed to determine the
concentration of total fluoride concentrations upwind and downwind of the gypsum
pond. Vertical fluoride concentrations up to 25 meters were obtained through the
elevation of sampling probes with a crane. Together with the wet sampling
(conducted by TRC), a remote sensing measurement (conducted by EPA) was per-
formed using the ROSE system consisting of a long-path Fourier-transform IR
spectrometer system. A meteorological monitoring station was erected in the
middle of the gypsum pond area to collect meteorological data during the field
program.
-92-
-------
I
o
CO
_1
UJ
I—
QL
Z
o
Q
h—I
O
0.5
70 80 90 100
AVERAGE POND TEMPERATURE (°F)
no
Figure 6-5: Fluorine Emission Rates for Ponds with
Water Containing 0.628 g moles/liter
Fluorides V16 = Wind Speed at 16 Meters
in Meters per Second (7)
-93-
-------
TABLE 6-4. COMPARISON OF PREDICTED EMISSION FACTORS
AT VARIOUS TEMPERATURES (1)
Surface Wind Speed - M/Sec
Pond Temperature - °F 0.23 0.54 1.2
75
Process Water Study 0.41 0.86 1.9
Gypsum Ponds Study 0.96 1.6 2.9
85
Process Water Study 0.52 1.1 2.4
Gypsum Ponds Study 0.75 1.3 2.2
95
Process Water Study 1.50 3.2 7.3
Gypsum Ponds Study 0.92 1.6 2.8
-94-
-------
The data collected by the remote sensing method indicated that fluoride
emission from the gypsum ponds consisted almost entirely of HF; the SiF4 concen-
tration was below the detectable threshold of about 0.5 ppb.
The fluoride concentrations obtained by the two methods using different
principles were in the same order of magnitude. Downwind concentrations during
most of the runs were in tens of ppb while the upwind values were from zero to 10
ppb. Vertical traverses indicated that the maximum fluoride concentration was at
15 meters (probably due to local topography) and that the concentration ceiling was
at about 25 meters during the tests. Average fluoride emission rates from the
gypsum pond were estimated to be in the range of 0.2 to 713 Ib/acre-day during the
testing period. Measurement of ground level fluoride concentration downwind from
the ponds at the W.R. Grace plant showed somewhat lower values than at CF
Industries, probably due to lower fluoride concentration in the pond water. The
ROSE system proved to be a powerful tool for the measurement of fluoride
concentrations at the gypsum pond and the methodology is applicable to similar
measurement problems. The system gives essentially real time results requiring no
sample handling, which makes it ideally suited for the measurement of fugitive
emissions. An additional advantage over standard wet sampling/analysis in this
application is that it can distinguish between HF and SiF4 and can measure the
concentration of both species.
6.4 Seepage from the Gypsum Pond
In addition to fluoride emissions into the atmosphere, leaching presents
another important environmental problem.
Increased or decreased control of gaseous water-soluble fluorides will not
change the amount of liquid waste generated by the fertilizer industry. Most
present control systems use pond water for scrubbing, thereby eliminating the
creation of additional effluent. The pond water is reused and a discharge is needed
only when there is a rainfall in excess of evaporation (8). Figure 6-6 shows a typical
gypsum pond system. Natural soil from the surrounding area provides the base for
dikes. Gypsum is used to increase the height of the dike. A drainage ditch
surrounds the perimeter of the area to control contaminated water seepage through
earth and gypsum.
Although the magnitude of subterranean seepage and the potential migration
of dissolved fluoride, phosphate and uranium into aquifers is a major environmental
concern, virtually no published accounts are available that quantify these distribu-
tion patterns. This constitutes, therefore, one of the most poorly resolved facets of
fluoride distribution in storage ponds.
Percolation rates through compacted aged gypsum cake and the underlying
clayed sediments have been reported to be exceedingly low (9). It was concluded
that contamination of deep aquifers is not occurring. On the other hand, some local
contamination of the water table aquifer is occurring at some sites. This takes
place in the immediate vicinity of the gypsum stacks and ponds. The seepage
-95-
-------
FRESH H20
.-
^j^^^^i^^^^^^^
Figure 6-6:
The Storage Pond System: Interrelation of Fluorine
Distribution and Water Management (2)
-------
through relatively pervious surface sands frequently reaches open bodies of water
such as creeks, rivers and lakes, but this is a seepage problem that is usually
relatively easy to contain and control. One potentially serious problem regarding
the seepage is that gypsum pond water is highly radioactive. A concentration of 60-
100 picocuries per liter (p Ci/1) has been measured which exceeds by at least two to
three times the Atomic Energy Commission (AEC) standards for discharge to an
unrestricted environment (10).
The State of Florida Department of Environmental Regulation is currently
conducting a study at three phosphate fertilizer plants to determine distribution of
contaminants around the gypsum pond. Soil samples will be taken through holes
located along a line normal to the stack and located at distances of approximately
75, 100, 250, 500 and 1,000 feet from the zero reference point at the gypsum stack-
cooling pond-seepage ditch system. The samples will be analyzed for fluoride, total
phosphorus and radium 226. Ground water samples will be collected in piezometers
and will also be analyzed for fluoride, phosphorus, radium 226 and pH.
6.5 Present Control of Gypsum Pond Emissions
EPA effluent limitation guidelines require that any gypsum pond water dis-
charged to navigable water when rainfall exceeds evaporation meet the limitation
shown in Table 6-5. These limitations were maintained as best conventional pollu-
tant control technology (11). Beginning July 1, 1977 and effective when each plant's
wastewater discharge permits are subject to renewal, discharge of process waste-
water pollutants to navigable waters is allowed only under certain conditions (8).
A two stage liming combined with settling is sufficient control to meet regula-
tions shown in Table 6-5. A flowsheet for a two-stage lime treatment plant is shown
in Figure 6-7. At least two stages of liming are required; the first treatment raises
the pH from less than 2 to about 3.5 -
-------
TABLE 6-5. EPA EFFLUENT LIMITATIONS GUIDELINES FOR
GYPSUM POND WATER (8)
Maximum Average Of Daily
Values For Periods Of
Aqueous Maximum Daily Discharge Covering 30
Waste Concentration Consecutive Days
Constituent (mg/1) (mg/1)
Phosphorous as (P) 105 35
Fluoride as (F) 75 25
Total suspended
nonfilterable solids 150 50
-98-
-------
LIME HOPPER
CAR
v£>
I
DUST COLLECTOR
a—-
,HOTT
WATER
TANK
•L.P. STEAM
FEEDER
I
MILK OF
LIME
STORAGE
•—O
PhC
POND
WATER
•—Ophc
THICKENER /, ** TO GYPSUM POND
:ALCIUM PHOSPHATE
POND
t
TO RIVER OR
PROCESS UNITS
Figure 6-7: Flowsheet, "Double Liming" Treatment of Gypsum Pond Water (12)
-------
.20 -
15 -
10 -
.05 -
Figure 6-8:
Species Predominance Diagram for
0.4 M HF Solution (1)
-100-
-------
MODEL PHASE SYSTEMS PREDICTED SOLIDS
HF-CaO-H20 CaF2
CaO-P205-H20 DCP, OCP, APATITE
HF-CaO-P205-H20 F-APATITE
OBSERVED SOLIDS
DCP (OCP) _ OCP
O O ^
\ DEFECT APATITE
Ca.(AlFc)(SiFc)SO.(F, OH)-10H00
4 o b 4 c.
pH RANGE
Figure 6-9: Chemical Models of F and P?0r Removal by Lime (2)
L. \J
-101-
-------
The main disadvantage of the liming operation for continual use is the high
cost involved. Because the buffering capacity of the gypsum pond water is high at
pH 1.0 to pH 3.0, large amounts of lime are required to raise the pH initially to 3.0
relative to the amount required to raise the pH from 3.0 to 6.0 (1). An additional
disadvantage is the deposition of calcium fluoride on the lime particles, rendering
them chemically inactive. The use of high intensity agitators is required to prevent
this. It is also important to note that radium 226 is also precipitated by lime
treatment with increasing pH.
6.6 Identification of Control Techniques For Gypsum Ponds
As seen in previous sections, the gypsum pond is plagued with two emissions
that are presently uncontrolled: fluoride emissions into the atmosphere and
contaminant seeping into the water table aquifer. The obvious answer to preventing
these emissions is not to use ponds for fluoride removal and water cooling. This
approach is unfortunately prohibitively expensive and can be used only in isolated
cases, such as the Mississippi chemical plant in Pascagoula.
The purpose of this section is to review the available control processes for
reducing fluoride emissions from gypsum ponds. The following control options will
be briefly considered:
1. Two pond system
2. Kidde process
3. Swift process
4. Liming of cooling ponds
5. Dry conveyance of gypsum to stacks
6. Phosphate rock calcining
7. Change to hemi/dihydrate process
8. Use of algae
9. TESI dry system.
6.6.1 Two Pond System
The reasons for separate gypsum/cooling ponds given in Reference (1) are:
• The required size of the gypsum slurry pond is small (about 5 acres) since
no area is required for cooling. This water would be the most
contaminated and acidic water in the plant due to the presence of P2O5,
H2SQ<, iron and aluminum complexes, and fluorides from the filtration
operation.
• The size of the pond required for the barometric condensers is deter-
mined by the cooling requirements. This area is estimated to be 0.1
acre/TPD P2O5.
-102-
-------
Figure 6-10 shows the flow diagram for a two pond system (1). One pond
would service the filtration operation by receiving gypsum slurry. The other pond
would receive waters from the fume scrubbers, flash coolers and the evaporation
trains. The cooling ponds would require about 200 acres for a 1,000 TPD P^Os plant
while the gypsum pond would require approximately 5 acres. The costs required to
segregate the two ponds are considered negligible compared to other control costs.
Pond segregation is considered tantamount in applying control options 2 through 5.
6.6.2 Kidde Process
The Kidde process is used to defluorinate wet-process phosphoric acid and
convert the extracted fluorine into various fluorine compounds. Fluorine compounds
that may be produced include aluminum fluoride, sodium aluminum flouride (cryo-
lite), hydrogen fluoride and calcium fluoride (fluorspar).
The basic reaction in the Kidde process occurs between reactive silica and
hydrogen fluoride producing silica tetrafluoride. SiF< is then condensed and reacted
with bifluoride. Barometric condenser and scrubber water is combined and neutral-
ized with ammonia, and the product, ammonium silicofluoride, is then stored for
shipment or reacted further to produce additional products, with aluminum fluoride
being the principal salable byproduct. The approximate fluoride reduction in the
Kidde process is about 95%. The major benefits of the process are (1):
• Reduction or elimination of liming for pond water discharged from the
plant's cooling ponds.
• Improvement in the physical properties of the phosphoric acid due to a
reduction in its fluorine content.
• Increased recovery of fluoride otherwise lost in the evaporation stage.
An economic evaluation indicates that a Kidde process capable of servicing a
1,000 TPD P2O5 facility would cost approximately 9.3 million dollars (1). The total
investment is of the same order of magnitude as that required for the P2O5 complex
itself. No full-scale plant actually exists today or is under construction.
6.6.3 Swift Process
The Swift process involves the removal of fluorine compounds from wet
process phosphoric acid manufacturing in a manner similar to the Kidde process.
The vapors from three evaporation units used for concentration of phosphoric
acid are scrubbed by an absorber which uses cooling water and removes 90% of the
fluorides. The resulting diluted H2SiF6 is removed and concentrated in the absorbers
in a countercurrent manner resulting in 25% H2SiP6 as a final product.
-103-
-------
o
-p-
I
FILTER HASH WATER
CAKE
SLUICE
WATER
GYPSUM
SLURRY
f
-—j
\ r\
GYPSUM
nri i-
1200 ACRES)
EVAPORATORS
BAROMETRIC
CONDENSER
FLASH
COOLERS
1
FUME
SCRUBBER
PUMPS
o
o
o
PUMPS
GYPSUM
POND
'3 TO 4 ACRES)
OVERFLOW
COOLING
POND
(200 ACRES)
Figure 6-10: Two Pond System for Phosphoric Acid Plant (1)
-------
The original purpose of the Swift process was for the economical recovery of
byproduct fluorides, but when coupled with a two pond system it is now recom-
mended as a pollution control system. The cost of retrofit to an existing plant is
$500,000 for the equipment servicing each evaporator or $1,500,000 total capital
investment for three modules (1). If all fluosilicic acid produced can be sold, which
is doubtful, the Swift process can operate at a profit.
6.6.^ Liming of the Cooling Pond
If gypsum and cooling ponds are segregated, the latter will contain fluorides
and small amounts of ?2O5. Thus it is necessary only to add significant lime to
precipitate out fluoride compounds as fluorspar. In bringing the pH of the cooling
water from 1.4 to 3.9, the emission of fluorides can be reduced by 90%. This can be
achieved by single stage liming shown in Figure 6-11. It will be necessary to lime
the pond only once and thereafter to add sufficient lime to handle the theoretical
amount of fluorides entering the cooling pond. The pond with an initial pH of l.*f
will need the addition of 0.157 Ib/gallon to reduce fluoride emissions by 90% (1).
Installation of a single-liming system at existing plants should pose few prob-
lems since most of them presently use double liming for pond water overflow
discharges.
During steady state operation 135 Ib CaO per ton P2O5 will be required. Since
a relatively large quantity of lime is required, it will be more economical for the
plant to produce its own lime from limestone. Total approximate annualized costs
using this strategy will result in an estimated cost increase of $3/ton P2O5.
6.6.5 Dry Conveyance of Gypsum to Stacks
Transporting the gypsum to stacks using a dry conveyor belt rather than a
slurry pipeline could reduce the fluoride emissions. The benefits/disadvantages are
not as easily determined as the annual costs are calculable. Apparently it is
economically justified where the gypsum can be marketed for agricultural or other
uses.
Annualized operating costs of approximately $l.^/ton P2O5 were estimated (1).
6.6.6 Phosphate Rock Calcination
Defluorination of phosphate rock through precalcination prior to acidulation
with sulfuric acid occurs in two regimes. At low temperatures (up to 2000° F) 66%
of the fluorine is removed, while heating to fusion temperature (2500°F) achieves
90% removal. The off-gases must be scrubbed, which is a major disadvantage of this
option, making the economics very unattractive.
