DRAFT GUIDELINE DOCUMENT:
CONTROL OF FLUORIDE EMISSIONS
FROM EXISTING
PHOSPHATE FERTILIZER PLANTS
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
DRAFT GUIDELINE DOCUMENT:
CONTROL OF FLUORIDE EMISSIONS
FROM EXISTING
PHOSPHATE FERTILIZER PLANTS
NOTICE
This draft guideline is now being published for'
comment. A final guideline will be published after
consideration of these comments.
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Emission Standards and Engineering Division
Research Triangle Park, North Carolina 27711
Telephone: (919) 688-8146
April 1976
-------
CONTENTS
1. INTRODUCTION AND SUMMARY 1-1
1.1 INTRODUCTION 1-1
1.2 HEALTH EFFECTS OF FLUORIDES 1-5
1.3 FLUORIDES AND THEIR CONTROL 1-6
1.4 EMISSION GUIDELINES 1-7
1.5 COMPLIANCE TIMES 1-10
1.6 ASSESSMENTS 1-11
1.6.1 Economic 1-11
1.6.2 Environmental 1-15
1.6.3 Energy 1-16
1.7 REFERENCES 1-17
2. HEALTH AND WELFARE EFFECTS OF FLUORIDES 2-1
2.1 INTRODUCTION 2-1
2.2 EFFECT OF FLUORIDES ON HUMAN HEALTH 2-3
2.2.1 Atmospheric Fluorides 2-3
2.2.2 Ingested Fluorides 2-3
2.3 EFFECT OF FLUORIDES ON ANIMALS 2-5
2.4 EFFECT OF ATMOSPHERIC FLUORIDES ON 2-6
VEGETATION
2.5 EFFECT OF ATMOSPHERIC FLUORIDES ON 2-7
MATERIALS OF CONSTRUCTION
2.5.1 Etching of Glass 2-7
2.5.2 Effects of Fluorides on Structures 2-9
2.6 RATIONALE 2-10
2.7 REFERENCES 2-10
-------
Page
3. PHOSPHATE FERTILIZER INDUSTRY ECONOMIC PROFILE AND 3-1
STATISTICS
3.1 INDUSTRY STRUCTURE 3-1
3.2 EXISTING PLANTS 3-4
3.3 CAPACITY UTILIZATION 3-18
3.4 CONSUMPTION PATTERNS 3-20
3.5 FUTURE TRENDS 3-25
3.6 PRICES 3-29
3.7 WORLD STATISTICS ON P00C 3-33
i b
3.8 REFERENCES 3-36
4. PHOSPHATE FERTILIZER PROCESSES 4-1
4.1 INTRODUCTION 4-1
4.2 WET-PROCESS PHOSPHORIC ACID MANUFACTURE 4-3
4.3 SUPERPHOSPHORIC ACID MANUFACTURE 4-11
4.4 DIAMMONIUM PHOSPHATE MANUFACTURE 4-17
4.5 TRIPLE SUPERPHOSPHATE MANUFACTURE AND STORAGE 4-21
4.5.1 Run-of-Pile Triple Superphosphate 4-21
Manufacture and Storage
4.5.2 Granular Triple Superphosphate 4-24
Manufacture and Storage
4.6 REFERENCES 4-29
5. EMISSIONS 5-1
5.1 NATURE OF EMISSIONS 5-1
5.2 UNCONTROLLED FLUORIDE EMISSIONS 5-3
5.2.1 Emissions from Wet-Process Phosphoric 5-3
Acid Manufacture
5.2.2 Emissions from Superphosphoric Acid 5-7
Manufacture
ii
-------
Page
5.2.3 Emissions from Diammonium Phosphate 5-8
Manufacture
5.2.4 Emissions from Triple Superphosphate 5-10
Manufacture and Storage
5.3 TYPICAL CONTROLLED FLUORIDE EMISSIONS 5-12
5.3.1 Emissions from Wet-Process Phosphoric 5-12
Acid Plants
5.3.2 Emissions from Superphosphoric Acid 5-13
Manufacture
5.3.3 Emissions from Diammonium Phosphate 5-13
Manufacture
5.3.4 Emissions from Triple Superphosphate 5-13
Manufacture and Storage
5.4 GYPSUM POND EMISSIONS 5-15
5.5 REFERENCES 5-18
6. CONTROL TECHNIQUES FOR FLUORIDES FROM PHOSPHATE 6-1
FERTILIZER PROCESSES
6.1 SPRAY-CROSSFLOW PACKED BED SCRUBBER 6-1
6.1.1 Description 6-1
6.1.2 Emission Reduction 6-5
6.1.3 Retrofit Costs for Spray-Crossflow 6-7
Packed Bed Scrubbers
6.2 VENTURI SCRUBBER 6-71
6.2.1 Description 6-71
6.2.2 Emission Reduction 6-74
6.2.3 Retrofit Costs for Venturi Scrubbers 6-75
6.3 SPRAY TOWER SCRUBBER 6-78
6.3.1 Description 6-78
6.3.2 Emission Reduction 6-78
m
-------
Page
6.3.3 Retrofit Costs for Cyclonic Spray Towers 6-79
6.4 IMPINGEMENT SCRUBBER 6-86
6.5 SUMMARY OF CONTROL OPTIONS 6-87
6.6 DESIGN, INSTALLATION, AND STARTUP TIMES 6-88
6.7 REFERENCES 6-95
7. ECONOMIC IMPACT 7-1
7.1 INTRODUCTION 7-1
7.2 IMPACT ON MODEL PLANTS 7-2
7.3 CRITERIA FOR PLANT CLOSURES 7-4
7.4 IMPACT ON THE INDUSTRY 7-6
7.5 IMPACT ON EMPLOYMENT AND COMMUNITIES 7-10
7.6 SUMMARY 7-10
7.7 REFERENCES FOR 7-12
8. B1ISSION GUIDELINES FOR EXISTING 8-1
PHOSPHATE FERTILIZER PLANTS
8.1 GENERAL RATIONALE 8-1
8.2 EVALUATION OF INDIVIDUAL EMISSION GUIDELINES 8-4
8.2.1 Wet-Process Phosphoric Acid Plants 8-4
8.2.2 Superphosphoric Acid Plants 8-6
8.2.3 Diammonium Phosphate Plants 8-7
8.2.4 Run-of-Pile Triple Superphosphate Production 8-8
and Storage Facilities
8.2.5 Granular Triple Superphosphate Production 8-9
Facilities
8.2.6 Granular Triple Superphosphate Storage 8-11
Facilities
8.3 REFERENCES 8-13
iv
-------
9. ENVIRONMENTAL ASSESSMENT 9-1
9.1 ENVIRONMENTAL ASSESSMENT OF THE EMISSION 9-1
GUIDELINES
9.1.1 Air 9-1
9.1.2 Water Pollution 9-9
9.1.3 Solid Waste Disposal 9-12
9.1.4 Energy 9-13
9.1.5 Other Environmental Concerns 9-18
9.2 ENVIRONMENTAL ASSESSMENT OF ALTERNATIVE 9-18
EMISSION CONTROL SYSTEMS
9.3 SOCIO-ECONOMIC EFFECTS 9-19
9.4' REFERENCES 9-20
-------
LIST OF FIGURES
Figure Page
3-1 Wet-Process and Superphosphoric Acid Plant Locations 3-16
3-2 Triple Superphosphate and Ammonium Phosphate Plant 3-17
Locati ons
3-3 Capacity Utilization of Wet-Process Phosphoric Acid 3-21
3-4 Capacity Utilization of Ammonium Phosphates 3-22
3-5 Wholesale Prices for Triple Superphosphate and 3-31
Diammonium Phosphate
4-1 Major Phosphate Rock Processing Steps 4-2
4-2 Flow Diagram Illustrating a Wet-Process Phosphoric 4-5
Acid Plant
4-3 Flow Diagram for Prayon Reactor 4-6
4-4 Operating Cycle of Rotary Horizontal Tilting 4.9
Pan Filter
4-5 TVA Evaporator for Producing Superphosphoric Acid 4-13
4-6 Submerged Combustion Process for Producing Super- 4-14
phosphoric Acid
4-7 Stauffer Evaporator Process 4-16
4-8 Swenson Evaporator Process 4-16
4-9 TVA Diammonium Phosphate Process 4-19
4-10 Run-of-Pile Triple Superphosphate Production and 4-22
Storage
4-11 TVA Cone Mixer 4-23
4-12 TVA One-Step Process for Granular Triple Super- 4-25
phosphate
4-13 Dorr-Oliver Slurry Granulation Process for Triple 4-27
Superphosphate
4-14 Granular Triple Superphosphate Storage 4-28
vi
-------
Fj gure Page
6-1 Spray-Crossflow Packed Bed Scrubber 6-2
6-2 Manufacture of Wet-Process Phosphoric Acid 6-15
6-3 Existing Control Equipment Layout for Model WPPA Plant 6-17
6-4 Retrofit Control Equipment Layout for Model WPPA Plant 6-20
6-5 Retrofit Control Equipment Layout for Model SPA Plant 6-31
6-6 Existing Control Equipment Layout for Model DAP Plant 6-38
6-7 Retrofit Control Equipment Layout for Model DAP Plant 6-39
6-8 Existing Control Equipment Layout for Model ROP-TSP 6-47
Plant, Case A
6-9 Retrofit Control Equipment Layout for Model ROP-TSP 6-51
Plant, Case B
6-10 Existing Control Equipment Layout for Model GTSP 6-58
Plant
6-11 Retrofit Control Equipment Layout for Model GTSP 6-59
Plant
6-12 Gas Actuated Venturi Scrubber with Cyclonic Mist 6-73
Eliminator
6-13 Water Actuated Venturi 6-73
6-14 Cyclonic Spray Tower Scrubber 6-79
6-15 Retrofit Control Equipment Layout for Model ROP-TSP Plant 6-82
6-16 Doyle Scrubber 6-86
6-17 Time Schedule for the Installation of a Wet Scrubber on 6-89
a Wet-Process Phosphoric Acid Plant
vii
-------
LIST OF TABLES
Table Page
1-1 Fluoride Emission Guidelines for 1-8
Existing Phosphate Fertilizer Manufacturing Plants
1-2 Performance of Aqueous Scrubber Emission Control 1-9
Equipment in Phosphate Fertilizer Manufacturing Plants
1-3 Increments of Progress for Installation of Wet 1-10
Scrubber for a Wet Process Phosphoric Acid Plant
1-4 Economic Impact of Fluoride Emission Guidelines for 1-12
Existing Phosphate Fertilizer Manufacturing Facilities
1-5 Summary of Retrofit Control Cost Requirements for 1-13
Various Phosphate Fertilizer Manufacturing Processes
2-1 Examples of HF Concentrations (PPB) and Exposure 2-8
Durations Reported to Cause Leaf Damage and Poten-
tial Reduction in Crop Values
3-1 Ten Largest Phosphate Rock Producers 3-2
3-2 Ten Largest Phosphoric Acid Producers 3-3
3-3 Production Capacity of Wet-Process Phosphoric 3-5
Acid (1973)
3-4 Production Capacity of Superphosphoric Acid (1973) 3-8
3-5 Production Capacity of Triple Superphosphate (1973) 3-10
3-6 Production Capacity of Ammonium Phosphates (1973) 3-12
3-7 Production as Percent of Capacity 3-19
3-8 U.S. Phosphate Consumption, 1960-1973 (1000 tons 3-24
3-9 U.S. Production of Three Commodities in the Phosphate 3-26
Industry, 1950-1973
3-10 Summary of List Prices as of July 1974 and Basis 3-32
for Quotation
3-11 United States and World Consumption of Phosphate 3-34
Fertilizer
3-12 World Reserves of Phosphate Rock and Apatite 3-35
vi i i
-------
Table Page
4-1 P205 Content of Phosphate Fertilizers 4-3
4-2 Components of Typical Wet-Process Acid 4-10
4-3 Comparison of Orthophosphoric to Superphosphoric 4-11
Acid
5-1 Fluoride Emissions from an Uncontrolled Wet-Process 5-4
Phosphoric Acid Plant
5-2 Typical Material Balance of Fluoride in Manufacture 5-6
of Wet-Process Phosphoric Acid
5-3 Fluoride Emission Factors for Selected Gypsum Ponds 5-17
at 90°F; Lbs/Acre Day
6-1 Calculated Equilibrium Concentrations of Fluorine in 6-5
the Vapor Phase Over Aqueous Solutions of Fluosilicic
Acid
6-2 Scrubber Performance in Wet-Process Phosphoric Acid 6-6
Plants
6-3 Spray-Crossflow Packed Bed Scrubber Performance in 6-8
Diammonium Phosphate and Granular Triple Super-
phosphate Plants
6-4 Installed Cost Indices 6-10
6-5 Flow Rates and Fluoride Concentrations of WPPA Plant 6-18
Effluent Streams Sent to Existing Controls (Case A)
6-6 Flow Rates and Fluoride Concentrations of WPPA Plant 6-19
Effluent Streams sent to Retrofitted Controls (Case A)
6-7 Pond Water Specifications 6-21
6-8 Major Retrofit Items for Model WPPA Plant (Case A) 6-22
6-9 Operating Conditions for Spray-Crossflow Packed 6-23
Bed Scrubber for Model WPPA Plant, Case A (500
Tons/Day P205)
6-10 Retrofit Costs for Model WPPA Plant, Case A (500 6-24
Tons/Day P20g)
6-11 Flow Rates and Fluoride Concentrations of WPPA Plant 6-25
Effluent Streams Sent to Existing Controls (Case B)
IX
-------
Table
Page
6-12 Flow Rates and Fluoride Concentrations of WPPA 6-26
Plant Effluent Streams Sent to Retrofitted Controls
(Case B)
6-13 Major Retrofit Items for WPPA Plant (Case B) 6-26
6-14 Operating Conditions for Spray-Crossflow Packed Bed 6-27
Scrubber for Model WPPA Plant, Case B
(500 Tons/Day PgOg)
6-15 Retrofit Costs for Model WPPA Plant, Case B 6-28
(500 Tons/Day P205)
6-16 Major Retrofit Items for Model SPA Plant 6-32
6-17 Operating Conditions for Spray-Crossflow Packed Bed 6-33
Scrubber for Model SPA Plant (300 Tons/Day P205)
6-18 Retrofit Costs for Model SPA Plant (300 Tons/Day 6-34
P2°5>
6-19 Flow Rates and Fluoride Concentrations for DAP Plant 6-36
Emission Sources
6-20 Major Retrofit Items for Model DAP Plant 6-40
6-21 Operating Conditions for Spray-Crossflow Packed 6-41
Bed Scrubbers for Model DAP Plant (500 Tons/Day
W
6-22 Retrofit Costs for Model DAP Plant (500 Tons/Day 6-42
P2°5>
6-23 Flow Rates and Fluoride Concentrations for ROP-TSP 6-44
Plant Emission Sources
6-24 Major Retrofit Items for Model ROP-TSP Plant 6-45
(Case A)
6-25 Operating Conditions for Spray-Crossflow Packed Bed 6-46
Scrubber for Model ROP-TSP Plant, Case A (550 Tons/
Day P205)
6-26 Retrofit Costs for Model ROP-TSP Plant, Case A (550 6-48
Tons/Day P20g)
6-27 Flow Rates and Fluoride Concentrations of Effluent 6-49
Streams Sent to Existing Controls
-------
Table Page
6-28 Major Retrofit Items for Model ROP-TSP Plant (Case B) 6-52
6-29 Operating Conditions for Spray-Crossflow Packed Bed 6-53
Scrubber for Model ROP-TSP Plant, Case B (550 Tons/
Day P205)
6-30 Retrofit Costs for Model ROP-TSP Plant, Case B 6-54
(550 Tons/Day P205)
6-31 Flow Rates and Fluoride Concentrations for GTSP 6-57
Plant Emission Sources
6-32 Major Retrofit Items for Model CTSP Plant 6-61
6-33 Operating Conditions for Spray-Crossflow Packed 6-64
Bed Scrubbers for Model GTSP Plant (400 Tons/Day
P2°5>
6-34 Retrofit Costs for Model GTSP Plant (400 Tons/Day 6-65
P2°5'
6-35 Operating Characteristics of Scrubbers in Retrofit 6-68
Case A
6-36 Case B Retrofit Project Costs 6-70
6-37 Venturi Scrubber Performance in Superphosphoric 6-74
Acid and Diammonium Phosphate Plants
6-38 Major Retrofit Items for Model DAP Plant 6-75
6-39 Retrofit Costs for Model DAP Plant (500 Tons/Day 6-77
P2°5>
6-40 Cyclonic Spray Tower Performance in Wet-Process 6-80
Phosphoric Acid, Diammonium Phosphate, and Run-of-
Pile Triple Superphosphate Plants
6-41 Major Retrofit Items for Model ROP-TSP Plant 6-83
6-42 Operating Conditions for Cyclonic Spray Tower 6-84
Scrubbers for Model ROP-TSP Plant (550 Tons/Day
P2°5>
6-43 Retrofit Cost for Model ROP-TSP Plant (550 Tons/Day 6-85
P205>
XI
-------
Table Page
6-44 Estimated Total Capital Investment and Annualized 6-87
Cost for DAP and ROP-TSP Retrofit Models using
Spray-Crossflow Packed Bed and Alternative Scrubbers
6-45 Description of Individual Activities Involved in the 6-90
Procurement, Installation, and Startup of Control
Equipment
7-1 Summary of Retrofit Control Cost Requirements for 7-2
Various Phosphate Fertilizer Manufacturing Processes
9-1 Annual U.S. Fluoride Emission Reduction Due to Instal- 9-2
lation of Retrofit Controls Capable of Meeting
Emission Guidelines
9-2 Typical 1974 Fluoride Emissions Source Strengths Be-' 9-3
fore and After Installation of Retrofit Controls
Capable of Meeting Emission Guidelines
9-3 Existing Controls and Emissions for Model Phosphate 9-5
Fertilizer Complex
9-4 Retrofit Controls and Emissions for Model Phosphate 9-6
Fertilizer Complex
9-5 Estimated 30-Day Average Ambient Fluoride Concentra- 9-8
tions Downwind of a Phosphate Fertilizer Complex
9-6 Comparison of Emission Guidelines and an Alternative 9-10
Standard
9-7 EPA Effluent Limitations for Gypsum Pond Water 9-11
9-8 Incremental Power Requirements for Fluoride Control 9-14
Due to Installation of Retrofit Controls to Meet
Emission Guidelines
9-9 Increase in Phosphate Industry Energy Requirements 9-16
Resulting from Installation of Retrofit Controls
to meet Emission Guidelines
9-10 Increased Electrical Energy Demand by the Phosphate 9-17
Industry as a Result of Installation of Retrofit
Controls
xn
-------
1. INTRODUCTION AND SUMMARY
1.1 INTRODUCTION
Section lll(d) of the Clean Air Act, 42 U.S.C. 1857c-6(d), as
amended, requires EPA to establish procedures under which States submit
plans to control certain existing sources of certain pollutants. On
November 17, 1975 (40 FR 53340), EPA implemented section lll(d) by
promulgating Subpart B of 40 CFR Part 60, establishing procedures and
requirements for adoption and submittal of State plans for control of
"designated pollutants" from "designated facilities." Designated
pollutants are pollutants which are not included on a list published
under section 108(a) of the Act (national Ambient Air Quality Standards)
or section 112(b)(l)(A) (Hazardous Air Pollutants), but for which
standards of performance for new sources have been established under
section lll(b). A designated facility is an existing facility which
emits a designated pollutant and which would be subject to a standard
of performance for that pollutant if the existing facility were new.
Standards of performance for five categories of new sources in
the phosphate fertilizer industry were promulgated in the FEDERAL
REGISTER (40 FR 33152) on August 6, 1975, to be incorporated into the
Code of Federal Regulations under 40 CFR Part 60. New subparts T, U,
V, W, and X were added to set standards of performance for fluoride
emissions from new plants manufacturing wet-process phosphoric acid
(WPPA), superphosphoric acid (SPA), diammonium phosphate (DAP),
triple superphosphate (TSP), and for storage facilities used in the
manufacture of granular triple superphosphate (GTSP). The States,
therefore, are required to adopt fluoride emission standards for
1-1
-------
existing phosphate fertilizer plants which would be subject to the
standard of performance if they were new.
Subpart B of 40 CFR Part 60 provides that EPA will publish a
guideline document for development of State emission standards after
promulgation of any standard of performance for a designated
pollutant. The document will specify emission guidelines and times
for compliance and will include other pertinent information, such as
discussion of the pollutant's effects on public health and welfare
and a description of control techniques and their effectiveness and
costs. The emission guidelines will reflect the degree of emission
reduction attainable with the best adequately demonstrated systems of
emission reduction, considering costs as applied to existing facilities.
After publication of a final guideline document for the pollutant
in question, the States will have nine months to develop and submit
plans for control of that pollutant from designated facilities. Within
four months after the date for submission of plans, the Administrator
will approve or disapprove each plan (or portions thereof). If a
state plan (or portion thereof) is disapproved, the Administrator will
promulgate a plan (or portion thereof) within six months after the
date for plan submission. These and related provisions of subpart B
are basically patterned after section 110 of the Act and 40 CFR Part
51 (concerning adoption and submittal of state implementation plans
under section 110).
As discussed in the preamble to subpart B, a distinction is drawn
between designated pollutants which may cause or contribute to
endangerment of public health (referred to as "health-related pollutants")
1-2
-------
and those for which adverse effects on public health have not been
demonstrated (referred to as "welfare-related pollutants"). For
health-related pollutants, emission standards and compliance times in
state plans must ordinarily be at least as stringent as the corresponding
emission guidelines and compliance times in EPA's guideline documents
As provided in Subpart B, States may apply less strinaent requirements
for particular facilities or classes of facilities when economic
factors or physical limitations make such application sicinificantly
more reasonable.
For welfare-related pollutants, States may balance the emission
guidelines, times for compliance, and other information provided in
a guideline document against other factors of public concern in
establishing emission standards, compliance schedules, and variances,
provided that appropriate consideration is given to the information
presented in the guideline document and at public hearing(s) required
by subpart B and that all other requirements of subpart B are met.
Where sources of pollutants that cause only adverse effects to crops
are located in non-agricultural areas, for example, or where residents
of a community depend on an economically marginal plant for their
livelihood, such factors may be taken into account (in addition to
those that would justify variances if a health-related pollutant
were involved). Thus, States will have substantial flexibility to
consider factors other than technology and cost in establishing plans
for the control of welfare-related pollutants if they wish.
For reasons discussed in section 2 of this document, the
Administrator has determined that fluoride emissions from phosphate
1-3
-------
fertilizer plants may cause or contribute to endangerment of the
public welfare but that adverse effects on public health have not
been demonstrated. As discussed above, this means that fluoride
emissions will be considered a welfare-related pollutant and the
States will have greater flexibility in establishing plans for the
control of fluorides than would be the case if public health might
be affected.
This guideline document provides a brief description of the
phosphate fertilizer industry, the five manufacturing categories
for which fluoride emission guidelines are established, and the
nature and source of fluoride emissions. Also, information is provided
regarding the effects of airborne fluorides on health, crops, and
animals.
Emphasis has been placed on the technical and economic evaluation
of control techniques that are effective in reducing particulate and
gaseous fluoride emissions, with particular emphasis on retrofitting
existing plants. Some costs were frequently not available and were
fragmentary. Therefore, the cost basis for adoption of State
standards based on the emission guidelines is instead developed by
engineering cost estimates on a hypothetical phosphate fertilizer
plant complex where assumed marginally acceptable controls are replaced
with controls based on the emission guidelines. These retrofits are
called retrofit models and are presented in Section 6.1.3.1.
The emission guidelines and the control equipment on which
they are based are discussed in Sections 7 and 8. The environmental
1-4
-------
assessment of the emission guidelines is presented and discussed in
Section 9. The remainder of this introductory section summarizes
information presented in subsequent sections.
1.2 HEALTH AND WELFARE EFFECTS OF FLUORIDES
Fluoride emissions from phosphate fertilizer plants have been
determined to be welfare-related [i.e. no demonstrated impact upon
public health for purpose of section lll(d)]. The daily intake of
fluoride inhaled from the ambient air is only a few hundredths of a
milligram - a very small fraction of the total intake of the average
person. If a person is exposed to ambient air containing about
eight micrograms (yg) of fluoride per cubic meter, which is the
maximum average concentration that is projected in the vicinity of a
fertilizer facility with only moderate control equipment (Table 9-5),
his total daily intake from this source is calculated to be about 150
yg. This is very low when compared with the estimated daily intake
of about 1200 yg from food, water and other sources for the average
person. Also, the intake of fluoride indirectly through standard
food chains is insignificant. Fluorides are not passed into dairy
products and are only found in farm produce in very small amounts.
Fluorides do, however, cause damage to livestock and vegetation
in the immediate vicinity of fertilizer plants. Ingestion .of
fluorides by livestock from hay and forage causes bone lesions,
lameness and impairment of appetite that can result in decreased
weight gain or diminished milk yield. It can also affect developing
teeth in young animals, causing more or less severe abnormalities
1-5
-------
in permanent teeth. Exposure of plants to atmospheric fluorides can
result in accumulation, foliar lesions, and alteration in plant
development, growth, and yield.
1.3 FLUORIDES AND THEIR CONTROL
For purposes of standards of performance for new stationary
sources (SPNSS) and the attendant requirements of section lll(d),
emissions of "total fluorides," rather than specific fluorides are
limited. Total fluorides means elemental fluorine and all compounds
of fluorine measured by reference methods identified in subparts T,
U, V, W, and X and specified in Appendix A of 40 CFR, Part 60, or
equivalent or alternative test methods.
Good control of fluoride emissions from phosphate fertilizer
manufacturing operations is achievable by water scrubbers which are
properly designed, operated, and maintained. The most satisfactory
scrubber for general use seems to be the spray crossflow packed
scrubber. Other scrubbers, such as the venturi and the cyclonic
spray tower can give satisfactory results when used in series. The
spray-crossflow packed scrubber, shown diagramatically in Figure 6-1,
owes much of its success to its greater fluoride absorption capability
and its relative freedom from solids plugging. This plugging has qiven
some trouble in the past in DAP and GTSP plants, but current designs
are available which have acceptable turnaround periods . One design
involves a venturi ahead of, and integral with, the scrubber.
A description of the performance of water scrubbers in fluoride
emission control is given in Table 1-1. The industry-wide range of
control is given by a variety of scrubbers and is discussed in Chapter
1-6
-------
6. The scrubber data associated with best control technology was
obtained from EPA sponsored tests conducted during the development
of SPNSS. Most of the scrubbers tested were the spray crossflow
packed type, but a few venturi were tested.
1.4 EMISSION GUIDELINES
Emission guidelines for existing phosphate fertilizer manufacturing
facilities for control of fluoride emissions are described in this
Section. Table 1-1 gives the fluoride emission levels that may be
achieved by application of best adequately demonstrated technology to
existing facilities, including five manufacturing processes and the
storage facilities for granular triple superphosphate. Comparison of
these emission guidelines with the ranges shown for well-controlled
plants (Table 1-1) shows that equivalent control of fluoride emissions
can be achieved by application of best adequately demonstrated technology
for either new or existing sources.
Adoption of these controls would result in fluoride emission
reductions ranging from about 50 percent for granular triple super-
phosphate (6TSP) production facilities to around 90 percent for
run-of-pile triple superphosphate (ROP-TSP) plants. Overall nationwide
emissions would be reduced by about 75 percent.
The emission levels of Table 1-2 are identical to the standards
of performance for new stationary sources (SPNSS) since the best
adequately demonstrated technology applicable is the same type of
control equipment. The justification for application of this equipment
to existing as well as new sources is summarized in Section 1.6.1
and discussed more completely in Section 8.
1-7
-------
TABLE 1-1 PERFORMANCE OF AQUEOUS SCRUBBER EMISSION CONTROL EQUIPMENT
IN PHOSPHATE FERTILIZER MANUFACTURING PLANTS.
Fluoride Source
00
Wet-Process Phosphoric Acid
Superphosphoric Acid
Submerged Combustion
Vacuum Evaporation
Diammonium Phosphate
Triple Superphosphate
(run-of-pile - ROP)
Granular Triple Superphosphate
Granular Triple Superphosphate
Storage
Fluoride Emissions from Control Equipment
g TF/kg of PO input
Industry-Wide Ranne
0.01 - 0.030
0.06
2.5 x 1.0"3
0.03 - 0.25
0.10 - 1.30
0.10 - 1.30
2.5 x 10"4 - 7.5 x 10"4*
Best-Controlled Senment
0.001 - 0.0095
2.05 x 10"4 - 7.5 x 10~4
0.0125 - 0.03
0.015 - 0.1505
0.02 - 0.135
0.25 x 10"4 - 2.75 x 1C"4
"Units are g TF/hr/kg of P205 stored,
-------
TABLE 1-2 FLUORIDE EMISSION GUIDELINES FOR EXISTING
PHOSPHATE FERTILIZER MANUFACTURING PLANTS.
Process Source
of Fluorides
Wet-Process Phos-
phoric Add
Supcrphosphoric Add
Diammonium Phosphate
Triple Superphosphate
(ROP)
Granular Triple
Superphosphate
Granular Triple
Superphosphate Storage
Emission Guidelines
Total Fluorides
^
0.01
0.005
0.030
0.100
0.100
g/hr kilogram
- weiqht ppr unit nf P*>fic
Ibs/ton
0.02
0.01
0.06
0.2
0.2
input
Ibs/hr ton*
2.5 x 10
-4
5 x 10'4
*These denominator units are 1n terms of P205 stored,
-------
1.5 COMPLIANCE TIMES
The compliance times for installation of a wet scrubber are given
in Table 1-3, which is derived from Figure 6-15. Milestones in the
compliance schedule are also shown. The first milestone can increase
to 18 weeks if justifiable source tests must be run and control
alternatives evaluated. This is rather unlikely, since the spray-
crossflow packed scrubber is the one most widely specified for new
controls. The interval between milestones two and three is that required
for fabrication and shipping. The fabrication time is virtually beyond
the control of either the customer or the air pollution control
official. For this reason, a range of elapsed time must be understood
for fabrication. The compliance time can exceed 100 weeks and depends
upon availability of materials of construction, labor factors, work
TABLE 1-3
COMPLIANCE TIMES FOR INSTALLATION OF WET SCRUBBER FOR
A WET PROCESS PHOSPHORIC ACID PLANT
Milestone Elapsed Time. Weeks
Submit final control plan 6
to Agency
Award scrubber contract 26
Initiate scrubber 52
installation
Complete scrubber 72
installation
Final compliance achieved 74
backlogs, and many other things. If a given fertilizer complex has
to install several scrubbers, the total time for compliance may exceed
1-10
-------
that for only one scrubber. In practice, enforcement officials should
try to consider each plant on a case-by-case basis and should require
proof for the time requirements claimed for each milestone.
1.6 ASSESSMENTS
1.6.1 Economic
The information shown in Table 1-4 provides a major portion of
the justification for the emission guidelines. The costs in the
table were derived from retrofit models (section 6.1.3.1). The capital
and annualized costs shown in Table 1-4 represent emission controls
for each separate process.
Actual total expenditures for emission controls of a process
have to take into account the control costs allocated to its feed
materials. Table 1-5 summarizes retrofit control costs for fertilizer
plants of the capacities shown. These costs (see Table 7-1) include
prorated WPPA plant control costs according to the amount of acid
used. For example, the ROP plant control cost includes the control
cost for the 330 tons/day of wet process phosphoric acid required to
make 550 TPD of ROP, both on a PgOg basis. Therefore, the annualized
control costs, as a percent of sales, differ from those shown in
Table 1-4, except for the WPPA plant taken alone. The greatest unit
basis cost is for the combination of processing and storage of GTSP.
About 75,percent of GTSP production is believed to be already
sufficiently controlled while five of eight storate facilities may
need to be retrofitted if the States establish emission standards as
stringent as the emission guidelines. This would not have a great
effect on GTSP manufacture. About 60 percent of DAP plants would
1-11
-------
TABLF 1 4 ECONOMIC ASSESSMENT Of FLUORIDE EMISSION GUIDELINES FOR EXISTING
PHOSPHATE FERTILIZER MANUFACTURING FACILITIES.*
Process Source of
Fluorides
Wet- Process
Phosphoric Acid
jSuperphosphoric
jAcid
|Di ammonium
; Phosphate
Triple Super-
Phosphate (ROP)
Granular Triple
[Superphosphate
^Granular Triple
Superphosphate
Storage
Annual i zed
Control Cost
% of Sales
0.19 - 0.23
0.3
0.37
0.40 - 0.70
0.44
0.40
Capital Control
'• Cost of Equipment
$/short ton of
P2°5
1.26 - 1.51
1.04
4.00
4.00 - 6.85
4.55
4.10
Pprcent of Plants
* Not Meeting This
Emission Guidelines
i
47
21
60
40
25
70
Applicable
Emission
Guideline
grams/kilogram P2o5 input
0.01
0.005
0.03
0.1
0.1
2.5 x 10 *
ro
* Derived from EPA retrofit models.
** Based upon total annual production at capacity for 330 days/year.
** Units are grams F/hr/kilogram of PJ)- stored. This facility is
assumed to accompany a 400 short ton°P205/day GTSP plant.
-------
TABLE 1-5
SUMMARY OF RETROFIT CONTROL COST REQUIREMENTS FOR VARIOUS PHOSPHATE FERTILIZER MANUFACTURING PROCESSES
End Product
Design Rate, short
tons/day
(Pj>0s Basis)
Capital Control Cost,
$/short ton P205
Sales Price
($ per ton product)
Annual i zed Costs
Unit Basis
($ per ton product)
As a % of
Sales Price
Phosphoric
Acid
500
1.26 - 1.51
105
0.19 - 0.23
0.2
Superphosphoric
Acid
i
300
2.42
152
0.48
0.3
DAP
500
5.35
145
0.68
0.5
ROP-TSP ' GTSP
1
i
i
550
4.80 - 8.05
126
0.66 - 1.03
0.5 - 1.0
400
9.35
130
1.18
0.9
Based upon 90 percent capacity factor.
-------
possibly need to be retrofitted. Although this segment of the industry
requires the most control effort, control costs are only 0.5 percent of
sales.
The capital retrofit costs shown in Table 1-5, while significant,
are moderate. Annualized costs as a percent of sales are small,
showing that all the control costs can be readily recovered.
Cyclonic spray and venturi scrubbers, alone, do not have more
than about two transfer units, whereas the spray-crossflow packed
scrubber (SCPS) is furnished in the 5-9 transfer unit.range. The
former controls would require two or more scrubbers in series to
achieve the performance of one spray-crossflow packed scrubber. This
scrubber multiplication would cost more in comparison to the SCPS
and would not be selected for high degrees of fluoride removal when
costs are taken into account. Having made this choice, there is no
reason to design short of the SPNSS. A SCPS designed to achieve 0.08
Ibs F/ton for DAP can achieve 0.06 Ibs TF/ton with a little additional
packing. Therefore, the fluoride emission guidelines given in Table
1-1 reflect the performance of a control system which is judged to be
the best when costs are taken into account, and they are identical to
the SPNSS.
If the States establish emission standards as stringent as the
emission guidelines, the financial impact upon most existing plants
will be moderate, as shown in Tables 1-4 and 1-5. The only plants
likely to be financially burdened will be: small plants of less than
about 170,000 tons per year capacity; plants that are 20 years or more
of age; and plants isolated from raw materials, i.e. certain DAP plants
that purchase merchant phosphoric acid and ammonia.
1-14
-------
1.6.2 Environmental
The environmental assessment provided here is an assessment of
the difference between two degrees of control: 1) the reduction on
fluoride emissions resulting from application of the emission
guidelines and 2) the normal reduction in fluoride emissions resulting
from State Implementation Plans (SIP), local regulations, etc.
The adoption of fluoride emission standards would have a
beneficial impact upon air quality. Installation of retrofit controls
similar to those described in section 6.1.3.1 can reduce fluoride
emissions from existing sources by amounts ranging from 50 percent
for GTSP storage to 90 percent for ROP-TSP plants. The projected
average nationwide emission reduction that would result from applica-
tion of the emission guidelines is 74.5 percent or 1250 tons F/year.
The method of deriving these results is described in section 9.1.1.
The removal of fluoride pollutants from fertilizer plant emissions
would have a beneficial effect on the environment. The threshold
average concentration of fluoride in foliage that results in harmful
effects to animals when ingested is 40 ppm. The available data
suggest that a threshold for plant deterioration (foliar necrosis)
on sensitive plant species is also 40 ppm. As discussed in detail
in Chapter 2, an accumulation of fluoride in foliage of more than 40
ppm would result from exposure to a 30 day average air concentration
of gaseous fluoride of about 0.5 micrograms per cubic meter (yg/m3).
In order to evaluate potential ambient concentrations of fluoride,
atmospheric dispersion estimates were made for a typical phosphate
fertilizer complex. Groundlevel fluoride concentrations were compared
for marginally acceptable controls and for controls essentially
1-15
-------
similar to the emission guidelines shown in Table 1-1. At a distance
of about 2.5 kilometers (Table 9-5) from the complex, the 30-day
average fluoride ground-level concentration was 3.5 ug/m for the
marginally acceptable controls, and it was 0.5 ug/m for the good
retrofit controls. The conclusion is apparent that for protection
of public welfare (i.e. foliage, animals, etc.) marginally acceptable
controls are effective for protection of property beyond 15 km (9.3
miles) and best controls are effective beyond 2.5 km (1.5 miles)
relative to the fertilizer facility location.
Increased or decreased control of fluorides would not change
the volume of aqueous waste generated in a phosphate fertilizer
complex. Gypsum pond water is used and re-used, and a discharge is
needed only when there is rainfall in excess of evaporation.
Any solid waste generated by scrubbing fluorides would be in
the form of fluorosilicates of CaF2 in the gypsum ponds. Section
9.1.3 shows that the increase in solids discharged to the gypsum
pond due to scrubbing in a WPPA plant is only about 0.06 weight
percent, a negligible amount. The total fluoride solids increase
from a fertilizer complex to the gypsum pond would be nearer four
percent of the gypsum discharge, but much of this is from sources
other than scrubbing and certainly cannot be charged to small
increments in emission standards.
1.6.3 Energy
Energy requirements for State controls based on the
emission guidelines, in excess of existing controls, would be small
and varying from 0.4 to 25 KWH per ton PoOci depending on the
process. Raising the allowable emission levels would have only a
1-16
-------
small effect on these power figures. Section 9.1.4 estimates the
total incremental energy demand for the phosphate fertilizer industry.
This total incremental electrical energy demand that would result from
installation of retrofit controls to meet State controls based on the
guidelines is estimated as 27 x 106 KWH/yr, which is energy enough to
operate one SPA plant of 300 tons/day P205 for 115 days/year. Although
this energy number can be only an approximation, it puts the
incremental energy demand into perspective and shows that it is very
small compared to the total annual energy demand for the industry.
1.7 REFERENCES
1. Private communications, George B. Crane and Teller Environmental
Systems, Inc., December 13, 1974.
2. Biologic Effects of Atmospheric Pollutants; Fluorides. National
Academy of Sciences. Washington, D. C. Contract No. CPA 70-42.
1971.
3. Beck, Leslie L., Technical Report: An Investigation of the Best
Systems of Emission Reduction for the Phosphate Fertilizer
Industry. U. S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park, North
Carolina. April 1974.
1-17
-------
2. HEALTH AND WELFARE EFFECTS OF FLUORIDES
2.1 INTRODUCTION
In accordance with 40 CFR 60.22(b), promulgated on November 17,
1975 (40 FR 53340), this chapter presents a summary of the available
information on the potential health and welfare effects of fluorides
and the rationale for the Administrator's determination that is is a
welfare-related pollutant for purposes of section lll(d) of the Clean
Air Act.
The Administrator first considers potential health and welfare
effects of a designated pollutant in connection with the establishment
of standards of performance for new sources of that pollutant under
section lll(b) of the Act. Before such standards may be established,
the Administrator must find that the pollutant in question "may
contribute significantly to air pollution which causes or contributes
to the endangerment of public health or welfare" [see section
lll(b)(l)(a)]. Because this finding is, in effect, a prerequisite
to the same pollutant's being identified as a designated pollutant
under section lll(d), all designated pollutants will have been
found to have potential adverse effects on public health, public
welfare, or both.
As discussed in section 1.1 above, Subpart B of Part 60
distinguishes between designated pollutants that may cause or
contribute to endangerment of public health (referred to as "health-
related pollutants") and those for which adverse effects on public
health have not been demonstrated ("welfare-related pollutants").
In general, the significance of the distinction is that States
have more flexibility in establishing plans for the control of
2-1
-------
welfare-related pollutants than is provided for plans .involving
health-related pollutants.
In determining whether a designated pollutant is health-related
or welfare-related for purposes of section lll(d), the Administrator
considers such factors as: (1) Known and suspected effects of the
pollutant on public health and welfare; (2) potential ambient
concentrations of the pollutant; (3) generation of any secondary
pollutants for which the designated pollutant may be a precursor;
(4) any synergistic effect with other pollutants; and (5) potential
effects from accumulation in the environment (e.g., soil, water and
food chains).
It should be noted that the Administrator's determination
whether a designated pollutant is health-related or welfare-related
for purposes of section lll(d) does not affect the degree of control
represented by EPA's emission guidelines. For reasons discussed in
the preamble to Subpart B, EPA's emission guidelines [Vike standards
of performance for new sources under section lll(b)] are based on the
degree of control achievable with the best adequately demonstrated
control systems (considering costs), rather than on direct protection
of public health or welfare. This is true whether a particular
designated pollutant has been found to be health-related or welfare-
related. Thus, the only consequence of that finding is the deqree
of flexibility that will be available to the States in establishing
plans for control of the pollutant, as indicated above.
2-2
-------
2.2 EFFECT OF FLUORIDES ON HUMAN HEALTH.
1
2.2.1 Atmospheric Fluorides
The daily intake of fluoride inhaled from the ambient air is
only a few hundredths of a milligram --a very small fraction of the
total intake for the average person. If a person is exposed to
ambient air containing about 8 micrograms (yg) of fluoride per cubic
meter, which is the maximum average concentration that is projected
in the vicinity of a fertilizer facility with only mediocre control
equipment (Table 9-5), his total daily intake from this source is
calculated to be about 150 yg. This is very low compared with the
estimated daily intake of about 1200 yg from food, water, and other
sources for the average person.
Few instances of health effects in people have been attributed
to community airborne fluoride, and they occurred in investigations
of the health of persons living in the immediate vicinity of fluoride-
emitting industries. The only effects consistently observed are
decreased tooth decay and slight mottling of tooth enamel when compared
to control community observations. Crippling fluorosis resulting from
industrial exposure to fluoride seldom (if ever) occurs today, owinq
to the establishment of and adherence to threshold limits for exposure
of workers to fluoride. It has never been seen in the United States.
Even persons occupationally exposed to airborne fluoride do not usually
come in contact with fluoride concentrations exceeding the recommended
industrial threshold limit values (TLV). The current TLV for hydrogen
fluoride is 3 parts per million (ppm) while that for particulate
fluoride is 2.5 milligrams per cubic meter (mg/m3) expressed as elemental
fluorine.
2-3
-------
There is evidence that airborne fluoride concentrations that
produce no plant injury contribute quantities of fluoride that are
negligible in terms of possible adverse effects on human health and
offer a satisfactory margin of protection for people.
Gaseous hydrogen fluoride is absorbed from the respiratory tract
and through the skin. Fluoride retained in the body is found almost
entirely in the bones and teeth. Under normal conditions, atmosnheric
fluoride represents only a very small portion of the body fluoride
burden.
2.2.2 Ingested Fluorides
Many careful studies, which were reviewed by the National Academy
of Sciences, have been made of human populations living in the vicinity
of large stationary sources of fluoride emissions. Even in situations
where poisoning of grazing animals was present, no human illness due
to fluoride poisoning has been found. In some of these areas much of
the food used by the people was locally produced. Selection, processing
and cooking of vegetables, grains and fruits gives a much lower fluoride
intake in human diets than in that of animals grazina on contaminated
pasture.
In poisoned animals, fluorine levels are several thousand times
normal in bone, and barely twice normal in milk or meat. Calves and
lambs nursing from poisoned mothers do not have fluorosis. They do not
develop poisoning until they begin to graze. Meat, milk and eqqs from
local animals contain very little more fluoride than the same foods
from unpoisoned animals. This is due to the fact that fluorine is
deposited in the bones almost entirely.
2-4
-------
2.3 EFFECT OF FLUORIDES ON ANIMALS.1
In areas where fluoride air pollution is a problem, high-
fluoride vegetation is the major source of fluoride intake by livestock,
Inhalation contributes only a negligible amount to the total fluoride
intake of such animals.
The available evidence indicates that dairy cattle are the
domestic animals most sensitive to fluorides, and protection of
dairy cattle from adverse effects will protect other classes of live-
stock.
Ingestion of fluoride from hay and forage causes bone lesions,
lameness, and impairment of appetite that can result in decreased
weight gain or diminished milk yield. It can also affect developing
teeth in young animals, causing more or less severe abnormalities
in permanent teeth.
Experiments have indicated that long-term ingestion of 40 ppm
or more of fluoride in the ration of dairy cattle will produce a
significant incidence of lameness, bone lesions, and dental
fluorosis, along with an effect on growth and milk production.
Continual ingestion of a ration containing less than 40 ppm will give
discernible but nondamaginq effects. However, full protection
requires that a time limit be placed on the period during which high
intakes can be tolerated.
It has been suggested that dairy cattle can tolerate the
ingestion of forage that averages 40 ppm of fluoride for a year,
60 ppm for up to 2 months and 80 ppm for up to 1 month. The usual
food supplements are low in fluoride and will reduce the fluoride
concentration of the total ration to the extent that they are fed.
2-5
-------
Fluoride-containing dusts can be non-injurious to vegetation
but contain hazardous amounts of fluoride in terms of forage for
farm animals. Phosphate rock is an example of a dust that seemingly
has not injured plants but is injurious to farm animals. This was
made evident forty years ago when an attempt was made to feed
phosphate rock as a dietary supplement source of calcium and phosphate.
Fluoride injury quickly became apparent.2 Phosphate rock is used
for this purpose today, but only after defluorinating by heat treat-
ment. Phosphate rock typically contains up to about 4 weight percent
fluorine.
2.4 EFFECT OF ATMOSPHERIC FLUORIDES ON VEGETATION.1
The previous sections state that atmospheric fluorides are
not a direct problem to people or animals in the United States, but
that animals could be seriously harmed by ingestion of fluoride from
forage. Indeed, the more important aspect of fluoride in the ambient
air is its effect on vegetation and its accumulation in foraqe
that leads to harmful effects in cattle ana otner animals, fne
hazard to these receptors is limited to particular areas: industrial
sources having poorly controlled fluoride emissions and farms located
in close proximity to facilities emitting fluorides.
Exposure of plants to atmospheric fluorides can result in
accumulation, foliar lesions, and alteration in plant development,
growth, and yield. According to their response to fluorides, plants
may be classed as sensitive, intermediate, and resistant. Sensitive
plants include several conifers, several fruits and berries, and some
grasses such as sweet corn and sorghum. Resistant plants include
2-6
-------
several deciduous trees and numerous veaetable and field crops. Most
forage crops are tolerant or only moderately susceptible. In
addition to differences among species and varieties, the duration of
exposure, stage of development and rate of growth, and the environmer.tpl
conditions and agricultural practices are important factors in
determining the susceptibility of plants to fluorides.
The average concentration of fluoride in or on foliage that appears
to be important for animals is 40 ppm. The available data suggest
that a threshold for significant foliar necrosis on sensitive
species, or an dccumulation of fluoride in forage of more than 40 ppm
would result from exposure to a 30-day average air concentration of
gaseous fluoride of about 0.5 micrograms per cubic meter (yg/m3).
Examples of plant fluoride exposures that relate to leaf
damage and crop reduction are shown in Table 2-1.2 As shown, all
varieties of sorghum and the less resistant varieties of corn and
tomatoes are particularly susceptible to damage by fluoride ambient
air concentrations projected in the immediate vicinity of fertilizer
facilities (See Table 9-5).
2.5 THE EFFECT OF :.T-OSPHERIC FLUORIDES ON MATERIALS OF CONSTRUCTION.
2.5.1 Etching of Glass2
It is well known that glass and other high-silica materials
are etched by exposure to volatile fluorides like HF and SiF4. Some
experiments have been performed where panes of glass were fumigated
with HF in chambers. Definite etching resulted from 9 hours ex-
posure at a level of 590 ppb (270 ug/m3). Pronounced etching resulted
14.5 hours exposure at 790 ppb (362 ug/m3). Such levels would, of
2-7
-------
Table 2-1.,EXAMPLES OF HF CONCENTRATIONS AND EXPOSURE DURATIONS REPORTED
TO CAUSE LEAF DAMAGE AND POTENTIAL REDUCTION IN CROP VALUES
ro
C3
Plant
Sorghum
Corn
Tomato
Alfalfa
Concentration and Time*
Most sensitive varieties - most resistant varieties
0.7 ppb [0.32 ug/m3) for 15 days - 15 ppb (6.9 »g/m3) for 3 days
2 ppb (0.92 ng/m3) for 10 days - 800 ppb (366 Mg/m3) for 4 hrs.
f\
10 ppb (4.6 yq/iii3 for 100 days - 700 ppb (321 yg/m ) for 6 days
100 ppb (45.8 Mg/m3) for 120 days - 700 ppb (321 yg/m3} for 10 days
Concentrations are expressed In terms of parts per billion (ppb) with the equivalent
concentration expressed in micrograms per cubic meter (ug/m3) given in parenthesis.
-------
course, cause extensive damage to many species of vegatation. However,
ambient concentrations of this magnitude are improbable provided that
a fertilizer facility properly maintains and operates some type of
control equipment for abating fluoride emissions.
2.5.2 Effects of Fluorides on Structures
At the relatively low gaseous concentrations of fluorides tn
emissions from industrial processes, luuu ppm or less, tne damage
caused by fluorides is probably limited mostly to glass and brick.
Occasionally, damage to the interior brick lining of a stack has
been attributed to fluorides.
Considerable experience is available on corrosion in wet process
phosphoric acid plants, where the presence of fluoride increases the
3 5
corrosive effects of phosphoric acid. This experience applies to
the liquid phase; the effects of fluoride air emissions need more
study. Entrained crude phosphoric acid will corrode structural
steel and other non-resistant materials that it settles on, The
corrosive effects of "fumes" from the digestion of phosphate rock
have been acknowledged and good design and maintenance practices
for plant structural steel have been outlined. More information is
needed about effects of gaseous fluorides in low concentration outside
of the mill. It is usually difficult to separate the corrosive
effects of airborne fluorides from those of other local and back-
ground pollutants.
2-9
-------
2.6 RATIONALE
Based on the information provided the preceding sections of
chapter 2, it is clear that fluoride emissions from phosphate
fertilizer facilities have no significant effect on human health.
Fluoride emissions, however, do have adverse effects on livestock
and vegetation. Therefore the Administrator has concluded that
fluoride emissions from phosphate fertilizer facilities do not
contribute to the endangerment of public health. Thus fluoride
emissions will be considered a welfare-related pollutant for
purposes of section lll(d) and Subpart B of Part 60.
2.7 REFERENCES
1. Biologic Effects of Atmospheric Pollutants; Fluorides. National
Academy of Sciences. Washington, D.C. Contract No. CPA 70-42.
1971.
2. Engineering and Cost Effectiveness Study of Fluoride Emissions
Control. Resources Research Inc. and TRW Systems Group.
McLean, Va. Contract No. EHSD 71-14. 1972. p. 5-1 to 5-11.
3. Leonard, R.B. Bidding to Bulk Corrosion in Phos-Acid Concentration.
Chem. Eng. 158-162, June 5, 1967.
4. Dell, G.D. Construction Materials for Phos-Acid Manufacture.
Chem. Eng. April 10, 1967.
5. Pelitti, E. Corrosion: Materials of Construction for Fertilizer
Plants and Phosphoric Acid Service. In: Chemistry and Technology
2-10
-------
of Fertilizers, Sauchelli, V. (ed). New York, Reinhold Publishing
Corporation, 1960. p. 576-631.
6. Peletti, E. Corrosion and Materials of Construction. In:
Phosphoric Acid, Vol. I, Slack, A.V. (ed). New York, Marcel
Dekker, Inc., 1968. p. 779-884.
2-11
-------
3. PHOSPHATE FERTILIZER INDUSTRY ECONOMIC PROFILE AND STATISTICS
3.1 INDUSTRY STRUCTURE
The phosphate fertilizer industry is a segment of the agricultural
chemical industry that is devoted to the production and marketing of
commodities bearing the basic nutrients—nitrogen, phosphorous, and
potash—for crop production. From the perspective of end-use products,
the scope of the agricultural chemical industry includes ammonia,
ammonium nitrate, urea, ammonium phosphates, nitrophosphates, mixed plant
foods (in varying N-P-K combinations), superphosphates, phosphoric acid,
and potash. The phosphate production segment of the agricultural chemical
industry begins with the mining of phosphate rock; proceeds with the basic
chemical production of phosphoric acid and its subsequent processing to
diammonium phosphate (DAP), superphosphoric acid (SPA), and triple super-
phosphate (TSP); and culminates at the retailer level where the numerous
blends of fertilizers are formulated to satisfy the diverse interests of
consumers. There are three basic types of retailers - the granular NPK
producers (manufacturers of chemical formulations), the liquid fertilizer
manufacturers, and the mechanical blenders (dry bulk). These groups compete
with each other in some markets (mixed fertilizers).
The basic chemical producers in the industry sell merchant phosphoric
acid and products derived from phosphoric acid, such as SPA, DAP, and TSP.
NPK producers can therefore buy from a choice of raw materials to produce
a specific product. For example, the typical NPK plant operator can buy
DAP or produce it from wet-process phosphoric acid. Therefore, some com-
petition can be expected among the various phosphate concentrates.
3-1
-------
The basic chemical producers, which are the focus of this
analysis, are generally not identifiable as single product firms.
Very few firms are totally dependent on fertilizer production for their
business. Most fertilizer production is conducted as a subsidiary
activity in well diversified, often-times large, corporations. These
firms are chemical manufacturers or petrochemical companies. Some
companies are farm cooperatives, vertically integrated from production to
marketing, in geographic areas in which they are economically based.
These latter firms are primarily engaged in serving farm customers by
retailing fertilizers, by purchasing and shipping grains and other
agricultural products to regional centers, and by providing necessary
supplies and services. Finally, there are firms engaged in fertilizer
production that derive the main portion of their revenues from totally
unrelated activities, such as steel manufacture, pipeline construction,
etc.
Generally, the basic chemical producers own the sources of
their raw materials (phosphate rock mines). According to 1970
production statistics, the ten largest firms in rock mining are ranked
as follows:
TABLE 3-1
TEN LARGEST PHOSPHATE ROCK PRODUCERS1
Production
Firm (1000 Short Tons)
International Minerals & Chemicals 8,000
Williams Co. (was Continental Oil Co.) 6,500
Mobil Chemical Company 5,900
Occidental Chemical Company 3,750
American Cyanamid Corp. 3,650
U.S.S. Agrichemicals 3,640
3-2
-------
TABLE 3-1 (CONTINUED)
Production
Firm (1000 Short Tons)
Swift & Company 3,000
Texas Gulf, Inc. 3,000
Stauffer Chemical Company 2,500
Gardinier, Inc. (was Cities Service Co.) 2,000
Total 41,940
U.S. Production 50,640
Percent of total production of ten largest
fi rms 83%
Based on the production of wet-process phosphoric acid, the
cornerstone of the basic chemical production in the phosphate fertilizer
industry, the ten largest firms in terms of 1972 production are as follows;
TABLE 3-2
TEN LARGEST PHOSPHORIC ACID PRODUCERS2
Production Capacity
Firm (1000 Short Tons PpQ5)
CF Industries 880
Freeport Minerals Co. 750
Gardinier , Inc. 544
Farmland Industries 455
Beker Industrial Corp. 411
Texas Gulf Inc. 346
01 in Corporation 337
W.R. Grace & Co. 315
U.S.S. Agri-Chemicals Inc. 266
Occidental Chemical Co. 247
Total 4,551
U.S. Production 6,370
Percent of total production of ten 71%
largest firms
3-3
-------
A review of the above tabulations reveals vertical
integration from the mine through the c*i°m-'c?1 "»r«Huction
within several corporations. Each of the precerilnn
phosphate rock producers owns basic chemical production facilities
directly, or through equity interest in chemical producing companies.
CF Industries and Farmland Industries are integrated from the chemical
production stage forward to the ultimate retailing of fertilizers.
Freeport Minerals is strong in ownership of sulfur reserves, an
important raw material for production of phosphoric acid. Beker
Industries is a newcomer into the fertilizer industry, as they purchased
the fertilizer assets of Hooker Chemical (Occidental Petroleum) and El
Paso Products Company.
3.2 EXISTING PLANTS
The United States is the world's leading producer and consumer of
phosphate fertilizer with an annual consumption of nearly 20 percent of
the world's total. Phosphate fertilizers are produced by several
processes and consumed in various product forms. Plant statistics are
available for those processes of interest under the following classifications:
wet-process phosphoric acid, superphosphoric acid, triple superphosphate,
and ammonium phosphates.
Tables 3-3 through 3-6 list the company, location, year brought on
stream, and annual production capacity of all wet-process phosphoric
acid, superphosphoric acid, triple superphosphate, and ammonium phosphate
facilities in the United States. Figures 3-1 and 3-2 show the geographic
distribution of these plants.
3-4
-------
TABLE 3-3
PRODUCTION CAPACITY OF WET-PROCESS PHOSPHORIC ACID (1973)
4,5
CO
Company
Allied Chen. Corp.
Union Texas Petroleum Div.
Agricultural Dept.
Arkansas Louisiana Gas Co.
Arkla Chem. Corp., subsid.
Atlantic Richfield Co.
ARCO Chem. Co., Div.
Beker Indust. Corp.
Agricultural Products Corp.,
subsid.
National Phosphate Corp.,
subsid.
Borden Inc.
Borden Chem. Div.
Smith-Douglass
CF Indust., Inc.
Bartow Phosphate Complex
Plant City Phosphate Complex
Location
Geismar, La.
Helena, Ark.
Fort Madison, Iowa
Conda, Idaho
Marseilles, 111.
Taft, La.
Piney Point, Fla.
Streator, 111.
Bartow, Fla.
Plant City, Fla.
Date on Stream Annual Capacity
{Thousands of Tons
1967
1967
1968
1972
1962
1965
1966
1953
1961
1965
160
50
225
125 (adding 125)
105
185 (adding 30)
165
25
650
250 (adding 375)
-------
co
i
O")
Company
Conserv Inc.
Farmland Indust., Inc.
Freeport Minerals Co.
Freeport Chem. Co., Div.
Gardinier, Inc.
W. R. Grace & Co.
Agricultural Chems. Group
International Minerals and
Chemicals Corp.
Mississippi Chem. Corp.
Mobil Oil Corp.
Mobil Chem. Co.
Agricultural Chemicals, Div.
North Idaho Phosphate Co.
Occidental Petroleum Corp.
Occidental Chem. Co., subsid.
Occidental of Florida Div.
Western Div.
TABLE, 3-3
(CONTINUED)
Location
Nichols, Fla.
Greenbay, Fla.
Uncle Sam, La.
Tampa, Fla.
Bartow, Fla.
New Wales, Fla.
Pascagoula, Miss.
Depue, 111.
Kellogg, Idaho
White Springs, Fla,
Lathrop, Calif.
Date on Stream
1973
1965
1968
1961
1962
1975
1958
1966
1960
1966
1954
Annual Capacity
(Thousands of Tons
""""^™^^"^^^^^^^"^^™
150
500
750
490
315 (adding 250)
(600)
130
130
30
225 (adding 350)
17 (adding 23)
-------
CO
I
Company
01 in Corp.
Agricultural Chems. Div.
Indust. Products and Services Div.
Pennzoil Co.
Pennzoil Chem., Inc., subsid.
Royster Co.
J. R. Simplot Co.
Minerals and Chem. Div.
Stauffer Chem. Co.
Fertilizer and Mining Div.
Texas Gulf, Inc.
Agricultural Div.
Union Oil Co. of California
Collier Carbon & Chemical
Corp., subsid.
United States Steel Corp.
USS Agri-Chemicals, Div.
Valley Nitrogen Producers, Inc.
The Williams Companies
Agrico Chem. Co., subsid.
TABLE 3-3
(CONTINUED)
Location Date
Pasadena, Tex.
Joliet, 111.
Hanford, Calif.
Mulberry, Fla.
Pocatello, Idaho
Pasadena, Tex.
Salt Lake City, Utah
Aurora, N. C.
Nichols, Calif.
Bartow, Fla.
Ft. Meade, Fla.
Helm, Calif.
Edison, Calif.
South Pierce, Fla.
Dona Idsonvi lie, La.
on Stream
1965
1972
1968
1962
1966
1954
1966
1961
1964
1962
1959
1966
1965
1974
Annual Capacity
(Thousands of Tons ?£§)
230 (adding 14)
160
10 (adding 10)
140
145 (adding 80)
60
65
350 (adding 350)
8
95
190
35 (addinq 83)
8
280
(400)
TOTAL 6,453 (adding 2,690)
-------
TABLE 3-4
PRODUCTION CAPACITY OF SUPERPHOSPHORIC ACID (1973) 4'5
CJ
00
Company
Allied Chem. Corp.
Union Texas Petroleum Div.
Agricultural Dept.
Beker Indust. Corp.
Agricultural Products Corp.
subsid.
Farmland Indust., Inc.
Location
Geismar, La.
Conda, Idaho
Greenbay,, Fla.
Internat'l Minerals & Chem. Corp. Bartow, Fla.
North Idaho Phosphate Co.
Occidental Petroleum Corp.
Occidental Chem. Co., subsid.
Occidental of Florida Div.
J. R. Simplot Co.
Minerals and Chem. Div.
Stauffer Chem. Co.
Fertilizer and Mining Div.
Kellogg, Idaho
White Springs, Fla.
Pocatello, Idaho
Pasadena, Tex.
Salt Lake City, Utah
Date on Stre<
1967
—
1971
1963
1967
1964
1966
1964
1966
im Annual Capacity
{Thousands of Tons P-Og)
127
45
138
52
87
139"
11
69
32 (adding 23)
22
34
Process &
Remarks3
submerged
combustion
vacuum
vacuum
vacuum: acid
Is redl luted
and used
captlvely to
make feed
phosphates
vacuum
submerged
combustion
vacuum
vacuum
vacuum
-------
TABLE 3-4
(CONTINUED)
Company
Texas Gulf, Inc.
Agricultural Div.
Location
Aurora, N. C.
Manufactured from wet process phosphoric acid
Date on Stream
1967
1970
Annual Capacity Process &
(Thousands of Tons PO^C) Remarks
83 vacuum
83
155" (adding 82)
TOTAL
783 (addi nn 105)
CO
vo
-------
to
O
TABLE 3-5
PRODUCTION CAPACITY OF TRIPLE SUPERPHOSPHATE (1973)
4-7
Company
Beker Indust. Corp.
Agricultural Products
Corp., subsid.
Borden Inc.
Borden Chen. Dlv.
Smith-Douglass
CF Indust., Inc.
Plant City Phosphate
Complex
Conserv Inc.
Farmland Indust., Inc.
Gardlnler Inc.
Location
Conda, Idaho
Date on Annual Capacity3
Stream (Thousands of Tons Product)
1974-75
Plney Point, Fla. 1966
Plant City, Fla. 1965
Nichols, Fla.
Greenbay, Fla.
Tampa, Fla.
W. R. Grace & Co. Bartow, Fla.
Agricultural Chems. Group
(340)
70
530 (adding 400)
1973
1965
1952
1972
1954
1958
280
190
395
350
74T
390
275
55T
Product
Granular
ROP - granulate
portion of pro-
duction
ROP
Granular
ROP and granular
ROP and granular
Joplln, Mo.
1953
100
ROP
-------
Company
Mississippi Chem. Corp.
Occidental Petroleum Corp.
Occidental Chem. Co.,
subsId.,
Occidental of Florida
Dlv.
Royster Co.
J.R. Simplot Co.
Minerals & Chem. 01v.
Stauffer Chem. Co.
Fertilizer & Mining Dlv.
Texas Gulf, Inc.
Agricultural Dlv.
USS Agrl-Chemicals, Dlv.
The Williams Companies
Agrico Chem. Co., subsid.
Capacities are for gross weight.
TABLE 3-5
(CONTINUED)
Location Date on
Stream
Pascagoula, M1ss. 1972
White Springs, Fla. 1966
Mulberry. Fla. 1968
Pocatello, Idaho 1954
Salt Lake City, 1954
Utah
Aurora, N.C. 1966
Fort Meade, Fla. 1962
South Pierce, Fla. 1965
Annual Capacity9
(Thousands of Tons Product)
300
460
210
120
35
370 (adding 130)
295
600
Product
Granul ar
Granular
ROP
ROP- aranulate
portion of pro-
duction
ROP- granulate
portion of pro-
duction
ROP and granular
Granular
ROP- qranul ate
TOTAL 4,970 (adding 870)
• *W *t I VII IV* I VL WW
portion of pro-
ductIon
-------
TABLE 3-6
PRODUCTION CAPACITY OF AMMONIUM PHOSPHATES (1973)
4-6
Company
Location
Annual Capacity6
Date on Stream (Thousands of Tons Product) Remarks
CO
I
ro
Allied Chem. Corp.
Union Texas Petroleum Div.
Agricultural Dept.
American Plant Food Corp.
Arkansas Louisiana Gas Co.
Arkla Chem. Corp., subsid.
Beker Indust. Corp.
Agricultural Products Corp.,
subsid.
National Phosphate Corp.,
subsid.
Borden Inc.
Borden Chem. Div.
Smith-Douglass
Brewster Phosphates
C F Indust., Inc.
Bartow Phosphate Complex
Plant City Phosphate Complex
Geismar, La.
Galena Park, Tex.
Helena, Ark.
Conda, Idaho
Marseilles, 111.
Taft, La.
Piney Point, Fla.
Streator, 111.
Luling, La.
Bartow, Fla.
Plant City, Fla.
1967
150
DAP, leased to
Brewster
Phosphates
1966
1967
1972
1962
1965
1966
1965
1961
1974
175
150
270
200
395 (adding 70)
130
90
385
1,000
(390)
Mostly
DAP and
DAP
DAP
DAP
Mostly
DAP
DAP
DAP
mixtures
mixtures
mixtures
-------
CO
CO
Company
Conserv Inc.
Farmland Indust., Inc.
First Mississippi Corp.
Gardinier Inc.
W. R. Grace & Co.
Agricultural Chems. Group
Internat'l Minerals & Chem.
Corp.
Kaiser Steel Corp.
TABLE 3-6
(CONTINUED)
Location
Nichols, Fla.
Greenbay, Fla.
Joplin, Mo.
Fort Madison, Iowa
Tampa, Fla.
Bartow, Fla.
Bartow, Fla.
New Wales
Fontana, Calif.
Date on Stream
1973
1965
1954
1968
1959
1966
1962/63
1975
1955
Annual Capacity3
(Thousands of Tons Product)
200
390
245
495
525
235
50
(490)
25
Remarks
MAP
DAP
Mixtures
DAP and
Mixtures
DAP, MAP
DAP, MAP
Feed grade
DAP and MAP
DAP and MAP
Switches
between
ammonium sul-
fate and DAP.
-------
CA.
I
-p.
TABLE 3-6
CONTINUED)
Company
Lone Star Gas Co.
Nipak, Inc., subsid.
Mississippi Chem. Corp.
Mobil Oil Corp.
Mobil Chem. Co.
Agricultural Chemicals Div.
Location
Kerens, Tex.
Yazoo City, Miss.
Depue, 111.
Monsanto Co. Trenton, Mich.
Monsanto Indust. Chems. Co.
North Idaho Phosphate Co. Kellogg, Idaho
Occidental Petroleum Corp.
Occidental Chem. Co., subsid.
Occidental of Florida Div.White Springs, Fla.
Western Div. Lathrop, Calif.
Plainview, Tex.
01 in Corp.
Agricultural Chems. Div. Pasadena, Tex.
Pennzoil Co. Hanford, Calif.
Pennzoil Chem., Inc. subsid.
Royster Co.
J. R. Simplot Co.
Minerals and Chem. Div.
Mulberry, Fla,
Pocatello, Idaho
Annual Capacity3
Date on Stream (Thousands of Tons Product) Remarks
1964 no
1958
1966
1965
1966
1973
1968
1961
630
240
40 - 45
65
575 (adding 350)
165
25
800
Mostly mixtures
Mostly mixtures
DAP
DAP, MAP, and
mixtures
270
190 (adding 50)
Mostly mixtures
Mostly mixtures
Mostly mixtures
PiAP
DAP
DAP and MAP
-------
CO
u
Company
Standard Oil Company of Calif.
Chevron Chem. Co., subsid.
Stauffer Chem Co.
Fertilizer & Mining Div.
Tennessee Valley Authority
Texas Gulf, Inc.
Agricultural Div.
Union Oil Co. of California
Collier Carbon and Chem. Corp.
subsid.
United States Steel Corp.
USS Agri-Chemicals, Div.
TABLE 3-6
(CONTINUED)
Location
Fort Madison, Iowa
Kennewick, Mash.
Richmond, Calif.
Pasadena, Tex.
Salt Lake City, Utah
Muscle Shoals, Ala.
Aurora, N. C.
Nichols, Calif.
Cherokee, Ala.
Annual Capacity
Date on Stream (Thousands of Tons Product) Remarks
Valley Nitrogen Producers, Inc. Bakersfield, Calif.
Helm, Calif.
Arizona Agrichemical Corp., Chandler, Arizona
subsid.
The Williams Companies
Agrico Chemical Co., subsid.
Donaldsonville, La.
1962
1959
1957
1966
1965
1966
1966
1957
1962
1960
1959
1967
1969
200
75
100
135
65
33
220
55
245
10
Mixtures
Mixtures
Mixtures
Mostly DAP & MAP
Mostly DAP & MAP
Solid ammonium
polyphosphates
DAP
Mostly mixtures
DAP & mixtures
8-24-0
TOTAL
140 (adding 150) MAP & mixtures
60 Mao, 16 - 20 - 0
700 (adding 840) DAP
10,288 (adding 2,340)
Capacities are for gross weight of product and includes diammonium phosphate (DAP), monammonium phosphate (MAP),
ammonium phosphate sulfate and ammonium phosphate nitrate.
-------
\
\\1
TW5^-
rlbUKt J-I
TRIPLE SUPERPHOSPHATE AND AMMONIUM PHOSPHATE PLANT LOCATIONS
/ r
/ >
fiSnTo
LEGEND
• TRIPLE SUPERPHOSPHATE PLANT
A AMMONIUM PHOSPHATE PLANT
(INCLUDES NITRIC PHOSPHATE
PRODUCERS)
TRIPLE SUPERPHOSPHATE AND SUPER-
PHOSPHATE PLANT AT SAME LOCATION
-------
/ PSSBB-
FIGURE 3-2
WET-PROCESS AND SUPERPHOSPHORIC ACID PLANT LOCATIONS
pSSKS J
LEGEND
A WET-PROCESS AND SUPERPHOSPHORIC
ACID PLANT AT SAME LOCATION
• WET-PROCESS PHOSPHORIC ACID PLANT
-------
As might be expected, the majority of the plants are located either near
the phosphate rock deposits of Florida, Idaho, and Utah; the sulfur deposits
of Texas and Louisiana; or the farming outlets.
As of 1973, there were 34 operating wet-process phosphoric acid
plants with an annual capacity of 6,435,000 tons of P205; 10 super-
phosphoric acid facilities with an annual capacity of 783,000 tons of
PoOcl 15 triple superphosphate facilities with an annual capacity of
4,970,000 tons of product, and 44 ammonium phosphate facilities with an
annual capacity of 10,280,000 tons of product. The production capacity
attributed to wet-process acid plants in Table 3-3 is about 80 percent
of the total United States phosphoric acid production. The balance is
produced from elemental phosphorous made by the furnace method, which is
not covered by the standards of performance for new stationary sources
(SPNSSfr for the phosphate fertilizer industry. Table 3-5 presents statistics
for facilities producing both run-of-pile triple superphosphate and granular
triple superphosphate; it is estimated that between 60 and 70 percent of
the total capacity is associated with granular TSP. Approximately 70
percent of the production capacity of ammonium phosphates listed in
Table 3-6 can be attributed to diammonium phosphate.
3.3 CAPACITY UTILIZATION
The phosphate fertilizer industry has followed a cyclic pattern
of capital investment in new plants. This pattern is demonstrated by
the graphs for phosphoric acid and ammonium phosphate production
presented in Figures 3-3 and 3-4. As shown in the graphs by the
duration between peak utilization (operating near 100 percent), the
3-18
-------
cycle length appears to be 6 to 7 years. During the 1965 to 1972 cycle,
expansion peaked in 1969. Slackened demands prompted price cutting
and eventual temporary shutdown of some facilities. At the end of the
cycle, supply of plant capacity came in balance with production.
For additional insight into'the cyclic trend of capacity
utilization, Table 3-7 lists operating ratios for phosphoric acid and
diammonium phosphate production.
TABLE 3-7
PRODUCTION AS PERCENT OF CAPACITY8
Year
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
WPPA
100
92
80
77
69
84
96
96
89
89
83
82
DAP
72
63
66
56
54
78
96
96
—
—
—
~~
During mid-1973, the industry was operating near capacity. Idle
plants that had been shutdown during the 1968-1970 recession were being
refurbished for production. Beker Industries is one example of a firm
that purchased idle phosphate facilities from petroleum companies for
acid and ammonium production. New plant construction, as announced
3-19
-------
by Agrico Chemical and IMC, will not provide significant additions
to supply of phosphates until 1975 or 1976. By inspection of the
profiles in Figures 3-3 and 3-4 and the operating ratios presented
in Table 3-7, planned plant capacity for phosphoric acid seems
sufficient through 1976; ammonium phosphate capacity, on the other
hand, will have to be increased to cope with the projected demand.
3.4 CONSUMPTION PATTERNS
For an understanding of the historical consumption patterns of
WPPA, SPA, DAP, and TSP, an overview of consumption of all phosphate
fertilizers is presented. Although some superphosphoric acid is consumed
in the form of animal feed supplements, most phosphate production from
wet-process phosphoric acid ends up in fertilizers.
Historical data are presented for U.S. consumption in Table 3-8.
Liquids and solids (bulk and bagged) are all included in these data.
Total consumption includes phosphate values derived from wet-process
phosphoric acid to produce triple superphosphate, and phosphate rock
reacted with sulfuric acid to produce normal superphosphate.
Overall, the growth trend in total consumption has been at a rate
of 6.5 percent compounded annually from the base year 1960. However,
normal superphosphate production has declined steadily from 1,270,000
tons (P205) in 1960 to 621,000 tons (P20g) in 1973.9 The gap in
phosphate values generated by the decline in NSP has been mostly taken
up by diammonium phosphate production, as well as wet-process phosphoric
acid, the intermediate product. Hence, consumption of wet-process
phosphoric acid and diammonium phosphate production have grown at a
more rapid rate than total consumption of phosphates.
3-20
-------
•^ FIGURE 3-3. CAPACITY UTILIZATION OF WET-PROCESS PHOSPHORIC ACID10'11
(.
rt.i~"~-
i;-.zn:"".:r.
3^1 4Tt JTi--^-"--^ — p—114;
eoduetfon;
- Actual data
3-21
-------
^^:r-.v IfcSSfe&ife^ii^^^^
FIGURE 3-4. CAPACITY UTILIZATION OF AMMONIUM PHOSPHATES12
FT** r*T" -^r-r-r- • r-t-rr
iii 11 x-— . •'. --.- ——^—
... ...-
rittH.rtrja.iir.-.'f-^;-- - ht:t^-
3-22
-------
The two other major categories presented in Table 3-8 separate
the basic chemicals that are applied directly to the soil from those
that receive further processing into mixtures; foods containing at least
two of the nutrients basic to plant growth. Some duplication of reporting
is evident in the statistics as some undetermined amount appears twice,
in "mixtures" and "direct applications".
Review of the data in Table 3-8 shows that the demand for
normal superphosphate has decreased drastically in recent years.