-105-
-------
o
ON
I
LIME
CAR
FROM BAROMETRIC
CONDENSER
COOLING POND
Figure 6-11: Proposed Single Liming System (1)
-------
6.6.7 Change to Hemi/Dihydrate Process
The hemihydrate-dihyrate process for making wet-process phosphoric acid has
been in use in Japan for several years. Its primary advantages are the higher overall
yield of P2O5 and the production of high quality gypsum, suitable for gypsum plaster
and gypsum boards. About 40% of the fluorides present in the rock are volatilized in
the acidulation stages and may be recovered as f luosilicic acid.
The variations of this process differ from the common dihydrate process in
that hemihydrate is formed during decomposition of rock and is recrystallized as
dihydrate before filtration. It is estimated that 70-95% of the fluorides can be
recovered as H2SiF6.
In order to use the hemi/dihydrate process, the existing plants would have to
be extensively changed and markets found for the gypsum and fluosilicic acid
produced. The economics with low cost U.S. rock are questionable.
6.6.8 Use of Algae
Initial work with algae indicates that the pH can be increased to 3 and the
fluoride concentration decreased by a factor of 2 to 7 within 3 hours. Higher plant
life, such as 3uncus and autotrophic bacteria have given indications of providing a
similar effect (4).
This option is at the for scientific feasibility evaluation stage.
6.6.9 TESI Dry System (»)
In this process the gaseous fluorides are removed at elevated temperatures by
a dry chromatographic technique (TESISORB). Fine particulate is also collected in a
dry reactor. Since no cooling is required, a heat load normally imposed on the pond
is circumvented.
The cost of TESISORB is $0.07/ton P2O5. The capital cost for the hardware is
estimated to be one-half of the cost for wet scrubbing.
Testing recently completed by EPA of the TESI dry system at a secondary
aluminum installation demonstrated good performance. Several secondary aluminum
plants now use this system which might be applicable to fertilizer plants. A
recommended demonstration project will be described in Section 9.
-107-
-------
6.6.10 Comparison of Control Options
The cost and efficiency of control options 2 through 6 is shown in Table 6-6.
The Swift process appears to be the most cost-effective method in conjunction
with a segregated two pond system. The economics of this process depend on the
H2SiF6 market. Process complexity is another obstacle to its commercialization.
The second most cost effective method involves liming the pond waters to
pH 4. The advantage of this option is its simplicity. The disadvantages are
connected with secondary environmental impact caused through handling of large
quantities of limestone.
The Kidde process has the highest potential recovery and the greatest tech-
nical merit. Unfortunately, the high annualized cost and process complexity makes
its application rather dubious. The other three processes suffer from the disad-
vantage of needing major plant changes and none is currently used on a large scale
in the U.S.
The use of algae looks interesting but needs more research before it can be
considered as an engineering option.
The TESI dry system has been successfully demonstrated in the secondary
aluminum industry and a demonstration of this method in a phosphate fertilizer
plant is proposed.
6.7 References
1. Evaluation of Emissions and Control Techniques for Reducing Fluoride
Emissions from Gypsum Ponds in the Phosphoric Acid Industry, EPA-
600/2-78-124, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, June 1978, 218 pp.
2. Lehr, J.R., Fluorine Chemical Redistribution in Relation to Gypsum
Storage Pond System, The Fertilizer Institute Environmental
Symposium - 1979, New Orleans, Louisiana, March 6-8, 1978, 16 pp.
3. Environmental Science and Engineering, Inc., Personal Communication
from 3.C. Knott to 3. Sosebee (1975).
4. Teller, AJ., New Technologies in Control of Fertilizer Plant Emissions;
Pond Control - Fluoride Products, Presented at the Fertilizer Round
Table, November 1975, 16 pp.
5. Cross, F.L., Ross, R.W., 3. Air Pollution Control Assoc., 19, No. 1,
January 1969, 3 pp.
6. Tatera, B.S., Parameters which Influence Fluoride Emissions from
Phosphoric Acid Gypsum Pond, Ph.D. Dissertation, University of
Florida, 1970.
-108-
-------
TABLE 6-6. CAPITAL INVESTMENT AND OPERATING COSTS FOR FLUORIDE CONTROL OF 1,000 TPD P2O5 PLANT (J
Process
Kidde
Swift Absorption
Liming
Conveyor
i
o
Pre-Calcination
of Rock
Where Applied
Barometric
Condensers
Barometric
Condensers
Ail Cooling
Pond Water
Gypsum Filter After
Aciduiation
Crushed Phosphate
Rock
Fluorine Removal
Efficiency
95-98%a
90%b
90%b
a
b
Total Capital
Investment
$MM
2.57
1.30
2.10
1.07
29.81
Annualized Operating
Costs
Total, MM
2.31
0.17
0.97
0.43
7.57
$/Ton P205
7.46
1.25°
3.13
1.40
24.42
By-Product
(NH4)2SiF6
H2SiF6
None
None
None
Not calculated due to uncertainties in Fluorides evolved from filter cake.
90% removal of fluoride in rock is achieved. However, the fluorides evolved are scrubbed and transferred to ponds.
A credit of $2.26/ton P2O5 is realized if all fluorilicic acid produced is sold at $60/ton (100% H2SiF6).
-------
7. King, W.R., Farrell, J.K., Fluoride Emissions from Phosphoric Acid
Gypsum Ponds, EPA 650-2-74-095, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, October 1974.
8. FEDERAL REGISTER, (41 FR 20582) May 19, 1976, p. 20582.
9. Wissa, A.E.Z., "Gypsum Stacking Technology" presented at the 1977
Annual Technical Meeting, Central and Peninsular Florida Sections,
AlChE, Clearwater, Florida, May 22, 1977, 44 pp.
10. Reconnaissance Study of Radiochemical Pollution from Phosphate Rock
Mining and Milling, Office of Enforcement, U.S. Environmental Protec-
tion Agency, Denver, Colorado, December 1973, 46 pp.
11. FEDERAL REGISTER, (43 FR 37582), August 23, 1978, p. 37582.
12. Martin, E.E., Development Document for Effluent Limitations Guidelines
and New Source Performance Standards for the Basic Fertilizer Chemi-
cals Segment of the Fertilizer Manufacturing Point Source Catgegory.
EPA-440/I-74-011-a, (PB 238652) U.S. Environmental Protection
Agency, Washington, D.C., March 1974, 170 pp.
-110-
-------
7.0 SOLID WASTE CONTROL
7.1 Sources and Amount of Solid Waste
There are three sources of solid waste in the phosphate fertilizer industry (1):
• Gypsum from the filtration of wet process phosphoric acid
• WPPA sludge
• Wet scrubber liquor
The quantity of gypsum produced in a WPPA plant ranges from 4-.6 to 5.2
metric tons of gypsum/metric ton PjOs; therefore, approximately 1,360 m3 of
gypsurn will be accumulated yearly per metric ton of P2O5 produced per day (2).
A second source of solid residue is phosphate rock from ciarifiers (sludge).
Effluent from ciarifiers ranges from 0.7 m3 to 3.2 m3/metric ton P2O5 (3).
The third source of solid waste is the liquor from wet scrubbers which are used
throughout the phosphate fertilizer industry. At ammonium phosphate plants,
scrubber liquor piped to the gypsum pond contains about 10 g of solid waste (con-
sisting primarily of silicon hydroxide) per kilogram P2O5.
7.2 Special Waste Regulation (4)
As a result of wind and water erosion, the large piles of by-product gypsum in
the vicinity of WPPA plants serve as sources of airborne and waterborne emissions
for years after the time of abandonment. Furthermore, these piles contain
thousands of curies of radium 226 in a readily leachable form (5). The piles are
sources of low radioactivity and EPA has addressed them under Resource Conserva-
tion and Recovery Act (RCRA) regulations. These regulations are controversial and
opposed by the industry.
In the course of preparing its hazardous waste regulations under SubtitleC of
RCRA, EPA realized that some portions of certain very large volume wastes would
come within the purview of the regulatory scheme. These wastes include cement
kiln dust, utility waste (fly ash, bottom ash, and scrubber sludge), phosphate mining
and processing waste, uranium and other mining waste, and gas and oil drilling muds
and oil production brines.
The Agency has very little information on the composition, characteristics, or
the degree of hazard posed by these wastes. Nor does the Agency yet have data on
the effectiveness of current or potential waste management technologies or the
technical or economic practicability of imposing the proposed RCRA standards on
facilities managing such waste.
The limited information EPA does have indicates that such waste occurs in
very large volumes, that the potential hazards posed by the waste are relatively low,
-111-
-------
and that the waste generally is not amenable to the control of techniques developed
under the proposed RCRA §3004 regulations. The Agency is calling such high-
volume hazardous waste "special waste" and is proposing to regulate it with special
standards.
Table 7-1 provides some information about those wastes which, when
hazardous, EPA proposes to regulate with special standards. With two exceptions,
EPA does not know how much of the total amount of waste generated in these
categories is, in fact, hazardous and thus subject to Subtitle C regulations. Only
waste which is "hazardous" as defined under EPA's §3001 rules , however, will be
treated as "special waste." Any portion of the waste on the following table which is
not hazardous under the 53001 standards is not regulated at all under Subtitle C of
RCRA and thus is not "special waste."
Under EPA's proposed "special waste" regulations, the treatment, storage and
disposal of phosphate fertilizer solid waste will have to be managed in accordance
with standards set forth in Section 250.46-3 which provides as follows:
(a) The treatment, storage, and disposal of hazardous waste listed below
(and which is listed as hazardous waste in 250.14 of Subpart A) are
subject to the requirements specified in paragraphs (b) and (c):
(1) Overburden, slimes (phosphoric clays) and tailings from phosphate
rock mining;
(2) Waste gypsum from phosphoric acid production; and
(3) Slag and fluid bed prills from elemental phosphorus production.
(b) The requirements of the following Sections of the proposed 3004 stan-
dards are applicable to waste listed under paragraph (a):
250.43(f) (General Facility Standards—waste analysis);
250.43-1 (General Site Selection—for new sources only);
250.43-2 (Security);
250.43-5(a), (b)(l), (b)(2)(i), (b)(6), and (c) (Manifest System, Recordkeep-
ing, and Reporting);
250.43-6 (Visual Inspections);
250.43-7(1), and (m) (n) (Closure and PostClosure); and
250.43-8(a), and applicable requirements of (c) and (d) which relate to
groundwater monitoring, (Groundwater and Leachate Monitoring-for
groundwater monitoring only).
-112-
-------
TABLE 7-1. SPECIAL WASTE (Metric tons/yr) (4)
Waste
Cement Kiln Dust
Utility Waste (fly ash, bottom ash,
scrubber sludge)
Phosphate Mining, Beneficiation,
and Processing Waste
Uranium Mining
Other Mining Waste
Gas and Oil Drilling Muds and
Oil Production Brines
Quantity
Possible Hazard
12 million* Alkalinity and heavy metals
66 million* Heavy metals (trace)
400 million Radioactivity (low levels)
150 million Radioactivity
- billion Heavy metals, acidity
5 million Alkalinity, heavy metals,
toxic organics, salinity
* Hazardous waste portion unknown.
-113-
-------
(c) Additionally, the following requirements are applicable to waste listed
under paragraph (a):
(1) Location of waste deposits shall be recorded on reference maps
which shall be maintained through the operating and post-closure
periods.
(2) Land reclaimed by filling with waste listed in paragraph (a) shall be
used for residential development only where provisions have been
made to prevent alpha radiation exposure from Radon 222 inhala-
tion from exceeding background levels by 0.03 Working Level Units
and gamma radiation from exceeding background levels by 5 micro
Roentgens/hour. The possible need for special construction
methods for structures on such reclaimed land shall be identified to
any future land owner(s) by recording a stipulation in the deed of
the reclaimed land.
(3) Building products manufactured from waste listed in paragraph (a)
shall not be used if such products cause alpha radiation exposure
from Radon 222 inhalation to exceed background levels by 0.03
Working Level Units or cause gamma radiation to exceed back-
ground levels by 5 micro Roentgens/hour. Purchasers of waste and
of products manufactured from waste shall be advised of this
requirement by the seller.
(4) Analysis required under §250.43-8(c)(5) shall also include determin-
ation of Radium concentration in picocuries/gram.
(5) Analysis required under §250.43-8(c)(6) shall also include the
following:
(i) Radium, picocuries/gram
(ii) Phosphate, mg/liter
(iii) Fluoride, mg/liter
7.3 Phosphate Fertilizer Solid Waste Control
The solid waste originating from the three sources discussed in Section 7.1 can
be disposed in three ways (1):
1. Gypsum ponds and piles
2. Abandoned mine pits
3. Sea disposal
In the United States, more than 90% of the plants use gypsum ponds. As one
gypsum pond becomes filled, the gypsum slurry stream is diverted to another area
and the original pond is dried. The resulting gypsum piles range in height from 30 to
-------
36 meters. The disposal of gypsum in abandoned mine pits is practiced in Central
Florida. The only environmental hazard associated with such disposal is leaching of
fluorides, phosphates and radium 226 in the aquifer. Since mine pits are closer to
the subsurface of the aquifer, a leaching hazard is greater than that found in gypsum
ponds.
Sea disposal is mostly practiced in foreign countries while less than 296 of U.S.
plants use it.
For every ton of phosphoric acid produced from phosphate rock, about five
tons of waste gypsum are generated. Since most of the gypsum is disposed of
through a gypsum pond-pile system, the stacking technology will be discussed next.
Planning of a gypsum stack and acid water system is as important as the
detailed engineering design (6). Figure 7-1 shows a planning flowchart for gypsum
stack and pond design illustrating necessary steps and their interrelations. The first
step is the establishment of the process and climatic constraints. Process
constraints include production estimates, plant and stack life, process water
temperature, etc. Climatic constraints may influence process constraints, in
particular heat and water balances.
Sizing of the stack and pond takes place in the next step. The area required
for the gypsum stack is dictated by its design maximum height. After the sizing a
field investigation should be carried out of the subsurface condition of a tentative
site. The final selection of the site can be made in a rational manner by weighing
the engineering pros and cons of each site. In designing a gypsum pond-stack sys-
tem, one can use an idealistic layout based on experience as described in Refer-
ence (6). The following features have been included in this idealistic layout:
1. The gypsum stack is square in plan.
2. Two gypsum ponds are used. One is active while the other one is
drained.
3. The acid water cooling pond and surge area is built entirely around the
perimeter of the gypsum stack.
4. An exposed clay starter dike is used.
5. An internal drainage system is used to control seepage.
6. The decant structures are used in the center of the divider dike so that
the finer gypsum settles at the center of the stack around the decant
structure.
The detailed engineering design of the gypsum pond-stack system is described
in Reference (6). The major environmental concern regarding the gypsum pond-
stack system is groundwater contamination. This problem was discussed in Section 6
with a conclusion that, pending the results of ongoing field programs, contamination
is probably limited to the water table aquifer.
-115-
-------
PROCESS
CONSTRAINTS
CLIMATIC
CONSTRAINTS
SIZING
SURGE
POND
AND
LIMING
STATION
ACID
EVAPORATION/
COOLING
POND
GYPSUM
STACK
PRELIMINARY
SITE
SELECTION
I
PRELIMINARY
FIELD
INVESTIGATION
FINAL
SITE
SELECTION
i
CONCEPTUAL
DESIGN
ZONING
AND
PERMITS
DETAILED
ENGINEERING
DESIGN
Figure 7-1: Gypsum Stack and Acid Water System (6)
-------
An important question regarding phosphate fertilizer solid waste is the
possibility of resource recovery. The gypsum has been used in the United States on
calcium-deficient soils, alkali soils and for land reclamation. In Europe and Japan
relatively pure gypsum obtained from the hemihydrate process is successfully used
for wallboard production.
While several potential resource recovery methods are technically feasible,
less than 1% of the gypsum waste in the United States is utilized because its
recovery is not economically feasible.
7 A References
1. Nyers, J.M., et al, Source Assessment: Phosphate Fertilizer Industry,
EPA 600/2-78-004, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, May 1978, 185 pp.
2. Lutz, W.A., Pratt, C.3., Principles of Design and Operation. In:
Phosphoric Acid, Volume I, Slack, A.V. (ed.), Marcel Dekker, Inc., New
York, NY, 1968. pp. 158-208.
3. Martin, E.E., Development Document for Effluent Limitations Guidelines
and New Source Performance Standards for the Basic Fertilizer Chemi-
cals Segment of the Fertilizer Manufacturing Point Source Category.
EPA 440/1-74-011-e (PB-238 652), U.S. Environmental Protection
Agency, Washington, D.C., March 1974, 170 pp.
4. Proposed §3004 rules under the Resource Conservation and Recovery Act
(40 CFR 250, Subpart D, 43 FR 58991), December 18, 1978.
5. Reconnaissance Study of Radiochemical Pollution from Phosphate Rock
Mining and Milling, U.S. Environmental Protection Agency, Office of
Enforcement, Denver, Colorado, December 1973, 45 pp.
6. Wissa, A.E.Z., Gypsum Stacking Technology, Presented at 1977 Annual
Technical Meeting, Central Florida and Peninsular Florida Sections,
American Institute of Chemical Engineers, Clearwater, Florida, May
1977, 44 pp.
-117-
-------
8.0 REVIEW OF CONTROL EQUIPMENT VENDORS AND ENGINEERING FIRMS
Sections 5 and 6 discussed various air and water control techniques and equip-
ment but no specific mention of any vendors was made. The wastewater treatment
design is done by local engineering and consulting firms and the plants themselves.
Since all phosphate fertilizer processes use scrubbers for air emission control,
the discussion will be limited to scrubber vendors. Contacts with the phosphate
fertilizer industry revealed that a large number of vendors have their equipment
installed and operated in this industry. Since it is beyond the scope of this study to
make a complete inventory of pollution control equipment vendors, only four
selected vendors will be reviewed. Selection of the following representative vendors
was based on discussions with the Industrial Gas Cleaning Institute.
Large installations are represented by:
1. Environmental Elements Corporation
2. The Ducon Company, Inc.
Smaller installations are represented by:
3. The Ceilcote Company
4. Heil Process Equipment Company.
_ purpose of this review is to illustrate briefly the type of control equipment
furnished by these vendors and the process application. In preparing the discussion
on these four scrubber vendors, vendors' bulletins, open literature, discussion with
vendors and industry and the Mcllvaine Scrubber Manual (1) were used.
8.1 Environmental Elements Corporation (ENELCO), Baltimore, MD
ENELCO is a subsidiary of Koppers Company with annual scrubber sales of $1
to 5 million (1). The company's scrubber expertise has been acquired through
acquisition of Poly Con Corporation which is a subsidiary of ENELCO. ENELCO is
represented by Nosun Engineering Sales, Inc., Lakeland, Florida which serves as a
liaison with the phosphate fertilizer industry. Three major scrubber types manu-
factured by ENELCO and their principles of operation are shown in Figures 8-1, 8-2,
and 8-3 (2). The materials of construction vary with application but the most
commonly used are: mild steel, rubber lined steel, FRP and acid brick lined steel.
Mcllvaine lists the following design advantages and disadvantages for Venturi
and packed towers manufactured by Poly Con which apply to almost all conventional
designs (1).
VENTURI
DESIGN ADVANTAGES: The annular type and full throat Venturi have those
design advantages common to high energy scrubbers. High efficiency can be
-118-
-------
PRINCIPLES OF OPERATION
A - The contaminated gas enters
the Venturi and is accelerated in
the converging section.
B - The scrubbing liquid is intro-
duced, uniformly, at the top of
the converging section and cas-
cades by gravity and velocity
pressures towards the throat.
C - The contaminated gas and the
scrubbing liquid enter the
Venturi Throat where they are
mixed at high energy and extreme
turbulence.
D - The scrubbed gas and entrained
droplets (with contaminants en-
trapped) enter the diverging
section where further collisions
and agglomeration take place
creating larger drops.
E - The gases then proceed to the
separator where all liquid drops
are easily removed from the gas
stream and collected.
GAS OUTLET
GAS INLET
ANNULAR TYPE VENTURI
WITH FLOODED ELBOW
VENTURI WITH
CYCLONIC SEPARATOR
Figure 8-1: Enelco Venturi Gas Scrubber (2)
-119-
-------
PRINCIPLES OF OPERATION
The contaminated gas enters the involute section where it is
brought into intimate contact with positive curtains of liquid
spray. The treated gases proceed along the involute where the
bulk of scrubbing liquid is separated. The accelerated gases
then enter the spin zone where the remainder of the entrained
droplets are removed. The liquid travels by gravity to the
bottom cone drain and the cleaned gases leave the scrubber
at the top.
CLEAN GAS OUTLET
1 - INVOLUTE GAS INLET
2 -- WALL MOUNTED SPRAYS
3 - REDUCED HEIGHT
4 -- NO INTERNALS
5 - LOW PRESSURE DROP
DRAIN
GAS INLET
Figure 8-2: Enelco Cyclon-Spray Gas Scrubber (2)
-120-
-------
PRINCIPLES OF OPERATION
The contaminated gas enters the bottom of the tower and flows
upward through the packed section counter to the scrubbing
liquid. The packing forces the gas to follow a tortuous path
over the contact surfaces and interstices creating intimate
mixing with the descending liquid. The spray nozzles are
mounted on retractable manifolds. Demister located over the
spray manifolds prevents liquid entrainment. The scrubbing
liquid is collected and stored at the bottom of the scrubber.
CLEAN
GAS OUTLET
DEMISTER PAD
CONTACT BED
CONTAMINATED
GAS INLET
LIQUID INLET
RECYCLE SECTION
Figure 8-3: Enelco Poly Pack Tower Gas Scrubber (2)
-121-
-------
achieved at some energy expenditure. Pressure drop can be varied in both
units. Both types can be combined with the cyclonic type, separator/cooling
tower type, and demister type entrainment separators.
DESIGN DISADVANTAGES: These units have the disadvantages common to all
Venturi scrubbers. High energy requirements, and potential abrasion of
internals due to the high velocity gas stream are the most significant.
PACKED TOWER
DESIGN ADVANTAGES: Like other packed towers these units have relatively
high absorption efficiency for the energy consumed. Initial cost is also low
compared to other scrubber types.
DESIGN DISADVANTAGES: As in other packed towers scaling can occur
easily with precipitation or introduction of participate.
Figure 8-4 shows a WPPA process schematic with special emphasis on the air
emission control system (3). Since the author of the referenced article is now with
ENELCO, the pollution control schematic in Figure 8-4 would be representative of
ENELCO's approach. The same holds for Figure 8-5, showing an ROP-TSP process
schematic with emphasis on air pollution control.
8.2 The Ducon Company, Inc., Mineola, NY
Ducon is a subsidiary of the U.S. Filter Corporation which subsequently
acquired MikroPul. MikroPul already had a scrubber group called Airtron with a
number of installations in the phosphate fertilizer industry. Since the decision was
made to transfer all scrubber activities to Ducon, the Airtron Scrubber design was
phased out. Ducon is represented by Linder Industrial Machinery Company,
Lakeland, Florida which serves as a liaison with the phosphate fertilizer industry.
Figure 8-6 shows a cyclone, five wet scrubbers, and a brief feature description
of typical Ducon units (4). Mcllvaine states the following advantages and disadvan-
tages for the scrubbers: The VO Oriclone has a design advantage in that it is a
compact, low head room unit. The U.W-4 dynamic scrubber has the design
advantage that the wet fan stays clean on some applications where a dry fan would
not; however, unit height is a disadvantage. Figure 8-7 shows an Oriclone VVO
Venturi which is well suited for use in the phosphate fertilizer industry (4).
According to the vendor's bulletin, this scrubber has the following features (4):
1. Simple, trouble-free design—There are no spray nozzles or distribution
jets in which solids can collect; therefore, slurries, of any solids content
capable of being pumped, can be handled.
2. No wet-dry line build-up—Scrubbing liquid, introduced through open pipes
on the internal surface of the convergent section, swirls down in a flow
pattern that assures thorough wetting of the complete surface.
-122-
-------
CO
SULFURIC ACID
RECYCLE ACID
GROUND PHOSPHATE ROCK
ACID
WASH
FILTER CAKE
GYPSUM
TO WASTED*
CYCLONSPRAY
SCRUBBER
FILTRATE RECEIVERS
FILTRATE SEAL TANKS
POND WATER
RETURN TO POND
(30-32v2P6)
Figure 8-4: Wet Process Phosphoric Acid Flow ^heet (3)
-------
GROUND PHOSPHATE ROCK
•PHOSPHORIC ACID
TAIL GAS
CYCLONIC
SCRUBBER
VFNTURI CYCLONIC
.SCRUBBER SCRUBBER
16-20% 2-55',
ACID ACID
RECYCLE RECYCLE
TANK TANK
FLUORINE
TAIL GAS
SCRUBBER
FAN
STORAGE (CURING) PILE
STORAGE
BUILDING
Figure 8-5: ROP Triple Superphosphate Flow Sheet (3)
-124-
-------
ro
Ul
Multivane Scrubber
The Multivane Centrifugal
Scrubber is particularly suited
for applications involving
heavy dust loads. This unit
operates with centrifugal ac-
tion between gas stream and
liquid.
Wet Approach
Venturi Scrubber
This high energy scrubber
features a wetted wall inlet
which eliminates wet-dry line
build-up and permits direct
recycle of high-solids liquid,
Oriclone
Venturi Scrubber
This high energy Venturi Scrubber
provides ultra-efficient dust and
fume control with 99+% efficiency
into the sub-micron range. It
combines venturi agglomeration
and seperation.
Figure 8-6:
Cyclonic and Wet Scrubbers Manufactured
by Ducon (4)
-------
N)
ON
i
Duclone-Cyclones
Duclone Cyclones are designed
for high efficiency and low
resistance to assure maximum
recovery of industrial dusts
at minimum expense.
UW4-Dynamic Scrubber
This scrubber features three-
way scrubbing action and in-
corporates a wet fan as an
integral part of the unit.
Dynamic scrubbing action of-
fers 99+% cleaning efficiency
in the 1-2 micron range, with
low water requirements and
low pressure drop.
UW-3 Dynamic Scrubber
This compact, unitized scrubber
with self-cleaning wet fan fea-
tures high dust collection ef-
ficiency in minimum space with
low water requirements.
Figure 8-6: Cyclonic and Wet Scrubbers Manufactured by Ducon (4)
(continued)
-------
CLEAN GAS OUTLET
CYCLONIC SEPARATOR
SCRUBBING
LIQUID INLET
WET
APPROACH
VENTURI
THROAT
(ADJUSTABLE
OR FIXED)
Figure 8-7: Oriel one WO Venturi (5)
-127-
-------
3. Low water requirements—In applications where liquid is recycled, the
recycle tank can be made as an integral part of the entrainment separa-
tor. A fraction of the circulation load can be drawn off to a thickener or
centrifuge for clarification and the remaining liquid returned to the
scrubber. Scrubbing liquid lost in separation and through evaporation is
made up by direct addition to the recycle tank.
4. Handles high temperature gases—WO Scrubbers with the wet approach
design feature can handle inlet gas temperatures to 1000 °F and higher
without the requirement of separate quenchers.
5, Special construction for unusual environments—Scrubbers can be con-
structed of stainless steel, special alloys, plastics, or with special linings
for severe corrosive conditions.
Figure 8-8 shows a schematic of the DAP process emphasizing Ducon's emis-
sion control (5). Since cyclones are normally considered process equipment, the
emission control consists of individual Venturi scrubbers on four sources and a
common tail gas scrubber.
3.3 The Ceilcote Company, Berea, Ohio
Since 1975 Ceilcote has been a unit of General Signals. Its Air Pollution
Control Division has a line of wet scrubbers. The material of construction is usually
FRP and thermoplastics. Tellerette tower packing is also manufactured and can be
sold separately or as a packing in the scrubbing equipment. One of the company's
most important developments has been the IWS scrubber which will be discussed.
Ceilcote also provides conventional crossflow, Venturi and packed scrubbers which
can be applied in the phosphate fertilizer industry.
The IWS ("Ionizing Wet Scrubber") is a new type of air pollution control device
which collects aerosols and gaseous pollutants. The device combines the advantages
of electrostatic precipitators and wet scrubbers within one unit to achieve a high
rate of efficiency, while consuming a low amount of energy (6). Figure 8-9 shows a
simplified cross-section and side view of the IWS scrubber (6). According to vendor
literature, there is no defined lower size limit to aerosol collection, so that virtually
the same collection efficiency is achieved on the very small aerosols (down to 0.025
microns) as it is on the large aerosols. As described in Section 5, aerosols of 2-5
microns and larger are collected through inertial impaction. Aerosols smaller than 3
microns become attracted and attached to collecting bodies through impaction and
image force attraction. Image force attraction is a phenomenon in which the
charged aerosols come close to the neutral surface (Tellerette packing/liquid
droplets), induce a charge of opposite polarity, and are ultimately attracted to these
surfaces. Collected particles are washed from the scrubber by the scrubber liquid.
The vendor lists the following features for the IWS scrubber (6):
1. Collects most any type of micron/submicron particulate with equal
efficiency.