During this same time period, the use of ammonium phosphates (other
than DAP) and triple superphosphate have sftowed while the demand for
DAP has grown steadily. Almost all direct application materials are
now DAP or GTSP. Demand for these materials appears to have grown
more rapidly than total consumption. Additional factors contributing
to this trend are the rise of bulk blending operations and intensive
cultivation (emphasis on increased yield per acre).
Fanners have lately realized that mechanical blends of grandulated
concentrates do just as well as a grandulated, chemically produced
NPK food and are available at lower costs. A shift from normal
superphosphate and run-of-pile triple superphosphate production to the
grandulated concentrates, DAP, and GTSP, Is occurring.
The shift in product usage has also been accompanied by a shift
in raw materials for NPK plants. Run-of-pile triple superphosphate
has been replaced by wet-process phosphoric acid as a raw material.
Improvement in phosphoric acid technology has made it possible to inhibit
the precipitation of impunities during shipping, as most NPK plants
are far removed from the areas of acid production.
3-23
-------
TABLE 3-8. U.S. PHOSPHATE CONSUMPTION, 1960-1973
(1000 tons P0
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973C
Total
Consumption
2572
2645
2807
3073
3378
3512
3897
4305
4452
4666
4574
4803
4873
5072
Mixtures
2033
2069
2219
2474
2705
2816
3111
3503
3579
3724
3709
3943
4007
4200
Di ammonium
Phosphates
35
63
110
177
244
302
418
451
608
724
726
814
884
~
Direct Application
Normal
Superphosphate
103
100
97
98
93
95
94
86
79
72
62
55
44
35
Materials
Triple
Superphosphate
185
203
217
220
289
309
413
432
487
585
546
556
577
569
Ammonium b
Phosphates
171
188
205
205
216
204
221
224
227
207
184
179
174
-
ro
alncludes grades 18-46-0 and 16-48-0
Includes grades 11-48-0, 13-39-0, 16-20-0, 21-53-0, and 27-14-0
Preliminary
-------
Consumption of superphosphoric acid is only recently beginning to
expand. To date, it has been used primarily for the production of liquid
fertilizers with some secondary end-use in the production of animal feed
supplements. Data for consumption is limited. Superphosphoric acid con-
sumption is currently estimated at only 15 percent of overall phosphate
consumption.
Several reasons are presented to explain the expected expansion .of
superphosphoric acid consumption. Technology has made it possible to
produce a product which eliminates the problems of sludge formation en-
countered during shipping and storage of wet-process acid. Increased crop
yield per unit PgOg applied from liquid fertilizers has been claimed.
Transportation costs per ton of PJ^c are less for liquid than for solid
fertilizers.
The implications of the shifting patterns in the industry in
response to demands for cheaper, better quality products are as follows:
1. Granular concentrates will continue to expand in production;
these include DAP and GTSP.
2. Run-of-pile TSP production will decline and be replaced by
GTSP and DAP.
3. Superphosphoric acid will have the largest growth rate of all
phosphate commodities.
3.5 FUTURE TRENDS
The phosphate fertilizer industry has experienced dynamic growth
in recent years. Table 3-9 provides production statistics for wet
process phosphoric acid, triple superphosphate, and ammonium phosphates
3-25
-------
TABLE 3-9
U.S. PRODUCTION OF THREE COMMODITIES IN THE
PHOSPHATE INDUSTRY, 1950-1973
Year
1950
1955
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
Wet Process
Phosphoric Acid
(Thousand
299
775
1,325
1,409
1,577
1,957
2,275
2,896
3,596
3,993
4,152
4,328
4,642
5,016
5,594b
5,621b
Triple
Superphosphate
tons of P205)
309
707
986
1,024
960
1,113
1,225
1,466
1,696
1,481
1,387
1,354
1,474
. 1 ,503
1,659
l,716b
Ammonia3
Phosphates
_
_
269
370
536
-
-
1,081
1,376
1,747
1,633
1,844
2,092
2,395
2,577
2,665b
Includes diammonium phosphate, monammoniiim phosphate, ammonium
phosphate sulfate, ammonium phosphate nitrate, and other phosphate
fertilizers.
Preliminary.
3-26
-------
from 1950 to 1973. During this period, wet-process phosphoric acid
has shown a strong steady growth because of its role as an intermediate
in the production of ammonium phosphates, triple superphosphate, and
other phosphate products. Production, of wet acid has grown at an average
annual rate of 14 percent since 1960. Table 3-3 lists announced con-
struction of wet acid plants through 1975. This new construction will
increase total capacity by 41.6 percent. An average annual growth rate
15
of ,6.0 percent is expected for the period from 1976 to 1980.
Documentation of superphosphoric acid production is very limited.
The usual reporting groups, such as Department of Commerce and TVA, do
not report production figures. The Fertilizer Institute reports
production in its Fertilizer Index but privately concedes that its
published figures for the years of 1969-1971 are below estimates of
actual production.
A 40 percent saving in freight costs per unit weight of PgOg is
realized when phosphoric acid is shipped in the concentrated super-
acid form. Anticipated growth for superphosphoric acid is largely
due to this reduced shipping cost and the availability of merchant
grade wet-process acid will be a major factor affecting expansion. Announced
construction through 1975 will increase existing capacity by approximately
13 percent. Rapid growth during the remainder of the decade is expected.
By definition, ammonium phosphates are products manufactured directly
from ammonia, phosphoric acid, and sometimes other acids, in contrast
to those ammoniated phosphates that are produced in NPK granulation plants from
ammonia and run-of-pile triple superphosphate. "Diammonium" phosphates
3-27
-------
include 16-48-0 (N, PgOg, and KgO content) and 18-46-0 grades. Monam-
monium phosphates are 11-48-0. These two generic products are produced
strictly from ammonia and phosphoric acid; other ammonium phosphates are
produced from a mixture of ammonia, phosphoric acid, nitric acid, and
possibly sulfuric acid.
The growth of ammonium phosphates has been more rapid than that of
triple superphosphates - 20 percent annual growth since 1960 - because
of several inherent advantages of ammonium phosphates (see Section 4.4).
New construction through 1975 will increase production capacity by 22.7
percent. Annual growth from 1975 to 1980 is projected at 6 percent.15
Production of triple superphosphate has grown at an average annual
rate of 4 percent since 1960. Triple superphosphate is produced by
two methods; the den method and the granulator method. The den method
produces a material (run-of-pile) that is non-uniform in particle
size. This material is stored, pulverized, and shipped to NPK plants
for ammoniation. The granulator method produces a granular product that
is sold to bulk blender retailers for mixing or for direct application
(as a 0-46-0 fertilizer) to the soil.
No statistics are available as to the breakdown of run-of-pile
versus direct granulator production. In the industry, run-of-pile
production by the primary producer may be granulated and sold as GTSP
to bulk blender retailers as a direct application fertilizer. Ultimately,
essentially all run-of-pile production becomes granulated, either by the
primary producer or by the NPK plant. Only granulated TSP is expected
to be of importance in the future.
3-28
-------
Announced new construction through 1975 will result in a 17.4
percent increase in triple superphosphate production capacity, however,
this apparent growth does not take into consideration the possible
closings of existing run-of-pile facilities. Granular triple super-
phosphate production should experience an average annual growth of 4
15
percent from 1975 to 1980.
There appears to be a trend toward larger production facilities in
the phosphate fertilizer industry. Average plant life is from 10 to 15
years and older plants are generally replaced by larger ones employing
the latest proven technology. A number of small experimental plants
have been built that produce such products as ultraphosphoric acid (83
percent ^2^5)' ammoir>UIT1 polyphosphate (15-61-0, NPK content) and high
analysis superphosphate (54 percent ^2®$) l)ut t'11s experimental technology
has not yet been applied to large scale production. All indications are
that the phosphate fertilizer industry will continue to grow rapidly
throughout the 1970-1980 decade.
3.6 PRICES
Price competition in the fertilizer industry has been very intense
historically because of the large number of participants in all facets
of manufacturing—basic chemical producers, manufacturers of mixed
fertilizers, blenders, and retailers. No one chemical producer can be
said to be a price leader. The participation of farm cooperatives in the
manufacturing segment of fertilizers, including the basic chemicals, un-
doubtedly has been a steadying factor on prices, minimizing cyclic
fluctuations.
3-29
-------
List prices are available for (agricultural grade) wet-process
phosphoric acid, triple superphosphate (run-of-pile and granular),
diammonium phosphate, and superphosphoric acid in the Chemical Marketing
Reporter published by Snell Publishing Company of New York. These
prices are not firm indicators of actual prices paid, however, since
discounts, variability in credit terms to buyers, and service fees
combine to determine the realized price available to the producer.
The long term profiles of wholesale prices for granular triple
superphosphate and diammonium phosphate are presented in Figure 3-5.
The estimates of prices realized by manufacturers are plotted against the
ranges of listed quotations for the same products for 1971 and 1972.
The spreads in prices reflect the difference in quotations by various
manufacturers at any given time. No long term profiles of prices are
available for wet-process phosphoric acid, superphosphoric acid, and
triple superphosphate.
July 1974 phosphate fertilizer list prices are presented in
Table 3-10. The prices presented later in the text (Table 7-1) reflect
estimated averages for November 1974 developed from a more recent
economic study. These averages reflect more closely prices realized
by the producers and will be used in measuring the economic assessment
of emission guidelines in Section 7.
3-30
-------
FIGURE 3-5. WHOLESALE PRICF.S FOR TRIPLE SUPERPHOSPHATE AND DIAMMONIUM PHOSPHATE
.
' 0
S-
i OJ
—04
-
D.
fO
t/>
QJ
...o
40
1964
— .
4
.
1 •
'
.... J
1 • t
i ~~
•
965 ; 1
i
:
.._...;..... j-_ ...j ......
: i
! •
966 •• 19fe7 19
. - i... YEA
68 19
R '.'..'.
i .
i [
:
59. 19
. ..
. . . ! .
: | : i. . i
70 .19
:.. • i . : •
i
... !
7i: . i 19;
• ' • 1
-
| • • • 1 •' • 1 • • :
• i ! ! :
• • . i ; !
-------
TABLE 3-10. SUMMARY OF LIST PRICES AS OF JULY 1974 AND BASIS FOR PHOSPHATE QUOTATION19
I
o<
ro
Commodity
Wet-process phosphoric acid (WPPA)
Superphosphoric acid (SPA)
Diammonium Phosphate (DAP)
Run-of-Pile Triple Superphosphate
Granular Triple Superphosphate
Price
$ per actual ton)
$105
$150 - $158
S145 - $165
$38 - $86.50
$55 - $91
Production Quality
52-54% P205
7Q% P205
18%N
46% P205
I
46% P205 nin
46% P205 min
Quotation Basis
Delivered in Tanks,
F.O.B. Florida works
Same as WPA
Bulk Delivered, Railroad
car lots, F.O.P. Florida
Same as PAP
Same as DAP
-------
3.7 WORLD STATISTICS ON PgOg
The levels of crop yields per acre have greatly increased during
the past generation. This increase has depended upon the generous
application of fertilizers containing the elements phosphorus, nitrogen,
and potassium. No two of these elements together could maintain high
crop levels; therefore, plentiful application of P205 will continue to
be necessary even to maintain food production at its current level.
Table 3-11 shows U.S. consumption of phosphate fertilizer-expressed
as P205 and the corresponding consumption for the entire world is given
for comparison. The data from the reference are adapted to this table
and are rounded off.
Phosphate fertilizer is made almost entirely from phosphate rock
and this is the only practical source for the quantities required.
Table 3-12 shows the total known world reserves of phosphate rock.
The United States has 30 percent of the supplies which are considered
mineable and beneficiable by current technology. The Arab Nations
possess 50 percent of world reserves and the Soviet Union has an
additional 16 percent. It must not be inferred that reserves within
a country are uniform in quality; the higher grades are mined first, and
successfully poorer grades follow at increased energy consumption and
cost rates.
3-33
-------
TABLE 3-11
UNITED STATES AND WORLD CONSUMPTION OF PHOSPHATE FERTILIZER
Fiscal
Year Consumption of Phosphate Fertilizer Million Short Tons
U.S. World
1950 1.950 6.45
1955 2.284 8.33
I960 2.572 10.52
1965 3.512 15.03
1970 4.574 20.40
1975 5.800*
'Estimated
3-34
-------
TABLE 3-12
WORLD RESERVES OF PHOSPHATE ROCK 2°
Country Million Short Tons
French Morocco 23,500
U.S. 16,250
U.S.S.R. 8,500
Tunisia 2,240
Algeria 1,120
Brazil 670
Peru 500
Egypt 220
Togo 130
Spanish Sahara 110
Islands - Pacific & Indian Ocean 45
Senegal 45
Other Countries 800
3-35
-------
3.8 REFERENCES
1. Harre, E.A. Fertilizer Trends 1969. Tennessee Valley
Authority. Muscle Shoals, Alabama. 1970. p. 37.
2. David, Milton L.t J.M. Malk, and C.C. Jones. Economic
Analysis of Proposed Effluent Guidelines for the Fertilizer Industry.
Development Planning and Research Associates, Inc. Washington,
D.C. Publication Number EPA-230-1-73-010. November 1973. p.1-8.
3. Harre, E.A. Fertilizer Trends 1973. Tennessee Valley Authority.
Muscle Shoals, Alabama. 1974. p. 5,7.
4. 1973 Directory of Chemical Producers, United States of America.
Stanford Research Institute. Menlo Park, California. 1973.
p. 417-418, 765-766, 860.
5. Osag, T. Written communication from Mr. T.A. Blue, Stanford
Research Institute. Menlo Park, California. November 29, 1973.
6. Blue, T.A. Phosphorous and Compounds. In: Chemical Economics
Handbook. Menlo Park, Stanford Research Institute, 1973.
p. 760.4003A - 760.4003E, 760.5003B - 760.5003K.
7. Beck, L.L. Recommendations for Emission Tests of Phosphate
Fertilizer Facilities. Environmental Protection Agency. Durham,
North Carolina. September 28, 1972. p. 12, 13, 16.
3-36
-------
8. Initial Analysis of the Economic Impact of Water Pollution
Control Costs upon the Fertilizer Industry. Development Planning
and Research Associates, Inc. Manhattan, Kansas. Contract No.
68-01-0766. November 1972.
9. Reference 3, p. 22.
10. Reference 3, p. 16.
11. Reference 2, p. 111-34.
12. Reference 2, p. 111-38.
13. Reference 3, p. 19.
14. Reference 3, p. 21, 22.
15. Bunyard, F.L. and P.A. Boys. The Impact of New Source Performance
Standards upon the Phosphate Fertilizer Industry. Environmental
Protection Agency. Durham, North Carolina. August 25, 1973.
16. Striplin, M.M. Jr. Production by Furnace Method. In: Phosphoric
Acid, Vol. 1., Slack, A.V. (ed). New York, Marcel Dekker, Inc.,
1968. p. 1008.
17. Reference 2, p. 111-49.
18. Chemical Marketing Reporter. June 1971 through December 1972.
19. Chemical Marketing Reporter. July 22, 1974.
20. Mineral Facts and Problems. Bulletin 630. United States Bureau
of Mines. 1965.
3-37
-------
4. PHOSPHATE FERTILIZER PROCESSES
4.1 INTRODUCTION.
The phosphate fertilizer Industry uses phosphate rock as its
major raw material. After preparation, the rock Is used directly 1n
the production of phosphoric acid, normal superphosphate, triple
superphosphate, nitrophosphate, electric furnace phosphorous and
defluorinated animal feed supplements. In addition to those products
made directly from phosphate rock, there are others that rely on
products prod.uced from phosphate rock as a principal ingredient.
Figure 4-1 Illustrates the major processing steps used to transform
phosphate rock into fertilizer products and Industrial chemicals.
The primary objective of the various phosphate fertilizer processes
is to convert the fluorapatite (Ca10(P04)gF2) in phosphate rock to soluble
P205, a form readily available to plants. Fluorapatite is quite
insoluble in water and, in most farming situations, 1s of little
value as a supplier of nutrient phosphate. The most common method
of making the P205 content of phosphate rock available to plants is
by treatment with a mineral acid - sulfurlc, phosphoric, or nitric.
Table 4-1 lists the available PgOg content of several phosphate
fertilizers. Available P20g is defined as the percent soluble PgOg
in a neutral citrate solution.
4-1
-------
FIGURE 4-1. MAJOR PHOSPHATE ROCK PROCESSING STEPS
PHOSPHATE
ROCK
Defluorination
Grinding
Acidulation (H2S04)
Acidulation (HNO3)
Acidulation (H3PO4)
Elemental
Phosphorus
Phosphoric
Acid
Various
ANIMAL FEEDS
FERTILIZERS:
Direct Application
Normal Superphosphate
Nitric Phosphates
Triple Superphosphate
Ammonium Phosphates
Direct Application
INDUSTRIAL AND
FEED CHEMICALS
4-2
-------
TABLE 4-1. P205 CONTENT OF PHOSPHATE FERTILIZERS2
FERTILIZER PERCENT SOLUBLE
Normal Superphosphate 16-22
Triple Superphosphate 44 -,47
Monammonlum Phosphate 5?.
D1 ammonium Phosphate 46
4.2 WET PROCESS PHOSPHORIC ACID MANUFACTURE.
Phosphoric add is an Intermediate product in the manufacture
of phosphate fertilizers. It is subsequently consumed in the
production of triple superphosphate, ammonium phosphates, complex
fertilizers, superphosphorlc acid and dicalcium phosphate.
Most current process variations for the production of wet-
process phosphoric acid depend on decomposition of phosphate rock by
sulfuric acid under conditions where gypsum (CaSO, • 2H20) is
precipitated. These variations are collectively referred to as
dihydrate processes since the calcium sulfate is precipitated as
the dihydrate (gypsum). Calcium sulfate can also be precipitated
in the semihydrate (Ca SO^ • 1/2 H^O) and anhydrite (CaSO^) forms.
Processes which accomplish this are commercially less important than
the dihydrate processes, however, since they require more severe
operating conditions, higher temperatures, and a greater degree of control.
4-3
-------
The overall reaction In the dlhydrate processes is described by the
following equation. (4-1)
(P04)6 F2 + 30H2S04 + S102 + 58H20 * 30CaS04 ' 2
18H3P04 +
In practice, 93 or 98 percent sulfuric acid is normally used for
digestion of the rock. Calcium sulfate precipitates, and the liquid
phosphoric acid 1s separated by filtration.
Several variations of the di hydrate process are currently in use
by the phosphate fertilizer Industry. The Dorr-Oliver, St. Gobain,
Prayon, and Chemico processes are among the better known designs.
Fundamentally, there is little difference among these variations -
most differences are in reactor design and operating parameters.
Figure 4-2 presents a flow diagram of a modern wet-process phosphoric
acid plant.
Finely-ground phosphate rock is continuously metered
into the reactor and sulfuric acid is added. Because
the proper ratio of acid to rock must be maintained as closely as
possible, these two feed streams are equipped with automatic controls.
Some years ago, plants were built with several separate reaction
tanks connected by launders, which are channels for slurry flow. The
tendency now is to use a single tank reactor that has been divided
Into several compartments. In most of these designs, a series of
baffles is used to promote mixinp of the reactants.
4-4
-------
-^
I
en
• ASM
NATCH
GYPSUM
POND WATER
CTPJUM
TO fOHO
•TO 5CRUB8ER
HTOROFLUOSILICIC AGO
FIGURE 4-2. FLOW DIAGRAM ILLUSTRATING A WET-PROCESS PHOSPHORIC ACID PLANT
-------
The single-tank reactor (Dorr-Oliver design) Illustrated in
Figure 4-2 consists of two concentric cylinders. Reactants
are added to the annul us and digestion occurs in this outer compart-
ment. The second (central) compartment provides retention time for
gypsum crystal growth aid prevents short-circuiting of rock.
The Prayon reactor has been a widely used design. This process
variation involves the use of a rectangular, multicompartment attack
tank - typically 10 compartments - as indicated in Figure 4-3. The
compartments are arranged in two adjacent rows with the first and
tenth located at one end of the reactor and the fifth and sixth at
the other. In operation, digestion of the rock occurs in the first
four compartments, the next four provide retention time for the growth
of gypsum crystals, the ninth supplies feed for the vacuum flash
cooler, and the tenth receives the cooled slurry from the flash
cooler and splits the flow between the filter and a recycle stream.
BAROMCTRIC
CONDENSER
OCK
^1
\
^
1
V
:
> :
H
J
!
,sn
>
o
I
5
*
f
' II
FUME
1
r
**-* 4
1
_y
*
^»-
5
^
*
-]
ATE
|
1
TO
5EWI
L
5 ^
^
|
R
SCRUBBEI
:R
lr\
* 6
n
^SPLITTER
TANK
TO ^JL
STACK r ^
FLASH
f COOLER
:M k -1
r^
Q COLO
^ SLURR
i^—
7 W fl ^^_
ATTACK TANKS (10)
f WEAK ACID
ZZVPjO,
TO
FILTER
RECYCLE
FROM FILTER
FIGURE 4-3. FLOW DIAGRAM FOR PRAYON REACTOR3
4-6
-------
Proper crystal growth depends on maintaining sulfate ion
concentration within narrow limits at all points in the reaction
slurry. The proper sulfate ion concentration appears to be slightly
more than 1.5 percent. Lower levels give poor crystals that are
difficult to filter; higher concentrations interfere with the reaction
4
by causing deposition of calcium sulfate on unreacted rock.
Good reactor design will prevent sudden changes of sulfate ion concen-
tration, will maintain the sulfate ion concentration and temperature
near optimum, and will provide sufficiently long holdup time to allow
growth of large, easily filterable crystals without the formation of
excessive crystal nuclei.
Impurities In small amounts often have a marked effect on crystal
growth when they are present in a medium where crystallization Is
taking place. Usually this impurity effect is detrimental. Such
Impurities are likely to cause crystal fragmentation, small crystal
size, or a shift to needles or other hard-to-filter forms.
Concentrated sulfurlc acid is usually fed to the reactor. If
dilute acid is used, its water content must be evaporated later. The
only other water entering the reactor comes from the filter-wash
water. To minimize evaporation costs, it is important to use as little
wash water as is consistent with practical P205 recoveries.
Considerable heat of reaction 1s generated in the reactor and
must be removed. This is done either by blowing air over the hot
slurry surface or by vacuum flash cooling part of the slurry and
4-7
-------
sending it back into the reactor. Modern plants use the vacuum
flash cooling technique illustrated in Figures 4-2 and 4-3,
The reaction slurry is held in the reactor for up to 8 hours,
depending on the type rock and the reactor design, before being sent
to the filter. The most common filter design in use is the rotary
horizontal til ting-pan vacuum filter shown in Figures 4-2 and 4-4.
This type unit consists of a series of individual filter cells mounted
on a revolving annular frame with each cell functioning essentially
like a Buchner funnel. Figure 4-4 illustrates the operating cycle
of a rotary horizontal tiltinq-pan filter.
Product slurry from the reactor is Introduced into a filter cell
and vacuum 1s applied. After a dewatering period, the filter cake
undergoes 2 or 3 stages of washing with progressively weaker solutions
of phosphoric add. The wash-water flow is countercurrent to the
rotation of the filter cake with heated fresh water used for the
last wash, the filtrate from this step used as the washing liquor for
the preceding stage, etc.
After the last washing, the cell is subjected to a cake
dewatertng step and then inverted to discharge the gypsum. Cleaning
of the filter media occurs at this time. The cell is then returned
to Its upright position and begins a new cycle.
4-8
-------
CAKE WASHING
CAKE OCWATERiNQ
FEED SLURRY
CAKE DISLODGING
AND DISCHARGING
FIGURE 4-4. OPERATING CYCLE OF ROTARY HORIZONTAL
TILTING PAN FILTERS
The 32 percent acid obtained from the filter generally needs
concentrating for further use. Current practice is to concentrate
it by evaporation in a two or three-stage vacuum evaporator system.
Wet process acid is usually not concentrated above 54 percent, because
the boiling point of the acid rises sharply above this concentration.
Corrosion problems also become more difficult when concentration
exceeds 54 percent. In the evaporator, Illustrated in Figure 4-2,
provision 1s made for recovery of fluoride as fluosilicic acid. This
recovery feature is not necessary to the evaporation and its
inclusion is a matter of economics. Many evaporation plants have not
installed this device.
4-9
-------
Table 4-2 shows a typical analysis of commercial wet-process
phosphoric acid. In addition to the components listed in Table 4-2,
other trace elements are commonly present. Impurities, those listed
in Table 4-2 as well as trace elements, affect the physical properties
of the acid. Commercial wet-process acid has a higher viscosity than
pure orthophosphoric acid of the same concentration. This tends to
increase the difficulty of separating the calcium sulfate formed
during acidulation of the phosphate rock.
TABLE 4-2
COMPONENTS OF TYPICAL WET-PROCESS ACID7
Component
P2°5
CA
Fe
Al
Mg
Cr
V
H-O and other
Weight, %
53.4
0.1
1.2
0.6
0.3
0.01
0.02
37.56
Component
Na
K
F
so3
Si02
C
solid
Weight, %
0.2
0.01
0.9
1.5
0.1
0.2
2.9
4-10
-------
4.3 SUPERPHOSPHORIC ACID MANUFACTURE.
Superphosphoric acid (also referred to as polyphosphoric acid)
is a mixture containing other forms of phosphoric acid in addition
to orthophosphoric acid (H3P04). At least one-third of the P205
content of Superphosphoric acid are polyphosphates such as pyro-
phosphoric acid (H4P207), tripolyphosphoric acid (H5P301Q). tetra-
polyphosphorlc acid (HgP4013), etc. Pure orthophosphoric acid
converts to polyphosphates when the P20g concentration exceeds 63.7
Q
percent. Concentrating above this level dehydrates orthophosphoric
acid to form polyphosphates. Superphosphoric acid can have a minimum
of 65 percent PgOg which represents an orthophosphoric concentration of
just over 100 percent. Commercial Superphosphoric acid, made by
concentrating wet-process or furnace orthophosphoric acid, normally
o
has a PgOg concentration between 72 and 76 percent. Table 4-3 compares
the properties of 76 percent Superphosphoric acid to 54 percent ortho-
phosphoric acid.
TABLE 4-3. COMPARISON OF ORTHOPHOSPHORIC TO SUPERPHOSPHORIC ACID9
Orthophosphoric Superphosphoric
Acid Acid
Concentration of Commercial
Acid, % P205 54 76
H3P04 equivalent, % 75 105
Pounds P20g/gal 7.1 12.2
Percent of P205 as Polyphosphates 0 51
Viscosity, CP
at 100°F 12 400
at 200°F 4 45
4-11
-------
Superphosphoric acid has a number of advantages over the more
dilute forms of phosphoric acid, the foremost being economy In
shipping. Since phosphoric acid of any concentration Is usually
transported at the same price per ton, a 40 percent savings In freight
per unit weight of P^Oc results when superphosphorlc acid Is transported
g
Instead of ordinary phosphoric add. Superphosphorlc acid may be
diluted to orthophosphorlc acid at its destination.
In addition to freight savings, superphosphorlc acid offers
several other advantages. It 1s less corrosive than orthophosphorlc
acid, which reduces storaae costs. Finally, the con-
version of wet-process acid has a special advantage. Unlike furnace
acid, wet-process phosphoric acid contains appreciable quantities
of Impurities which continue to precipitate after manufacture
and form hard cakes 1n pipelines and storage containers. When wet-
process acid 1s converted to superphosphorlc add, the polyphosphates
sequester the Impurities and prevent their precipitation. Therefore
shipment and storage of wet-process acid 1s far more attractive after
conversion to superphosphorlc add.
Two cornnerdal processes are used for the production of super-
phosphoric add: submerged combustion and vacuum evaporation. The
submerged combustion process was pioneered by the TVA; dehydration
of the acid Is accomplished by bubbling hot combustion gas through a pool
•of the acid.
4-12
-------
The hot gases are supplied by burning natural gas in a
separate chamber. The combustion gases are diluted
with air to maintain a gas temperature of 1700°F for intro-
duction into the acid evaporator. Figure 4-5 depicts an
acid evaporator and Figure 4-6 the general process. After
passage through the acid, the hot gases are sent to a sepa-
rator to recover entrained acid drpplets and then to emission
control equipment.
Clarified acid containing 54 percent P?05 Is continuously
fed to the evaporator from storage, and acid containing 72 percent
PpOj- is withdrawn from the evaporator to product holding
tanks. Cooling is accomplished by circulating water through
stainless steel cooling tubes in the product tanks. The process
can be controlled by regulating the natural gas and air flows to
the combustion chamber, the dilution air to the combustion stream,
or the amount of acid fed to the evaporator.
FIGURE 4-5. TVA EVAPORATOR FOR PRODUCING SUPERPHOSPHORIC
ACID
HOT
GASES
CARBON
INSERT
AGIO
FEED
•- PRODUCT
DISCHARGE
4-13
-------
FUEL
TEMPERING
AIR
AIR
COMBUSTION
CHAMBER
EVAPORATOR
54°, CLARIFIED
ACID
SEPARATOR
m
WATER
r
CONTROLS
ACID MIST, SiF,,
HF
Fl' 722 ACID
!
PRODUCT'
WATER STORAGE
ACID COOLER
FIGURE 4-6. SUBMERGED COMBUSTION PROCESS FOR PRODUCING
SUPERPHOSPHORIC ACID
4-14
-------
In addition to the TVA process, a number of other submerged
combustion processes have been developed. Among them are the
Collier Carbon and Chemical Process, the Albright and Wilson Process,
the Occidental Agricultural Chemicals Process, and the Armour Process.
The latter process produces superpnosphoric acid of about 83 percent
PgOj- which 1s sometimes referred to as ultraphosphoric acid. The
Occidental and TVA designs are currently in use in the United States.
Vacuum evaporation is by far the more Important commercial
method for concentrating wet-process phosphoric acid to superphosphoric
acid. There are two commercial processes for the production of super-
phosphoric add by vacuum evaporation:
1. The falling film evaporation process (Stauffer Chemical
Co.) and
2. The forced circulation evaporation process (Swenson
Evaporator Co.).
Feed acid clarification 1s required by both processes. Clarification
is usually accomplished by settling or by a combination of ageing and
settling.
In general, both processes are similar 1n operation. Both use
high-vacuum concentrators with high-pressure steam to concentrate add
to 70 percent PpOg and both Introduce feed acid into a large volume
of recycling product acid to maintain a highly concentrated process
acid for lower corrosion rates. In both systems, product acid
iis pumped to a cooler before being sent to storage or shipped.
Figures 4-7 and 4-8 show the Stauffer and Swenson processes
respectively. The Stauffer process adds 54 percent feed acid to
the evaporator recycle tank where it mixes with concentrated product
4-15
-------
FIGURE 4.7 STAUFFER EVAPORATOR PROCESS
10
High-pressure
steam from
package boiler
FALLING-FILM
EVAPORATOR
Condensate,«
to package
steam boiler
'.'.'el-process
phosphoric
ocid(S4%PzO,)
Concentrated
acid
To ejectors
FEED TANK
EVAPORATOR
RECYCLE
TAMK
BAROMETRIC
CONDENSER
Superphosphoric
acid
Coolant
discharge
Superphosphoric
acid
(68-72%PtOj)
FIGURE 4-8 SWENSON EVAPORATOR PROCESS10
TO AIR HECTOR
VArnt
STOBACt
4-16
-------
acid. This mixture Is pumped to the top of the evaporator and
distributed to the Inside wall of the evaporator tubes. The
acid film moves down along the inside wall of the tubes receiving
heat from the steam on the outside. Evaporation occurs and the
concentrated acid is separated from the water vapor in a flash
chamber located at the bottom of the evaporator. Product acid flows
to the evaporator recycle tank and vapors to the barometric condenser.
To insure minimum PpOc loss, the separator section contains a mist
eliminator to reduce carryover to the condenser.
The Swenson process, uses acid in the tube side of a forced
circulation evaporator (Figure 4-8). Feed acid containing 54 percent
PpOg is mixed with concentrated acid as it is pumped into the
concentrator system. As the acid leaves the heated tube bundle
and enters the vapor head, evaporation occurs and the acid disengages
from the water vapor. The vapor stream is vented to a barometric con-
denser while the acid flows toward the bottom of the vapor hrad tank
where part of it is removed to the cooling tank and the remainder is
recycled to the tube bundle.
4.4 DIAMMONIUM PHOSPHATE MANUFACTURE.
D1ammonium phosphate is obtained by the reaction of .ammonia
with phosphoric acid. In addition to containing the available
phosphate of triple superphosphate, diammonium phosphate has the
advantage of containing 18 percent nitrogen from ammonia.
4-17
-------
The Importance of di ammonium phosphate produced by wet-process
acid has increased as it continues to replace normal superphosphate as
a direct application material. The shift to diammonium phosphate is
most evident on the supply side. Ammonium phosphate production now
exceeds 2.7 million tons of PgOg a year while normal superphosphate
production has declined 32 percent since 1968 to 0.6 million
tons. Increasing amounts of diammonium phosphates are also being
used in bulk blends as these Increase ir. popularity.
The increased use of diammonium phosphate is attributable to
several factors. It has a high water solubility, high analysis
(18 percent nitrogen and 46 percent available P205)» good physical
characteristics, and low production cost. In addition, the phosphate
content of diammonium phosphate (46 percent) is as high as triple-
superphosphate, so by comparison, the 18 units of nitrogen can be
shipped at no cost.
The TVA process for the production of diammonium phosphate
appears to be the most favored with several variations of the original
design now in use. A flow diagram of the basic process is shown in
Figure 4-9.
Anhydrous ammonia and phosphoric acid (about 40 percent P?0R)
are reacted in the preneutralizer using a NH~ / H,PO^ mole ratio
of 1.35. The primary reaction is as follows:
2 NH3 + H3P04 * (NH4)2 HP04 (4-2)
The use of a 1.35 ratio of NhL / H-jPO* allows evaporation to a water
4-18
-------
TO AIR POLLUTION
CONTROL SYSTEM
PHOSPHORIC
ACID
r
AMMONIA, SiF4, HF
i i
n
DRYER
CYCLONES
V
7
COOLER
CYCLONES
u
\\
3
-
! \
DRYER
AT'ONIA
FLLD
SCREE:;
/\
./ !
PARTICOLATE
i
^ /
OOVCRSIZE
V'LL
FIGURE 4-9. TVA DIAMMONIUM PHOSPHATE PROCESS.
-------
content of 18 to 22 percent without thickening of the DAP slurry to
a nonflowing state. The slurry flows Into the ammoniator-qranulator
and is distributed over a bed of recycled fines. Ammoniation to the
required mole ratio of 2.0 takes place in the granulator by injectinq
ammonia under the rolling bed of solids. It is necessary to feed excess
ammonia to the granulator to achieve a 2.0 mole ratio. Excess
ammonia and water vapor driven off by the heat of reaction are directed
to a scrubber which uses phosphoric acid as the scrubbing liquid. The
ammonia is almost completely recovered by the phosphoric acid scrubbing
liquid and recycled to the preneutrallzer. Solidification occurs
rapidly once the mole ratio has reached 2.0 making a low solids recycle
ratio feasible.
Granulated diammonium phosphate Is next sent to the drier,
then screened. Undersized and crushed oversized material are
recycled to the granulator. Product sized material Is cooled and
sent to storage.
In addition to the TVA process, a single-step drum process
designed by the Tennessee Corporation and the Dorr-Oliver granular
process are used for the manufacture of diammonium phosphate. The
single step drum process 1s designed so that the entire neutralization
reaction occurs in the granulator drum - phosphoric add 1s fed
directly onto a rolling bed of fines while the ammonia 1s injected
under the bed. In the case of the Dorr-Oliver design, a two-stage
continuous reactor is used for the neutralization step. The reaction
slurry 1s then combined with recycled fines in a pugmlll.
4-20
-------
4.5 TRIPLE SUPERPHOSPHATE MANUFACTURE AND STORAGE.
Triple superphosphate* also referred to as concentrated
superphosphate, Is a product obtained by treating phosphate rock
with phosphoric add. According to the grade of rock and the
strength of add used the product contains from 44 - 47 percent
available P205.
Like diammonium phosphate, the Importance of triple super-
phosphate has Increased with the declining use of normal super-
phosphate. Triple superphosphate production now Is around 1.7 million
tons of P205 which is more than double that of normal super-
phosphate.11 It is used in a variety of ways - large amounts are
Incorporated Into high analysis blends, some are ammoniated. but
the majority are applied directly to the soil.
4.5.1 Run-of-Pile Triple Superphosphate Manufacture and Storaqe
Figure 4-10 Is a schematic diagram of the den process for the
manufacture of run-of-pile triple superphosphate. Phosphoric
add containing 52 - 54 percent P20g is mixed at ambient tempera-
ture with phosphate rock which has been ground to about 70 percent
minus 200 mesh. The majority of plants in the United States use the
TVA cone mixer which is shown in Figure 4-11. This mixer has
no moving parts and mixing is accomplished by the swirling action
of rock and acid streams introduced simultaneously into the cone.
The reaction that takes place during mixing is represented by the
following equation:
4-21
-------
PHOSPHATE
ROCK
PHOSPHORIC
ACID_ _ T^CONTROLS
I
rv>
ro
SiF4, PARTICULATE
FIGURE 4-10, RUN-OF-PILE TRIPLE SUPERPHOSPHATE PRODUCTION AND STORAGE
-------
Ca
10
F2 + 14H3P04 + 10H2° •* 10CaH4(P04)2 *
2HF
After mixing, the slurry is directed to a "den" where
solidification occurs. Like mixers, there are a number of den
designs, one of the most popular continuous ones being the Broadfield.
This den is a linear horizontal slat belt conveyor mounted on rollers
with a long stationary box mounted over it and a revolving cutter at
the end. The sides of the stationary box serve as retainers for the
slurry until it sets up.
FIGURE 4-11. TVA CONE MIXER
ClD LINC
4-23
-------
The solidified slurry which exits from the den 1s not a
finished product. It must be cured - usually for 3 weeks or more -
to allow the reactions to approach completion. The final curing stage
is depicted in Figure 4-10 by the conveying of product to the sheltered
storage pile.
4.5.Z Granular Triple Superphosphate Manufacture and Storage
Two processes for the direct production of granular triple
superphosphate will be briefly presented. A third process uses
cured run-of-pile triple superphosphate, treats it with water and
steam in a rotary drum, then dries and screens the product. A
large amount of granulated triple superphosphate is produced by
this method but product properties are not as good as that
produced by other processes.