-128-
-------
r-o
vD
t
PHOSPHORIC ACID
*Kj €7*
DUCON VENTURI
AND CYCLONIC SCRUBBER
FINISHED
PRODUCTS H STORAGE OR SHIPPING
OUT
RAW I ANHYDROUS AMMONIA-/- -,
MATERIALS-H PHOSPHORIC ACID ,! I
IN SULPHURIC ACID--i |,
^POND WATER
TO POND
OUCON "
ABATEMENT SCRUBBER
MATERIAL FLOW
LIQUID FLOW
PHOSPHORIC ACID
'DUCON V^J'DUCON VENTURI
'DRY CYCLONEI^ ANO CYCLON|c SCRUBBER
TO AMMONIATOR
TO RECYCLE HOPPER
DRY AMMONIA
COLLECTION • RECOVERY
I FANS 1
FLUORINE ABATEMENT
Figure 8-8: DAP Process Schematic Including Ducon Scrubbers (5)
-------
> Inlel-To 50.000 ACF1
Pjrticit Cnarjmj Sect ton—Short. intensely charged High voltage DC
power Low energy consumption
i Ionizer Elements—Continuous Hustling prevents solids buildup
Factory AisemWed Modules -Fast in* i pensive held installation
All partictei charged as they enter scrubber wction
PUH.c W»*ii and Inttriwls—Corrouon Ren&Unt Minimum nainleiunce
Eihjusl ol IWS CM b*
connected directly lo stick
fin or ID mother IWS lof
increjsed collection
efficiency
I No mo*me Parts (only
L Recycled LIQUK* -For fit «btorptiafl. solids
Rmhuif
^— Collection Surfacet-Cofldnweusly flushed in
scrubber lection
All trfubtoer uirfares (packing «j|ef drops shell)
tutUcet
of rollert paflitta thtoufh imaee loicc jtlurtion
T»ll*rr!!.
and hi{fii,
wjtrf dropleli to* hfcM.smj1 tfnpme«ment «nd
H>rh v •'!»». DC Pdwet Suao*»
Figure 8-9:
Simplified Cross-Section and Side-View
of the IWS Scrubber (6)
-130-
-------
2. Removes solid and liquid particulates and noxious gases simultaneously.
3. Acts as a fractional collector. Percent of particles removed varies little
with load.
4. Collection efficiency remains constant with turndown from 10096 to 0%
load.
5. Consumes little energy (2.0 to 2.5 BhP per 1000 CFM per stage).
6. Operates at low static pressure drops (less than 2.0 inches W.G. per
stage).
7. Is impervious to corrosive atmospheres because of predominance of
plastic construction.
8. Provides long operating life and requires little maintenance.
9. Modular construction simplifies on-site erection.
10. Handles large variations in process flows without decreased efficiency.
11. Can be continually upgraded by adding on additional modules as require-
ments change.
Mcllvaine lists the following advantages and disadvantages for the IWS Scrub-
ber (1).:
DESIGN ADVANTAGES: Unit will not only remove coarse particles, but will
remove fine particles at high efficiency and low energy consumption (e.g., fine
particles 0.02-2 microns are removed nearly as efficiently as > 2 micron
particles and energy required is approximately equal to a Venturi scrubber A P
of 10 inches water gauge). Collection efficiency remains nearly constant with
turn down from 100% to 0% load. The fiberglass reinforced polyester material
of construction provides a lightweight, corrosion-resistant scrubber. Factory
assembled modules up to 50,000 CFM reduce installation costs.
DESIGN DISADVANTAGES: High voltage-liquid combinations necessitate
strict safety precautions and maintenance and operating procedures. Plastic
materials of construction must be protected from high temperature gases.
Clean scrubbing liquor must be used so a clarifier is required. Installation
space requirements for the system could be greater than for a similar conven-
tional modular wet scrubber because of the auxiliary electrical requirements.
Since this unit is relatively new, it has not been widely applied in the phos-
phate fertilizer field. It has a good potential in DAP emission control where new
evidence (7) shows that fluorides are emitted in the form of submicron aerosols
(mostly ammonium bifluorides).
-131-
-------
8.4 Heil Process Equipment Company, Avon, Ohio
Heil, a Division of Dart Industries, Inc., has 40 years of experience in
corrosion-resistant control equipment. It has a number of fume scrubber installa-
tions in the phosphate fertilizer industry. The majority of applications involve
removal of dust combined with acidic gases and ammonia. Fume scrubbers have also
been used for scrubbing and in some cases recovery of hydrofluoric and fluosilic acid
from acidulation operations.
Figure 8-10 shows three arrangements of Heil 720 Venturi scrubbers. Although
these units of conventional design represent a small market share of the total
Venturi market, their share in FRP is significant. Heil also has a significant market
share in water jet eductor-Venturi scrubbers (9). Applications in the phosphate
fertilizer industry involve recovery and scrubbing of hydrofluoric and fluosilic acid.
The design advantages include compact size and elimination of the fan, while
the disadvantages consist of a high water requirement and a need for high pressure
sumps (1).
i-5 Major Engineering Firms Designing Phosphate Fertilizer Plants
There are a number of engineering firms involved in the design of phosphate
fertilizer plants. The following three appear to be the leading ones:
1. Davy Powergas (including Wellman-Lord technology)
2. Jacobs Engineers (including Dorr-Oliver technology)
3. Waverly
Davy Powergas has designed and constructed the largest number of phosphate
fertilizer plants in North America. According to the Davy Powergas brochure,
contracts for 38 out of the last 43 WPPA plants and 31 out of the last 34 GTSP-DAP
plants were awarded to Davy Powergas (10). Consequently, only Davy Powergas will
be briefly described as representative of engineer-constructor firms dealing with the
Phosphate Fertilizer Industry. It is also important to mention that Davy Powergas
designs its own air pollution control equipment for the industry.
The wet process phosphoric acid plants designed by Davy Powergas are based
on the Prayon process. Figure 8-11 shows a typical WPPA process flowsheet (11).
Since this process is similar to one described in Section 4, it will not be described
here. According to Davy, the economic advantages of the Prayon process are (10):
1. Overall reliability and onstream factor is greater than 90% which is
superior to other processes, making the Prayon Process the industry
standard in North America.
2. Very adaptable to the use of any type of rock. Prayon has the expertise
in testing all grades of rock to enable optimum design.
-132-
-------
VERTICAL VENTURI SCRUBBER
WITH BLOWER LOCATED ON
DISCHARGE SIDE &
RECIRCULATION SYSTEM
HORIZONTAL VENTURI SCRUBBER
WITH BLOWER LOCATED ON
DISCHARGE SIDE WITH TYPICAL
RECIRCULATION TANK, PUMP,
AND PIPING.
VERTICAL VENTURI SCRUBBER
WITH BLOWER LOCATED ON
INLET SIDE INCLUDING
RECIRCULATION SYSTEM
Figure 8-10: Three Arrangements of Heil 720 Venturi Scrubbers (8)
-133-
-------
^SULPHURIC ACID
i-GROUND PHOSPHATE ROCK
-TO GAS SCRUBBING FILTER WASH WATER
SYSTEM
tf
OO
OD
REACTOR
t f
\J/\I/
I I I
FILTER
WATER
GYPSUM TANK
CONDENSER
FLASH
EVAPORATOR
L.P. STEAM
L-4
GYPSUM SLURRY
HEATER
30% P205
PHOSPHORIC ACID
l ,
WEAK ACID
STORAGE TANK
WATER
Figure 8-11: Typical Phosphoric Acid Process Flow Sheet (il)
STEAM
EJECTOR
50% P205
PRODUCT ACID
-------
3. Plant design and layout minimizes maintenance costs, less than 5% of
capital per year.
k. Defoamer usage is much less than competing processes, less than 25£/ton
P2O5 as compared to $1 to $2.00/ton P2O5 for others.
5. The adaptability of the Prayon Process is very suitable for the construc-
tion of large plants, (e.g., International Minerals & Chemical Corp., 2
trains @ 912 TPD and Agrico, one train (9 1200 TPD).
In designing granulation plants Davy uses DPG and Fisons technology. Fig-
ure 8-12 shows a granulation plant process flowsheet (11). This flowsheet is typical
for a range of nitrogen-phosphorus and nitrogen-potassium-phosphorus granulated
fertilizers. Typically, these plants are used for manufacture of GTSP and DAP;
however, depending on the solid raw materials available—powder MAP, ground
phosphate rock, ground TSP, potassium chloride, etc—a range of NPK fertilizers
can be manufactured in a similar way. The advantages of the Davy Powergas design
of the granulation plants are (10):
1. Overall reliability and onstream factor is greater than 90%.
2. The use of the Slurry Process and Rotary Granulator lowers capital cost
installation as compared to Pug Mills.
3. Plant design and layout minimizes maintenance costs, 5% or less of
capital per year.
Davy Powergas designs its own air pollution control equipment but no informa-
tion on design specifics is available. In the past Davy Powergas also worked closely
with Teller Environmental System on the design of emission controls, specifically
the crossflow scrubber. Figure 8-13 shows Teller's design of a crossflow packed
scrubber which was applied on various phosphate fertilizer processes and, together
with similar designs, represents best available control technology. This scrubber
permits high gas and liquid flowrates, is modular in design and has a maximum
capability for particuiate feed or deposition (12). As reported by EPA, the
emissions for a well designed Teller Scrubber range from 1-3 PPM of fluorides,
essentially in equilibrium with the pond water in the order of 0.003 to 0.01 Ib F/ton
P205 (12).
8.6 References
1. Mcllvaine Scrubber Manual, The Mcllvaine Company, Northbrook,
Illinois.
2. Environmental Elements Corporation, ENELCO Scrubbers, Form DVGM,
Rev. 10/76.
-135-
-------
WATER-
SCRUBBER
TANK
30% P205 ACID
RETURN SCRUBBER LIQUOR
STACK GAS
FEED SCRUBBER LIQUOR
TO ATMOSPHERE
WATER
GRANULATOR GAS
CYCLONES
SCREENS
SUMP
DRYER ELEVATOR
7
RECYCLE
ELEVATOR
"EQUIPMENT
VENTS
WATER
DRYER FAN
STACK
COOLER FAN
EFFLUENT
WATER
-* SOLID RAW
DRY MATERIALS
BINS
BIN
V O'r**^
RECYCLE CONVEYOR ~* COOLER'"N/U r=
COATING AGENT
COATING!
COATING
FEEDER
SCALE
CONVEYOR
Figure 8-12: Granulation Plant Process Flow Sheet (11)
-------
GAS INLET
GAS
INLET
ALTERNATIVE
IRRIGATING
LIQUID OUTLET
IRRIGATING LIQUID INLET
SEPARATE OR COMBINED SUMP
IRRIGATING
LIQUID OUTLET
Figure 8-13: Teller Crossflow Scrubber (12)
-------
3. Hill, L.J., Install Adequate System at Start to Control Fertilizer
Emission, Croplife, March 1976.
4. Ducon Bulletin No. W 7572, Dynamic Gas Scrubber Type UW-4.
5. Ducon Bulletin No. W 9075, Ducon Venturi Scrubber Type WO.
6. Ceilcote Bulletin 1201, Guide to Selection of Air Pollution Control
Equipment.
7. Private Communication with Dr. A. Teller.
8. Heil Bulletin No. B-720-2, Heil Venturi Scrubbers.
9. Heil Bulletin No. B-770-1, Water :et Eductor Venturi Scrubbers.
10. Davy Powergas, Qualification Summary for Phosphate Chemicals.
11. Davy Powergas, Bulleting D-20, Phosphate Fertilizer Plants.
12. Teller, A.3., Scrubbers in the Fertilizer Industry, Their Success, Near
Future and Eventual Replacement, presented at the Fertilizer Round
Table, Washington, D.C., November 1973, 8 pp.
-138-
-------
9.0 RECOMMENDED RD&D PROJECTS
One of the major objectives of this project was to identify gaps in information
and determine needs for future research and development and full scale demonstra-
tion. This section will recommend RD&D programs needed for a better understand-
ing of the phosphate fertilizer industry's environmental problems and available
solutions. The discussion of proposed RD&D programs outlines background informa-
tion and generalized methodology as to how such programs can be carried out. It
was beyond the scope of this wor k to prepare detailed descriptions of specific RD&D
programs. A detailed scope of work for such projects can be prepared after the
decision is made as to which studies to pursue.
The following recommended RD&D projects are described in this section:
1. An epidemiological study of the phosphate fertilizer industry
2. Studies of gypsum pond emissions and chemistry
3. An evaluation and optimization of wet scrubbers
4. An ammonia-sulfuric acid mist interaction atmospheric chemistry and
dispersion modeling study
5. A demonstration of a dry system for fluoride removal
6. An evaluation of the Kimre mist eliminator.
9.1 Epidemiological Study of the Phosphate Fertilizer Industry
9.1.1 Background
The phosphate fertilizer industry spends a considerable amount of money for
the purchase, installation and operation of pollution control equipment. Some
industry spokesmen consider that industry has gone as far as necessary in emission
control and that stricter control is warranted only if a health hazard is demons-
trated. There is increasing concern that environmental pollution plays an important
role in the development of cancer, even though the etiology of most cancers is
unknown. The relationship of environmental agents to the carcinogenic risk to man
has been discussed by the International Agency for Research on Cancer (1):
Evaluation of the carcinogenic risk to man of suspected environmental
agents rests on purely observational studies. Such studies must cover a
sufficient variation in levels of human exposure to allow a meaningful
relationship between cancer incidence and exposure to a given chemical to be
established. Difficulties arise in isolating the effects of individual agents,
however, since people are usually exposed to multiple carcinogens.
-139-
-------
The initial suggestion of a relationship between an agent and disease
often comes from case reports of patients with similar exposures. Variations
and time trends in regional or national cancer incidences, or their correlation
with regional or national 'exposure1 levels, may also provide valuable insights.
Such observations by themselves cannot, however, in most circumstances be
regarded as conclusive evidence of carcinogenicity.
In the case of the phosphate fertilizer industry, the major concern would be
low level radioactivity from Uranium in phosphate rock, gypsum ponds, and stacks.
Natural uranium contains approximately 99.82% uranium 238 and 0.71% uranium 235.
The decay chain of uranium 238 (known as the uranium-radium family), shown in
Figure 9-1, is of primary importance (2). The isotopes formed from uranium 238
present varying degrees of hazard. From the standpoint of water pollution, the
radioisotope radium 226 is the most hazardous. It has a maximum permissible
concentration (MFC) an order of magnitude less than any of the other decay
isotopes. Phosphate rock contains approximately 50 microcuries of radium per
metric ton (2).
Radium tends to replace calcium in bone. Persons working in industry may
have been exposed for a number of years to low radiation. The methodology of
studying the health effects of workers in industry have been well defined and applied
to specific problems (3,4).
Of equal concern is the possibility that effluents from industrial waste may
lead to adverse health effects in persons Jiving in areas near the plants.
While the doses of exposure to such general populations are undoubtedly less than
those for workers, such exposures impact on a larger population for longer periods of
time. Any effects seen are likely to be of a lower magnitude and thus are more
difficult to detect.
In the United States, 83% of the phosphate mining in 1968 was done in Florida
and North Carolina with the bulk done in Florida. In Florida the industry is concen-
trated in Polk and Hillsborough counties. Figure 9-2 shows the epidemiological data
for specific cancers for white males in the State of Florida. Table 9-1 ranks the
counties in Florida based on the cancer mortality rate from respiratory tract
cancers. Although these data may appear to indicate that there is no problem
regarding respiratory types of cancer in Polk County especially, the conclusive
evidence of carcinogenicity can be derived only through a careful epidemiological
study. The following generalized methodology, with special emphasis on bone
cancer, is recommended.
9.1.2 Generalized Methodology for Epidemiological Study
In collecting information on human populations to assess the health effects of
pollution, either mortality data (the proportion of deaths to population) or morbidity
data (relative incidence of disease) may be assembled. Since many cancers lead
eventually to death, the use of mortality data in assessing the occurrence of cancer
is logical. Further, some mortality data are routinely collected by governmental
-------
ALPHA BETA BETA ALPHA
4.5 x 109 YR 24.1 DAY 1.1 MIN 2.5 x 105 YR
ALPHA ALPHA ALPHA ALPHA )
8.0 x 104 YR 1620 YR 3.8 DAY 3.05 MIN /
>>^ ^^ D-;1--1- ' -^^ Dr.1--1- ' ^_ m t. j.w ^^ Ri""""
^^ 83B1 -^
Rl-214 ^_ p 214 ^_ 210 ^^ Ri210
83Bl ^^84P° ^^82Pb ^~ Bl
BETA BETA ALPHA BETA
26.8 MIN 19.7 MIN 1.6 x 10~4 SEC 22 YR
C
210 - 206
BETA ALPHA
5.0 DAY 140 DAY STABLE
Figure 9-1: Uranium-Radium Family Mode of Decay (2)
-------
State of Florida, by County
EPIDEMIOLOGICAL DATA FOR WHITE MALES:
Average annual age-adjusted mortality
rates (per 100,000) due to malignant
neoplasm of the trachea, bronchus and
lung, for the years 1950-1969.
Source: U. S. Dept. of Health, Education
and Welfare (1974).
Average, U.S.A. = 37.98
Florida - 44.93
Figure 9-2: Epidemiological Data for White Males,
State of Florida (5)
-142-
-------
TABLE 9-1. CANCER MORTALITY RATE, STATE OF FLORIDA (5)
Rank
1
2
3
it
5
6
7
8
9
10
11
12
13
1*
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
3*
County
Flagler
Franklin
Monroe
Duval
Dixie
Marion
Hillsborough
Nassau
Wakulla
Citrus
Okeechobee
Escambia
Bay
Dade
Alachua
St. Johns
Leon
Charlotte
Columbia
Palm Beach
Gulf
Brevard
DeSoto
Taylor
Pinellas
Broward
Madison
Orange
Putnam
Hamilton
Polk
Volusia
Highlands
Collier
Cancer Mortality Rate
Lung, Trachea, Bronchus
85.8
62.2
57.1
56.8
55.2
53.7
52.5
52. 1
52.2
51.2
50.1
50.0
49.7
49.2
47.5
it7.it
47.2
45.0
44.9
44.2
44.1
44.1
43.9
43.7
43.2
43.0
42.8
42.6
42.5
42.2
41.8
41.7
41.2
41.0
Rank
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
County
Sarasota
St. Lucie
Baker
Indian River
Levy
Seminole
Okaloosa
Suwanee
Hardee
Lake
Sumter
Lafayette
Hernando
Manatee
Osceola
Bradford
Pasco
Glades
Walton
Hendry
Lee
Clay
Martin
Gilchrist
Jackson
Calhoun
Santa Rosa
Liberty
Union
Holmes
Washington
Gadsden
Jefferson
Cancer Mortality Rate
Lung, Trachea, Bronchus
40.0
39.8
39.1
38.9
38.6
38.4
37.7
37.2
36.3
36.3
36.2
36.1
36.1
36.0
35.8
34.9
34.6
34.2
34.0
33.1
33.1
33.0
32.7
32.6
32.5
32.5
28.8
25.7
25.5
24.8
24.6
24.3
18.6
NOTES: 1) Average annual age-adjusted mortality rates (per 100,000) due to malignant neoplasm of the trachea,
bronchus, and lung, for the years 1950-1969.
2) Source: U.S. Department of Health, Education and Welfare (1974).
-------
agencies and thus may be assembled at relatively low cost. Many other diseases,
such as cardiovascular disease and mental illness, may not lead to death; therefore,
it is necessary to evaluate morbidity data.
In categorizing the methods to be used in assessing the possible effects on
health of living in the area of a phosphate fertilizer plant, a primary question is
whether data collected for other reasons may be used. A secondary question is
whether the data are on mortality or on morbidity. Finally, data can be collected by
following groups of exposed and non-exposed persons or by evaluating past exposures
in groups of diseased and non-diseased persons. Not all of the possibilities are
realistic, and only six strategies will be considered.
These strategies can be outlined as:
1. Evaluation of routinely collected data
a. Mortality
b. Morbidity
2. Evaluation of specifically collected data
a. Mortality follow-up of exposed groups
b. Mortality evaluation of decreased persons
c. Morbidity ad hoc evaluation
d. Morbidity case-control surveillance
A short description of the above six strategies follows.
1. Evaluation of Routinely Collected Data
a. Mortality
Mortality rates have been used to evaluate the associations of
cancer with counties in which chemical (6) or petroleum (7) plants
are located. A similar evaluation of counties with phosphate
fertilizer plants should be carried out.
The methodology for such an evaluation is straightforward. The
first step is to select counties in which there are phosphate fertil-
izer plants, complete the cancer-specific mortality rates for the
entire set of such counties, and compare these rates to those of the
entire U.S. population. If higher rates of specific cancers are
found in counties where these plants are located, the possibility
that the cancer is a result of the industry's presence must be
considered.
Such evidence, however, is extremely tenuous. There may be other
characteristics of the general population in those counties which
-------
are associated with the specific cancer(s). For this reason, a
second step should be performed to evaluate this possibility.
Control counties, similar to the "exposed" counties, should be
selected, and the cancer experience in these counties should be
evaluated.
b. Morbidity
Morbidity data are routinely collected by the National Health
Survey. However, these data are based on a relatively small
sample and are unlikely to provide specific data on areas in which
phosphate fertilizer plants are located.
Morbidity data are also collected by the Professional Activities
Service (PAS) of the Commission on Professional Activities
(CPHA). Roughly 40% of the hospitals in the United States provide
diagnostic information to the PAS. As an example of the manner
in which morbidity data can be used in an epidemiological study,
data from the PAS have been used as the starting point for an
evaluation of the relationship between estrogen usage and myo-
cardial infarction (8).
It is possible that the PAS serves many hospitals in counties where
the phosphate fertilizer industry is located. Such counties could be
selected and compared to control counties in the distribution of
diagnoses reported to the PAS.
2. Evaluation of Specially Collected Data
a. Mortality Follow-Up of Exposed Groups
If persons live for many years in the environs of a phosphate fertil-
izer plant, and if effluent from the plant affects their health, it is
possible that such an effect may be detectable from the pattern of
cause of death among these people. Through the use of census
records or other population lists extending back many years, an
"exposed" population is assembled. A control population can be
also assembled from the same records. The methodology of such a
study is uncomplicated but the approach may be too insensitive to
detect small casual associations.
b. Mortality Evaluation of Deceased Persons
Rather than using mass data collected on populations, the cause of
death can be evaluated by examining individual death certificates.
For counties where the phosphate fertilizer industry is located, and
for control counties, all death certificates for a number of years
can be assembled. Two basic comparisons would be from such
data: between counties and within counties.
-145-
-------
c. Morbidity Ad Hoc Evaluation
Specially-collected morbidity data would be based on patient
records from hospitals in the phosphate fertilizer industry area.
Rather than reviewing all records, only specific diseases linked to
this industry exposure would be reviewed.
d. Morbidity Case-Control Surveillance
Hospitals serving populations where phosphate fertilizer plants are
located could be monitored. Patients could be interviewed, both
about their occupational exposure to the fertilizer industry and
about their residence pattern. Analysis would include the rela-
tionship between specific illness and residence near the plant.
The various types of studies outlined are discussed only in general terms. Each
of the studies discussed would seem to be feasible. The studies in which existing
data are utilized could be done in one or two years. The studies in which data is to
be collected would require three to five years at least. Even larger programs may
be needed if patterns of change with time are to be evaluated.
Whether an association in general populations between disease and industrial
pollution can be found is problematic. Data on human populations are subject to
much variation and weak associations may not be detectable. If such associations
are to be found, relatively large amounts of data must be collected and fairly exten-
sive analysis must be done. Associations seen must be duplicated in independent
populations. Interpretation of any association must be conservative.
The most cost-effective way to proceed with an epidemiological study for the
phosphate fertilizer industry appears to be a pilot study. The main targets would be
the plants with gypsum ponds near populated areas. Only a dozen plants would fall
in this category. As a first cut, the mortality rates for the counties where these
plants are located should be compared with rates for the entire U.S. population.
Simultaneously, control counties which have phosphate fertilizer plants that do not
have gypsum ponds should be selected and rates examined. The comparison of these
mortality rates can serve as a preliminary indication of risk associated with the
phosphate fertilizer plants. If the pilot study shows significant association, a
follow-up study to explain its causes should be planned. A more comprehensive
epidemiological study of the complete phosphate fertilizer industry does not appear
to be justified in the light of relatively high cost and possible inconclusive results.
The costs and time for a pilot epidemiologicai study vary greatly depending on the
availability of information.
9.2 Studies of Gypsum Pond Emissions and Chemistry
One concern about wet scrubbing for air emission control has been that this
technique in most cases does not result in the ultimate disposal of pollutants. Fre-
quently air pollution is simply replaced by water pollution or a solid waste manage-
ment problem. This concern fully applies to fluoride scrubbing in the phosphate
fertilizer industry. The fluorides are removed from stationary sources with a high
-------
efficiency (as a rule over 95%) and they are then transferred to a gypsum pond
where most of them become fixated. The remaining portion may be released to the
atmosphere. These emissions are governed by liquid/gas equilibria which have not
been thoroughly investigated. Table k-2 indeed indicates that emissions of fluorides
from gypsum ponds by far surpass the emissions from other sources in a WPPA plant.
Section 6 discussed at length all alternatives for gypsum pond emission
control, but no economical means of control surfaced.
9.2.1 Study of Fluorine Distribution
Section 6-2 discussed gypsum pond chemistry with special emphasis on fixation
of fluorides in the pond. A simplified model showing the reaction pathway (shown in
Figure 6-3) has been developed assuming the chemistry of relatively pure phosphate
rock. Today's lower quality feed rock permits only a qualitative modeling of the
chemical redistribution between product and waste stream.
Unless properly recognized and researched, there is the inherent risk that
environmental assessments, particularly remedial actions for fluoride containment,
will be fatally flawed by false chemical assumptions (9).
A thorough study of gypsum pond chemistry, namely fluorine distribution, is
recommended since this can lead the way to reduced fluoride emission. The major
objective of the study should be an understanding of the factors fixating fluoride in
the pond or even preferably in the cake. If the redistribution of fluoride in the
acidulation stage could be understood, most fluorides could be fixated in the cake
where it would be ultimately disposed.
Table 9-2 shows the controlling variables of the WPPA process (9). Precipita-
tion of fluoride dissolved from the rock is governed primarily by the interacting
concentration levels of Al, Si, Na, Mg, Ca, and F released into solution from the
feed rock. The quantities of these controlling constituents vary considerably and
the chemistry of rock is getting more complex as the quality of rock is decreasing.
In the past, with good quality rock the predominant fluoride salt was Na2SiF6. Rock
presently used contains other fluoride salts which are frequently more soluble (e.g.,
NaKSiF6). Precipitation of either ralstonite or chuckrovite in place of alkali
fluosilicates greatly increases the amount of fluoride in the byproduct cake. The
"recipe" for addition of materials leading to a chemical balance favorable to fixa-
tion of fluoride in cake would be an ultimate goal of a fluorine distribution chemis-
try study. A study of this nature with a limited scope is currently underway at TVA
but an additional and complementary study is needed to make findings usable on a
commercial level. It is not possible to estimate the cost and time-frame for such a
study until a detailed scope of work is prepared.
A study of gypsum pond fluoride liquid-vapor equilibria is also recommended.
Figure 9-3 shows competing equilibria governing the chemical state of fluorine (9).
Assuming that SiO2 is not a limiting constituent to establishing equilibrium, the
overall reaction in solution can be expressed as:
-147-
-------
TABLE 9-2. PARAMETERS OF F DISTRIBUTION IN
DIGESTER STAGE (9)
Controlling dissolved components
A13+, Si4+, Na+, Mg2+, F-, (Ca) A13+ Mg2+
Sources
Apatite: Na Mg F
Accessories: Na Mg Al Si
Variability (C.F.* 70 BPL)
Solubility in F.G
Na203
MgO
A1203
Si02
F
. WPA
Apatite source: ~100%
Accessory source: Highly
Mean wt. %
0.60
0.30
1.07
4.69
3.78
variable
Obs. range
0.28-0.82
0.15-0.52
0.30-1. 42
2.50-7.40
3.60-3.90
*Central Florida
-148-
-------
COMPONENTS
VAPOR PHASE:
PW PHASE:
SiF4, HF, H20
F-, Al3*, Mg2+, Si4+, (Na,
Fe3+, Ca2+, P043~, SO/"
SOLID PHASE: GYPSUM, F SALTS, INERTS
Figure 9-3: Competing Fluorine Chemical Equilibria (9)
-------
6 HF + SiO^=?rH SiF + 2 H 0
2 26 2
and Henry's law should be applicable:
PFv = K • CFs °r PFv - L58 x 10"7(Cppm)
in which
Pp = pressure (mmHg0) of total F in vapor phase
Cps = concentration (moles/liter) F in solution
C = concentration (ppm) F in solution
K = proportionality constant
The overall reaction sequence appears to favor HF emissions from ponds with
5,000-10,000 ppm F. Except for Reference (10) most literature examples deal with
higher fluoride concentration which favors higher SiF4 emission. It is a probable
source of speculation still maintained by some industry's chemists that SiF4 is a
major emission.