The TVA one-step granular process 1s shown in Figure 4.12. In
this process, phosphate rock, ground to 75 percent below 200 mesh,
and recycled process fines are fed Into the acidulation drum along
with concentrated phosphoric acid and steam. The use of steam helps
accelerate the reaction and ensure an even distribution of moisture in
the mix. The mixture is discharged into the granulator where solidifi-
cation occurs, passes through a rotary cooler, and Is screened. Over-
sized material Is crushed and returned with undersized material to
the process. The reaction for the process 1s the same as that of
ruh-qf-pile triple superphosohate.
4-24
-------
1 SiF4 *S1F4,
! i
W2lcK*TV 1 h 1 "CYCLED FINES !
\ — TS; — 7
.>/>/ 1 STEAM
CT _> i i
r-^C
sTEAM-n L-P- u.,.^1
ACIOUl ATIO^
PHOSPHOillC HtATCR OHUM
U(A] 1 — i ~_
JMETER M— CONO
' *ca
PUMP
^
SCREENS ^
i
^L
1
J
1
I 1
— r4-> 1
n_,
GliANULATOH |l
r"i
COOLER
l~ ~
/ 1 | OVE
RSIZE
— 1
r- — CAGE
FINES JL. MILL
&-O
ROLL 1 , f
CRUSHER
- •*
PARTICULATE
->S1F.t PARTICULATE
_ >SU4. PARTICULATE
PRODUCT
1 to- TO
STORAGE
FIGURE 4-12. TVA ONE-STEP PROCESS FOR
GRANULAR TRIPLE SUPERPHOSPHATE
4-25
-------
The Dorr-Oliver slurry granulation process is shown in
Figure 4-13. In this process, phosphate rock, ground to an
appropriate fineness is mixed with phosphoric acid (40% P2°5) in a
series of mixing tanks. A thin slurry is continuously removed, mixed
with a large quantity of dried, recycled fines in a pugmill mixer
(blunger), where it coats out on the granule surfaces and builds up
the granule size. The granules are dried, screened, and mostly (about
80 percent) recycled back into the process. Emissions from the drier
and screening operations are sent to separate cyclones for dust removal
and collected material is returned to the process.
After manufacture, granular triple superphosphate 1s
sent to storage for a short curing period. Figure 4-14 illustrates
the activities in the storage building. After 3 to 5 days, during
which some fluorides evolve from the storage pile, the product is
considered cured and ready for shipping. Front-end loaders move the
GTSP to elevators or hoppers where it is conveyed to screens for size
separation. Oversize material is rejected, pulverized, and returned
to the screen. Undersize material is returned tc the GTSP production
plant. Material within specification Is shipped as product.
4-26
-------
PHOSPHATE ROCK
i
ro
PHOSPHORI-:: ACID
S1F.
TO AIR POLLUTION
C0.1TROL SYSTEM
SiF4. PARTICULATE Sii F4, PARTICULATE
n
DUST II DRYER I
•CYCLONE, CYCLONE
-1 r •
V V
OVERSIZE
SCREEN
PUG LULL
GRANULATOR
(ROTARY TYPE
ALSO USED)
PRODUCT
SCREEN
OVERSIZE
WILL
PRODUCT TO COOLING
AMD STORAGE
ACIDULATORS
FIGURE 4-13. DORR-OLIVER SLURRY GRANUUTION PROCESS FOR TRIPLE SUPERPHOSPHATE
-------
S1F4, PARTICULATE
i
rv>
CO
GTSP FROM
PROCESS
^'^-'^'-:^i-"';'^r'i'l'V''"'"^"'- • • ••''••'•
STORAGE PILE ^'':',-"
"'
1
ce
0
A
SCREENS}
r
T1 »
MILLS
;
UJ
_l
LU
SHIPPING
FIGURE 4-14. GRANULAR TRIPLE SUPERPHOSPHATE STORAGE.
-------
4.6 REFERENCES
1. Blue, T.A. Phosphate Rock. In: Chemical Economics Hand-
book. Menlo Park, Stanford Research Institute, 1967.
p. 760.2011 F.
2. Slack, A.V. Fertilizers. In: Klrk-Othmer Encyclopedia of
Chemical Technology, Vol. 9V Standen, A. (ed). New York",
John Uiley & Sons, Inc., 1966. p. 100, 106, 125.
3. Slack, A.V. Dihydrate Processes - Prayon. In: Phosphoric Acid,
Vol. 1, Slack, A.V. (ed). New York, Marcel Dekker, Inc.,
1968. p. 254.
4. Noyes, R. Phosphoric Acid by the Wet Process. Park Ridge,
Noyes Development Corporation, 1967. p. 10-11.
5. Roos, J.T. Commercial Filters - Bird-Prayon. In: Phosphoric
Acid, Vol. I, Slack, A.V. (ed.). New York, Marcel Dekker,
Inc., 1968. p. 446.
6. Atmospheric Emissions from Wet Process Phosphoric Acid Manufacture.
National Air Pollution Control Administration. Raleigh, N.C.
Publication Number AP-57. April 1970. p. 13-14.
7. Reference 6, p. 11.
8. Reference 4, p. 174.
9. Striplin, M.M., Jr. Production by Furnace Method. In:
Phosphoric Acid, Vol. I., Slack, A.V. (ed.). New York,
Marcel Dekker, Inc., 1968. p. 1008.
4-29
-------
10. Reference 4, p. 222.
11. Harre, E.A. Fertilizer Trends 1973. Tennessee Valley
Authority. Muscle Shoals, Alabama. 1974. p. 22.
4-30
-------
5. EMISSIONS
5.1 NATURE OF EMISSIONS.
In assessing the environmental effect of the emissions from
the various phosphate fertilizer processes, fluorides - which are largely
emitted in gaseous form, were considered to be the most significant
and were chosen for regulation as discussed in Section 1.2.
Gaseous fluorides emitted from phosphate fertilizer processes
are primarily silicon tetrafluoride (SiF.) and hydrogen fluoride
(HF) . The origin of these gases may be traced to the reaction
between phosphate rock and sulfuric acid represented by equation 4-1.
3Ca1Q (P04)6F2 + 30H2S04 + Si02 + 58H20 - (4-1)
30CaS04 • 2 H20 + 18 H3P04 + HgSiFg
Under the existing conditions of temperature and acidity,
excess fluosilicic acid decomposes as follows:
H2S1F60) * SiF4(g) + 2HF(g) (5-1)
Actually, the mole ratio of hydrogen fluoride to silicon tetra-
fluoride In the gases emitted during the decomposition of phosphate
rock change with conditions (e.g., the amount of excess silica
5-1
-------
in the reaction mixture) and Is seldom equal to the stoichlo-
metric value. At high levels of excess silica, the hydrogen
fluoride evolved will react to form silicon tetrafluoride according
to equation 5-2:
4HF + Si02 -* SiF4 + ZHgO (5-2)
At low concentrations of silica* emissions will be rich in
hydrogen fluoride.
Not all of the fluorides are driven off during the digestion
of the phosphate rock. A certain amount is retained in the product
acid depending upon the type of rock treated and the process used.
These fluorides can be emitted during the manufacture of super-
phosphoric acid, diammonium phosphate, or triple superphosphate.
Fluoride emissions from superphosphoric acid and diammonium
phosphate processes depend solely on the fluoride content of the
feed acid. In the manufacture of triple superphosphate, fluoride
emissions can also be attributed to the release of fluorides from
the phosphate rock. Calcium fluoride and silica in the rock react
with phosphoric acid to form silicon tetrafluoride according to the
2
following reaction :
2CaF2 + 4H3P04 + Si02 * SiF4 + 2CaH4(P04)2 • 2H20 (5-3)
Scrubbing with water is an effective fluoride control technique
because of the high water solubility of most gaseous fluorides.
5-2
-------
This straight- forward approach 1s somewhat complicated, however,
by the presence of silicon tetrafluorlde. Silicon tetrafluoride will
react with water to form hydrated silica (S1(OH)4) and fluosillclc
add (H2 SIFg) as Indicated by equation 5-4:
3S1F4 + 4 H20 -•> 2H2S1F6 + S1(OH)4 (5-4)
Hydrated silica precipitates forming deposits on control equipment
surfaces which plug passageways and tend to absorb additional
silicon tetrafluorlde. The nature of the precipitate, In the
presence of hydrogen fluoride, 1s temperature dependent. Below
125°F, the precipitate is 1n the form of a gel. Above this
temperature, 1t 1s a solid. Control systems should be designed
to minimize plugging and to allow removal of silica deposits.
Entrapment of scrubbing liquid must be kept to a minimum to
prevent the escape of absorbed fluorides. Fluorides can also
be emitted as parti cul ate from some fertilizer processes.
Particulate emissions can be effectively controlled by using
cyclones in combination wtth water scrubbers.
5.2 UNCONTROLLED FLUORIDE EMISSIONS.
5.2.1 Emissions from Wet-Process Phosphoric Acid Manufacture
Fluoride emissions from wet-process acid manufacture
gaseous silicon tetrafluoride and hydrogen fluoride. The reactor
Is the major source of fluoride emissions from the process accounting
for as much as 90 percent of the fluorides emitted from an uncontrolled
5-3
-------
plant. Additional sources are the filter, the filtrate feed and
seal tanks, the flash cooler seal tank, the evaporator system
hotwell, and the acid storage tanks. Table 5-1 lists reported
emission factors for the various sources. Fluoride emissions will vary
depending upon the type of rock treated and the process used.
Table 5-1 Fluoride Emissions from an Uncontrolled
Wet-Process Phosphoric Acid Plant4
Source Evolution Factor
(IbTF/ton Po(L
Reactor 0.04 - 2.2
liter 0.01 - 0.06
Miscellaneous(filtrate feed and up to 0.26
seal tanks, hotwells, etc.)
Modern reactors emit fluorides from two sources; the reaction
vessel and the vacuum flash cooler. The primary source is the
reactor tank, where silicon tetrafluoride and hydrogen fluoride are
evolved during the digestion of the phosphate rock.
To prevent an excessive temperature rise in the reactor, the
heat of reaction is removed by cycling a portion of the reaction
slurry through a vacuum flash cooler. Vapors from the cooler are
condensed in a barometric condenser and sent to a hot well while
the non-condensables are removed by a steam ejector and also vented
to the hot well. This arrangement is illustrated in Figure 4-2.
The majority of the fluorides evolved in the flash cooler are
absorbed by the cooling water in the barometric condenser. If air
cooling is utilized, fluoride evolution can be considerably areater
than indicated in Table 5-1.
5-4
-------
The filter 1s the second largest source of fluoride emissions.
Most of the fluorides are evolved at the points where feed add
and wash liquor are Introduced to the filter. These locations
are usually hooded and vented to the digester scrubber.
A third source of fluoride emissions Is the multiple effect
evaporator used to concentrate the phosphoric acid from 30 percent
P205 to 54 percent PgOg. It has been estimated that 20 to 40 percent
of the fluorine originally Introduced Into the process with the rock
Is vaporized during this operation. Most of these fluorides are
collected 1n the system's barometric condensers. The remainder
exit with the non-condensables and are sent to the hot well
which becomes the emission source for this operation.
In the plant design Illustrated In Figure 4-2. the vapor stream
from the evaporator Is scrubbed with a 15 to 25 percent solution
of fluoslHcIc add at a temperature at which water vapor, which would
dilute the solution. Is not condensed. The water vapor 1s then
removed by a barometric condenser before the non-condensables are
ejected from the system. Almost all of the fluoride 1s recovered
as by-product fluoslllclc acid.
In addition to the preceding sources of fluoride emissions,
there are several minor sources. These Include the vents from such
points as sumps, clarlflers, and acid tanks. Collectively, these
sources of fluoride emissions can be significant and are often
ducted to a scrubber.
5-5
-------
Table 5-2 illustrates a typical material balance for the
fluorine originally present in ohosohate rock. It should be
noted that the results in any given wet-orocess acid plant may differ
considerably from those shown in the table. Fluorine distribution
will depend upon the type of rock treated, process used, and kind of
operation prevailing.
TABLE 5-2
TYPICAL MATERIAL BALANCE OF FLUORIDE IN MANUFACTURE
OF WET-PROCESS PHOSPHORIC ACID
Fluoride Input
I F/1QO I Feed Rock
Feed
Fluoride Output
3.9
F/10C # Feed Rock
Product acid
Gypsum
Barometric condensers
Air*
1.0
1.2
1.67
0.03
Total
*
3.9
Typical emission from an uncontrolled plant.
Fluoride-bearing water from the barometric condensers as well as
the gypsum slurry is sent to the gypsum oond. In the gypsum pond,
silica present in the soil converts hydrogen fluoride to fluosilicates
Limestone or lime may be added to ponds to raise the pH and convert
fluoride to insoluble calcium fluoride. Fluoride associated with the
gycsum slurry ->s already in the insoluble forr-i before being sent EG
the oond.
5-6
-------
5.2.2 Lnissions fron Superphosc'-:oric Acid 'Manufacture
5.2.2.1 Submerged combustion process
T!;e direct contact evaporator is the major source of fluoride
emissions from the submerged combustion process. Fluoride
evolution is in tiie forn of silicon tetrafluoride and hydrogen fluo-
ride with a substantial portion ?s the latter. The amount of
fluorides evolved will depend on the fluoride content of the feed
acid and the final concentration of phosphoric acid produced. Feed
acid containing 54 percent P2r& has a typical fluoride content (as F)
of from 0.4 to 0.8 percent.
Control of evaporator off-pases is complicated by the presence of
large amounts of entrained ohosphoric acid - amounting to as nuch as
o
5 percent of the PgOg input to the concentrator. /".n entrainment
separator is used to recover acid and recycle it to the process. Some
entrained acid exits the separator, however, and tends to for?, a diffi-
cult to control acid aerosol. The formation of this aerosol can be
minimized by reducing the temperature of the combustion oases before
g
they contact the acid.
The acid sump and product Molding tank are secondary sources of
fluoride emissions from the submerged combustion process. These
emission points are identified in Figure 4-6. Uncontrolled emissions
from the submerged combustion process range frof 13 to 22 pounds of
fluoride per ton of P"- input.
5-7
-------
5.2.2.2 Vacuur evaporation orocess
Tlie ^aronatric condenser ,'ict^ell. tlio evaporator recycle tank,
and the product coolino tank are the three sources of fluoride
erissions from the vacuum evaporation nrooess. These emission noints
are identified in ^inures 4-7 ind ^-8. "ost of the fluorides
evolved during evaporation are absorbed by the cool in? v.-ater in the
barometric condensers resulting in a necjlicible emission to the
atmosphere frorr this source. Moncnndensables are elected fron the
condenser systen and sent to the hotv/ell alono v/ith the :ondenser
water. This results in the hotwell beconino the naior source of
emissions from the process. The evanorator recvcle tank and the
oroduct cooling tank are lesser sources of fluoride emissions.
Total erissions fron an uncontrolled plant are estinated at 0.005
oer ton P20g input.
5.2.3 LITissions from Diamnoniun Phosphate Manufacture.
Fluorides are introduced into the DAD nrocess with the wet orocess
pliosp.'ioric acid feed and are also evolved frorr the c'losohoric acid
scrubbing solution used to recover ammonia. Wet process acid v/hic'i
has been concentrated to 54 oercent PgCg typically contains n.l to 0.8
percent fluorides (as F) while filter acid (26-307. Pj/%.) v/ill contain
fron 1.8 to 2.0 oercent. ' D1iosphoric acid ccntainin" about 4T
percent f^Pr - obtained by nixir.n 5^ percent acid fron the
-------
"ajcr sources of fluoride em'ssions fron dirtT^oniu^ ohosnhnte
include the reactor, "ranulatcr, dryer, cooler, screens and
pills. Tho locations nf these emission points are de-Dieted in
Finure 4-9. Ventilation streams .from these sources are corrhined
for nurnoses of control accordin" to the follnv.'ino scheme: 1)
reactor-oranulator cases, 2) drvsr qases, and 3) cooler an? screening
qases.
Fluorides and arronia are the najor eniissions fro™ both the
reactor and the oranulator. Reactor-oranulator rases are treated
for annonia recovery in a scrubber that uses ohosohnric acid as
tha scrubber liquid. The phosohoric acid reacts with the ammonia and
the resulting product is recycled back to the process. Flur-rides
can be strinoed fror the nhosrshoric acid and a secnndar' scrubber is
usually nauired for fluoride control. Removal of evolved *lunridss
can be comolicated by their reaction with anmonia to forir a aarticu-
late.
Drier emissions consist of ammonia, fluorides, and oarticulate.
Gases are sent through a cyclone for oroduct recovery before beinq
treated for armonia or fluoride renoval. rddition?l fluorides csn
be striooed fron the ohosphoric acid scrubbing if arprni? recovery is
practiced.
Emissions from the screens, pills, and cooler consist orirarily
of oarticulate and ciaseous fluorides. Ml gases are treated ^or
product recovery before enterinc ^luoride control enuiorient, fvolutinn
of fluorides fror1 the oroducti^n nf c*iannr''oniui~ "'losoh^t? is about 0.3
nounds of fluorides ^er ton of p pc fron tie reactor ?rd r-rcinulatnr,
5-9
-------
and 0.3 pounds of fluoride ncr ton p-f D?n5 fror. the drver, cooler
14
and screens.
5.2.4 Emissions from Triple Sun?rohosnhate 'Manufacture and ftorane
5.2.4.1 Run-of-nile triple sunernhpsnhate
fluorides can be released fmn both the nhosnhnric acid gnd t
phosnhate rock durino the aciYulation reaction. Maior sources
fluoride enissions include the nixinn cone, curinn belt (den),
transfer convevors. and storacie m'les. These enission nrints
shown in fioure 4-10.
The nixinn cone, curinn belt, and transfer convevors are tvnicallv
hooded with ventilation streams sent to a common ^luori^e control
system. Storane 'ouildinns are usually sealed and ventilate^ bv
aonroxinatelv five air chanqes ner hour. The ventilation stream
from the storage facilitv nav either be combined vnth the ^i
and den oases for treatment or sent to separate controls.
Fluoride emissions are nrinarily silicon tetra^luoride -
35 to 55 oercent of the total fluoride content of the acid and rock
is volatilized as silicon tetrad uo ride.16 ^a*or sources o* fluoride
are the rr.ixina cone, curina belt, nroduct convevors. and stnrane
facilities. Distribution of enissions anonn these sources will v^rv
denendinn on the reactivitv nf the rock and the snecific oneratin" con-
ditions ennloved. Emissions fron the cone, curinn belt, and con-
vevors can account for as nuch as 00 nprccnt nf VIP t^tcl 'lu^rides
released. Conversely, it h?s been claimed that annroxi'"3tel»' °°
5-10
-------
percent ?f '.he fluoride emissions from certain ROP olants are from
the storage area. Emissions from the storage area denend on such
factors as the turnover rate and the age and nuantity of PPP-TSP
in storage.
evolution of fluorides from pnp-TSP production and storage has
been estimated at 31 to 48 nounds ner ton of ^5- This estimate
is based on the follov/ing assumptions: 1) silicon tetra fluoride is
the only fluoride emitted in annreciable quantities and 2) the feed
acid and rock contain typical amounts of fluorine.
5.2.4.2 Granular triple sunerphosohate
Manufacture
The major sources of fluoride emissions from granular triole
suoernhosphate plants usino the TV<\ one step process are the
acidulation drum, the granulat.or, the cooler, and the screening and
crushing operations. Ma.ior sources of emissions for the Dorr-Cliver
process include the mixinq tanks, the blunger, the drier, and the
screens. These emission ooints are indicated in Figures 4-12 and
4-13. In addition to gaseous forms, fluorides are emitted as
parti cul ate from the granulator, blunger, dryer, screens, and mills.
The acidulation drum and granulator (TVA orocess) and the
mixing tanks and blunger (Dorr-Cliver process) account for about 38
percent of the fluoride emissions, the drier and screens account for
I g
50 percent, and the storage facilities account for the remainder.
It has been estimated that an uncontrolled production facility would
18
emit approximately 21 pounds of fluorides per ton of P2^5 i
5-11
-------
Storage
GTSP storage facilities can emit both particulate and oaseous
fluorides. Uncontrolled emissions are estimated to be three pounds
1 o
per ton of PpOg input.
5.3 TYPICAL CONTROLLED FLUORIDE EMISSIONS
5.3.1 Emissions from Wet-Process Phosphoric Acid Manufacture
Almost all existing wet-process phosphoric acid olants are equipped
to treat the reactor and filter qases. A large number of installa-
tions also vent sumps, hotwells, and storage tanks to controls.
Typical emissions range from 0.02 to 0.07 pounds of fluoride per ton
of P«0c input, however, emission factors as high as 0.60 pounds fluoride
in on
per ton PgOg have been reported for a few poorly controlled plants. '
It is believed that approximately 53 percent of the wet-process
add plants - accounting for 74 percent of the production caoacity -
are either sufficiently controlled at present to meet the SPNSS
emission level of 0.02 pounds of total fluorides (as F) per ton of
PgOg input to the process or will be required to attain that level
by July 1975 to satisfy existing State regulations. This estimate is
based on the following: 1) all wet-process acid plants located in
Florida are required to meet an emission standard equivalent to the SPNSS
as of July 1975 and 2) all wet process plants built since 1967 are
assumed to have Installed spray-crossflow packed bed scrubbers or their
equivalent as a part of the original design.
5-12
-------
5.3.2 Emissions from Superphosphorlc Acid Manufacture
Two types of processes are used for Superphosphorlc acid
manufacture; the vacuum evaporation (VE) process and the direct
contact evaporation (DCE) or submerged combustion process. Emissions
from the VE process, are very low in- comparison to the DCE process.
Emissions from a VE process using a water actuated venturi to treat
hotwell and product cooler vent gases have been reported to range
from 4.1"X 10~4 to 15 X 10"4 pounds fluoride per ton P205 input.21
However, uncontrolled emissions from this process are also less than
the 0.01 pound per ton of P20g input emission guideline.
Since most of the existing superphosphoric acid plants use the VE
process, approximately 78 percent of these plants are currently
meeting the emission guideline,
Since the DCE process has much higher emissions, the emission
guideline was established at 0.01 Ib. F/ton P205 input.
This guideline 1s consistent with the level of emission control
achievable by application of best control equipment to a DCE process.
Typical controls used are a primary scrubber for removal of entrained
99
acid and one «r more additional scrubbers for fluoride control.
Emission from an existing facility weee reported at 0.12 pounds
fluorl* per ton
6.3.3 Emission* fn» 01 ammonium Phosphate Manufacture
Most existing plants are equipped with ammonia recovery
scrubbers (venturi or cyclonic) on the reactor-granulator and
drier streams and participate controls (cyclones or wet scrubbers)
on the Cooler stream^ Additional, scrubbers for fluoride removal are
lUNUfl, but not typical. Only about 15-20 percent of the instal-
lations contacted by EPA during the development of the-SPNSS were
5-13
-------
equipned with spray-crossflow packed bed scrubbers or their eauiva-
lent for fluoride removal. Fluoride emissions ranoe fro"! 0.0? to 0.5
2/i
oounds per ton P-Cg deoendino uoon the decree of control orovided."
5.3.4 Emissions from Triple Superphosphate Manufacture and Storage
5.3.4.1 ROP triple superphosphate (manufacture and storage)
All run-of-pile triple superphosphate production facilities and
70 percent of the storage facilities are eouioped with so^e *orw o^
25
control. Emissions from those slants which control bo^h nrcduction
and storage areas .an ranqe from 0.2 to 3.1 pounds of -fluoride ner
pc ?7
ton of P20g innut dependina unon the dearee of control orovideri. "
Plants with uncontrolled storaoe facilities could emit as puch as 12.7
oounds of fluoride oer ton of Pp^g inout. "t least 6? oercent o^ ts.e
industry will be reauired to meet State emission standards equivalent
to the SPNSS by July 1975.
5.3.4.2 Granular triple superphosphate (manufacture)
Existing State regulations will require 75 oercent of the industry
to neet an emission standard of 0.20 pound fluoride per ton Pg0^ bv
July 1975. Emission factors for the industry ranoe fror 0.20 to 0.60
op
nounds per ton P2^5-
5.3.4.3 Granular triole suoerohosphate (storaoe)
Aooroxinately 75 oercent of the r*TSP storaoe facilities ere
23
thought to be equipped with some ^errr of control. '' Poorlv con-
-d
trolled buildinns can release as nuch as 15 x 10 oounds of
fluoride per hour per ton of P-,' 5 in stor^ce ' V-'ell-ccntrn'Med
-4
storage facilities can reduce emissions to less than 5 x 10 nounds
5-14
-------
on
fluoride per hour per ton of P2^5 in storage. It is estimated
p
th?.t 33 oercent of the controlled bin'ldinos could reet fnf!SS °n'ssion
20
level.
5.4 GYPSUM POND EMISSIONS
.', wet process phosphoric acid plant produces gypsuT in slurry
forn, according to the cherrical reaction indicated in equation 4-1.
The reaction also volatilizes fluorides which are largely absorbed
in scrubber and condenser water and is then sent with the qt'osum to
large storage ponds, known as oypsun ponds or "gyp" conds. nver 7^
percent of the fluorine content of the rock used in the wet-acid
process may pass over to the gyp pond. If the same plant also pro-
duces DAP or TSP, a larga part of the fluorine content of the phosphoric
ecid will also pass to the gyp pond through the use of water scrubbers
in these additional processes. Thus, 85 percent or more of the fluo-
rine originally present in the phosphate rock may find its way to the
gy? pond.
T;,2 water of the gyp pond is normally acic1, Kavir.a a pH ar-u.n'
1.5. This acidity is.probably due to inclusion of phosphoric acid in
the -..ashed gypsum from the gyr»sm« filter. It is impractical to remove
ell of the acid from the filter cake by washinn. For this reason,
?yo ponds around the country have bee.i found to have a fluoride concen-
tration of 2000-12,500 ppm.31"34 The fluoride concentration of a given
oond does not continue rising, ;--jt tends to stabilize. Tin's nay te
c!u£ to orecinitation cf co— !•};• c-lcium silicofluorides in the ocnd
./at&r. Tfiare /ould oe an equilibrium invclvin" tiipse conlexes,
iiydrogen ion, and soluble or volatile dissolved fluorides.
5-15
-------
It has been observed that the above concentrations of fluoride
exert a partial pressure out o* OVP onnd water anc1 that volatile
fluorides tend to evolve fror ayp ponds. Based on wet orocess
Dhoso.horic acid production. plants have pyo oonds of surface areas
in the ranoe of 0.1-0.4 acres per daily ton of ?2f5-34 This reans
that a large plant may have a CJVD pond vn'th surface area of 200 acres
or more.
Emission factors have been estimated, measured and calculated for
ayp ponds. These factors vary from about 0.2 to 10 Ibs F/acre dav.?1~*?/!
The most comprehensive work on ovp oond emission ^actors is that
recently done in EPA Prant No. R-800950. The experimental and
mathematical procedures are quite detailed and the entire report should
be examined by those needinn to understand the methods user!. The
partial pressure of fluorides out of actual oond water was deterrined
in the laboratory. The evaporation rates of dilute fluoride solutions
were derived from known data ^or flat water surfaces, usinn established
mass transfer principles. Also, ambient air fluorides were measured
downwind of the same gyp oonds which Burnished the above water samoles
for fluoride partial oressure measurements. Finally, the contribution
of the gyp pond to the fluoride peasurenent at the downwind sensor
was calculated, usina a variant o^ the Pasauill division eauation.
The source strength in this eouation was, of course, calculated
with the partial pressure data and mass transfer coefficient previouslv
develooed. There were a total of 95 useable downwind peasure^ents for
5-16
-------
two nond sites, and the estipated and the neasured downwind fluoride
'concentrations showed need frree^ent. The calculated v?lue *f the
arbient air fluorine concentration downvind of the oonri -,'?s fnunH
to be statistically the sane as the.neasured value.
Tone emission factors fro**1 the above investication ere niven in
Table 5-3. Data at other temosratures mav be found in the orim'nal
reference.
Table 5.3 FLUORIDE EMISSION FACTORS FOR SELECTED RYP.SU''1
90°F; IDs/acre day.3*
Nind velocity
at 16 ft elevation,
m/sec
1 2
fond 10 0.8 1.3
6/00 pprn F
Pond 20 0.8 1.3
12,000 pir F
4 P
2.3
2.3 3.2
For the two plants studied, the emission rates were nearlv
identical. There r,ay be significant differences if other oonds are
considered, but more neasurenents would be required to establish this.
The most effective v/av to reduce fluoride evolution fror nvn oonds
vould be to reduce their fluoride oartial pressure in sone way. The
nost effective nethod now knovn v/ould be liruina, to raise the nH.
Lininq to a pH of 6.1 has redo:ed the nartial oressure of fluoride 30-
fold.31 The indicated li>e cost -.-ould bs hi oh for the case described,
but this cost can be reduced i* ? rethod can be *ound to reduce
pnosnhoric acid loss to the nvp pond.
5-17
-------
5.5 REFERENCES
1. Teller, A.J. Control of Gaseous Fluoride Emissions. Chsrical
Engineer! no Proqress. 63! : 75-79, March 1967.
2. Lutz. W.A. and C.J. Pratt. Kanu^actiire of Triple Smerohcsnhate.
In: Chemistry and Technoloay of Fertilizers. Sauchelli, V. (ed.)
New York, Reinhold Publishing Corporation, 19fQ. D. 175.
3. Teller, A.J. and D. Reeve. Scrubbing of Gaseous F.
In: Phosphoric acid, Vol. I, Slack, A.V. (ed.). F!ew Yori-,
?iarcel Dekker, Inc., 19F8. p. 752.
A. Engineer! nq and Cost Effectiveness Study of Fluoride Enissions
Control. Resources Research, Ire. McLean, Viroim'a. rD/1
Contract EHSD 71-14. January 1972. p. 3-152.
5. Atmospheric Emissions from Het-Process Dhosohnric ^cid Manufacture.
National Air Pollution Control Administration. Paleinh, north
Carolina. Publication Number AP-57. Aoril 1970. p. 1C.
6.. Control Techniques for cluoride Enissions. Environmental Health
Service. Second Draft. September 1970. p. 4-71. (Unoublished).
7. Noyes, R. Phosphoric Acid by the Met Process. park Hidoe,
New Jersey. Koyes Development Corporation, 1957. o. 2?^t
231.
5-18
-------
P. Scott, W.C. Jr. "reduction by Vet Process. In:
P.'sosohoric Acid, Vol. !. Slack. A.V. (ed). fie-.' YO-I-,
•iarcel Dehker. Inc., 1968. D. 1080.
?. reference 7, D. 191 .
10. Reference 5, p. 4-71.
11. Air Dollution Control Technolociy and Costs in Seven Selected
/Teas, Phase I. Industrial r^s Cleaninn Institute, ^tan^ord,
Connecticut. EPA Contract 63-02-0289. '.arch 1973. P. 86.
12. Reference 7, p. 256.
13. reference 6, p. 4-106.
14. Reference 4, p. 3-161.
15. Tin'berlake, R.C. Fluorine Scrubber. Southern Fnnineer.
June 1967. p. 62-64.
16. Jacob, K.D. et al. Comoosition and Pronerties of Suierohosohf»t<5,
Ind. and Ena. Chem. _34_: 7*7. June
17. Reference 2, o. 180.
18. Reference 4, o. 3-167.
19. Referen- •• 5, D. 3.
5-19
-------
20. Technical Report: Phosohate Fertilizer Industry. In: An
Investigation of t'ie Pest cystens if Erission. "aduction *or "i/
Phosphate Fertilizer Processes. Environmental Protection "nenc".
Research Triangle park, North Carolina. .April 197*. n. 22.
21. Reference 20, o. 33.
22. Reference 6. p. 4-74.
23. Good/in, D. Written corrmunication fro^ *1r. ^.D. frith, Acci-
dental Chenical Conpany. Houston, Texas, /ioril 3"), 1973.
24. Reference 20, p. 36, 38.
25. Beck, L.L. Reconmendations for Emission Tests ^f phosnhate
Fertilizer Facilities. Environmental Protection ftnencv.
Durham. North Carolina. September 28, 1972. o. 1^-16.
26. Reference 20, p. 47.
27. Reference 4, p. 3-1C7.
28. Reference 20, o. 52, 53.
29. Reference 25, D. 10-13.
30. Reference 20. o. 57.
31. Reference 5. on. 15-16.
5-20
-------
32. Tatera, B.J. Parameters '.-:hich Influence Fluoride Erissions from
Synsun Donds. D;-iD Thesis. University c-f Florida, <-?ir.=svillc.
1973. (University •-icrofilr-s. "nn ,'rjor, 'iicr.., liur.ber 71-275.)
33. Elfers, L.A., IIAPC'., to ^, .\J. and Crane, G.B. cistec!
Decerbsr 31, 1968. Fluoride .-nalyses of Gyp Pond Hater from
Texas Gulf Sulfur Corporation.
34. Kino. h'.R. Fluorine -r-ir Pollution from Wet-Process Phosphoric
,"cid Plant Process - "ater Ponds. TIO T'-.esis. north Carolina
State University. !>lein:i, -.r-. K71, st'-jported bv EP,n. Research
Grant i;o. 1-800950.
35. Teller, /"-.J. Communication at fAPCTAC msetinci in '-.aleiah, r.'.C.
on February 21, 1373.
5-21
-------
6. CONTROL TECHNIQUES FOR FLUORIDES FROM PHOSPHATE FERTILIZER PROCESSES
6.1 SPRAY-CROSSFLOW PACKED BED SCRUBBER
6.1.1 Description
The spray-crossflow packed bed scrubber has been accepted for
several years as the most satisfactory fluoride control device available
for wet-process phosphoric acid plants.1 Most wet-process acid plants
built since 1967 probably have installed this scrubber as part of the
original design. During this same time, however, the spray-crossflow
packed bed design has seen less general use in processes other than wet
acid manufacture. The reluctance of the fertilizer industry to fully
adopt the spray-crossflow packed bed scrubber can be traced primarily
to concern about its operational dependability when treating effluent
streams with a high solids loading. Such effluent streams can be
handled by placing a venturi scrubber in series with and before a spray-
crossflow packed bed scrubber; the EPA has tested a number of DAP and GTSP
plants having this dual scrubber arrangement. Also, improvements in spray-
crossflow packed scrubber design have alleviated the initial problem of
plugging and allow a greater solids handling capacity. The development
of stricter fluoride emission standards should provide incentive for more
widespread use of this scrubber design.
Figure 6-1 is a diagrammatic representation of the spray-crossflow
packed bed scrubber. It consists of two sections - a spray chamber and
a packed bed - separated by a series of irrigated baffles. Scrubber
size will depend primarily upon the volume of gas treated. 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.2
6-1
-------
PRIMARY GAS INLET
en
i
rxj
POND WATER
V
I!
1
i
xU SECONDARY
r^\^f
V
GAS INLET
SPRAYS
CQ
Q
LU
f—
-------
All internal parts of the scrubber are constructed of
corrosion resistant plastics or rubber-lined steel. Teflon can be
used for high temperature service. General maintenance consists
of replacement of the packing once, or twice a year. Exoected life
of the scrubber is 20 years.
Both the spray and the packed section is equipped with a aas
inlet. Effluent streams with relatively hioh fluoride concentrations -
particularly those rich in silicon tetrafluoride - are treated in the
spray chamber before entering the packing. This preliminary scrubbing
removes silicon tetrafluoride thereby reducing the danger of plugaino
the bed. At the same time, it reduces the loading on the packed staoe
and provides some solids handling capacity. Gases low in silicon tetra-
fluoride can be introduced directly to the packed section.
The spray section accounts for approximately 40 to 50 percent
of the total length of the scrubber. It consists of a series of
countercurrent spray manifolds with each pair of soray manifolds followed
by a system of irrigated baffles. The irrioated baffles remove pre-
cipitated silica and prevent the formation of scale in the spray chamber.
Packed beds of both cocurrent and crossflow design have been
tried with the crossflow design proving to be the more dependable.
The crossflow design operates with the gas stream moving horizontally
through the bed while the scrubbing liguid flows vertically through
the packing. Solids tend to deposit near the front of the bed where
they can be washed off by a cleaning spray. This design also allows the
6-3
-------
use of a higher irrigation rate at the front of the bed to aid in
solids removal. The back portion of the bed is usually operated dry
to provide mist elimination.
The bed is seldom more than 3 or 4 feet in length, but this can
be increased if necessary with little change in capital or operating cost.1
Several types of ceramic and polyethylene packing are in use with
Tellerettes probably the most common. Pressure loss through the scrubber
ranges from 1 to 8 inches of water with 4 to 6 being average.1'3
Recycled pond water is normally used as the scrubbing liquid
in both the spray and packed sections. Filters are located in the
water lines ahead of the spray nozzles to prevent plugging by suspended
solids. The ratio of scrubbing liquid to gas ranges from 0.02 to 0.07
gpm/acfm depending upon the fluoride content - especially the silicon
tetrafluoride content - of the gas stream.3*4 Aporoximately one-third
of this water is used in the spray.section while the remaining two-thirds
is used in the packing.
The packed bed is designed for a scrubbing liquid inlet pressure
of about 4 or 5 pounds-per-square-inch (qauge). Water at this pressure
is available from the pond water recycle system. The spray section
requires an inlet pressure of 20 to 30 pounds-per-square inch (qauqe).
This normally necessitates the use of a booster pump. Spent scrubbina
water is collected in a sump at the bottom of the scrubber and pumped
to the gypsum pond.
6-4
-------
6.1.2 Emission Reduction
The use of gypsum pond water as the scrubbing solution com-
plicates the task of fluoride removal regardless of the scrubber
design. Gypsum pond water can be expected to contain from 0.2 to 1.5
percent fluosilicic acid (2000-12,500 pom F) or most often, 5000-
6000 ppm F. Decomposition of fluosilicic acid to silicon tetrafluoride
and hydrogen fluoride results in the formation of a vapor-liquid
equilibrium that establishes a lower limit for the fluoride concentra-
tion of the gas stream leaving the scrubber. This limit will vary
with the temperature, pressure, and fluosilicic acid concentration of
the water. Table 6-1 presents equilibrium concentrations (y1) calcu-
lated from experimentally obtained vapor pressure data at three
temperatures and several fluosilicic acid concentrations.