The ultimate goal of pond water vapor equilibria study would be to determine
the correlation between various pond composition and corresponding emissions.
Such a study is well suited for an academic institution.
9.2.2 Measurement of Gypsum Pond Emissions
As mentioned in Section 6.3, actual field measurements of fluorides were
carried out in 1977. The fluoride emissions from two gypsum ponds were measured
by remote sensing and point sampling/analysis methods. The concentration of
fluorides (primarily HF) over and at the edge of gypsum ponds was on the order of
tens of ppb. Vertical traverses of fluoride concentration were obtained by a wet
sampling/analysis method using a crane to position the sample. Average total
fluoride emission rates for the gypsum pond were estimated to be in the range of 0.2
to 7.3 Ib/acre-day.
The initial purpose of this study was to estimate the fluoride emissions rate
(flux) from gypsum ponds as a function of plant operations, atmospheric conditions
and gypsum pond characteristics. The data obtained during the field program were
too scattered and too few to make any correlations. An additional objective of the
above study was to prepare a dispersion model of fluoride emissions from gypsum
ponds.
-150-
-------
Since all of the above information is very important for further clarification
of gypsum pond emissions and control, it is recommended that subsequent field
measurements be carried out. The purpose of such a measurement program would
be to expand fluoride emission data from the gypsum pond data base and allow
correlation with atmospheric conditions and plant specifics. The upwind-downwind
sampling methodology applied in TRC's study should be refined and used (11). A
Gaussian dispersion model would be used in this study to simulate dispersion of
fluorides emitted from the gypsum pond. The diffusion coefficients may be adjusted
after the field evaluation.
A rough cost estimate of such a project is $70,000 and it could be carried out
in six months.
9.3 Evaluation and Optimization of Wet Scrubbers
Wet scrubbers installed in the phosphate fertilizer industry represent the
major investments for pollution control and costs run well into millions of dollars.
These scrubbers control particulate, fluorides and other gaseous emissions and, as a
result, most of the plants are in compliance with emission regulations. Unfortun-
ately, most of the plants are satisfied with this fact and do relatively little to
further improve scrubber efficiency and/or reduce operating cost. Although some
firms collect data on scrubber performance, practically no open literature data exist
that would benefit the entire phosphate fertilizer industry. There is even some
doubt whether the scrubbers perform as well as claimed by the equipment vendors.
Since reliable information on wet scrubber performance can result in better
efficiency and lower operating and maintenance costs, two R&D programs are
recommended. The first one is on scrubber optimization through a factorial design
experiment; the second one includes a thorough evaluation of the IWS scrubber
discussed in Section 8.3.
9.3.1 Factorial Design Experiments
The evaluation and optimization of scrubbers should be carried out on a full
scale unit in a representative plant. Since the testing is relatively costly and time
consuming, the testing program should be carefully planned. If the experiments are
planned wisely and incorporate statistical design, correlation is often apparent and
straightforward to extract (12). In order to reduce the number of runs and obtain
interaction of process variables, fractional factorial design experiments are recom-
mended. If a one-variable-at-a-time experiment is used, the interaction of
important variables can be easily overlooked.
The following variables are considered important in evaluation of the effi-
ciency of wet scrubbers:
1. Gas volumetric flow rate
2. Liquid volumetric flow rate
-151-
-------
3. Pressure drop
4. Residence time
5. Fluoride liquid concentration
The first step in experiment planning is the screening of variables. This is a
crucial step in testing where a good understanding of process variables is needed.
Two or three variable levels are selected and a statistical design is carried out.
Connor described the fractional factorial experiment design for factors at three
levels (13). In scrubber testing two level, five variables statistical design (2s-2) is
recommended. Table 9-3 shows the statistical design for scrubber testing. It can be
seen that eight runs are needed for the screening of variables. The following three
responses are considered to be important: fluoride removal efficiency, particulate
removal efficiency and energy consumption. The results of these tests could be
interpreted visually for most responses. A more rigorous analysis would consist of
averaging the results at the high level of the variable and subtracting the average at
the low level of the variable. Thus, an estimate of the variable's main effect can be
obtained.
In a full factorial design (e.g., 25=32 runs) this estimate is independent of
estimates of other effects. However, in a fractional factorial design the effects are
confounded, which means that main effects are tied-up with interaction. As can be
seen in Table 9-4, for example, the main effect of variable 1 is confounded with the
interaction between variables 3 and 5. However, it is reasonably certain that some
interaction in wet scrubber tests will be negligible and that most of the combined
effects can be attributed to the main effect rather than an interaction.
The main objective of a screening study is to determine variables with the
largest effect so that these can be optimized. If the optimum has not been
approached after an initial screening study, the method of steepest ascent should be
used to find the optimum region (14). Since wet scrubbers have been in operation in
fertilizer plants for a number of years, it is reasonable to expect that the selection
of variable values will result in near optimum conditions. An assumption is made
that two variables, e.g., gas volumetric flow rate and fluoride liquid concentration,
were found to have the largest effect. The optimization of these two variables will
be the next step. A simple design for studying both the linear and quadratic effect
of these two variables is a 32 equal 9 run design. Each variable is studied at three
levels as shown in Table 9-5.
Based on these tests, contour surfaces for all three responses can be
generated. If the optimum is really close, the contours would represent concentric
ellipses with unique optimum values of the variables at the center of the ellip-
ses (12). Again, the test results can be interpreted visually. A more rigorous
statistical analysis consists of fitting the desired response with an empirical
graduating function (e.g., polynomial). Such a model or regression equation would
contain linear terms in the variables, an interaction term, and quadratic terms in
both variables. This equation can be used as a prediction equation within the range
of variables studied and would be applicable to wet scrubbers and processes similar
to the one studied.
-152-
-------
TABLE 9-3. STATISTICAL DESIGN FOR SCRUBBER TESTING
Run XT
1
2 +
3
14 +
5
6 +
7
8 +
Fluoride Removal Particulate Removal Energy
X2 X3 X4 X5 Efficiency Efficiency Consumption
+ + f f
Variables
Hypothetical Variable Levels
1.
2.
3.
*-
5.
XT =
X2 =
X3 =
X4 =
X5 =
Gas volumetric flow rate
Liquid volumetric flow rate
Pressure drop
Residence time
Fluoride liquid concentration
20,000 ACFM
600 GPM
8 in WC
0.4 sec
5,000 ppm
25,000 ACFM
800 GPM
12 in WC
0.6 sec
8,000 ppm
-153-
-------
TABLE 9-4. CONFOUNDING PATTERNS FOR DESIGN (2)
1 +3x5+= (Effect of variable 1) + (Interaction between variables 3 & 5)
2 + 4x5
3 +1x5
4 +2x5
5 +1x3+2x4
1 x 2 + 3 x <4
1 x 4 + 2 x 3
All 3-factor interactions (e.g., 1x2x3) and higher are assumed negligible.
TABLE 9-5. OPTIMIZATION DESIGN FOR SCRUBBER TESTING
Run
9
10
11
12
13
14
15
16
17
x,
_
-
0
0
0
+
+
+
X2
0
+
-
0
+
-
0
+
Variables
Fluoride Removal Particulate Removal Energy
Efficiency Efficiency Consumption
Hypothetical Variable Levels
1. X, = gas volu-
metric flow rate
2. X2 = fluoride
liquid concen-
tration
20,000 CFM
23,000 ACFM
25,000 ACFM
-154-
-------
Cross-flow packed scrubbers are considered to be the most universal wet
scrubbers for the phosphate fertilizer industry and the factorial design experiments
should be carried out on one of them.
It appears that six months would be sufficient to perform such experiments.
The cost for such a program is estimated to be about $100,000.
9.3.2 Performance Evaluation of Ionizing Wet Scrubbers
In the light of increased emphasis on air pollution health effects, this type of
wet scrubbing equipment will be reviewed from the point of view of the respirability
of effluent particles (15). Regarding particulate emissions, the main concern from
the industrial hygiene standpoint is the amount of particulate deposited in the
alveoli and terminal bronchioles of the respiratory system. The deposition of
particulate in the pulmonary space is the function of particle size. The larger
particles (above 2 microns) are readily removed in the upper respiratory system
while the smaller particles (under 0.2 microns) are inhaled and exhaled. Conse-
quently, a particle size of about 1 micron presents the main health hazard, as such
particles have the maximum deposition in the alveoli (16, 17).
Section 5 of this report discussed the theory and practice of wet scrubbers
with a special emphasis on particulate and gaseous removal efficiency. The particu-
late removal efficiency is based on two basic mechanisms: inertial impaction and
diffusion. The larger particles are removed through inertial impaction while smaller
ones are removed through diffusion. Because of these basic removal mechanisms,
the efficiency is a function of the particle size and scrubbers operating under
standard conditions have the lowest efficiency on particulate in the range of 0.1 to 1
micron. The efficiency is, therefore, the lowest on the particle size that is the most
hazardous to human health.
It is important to note that dry gas cleaning devices like cyclones and filter
bags have the same limitation since particulate removal in these devices is based on
the same unit mechanisms. Electrostatic precipitators have also been reported to
have minimum efficiency in this particle size range (18). Most of the wet scrubbers
on the market today claim an efficiency of 99% and over. Unfortunately, in many
cases, the basis on which the high collection efficiency claims are made is not fully
described. Since the particulate emission regulations are as a rule based on a weight
basis, most removal efficiencies are also reported on a weight basis. However, some
scrubbers which may have a 99% efficiency on a weight basis can have a very low
efficiency on a particle size basis. The best way to improve the efficiency of wet
scrubbers on small particles (0.2 to 2 micron range) is to apply some basic
mechanisms which are not particle size dependent. One such principle is the
electrostatic charging of particulate or collecting bodies. There are a number of
wet scrubbers where this principle has been applied and the ionizing wet scrubber
(IWS) has been described in Section 8.4.
-155-
-------
The evaluation of IWS scrubber efficiency in particulate removal as a function
of particle size is highly recommended. It appears that this scrubber demonstrates
high particulate removal efficiency regardless of particle size distribution. The
verification of this fact is very important for the reduction of the above described
health hazards and its application, as well as the application of similar devices,
could go well beyond the phosphate fertilizer industry.
The major purpose of this evaluation will be to measure the particulate
removal efficiency as a function of the particle size distribution and determine
optimum operating conditions. The test should be carried out through simultaneous
sampling of the IWS scrubber inlet and outlet and subsequent measurement of the
particle size distribution. The important range of particle size distribution is in the
range of 0.02 to about 5 microns. The determination of particle size range is not
straightforward and the reliability of this experiment will probably be limited by
such measurements. A recently developed cascade impactor system for sampling
0.02 to 20 micron diameter particles can be applied to this study (19). The new
sampling train components were designed to accompany the UW Mark 3-4 Source
Test Cascade Impactor. This train was designed to sample in-situ at both the inlet
(high concentration) and the outlet (low concentration) of particulate control
devices applied to coal-fired utility boilers. Figure 9-4 shows the actual particle
size distribution for tests at a Generating Station (19). This instrument can be
applied for measurement of particulate size distribution in a phosphate fertilizer
plant. Since there is a growing concern that fluoride emissions from DAP plants are
submicron particles, the IWS tests should be performed on DAP plant emissions. The
operating parameters should be varied in the course of testing in order to obtain
optimum conditions for particulate removal. Possibly fractional factorial design
experiments described in a previous section could be used for this evaluation.
These tests could be performed within six months; the cost is unknown as it
will probably depend on the complexity of particle size distribution measurements.
9.4 Ammonia-Sulfuric Acid Mist Interaction Atmospheric Chemistry and
Dispersion Modeling Study
Frequently, manufacturing activities for inorganic acids (sulfuric and nitric
acid) and inorganic fertilizers are concentrated within one industrial area (fre-
quently even within one plant). Because of the proximity of points of emissions, the
interaction between the various air pollutants can have synergistic effects on the
environment and on air quality. Of particular interest and potential for interaction
are sulfuric acid mist emitted from sulfuric acid manufacturing plants, and ammonia
emitted from DAP plants and nitrate fertilizer facilities. Possible uncontrolled
emissions from each manufacturing activity are given in Table 9-6. Interaction
between ammonia and sulfuric acid mist will result in formation of ammonium
sulfate particles which will lead, at high relative humidity, to particle growth due to
condensation and thus to visibility impairment.
Since ammonium sulfate, ammonium nitrate and sulfuric acid have different
deliquescence points (62% RH for NH4NO3 80% RH for (NH4)2SO4 and «70% RH for
-156-
-------
10.0
o
OL
O
O
LT>
O
1.0
Q
Q
O
o:
0.1
a:
-------
TABLE 9-6. EMISSION FACTORS OF SOME INDUSTRIAL ACTIVITIES (20)
Manufacturing
Activity
H2SO4
Ammonia
Ammonium Phosphate
Ammonium Nitrate
Urea
Triple Phosphate
Pollutant
Emitted
Acid Mist
Ammonia
Ammonia
Nitric Oxides (as NO3)
Ammonia
Ammonia
Ammonium Chlorine
Emission Factor
kg/MT
3.75
100.
2
2.5 - 3.
25 - 50
-158-
-------
HjSOj, the transformation from one species to the other can have a marked effect
on the particle size distribution within the plume. Since the scattering coedfficient
(b ) is a function of the square of the particle diameter, a change in particle size
distribution will cause a significant effect on visibility.
In order to fully understand the potential effect of the interaction of ammonia
and sulfuric acid mist on visibility, the following studies are necessary:
1. Development of a model that will describe the interaction of two plumes
and will incorporate an aerosol dynamics model. The required model will
describe the dispersion of two plumes as independent entities until
"point" of interaction. After interaction the model should incorporate
the chemical reaction between ammonia and sulfuric acid (the reaction
will, probably, be diffusion control.) At this stage, the model should also
include an aerosol dynamics model that will calculate the evolution with
time of the aerosol size distribution as a function of aerosol composition
and relative humidity.
2. Experimental studies on the interaction between gaseous ammonia and
sulfuric acid mist and its effect on the optical properties of the
atmosphere, at different relative humidities. Experimental studies
should be conducted in a chamber with controlled relative humidity into
which acid mist aerosol could be injected (via a vibrating needle aerosol
generator, for example) together with gaseous ammonia. The change, in
time, of transmitted light attenuation and/or light scattering should be
measured. This study can be accomplished within 6 months at an
approximate cost of $100,000.
9.5 Demonstration of Dry Fluoride Removal System
One of the major roles of the gypsum pond system is to cool and recycle
scrubbing liquid used for emission control. Recycled scrubbing liquid is used for the
reduction of the gas temperature needed to reach the low levels of fluoride
emissions. It is estimated that approximately half the thermal load in the fertilizer
complex results from the emission control (21). The dry chromatographic technique
can accomplish the same function as wet scrubbers, namely simmultaneous emission
of particulate and gaseous emissions. In this technique no cooling is necessary so
that a pond is not needed.
The dry chromatographic system has been used in the secondary aluminum and
glass industries. Figure 9-5 shows the system as used on secondary aluminum
smelters (22). This dry chromatographic system consists of an evaporation chamber,
baghouse and blower. The blower induces discharge gases for the stack hoods into
the inlet duct work. The gases are respectively cooled by water punching and air
blending to a range of 160°F to 220°F before entering the baghouse. The filter bags
are protected from blinding and burning by application of a chromatographic
coating. The coating absorbs and neutralizes acid gases. It is applied batch wise to
the bags through precoated inlet ducts. Once applied, the coating is effective for
-159-
-------
COMBUSTION
ZONE GAS
EXHAUST FAN
EVAPORATION
CHAMBER
REVERBERATORY
FURNACE
VENTURI
DISCHARGE CHUTE
CHARGING
ZONE GAS
CHARGING SCRAP
Figure 9-5: Dry Chromatographic System (22)
-160-
-------
several days, during which time the system may be left unattended. Table 9-7
shows the performance of a TESI dry chromatographic system on three applications.
The application of this system to phosphate fertilizer emissions is expected to
reduce pond size 50 to 67%. Total fluoride pond emissions are expected to be
reduced as much as 85%. Inasmuch as no cooling is required for this application, a
load of 150,000,000 BTU/hr normally imposed on the ponds is circumvented. The
gaseous fluorides are removed at elevated temperatures using a dry collector as the
reactor. Submicron particulates (i.e., NH4F • HF) are also agglomerated by the
chromatographic technique. Gaseous fluoride emissions are expected to be reduced
to 1 ppm and particulate to less than 0.01 gr/sdcf.
Table 9-8 shows the performance of existing wet scrubbing TESI equipment at
two plants and the projections for the dry chromatographic system (21). The cost of
dry chromatograph material (TESISORB) is $0.07/ton P2O5. The consumption of
TESISORB is estimated to be less than $50/day and the quantity of neutralized
product about a ton per day. The neutralized product can be mixed with the fertil-
izer product and discharged from the plant. The power consumption should be
reduced 20% because of reduction in liquid pumping.
Figure 9-6 shows the flow sheet for a proposed demonstration pilot dry flu-
oride removal system. It is recommended for use on emissions from a diammonium
phosphate plant. This way its effectiveness in the removal of fluoride and sub-
micron particulate can be demonstrated.
The gases from the process will be first introduced into a quench reactor
where they will be treated with a mixture of water and calcium hydroxide. Some
fluorides will be absorbed by the alkaline solution and gas will be cooled on its way
to the chromatographic separator. Gas will then be passed through a dry Venturi. A
dry Venturi exhibits high efficiency in removal of submicron particulate. TESISORB
particles are injected into a gas stream where they serve as the collection bodies for
the collection of particulate.
Although most of the particulate is in the submicron range, high efficiency is
achieved at a very low pressure drop. The reasons for this have been previously
presented in Section 5.1. The efficiency of particulate removal in the Venturi
throat is governed by the inertial impaction parameter. Since the diameter of
collection bodies (TESISORB particulate) is very small, a high efficiency can be
achieved. The final removal of TESISORB and particulate is accomplished in a
chromatographic separator. The solids from the chromatographic separator will be
removed from the separator. The solids can then be washed on a rotary filter, dried
and returned to the dry Venturi.
The major purpose of this demonstration project will be to verify applicability
of dry fluoride removal in the phosphate fertilizer industry. Experimental and
measurement techniques described in previous recommended projects can be also
applied in this demonstration. The design parameters for a full scale system will be
obtained in this project. The cost for the installation will be in the range of
$100,000 to $200,000 depending on the complexity of the auxiliary equipment. The
unit can be built in six months and testing carried out during the subsequent six
months.
-161-
-------
TABLE 9-7. PERFORMANCE OF CHROMATOGRAPHIC SYSTEMS (23)
Secondary Aluminum
Inlet, 0.03-0.35 gr/dscf
0.69-80 g/dncm
30 percent Oils
Hcl, 10-150 ppm
C12, 0-100 ppm
HF, 0-30 ppm
Container Glass
Inlet, 0.15-0.3 gr/dscf
0.34-0.69 g/dncm
SO3, 20-40 ppm
SO2, 100-180 ppm
Fiberglass
Inlet, 0.15-1.3 gr/dscf
0.34-2.98 g/dncm
HF, 100-150 ppm
SOx, 200-300 ppm
PARTICULATE
gr/dscf
0.00056
to 0.008
g/dncm
0.0013
0.0183
ACID GAS
PPm
0.5
Opacity
Percent
Zero
0.005-0.01
0.011-0.023
SO3< 1
SO2 10-60
Zero
0.001-0.008
0.002-0.018
HF<2
SOX
Zero
-162-
-------
TABLE 9-8. PERFORMANCE OF TESI EQUIPMENT
Acidulation Ib/ton P2O5
DAP Ib/ton P2O5
GTSP Ib/ton P2O5
EPA
0.02
0.06
0.2
Plant 1
1967
(550 TPD)
0.003
0.04
Plant 2*
1973
(1100 TPD)
0.02
0.04
Third***
Generation
TESI
0.002
0.010
0.020
GTSP Storage
Ib/ton P;O5 in storage
Pond
Fluoride emission
Ib/day
Total Ib/day F (stacks)
5 x 10"
30
27
2 x 10
30
45-60**
1 x 10"
15 (Gyp Only)
*No ammonia to pond
**Plant reports 35-50% of guaranteed emissions.
***Dry chromotographic technique.
-163-
-------
•c-
i
QUENCH
REACTOR
CHROMATOGRAPHIC
SEPARATOR
I
) ,
Ca(OH)2 SOLUTION
TESISORB
SYSTEM FAN
SOLIDS
BLOWER
SOLIDS
Figure 9-6: Pilot Dry Fluoride Removal System (25)
-------
9.6 Evaluation of the Kimre Mist Eliminator
In the past, one of the major problems with wet scrubbers in the phosphate
fertilizer industry has been maintenance. Packing and mist eliminators have a
tendency to plug and cleaning is a tedious and time consuming process. During
recent years many wet scrubbers have been retrofitted with Kimre mist eliminators.
These mist eliminators are made of various materials and utilize interlaced
monofilaments with essentially all of the filaments oriented perpendicular to the gas
flow. They are very stable, both dynamically and statically, since each filament
reinforces those closest to it. This makes these eliminators very easy to clean, and
is a major reason for their popularity.
The amount of liquid passing through the eliminator without being captured
may be expressed as:
Pt = exp. (-K,N)
Where Pt = penetration (1-efficiency); fractional amount passing
N = number of layers
K, = variable relating the geometry of the material, droplet size, and
gas velocity with drop removal effectiveness
KI must be determined experimentally. For one Kimre mist eliminator, the
value of KI for the removal of 3 micron droplets at a gas velocity of 5 feet per
second has been determined to be 0.452 (24). Using this value, N may be determined
from:
K
For an efficiency of 99.9%,
tn(l -0.999)
N = - 0.452 = 15'28'
so that 16 layers must be used.
The pressure drop through a mist eliminator may be expressed as:
Ap = K2 • V2« N,
where^p = pressure drop, includes H20
K2 = variable relating material geometry, liquid loading and pressure
drop
-165-
-------
V = gas velocity, feet per second
N = number of layers.
K2 for low liquid rates and air at standard atmospheric conditions is 1.71 x 10 .
The pressure drop for the eliminator in the example cited above is:
Ap = 1.71 • 10 • 52 • 16 = 0.68 in. H20
Although Kimre mist eliminators have been widely used in the phosphate fertilizer
industry, there is little information on their performance. According to the manu-
facturers these mist eliminators have performed substantially better than theory
would predict for the small droplets.
An evaluation of the Kimre mist eliminator is recommended in order to estab-
lish its performance and obtain optimum operating conditions. The evaluation can
be carried out on a prototype scrubber/mist eliminator shown in Figure 9-7. The
advantage of this unit is that it consists of multiple layers of different size with the
larger being in front to prevent plugging and insure easy cleaning. The main purpose
of the evaluation would be to obtain reliable performance data applicable to the
phosphate fertilizer industry.
The cost of the unit is unknown, but testing can be accomplished at a cost of
$30,000 within four months.
9.7 References
1. International Agency for Research on Cancer (IARC), Monograph on the
Evaluation of Carcinogenic Risk of Chemicals to Man, Vol. 13, Some
Miscellaneous Pharmaceutical Substances. Lyon, France, 1977, 255 pp.
2. Reconnaissance Study of Radiochemical Pollution from Phosphate Rock
Mining and Milling, U.S. Environmental Protection Agency, National
Field Investigations Center, Denver, Colorado, December 1973, 46 pp.
3. Monson, R.R., Effects on Health of Working Industry, Environmental
Law, In press, 1978.
H. Monson, R.R., Fine, L.3., Cancer Mortality and Morbidity Among Rubber
Workers, Journal National Cancer Institute, 61:1047, 1978.
5. Manson, T.S., McKay, F.W., U.S. Cancer Mortality by County 1950-1969.
U.S. Department of Health, Education and Welfare, Publication No. 74-
615, Washington, D.C., 1974.
-166-
-------
37
4 LAYERS
mil FILAMENT
94% VOID
4 LAYERS
37 mil FILAMENT
97% VOID
I
ON
I
-16
4 LAYERS
mil FILAMENT
96% VOID
3 LAYERS 4 LAYERS
mil FILAMENT 16 mil FILAMENT
94% VOID 96?= VOID
Figure 9-7: Four Stage Scrubber/Mist Eliminator (26)
-------
6. Hoover, R., Fraumeri, J.F., Jr. Cancer Mortality in U.S. Counties with
Chemical Industries, Environ. Research, 9:196, 1975.
7. Blot, W.J., et al, Cancer Mortality in U.S. Counties with Petroleum
Industries, Science, 198:51, 1977.
8. Jick, H., et al, Oral Contraceptives and Nonfatal Myocardial Infarctions,
Journal of American Medical Associations, 239:1403-1406, 1978.
9. Lehr, J.R., Fluorine Chemical Redistribution in Relation to Gypsum
Storage Pond System, The Fertilizer Institute Environmental Symposium-
1979, New Orleans, Louisiana, March 6-8, 1978, 16 pp.
10. Tatera, B.S., Parameters Which Influence Fluoride Emissions from
Gypsum Pond, Ph.D. Dissertation, University of Florida, 1970.
11. Measurement of Fluoride Emissions from Gypsum Ponds, TRC Project
82915 carried out for U.S. Environmental Protection Agency under
Contract No. 69-01-4145, Task 10, Division of Stationary Source
Enforcement, Washington, D.C., 1978.
12. Hill, W.3., Demler, W.R., More on Planning Experiments to Increase
Research Efficiency, Industrial and Engineering Chemistry, Vol. 62, No.
10, October 1970, 6 pp.
13. Connor, W.S., et al, Fractional Factorial Experiment Designs for Factors
at Three Levels, National Bureau of Standards, Washington, D.C., PB-
COM-73-11111, May 1959, 37 pp.
14. Davies, D.L., The Design and Analysis of Industrial Experiments, Hafner
Publishing Company, New York, NY, 1956.
15. Boscak, V., Koncar-Djurdjevic, S., Gas Cleaning Equipment Revisited
from an Industrial Hygiene Point of View, International Congress on
Chemical Engineering at the Service of Mankind, Paris, France, 1972.
16. Brown, J., Hatch, T., American Journal of Public Health, 40 A, 450,
1950.
17. Wilson, J., LaMer, V., Journal of Industrial Hygiene and Toxicology, 30,
265, 1948.
18. Strauss, W., Industrial Gas Cleaning, Pergamon Press Ltd., Oxford 1966.
19. Pilat, M.J., et al, Development of a Cascade Impactor System for Sam-
pling 0.02 to 20 Micron Diameter Particles, Electrical Power Research
Institute, EPRI FP-844, Volume 1, Palo Alto, California
20. Compilation of Air Pollutant Emission Factors, Second Edition, U S
Environmental Protection Agency, Office Of Air Quality Planning and
Standards, Research Triangle Park, NC, March 1975/392 pp.
-168-
-------
21. Teller, A.J., New Technologies in Fertilizer Emission Control, Presented
at the Central Florida Section AIChE Meeting, Daytona, Florida,
May 14-16, 1976, 11 pp.
22. Gerstein, S.M., Franza, M.E., Control Technology for Secondary
Aluminum Smelters, Presented at the 68th APCA Meeting, Boston,
Massachusetts, June 15-20, 1975, U pp.
23. Teller, A.3., New Systems for Municipal Incinerator Emission Control,
Proceedings of 1978 National Waste Processing Conference-Energy Con-
servation through Waste Utilization, ASME Solid Waste Processing
Division, pp. 179-185.
24. Kimre Bulletin on Mist Eliminator and Tower Packing, 976-SM-1P, 1976.
25. Private Communication, Dr. A.J. Teller, Teller Environmental Systems, Inc.
to Dr. V.G. Boscak, TRC, Feburary 1979.
26. Drawing TI27, Kimre, Inc., Perrine, Florida.
-169-
-------
10.0 BIBLIOGRAPHY
This section lists major sources used in this report. For clarity and conveni-
ence, the bibliography is organized into five subsections:
1) Process Description
2) Air Emissions and Controls
3) Wastewater Sources and Treatment
4) Solid Waste
5) Miscellaneous
10.1 Process Description
Air Pollution Control Technology and Costs in Seven Selected Areas, EPA-
450/3-75-010, U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, March 1973, pp. 11-192.