Table 6-1. CALCULATED EQUILIBRIUM CONCENTRATIONS OF FLUORINE IN
THE VAPOR PHASE OVER AQUEOUS SOLUTIONS OF FLUOSILICIC
ACID6
Fluosilicic acid
content of solution (wt %)
0.1C5
0.550
1.000
2.610
2.640
5.050
7.470
9.550
11.715
14.480
Total fluorine concentration
in vapor phase (ppm F)
50°C
2.4
3.8
4.4
5.6
8.2a
12. 4a
13.5
19.1
-
60°C
3.8
4.43
7.1
9.8a
_
14.2?
19. 4a
25.6
34.6
83.5
70°C
_
10. 5a
15.4
•20. 7a
_
54. la
208.5
-
—
—
Average based on several vapor pressure measurements
6-5
-------
Providing that the solids loadinq of the effluent stream has
been reduced sufficiently to prevent plugging, the fluoride removal
efficiency of the spray-crossflow packed bed scrubber is limited
only by the amount of packing used and the scrubbing liquid. Efficiencies
as high as 98.5 and 99.9 percent have been measured for scrubbers
installed at separate wet-process acid plants.1'7 Table 6-2 lists the
levels of fluoride control reached by several wet acid plants tested
by the Environmental Protection Agency during the development of
SPHSS. All plants used a sprav-packed bed type scrubber to control
the combined emissions from the reactor, the filter, and several
miscellaneous sources and were felt to represent the best controlled
segment of the industry. Gypsum pond water was used as the scrubbing
liquid. Emission rates ranged from 0.002 to 0.015 nounds fluoride
(as F) per ton P20g input to the process.
Table 6-2. SCRUBBER PERFORMANCE IN WET-PROCESS PHOSPHORIC ACID
PLANTS8
Plant
A
B
C
D
Scrubber design
spray-cocurrent packed bed
spray-crossflow packed bed
spray-crossflow packed bed
spray-crossflow packed bed
Fluoride emissions3
(Ib F/ton P205)
0.015
0.006
0.002, 0.012b
0.011
aAverage of testing results
Second series of tests
6-6
-------
Spray-packed bed type scrubbers have seen only limited service in
diammonium phosphate and granular triple sunerohosphate plants and none
at all in run-of-pile triple superphosphate plants. Table 6-3 oressnts
performance data, collected during the development of SPNSS, for
spray-crossflow packed bed scrubbers treating effluent streams from
dianmonium phosphate, granular triple superphosphate production, and
granular triple superphosphate storage facilities. In mqst.cases, a
preliminary scrubber (venturi or cyclonic) was used to reduce the
loading of other pollutants (ammonia or solids) nrior to treatment in
the spray-crossflow packed bed scrubber. Gyosum nond water was used as
the scrubbing solution except where indicated. Fluoride emission rates
from dianmonium phosphate plants ranged from 0.029 to 0.039 nounds ner
ton PgOg input, while emissions from granular triple superphosphate pro-
duction facilities ranged from 0.06 to 0.18 pounds per ton PpOc- Cranular
triple superphosphate storage facility emissions were measured at 0.00036
pounds per hour per ton of P^O,. in storage.
6.1.3 Retrofit Costs for Spray-Crossflow Packed Bed Scrubbers
This section discusses the costs associated with retrofitting spray-
crossflow packed bed scrubbers in wet-process phosohoric acid, suoer-
phosphoric acid, diammonium phosphate, run-of-pile triple superphosphate,
and granular triple superphosphate plants. Two separate approaches -
retrofit models and retrofit cases • are used to present cost information.
Tne retrofit model approach is meant to estimate costs for an average or
typical installation. No specific plant is expected to conform exactly
to the description presented in these models. Where possible, the retrofit
model treatment is supplemented by retrofit cases - descriptions of specific
plants which have added spray-crossflow packed bed scrubbers to uonrade
their original control systems.
6-7
-------
Table 6-3. SPRAY-CROSSFLOW PACKED BED SCRUBBER PERFORMANCE
IN DIAMMONIUM PHOSPHATE AND GRANULAR TRIPLE
SUPERPHOSPHATE PLANTS9
Type of
facility
Sources controlled
Primary controls
Secondary controls
Fluoride emissions'
(Ib F/ton P205)
DAP
UAP
GTSP
GTSP
jGTSP
(storage
reactor, granulator,
drier, and cooler
reactor, granulator,
drier, and cooler
reactor, qranulator,
drier, and cooler
reactor, granulator,
drier, and cooler
storage building
3 venturi scrubbers
in parallel"
3 venturi scrubbers
in parallel6
3 venturi scrubbers
in parallel
process qases com-
bined and sent to 2
venturi scrubbers in
parallel followed by
a cyclonic scrubber
3 spray-crossflow
packed bed scrubbers
in parallel
3 spray-crossflow
packed bed scrubbers
in parallel
3 spray-crossflow
packed bed scrubbers
in parallel
spray-crossflow
oacked bed scrubber
spray-crossflow
packed bed scrubber
0.034, 0.029°
0.039
0.18, 0.06C
0.21
O.OQ036n
Average of testing results.
Weak phosphoric acid scrubbinq solution.
cSecond series of tests.
.4
Emission rate is in tenns of pounds T per hour per ton of P205 in storaqc.
-------
6.1.3.1 Retrofit Models
General Procedure
Each retrofit model provides the following information:
1. A brief description of the process in use,
2. A description of existing fluoride controls and the sources
treated,
3. A description of the retrofit project (including the reduction
in fluoride emissions achieved), and
4. A breakdown of estimated retrofit costs.
Items 1 and 2 are self-explanatory, however, items 3 and 4 will require
some discussion. In the case of item 3, all retrofit systems are designed
to meet SPNSS emission levels. A scaled plot plan of a model phosphate
fertilizer complex was used to estimate piping, ductwork, pumps, and fan
requirements.
The procedure used for development of costs is a module approach,
starting with the purchase cost of an item - such as a pump, scrubber,
fan, e£c. - and building up to a field installed cost by using an
appropriate factor to account for ancillary materials and labor. For
example, a pump of mild steel construction costing $10,000 is projected
to $17,600 field installed. The installation cost index in this case
is 1.76 and the installation cost is $7,600. If the pump were built
of stainless steel, the purchase cost would be $19,300 but the installa-
tion cost would remain at $7,600 since it is calculated for the element
of base construction - mild steel.
6-9
-------
The purchase cost of the various items on an equipment specifica-
tion list drawn up for each model plant were derived from literature,
manufacturer's bulletins, telephone quotations from suppliers, and
a report prepared by the Industrial Gas Cleaning Institute. Scrubber
costs were obtained by combining designer, manufacturer and user estimates.
Purchase costs were scaled up to field installed costs by using an
appropriate installed cost index. Table 6-4 is a list of the cost indices
assumed for this analysis.
Table 6-4. INSTALLED COST INDICES
Item Installed cost index
Pumps 1.76
Piping (except valves) 2.00
Scrubbers 1.20
Centrifugal fans 1.60
Stack 1.50
Ductwork 1.40
The sum of the field installed equipment cost is the direct
cost billed to a particular project. Other costs such as general
engineering, procurement of goods and services, equipmental rentals,
field supervision, labor burdens, contractor fees, freights, insurance,
sales taxes, and interest on funds used in construction are included
in the catch-all category of indirect costs. In this study, the indirect
cost is assumed to be 35 percent of the direct cost. In addition, a
6-10
-------
contingency factor is included in a capital project to account for
unforeseen expenditures. Due to the nature of the type retrofit
projects studied in this document, a factor of 25 percent of direct
costs has been incorporated in the capital estimates. The total
capital requirement of a project therefore is equal to the sum of
the direct cost, the indirect cost, and the contingency cost, as
indicated in equation 6-1:
I = D+0.35D+0.25D
where I = total capital
D = total direct cost
The following assumptions were used in the development of cost
estimates:
1. The purchase costs of scrubbers were determined from the most
recent manufacturer quotations, users wherever possible,
and the Industrial Gas Cleaning Institute. The purchase
cost of ductwork, stacks, and centrifugal fans were derived
12
from a manufacturer's published list prices. The costs
are 1974 estimates based, for the most part, on the use of
corrosion resistent fiber reinforced plastics (FRP) as the
material of construction.
2. Installed costs for scrubbers, ductwork, stacks, and centri-
fugal fans (including drivers) were derived by multiplying
the purchase costs by the appropriate cost index from
Table 6-4. An inherent assumption is that FRP is a base
6-11
-------
construction material suitable for application of the
listed indices.
3. Demolition costs were estimated from contractor Quotations to be
$2500/8-hour day.
4. Piping costs were derived for a corrosion resistant material
called Permastrand.
5. Pumps were assumed to be of stainless steel construction.
Cost estimates were obtained from the literature. These
costs» originally published in 1968, were increaser 28 oercent
(5% per year) to update to 1974 costs.
6. Costs for pump motors were obtained from the literature and
adjusted for inflotion usirg the same procedure described for
pumps.
7. Special compensatory factors for construction costs were
incorporated into the ROP-TSP and GTSP storage facilities.
Such factors appear under the headings of "sealing of storage
building", "curing belt hooding", and "structural steel suonorts/
bldg." The costs for these items v/ere pro-rated on the basis
of a recent engineering project study for a fertilizer producer.
8. Cost for performance tests were based on a telephone survey of
independent contractors.
6-12
-------
9. Annualized Costs
a. Capital charges are 16.3 percent of the total capital
outlay. This was derived from the capital recovery
factor equation,
i (1 + i)n
R = P (6-2)
(l+i)n-l
where: P = capital outlay (principal),
R = periodic capital charge,
i = annual interest rate (10$), and
n = number of payments (10)
b. Maintenance and repair charge were assumed to be 3
percent of the original investment.
c. Taxes, insurance, and administrative costs were assumed
to be 4 percent of the original investment.
d. Operating labor costs were estimated at $2,000 per
year for the simple operation (phosphoric acid plant
and GTS storage) $4000 for the more difficult operations
(DAP, ROP, and GTSP processing).15
e. Utilities (electricity only) were based on a rate of
$0.015 per kw-hr and 7,900 hours operation per year.
6-13
-------
Wet Process Phosphoric Acid Plant
The model plant uses the Prayon process for the manufacture of
wet process phosphoric acid. Figure 6-2 presents a basic flow dia-
gram of the operation. The reactor is a multicompartment unit (9
compartments) with a designed production rate of 500 tons per day
PoOc- Temperature control for the reactor is provided by a vacuum
flash cooler. Under normal conditions, the reactor is maintained
at a temperature of 160-180°F and produces an acid containing 30
percent P2°5-
Filtering and washing of the by-product gypsum is accomplished
with a Bird-Prayon tilting pan filter. The separated gypsum is re-
moved from the filter, slurried with water, and pumped to a settling
pond. Product acid from the reactor (30% P00C) is stored before
Z b
being sent to the concentration system. Three vacuum evaporators in
series are used to concentrate the acid to 54 percent PO^C- Evaporator
off gases are treated in barometric condensers for removal of conrten-
sables; a large percentage of the fluorides are also collected.
Retrofit costs for some wet-process phosphoric acid plants
could be substantially greater than those estimated for this plant.
The retrofit model is of moderate complexity and includes all of the
activities with which most installations are expected to become
involved; however, increases in the gas volume being treated, additions
to the scope 6f work, and space limitations are all factors capable
of inflating the project cost above that estimated. Modifications
to the plant drainage system and installation of a ventilation system
6-14
-------
STEAU
CONDENSER
~\ •• '~I
"-
TO A3 PCLU
i P.' A ^1 /a:
^'^FiE'tjL P ••••"'J':
,. ,....,..-. j...~: |
"•* LQ Hi Hi 111 T;ic io-rc^-iu
' _: , p.i-->; ' '':--"' '•
, . . . •;•-'!!•••• ^
I
DIGESTED
«'i vrp
J j i v u C.
ScrlTANXS
..
• i.. .H ••.•"./
FIGURE 6-2. MANUFACTURE OF MIT-PROCESS PhOS^iiORIC ACID.
-------
for the filter are two items which have not been included within
the scope of the model but which could be encountered by some plants.
Costs will be estimated for two effluent stream sizes - 25,000
and 35,000 scfm. The effluent stream from an actual 500 ton per day
plant could range from about 20,000 to 40,000 scfm dependino primarily
on the digester design.
Existing Controls (Case A)
Existing controls consist of a cyclonic spray tower used to treat
the digester and the filter ventilation streams. Gypsum pond water
is used as the scrubbing liquid. This scrubber has been in operation
for eight years. Figure 6-3 shows the location of the unit.
Volumetric flow rates and fluoride concentrations associated
with the various emission sources are listed in Table 6-5. The flow
rates are based on a combination of literature data, source test
information, and control equipment design data. Fluoride removal
efficiency of the cyclonic spray tower is 81 percent. Total emissions
to the atmosphere from the sources listed in Table 6-5 are 7.3 pounds
of fluoride per hour with existing controls. Several miscellaneous
sources of fluoride such as the flash cooler seal tank, the evaporator
hotwell, the filtrate sump, the filtrate seal tank, and the filter
acid storage tanks are uncontrolled. Emission rates from these
sources are unknown.
6-16
-------
DIGESTER AND FILTER
VENT GASES
CVCLONIC SPRAY TOMER
ooo
PHOSPHORIC (~^) (~*)
ACID STORAGE^-' ^^
EVAPORATORS
en
o o
FILTER _
ACID STORAGE
WPPA
PLANT
125'
T
100'
_L
ROCK mrj
FROM GRIPPING MILL
Figure 6-3. fXISTINR CONTROL EQUIPMENT LAYOUT FOD "ODEL 1,'PPA PLANT.
-------
fi_5. . FLOW RATES AND FLUORIt.-E CONCENTRATIONS OF VJPPA PLANT
EFFLUENT STREAMS SZNT TC EXISTING CCNTR?LS (CASE -)
Emission source
Digester vent gas
Filter vent gas
Flow rate
(SCF'-i)
10,000
7,500
Fluoride concentration
(mg/SCF) (ppm)
25
5.5
1050
23*
Retrofit Controls (Case A)
The retrofit consists of the reolacement of the cyclonic spray
tower with a crosstlow packed bed scrubber. Limitations imposed
by the arrangement of existing equipment require the new scrubber
to be installed at a site 50 feet from the one previously occupied
by the tower. Gypsum pond water will be used as the scrubbing liquid.
Several miscellaneous sources (flash cooler seal tank, evaporator
hot well,^filtrate sump, filtrate seal tank, and acid storage tanks)
will be vented to the new unit which is designed to meet SPNSS
requirements for wet-process phosphoric acid plants [0.02 pounds
fluoride per ton P20g input). This corresponds to an emission rate
of 0.42 pounds fluoride per hour. Table 6-6 summarizes tbe volumetric
flow rates and the fluoride concentrations associated with the
emission sources to be treated.
6-18
-------
Table 6-5. FLOU R.VTES \'!D FL'JORIDE CONCENTRATIONS OF WPPA PLANT
T ST°EAMS SFWT Tn prTn°FITTF.P C™!TPnl^
Emission source
(Digester vent gas
Filter vent gas
.''iscellaneous
Flow rate,
(SCR1)
10,000
7,500
7, 500
Fluoride concentration
(ng/SCF) (pom)
25
5.5
0.3
105"
230
13
Figure 6-4 provides a view of the plant layout folio-vino the com-
oletion of the retrofit oroject. Installation of the new scrubber
requires the rearrangement of the existing ductwork and the addition
of a new ventilation system to handle the miscellaneous sources. P
net1 fan vill be required for the digester-filter ventilation system
because of the hiqher pressure drop of the crossflow sacked bed scrub-
ber. Treated gases will be exhausted from a newly installed 75-foot
tall stack.
Scrubbing water will be obtained from existing plant water lines.
A booster pump is required to provide 40 psig water for the spray
section. Pond water is assumed to have the properties shown in
Table 6-7. All scrubbing water will be recycled to the gyosum pond in
the existing plant drainage system.
6-19
-------
FILTER, FILTRATE fir.P AND
FILTPATE SEAL TANK
EVAPORATOR HOTHEL
:|f \ SPPAY-CPP?
^^. ^ ' . CTflC
PHOSPHORIC V^y
^ ACID _
• STORAGE ( ^\
5 \J
EVAPORATORS
O
o.
ooo
!
. riu
Q WPPA
/C
FILTER
J PLANT
ACID i. . 1051
STORAGE 1
— i
PACKED PEP
1-pIGFSTF.P AMP FILTER
PPCK COMVEVER
FRO'* GPINDINR MILL
FIGURE 6-4. PETROFIT CONTROL EQUIPMENT LAYOUT FOR MnpEL WPP/» PLAf''T.
-------
Table 6-7. POND WATER SPECIFICATIONS
15
Pond Hater pH
Temp., °F
Sn wt %
O *j it 3 *• w /O
tt
P2o5, \/t %
H2SiF6, wt %
Fluoride, wt %
Design
2.0
80.0
0.15
0.1
0.63
0.5
fin.
1.2
55
-
-
0.25
0.2
Max.
2.2
88
-
-
1.0
0.8
Major retrofit items are listed in Table 6-8. All ducting, piping,
and motors are specified in terms of the nearest aporooriate standard
size. Table 6-? oresents typical ooerating conditions for the new
scrubber and the estimated number of transfer units (NTH) necessary
to meet emission requirements. The NTU were calculated
by using equation 6-3.
NTU required = In
y2
(6-3)
where: y2 = fluoride concentration o* gas stream at the
scrubber inlet
y, = fluoride concentration of gas stream at the
scrubber outlet
y' = fluoride concentration cf gas stream in
equilibrium with entering liquid stream
Table 6-10 lists the estimated capital and annualized costs of the
project.
6-21
-------
Table 6-8. MAJOR RETROFIT ITEMS FOR MODEL WPPA PLANT (CASE A)
1. Ductwork required to connect existing diqester-fliter ventilation
system with retrofit scrubber - 50 feet of 36-inch duct. New
ventilation system connecting miscellaneous sources with control
system. Requirements are - 175 feet of 9-inch duct, 50 feet of
10-inch duct, 125 feet of 12-inch duct, 75 feet of 16-inch duct,
100 feet of 20-inch duct, and 50 feet of 24-inch duct.
2. Pipe connecting spray-crossflow packed bed scrubber with existing
plant water "line - 150 feet of 6-inch pipe.
3. Booster pump for spray section - 190 gpm, 81 feet total dynamic
head (TDH), 7.5 horsepower motor.
4. Centrifugal fan for digester - filter ventilation system -
17,500 scfm, 620 feet TDH, 50 horsepower motor. Fan for miscel-
laneous sources - 7,500 scfm, 660 feet TDH, 20 horseocwer motor.
5. Removal of cyclonic spray tover *nd existing stack.
6. Spray-crossflow packed bed scrubber. Unit will be reouired to
reduce the fluoride concentration to C.13 nc/SCF (5.6 pom)
when using the pond water specified in Table 6-7 and treatina
the gases listed in Table 6-6.
7. Stack - 75-foot tall, 4-foot diameter.
6-22
-------
Table 6-9. OPERATING CONDITIONS FOR SPPAY-CROSSFLra I PACKED
EFD SC'^.BFP F-^P "fDEL 'PP/1 PLANT, CASE A
9 tons/da" P°)
Gas to Scrubber
Flow, SCFI1 25,000
Flow, DSCFM 22,725
Flow, ACFM 27,150
Temp., °F 116
Moisture, Vol . % 9-1
Fluoride (as F), Ib/hr 38.7
Fluoride (as F), ppm 492
Gas from Scrubber
Flow, SCFM 24,400
Flow, DSCFM 22,725
Flow, ACFI1 25,700
Temp., °F 100
Moisture, Vol . % 6.5
Fluoride (as F), Ib/hr 0.42
Fluoride (as F), ppm 5.6
Fluoride Removal, wt % 99
Estimated y', ppm (see 0.85
page 6-5)
Estimated NTU required 4.7
6-23
-------
Table 6-10. RETROFIT COSTS FOR MODEL WPPA PLANT, CASE A
(500 tons/day PgOg) November 1974
Cost ($)
A. Direct Items (installed)
1. Spray-crossflow packed bed scrubber 58,900
2. Ductwork 18,600
3. Piping 2,400
4. Pumps and motor 3,400
5. Centrifugal fan and motor 14,300
6. Removal of old equipment 12,500
7. Stack 15,800
8. Performance test 4,000
Total Direct Items 129,900
B. Indirect Items
Engineering construction expense, fee,interest on
loans during construction, sales tax, freight insurance.
(35% of A) 45,500
C. Contingency
(25% of A) 32,500
D. Total Capital Investment 207,900
E. Annualized Costs
1. Capital charges 33,900
2. Maintenance 6,200
3. Operating labor 2,000
4. Utilities 6,900
5. Taxes, insurance, administrative 8,300
Total Annualized Costs 57,300
6-24
-------
Existing Controls (Case B)
The existing control system is the same as described in case A;
a cyclonic spray tov;er is used to treat the digester and filter
ventilation streams. Fluoride collection efficiency of the tower is
81 percent, f'inor miscellaneous sources of fluoride are uncontrolled.
Volumetric flow rates and fluoride concentrations of the various
effluent streams being controlled are listed in Table 6-11. Emissions
from the sources listed are currently 11.0 oounds of fluoride per
hour.
Table 6-11. FLOW RATES AND FLUORIDE CONCENTRATIONS OF WPPA PLANT
EFFLUENT STREAMS SENT TO EXISTING CONTROLS (CASE B)
Emission Source
Digester vent gas
Filter vent gas
i
Flow Rate
(SCFM)
20,000
7,500
Fluoride Concentration
(mg/SCF) (pom)
20 840
5.5 230
Retrofit Controls (Case B)
Details of the retrofit oroject remain the same as in the initial
case. The cyclonic spray tower treating the digester-filter gases
will be replaced with a spray crossflow packed bed scrubber de-
signed to handle the sources listed in Table 6-12.
6-25
-------
Table 6-12. FLOW RATES flfJn FIJinoiPE CONCENTRATIONS OF UP PA PLAMT
EFFLUENT STRESS SI.'iT T? PIIT^FITTE" rnMTD«LS (rfiSE a)
JEmission Source
Digester vent gas
Filter vent gas
'iscellaneous
Flow pate , Fluoride Concentration •
(SCF") . (mcj/SCF) (n0irj)
i
| 20,000 20
7,500 5.5
i 7,500 ; 0.3
1
1
P-40
?30
13
•
£ list of major retrofit items is presented in Table 6-13 while
operating conditions for the new scrubber are provided in Table 6-14.
Estimated capital and annual i zed costs of the orrqram is listed in
Table 6-15. Increasing the capacity of the system by intnn^ SCF*'
has resulted in a 20 percent increase in the capital cost of the
program and a 21 percent increase in the annualized cost.
Table 6-13. MAJOR RETROFIT ITEMS FOR MODEL WPPA PLANT (utoE C)
1. Ductwork required to connect existing dipester-filter ventilation
system v/ith retrofit scrubber - 50 feet of "R-inch duct, few
ventilation system connect! no miscellaneous sources with control
system - 175 feet of 9-inch duct, 50 feet of 10-inch duct, 125
feet of 12-inch duct, 75 feet of 16-inch duct, 100 *eet of 2^-
inch duct, and 50 feet of 2^-inch duct.
?. Pioe connecting spray-crossflow packed bed scrubber yit> existing
plant water line - 150 feet of 8-inch oine.
3. Dooster pump for spray section - 269 apn , 81 feet total dynamic
head (TOH), 10 horsepower motor.
6-26
-------
ft. Centrifugal fan ^or dioester - filter ventilation system -
27,500 scfm, 604 feet TPM, 75 horsepower motor. Fan fr*r
miscellaneous sources - 7,500 scfm, 660 feet TDK, 20 horsenover
motor.
5. Removal of cyclonic soray tower and existing stack.
6. Soray-crossflow packed bed scrubber. Unit will be required
to reduce the fluoride concentration to 0.09 mg/scf (3.9 ppm)
when using the pond water specified in Table 6-7 and treating
the gases listed in Teble 5-11.
7. Stack - 75 foot tall, 5 foot diameter.
Table 5-14. CPERflTING CONDITION'S FOR SPRAY-CROSSFLOW PACKED BF.D
SCRUBBER FOR I10DEL WPPA PLANT, CASE B
(500 tons/day Pg^)
Gas to Scrubber
Flow, SCFM 35,000
Flow, DSCFH 31,800
Flow, ACFfl 37,600
Temp., °F 109
Moisture, vol. % 9.1
Fluoride (as F) , Ib/hr 58.1
Fluoride (as F), ppm 529
Gas from Scrubber
Flow, SCFM 34,000
Flov, DSCFf 31,800
Flow, ACFK 35,600
Temo. , °F 95
I'oisture, vol . % 6.5
Fluoride, Ib/hr 0.42
Fluoride, ppm 3.9
Fluoride removal, wt % 99.3
Estimated y1 , pom 0.85
Estimated NTU required 5.2
6-27
-------
Table 6-15. RETROFIT COSTS FOR MODEL WPPA PLANT, CASE B
(500 tons/day P2C5) November 1974
A. Direct Items (installed)
1. Spray-crosslow packed bed scrubber
2. Ductwork
3. Piping
4. Pump and motor
5. Centrifugal fans and motors
6. Removal of old equipment
7. Stack
8. Performance test
Total Direct Items
B. Indirect Items
Engineering construction expense* fee, interest on
loans during construction, sales tax, freight insurance.
(35% 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 ($)
78,800
20,000
3,300
5,300
16,000
12,500
15,800
4,000
155,700
54,500
38,900
249,100
40,600
7,500
2,000
9,300
10,000
69,400
6-28
-------
Superphosphoric Acid
Two processes are currently available for the manufacture of
superphosphoric acid - vacuum evaporation and submerged combustion.
All but two of the existing U.S. production facilities use the vacuum
evaporation process and it is belteved that new facilities will
favor vacuum evaporation. No retrofit model will be presented for vacuum
evaporation plants because the low level of fluori.de emissions from
these facilities do not require control equipment in order to meet the
emission guidelines.
Existing submerged combustion plants are expected to continue
operation with some expansion in capacity possible. Retrofitted control
equipment may be needed to meet the emission guidelines for this type
of process. A retrofit model is presented for a plant using the
submerged combustion process in order to estimate the costs of applying
control equipment. The costs are developed based upon control equip-
ment designed to meet the fluoride emission guideline of 0.01 pounds per
ton of P205 input.
The model plant uses the Occidental Agricultural Chemicals process
for the production of superphosphoric acid. Desianed production capacity
is 300 tons per day P,,05. Figure 4-6 is a basic flow diagram of the
process.
Wet-process acid containing 54 percent P?05 is fed to the
evaporator and concentrated product acid containinq 72 percent tf
is withdrawn. The acid is maintained at its boiling point bv intro-
ducing a stream of hot combustion pases into the acid Dool. Gaseous
6-29
-------
effluent from the evaporator is cooTed by direct contact with weak
phosphoric acid feed in the evaporator vapor outlet duct, treated for
phosphoric acid recovery, given additional cooling, and treated for fluoride
rejnoval.
Existing Controls
Exhaust gases from the evaporator are treated for the recovery
of entrained acid before being sent to fluoride controls. The phosphoric
acid recovery system consists of an initial cyclonic separator followed
by a baffled spray duct and a second cyclonic separator. Weak phosphoric
acid (30% P20g) i? used as the scrubbing liquid in the soray duct.
Fluoride controls consist of 3 spray chambers in series followed
by an impingement scrubber. The spray chambers are baffled and each
is followed by an entrainment separator. Pond water is used as the
scrubbing liquid in all cases. Emissions to the atmosphere are 1.56
pounds of fluoride per hour with existing controls.16
Retrofit Controls
The retrofit cost projection is based on reolaceraent of the
impingement scrubber with a spray-crossflow packed bed scrubber* Since
available space is usually limited, the new unit is assumed to be
installed at the site previously occupied by the impingement scrubber.
Figure 5-6 provides a schematic diagram of the plant following
completion of the retrofit project.
Gypsum pond water will be used as the scrubbing liquid. Pond water
characteristics are listed in Table 6-7. Retrofitted controls are
designed to reduce fluoride emissions to C.12 pounds fluoride oer hour.
6-30
-------
SPRAY-CROSSFLOW
PACKED BED _
SCRUBBERS t
100
STACK
cr>
CO
I
SPA
PLANT
TOO1
OACID FEED
XTANKS
PRODUCT HOLD
ING TANK
SPRAY-CROSSFLOW PACKED BED
SCRUBBER
Figure 6-5. RETROFIT CONTROL EQUIPMENT LAYOUT ^OR MODEL SPA PLANT
-------
Installation of the spray-crossflow packed bed scrubber will
require moderate alteration of existing ductwork and construction of a
new pipe line connecting the scrubber to the existing water supply. No
4
additional fans will be required. Treated oases will be exhausted from
the existing stack. Scrubbing water is to be recvcled to the ovosuro pond
in the existing drainage system.
A list of major items required for the retrofit oroject is
presented in Table C-16. Table 6-17 provides operating conditions for
the new scrubber. Retrofit cost estimates are listed in Table 6-18.
Table 6-16. MAJOR RETROFIT ITEMS FOR MODEL SPA PLANT
1. Ductwork - modification of existinc ducting to connect new sorav-
crossflow packed bed scrubber. Requirements are 100 feet of 30-inch
duct.
2. Line connecting scrubber to main pond water supply system - 150
feet of 4-inch pipe.
3. Centrifugal pump - 130 gpm,113 feet total dynamic head (TDM), 7.5
horsepower motor.
4. Removal of impingement scrubber.
5. Supports and foundations.
6. Spray-crossflow packed bed scrubber. Unit is required to reduce
the fluoride concentration to 0.09 mg/SCF (4 npm) when usinq pond
water specified in Table 6-7 and treating aas stream described in
Table 6-12.
6-32
-------
Table 6-17. OPERATING CONDITIONS FOR SPRAY-CROSSFLOW PACKED
BED SCRUBBER FOR MODEL SPA PLANT
(300 Tons/Day P20g)
Gas to Scrubber
Flow, SCFM 9,800
Flow, DSCFM 9,110
Flow, ACFM 10,600
Temp., °F 115
Moisture, vol. * 7.0
Fluoride (as F), Ib/hr 3.9
Fluoride (as F), ppm 126
Gas from Scrubber
Flow, SCFM 9,400
Flow, DSCFM 9,110
Flow, ACFM 9,760
Temp., °F 90
Moisture, vol. % 3.0
Fluoride (as F), Ib/hr 0.12
Fluoride (as F), ppm 4.0
Fluoride removal, wt % 96.7
Estimated y1, ppn 0.85
Estimated NTU required 3.7
6-33
-------
Table 6-18. RETROFIT COSTS FOR MODEL SPA PLANT
(300 tons/day P205) November 1974
Cost ($)
A. Direct Items (installed)
1. Spray-crossflow packed bed scrubber 37,600
2. Ductwork 5,000
3. Piping 1,900
4. Pump and motor 3,400
5. Removal of old equipment 12,500
6. Performance test 4,000
Total Direct Items 64,300
B. Indirect Items
Engineering construction expense, fee, interest on
loans during construction, sales tax, freight insurance.
(35% of A) 22,500
C. Contingency
(25% of A) 16,000
D. Total Capital Investment 102,800
E. Annualized Costs
1. Capital charges 16,800
2. Maintenance 3,000
3.. Operating labor 2,000
4. Utilities 700
5. Taxes, jnsurance, administrative 4,000
Total Annualized Costs 26,500
6-34
-------
Pi ammonium Phosphate
This plant uses the TVA process for the production of diammoniuro
phosohate. A flov diagram of the operation is provided in Figure 4-9.
The model plant has a designed production capacity of approximately
1080 tons per day diammonium phosphate (500 T/D P205).
A preneutralization reactor is used for the initial contacting
of the anhydrous ammonia and the phosphoric acid. Completion of
the reaction and solidification of the product occurs in the granula-
tor. Effluent gases from the preneutralization reactor and the granu-
lator are treated for ammonia recovery and fluoride control before
being vented to the atmosphere.
A gas-fired rotary drier is used to remove excess moisture from
the product. Drier flue gases are vented through dry cyclones for
product recovery before being treated for ammonia removal. Air
streams vented from accessory cooling and screening equipment are
treated for particulate removal before being exhausted.
Existing Controls
Exhaust gases from the preneutralization reactor and the granula-
tor are combined and vented to a venturi scrubber for ammonia re-
covery. Weak phosphoric acid (30% P205) serves as the scrubbing
liquid. Approximately 95 percent of the anmonia is recovered and
recycled to the reactor. Fluorides stripped from the phosphoric
acid in the venturi are removed by a cyclonic spray tower using
gypsum pond water as tiie aLr.nrbinq solution. Fluoride removal
efficiency is 74 percent.
6-35
-------
The drier flue gases are treated for product recovery before
being- sent to additional controls. Collected particulate is re-
cycled to the granulator. A venturi scrubber using weak phosohoric
acid is used for ammonia recovery. Ammonia removal efficiency is
approximately 94 percent. No additional scrubbing is practiced.
Air streams vented from product cooling and screening equip-
ment are sent through dry cyclones for product recovery, combined,
and treated in a venturi scrubber for particulate removal. Weak
phosphoric acid serves as the scrubbing solution. Collected DAP is
recycled to the reactor.
Volumetric flow rates and fluoride concentrations associated with
the three major emission sources are presented in Table 6-19. The
values listed are estimates based on source test results and data ob-
tained from a recent contract study of control equipment costs (5).
Fluoride concentrations presented for the reactor-granulator and the
drier gas streams are values at the outlet of the ammonia recovery
scrubbers. Total fluoride emissions from the sources identified in
Table 6-19 are 4.95 pounds per hour with existing controls.
Table 6-19. FLOW RATES AND FLUORIDE CONCENTRATIONS FOR DAP PLANT
EMISSION SOURCEST7.18
Emission source Flow rate Fluoride concentration
(SCFM) (mg/SCF) (ppm)
Combined reactor-granula-
tor vent gases 30,000 0.65 27
iDrier oases 45,000 0.36 15
Cooler and screening equip-
ment vent gases 45.000 0.36_ 15_
6-36
-------
etrofit Controls
The retrofit consiscs 'A the replacement o* the cyclonic spray
to-..er cr, the reactcr-c-anulator stream \-ith e soray-crossflcv. packed
bed scrubber and' the adclicion cf spray- crossflo1.. packed bad scrubbers
as tail gas units tc the drier and cooler streans. Sypsur.i pond
v/atcr v;111 be used as the scrubbing liquid. Pond water is available
at 80°F with the properties listed in Table 6-7. Tha control system
is designed to conform with the fluoride emission guideline of 0.06
pounds cr fluoride per ton P2?5 Input - 1.25 pounds fluoride per hour.
Existing controls are located as depicted in Figure 6-6. The
rrrangenent of equipment is such that the spray-crossflov/ packed bed
seniors can be installed adjacent to the venturi scrubbers after
i-iodarata alteration of the ductvork. A new water line must bs in-
stalled to satisfy the increased da*and caused by the retrofitted scrub-
bers. A new fan will also be required for both the drier and the cooler
streaRi to compensate for the pressure drop of the secondary scrubber.
Treated gases will be exhausted from the existing stack. Spent scrub-
bing water is to be recycled in the existing drainage system.
Figure 6-7 provides a view of the plant layout after the instal-
lation of new controls. A list of major retrofit items is provided
in Table 6-20. Table 6-21 presents operating conditions for the sprsy-
crossflow packed bed scrubbers. Total capital cost and annual i zed
cost estimates for the project are presented in Table 6-22.
fi-37
-------
GYPSUM
POND
CD
CO
03
1200'
MISCELLANEOUS
DRIER STACK
l> l>
DAP
PRODUCTION
r> n.
DAP
STORAGE
REACTOR - GRANULATOR
L 150'- vJ *— 300' '• *>
125
v _ VENTURI SCRUBBER
O CYCLONIC SCRUBBER
FIGURE 6-6. EXISTING CONTROL EQUIPMENT LAYOUT FOR MODEL DAP PLANT.
-------
GYPSUM
POND
1200'
DRIER
MISCELLANEOUS
ic— a
DAP
PRODUCTION
STACK
DAP
STORAGE
k-a-
REACTOR - GRANULATOR
-150'
1
125'
1
300'
v_ VENTURI SCRUBBER
o — SPRAY CROSS-FLOW PACKED BEJ SCRUBBER
FIGURE 6-7. RETROFIT CONTROL EQUIPMENT LAYOUT FOR MODEL DAP PLANT,
-------
Table 6-20. MAJOR RETROFIT ITEMS FOR MODEL DAP PLANT
1. Ductwork - removal of cyclonic spray tower from service and
connection of three spray-crossflow packed bed scrubbers.
Requirements are 100 feet of 60-inch duct and 50 feet of 54-
inch duct.
2. Water line connecting gypsum pond with spray-crossflow packed
bed scrubbers - 1200 feet of 16-inch pipe with a 200-foot branch
of 14-inch pipe and a 150-foot branch of 6-inch pipe.
3. Two centrifugal pumps {one -spare) - 2550 gpm, 105 feet
total dynamic head (TDH), 125 horsepower motor. Booster pump
for.spray section of both-the drier and the cooler stream scrubber
345 gpm, 89 feet TDH, 7.5 horsepower motor.
4. Two centrifugal fans - 45,000 scfm, 285 feet TDH, 50 horsepower
motor.
5. Removal of cyclonic spray tower.
6. Supports and foundations.
7. Three spray-crossflow packed bed scrubbers. When using specified
pond water and treating gases described in Table 6-19, scrubbers
are required to obtain performance indicated in Table 6-21.