Atmospheric Emissions from Wet-Process Phosphoric Acid Manufacture, AP-
57 (PB 192 222), U.S. Department of Health, Education, and Welfare, Raleigh, North
Carolina, April 1970, 86 pp.
Final Guideline Document: Control of Fluoride Emissions from Existing
Phosphate Fertilizer Plants, EPA-450/2-77-005, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, March 1977.
Nyers, J.M., et al, Source Assessment: Phosphate Fertilizer Industry, EPA-
600/2-78-004, U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, May 1978, 184 pp.
Rawlings, G.D., Reznik, R.B., Source Assessment: Fertilizer Mixing Plants,
EPA 600/2-76-032c, U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, March 1976, 187 pp.
Riegel's Handbook of Industrial Chemistry, Seventh Edition, T.A. Kent (ed).
Van Nostrand Reinhold Co., New York, NY, 1974, pp. 551-569.
Robinson, J.M., et al, Engineering and Cost Effectiveness Study of Fluoride
Emissions Control, PB 207-506, prepared by TRW Systems Group for U.S. Environ-
mental Protection Agency, January 1972.
Scott, W.C., et al, Status of Modern Wet-Process Phosphoric Acid Technology,
Process Engineering Branch, Division of Chemical Development, Tennessee Valley
Authority.
Shreve, R.N., Chemical Process Industries, Third Edition, McGraw-Hill Book
Co., New York, NY, 1967, pp. 274-277.
-170-
-------
Weber, W.C., Pratt, C.3., Wet Process Phosphoric Acid Manufacture, In:
Chemistry and Technology of Fertilizers, Sauchelli, V. (ed). Reinhold Publishing
Corporation, New York, NY, 1960, p. 224.
10.2 Air Emissions and Controls
Air Pollution Control Technology and Costs in Seven Selected Areas, EPA-
450/3-73-010, U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, December 1973.
Atmospheric Emissions from Wet-Process Phosphoric Acid Manufacture,
National Air Pollution Control Administration, Raleigh, North Carolina, Publication
Number AP-52. April 1970, p. 25-26.
Background Information for Standards of Performance: Phosphate Fertilizer
Industry, Volume 1: Proposed Standards, EPA-450/2-74-019A, U.S. Environmental
Protection Agency, Research Triangle Park, October 1974, 119 pp.
Balazova, C., et al, Effect of Atmospheric Fluorine Pollution on the Living
Organism, Nutr., Proc. Int. Congr., 8th, 1969.
Bennet, F.W., Spall, B.C., Review of Effluent Problems in Fertilizer
Manufacture, In: Symp. Fertilizers and the Environment, Pert. Soc., Proc. No. 156,
pp 5-54.
Calvert, S.J., et al, Scrubber Handbook, EPA-R2-72-118-a, PB 213-061, 1972.
Final Guideline Document: Control of Fluorides Emissions from Existing
Phosphate Fertilizer Plants, EPA 450/2-77-005, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, March 1977.
Inspection Manual for Enforcement of New Source Performance Standards:
Phosphate Fertilizer Plants, EPA 340/1-77-009, U.S. Environmental Protection
Agency, Washington, D.C., February 1977, 80 pp.
Lunde, K.E., Performance of Equipment for Control of Fluoride Emissions, Ind.
Eng. Chem., Vol. 50, No. 3, March 1958, p. 293-298.
The Mcllvaine Scrubber Manual, The Mcllvaine Co., Northbrook, Illinois, 1974.
Osag, T.R., Fluoride Emission Control Cost, Chem. Eng. Prog., Vol. 12, No. 2,
December 1976, pp. 3-6.
Partridge, 3.E., et al, Population Radiation Dose Estimates from Phosphate
Industry Air Particulate Emissions, Technical Note, ORP/EERF-77-1.
Semrau, K.T., Emission of Fluorides from Industrial Processes, A review,
APCA Journal, Vol. 7, No. 29, August 1957, pp. 92-108.
-171-
-------
Technical Report: Phosphate Fertilizer Industry. In: Group III Background
Document, U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina.
Technical Report: Phosphate Fertilizer Industry. In: An Investigation of the
Best Systems of Emission Reduction for Six Phosphate Fertilizer Processes, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina, April
1974, p. 25.
Teller, A.3., Scrubbers in the Fertilizer Industry; Their Success, Near Future
and Eventual Replacement, Presented at Fertilizer Round Table, Washington D.C.,
November 1973.
World Guide to Pollution Control in the Fertilizer Industry, British Sulphur
Corp. Ltd., London, 1975, 125 pp.
10.3 Wastewater Sources and Treatment
Arora, H.C., Chattopadhya, S.N., A Study on the Effluent Disposal of
Superphosphate Fertilizer Factory, Indian 3. Environmental Health, Vol. 16, No. 2,
April 1974, pp. 140-50.
Douglas, L.A., the Potential Contribution of Fertilizers to Water Pollution,
Water Resources Research Institute, Rutgers University, New Brunswick, NJ, June
1976.
Evaluation of Emissions and Control Techniques for Reducing Fluoride
Emissions from Gypsum Ponds in the Phosphoric Acid Industry, EPA-600/2-78-124,
U.S. Environmental Protection Agency, Research Triangle Park, North Carolina,
June 1978, 218 pp.
Gupta, B.S., Behaviour of Some Phosphatic Fertilizers in Water, J. Indian Soc.
Soil Sci., Vol. 21, No. 4, 1973, pp. 413-420.
Hartig, R.G., Prevention of Surface Water Contamination and Air Pollution by
Fluorine Compounds from Phosphate Plants, U.S. Patent 3,642,438, February 15,
1972.
Inorganic Fertilizer and Phosphate Mining Industries Water Pollution and
Control, Water Pollution Control, Research Series No. 12020 FPD, September 1971.
Jones, W.E., Olmsted, R.L., Waste Disposal at a Phosphoric Acid and
Ammonium Phosphate Fertilizer Plant, Purdue Univ., Eng. Bull., Ext. Ser. No. 112,
1962, pp. 198-202.
Kaufman, R.F., Bliss, J.D., Effects of the Phosphate Industry on Radium -226
in Ground Water of Central Florida, 1977.
-172-
-------
Kawahara, K., et al, Treatment of Waste Water Containing Fluoride and
Alkali, Japan, Kokai 75, 15,355, February 18, 1975.
Khripunov, N.F., et al, Ways of Lowering the Amount of Fluorine-Containing
Waste Water in the Production of Wet-Process Phosphoric Acid, Khim. Prom-st
(Moscow), Vol. 2, 1975, pp. 110-12.
King, W.R., Farrell, CJ.K., Fluoride Emissions from Phosphoric Acid Gypsum
Ponds, EPA/650 2-74-095, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, October 1974.
Lehr, J.R., Fluorine Chemical Redistribution in Relation to Gypsum Storage
Pond System, The Fertilizer Institute Environmental Symposium - 1979, New
Orleans, Lousiana, March 6-8, 178, 16 pp.
Martin, E.E., Development Document for Effluent Limitations Guidelines and
New Source Performance Standards for the Basic Fertilizer Chemicals Segment of
the Fertilizer Manufacturing Point Source Category. EPA 440/1-74-011-a,
(PB 238652), U.S. Environmental Protection Agency, Washington, D.C., March 1974,
170 pp.
Reconnaissance Study of Radiochemical Pollution from Phosphate Rock Mining
and Milling, Office of Enforcement, U.S. Environmental Protection Agency, Denver,
Colorado, December 1973, 46 pp.
Tatera, B.S., Parameters which Influence Fluoride Emissions from Phosphoric
Acid Gypsum Pond, Ph.D. Dissertation, University of Florida, 1970.
Teller, A.3., New Technologies in Control of Fertilizer Plant Emissions; Pond
Control - Fluoride Products, Presented at the Fertilizer Round Table, November
1975, 16 pp.
10.4 Solid Waste
Lutz, W.A., Pratt, C.J., Principles of Design and Operation. In: Phosphoric
Acid, Volume I, Slack, A.V., (ed.), Marcel Dekker, Inc., New York, NY, 1968, pp.
158-208.
Martin, E.E., Development Document for Effluent Limitations Guidelines and
New Source Performance Standards for the Basic Fertilizer Chemicals Segment of
the Fertilizer Manufacturing Point Source Category, EPA 440/1-74-011-e, (PB-238
652), U.S. Environmental Protection Agency, Washington, D.C., March 1974, 170 pp.
Nyers, J.M., et al, Source Assessment: Phosphate Fertilizer Industry, EPA
600/2-78-004, U.S. Environmental Protection Agency, Research Triangle park, North
Carolina, May 1978, 185 pp.
-173-
-------
Reconnaissance Study of Radiochemical Pollution from Phosphate Rock Mining
and Milling, U.S. Environmental Protection Agency, Office of Enforcement, Denver,
Colorado, December 1973, 45 pp.
Wissa, A.E.Z., Gypsum Stacking Technology, Presented at 1977 Annual
Technical Meeting, Central Florida and Peninsular Florida Sections, American
Institute of Chemical Engineers, Clearwater, Florida, May 1977, 44 pp.
10.5 Miscellaneous
10.5.1 Marketing
Fertilizer Situation, Econ. Res. Serv., USDA, Report FS-7, January 1977, 26
pp.
Fertilizer Supply, Exports to be Monitored, Chemical Marketing Report, Vol.
204, No. 22, November 26, 1973, pp. 3, 19.
King Lee Wang, et al, Economic Significance of the Florida Phosphate
Industry, Economic Analysis, Mineral Supply, Washington D.C., U.S. Bureau of
Mines, I.C. 8653, 1974, 51 pp.
Kronseder, J.G., Economics of Phosphoric Acid Process, Chem. Eng. Prog.,
Vol. 64, No. 9, September 1968, pp. 97-102.
Trends in U.S. Fertilizer Consumption, Phosphorus Potassium, No. 81, January-
February 1976, pp. 27-34.
World Fertilizer Position Forecast, Chem. Mark. Rep., Vol. 209, No. 29, June
14, 1976, pp. 5, 37.
World Fertilizer Atlas, Fifth Ed., The British Sulfur Corp. Ltd., London, 1976,
108 pp.
10.5.2 Energy Consumption
Environmental Considerations of Selected Energy Conserving Manufacturing
Process Options, Vol. XV. Fertilizer Industry Report, EPA 600/7-76-0340, PB-264
281, U.S. Environmental Protection Agency, Industrial Environmental Research
Laboratory, Cincinnati, Ohio, December 1976, 59 pp.
Haycocks, C., Impact of Rising Energy Cost on the Domestic Production of
Selected Commodities, U.S. Bureau of Mines, OFR 88-76, May 1976, 124 pp.
Sherff, J.L., Energy Use and Economics in the Manufacture of Fertilizer. In:
Energy, Agriculture and Waste Management, Ann Arbor Science Publishers, 1975,
pp. 433-41.
-174-
-------
10.5.3 By-product Recovery
Aluminum Fluoride Made from Fluosilicic Acid, Chemical and Process
Engineering, Vol. 53, No. 11, June 1972.
Blake, H.E., et al, Utilization of Waste Fluosilicic Acid, 1. Laboratory
Investigations, 2. Cost Evaluation, U.S. Bureau of Mines, RI-7502, 1971, 60 pp.
Ellis, D.A., Recovery of Uranium from Industrial Phosphoric Acids By Solvent
Extraction, I. Summary Status Report, II Index of Monthly Progress Reports, U.S.
Atomic Energy Commission, DOW-81, 1952, 81 pp.
Greek, B.F., et al, Uranium Recovery from Wet-Process Phosphoric Acid, Ind.
Eng. Chem., Vol. 49, 1957, pp 628-38.
Hazen, W.C., et al, Production of Phosphates from Phosphates Slimes, U.S.
Patent 3,425,799, February 4, 1969.
Johnson, R.C., et al, Economic Availability of Byproduct Fluorine in the
United States, I. Utilization of Byproduct Fluosilicic Acid in the Manufacture of
Aluminum Fluoride. II. Utilization of Byproduct Fluosilicic Acid in the Manufacture
of Calcium Fluoride, U.S. Bureau of Mines, 1C 8566, 1973, 97 pp.
Ring, R.J., Manufacture of Phosphatic Fertilizer and Recovery of Byproduct
Uranium - A Review, (Aust. At. Energy Comm. Res. Establ., Luces Heights,
Australia). Report AAEC-E-355, 1975, 99 pp.
Yamaguchi, T., et. al., Improvement of Quality of Byproduct Gypsum, Japan,
Kokai 75 120, 491, September 20, 1975.
-175-
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-79-169
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Evaluation of Control Technology for the Phosphate
Fertilizer Industry
6. REPORT DATE
1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Vladimir G. Boscak
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING OROANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
TRC--The Research Corporation of New England
125 Silas Deane Highway
Wethersfield, Connecticut 06109
1AB604B
11. CONTRACT/GRANT NO.
68-02-2615, Task 10
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 7/78 - 6/79
14. SPONSORING AGENCY CODE
EPA/600/13
16. SUPPLEMENTARY NOTES JERL-RTP project officer is Ronald A. Venezia, Mail Drop 62,
919/541-2547.
16. ABSTRACT The repOrj. gives results of a. phosphate fertilizer industry study to evaluate
multimedia control technology, identify information gaps, and define needed RDandD
projects. The following manufacturing processes were covered: wet process phos-
phoric acid, superphosphoric acid, diammonium phosphate, and normal and triple
superphosphate. Air emission control technology, based largely on using wet scrub-
bers , is adequate for control of fluoride and particulate and is used throughout the
industry. The cross-flow packed scrubber appears to be the best and applies to all
processes. The gypsum pond appears to be the major environmental concern: main
problems are fluoride emissions to the atmosphere and possible leaching of fluoride,
phosphate, and radioactive substances.None of the several control alternatives are
economically attractive. RDandD projects identified and defined in this study are:
(1) an epidemiologic study of the industry; (2) studies of gypsum pond emissions and
chemistry; (3) an evaluation and optimization of wet scrubbers; (4) a study of the
atmospheric interaction and dispersion modeling of ammonia/sulfuric acid mist;
(5) a demonstration of a dry system for fluoride removal; and (6) an evaluation of the
Kim re mist eliminator.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. cos AT I Field/Group
Pollution
Fertilizers
Phosphoric Acids
Fluorides
Dust
Scrubbers
Radioactive Wastes
Pollution Control
Stationary Sources
Phosphate Fertilizers
Diammonium Phosphate
Particulate
Superphosphates
13B
02A
07B
11G
07A,13I
18G
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
186
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 («-73)
-176-
------- |