-------
Table 6-21. OPERATING CONDITIONS FOR SPRAY-CROSSFLOW PACKED
BED SCRUBBERS FOR MODEL DAP PLANT
(500 Tons/Day P90c)
Gas to scrubber
Flow, SCFM
Flow, DSCFM
Flow, ACFM
Temp. , °F
Moisture, vol. %
Fluoride (as F), Ib/hr
Fluoride (as F), ppm
Gas from scrubber
Flow, SCFM
Flow, DSCFM
Flow,. ACFM
Temp., °F
Moisture, vol. %
Fluoride (as F), Ib/hr
Fluoride ('as F), ppm
Fluoride removal, wt %
Estimated y1 , ppm
Estimated NTU required
Reactor-
granulator
stream
30,000
18,000
34,000
140
40
2.58
27.1
19,400
18,000
23,600
100
7
0.44
5.9
83
1.05
1.69
Dryer
stream
45,000
29,200
52,700
160
35
2.14
15.0
31 ,500
29,200
38,400
100
7
0.36
3.0
83.5
1.25
2.06
Cooler
stream
45,000
43 ,600
49,600
125
3
2.14
15.0
45,400
43,600
48,000
100
4
0.45
3.0
79
1.05
1.94
6-41
-------
Table 6-22. RETROFIT COSTS FOR MODEL DAP PLANT
(500 tons/day P20g) November 1974
Costs ($)
A. Direct Items (Installed)
1. Spray-crossflow packed bed scrubbers (3) 285,000
2. Ductwork 16,700
3. Piping 26,200
4. Pumps and motors 34,500
5. Centrifugal fans and motors 33,000
6. Removal of old equipment 12,500
7. Performance test 4,000
Total Direct Items 411,900
B. Indirect Items
Engineering construction expense, fee, interest on
loans during construction, sales tax, freight insurance.
(35X of A) 144,200
C. Contingency
(25% of A) 103,000
D. Total '• Capi tal. Invetteent 659,100
E. Annualtzed Costs
1. Cpaltal Charges 107,400
2. Maintenance 20,000
3. Operating labor 4,000
4. Utilities 21,200
5. Taxes, Insaeance, administrative 26,400
Total Annual1zed Costs 179,000
6-42
-------
Pun-of-Pile Triple Superphosphate
The plant uses the conventional TVA cone process for the pro-
duction of run-of-pile triple superphosphate. Rated production
capacity is approximately 1200 tons of triple superphosphate per day
(550 T/D P205)- Actual production averages approximately 800 tons
of triple superphosphate per day.
Figure 5-10 provides a flow diagram of the operation. Ground
phosphate rock is contacted with phosphoric acid (54 percent Pp^c)
in a TVA cone mixer. The resultant slurry is discharged to the den
where solidification of the product occurs. Cutters are used to
break up the product before it is sent to storage. A curing period of
approximately thirty days is required to allow the reaction to po to
completion.
Two initial levels of control will be assumed for the model POP
triple superphosphate plant and retrofit costs estimated for each
case. Most actual costs should fall somewhere between the two estimates,
Existing Controls (Case A)
In this case, it is assumed that the plant is in a relatively
good state of repair, that necessary ducting and piping changes are
moderate, and that the existing ventilation system does not require
modification. Replacement of an existing scrubber is assumed to be
the major item in the retrofit program.
Gases vented from the cone mixer and the den are currently treated
in a 20,000 cfm venturi, combined with the storage building ventila-
tion stream, and sent to a spray tower. The storage building ventila-
6-43
-------
tion air is sent directly to the spray tov/er. This control system
has been in'operation for approximately five years.
Gypsum pond water serves as the scrubbing liquid for both the
venturi and the spray tower. Mater is available at 80°F with a fluo-
ride content (as F) of 0.5 weight percent. Additional information
regarding the scrubbing liquid is provided in Table 6-7.
Ventilation flow rates and fluoride concentrations for the
various sources are listed in Table 6-23. The values listed in this
table are estimates based on source test results and control equip-
ment design data. Fluoride removal efficiencies are 86 percent for
the venturi treating the combined cone mixer - den gases and 71 percent
for the spray tcwer. Total fluoride emissions from the production
and storage facilities are 127 pounds per hour.
Table 6-23. FLOW RATES AND FLUORIDE CONCENTRATES FOR ROP-TSP
PLANT EMISSION SOUPCES19-21
Emission Source
Cone mixer vent gases
Curing belt (den) vent
gases
Storage building vent
gases
1
Flow Rate
(SCR1)
50Q
24,500
125,000
Fluoride Concentration
(mg/SCF) (opm)
0.71
95
24
30
4000
1000
Retrofit Controls
The proposed retrofit involves the replacement of the spray tower
with a spray-crossflow packed bed scrubber designed for 9? percent
fluoride removal. Installation of the new scrubber will reduce
6-44
-------
fluoride emissions to 4.6 pounds per hour. This emission level is
equivalent to the emission guideline of 0.2 pounds fluoride per ton P,
input.
Moderate rearrangement of the ductwork will be reouired to
install the new scrubber. Existing controls are located as deoicted
in Figure 6-8. The spray tower vn'll be removed and the spray-cross-
flow packed bed scrubber installed in the vacated area. A new fan
will be required to compensate for the higher pressure drop of the
spray-crossflow packed bed scrubber. Existing water lines and pumps
will be used to supply gypsum oond water at 40 psig to the spray
section. A 16-inch line will be required to supply 2400 qom of water
at 5 psig for the packed bed. Spent scrubbing water is to be re-
cycled to the gypsum pond in the existing drainage system. Treated
gases will be emitted from a newly installed 75 foot stack.
Table 6-24 lists the major cost items involved in the retrofit
project. Operating conditions for the spray-crossflow packed bed
scrubber are presented in Table 6-25. A breakdown of the estimated
cost of the project is orovided by Table 6-26.
Table 6-24. MAJOP RETROFIT ITEMS FOP "n*EL MP-TSP PL^NT (C/V-SE A)
1 Rearrangement of ductv/ork - removal of spray tower from service
and connection of spray-crossflow packed bed scrubber and stack.
Requirements are 50 feet of 96-inch* duct.
2. Water line connecting gypsum pond with spray-crossflow packed
bed scrubber - 1600 feet of 16-inch pipe.
*flot necessarily circular, but of equivalent cross-sectional area.
6-45
-------
3. Two centrifugal pumps (one spare) - 2400 gpm, 76 feet total
dynamic head (TDH), 100-horsepov:er motor.
4. Removal of spray tower.
5. Centrifugal fan - 150,000 SCFM, 355 feet TDH, 200-horsepower
motor.
6. Spray-crossflow packed bed scrubber. Unit is designed to
handle 158,000 acfm. Using pond water at specified conditions,
scrubber must reduce fluoride concentration to 0.23 mg/scf
(9.7 ppm) when treating streams listed in Table 6-23.
7. Stack - 75 f^et tall, 9 feet diameter.
8. Supports and foundations.
Table 6-25. OPERATING CONDITIONS FOR SPRAY-CP.OSSFLOW PACKF.D
BED SCRUBBEP FOP MOPEL POP-TSP PL^IT, C«?F «
(550 Tons/Day P205)
Gas to scrubber
Flow, SCFM 150,000
Flow, DSCFM 145,500
Flow, ACFM 158,000
Temp., °F 100
Moisture, Vol. % 3.0
Fluoride (as F), Ib/hr 439
Fluoride (as F), pom 928
Gas from scrubber
Flow. SCFM 150,000
Flow, DSCFM 145,500
Flow, ACFM 156,000
Temp., °F 90
Moisture, Vol. % 3.0
Fluoride (as F), Ib/hr 4.6
Fluoride (as F), ppm 9.7
Fluoride removal,• wt 99.°
Estimated y1, ppm 0.8
Estimated NTU required 4.7
-------
GYPSUM
POND
1300'
MIXFP TONE AND
DEN VENTILATION
STOPARE BUILDINR
I
ion1
i
l>i
POP - TSP
PPODUCTION
L_:_ n nn i ^
|^- -VENTILATION
POP - TSP
STHRAFF
L^ 375' \' - >^
7 - VE.NTURI • SCRUBBER
c; - SPRAY TOUER SCRUBBER
Figure 6-8. EXISTING CONTROL EQUIPMENT LAYOUT FOP. ?10PEL pop-TSP PLANT, CASH A
-------
Table 6-26. RETROFIT COSTS FOR MODEL ROP-TSP PLANT, CASE A
(550 tons/day PgOg) November 1974
Cost ($)
A. Direct Items (installed)
1. Spray-crossflow packed bed scrubber 294,000
2. Ductwork 9.800
3. Piping 33,300
4. Pumps and motors 26,500
5. Centrifugal fan and motor 28,800
6. Removal of old equipment 12,500
7. Stack 44,000
8. Performance test 4»000
Total Direct Items 452,900
B. Indirect Items
Engineering construction expense, fee,interest on
loans during construction, sales tax, freight insurance.
(35% of A) 158,500
C. Contingency
(25% of A) 113,200
D. Total Capital Investment 724,600
E. Annualized Costs
1. Capital charges 118,100
2. Maintenance 21,700
3. Operating labor 4,000
4. Utilities 26,500
5. Taxes, insurance administrative 29,000
Total Annualized Costs 199,300
6-48
-------
Existing Controls (Case B)
In th.' .sse, it is assumed that only the iroduction area is
originally equipped vi th controls. /* Doyle scrubber is used to
treat the combined ventilation streams from the nixing coni e~J
the den. Ventilation flov/ rates and fluoride concentrations for
these sources are presented in Table 6-27. Fluoride removal efficiency
of the Doyle scrubber is approximately 59 percent. Emissions from the
production area are 95.2 pounds of fluoride pen hour with existing
controls.
The ROP-TSP storage area is currently uncontrolled. Estimated
fluoride emissions from this source are 198 oounds oer hour.
Table 6-27. FLOW RATES AND FLUORIDE CONCENTRATIONS OF EFFLUENT
STREAMS SENT TO EXISTING CONTROLS.
Emission Source i
i
Coae mixer vent gases
Curing be 1 t^y en,^ gases ('
Flow Pate
(SCFM)
500
i
14,500 j
i
Fluoride Concentration
(mg/scf) (DOPI)
I
0.71 ! 3^
160 . 68nn
I' II 'I* IH | ' | 111 1 -t'fi 1
Fxetrofit Controls (Case B)
The hooding on the curing belt is in a poor state of reoair and
will be replaced. A new hooding crrangement utilizing a flat
stationary air tight top and plastic side curtains will be used.
The ventilation rate for the belt will be increased to 24,500 SCFM.
This higher flow rate will necessitate the replacement of existing
-------
ductwork and fans. The mixing cone will continue to be ventilated
at a rate of 500 SCFK.
Control of emissions from the storage area requires the
sealing of the building (roof monitor and sides) and the installation
of a ventilation system designed to handle 125,000 SCFIi. All
associated fans, pumps, piping, and ductwork must be installed. The
ventilation stream from the storage area will be combined with the
effluent stream from the oroduction area and sent to controls. Flow
rates and fluoride concentrations associated with the various emission
sources are the s ime as listed in Table 5-23.
Fluoride emissions must be reduced to 4.6 pounds per hour in
order to meet the emission guideline of 0.2 pounds fluoride per ton
P Q input. This will be accomplished by removing the Doyle Scrubber
£. J
and installing a spray-crossflow packed bed scrubber designed for
99.3 percent fluoride removal. Figure 6-9 indicates the placement
of the retrofit scrubber. Treated gases will be emitted from a newly
installed 75-foot stack.
Gypsum pond water will be used as the scrubbing liquid. Pond
water characteristics are listed in Table 6-7. An 18-inch line will
be installed to supply the required 3450 gpm of pond water. Spent
scrubbing water is to be recycled to the gypsum pond in an existing
drainage system.
Table 6-2"; identifies the major cost items involved in the
retrofit project. Operating'condi tioris for the hevi scrubber are'
listed in Table 6-2° Estimated costs are provided in Table *-3n.
6-50
-------
GYPSUM
POilD
V
"X 1300'
I
en
MUER CONE AND
DEN VENTILATION
ROP - TSP
PRODUCTION
TOO'
STORAGE BUILDING
VpT RATION SYSTEM
ROP - TSP STORAGE
TOO1
i.
375'
SPRAY CROSSFLOW PACKED
BED SCRUBBER
Figure 6-9. RETROFIT CONTROL EQUIPMENT LAYOUT FOR MODEL ROP-TSP PLANT,
CASE B
- STACK
-------
Table 6-2S. MAJOR RETROFIT ITEMS FOR MODEL ROP-TSP PLANT (CASE B)
1. Ductwork - replacement of the curing belt ventilation system
and installation of a storage building ventilation system.
Curing belt ventilation system --175 feet of 42-inch duct
with a 50 foot branch of 6-inch duct connecting the mixing
cone. Storage building ventilation system - 150 feet of 96-
inch duct with two 160-foot branches of 66-inch duct.
2. Water line connecting gypsum pond with spray-cro.,sflo;j packed
bed scrubber - 1700 feet of 18-inch pipe.
3. Two centrifugal pumps (one spare) - 3450 gom, 74 feet
total dynamic head (TDK), 125-horsepower motor. Booster pump
for spray section - 1150 gpm, 81-feet TDH, 40-horsepower motor.
4. Centrifugal fan for curing belt ventilation system - 25,000
SCFM, 760 feet TDH, 75-horsepower motor. Fan for storage
building ventilation system - 125,000 SCFM, 725 feet TDH,
350 horsepower motor.
5. Removal cf - 1) old hooding system from curing belt and
2) 2oyle scrubber.
6. Installation of a new hooding system consisting of a wooden air-
tight top and plastic side curtains on the curinc belt.
6-52
-------
7. Sealing of the storage building - roof monitor and sides of
building.
8. Spray-crossflo;/ packed bed scrubber. Unit is designed to
handle 158,000 acfn. Us ing- pond water at specified conditions,
scrubber must reduce fluoride concentration to 0.23 mg/scf
(9.7 ppm) when treating streams listed in Table 6-23.
9. Stack - 75 feet tall, 9 foot diameter.
10. Supports and foundations.
Table 6-29. OPERATING CONDITIONS FOR SPRAY-CRQSSFLOW PACKED BED
SCRUBBER FOR MODEL ROP-TSP PLANT, CASE B
(550 Tons/Day P0)
Gas to Scrubber
Flow, SCFM 150,000
Flow, DSCFI1 145,500
Flow, ACFM 158,000
Temp. , °F 100
Moisture, Vol . % 3.0
Fluoride (as F), Ib/hr 703
Fluoride (as F), ppm 1490
Gas from Scrubber
Flow, SCFM 150,000
Flow, DSCFM 145,500
Flow, ACFM 156,000
Temp., °F 90
Moisture, Vol . % 3.0
Fluoride (as F), Ib/hr 4.6
Fluoride (as F) , ppm 9.7
Fluoride removal, wt % 99.3
Estimated y' , ppm 0.8
Estimated NTU required 5.1
6-53
-------
Table 6-30. RETROFIT COSTS FOR MODEL ROP-TSP PLANT, CASE B*
(550 tons/day PgOg) November 1974
Cost ($)
A. Direct Items (installed)
1. Spray-crossflow packed bed scrubber 294,000
2. Ductwork 89,200
3. Piping 39,800
4. Pumps and motors 35,000
5. Centrifugal fans and motors 40,800
6. Curing belt hooding 26,700
7. Sealing of storage building 80,000
8. Removal of old equipment 20,000
9. Stack 44,000
10. Performance test 4,000
11. Structural steel supports/bldg. 100,000
Total Direct Items 773,500
B. Indirect Items
Engineering const)uction expense, fee, Interest on
loans during construction, sales tax, freight insurance.
• (35% of A) 270,700
C. Contingency
(25% of A) 193,400
D. Total Capital Investment 1,237,000
E. Annualized Costs
1. Capital"charges 201,600
2. Maintenance 37,100
3. Operating labor 4,000
4. Utilities 48,200
5. Taxes, insurance, administrative 50,000
Total Annualized Costs 340,900
*In costing this model, extensive use was made of a project report dated
June 27, 1974, prepared by Jacobs Engineering company for J. R. Simplot
Co., Pocatello, Idaho.
6-54
-------
j.-anular Triple Superphosphate Procuctio-'i and Storage
Tie :iodel slant uses the "5crr-G river process '"or the production
of granular triple superphcschate. Designed production capacity is
870 tons of triple superphosphate per day (409 T/D P9Cg). Figure
4-13 orcvides a scnematic diagram of the operation.
Ground phosphete rock and phosphoric acid (38 percent Pp^c) are
contacted in a series of reactors. Tha reaction mixture is then
pumped to the granulator <:here it is mixed v:ith recycled material
from the cyclone dust ccllectcrs and the screening operations to pro-
duc-2 product sized granules of triple superphosphate. A rotary
drier is used to reduce the product moisture content to about 3 per-
cent.
Dried triple superphosphate is cooled and screened before being
sent to storage. A curing period of 3 to 5 days is provided before
the product is considered ready for shipping. Shipping cf GTSP
is on a seasonal basis, therefore, a large storage capacity is re-
quired. The storage facility has a capacity of 25,000 tons of a
triple superphosphate 01,500 tons PgOg). This building is venti-
lated at a rate of 75,000 scfm using a roof rronitor.
Existing Controls
Sases vented from the reactors and the granulator are combined
and treated in a tiro-stage system consisting of a venturi and a
cyclonic spray tower. Gypsum pond -./ater serves as the scrubbing
liquid in both units. Pond v.-ater is available at 80°F with a fluo-
6-55
-------
ride content of 0.5 percent. Additional properties are listed in
Table 6-7. Fluoride removal efficiency is 83 percent ^or tha ven-
turi scrubber and 82 percent for the cyclonic spray tower.
The drier gases are passed through cyclones for product
recovery and then treated for fluoride' removal by a tvio-stage
scrubbing system (venturi-cyclonic spray toi:er) similar to that de-
scribed for the reactor-granulator cases. Fluoride collection is 85
percent in the venturi and 86 percent in the cyclonic scrubber.
Gypsum pond water is used as the scrubbing liquid.
Miscellansois gas streams vented from the product cooling and
screening operations are a third source of emissions from ths GTSP
production facility. These streams are combined and treated for
product recovery (dry cyclone) and fluoride removal (cyclonic spray
tower). Fluoride collection efficiency of the cyclonic spray tower
is 87 percent.
Existing controls have been in operation for five years. Flow
rates and^fluoride concentrations for the various emission sources
are listed in Table 6-31. All values are estimates based on a com-
bination of source test results and published data. Total fluoride
emissions from the production facilities are 31.0 pounds per hour.
Ventilation air from the storage building is presently emitted
uncontrolled. Table 6-31 lists the estimated volumetric flov/ rate
and fluoride concentration based on source test data. Fluoride
emissions from the storage building'are 13.2 pounds per hour.
6-56
-------
Table 6-31. FLOI' RATES AND FLIDPIDE CCSCENTP*.TIONS FOR GTSP PLA.'J
EJ'ISSKT! SCURCES22-24
I.-.iission source
P.eactor-granulator gases
Drier vent gases
Cooler S screening equip-
ment gases
Storage building ventilation
Flow rat*
(SCF")
18,000
48,000
51 ,000
75,000
Fluoride concentration
(rcg/SCF) (cr>n)
84
84
16.8
1.3
3500
3500
700
54
Retrofit Controls
The retrofit project for the GTSP production facility involves
tha replacement of the cyclonic scray tover on the reactor-granula-
tor stream and on the drier stream vnth a spray-crossflov; packed bed
scrubber. £ third spray-crossflow packed bed unit will be installed
on the miscellaneous stream to provide secondary scrubbing. The
A
new control system is designed to reduce fluoride emissions from the
production operation to 3.34 pounds per hour. This emission rate is
equivalent to the emission guideline of 0.2 pounds fluoride per ton
input.
Figure 6-10 shows the position of existing controls. Retrofit
plans call for the removal of the cyclonic spray towers treating the
reactor-granulator and the drier gases and the installation of spray
crossflow necked bed scrubbers in the vacated areas. The soray-
crossflo1./ packed bad scrubcer for the piscellaneous stream ---ill
oe located adjacent to the preliminary scrubber as indicated in
Ficure 6-11. 6-5/
-------
GYPSUM
POND
01
00
1200'
DRIER
MISCELLANEOUS
®
GTSP
PRODUCTION
STACK
GTSP
J
HH STORAGE
\
125'
1
REACTOR - GRANULATOR
k-
— Venturi Scrubber
— Cyclonic Scrubber
FIGURE 6-10. EXISTING CONTROL EQUIPMENT LAYOUT FOR MODEL GTSP PLANT.
-------
GYPSUM
POND
(Tl
I , '
cn
1200'
MISCELLANEOUS
DRIER ^° ET° ^ STACK
GTSP
PRODUCTION
GTSP
STORAGE
T
125'
1
REACTOR-GRANULATOR
-150-
300'
O — CYCLONIC SPRAY TOWER
v VENTURI SCRUBBER
D SPRAY CROSS-FLOW PACKED BED SCRUBBER
FIGURE 6-11. RETROFIT CONTROL EQUIPMENT LAYOUT FOR MODEL GTSP PLANT.
-------
Existing pumps, fans, piping and ductwork will be utilized
wherever possible. The existing piping system will be used to
supply water to the three preliminary scrubbers and the spray
sections of the secondary (spray-crossflow packed) scrubbers on the
reactor-granulator and the drier streams. Some minor alteration in
the piping arrangement will be required because of changes in the
scrubber geometry. A 16-inch line will be installed to provide ?]6p
gpm of water at 5 psig for the spray-crossflov packed bed unit on tfja
miscellaneous stream and the packed sections of the secondary scrub-
bers on the reactor-granulator and the drier streams. Duplicate
pumps, one on stand-by, will be provided for this service. In all
cases, the spent scrubbing liquid will be recycled to the gypsum
pond using the existing plant drainage system.
Some alteration of existing ductwork will be required to install
the retrofit scrubbers. A new fan will be installed on the miscellaneous
stream to compensate for the pressure loss caused by the secondary
scrubber.
Control of emissions from the GTSP storage facility requires
the sealing of the roof monitor and the installation of 350 feet of
ventilation ducting. Ventilation air will be treated in a spray-
cross flow packed bed scrubber before being emitted. The unit is
designed to reduce fluoride emissions to 1.25 pounds per hour; a rate
equivalent to, emission guideline under most conditions. All associated
fans, pumps, piping, and ductwork must be installed. The existing plant
6-60
-------
drainage system will be used to recycle gypsum oond water.
Fiaure 6-11 provides a vie*-/ of the equipment layout.
All major retrofit items are tabulated in Table 5-32.
Table 6-33 provides a list of operating conditions for the four
retrofitted spray-crossflow packed bed scrubbers. Table 6-34 pre-
sents the retrofit project, costs.
Table 6-32. WUOR RETROFIT ITE'IS FOP. MODEL GTSP PLANT
GTSP Production
1. Rearrangement of ductwork - removal of existing cyclonic scrubbers
on reactor-granulator and drier streams and connection of
replacement spray-crossflow packed bed scrubbers. Installation
of third spray-crossflov packed bed unit on miscellaneous
stream. Requirements are 150 feet of 60-inch diameter duct and
50 feet of 42-inch duct.
2. New water line connecting gypsum pond with retrofitted scrubbers -
* 1200 feet of 16-inch pipa with 200-foot branch of 14-inch pipe
to scrubbers treating the drier and miscellaneous streams and 150
foot branch of 5-inch pipe to the reactor-granulator scrubber.
3. Two centrifugal pumps, each 2160 gpm, 105 feet total dynamic
head (TDH), 100-horsepower motor. Booster pump for spray
section of spray-crossflow packed bed scrubber on rriscellaneous
stream - 374 gpm, 89 feat TDH, IC-horsepower motor.
5-61
-------
Table 6-32. MAJOR RETROFIT ITEMS FOR MODEL 6TSP PLANT (cont.)
4. Centrifugal fan for miscellaneous stream - 51,000 scfm,
356 feet TDK, 75-horsepower motor.
5. Removal of cyclonic scrubbers on reactor-granulator and
miscellaneous streams.
6. Three spray-crossflow packed bed scrubbers. Design parameters
are provided in Table 6-33. Using pond water at specified
conditions, the scrubbers are required to meet the indicated
emission levels when treating the gases described in Table 6-31.
7. Supports and foundations.
GTSP Storage
1. Sealing of roof monitor and installation of ducting - 350 feet of
78-inch ducting for ventilation of building and connection of
scrubber.
Z. Water line connecting gypsum pond with spray-crossflow packed
bed scrubber - 1700 feet of 12-inch pipe.
3. Centrifugal pump - 1730 gpm, 81 feet TDH, 60-horsepower motor.
Booster pump for spray section - 580 gpm, 89 feet TDH, 15-
Korsepower motor.
4, Centrifugal fan r 75.000 scfm, 630 feet TDH, 200 horsepower
motor.
$-62
-------
Table 6-32. MAJOR RETROFIT ITEMS FOR MODEL GTSP PLANT (cont).
5. Spray-crossflow packed bed scrubber. Using specified pond
water, scrubber must reduce fluoride concentration of venti
lation stream to 0.13 mg/scf (5.1) when treating the gases
described in Table 6-31.
6. Supports and foundations.
7. Stack - 50 feet tall, 6 foot diameter.
6-63
-------
Table 6-33.
OPERATING CONDITIONS FOP SPP.AY-CROSSFLOW
PACKED BED SCRUBBERS FOP flODEL GTSP PLANT
(400 Tons/Day P2^5)
-3 to Scrubber
Flow, SCFM
Flov, PSCP'
Flow, ACFfl
Tenp., °F
Moisture, vol. *
Fluoride (as F), Ib/hr
Fluoride (as F), ppm
Gas from Scrubber
Flow, SCFil
Flow, DSCFf
Flew, ACR1
Temp., °F
!ioisture, vol. %
Fluoride (as F), Ib/hr
Fluoride,, (as F), ppm
Fluoride removal , wt %
Estimated y1, ppm
Estimated MTU required
Product!
Reactor
18,000
16,560
19,^00
110
8.0
28
490
16,850
16,560
17,500
°0
2.0
1.00
17.5
96.5
n.n.5
3.38
on
Drier
48,000
44,160
52,500
120
8.0
79.8
525
45 ,050
44,160
46,800
90
2.0
1.76
11.5
97.8
0.95
3.90
Cooler
51 ,000
48,450
54,900
no
5.0
14.8
92
49,400
48,450
51 ,200
90
2.0
0.63
3.9
96.0
^.85
3.39
Storaos
"entilation
75,000
74 ,480
77,100
87
0.7
13.2
54.1
76 ,000
74,480
78,100
85
2.0
1.25
5.1
90.5
0.7
2.49
6-64
-------
Table 6-34. RETROFIT COSTS FOR MODEL
GTSP PLANT (400 tons/day P205) November 1974
Cost ($)
A. Direct Items (installed) ""•
1. GTSP Production
a. Spray-crossflow packed bed scrubbers (3) 261,000
b. Ductwork 22,800
S- "P1nS J 26,200
d. Pumps and, motors 19,700
e. Removal of old equipment 18*000
f. Performance test 4*.000
g. Centrifugal fan and motor 14*400
2. GTSP Storage ' '
a. Cross flow packed scrubber 150 000
b. Ductwork 56'600
c. Piping 27,800
d. Pumps and motors 15J200
e. Centrifugal fan and motor 23*000
f. Structural steel supports/bldg. 50*000
g. Sealing of storage building 10,000
h. Performance test 4,000
Total Direct Items 702,700
B. Indirect Items
Engineering construction expense, fee, interest on
loans during construction, sales tax, freight insurance.
(35% of A) 245,900
C. Contingency
(25% of A) 175j700
D. Total Capital Investment 1,124,400
E. Annualized Costs
1. Capital charges 183,300
2. Maintenance 33 300
3. Operating labor 6*000
4. Utilities ^O
5. Taxes, insurance, administrative 44,900
Total Annualized Costs 308,600
6-65
-------
6.1.3.2 Retrofit Case Descriptions
General Procedure
This section describes two actual cases in whlcii control
systems containing spray-crossflow oacked bed scrubbers were added to
existing production'facilities. Each case description provides the
following information:
1. A description of the process in use,
2. Identification of the original fluoride controls and sources
treated,
3. A description of the retrofit project, and
4. Retrofit costs.
Case A
Case A involve&*
-------
Original Controls
Fluoride control was initially provided by a spray towor installed
in 1953 as part of the orioinal plant desicn. Gypsum oond water was used
as the scrubbing liquid. Ventilation streams from the drier and the
product screens were sent to the spray tower while both reactor and
granulator gases were vented directly to the atmosnhere. The sprav
tower was improved in 1964 by the addition of more sprays and a mist
elimination section. Performance data for this system is not available.
Retrofit Controls
The spray tower was removed in 1966 as part of a retrofit project
and replaced by a three stage scrubbina system. Gases vented from the drier
(60,000 acfm) and the screens (40,000 acfm) are now treated in seoarate. venturi
scrubbers, combined, passed through a cyclonic scrubber, and finally
treated in a spray-crossflow packed bed scrubber. Operating characteristics
of these units are listed in Table 6-35. Pond water serves as the
scrubbing liquid for the entire system. Controls for the reactor and the
granulator were not added at this time.
*t
All associated fans, pumps, piping, ductwork, and stacks were installed
as part of the retrofit project. New pond water supply and drainaqe svstems
were also required.
Designed fluoride removal efficiency is 99+ percent. Tests
conducted by the Environmental Protection Aqency in June 1972 measured
fluoride removal efficiencies ranqino uo to 99.6 percent.
C-67
-------
Table 6-35. OPERATING CHARACTERISTICS OF SCRUBBERS IN RETROFIT C^SE A
Scrubber type
Scrubbing liquid
to gas ratio (pal/SCF)
Gas stream
pressure drop(in. H20
Drier venturi
Screen venturi
Cyclonic scrubber
Spray-crossflow
packed bed scrubber
0.008
0.006
0.007
0.002
12-15
8-13
4-6
2-6
Retrofit Costs
Total installed cost of the retrofit control equipment was $368,000,
however, this does not include the cost of removing old eouipment or of
adding new pond water supply and drainage systems. The annual operating
cost is reported to be $51,000.
Case B
Case B is similar to Case A in most respects. The facility involved
is a granular triple superphosphate plant built in 1953. This plant also
uses' the Dorr-Oliver process for GTSP. Annual capacity is approximately
200,000 tons triple superphosphate. Space limitations are similar to those
described in Case A.
6-68
-------
Original Controls
Emissions from the drier and the screening area ware controlled bv
a spray tower which had been installed as part of the original plant
design. Fluoride removal efficiency data is not available for this system.
Reactor and granulator gases were vented to the atmosphere without treatment.
Retrofit Controls
The retrofit project consisted of the removal of the spray tower and
its replacement by a system similar to that described in Case A. Controls
are in three stages - 3 Venturis in oarallel followed by a cyclonic scrubber
and a spray-crossflow, packed bed scrubber. Effluent streams from the drier
and the screens are treated in separate Venturis, combined with the oases
from the third venturi, and sent to the remaining controls. The third
venturi treats gases from either an adjacent wet acid olant or a nearby
run-of-pile triple superphosphate plant. Designed capacity of the control
system is 115,000 acfm. Gypsum pond water serves as the scrubbina liquid.
Controls for the reactor and the aranulator v/ere not installed as a part of
this project.
The retrofit controls were added in 1972. All associated fans, pumps,
piping, and ducting were installed as part of this project. Fluoride removal
efficiency of the system is reported to be 99+ percent.
Retrofit Costs
Total installed cost for the retrofit controls was reported to be
$760,000. Table 6-36 lists a breakdown of the cost. Demolition costs
and the cost of adding new pond water supnlv and drainage systems are
not included. i!o ooeratinn costs v/ere provided.
r-fg
-------
Table 6-36. CASE B RETROFIT PROJECT COSTS
Item
Installed Cost
(dollars)
Foundations
Structural steel
Blowers and motors
Wet scrubbers
Pumps, sumps and piping
Ducts and stack
Electrical and instruments
81,000
52,000
85,000
218,000
175,000
102,000
47,000
6-70
-------
6.2 VEi.'TURI SCRUBBER
6.2.1 Description
Venturi scrubbers are primarily participate collection devices,
however, tney are also applicable to .gas absorption work and are in
widespread use throughout the phosphate fertilizer industry. They are
particularly well suited for treating effluent streams containing large
amounts of solids or silicon tetrafluoride because of their high solids
handling capacity and self-cleaninn characteristics. Operational reliability
and low maintenance requirements are major reasons for the popularity of
this scrubber design.
A venturi provides a high degree of gas-liauid nixing but the
relatively short contact time and the cocurrent flow of the scrubbing
liquid tend to limit its absorption capabilities. When treating effluent
streams requiring a high degree of fluoride removal, Venturis are often
used as the initial component in a multiple-scrubber system.
Two types of venturi scrubbers, gas actuated and water actuated, are
in general use. In both cases, the necessary gas-liquid contacting is
obtained from velocity differences between the two phases and turbulence
in the venturi throat. Both types also reouire the use of a mist elimination
section for removal of entrained scrubbing liquid. The ma.icr difference
between the designs is the source of motive power for oneratinq the scrubber.
In the cast of the gas actuated venturi, the velocity of the gas stream
provides the energy required for gas-liauid contacting. The scrubbino
liquid is introduced into the oas stream at fie throat of the venturi
71
-------
and is broken into fine droplets by the acceleration cas
stream. Pressure drop across the scrubber is generally high - from
8 to 20 inches of water. A fan is required to compensate for this
loss in gas stream pressure. Figure 6-12 provides a schematic
diagram of a gas actuated venturi.
A water acutated venturi is pictured in Picture 6-13. In this
case, the scrubbing liquid is introduced at a high velocity through
a nozzle located upstream of the venturi throat. The velocity of the
water streams is used to pump the effluent gases through the venturi.
Drafts of up to 8 inches of water can be developed at hi oh liauid
25
flow rates.
The removal of the fan from the system makes the water actuated
venturi mechanically simpler, more reliable, and less costly
than the gas actuated type. An additional advantaae is its relative
insensitivity to variations in the gas stream flow rate?6 Gas
actuated Venturis rely upon t^e gas stream velocity for the energy
for gas-liquid contacting, therefore, variations in the aas flow can
greatly affect scrubber efficiency. The performance of the water-
actuated venturi depends mainly on the liquid strean- velocity.
Water actuated Venturis find application Drincioally as aas
25
absorption units. Their use is usually linited, however, to small
gas streams with moderate scrubbing requirements. The water-actuated
venturi is seldom used for gas flows greater than 5,000 ac^m because
2fi
of tne large water retirements.
5-72
-------
AIR
INLET
WATER
INLET
VEMTURI
AIR
OUTLET
CYCLONIC
MIST ELIMINATION
SECTION
WATER
OUTLET
FIGURE 6-12. GAS ACTUATED VENTURI SCRUBBER WITH CYCLONIC MIST ELIMINATOR.
SPRAY
NOZZLE
AIR
OUTLET
WATER
INLET
AIR
INLET
SEPARATOR
WATER
OUTLET
FIGURE 6-13.
WATER ACTUATED VENTURI.
6-73
-------
6.2.2 Emission Reduction
No wet-acid plant using a venturi scrubber was tested by the
Environmental Protection Agency, however, fluoride absorption efficiency
ranging from 84 to 96 percent have been reported for water-actuated
27
Venturis treating wet-acid plant effluent gases. Performance data was
obtained for venturi scrubbers installed in superphosphoric acid and
diammonium phosphate plants. This infornation is presented in Table 6-37.
Several additional plants (DAP, GTSP, ROP-TSP) were tested at which venturi
scrubbers were used as the preliminary scrubber in a two or three staqe
system. Performance data for the overall systems are presented in Tables
6-3 and 6-40.
Table 6-37 VENTURI SCRUBBER PERFORMANCE IN SUPERPHOSPHORIC ACID AND
DIAMMONIUV PHOSPHATE PLANTS 28
Type of plant
Vacuum evapora-
tion SPA
DAP
Sources controlled
barometric conden-
ser, hotwell , and
product cooling tank
reactor, granula-
tor, drier, and
cooler
Control
system
water
actuated
venturi
3 gas
actuated
Venturis
in para-
llel
Scrubbing
liquid
pond
water
weak acid
(20-22%
W
Fluoride emissions3
(Ib F/ton P205)
0.0009
0.129
Average of testing results
5-74
-------
6.2.3 Retrofit Costs for Venturi Scrubbers
This section evaluates the costs involved wit:-, retrofittlnp
venturi scrubbers in a diammonium phosohate plant. Venturis
might be used to provide fluoride control for this source because
of their high solids handling capaJilitv. Cnlv the ret.rrfi': model
approach will be used to provide costs.
The model plant is the same as described in section 6.1.3.1.
To avoid repetition, only a summary of retrofit controls, e list
of major retrofit items, and a Dreakdov:n of costs rill be orasented
here.
The general aspects of the retrofit project are the same as
described in Section 6.1.3.1. Gas-actuated Venturis will be used
as fluoride scrubbers on the reactor-granulator, the drier, and
the cooler streams. Pumping and fan requirements differ from those
presented in section 6.1.3.1. An existing line win be used to
supply part of the water requirement. Table 6-38 provides a list
of major retrofit Items required. Costs are presented in Table
6-39.
Table 6-3S. flAJOR RETROFIT ITEMS FOR MODEL DAP PLANT
1. Ductwork - removal of cyclonic spray tower from service and
connection of three gas-actuated venturi scrubbers. Reauire-
ments are 100 feet of 50-inch duct and 50 feet of 54-inch duct.
2. Hater line connecting gypsum pond with venturi scrubbers -
1200 feet of 16-inch pipe with 200-foot branch of H-inch
6-75
-------
pipe and 150-foot branch of 6-inch cine.
3. Two centrifugal Dunns (one spare) - 2550 gorc, 195
feet total dynamic head (TDK), 150 'lorsenower rotor.
4. Three centrifugal fans: one for the reactor-oranulator
stream, one for the dri?r stream, and one for the cooler
stream. Peactor-granulator fan - 30,000 scfm, 713 feet TDH,
75 horsepower motor. Drier stream fan and cooler stream
fan - 45,000 scfm, 713 feet TDH, 125 horsepower rotor.
5. Removal of cyclonic spray tower.
6. Three venturf scrubbers equipped with mist eliminator
sections. When using specified pond water and treatino
gases described in Table 6-19, scrubbers are reouired to obtain
performance indicated in Table 6-21.
7. Supports and foundations.
-------
Table 6-39. RETROFIT COSTS FOR MODEL DAP PLANT
(500 Tons/Day P205) November 1974
Cost ($)
A. Direct Items (installed)
1. Venturi scrubbers (3) 17 000
2. Ductwork 26'500
3. Piping 32 300
4. Pumps and motors ^p» ww
5. Centrifugal fans and motors «
.
6. Removal of old equipment QQQ
7. Performance test H'u
Total Direct Items 312,400
B. Indirect Items
Engineering construction expense,
fee, interest on loans during
construction, sales tax, freight
insurance (35% of A.) 109>JUU
C. Contingency (25% of A.) 78'100
D. Total Capital Investment 499,800
E. Annual!zed Costs
1. Capital charges 81,500
2. Maintenance -'OQQ
3. Operating labor 31'000
5: ^'insurance, administrative 20^,000
Total Annualized Costs 151,500
6-77
-------
5.3 SPRAY TOl'ER SCRUBBER
6.3.1 Description
Spray towers provide the interonase contacting necessary for
gas absorption by dispersing the scrubbing liouid in the gas phase
in the form of a fine spray. Several types of spray towers are in
general use. The simplest consists of an empty tower equipped with
liquid sprays at the top and a pas inlet at the bottom. Scrubbing
liquid is sprayed into the gas stream and droplets fall by -jravity
through a upv/ard flow of gas. This design has the advantages of a
very low pressure drop and an inexpensive construction cost but it can
provide only about ons transfer unit for absorption. Entrainment of
scrubbing liquid is also a problem.
Cyclonic spray towers eliminate the excessive entrainnent of
scrubbing liquid by utilizing centrifugal force to remove entrained
droplets. Figure 6-14 is a schematic diagram of a typical desicm.
In this case a tangential inlet ifused to.impart'the spinning
motion to the gas stream. Water sprays are directed parallel to the
gas flov: providing crossflow contacting of the gas and liquid streams.
Pressure drops across the scrubber ranges from 2 to 8 inches of water.
Solids handling capacity is high, however, a'.sorptic/1. caoacity is
29 ?0
limited to about two transfer units. '
6.3.2 Emission Reduction
Fluoride removal efficiencies ranging from 84 to 95 percent have
been reported for cyclonic spray towers treating wet acir* plant
6-78
-------
CLEAN GAS OUT
GAS IN
CORE BUSTER DISK
SPRAY MANIFOLD
DAMPER
WATER WATER
OUT IN
FIGURE 6-14. CYCLONIC SPRAY TOWER SCRUBBER.
31
effluent nases. Table 6-40 presents nerformance data obtained bv
the Environmental Protection Anencv for cvclonic soray tov/ers installed
in wet-process phosphoric acid, diannoniun ohosohate, and run-o^-oile
triple1 superphosphate plants. In most cases, the control system con-
sisted o^ a primary venturi scrubber or cyclonic sorav tower followed
by a secondary cyclonic spray tower. Gypsum oond water was used as
the scrubbing solution except where indicated.
6.3.3 Retrofit Costs for Cvclonic 5-irav Towers
This section will use the retrofit model approach to estimate
the costs involved with the installation of 'cyclonic snrav tov/ers in
? nOD-TSD olant. Control svstems utilizing cyclonic snrav tov/ers are
canable of providinn the collection efficiency necessary to reet
the emission guideline of 0.2 pounds fluoride per ton PpCv input.
f>-79
-------
Table 6-40. CYCLONIC SPRAY TOWER PERFORMANCE IN WET-PROCESS PHOSPHORIC ACID,
DIAMMONIUM PHOSPHATE, AND RUN-OF-PILE TRIPLE SUPERPHOSHATE PLANTS
Type of plant
Sources controlled
Primary controls
Secondary controls :luoride emissions9
Ib F/ton P205)
en
oo
o
WPPA
DAP
ROP-TSP
ROP-TSP
reactor, filter, and
miscellaneous sources
reactor, granulator,
drier, and cooler
mixing cone, den,
transfer conveyor,
and storage pile
mixing cone, den,
and storage pile
two-stage cyclonic
spray tower
3 cyclonic spray
tower scrubbers in
parallel. Scrub-
bers treating re-
actor-granular
and drier gases
use weak (28-30%
P205) acid
venturi scrubber
2 cyclonic spray
tower scrubbers
in parallel
2 cyclonic spray
tower scrubbers in
parallel treating
reactor-granulator
and drier gases
cyclonic spray tower
scrubber with packed
bed section
2 cyclonic spray tower
scrubbers in parallel
0.056
0.380
0.19/K 0.21V
OJ25
aAverage of testing results
bSecond series of tests
-------
The nodal plant is tne sane as described in section 6.1.3.1
(Case /"•). How rates and fluoride concentrations of the various
effluent streams ere listed in Table 6-23. -ases venter1 from ti-,e
cone rrixer and den are presently treated in a 20,000 cfm ver.turi,
combined with the storaqe buildinn ventilation stream and sent to a
spray tov/er. The storaae buildino ventilation air is sent directly
to the spray tower. Total fluoride emissions are 127 oounds oer
hour with existinp controls.
The'retrofit project involves the removal of the existinn scrubbers
and the installation of a new control system consist!"nq of orelim.inarv
cyclonic spray towers on the ventilation streams *rom the production
and storage areas followed by a secondary cvclonic sprav tower treatino
the combined effluent streams. This system will reduce fluoride
emissions to 4.6 pounds per hour which is equivalent to the emission
guideline.
-Retrofit controls will be located as shown in Flrure 6-15. Mod-
erate rearrangement of the ductwork is necessary to install the
cyclonic spray towers. Two new fans will be required because o* the
higher pressure drop associated with the retrofit system. Fxistino
water lines and numps will be use^ to supply the orelirinarv scrubbers.
A 14-inch line will be installed to provide 1725 cipn of oond water
for the secondary scrubber. Scent scrubbino v/ater will be recvcled
to the ayosum pond in the existina drainage svsteir. Treated oases
will be emitted from a newly installed 75 foot stacK
F-81
-------
HYPSUM
POND
cr>
i
oo
ro
.1300'
MIXEP. CONE AND
DEN VENTILATION
STORAGE PUILDIMR
/VENTILATION
pvnP_T$p
PRODUCTION
1 ^ -i nn 1 »j
pnp-TSP
STORAGE
rf __ 77R1 T"
f
inn1
1
O — CYCLONIC SPRAY TCWEP. SCRUBBER
Fiqure 6-15. RETROFIT CONTROL EOUIPMENT LAYOUT FOR MODEL RO^-TSP PLANT,
-------
Table 6-41 lists the major cost items involved in this retrofit
project. Operating conditions for the three cyclonic spray towers are
provided in Table 6-42. Retrofit costs are estimated in Table 6-43.
Table 6-41. MAJOR RETROFIT ITEMS FOR MODEL ROP-TSP PLANT
1. Rearrangement of ductwork - removal of venturi and spray tower
from service and connection of three cyclonic spray towers and
stack. Requirements are 50 feet of 42-inch duct and 125 feet
of 96-inch duct.
2. Water line connecting gypsum pond with cyclonic spray tower
treating the combined effluent streams from the production and
the storage area - 1600 feet of 14-inch pipe.
3. Centrifugal pump - 1725 gpm, 167 feet total dynamic head (TDH),
125-horsepower motor.
4. Removal of venturi and spray tower.
5. Centrifugal fan for the storage building ventilation system -
125,000 SCFM, 514 feet TDH, 250 horsepower motor. Centrifugal
fan for the combined effluent streams - 150,000 SCFM, 461 feet
TDH, 175 horsepower motor.
6. Three cyclonic spray tower scrubbers. When using pond water
specified in Table 6-7 and treating the effluent streams described
in Table 6-23, scrubbers are required to obtain the performance
indicated in Table 6-42.
6-83
-------
7. Stack - 75 feet tall, 9 feet diameter.
8. Supports and foundations.
Table 6-42.
OPERATING CONDITIONS FOR CYCLONIC SPRAY TOWER SCRUBBERS
FOR MODEL ROP-TSP PLANT
(550 Tons/Dav P00C)
2 b
Mixing cone and den
ventilation stream
Gas to scrubber
Flow, SCFM
Flow, DSCFM
Flow, ACFM
Temp., "F
Moisture, Vol. %
Fluoride Gas F), Ib/hr
Fluoride (as F),
ppm
Gas from scrubber
Flow, SCFM
Flow, DSCFM
Flow, ACFM
Temp., °F
Moisture, vol. %
Fluoride (as FL Ib/hr
Fluoride (as F), ppm
Fluoride removal, wt %
Estimated y*, ppm
Estimated NTU required
25,000
24,500
28,400
140
2
307
4,000
25,300
24,500
27,500
115
3
20.5
260
93
0.8
2.7
Storage building Combined
ventilation stream streams
125,000 150,000
122,500 1*5,500
128,200 154,000
85 85
2 3
396 50.5
1,000 107
126,000 150,000
122,500 145,500
128,500 153,000
80 80
3 3
30 4.6
76 9.7
92.5 91
0.8 0.8
2.6 2.5
8-84
-------
Table 6-43. RETROFIT COSTS FOR MODEL ROP-TSP PLANT
(550 Tons/Dav P2°5) November 1974
Cost ($)
A. Direct Items (installed)
1. Centrifugal spray tower scrubbers (3) 300,000
2. Ductwork 25,000
3. Piping 29,100
4. Pump and motor 19,100
5. Centrifugal fans and motors 54,400
6. Removal of old equipment 12,500
7. Stack 44,000
8. Performance test 4,000
Total Direct Items 488,100
B. Indirect Items
Engineering construction expense,
fee, interest on loans during
construction, sales tax, freight
insurance (35% of A.) 170,800
C. Contingency (25% of A.) 122,000
D. Total Capital Investment 780,900
E. Annualized Costs
1. Capital charges 127,300
2. Maintenance 23,400
3. Operating labor 6,000
4. Utilities 48 ',600
5. Taxes, insurance, administrative 31,400
Total Annualized Costs 236,700
6-85
-------
6.4 IMPINGEMENT SCRUBBER
Impingement scrubbers are primarily participate collection
devices but they also possess some absorption capability and have
been used with limited success to treat effluent streams from wet-
process acid and diammonium phosphate plants. The Doyle scrubber
pictured in Figure 6-16 is the type most commonly used by the
fertilizer industry.
Al* LOCK RELEASE
ELIMINATOR
LIQUID INLET
FIGURE 6-16. DOYLE SCRUBBER.
Effluent gases are introduced into the scrubber as shown in
Figure 6-11. The lower section of the inlet duct is equipped with a
axially located cope that causes an increase in gas stream velocity
prior to its impingement on the surface of the pond. The effluent
gases contact the pool of scrubbing liquid at a hinh velocity and under-
go a reversal in direction. Solids impinge on the liquid surface and
are retained while absorption of gaseous fluorides is promoted by the
interphase mixing generated bv impact. Solids handlinq capacity is
33
high, however, absorption capability is very limited.
6-86
-------
6.5 SUMMARY OF CONTROL OPTIONS
Sections 6.1 through 6.4 have examined the operational charac-
teristics of several scrubber designs commonly used in the phosphate
fertilizer industry. Only the spray-crossflow packed bed scrubber is
capable of providing the degree of fluoride control required to meet
SPNSS emission levels in all cases. In certain cases, cyclonic spray
tower scrubbers will meet the standards, but only at a higher cost as the
RDlMSf rdti«of1t example Illustrates (Table 6-44). Although retrofit
costs for installing venturi scrubbers in a DAP plant were lower than
those for spray-crossflow packed bed scrubbers, there is no data
available which substantiates that a venturi scrubber alone can achieve
SPNSS emission levels. The primary value of venturi scrubbers in
fluoride control is their higher solids handling capacity. This feature
is exploited in several spray-crossflow packed bed scrubber designs
which Incorporate a preliminary venturi scrubber.
Table 6-44. ESTIMATED TOTAL CAPITAL INVESTMENT AND ANNUAL IZED COST
FOR DAP AND ROP-TSP RETROFIT MODELS USING SPRAY-CROSS-
FLOW PACKED BED AND ALTERNATIVE SCRUBBERS.
November 1974.
Facility Type of Scrubber Capacity Total Capital Ann ua 11 zed
(tons/day Investment Cost
DAP
DAP
ROP-TSP
Spray-crossflow
packed bed
Venturi
Spray-crossflow
__•_•• •
500
500
550
$659,100
499,800
724,600
$179,000
151,500
199.300
packed bed
ROP-TSP Cyclonic spray 550 780,900 236,700
tower
6-87
-------
6.6 DESIGN, INSTALLATION, AND STARTUP TIMES
This section discusses the time required to procure and install
a wet scrubber on a phosphate fertilizer operation. Actual time
requirements can vary tremendously depending upon such factors as
space limitations, weather conditions, lack of available utilities,
delayj in equipment delivery, and lack of engineering data. The
Information presented in this section,-has to a limited extent,
attempted to take such factors into consideration. Since these
estimates are general, however, they should be used primarily as a guide*
Hne and may be modified as dictated by specific circumstances.
Figure 6-17 identifies the various steps involved in the procurement and
installation of a wet scrubber on a wet-process phosphoric acid plant. It
also provides an estimate of the total time requirement of the project. In
estimating this time requirement, it was assumed that those activities leadina
up to the finalization of control equipment plans and specifications had been
completed prior to the initiation of the retrofit project. The individual
steps shown in Figure *-l* are explained in more detail in Table 6-45.
6-88
-------
00
VO
FIGURE 6-17 TINE SCHEDULE FOR THE INSTALLATION OF A WET SCRUBBER ON A WET-PROCESS
PHOSPHORIC ACID PLANT34
0., P,
n
MILESTONES
2
3
4
5
ACTIVITIES
Design otion
A-C
A-B
C-0
D-E
E-f
F-G
G-l
1-H
Milestones
Activity and durarlan in weeks
Dare of lubmitlal of final control plan ro appropriate agency •
Date of award of control device contract.
Date of Initiation of on-site construction or Installation of emission control equipment.
Date by which on-iite construction or Installation of emission control equipment Is completed.
Date by which final compliance is achieved.
H-I
1-2
2-J
ELAPSED TIME (WEEKS)
Preliminary investigation
Source tests
Evaluate control alternatives
Commit funds for total program
Prepare preliminary control plan and compliance
schedule for agency
Agency review and approval
Finalize plans and ipecifications
Procure control device bidi
Evaluate control device bids
Award control device contract
Vendor prepares assembly drawings
Designation
J-K
K-L
L-M
K-N
N-0
0-P
P-3
3-M
M-Q
Q-4
4-5
Review and approval of assembly drawings
Vendor prepares fabrication drawings
Fabricate control device
Prepare engineering drawings
Procure construction bids
Evaluate construction bids
Award construction contract
On-site construction
Install control device
Complete construction (system tie-in)
Startup, shakedown, preliminary source test
-------
Table
DESCRIPTION OF INDIVIDUAL ACTIVITIES INVOLVED IN THE PROCUREMENT, INSTALLATION, AND
STARTUP OF CONTROL EQUIPMENT.35
ACTIVITY
CODE
ACTIVITY
DESCRIPTION
DETAILS OF ACTIVITY AND ESTIMATED TIME REQUIREMENT
G-l
Finalize plans and specification
1-H
Procure control device bids
en
i
vo
o
H-I
Evaluate control device bids
1-2
Award control device contract
The control system is specified in suffient detail for
control equipment suppliers and contractors to prepare
bids. A final.control plan summarizing this information
is also prepared for submittal to the appropriate aqency.
Two to six weeks are allocated for this activity. The
variation is dependent on the magnitude and complexity
of the project.
Transmittal of specifications for the control device and
request for bids from suppliers.
A minimum time of four weeks is required to procure bids
on small jobs. A maximum of twelve weeks should be allowed
for large non-standard units. Initial vendor quotations
frequently do not match bid specifications, thereby requiring
further contacts with each bidder.
The bids are evaluated and suppliers are selected.
Two to five weeks are required for evaluating control device
bids. Small, privately owned firms will require little time,
whereas in large corporations, the bid evaluation procedure
often involves several departments thereby increasing the
time requirements.
The successful bidder is notified and a contract is signed.
A minimum of two weeks should be allocated for preparing
the final contract papers and awarding contracts for the
control device and other major components. This activity
will take longer in large corporations where examination and
approval of the contract by several departments is required.
-------
.(continued). DESCRIPTION OF INDIVIDUAL ACTIVITIES INVOLVED IN THE PROCUREMENT,
INSTALLATION, AND STARTUP OF CONTROL EQUIPMENT.
35
ACTIVITY
_C_ODE
2-J
ACTIVITY
DESCRIPTION
Vendor prepares shop drawings
VO
J-K
Review and approval of
assembly drawings
K-L
Vendor prepares fabrication
drawings
DETAILS OF ACTIVITY AND ESTIMATED TIME REQUIREMENT
The vendor prepares the assembly drawings for the
control device. For the smaller and more common types of
control equipment, standard shop drawings which apply to
XEIL0?11*?1 e^1P"ent size ranges may be used with the
appropriate dimensions underlined or otherwise indicated
JISndel!1ce!£ U may be necessary *> Prepare drawings
ll£ f°rthe Eroject at hand- Tne drawings are
H Cl-6nt f°rn hl's approval Pr1or to ^nitrating
« " d!;aw:n9S. Depending on the complexity and
originality of the design, the time required by the vendor
to submit assembly drawings could vary from feJ weeks to
few months. Two to six weeks are estimated for this activity.
The client reviews the assembly drawings and gives approval
to begin fabrication drawings. The client also uses the
assembly drawings to prepare the necessary engineering drawings
One to two weeks are sufficient for review and approval of "
assembly drawings. The longer time is required for any delay
in approval as a result of revisions and modifications'".
Upon receipt of approval from client to proceed with con-
struction of the control device, the vendor prepares fab-
rication or shop drawings which will be used in the manu-
facturing and assembling of the control equipment. Three
to eight weeks are normally required for this task.
-------
Table «*«S(continued). DESCRIPTION OF INDIVIDUAL ACTIVITIES INVOLVED IN THE PROCUREMENT,
INSTALLATION, AND STARTUP OF CONTROL EQUIPMENT. 35
ACTIVITY
CODE
ACTIVITY
DESCRIPTION
ACTIVITY AND ESTIMATED TIME REQUIREMENT
Prepare engineering drawings
L-M
01
I
<£>
ro
Fabrication of Control
device
This is the time which is required by the client
(or his consultant) to prepare an engineering
drawings package for use by the construction company
for installing foundations, structures, ductwork,
electrical outlets and any other Items not supplied with
the control device. The drawings will also show the location
and tie-in of the control device. Estimated engineering ttme
for the project in question is 10 weeks.
On small size control devices which can be shop assembled,
this activity represents the fabrication, assembly, and
delivery of the control unit to the site. On larqe field
erected control devices, the time shown for this activity
indicates the fabrication and delivery of the first components
to the site. Delivery of the remaining components continues
throughout the construction phase. The duration of this
activity should be estimated after consultation with manufac-
turers of the appropriate air pollution control device. Es-
timate time requirement for this project is 24 weeks.
N-0
Procure construction bids
The bid package specifying the scope of work and specifications
of materials and including the drawings are mailed to selected
contractors. During this period, the contractors prepare their
bids for needed material and labor to install all ductwork,
piping, utilities, and control equipment. A minimum of four
weeks should be allocated for obtaining bids from the
contractors.
-------
Table faft(continued). DESCRIPTION OF INDIVIDUAL ACTIVITIES INVOLVED IN THE PROCUREMENT,
INSTALLATION, AND STARTUP OF CONTROL EQUIPMENT.35
ACTIVITY ACTIVITY
CODE DESCRIPTION
DETAILS OF ACTIVITY AND ESTIMATED TIME REQUIREMENT
0-P
P-3
Evaluate construction bids
Award construction contract
Ot
VO
CO
3-M
On-site construction
M-Q
Install control device
Construction bids are evaluated and the successful
bidder selected. Two weeks are estimated for this
activity..
Construction contract is prepared. In larae corporations,
it is reviewed and approved bv several departments prior
to its submission to the successful contractor. Two
weeks are allowed for this activity.
This consists of site clearance, pourinq of the foundation,
erecting structural members, ductwork, and installation of
auxiliary equipment. Twelve weeks were estimated for this
activity.
This activity is essentially an extension of the preceding
construction work. The time is primarily allocated for
installation of a shop assembled (or modular) control device.
In case of field erected unit, it represents the time which is
required to complete the installation of the remaining com-
ponents as they arrive on site. The installation time for
this case is estimated to be six weeks.
-------
Table.* & (continued). DESCRIPTION OF INDIVIDUAL ACTIVITIES INVOLVED IN THE PROCUREMENT,
INSTALLATION, AND STARTUP OF CONTROL EQUIPMENT.35
ACTIVITY
CODE
ACTIVITY
DESCRIPTION
Q-4
Complete construction
(system tie-*n)
C7I
I
IO
4-5
Start up, shakedown,
source test
DETAILS OF ACTIVITY AND ESTIMATED TIME REQUIREMENT
Tying the control device into the process requires that
the process be shut down. This shut down is usually
scheduled so that it will have the least imoact on the
operation. The contractors responsibility usually ends
at this point when the client and the vendors representative
accept the construction as beina complete. Two to six weeks
are allocated for tie-in. In large installations where the
process cannot be conveniently shut down at the end of
construction phase, longer times may be required.
The process is brought back on-line and any unforeseen
problems with the control system are resolved during this
time. Source testing may be performed to determine if
performance of the system is acceptable. Depending on the
type of control device installed, start UD, shake down, and
preliminary source testing would require from two weeks for
small and simple installation to about eiqht weeks for a large
and complicated system. '
-------
6.7 REFERENCES
1. Atmospheric Emissions from Wet-Process Phosphoric Acid
Manufacture. National Air Pollution Control Administration.
Raleigh, North Carolina. Publication Number AP-57. April 1970.
p. 25-26.
2. Reference 1, p. 31.
3. Technical Report: Phosphate Fertilizer Industry. In: Group III
Background Document. Environmental Protection Agency. Research
Triangle Park. p. -.
4. Reference 1, p. 30-32, 49, 51-52.
5. Air Pollution Control Technology and Costs in Seven Selected
Areas; Phase I. Industrial Gas Cleaning Institute. Stanford,
Connect!cuti EPA Contract 68-02-0289. March 1973. p. 52.
6. Reference 5, p. 41, 43.
7. Test No. 73-PSA-2; Texas Gulf, Inc.; Wet Process Phosphoric Acid;
Aurora, North Carolina; August 31-September 1, 1972. Environmental
Engineering, Inc. Gainesville, Florida. Contract No. 68-02-0232.
p. 4.
8. Technical Report: Phosphate Fertilizer Industry. In: An
Investigation of the Best Systems of Emission Reduction for Six
Phosphate Fertilizer Processes. Environmental Protection Agency.
Research Triangle Park, North Carolina. April 1974. p. 25.
9. Reference 3, p. -.
6-95
-------
10. Guthrie, K.M.; Capital Cost Estimating. In: Modern Cost-
Engineering Techniques, Popper, H. (ed). New York, McGraw-Hill
Book Co., 1970. p. 80-108.
11. Reference 5, 192 p.
12. Standards and Costs; Gas Absorption-and Pollution Control Equip-
ment. Ceil cote Company. Berea, Ohio. Bulletin 1200. 19 p.
13. Guthrie, K. Piping, Pumps, and Valves. In: Modern Cost-
Engineering Techniques, Popper, H. (ed). New York, McGraw-Hill
Book Co., 1970. p. 161-176.
14. Reference 5, p. 39.
15. Reference 5, p. 57.
16. Goodwin, D.R. Written communication from Mr. R.D. Smith,
Occidental Chemical Company. Houston, Texas. April 30, 1973.
17. Reference 5, p. 148.
18. Test No. 72-CI-25; Royster Company; Diammonium Phosphate;
Mulberry, Florida; May 17-18, 1972. Contract No. 68-02-0232.
p. 8.
19. Test No. 72-CI-18; Royster Company; Run-of-Pile Triple
Superphosphate; Mulberry, Florida; February 29-March 1, 1972.
Environmental Engineering, Inc. Gainesville, Florida. Contract
No. 68-02-0232. p. 4-5.
6-96
-------
20. Reference 3, p. -.
21. Reference 20, p. 4-95.
22. Reference 5, p. 114.
23. Control Techniques for Fluoride Emissions. Environmental Health
Service. Second Draft, September 1970. p. 4-95 (unpublished).
24. Reference 3, p. -.
25. Chatfield, H.E. and R.M. Ingels. Gas Absorption Equipment. In:
Air Pollution Engineering Manual, Daniel son, J.A. (ed). Research
Triangle Park, Ncrth Carolina. Environmental Protection Agency.
1973. p. 229.
26. Reference 5, p. 80.
27. Reference 1, p. 26.
28. Reference 3, p. -.
29. Reference 25, p. 228.
30. Emmert, R.E. and R.L. Pigford, Gas Absorption and Solvent
Extraction. In: Chemical Engineers' Handbook, Perry, R.H.,
C.H. Chilton, and S.D. Kirkpatrick (ed). New York. McGraw-
Hill, Inc. 1963. p. 14-37 to 14-39.
31. Reference 1, p. 27.
32. Reference 3, p. -.
6-97
-------
33. Reference 1, p. 27.
34. Technical Guide for Review and Evaluation of Compliance
Schedules for Air Pollution Sources. The Research Triangle
Institute and PEDCo - Environmental Specialists, Inc. Pre-
liminary Draft. June 1973. p. 3-39. (unpublished).
35. Reference 34, p. 2-4 to 2-8.
-------
7. ECONOMIC IMPACT
7.1 INTRODUCTION
This section describes the economic impact of adopting regulations
that require control of fluoride emissions from existing wet-process
phosphoric acid, superphosphoric acid, diammonium phosphate, run-of-pile
triple superphosphate, and granular triple superphosphate facilities.
The costs shown in Table 7-1 are based upon the installation and
operation of control equipment described in chapter 6.1.3. Installation
of other, less efficient control equipment is not expected to result
in any significant reduction in the economic impact incurred. The
capital costs and annualized costs of installing control equipment
represent expenditures needed to achieve the emission guidelines shown
in Table 1-1, but would also apply to the adoption of less stringent
fluoride emission regulations.
The economic impacts have been developed on a nrocess-by-process
basis since the national or industry-wide impact will be dependent
upon the collective actions of the states. To provide a perspective
on the significance of the costs incurred by adopting fluoride
emission regulations, they are related to unit production and product
sales price (Table 7-1). Additional insight on potential impacts
related to costs are given by a discussion on potential plant closures.
Criteria are presented that describe circumstances that could result
in plant closures, and the number of closures within the industry
that would result if all states adopted fluoride emission regulations
is estimated.
The information presented in this chapter is intended to assist
states in deciding on the advisability of adopting fluoride regulations.
7-1
-------
It is not expected that these emission guidelines would be
appropriate for all existing facilities.
7.2 IMPACT ON MODEL PLANTS
The total capital investment and annualized control cost ob-
tained from section 6.1.3.1 for each of the model facilities is
presented in Table 7-'l on a plant basis, on a unit product basis, and
as a percentage of the product sales price. For purposes of this
analysis, it is assumed that the wet-process acid plant sells all
acid production at prevailing merchant acid prices. The estimated
control costs for superphosphoric acid, diammonium phosphate, and
triple superphosphate plants reflect the retrofit requirements of
both the individual production facility and an associated wet-process
acid plant which produces the required intermediate phosphoric acid.
The captive acid plants are assumed to be sufficiently sized to
supply the needs of the various production units. For example, the
SPA plant is associated with a 300 ton P205/day acid plant while the
DAP plant requires a 500 ton/day unit. Control costs for the captive
units were obtained by prorating the cost developed for the model acid
plant.
A more detailed analysis of the potential financial effects of
control costs upon the phosphate industry could be obtained by cal-
culating the changes in profits and cash incomes for all plants or
firms in the industry if the necessary information were available.
Diammonium phosphate and granular triple superphosphate are the more
popular products sold and their processing will incur the higher
control costs on a unit basis. Industry statistics, representative
of 1973 performance, Indicate that after-tax profit margins ranged
7-2
-------
TABLE 7-1
SUMMARY OF RETROFIT CONTROL COST REQUIREMENTS FOR VARIOUS PHOSPHATE FERTILIZER MANUFACTURING PROCESSES
End Product
Design Rate, TPD
(P205 Basis)
Control Capital, $
Sales Price
($ per ton
product)
Annual i zed Costs
a. Total , $
b. Unit Basis
($ per ton
product)
c. As a % of
Sales Price
Phosphoric
Acid
500
208,000-
249,000
105
57,000-
69,000
0.19 -
0.23
0.2
Super-
Phosphoric
Acid
300
240,000
152
65,000
0.48
0.3
DAP
500
887,000
145
242 ,000
0.68
0.5
ROP-TSP-
(Case A)
550
875,000
126
262,000
0.66
0.5
ROP-TSP
(Case B)
550
1,465,000
126
404,000
1.03
1.0
GTSP
400
1,234,000
130
339,000
1.18
0.9
CO
Source of Price Quotations - Chemical Marketing Reporter, November 4, 1974.
-------
from 5 to 6 percent of sales and approximately doubled these per-
centages in 1974. Against this level of profitability, control costs
as shown in Table 7-1 appear to have minimal impact on a plant typical
of this profit performance. As long as product prices are unrestricted
(the Cost of Living Council removed price ceilings on domestic ferti-
lizers on October 25, 1973) and plant utilization remains at the cur-
rent level of approximately 90 percent, control costs could be ab-
sorbed by the industry without any price increases. On the other hand,
price increases to pay for the costs would be minimal.
An objective of this analysis is to highlight where the implemen-
tation of the emission guidelines might impose an economic
burden upon plants. A scenario for possible plant closures could be
presented in this fashion: overcapacity in spite of growing demand
develops in a particular segment of the industry resulting in under-
utilization at rates near 75 and 80 percent of capacity. Prices
and profits subsequently decline. In such a situation, plants
would probably close; however, the question is to what extent would
the impact of retrofit controls be responsible for plant closures.
In section 7.3, criteria are presented which can be used to pinpoint
the extent of plant closures.
7.3 CRITERIA FOR PLANT CLOSURES
Reasons for closing a facility are usually traced to the absence
of profitability for a specific site or facility. Managers of existing
plants faced with increased capital requirements for continuity of
operations will nave to decide whether the incremental investment will
"save" future cash income that otherwise would be lost by ceasing
operations. Plant managers will have the following options in such a
7-4
-------
1. Undergo Increased capital expenditures on the existing plant.
2. Shut down the plant and discontinue business.
3. Shut down the plant and replace it with a new plant.
The selection of an option is based on an interest or opportunity
cost for employing the required capital. There is usually a minimum
return that a plant manager will accept for employing funds—interest
cost for borrowing money or the interest cost of investing in short
term obligations. Since there is a risk with employment of capital,
businesses will require a higher rate of return for investing of
funds. A familiar tool for analyzing investments involves the deter-
mination of the >um of all future cash flow (income) streams over a
projected time span discounted (with the appropriate interest rate) to
the present. If the sum of these discounted residuals exceeds intended
cash outlay for investment, resulting in a positive term for net
present value, than the investment will be a good choice. Conversely,
if the discounted present value of projected cash flow streams results
in a negative value, then the proposed investment will be rejected.
The managerial tool of discounted cash flow analysis can be
applied to the retrofitting of control equipment to existing plants
in this manner. If the existing operations can only be continued in
the future by meeting a standard, then the investing of the control
capital has to be evaluated on the basis of the value of the future
income derived from continuing the operation of the present plant.
The merit of continuing operations after retrofitting a plant must be
evaluated in retrospect with the alternatives of discontinuing operations
and building a new plant.
7-5
-------
Guidelines for pinpointing plants as candidates for closure are
presented as follows. First, new plants to replace existing plants
of the comparable model size described in Table 7-1 v/ould require some
$10 to $20 million. In no instance could the construction of a new
plant be a better alternative than retrofitting controls requiring
the magnitude of capital, or even twice the values, shown in Table 7-1.
On the other hand, plants that have small or negative cash incomes
prior to retrofitting would certainly close. Plants that have small or
negative profits (after deducting depreciation charges) would eventually
become candidates for closure upon termination of their depreciation
schedules and subsequent increased tax liability.
The type of plants that would most likely face these circum-
stances are the following:
1. Small plants which generally suffer from the usual economies
of scale of production—less than 170,000 tons-per-year cap-
acity.1
2. Old plants which generally have outlived their useful or
economic lives—twenty years or more.
3. Plants isolated from raw materials—particularly diammonium
phosphate plants that purchase merchant phosphoric acid and
ammoni a.
4. Plants likely to suffer from a shift in the overall market
structure as a result of external forces.
Financial data on an individual plant basis necessary to evaluate
the impact of retrofit controls are unfortunately unavailable. Hence,
plant closures can De estimated only from a categorical approach, which
7-6
-------
classifies plants that possess characteristics of the nature of those
discussed above. Any estimate of plant closures has to be presented
with the usual qualifications.
7.4 IMPACT ON THE INDUSTRY
At the present time, the condition of the fertilizer industry is
healthy. Prices and profits in 1974 were the highest they have been
in years. The U.S. industry has become a leader in phosphate processing
technology and benefits from world trade in both rock and concentrated
phosphates. This position became more pronounced recently, in spite
of the fracture in the international monetary structure and con-
current high inflation. When the Cost of Living Council lifted
price ceilings on October 25, 1973, domestic prices heretofore con-
strained by CLC immediately arose 60 percent on the average reflecting
the foreign demand for domestic phosphate products. Demand for
fertilizers to increase agricultural production and yields has been
strong and will continue to be so, in spite of fluctuating international
currency values. Projected long-term demand for phosphate nutrients
2
is expected to grow a't an annual rate of 5-6 percent.
Historically, the fertilizer industry has experienced cyclic
patterns of overexpansion followed by plant shutdowns and product price
cutting. New phosphoric acid plant expansion scheduled to come on
stream in 1975-1976 may result in short term price declines until in-
creases in consumer demand restores equilibrium with capacitv. In
anticipation of overexpansion, producers will probably curtail con-
struction activity in the period beginning in 1976-1977. However,
during this slack period, retrofitting of existing plants for
7-7
-------
controls will be required in accordance with implementation plans.
Therefore, these retrofit projects should not hinder new construction.
Rather than resulting in plant closures, requirements for retro-
fitting fluoride emission control systsms will probably encourage some
improvements of marginal plants.
The nature of the impact of the lll(d) regulations for the
fertilizer industry will be geographical in scope. The state of
Florida, where most of the industry is located, has adopted regula-
tions for the existing industry that are equivalent m most instances
to the emission guidelines. Most of the remainina states with phos-
phate process facilities have no emission standards.
The greatest control cost - on a unit basis - for any process
subject to standards is for the combination of processing anc storage
of granular triple superphosphate. However, 75 percent of the industry
capability in GTSP production will be required to meet the
emission guideline by July 1975 regardless of Federal action.
a large portion of the production facilities will not require additional
retrofit controls, the impact upon the industry doesn't appear severe.
For run-of-pile triple superphosphate, the conclusion would be similar
to the GTSP as some 60 percent of the industry will be adequately con-
trolled because Of state standards.
The one segment of the industry where a wide-scale effort in
retrofitting would be required is for diammoniu"! phosphate plants.
Some 60 percent of industry cdunilty «ould be exnected to retrofit as a
result of Federal regulations. Control costs for this orocess,
7-8
-------
however, would amount to only 0.5 percent of sales. These costs alone
are not sufficient to close any plants.
Diammonium phosphate plants which incur water abatement costs as
great or greater than fluoride emission control costs would be likely
candidates for plant closures.3 There is no specific information
concerning plants which may fall into this category. The only
definitive statement that can be made is that those affected will
be outside the state of Florida and may amount to 3 to 5
plants, or approximately 10 percent of the total DAP manufacturing
capacity.
With regard to triple superphosphate plants, 1 to 3 plants (out-
side Florida) may close as a result of implementing the recommended emission
guidelines for control of aaseous flutHHp. This is likely to occur
in a geographical region where there is an oversupply of phosphate
processing capacity. An abundant supply of low-cost sulfuric acid
derived from non-ferrous smelters in the Rocky Mountains area could be
an incentive for construction of new phosphate facilities, ultimately
resulting in oversupply and price-cutting. Triple superphosphate capacity
does appear to be expanding rapidly in this area with a new 340,000
ton-per-year plant coming on-stream in 1975-1976.
Most of the control costs associated with a TSP complex are for
the solids manufacture and storage. Therefore, the closure of a TSP
facility as implied above does not mean that the entire complex
will be shut down. The plant manager has several options—(1) sell
merchant add, (2) convert to mixed fertilizers, or (3) produce
diammonium phosphate. However, if the same plant manager is faced
with Installing water abatement facilities, the overall abatement costs
will affect the entire facility.
7-9
-------
7.5 IMPACT ON EMPLOYMENT AND COMMUNITIES
The fertilizer industry is generally recognized as a capital
intensive industry; in other words, labor requirements for production
work and plant supervision are small, relative to plant sales.
Usually, those plants that may be affected by implementation of the
emission guidelines are widely dispersed throughout the
United States. Only in central Florida does the fertilizer industry
represent a substantial portion of overall community economic activity
and employment.
For purposes of illustrating the effects of plant closures on
employment, the shutdown of 1 to 3 triple superphosphate plants cited
in Section 7.4 might result in the loss of 10 to 50 jobs. Only those
jobs directly associated with the triple superphosphate plants would
be affected. Employment in supporting activities such as rock mining,
phosphoric acid production, and transportation services would remain
unaffected.
7.6 SUMMARY
An optimistic outlook for the phosphate fertilizer industry in
the next few years has been presented, but such an appraisal must be
cautionary after reviewing the historical chronic cyclic patterns
of product shortages and oversupply. Assuming that oversupply con-
ditions may occur in the next few years, some estimates of plant
closures have been made. In the triple superphosphate sector of
the industry, as many as three plants could close as a direct result
of the states adopting the emission guidelines. In the diammonium phosphate
a combination of e/penditures for retrofitting both fluoride emission
7-10
-------
controls and water effluenv: controls may result in as many as five
plant closures, or 10 percent of industry capacity.
However, fluoride emission controls alone would not cause these
closures. Associated costs for fluoride emission controls for wet-
process phosphoric acid plants that do not have attendant DAP or TSP
processes will not warrant plant closures. Similarly, costs for
superphosphoric acid plants do not present any apparent problems.
The number of predicted closures reflects the adoption of the
emission guidelines by all states; therefore, it reflects the maximum
number of closures that may occur.
7-11
-------
7.7 REFERENCES
1. David, Milton L., J.M. Malk, and C.C. Jones. Economic Analysis
of Proposed Effluent Guidelines for the Fertilizer Industry.
Development Planning and Research Associates, Inc. Washington,
D.C. Publication Number EPA 230-1-73-010. November 1973.
p. VI-12 to VI-15.
2. U.S. Industrial Outlook 1972 - With Projections to 1980. U.S.
Department of Commerce. Washington, D.C. Publication Number
BOC-704-08-72-005. p. 174-175.
3. Reference 4, p. V-13 to V-18.
4. Reference 4, p. VI-26.
7-12
-------
8. EMISSION GUIDELINES FOR EXISTING
PHOSPHATE FERTILIZER PLANTS
8.1 GENERAL RATIONALE
These emission guidelines represent the same degree of control
as is required by the standards of performance promulgated for new
plants [wet-process phosphoric acid, superphosphoric acid, diammonium
phosphate, run-of-pile triple superphosphate (production and storage),
and granular triple superphosphate (production and storage)]. The
emission guidelines were developed after consideration of the
following factors:
1. The degree of emission reduction achievable through the
application of the best adequately demonstrated svstem of
emission reduction (considering cost).
2. The technical and economic feasibility of applying the
best demonstrated technology to existing sources.
3. The impact of adopting the emission guidelines on annual
U. S. fluoride emissions.
4. The environmental, energy and economic costs of the
emission guidelines.
Identification of the best demonstrated control technology was
accomplished first. During the development of standards of
performance for new facilities in the phosphate fertilizer industry,
the spray-crossflow packed bed scrubber was found to represent the
best demonstrated control for total fluoride emissions. Historically,
the spray-crossflow packed bed scrubber was developed to control
fluoride emissions from the phosphate fertilizer industry. From this
8-1
-------
viewpoint, it is not unusual that this scrubber design is the best
demonstrated control technology. Many of the spray-crossflow packed
bed scrubbers tested by EPA were retrofitted. For this reason,
spray-crossflow packed bed scrubbers are recognized as the best
demonstrated control technology for both new and existing plants.
Alternative fluoride control technologies, such as the venturi
and cyclonic spray tower scrubbers, can only provide approximately
two transfer units for fluoride absorption unless two or more are used
in series, at multiplied costs. Spray-crossflow packed bed scrubbers
are not limited by the number of transfer units which they can provide;
in practice, five to nine transfer units per scrubber are provided. Con-
trol of gas streams with high particulate loadings has caused a plugging
problem for spray-crossflow packed bed scrubbers in the past. However,
use of a built-in venturi scrubber and other improvements in spray-
crossflow packed bed scrubber design have eliminated this problem. In
addition, all current fluoride control technologies involve some type of
scrubbing system, and consequently, they share any plugging tendencies,
as well as similar costs and energy requirements. With these considera-
tions in mind, it is not unreasonable to base fluoride emission guide-
lines on the one clearly superior scrubbing technology.
Evaluation of the problems and costs associated with a retrofit
project is complicated by the lack of actual data. Some of the
facilities equipped with spray-crossflow packed bed scrubbers installed
8-2
-------
the units as part of the original plant design. Retrofit information
that is-available is usually incomplete because of changes in plant
management and lack of cost breakdowns. Retrofit models were therefore
developed to evaluate the technical and economic feasibility of in-
stalling spray-crossflow packed bed scrubbers on existing WPPA, SPA,
DAP, ROP-TSP, GTSP processing, and GTSP storage facilities. The retro-
fit model approach was meant to estimate costs for"an" average"plant and
to clarify the technical problems involved in a typical retrofit pro-
ject. No technical problems, other than space limitations, were
foreseen for the average plant. • In all cases, the mannitude of the
estimated retrofit costs are minimal as is discussed in Section 7.
Table 9-1 indicates the impact of the emission ouidelinei.
on annual U.S. fluoride emissions. Adoption of the emission guidelines
would result in emission reductions ranging from 50 percent for fiTSP
storage facilities to 90 percent for ROP-TSP plants. Overall emissions
from the affected facilities would be renuced by 75 percent.
Environmental and energy costs associated with the
emission guidelines are minimal. With current spray-crossflow packed
bed scrubber designs, gypsum pond water can be used as the scrubbing
medium to meet the emission guidelines in practically all cases.
In the rare case where the partial pressure of fluoride out of pond water
is high, the emission guidelines can still be met. The aliquot of water
sent to the final section of scrubber packing nay be fresh or limed water.
*
This aliquot will only be a small fraction of the total water to the scrubber
8-3
-------
and will contain only a small fraction of the total fluoride absorbed
in the scrubber. This implies that no additional effluent need be
created. Any solids generated by fluoride scrubbing (e.g., in the WPPA
process) would go to the gypsum pond and cause no more than a 0.06
percent increase in the amount of solids normally produced.
The estimated total annual incremental electrical energy demand
which would be created by fluoride control to meet the
emission guidelines is only 26.9 X 106 KWH/yr. This is equivalent
to the amount of energy required to operate only one 300 tons/day
P205 SPA plant by the submerged combustion process 115 days/yr.
8.2 EVALUATION OF INDIVIDUAL EMISSION GUIDELINES
8.2.1 Wet-Process Phosphoric Acid Plants
Fluoride Emission Guideline
0.01 grams of fluoride (as F~) per kilogram of P,,0fi input to the
process.
Discussion
The emission guideline is equal to the promulaated
SPNSS. It is estimated that each will require removal of 99 percent
of the fluorides evolved from the wet-acid process. A sprav-crossflow
packed bed scrubber is capable of providing this collection efficiency.
8-4
-------
Rationale
1. The economic impact of the emission Guideline on the
industry should be negligible. Approximately 53 percent of the
existing wet process acid plants, ^ccountin^ for 7f- percent of the
production capacity, *ro either sufficiently controlled at present
to meet an emission level of 0.01 grans F/kilogram PgOs or will be
required to attain that level of control regardless of the proposed
emission guideline. This estimate is based on the assumption that
all wet-process acid plants built since 1967 have installed controls
capable of meeting an emission level of 0.01 grams of fluoride
par Lilograrc P^Oj, input as pare of t!i«2 original plant Jasign.
I'he retrofit costs for those plants that are affected, approximately
$230,000 for a 500 ton P205/day facility, can be successfully absorbed
within the existing cost structure. Annualized control costs for an
average sized plant, including capital charges, amount to approximately
0.2 percent of sales.
2. Relaxation of the guideline to allow emission increases of 50 to
100 percent would not alto* .additional control options or appreciably
reduce retrofit costs for the following reasons:
a. Only a packed bed scrubber is capable of providing the re-
quired fluoride removal efficiency = 99 percent. A tenfold
increase in the emission guideline would be required
8-5
-------
to allow the use of other commonly used scrubber designs -
Venturis, cyclonic spray towers, etc. with Gr)-rn percent
collection efficiency.
b. Packed bed scrubber cost will not vary significantly with
moderate changes in packing depth. The cost of additional
packing to increase scrubber efficiency is minor compared
to overall control costs.
3. Estimated impact of the emission guideline on annufl
fluoride emissions is significant - 73 percent reaction.
8.2.2 Superphosphoric Acid Plants
Fluoride Emission Guideline
0.005 grams of fluoride per kilogram of P20g input to the
nrncfiss.
Discussion
The emission guideline for existing SPA olants is equal
to the promulgated SPNSS. Compliance with this emission guideline
would require removal of. nnnrnyiaafrrly 90 percaot of thee-ftuarrdes
now being emitted from SPA plants using the submerged combustion process.
A spray-crossflow packed bed scrubber should be capable of providing
this performance. Three designers of control equipment have submitted
proposals to one operator for control to the level of the emission
guideline; venturi and other designs using the vacuum evaporation
process (79 percent of the SPA industry) will reouire no aJditional
control.
8-6
-------
Rationale
1. Impact on the industry snoulu be negligible. The two existino olants
using the submerged combustion process could be required to adc1 retrofit
controls.
2. Existing submerged combustion plants should be capable of meeting
tlie emission guideline by treating the exhaust stream from
controls with a spray-crossflow packed bed scrubber. This scrubber can
be acded to any existing mist separators, baffles, and spray chambers,
as was assumed in the SPA retrofit model, Figure 6-5.
3. Retrofit costs are expected to be acceptable - $103,000 for a 300
ton per day plant. Annualized control costs, including capital charges,
amount to only 0.3 percent of sales.
4. Relaxing the emission guideline to allow a three-fold increase in
emissions (0.015 grams F/kilograms PpOc) would be required to
accommodate the use of Venturis anH cyclonic snra*/ t.owprs, if t.hp rph.ro-
fit costs are to remain about the same.
8.2.3 Diammonium Phosphate Plants
Fluoride Emission Guideline
0.03 grams of fluoride (as F~) per kilogram of P205 input to the process,
Discussion
The emission Guideline for existing DAP plants is equal to the
promulgated SPNSS. Compliance would require removal of approximately
85 percent of the fluorides evolved from the DAP process. Snray-crossflow
packed bed scrubbers, added to any existing Venturis, are capable of
8-7
-------
providing the required collection efficiency. As pointed out in
section 8.1, new designs for these scrubbers are available and are
expected to overcome problems formerly associated with pluqqing by
excessive particulates (2).
Rationale
1. Relaxing the emission guideline to allow the use of alternative
scrubber technologies would increase fluoride emissions to the
atmosphere by 49 tons per year, a 50 percent increase.
2. Retrofit costs - $660,000 for a 500 ton P205/day plant - are not
considered excessive. Annualized cost, including capital charges,
would amount to 0.37 percent of sales.
3. Impact of applying the emission guideline on fluoride emissions
from U. S. DAP plants is significant - a 75 percent reduction (160
tons/yr).
8.2.4 Run-of-Pile Triple Superphosphate Production and Storage Facilities
Fluoride Emission Guideline
0.1 gram of fluoride (as F") per kilogram of PgOg input to the
process.
Discussion
The emission guideline is equal to the promulgated SPNSS. Only
40 percent of the industry is directly affected by the emission
guideline.
8-8
-------
Compliance with an 0.1 aram F/kilonran P?Cr> mission level would
recuire collection of about 99.2 percent of the fluorides evolved
from the process. This efficiency can be obtained by a two stage
system using Venturis and a soray-crossflow packed bee' scrubber.
Rationale
1. Economic impact on the industry should be moderate. Only 40 oercent
of the industry is directly affected by the emission aiiidPlinP.
The remaining 60 percent will be required to neet more stringent State
1 regulations.
2. No additional control options would be made availabl" hy relaxinr
the emission guideline by 50 to 100 percent. It would be necessary to
triple the emission guideline to allow the use of a venturi or cyclonic
spray tower as the secondary scrubber.
3. Retrofit costs - $725,000 for a typical 550 ton P^/day plant to
$1,240,000 for the extreme case - are not considered excessive. Annual-
ized control costs, including caoital charges, amount to 0.40 to 0.70
A
percent of sales. Although these costs are more severe than retrofit
costs for most other sources, they are expected to be manageable.
4. The emission guideline would reduce annual fluoride
emissions from existing ROP-TSP plants by 90 percent.
8.2.5 Granular Triple Superphosphate Production Facilities
Fluoride Emission Guideline
0.1 qram fluoride (as F") per kilooram of P205 input to the process.
8-9
-------
Discussion
The fluori:1e emission cjui-Jelino is equal to the nrori:ln?tpd
L^P.'SS. Conpliance with the emission ouic-'eline voulc- r^nuire
collection of about 99.6 percent of the- fluoride evolve:!
from the GTSP production process. This efficiency can be obtainecj by
a two-stage system consisting of a venturi and a spray-crossflow
packed bed scrubber.
Rationale
1. Economic impact of the emission guideline should he moderate. Only
25 percent of tr.e industry is directly affected by the
emission guideline. The remaining 75 percent will be required to meet
more stringent State regulations.
2. Relaxing the emission guideline by 50 percent would provide greater
flexibility with regard to the development of a control strategy,
however, it would also allow the emission of an additional 66 tons of
fluoride per year. A five-fold increase in the omission nuideline would
be necessary to allow the use of a venturi or a cyclonic spray tower
as the secondary scrubber in all effluent streams.
3. The estimated retrofit costs - $600,000 for a 400 ton P205/day
plant - are not considered excessive. Annualized control costs amount
to 0.44 percent of sales.
4. Tne emission guideline would reduce annual fluoride
emissions from GTSP production facilities by 51 percent.
8-10
-------
8.2.6 Granular Triple Superphosphate Storage Facilities
Fluoride Emission Guideline
2.5 X 10"4 gram fluoride (as F") per hour per kilogram of P205 in
storage.
Discussion
The fluoride emission guideline for existinq granular triple
superphosphate storage facilities is equal to the SPNSS. In order
to meet this emission level, a typical facility would be required to
remove approximately 90 percent of the fluorides evolved. Only 25
to 35 percent of the industry currently has this degree of control.
Twenty-five percent of the existing facilities are presently uncon-
trolled.
Rationale
1. It is estimated that 50 percent of the industry would still be
required to add retrofit scrubbers even if the allowable emissions
were increased by 50 percent.
2. The cost of retrofitting uncontrolled facilities would not vary
significantly with moderate (50 percent) relaxation of the emission
guideline. The major portion of the costs is associated with
refurbishing the building and is exclusive of the control device
itself.
8-11
-------
3. Retrofit costs for uncontrolled facilities - $540,000 for a 25,000
ton storage building - are not considered to be excessive. Such a
facility would accompany a 400 ton PgOg/day GTSP production facility.
Annualized control costs, including capital charges, would equal 0.4
percent of sales.
4- The emission guideline wouUi reduce annual fluoride
emissions from GTSP storage facilities by 50 percent.
8-12
-------
8.3 REFERENCES
1. Atwood, W. W., Occidental Chemical Company to Goodwin, D. dated
June 27, 1973. Fluorine Emissions from Submerged Combustion
Evaporation of Phosphoric Acid.
2. Crane, George B. Private communication with Teller Environmental
Systems, Inc. New York, N.Y. December 13, 1974.
8-13
-------
9. ENVIRONMENTAL ASSESSMENT
9.1 ENVIRONMENTAL ASSESSMENT OF THE EMISSION GUIDELINES
9.1.1 Air
Installation of retrofit controls similar to those described
in section 6.1.3.1 could reduce fluoride emissions from existing sources
by the amounts indicated in Table 9-1. Emission reductions range
from 50 percent for granular triple superphosphate storage facilities
to 90 percent for run-of-pile triple superphosphate plants. All estimates
are based on information presented in chapters 3, 5, and 6 of this study.
The following procedure was used to arrive at the estimates listed
in Tables 9-1 and 9-2. The percentage of existing facilities Cor capacity)
attaining emission levels equivalent to SPMSS was estimated in Chapter 5.
The remainder of the existing facilities were assumed to emit at a rate
midway between the SPNSS level and a level characteristic of a poorly
controlled plant. The retrofit models were used as a source of
information regarding poorly controlled plants.
Total emissions following the installation of retrofit controls
were estimated by applying the SPNSS level to the entire industry
which is identical to the lll(d) emission guidelines contained herein.
All estimates assume a 90 percent utilization of production capacity.
This general approach was altered in certain instances (SPA, DAP,
GTSP storage) either to make use of additional information or to com-
pensate for the lack of necessary data.
9-1
-------
Table 9-1, ANNUAL U.S. FLUORIDE EMISSION REDUCTION DUE TO INSTALLATION
OF RETROFIT COMTW&S OSRABlT OF MEETING EMISSION GUIDELINES
Segment of Industry
WPPA
SPA
DAP
ROP-TSP
GTSP
Production
Storage
Overall
Estimated 1974
Emissions (Tons F/Yr)
217
12.6
385
662
268
HO
1,685
Estimated Emissions Following
Installation of Retrofit Con-
trols (Tons F/Yr)
58
2.9
97
71
131
70
430
Fluoride Emission
Reduction
(% 1974 level)
73
77
75
90
51
5JL
74.5
-------
Table 9*8. TYPICAL 1974 FLUORIDE EMISSIONS SOURCE STRENGTHS BEFORE AND AFTER INSTALLATION OF
RETROFIT EONTWESTCAPABLE OF MEETING'EMISSION GUIDELINES
Type of Plant
WPPA
SPA
(Submerged combustion
f* process)
UJ
DAP
ROP-TSP
GTSP
Production
Storage
Capaci ty
(Tons/Day P20g)
500
300
500
550
400
25,000*
Emissions Before Retrofit
(Lb F/hr)
48.4
3.9
6.86
571
122.6
13.2
Emissions After Retrofit
(Lb F/hr)
.42
.12
1.25
4.6
3.34
1.25
*Tohs GTSP Stored
-------
As indicated in Table 9-1, an overall fluoride emission reduction of nearly
75 percent can be achieved by installation of retrofit controls capable of
meeting the emission guidelines. The corresponding reduction in
typical fluoride emission source strengths is illustrated by Table 9-2.
9.1.1.1 Atmospheric Dispersion of Fluoride Emissions
A dispersion analysis was made to compare qround-level fluoride
concentrations downwind of a phosphate fertilizer complex, before and
after retrofit of controls. The diffusion estimates were based on 30-
day average fluoride concentrations and extended to distances from the
plant where fluoride concentrations were less than 0.5 yg/m3. A 30-
day average ground-level fluoride concentration of 0.5 yg/m causes an
accumulation of more than 40 ppm fluoride in cattle forage, and this
concentration in their feed is a damage threshold for cattle.
The fertilizer complex being investigated represents no actual plant,
but contains all of the units discussed in Section 6.1.3.1 - Retrofit
Models - except the submerged combustion-superphosphoric acid plant.
Emissions from this complex are not necessarily typical of the emissions
used in the retrofit models of section 6, nor are they the same as the un-
controlled source strengths listed in Table 9-2. However, these emis-
sions fall within the range of emissions from actual plants. Specific
9-4
-------
Table 9-3. EXISTING CONTROLS AND EMISSIONS
FOR MODEL PHOSPHATE FERTILIZER COMPLEX
1C
1
tn
Product : Items
•Controlled
Gas Flow,
SCFM
KPPA Idigester, filter, ' 21,500
iflash cooler seal
:tank, evaporator !
•hotwell
DAP jreactor-granulator,
idrier, cooler-
, screen
TSP Icone mixer, den,
'storage bldg
GTSP |
'
1
i
j reactor-granu-
' lator, drier,
cooler-screen
1
110,000 i
1
182,000
75,000 from
uncontrolled
storage bide
112,000
i
Fluoride,
IL-s/hr
10.8
3.3
36.5
13.2
22.6
height,
i ft.
60
i
! 85
1
i
t
i 60
i
j bldg.
i louvers
' (a 45 ft
1 85
Stack
teir.o. , aas velocity
°F ft/sec
100 30
i
100 1 30
i
i
i
90 ! 30
1
85 i -
I
i
!
90 ' 30
i '
!
i
-------
Table 9-4. RETROFIT CONTROLS AND EMISSIONS
FOR MODEL PHOSPHATE FERTILIZER COMPLEX
Product
WPPA
DAP
TSP
GTSP
Items 1
Controlled
Table 9-3",
plus filtrate
sump and seal
tank, plus
acid storage
same as Table
9-3
same as Table
9-3
storage build-
ing
same as
Table 9-3
Gas Flow,
SCFM
25,000
96,000
182,000
76,000
111,000
Fluoride,
Ibs/hr
0.42
1.25
6.20
2.00
3.34
height,
ft
85
85
70
i — i
70
85
i
Stack
temperature,
or
100
100
r '
90
85
90
qas velocity
ft/sec
40
30
30
40
! 40
!
i
CTi
-------
fertilizer manufacturing units are pictured in Figures 6.3, 6.4, and
elsewhere. All of these units were assembled to scale on a olot plan
of the entire complex. From this plot plan the meteorologist could
measure the distance relationships of sources and of interferences such
as buildings and phosphate rock piles. The heights of these inter-
ferences were also tabulated. Additional information used is shown in
Tables 9-3 and 9-4. The former table indicates emissions from the
fertilizer complex having existing mediocre emission controls. The
latter table shows the emissions from the same sources after installation
of good controls.
The source data indicated that aerodynamic downwash was a problem
at the facility modeled, particularly for wind speeds in excess of 3 or
4 meters per seconds. At lower wind speeds, plume rise from some of the
stacks could be significant. Plume rise factors were consequently
developed, which accounted for the plume rise at low wind speeds and
downwash at higher speeds. Those factors were then incorporated into the
dispersion estimates.
The dispersion estimates were made through application of the
Climatological Dispersion Model (COM). The COM provides estimates of
long-term pollutant concentrations at selected ground-level receptors.
The model uses average emission rates from point and area sources and a
joint frequency distribution of wind direction, wind speed, and stability.
9-7
-------
One year of monthly stability-wind data from Orlando, Florida were
utilized in the COM dispersion estimates. The climatology of that lo-
cation is representative of that at facilities of concern in this docu-
ment. The COM estimates are typical high 30-day average ambient fluo-
ride concentrations. The results of the analysis are presented in
Table 9-5. A more general review of 5-year summaries of monthly stability-
wind data from the same location verified that the values presented in
Table 9-5 are representative of typical high 30-day average concentrations
for any given year.
Table 9-5 shows tjiat the best technology retrofit controls made a large
reduction in the qrounfl-level fluoride concentrations which has existed when
the mediocre controls were used on the four sources shown. At distances
greater than about 1-1/2 mile, the concentrations do not exceed 0.5 vg/m3,
even in the most unfavorable months when the emission guidelines herein are
applied.
Table 9-5. ESTIMATED 30-DAY AVERAGE AMBIENT FLUORIDE CONCENTRATIONS
DOWNWIND OF A PHOSPHATE FERTILIZER COMPLEX
Fluoride Sources
Existing Controls
WPPA DAP TSP GTSP
After Retrofit
WPPA DAP TSP GTSP
Estimate
Fluoride
1
6
0.8
2
4
0.6
sd 30-Day Average
5 Concentration
3
3
0.4
5
1.9
0.3
10
1.0
0.1
ug/m )
15 km
0.5
0.1
9-8
-------
9.1.1.2 Emission Guidelines vs. a Tvoical Standard
The Florida standard to take effect on July 1, 1975, has been
chosen as a typical standard to compare with the proposed emission
guidelines. Emission reductions listed in Table 9-1 have already
taken into account the effect of the Florida standard in reducing fluo-
ride emissions. Table 9-6 gives the incremental annual controlled fluo-
ride emissions when the emission guideline is substituted for
a typical standard. Emission figures in this table are based on the
data in Table 9-1. In all cases, the typical standard is as strict or
more so than the -antssion guidelines.
9.1.2 Water Pollution
Increased or decreased control of gaseous water-soluble fluorides
will not change the amount of liquid waste generated by the phosphate
industry. Most control systems now in use utilize recycled process
(gypsum pond) water as the scrubbing medium thereby eliminating the
creation of additional effluent. Phosphate fertilizer plants do not need
to discharge gypsum pond water continuously. The pond water is re-used in
the process, and a discharge is needed only when there is rainfall in excess
9-9
-------
Table 9-6. COMPARISON OF STATE GUIDELINES STANDARD AND AN ALTERNATIVE STANDARD
I
o
Process Source of
Fluorides
Wet Process
Phosphoric Acid
Superphosphoric
Acid
Di ammonium
Phosphate
Triple Super-
phosphate (ROP)
Granular Triple
Superphosphate
Granular Triple
Superphosphate
Storage
Percent of
Plants Probably
Affected by State
Guidelines Standard
47
21
60
40
25
70
Florida Standard
for
duly 1, 1975
Ibs/ton P205
0.02
Best Avail-
able Technology
0.06
Belt & Den 0.05
Storage 0.12
0.15
0.05**
Emission
Guidelines
input
0.02
0.01
0.06
0.2
0.2
5 x 10"4*
Increase in Esti-
mated Controlled
Annual Fluoride
Emissions if
State Guidelines
Standard is Sub-
stituted for
Florida Standard
(Tons F/Yr)
0
0
0
39
33
23
*Units are Ibs F/hr/ton of P205 stored.
**Units are Ibs F/hr/ton of P20g input to bldg.
-------
of evaporation. For this reason, the volume of effluent from phosphate
fertilizer plants is almost exclusively a function of rainfall conditions,
EPA effluent limitations guidelines require that any gypsum pond water
discharged to navigable waters when rainfall exceeds evaporation meet
the limitations in Table 9-7. A two-stage lime neutralization procedure
combined with settling is sufficient control to meet these limitations.
Table 9-7. EPA EFFLUENT LIMITATIONS GUIDELINES FOR GYPSUM POND WATER1
Aqueous Maximum Daily Maximum Average of Daily
Waste Concentration Values for Periods of
Constituent (mg/1) Discharge Covering 10 or
More Consecutive Days
(ng/1)
Phosphorus as (P)
Fluoride as (F)
Total Suspended
nonfilterable solids
70
30
50
35
15
25
The pH of the water discharged shall be within the range of 8.0 to 9.5
at all times.
The phosphate industry has voiced concern that the partial pressure
of fluoride out of pond water makes it infeasible in some cases to reach
SPNSS fluoride limitations with a scrubber using pond water. An equili-
brium fluoride concentration between 5000-6000 ppm seems to be estab-
lished in gypsum ponds - possibly because of a slow reaction between
234
gypsum and soluble fluosilicates. ' ' Even a pond with an apparent fluo-
ride concentration of 12,500 ppm has fallen within this equilibrium range
when the water was passed through a millipore filter., The excess fluoride
can be attributed to suspended solids. Pond water containing about 6000
9-11
-------
ppm of fluoride has a low enough partial pressure of fluoride to
allow scrubber vendors to design to meet emission guidelines. In all
cases, emission guidelines can be achieved with pond water
1f a well-designed spray-crossflow packed bed scrubber 1s used as the
5
control device.
9.1.3 Solid Waste Disposal
Any solid waste generated by scrubbing fluorides would be in the
form of CaF2 or similar precipitates in the gypsum ponds. The amount
of precipitate formed is negligible in comparison to the amount of
gypsum generated in producing wet process phosphoric acid, a required
intermediate throughout the phosphate fertilizer industry. An example
of the relative amounts of each of the solids produced in normal processing
with scrubbers which meet emission guidellres for a 500
tons/day P20g WPPA plant.is presented below:
Assumptions:
1. 6427# phosphate rock = 1 ton PoOe-
2. Phosphate rock is 35 weight percent Ca.
3. Uncontrolled emissions of 58.1 #F/hr are reduced to 0.42 #F/hr
by a scrubber. (See retrofit model WPPA plant, case B).
4. All of the F absorbed by the scrubber precipitates in the
gypsum pond as CaF2. (See Section 5.2.1, page 5-6).
5. The plant capacity is 500 tons/day P°-
9-12
-------
3Ca1Q (P04)g F2 + 30H2S04 + Si02 + 58H20 •* 30 CaS04
18H3P04
This reaction implies: 40#Ca -*- 172# gypsum.
500 x 6427 x 0.35 x 172
gypsum produced = = 201,510
# gypsurn/hr
24 x 40
From assumptions 3 and 4:
F absorbed in scrubber =58.1 - 0.42 # F/hr
= 57.68 # F/hr
Ca++ + 2F" •+ CaF2 4- (5-1)
CaF? + = 57.68 x 78 = 118.4 # CaF9/hr
^ 38 z
% increase in solids = 118.4 x 100 = 0.06
201,510
This example illustrates that the increase In solids due only to
scrubbing fluorides is negligible (0.06%). The disposal of the
Targe volume of gypsum is by depositing in mined-out areas, and by
lagooning, followed by drying and piling techniques. Such piles are
as much as 100 feet above grade in some areas.
9.1.4 Energy
Changes in fluoride control, electrical power requirements for the
spray-crossflow packed bed scrubber retrofit models in Section 6 are
presented in Table 9-8. Existing fluoride control power requirements
were estimated from the pump and fan requirements for the assumed existing
9-13
-------
Table 9-8. INCREMENTAL POWER REQUIREMENTS FOR FLUORIDE CONTROL DUE TO INSTALLATION OF RETROFIT
CONTROLS TO MEET EMISSION GUIDELINES.
Power Requirements
ran^,-4- Power Requirements for Retrofit Controls AD
Capacity for Existing Controls to Meet State Guide- APower A Enerqv
Type of Plant Ton/Day P205 (Hp) }?„„ SK6 bU1Qe HP KWH/Ton PJ
1
4^>
WPPA
SPA
DAP
ROP-TSP
(Case A)
GTSP
500
300
500
550
400
90
75
565
300
540
140
82.5
800
500
1100
50
7.5
235
200
560
1.8
0.4
8.4
6.5
25
-------
controls in the retrofit models. Power requirements for the retrofit
controls were obtained by adding the power ratings of the specified
retrofit fans and pumps to the existing power requirements and sub-
tracting the power for any fans or pumps removed in retrofitting.
The largest incremental power requirement for fluoride control
is for GTSP. This can be attributed to installing a spray-crossflow
packed bed scrubber for GTSP storage, a previously uncontrolled source
in the retrofit model which generates a very large volume of air having
a small concentration of fluoride. Raising the standard to allow larger
emissions from GTSP storage would not greatly reduce these power require-
ments. It would only allow the use of a scrubber with a fewer number of
transfer units. A less efficient scrubber would not reduce the volume
of gas to be scrubbed nor would it greatly reduce the amount of pond
water required for scrubbing. Only the pressure drop through the scrubber
would be reduced by raising the standard. In other words, raising the
GTSP storage standard by a factor of two would not reduce the power require-
ments proportionately.
Incremental increases in phosphate fertilizer processing energy
requirements are given in Table 9-9; such increases will vary from
plant to plant. Volumetric flow rates of fluoride-contaminated air
sent to the scrubbers can vary by a factor of two or three for the same
size and type of plant. Existing control schemes will also influence
incremental power requirements by the extent to which their pumping
and fan systems can be adapted. Therefore, the numbers presented in
Tables 9-8 and 9-9 should be considered approximate.
9-15
-------
Fertilizer processing energy requirements presented in Table 9-Q
are primarily based upon material in reference (6). The tynes of
energy utilized by the various processes vary. For example, approximately
50 percent of the energy required in GTSP processing can be attributed to
the 3 gallons of fuel oil used per ton PpOc processed while neerlv all
the energy used in the submorgsc' combustion process for SPA comes from
natural gas. All processing energy requirements listed in Table 9-9
include electrical power required for rock crushing and pumping.
Table 9-9. INCREASE IN PHOSPHATE INDUSTRY ENERGY REQUIREMENTS RESULTING
FROM INSTALLATION OF RETROFIT CONTROLS TO MEET EMISSION GUIDELINES
Fertilizer process
WPPA
DAP*
SPA*
ROP-TSP*
GTSP*
Existing energy
requirements
(KWH/Ton P205)
225
236
782
152
305
Fluoride control
incremental
energy require-
ments
(KWH/Ton P20g)
1.8
8.4
0.4
6.5
25
Percent
increase in
energy re-
auirements
0.8
3.6
0.05
4.3
8.2
*Existing energy requirements figures include energy needed to process WPPA
feed for process.
Annual incremental electrical energy demand for fluoride control is
presented in Table 9-10. These figures are based upon Tables 9-6 anc,
9-8 along with production statistics in section 3. The total incremental
9-16
-------
Table 9-10. INCREASED ELECTRICAL ENERGY DEMAND BY THE PHOSPHATE INDUSTRY AS A RESULT OF INSTALLATION
OF RETROFIT CONTROLS
Fertilizer Process
VO
1973 Production
(Thousand Tons P
% of Production Capacity
Affected by State
Guidelines Standard
Incremental Electrical
Energy Demand (Million
KWH/yr)
**
WPPA
DAP
SPA
ROP-TSP
GTSP
5,621
1,860
783
600
1,115
26
60
21
40
47*
2.6
9.6
0.06
1.6
13
*This is a fictitious average based upon a weighted average of GTSP production and storage
statistics (see Table 9-6).
**Total Incremental Electrical Energy Demand = 26.86 x 106 KWH/yr.
-------
electrical energy demand resulting from installation of retrofit con-
trol? to moot om-i^sfAn «iit/l<*lt^s ts ^utYa1 ent t.n the energy required to
operate one 300 ton/day P205 SPA plant 115 days/yr. It should be em-
phasized that these numbers can be only approximations. As mentioned
in the discussion of Tables 9-8 and 9-9, individual plant fluoride control
energy and power requirements will vary. This variability necessarily
constrains the accuracy of projections based upon single retrofit models.
9.1.5 Other Environmental Concerns
Due to the proposed method of fluoride control, namely, utilization
of a spray-crossflow packed bed scrubber with pond water as the scrubbing
medium, no other environmental concerns are anticipated. Scrubbing
fluorides with gypsum pond water produces a closed system effect for
phosphate fertilizer complexes. Although radioactive materials have been
detected in the wastewater at fertilizer complexes, recycling of the pond
water to the scrubber is not expected to contribute to this potential problem.7
9.2 ENVIRONMENTAL IMPACT UNDER ALTERNATIVE EMISSION CONTROL SYSTEMS
Analysis of the data I ase on which the emission guidelines are based
indicates that only the spray-crossflow packed bed scrubber can meet
emission guidelines in all cases. ROP-TSP plants can use cyclonic
spray tower scrubbers to meet the emission guidelines, but at a higher
cost than for a spray-crossflow packed bed scrubber (Table 6-44).
Tables 6-37 and 6-40 show that the ROP-TSP standard is the only one
substantiated by data which allows use of an alternative scrubber design.
Use of either scrubber design for controlling ROP-TSP plants would result
in similar environmental impacts. Except for ROP-TSP plants, raising
the emission guidelines to allow use of alternative scrubber designs.
would result in a 50 percent to 1000 percent increase in fluoride
emissions without causing any beneficial environmental impacts.
9-18
-------
9.3 SOCIO-ECONOMIC IMPACTS
Tha phosphate fcrti "h?..v industry is generally recofimzcci as 5
capital intensive industry, It'bor requirements for nroduction work anc
nlant suuervision arc small, compared to plant sales. Usually, those
fertilizer facilities which may be affected by the emission
guidelines are widely dispersed throughout the United States. Only in
central Florida does the fertilizer industry represent a substantial
portion of overall community economic activity and employment, and
Florida has enacted emission standards effective July 1, 1975 which are
at least as strict as the em'ssion guidelines. Therefore, any potential
plant closures as a result of the implementation of lll(d) regulations
will produce minimal community effects in terms of job losses and sales
revenues.
Retrofitting existing plants for controls should not imoede new
plant construction programs. During the years 1973 through 1974, the
phosphate industry entered an expansionary o!ase with the ccnsiruction
of several new fertilizer manufacturing comolexes. The construction
rate i$* expected to decrease after 1976 as these new plants come on-
stream. Installation of retrofit controls will consequently occur during
a period of slack construction activity and should not interruot the
long-term availability of phosphate fertilizers.
9-19
-------
9.5 REFERENCES
1. Martin, Elwood E. Development Document for Effluent Limitations
Guidelines and New Source Performance Standards for the Basic
Fertilizer Chemicals. Environmental Protection Agency. Washington,
D.C. Publication Number EPA-440/1-74-011 -a. March 1974.
2. Teller, A.J. Control of Gaseous Fluoride Emissions. Chemical
Engineering Progress. 63^75-79, March 1967.
3. Huffstutler, K.K. Pollution Problems in Phosphoric Acid
Production. In: Phosphoric Acid, Vol. I., Slack, A.V. (ed).
New York, Marcel Dekker, Inc., 1968. p. 728.
4. Weber, W.C. and C.J. Pratt. Wet-Process Phosphoric Acid Manu-
facture. In: Chemistry and Technology of Fertilizers,
Sauchelli, V. (ed). New York, Reinhold Publishing Corporation,
196U. p. 224.
5. Crane, George B. Telephone Conversation with Dr. Aaron Teller,
Teller Environmental Systems, Inc. New York, N.Y. December 13,
1974.
6. Bixby, David W. , Delbert L. Rucker, and Samuel L. Tisdale.
Phosphatic Fertilizers. The Sulphur Institute. Washington, D.C.
February 1964.
7. Rouse, J. V. Letter. In: Environmental Science and Technology.
Easton, Pa. October 1974.
9-20
-------
TECHNICAL REPORT DATA
'//far rcail Injunctions on the rei i-nc before rtimplcl'ns)
»: r-on i t*o
. T -|_. A.\O S'Jtii It I i
Draft Guideline Document: Control of Fluoride
DMSSIOIIS from listing Phosphate Fertilizer Plants
AUIHOrllS)
9 ff.JTF-OR.VHNG ORGANIZATION NAME AND ADUntSs"
U. S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
1? SPONSORING AGENCY NAMfc AND ADDRESS
3 FlfcCliMENl'S ACCLSi,iOf*NO
!j FIEPOR r DATE
April 1_976 _
G ft RFOHMING OROANI 'ATlOr.1 CODt"
8 PEHFORMING ORGANIZATION HtHORl NO
10 PROGRAM tLLMENT NO "
11 CONTHACf/GRANT NO
13 TYPE OF REPORT AND PERIOD COVERED
14 SPONSORING AGENCY CODE
15 SUPPLEMENT ARY NOTES
The U. S. Environmental Protection Agency is required under 40 CFR Part 60
to publish a uuideline document for development of State emission standards after
promulgating any standard of performance for a designated pollutant. Standards of
performance limiting emissions of such a designated"pollutant--fluorides--from new ancl
modified phosphate fertilizer plants were promulgated on August 6, 1975, necossitatinn
the development of this guideline document. The document includes the followino
information: (1) Emission guidelines and times for compliance; (2) A brief descrip-
tion of the phosphate fertilizer industry, the five manufacturinq categories for
which emission guidelines are established, and the nature and source of fluoride
emissions; (3) Information regarding the effects of airborne fluorides on health,
crops, and animals; and (4) Assessments of the environmental, economic, and enemy
impacts of the emission guidelines. This is a draft guideline document; the final
document will be published after receipt and consideration of public comments
solicited in tfie FEDERAL REGISTER notice announcing the document's availability.
17
KEY WORDS AND DOCUMENT ANALYSIS
DtSCRIPTORS
Phosphate Fertilizer Plants
Fluorides
Standards of Performance
h IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
c COSATI I icId/Group
Unlimited
19 SfcCURITY CLASS flhit Hipnrri
Unclassified
20 SECUFiiTY CLASS (This page)
Unclassified
21 NO OF PAGES
2.19
22 PRICE
EPA Form 2220-1 (9-73)
------- |