ATMOSPHERIC EMISSIONS
FROM
WET-PROCESS PHOSPHORIC
ACID MANUFACTURE
Cooperative Study Project
Manufacturing Chemists' Association, Inc.
and
Public Health Service
U. S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
Environmental Health Service
National Air Pollution Control Administration
Raleigh, North Carolina
April 1970
-------
; AP series of reports is issued by the National Air Pollution Con-
Administration to report the results of scientific and engineering
dies, and information of general interest in the field of air pollution.
>rmation reported in this series includes coverage of NAPCA intra-
al activities and of cooperative studies conducted in conjunction
ri state and lo,cal agencies, research institutes, and industrial
anizations. Copies of AP reports may be obtained upon request,
supplies permit, from the Office of Technical Information and
plications, National Air Pollution Control Administration, U. S.
>artment of Health, Education, and Welfare, 1033 Wade Avenue,
eigh, North Carolina 27605.
tional Air Pollution Control Administration Publication No. AP-57
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, B.C., 20402 , Price « cents
-------
PREFACE
To provide reliable information on the nature and quantity of
emissions to the atmosphere from chemical manufacturing, the Public
Health Service, United States Department of Health, Education, and
Welfare, and the Manufacturing Chemists' Association, Inc. , entered
into an agreement on October 29, 1962, to study emissions from
selected chemical manufacturing processes and to publish information
that would be helpful to air pollution control and planning agencies and
to chemical industry management. * Direction of these studies is vested
in an MCA-PHS Steering Committee, presently constituted as follows:
Representing PHS Representing MCA
Stanley T. Cuffet Willard F. Bixbyt
Robert L. Harris, Jr. Louis W. Roznoy
Dario R. Monti Clifton R. Walbridge
Raymond Smith Elmer P. Wheeler
Information included in these reports describes the range of emis-
sions during normal operating conditions and the performance of
established methods and devices employed to limit and control such
emissions. Interpretation of emission values in terms of ground-
level concentrations and assessment of potential effects produced by
the emissions are both outside the scope of this program.
*Reports in this series to date are Atmospheric Emissions from Sul-
furic Acid Manufacturing Process, Public Health Service Publication
No. 999-AP-13, Atmospheric Emissions from Nitric Acid Manufac-
turing Processes, Public Health Service Publication No. 999-AP-27,
and Atmospheric Emissions from Thermal-Process Phosphoric Acid
Manufacture, Public Health Service Publication No. 999-AP-48, and
Atmospheric Emissions from Hydrochloric Acid Manufacturing PrO-
ces see, National Air Pollution Control Administration Publication
No. AP-54.
tPrincipal representative.
-------
ACKNOWLEDGMENTS
Many companies and individuals in the wet-process phosphoric
cid industry have been helpful in executing this study, and for their
:ontributions the project sponsors extend sincere gratitude.
Special thanks are due the following organizations for their par-
icipation in a program of stack sampling and analysis specifically for
his study:
Allied Chemical Corp.
American Cyanamid Co.
Armour Agricultural Chemical Co.
Farmland Industries, Inc.
W. R. Grace and Company
Hooker Chemical Corp.
International Minerals and Chemical Corp.
Olin Mathieson Chemical Corp.
Phosphate Chemicals, Inc.
U. S. Industrial Chemicals Co.
George B. Crane and Donald R. Goodwin, of the National Air Pol-
ution Control Administration, and James H. Rook, of the American
Cyanamid Co. , were the principal investigators in this study and are
he authors of the report. The sponsors acknowledge the contribution
>f American Cyanamid Co. in providing the services of Mr. Rook.
-------
USE AND LIMITATIONS OF THE REPORT
This report, one of a series concerning atmospheric emissions
from chemical manufacturing processes, is designed to provide infor-
mation on phosphoric acid manufactured by the wet process.
Background information describing the importance of the
process phosphoric acid industry in the United States is included.
Basic characteristics of the industry are discussed, including growth
rate in recent years, uses for the product, and number and location of
producing sites.
The only important wet-process phosphoric acid manufacturing
procedure in the United States today involves treatment of phosphate
rock with sulfuric acid. Descriptions are given of the most commonly
used process variations that involve the formation of calcium sulfate
dihydrate, since these account for the greater part of U.S. production.
Process information includes: discussion of factors that affect the
quantity of emissions, the normal range of emissions, and methods for
controlling emissions. Supplemental material provides detailed des-
criptions of emission-sampling and analytical methods.
The emission data used herein represent results from approxi-
mately 20 percent of the present number of establishments. ^ Most of
the data are derived from a series of stack sampling programs con-
ducted during 1966 and 1967 by the Public Health Service at ten estab-
lishments, which produce about 48 percent of the wet-process phos-
phoric acid made in the United States.
Although this report is a technical review prepared primarily for
public offi.C3.ls concerned with the control of air pollution, it is expected
that it will also be helpful to chemical plant management and its tech-
nical staff. This report should be reviewed at intervals to determine
whether revision is necessary.
*Establishment A works having one or more wet-process phosphoric
acid plants or units, each being a complete production entity.
-------
Page Intentionally Blank
-------
TABLES
1. Growth of Wet-Process Phosphoric 'Acid Industry in United
States 6
2. Composition of High-Grade Florida Land Pebble 10
3. Impurities in Phosphate Rock 10
4. Components of Typical Wet-Process Acid 11
5. Effect of Liming on Fluoride Evolution from Gypsum-Pond
Water 16
6. Concentration of Fluorides from Uncontrolled Process
Equipment in Wet-Process Phosphoric Acid Plant 19
7. Summary of Emission Data on Performance of Control
Equipment in Wet-Process Phosphoric Acid Plants. ... 22
8. Summary of Emission Data on Performance of Control
Equipment in Wet-Process Phosphoric Acid Plants 24
A-l. Performance of Emission Control Equipment in Wet-
Process Phosphoric Acid Plants - Gaseous and
Particulate Fluoride Emission Data 46
A-2. Gaseous Fluoride Emission Data from Wet-Process
Phosphoric Acid Plants 48
A-3. Gaseous and Total Fluoride Emissions from Wet-Process
Phosphoric Acid Plants 54
A-4. Wet-Process Phosphoric Acid Plant Fluoride Emissions
After Control Units 56
C-l. Wet-Process Phosphoric Acid Establishments in United
States 75
D-l. Physical Properties of Aqueous Solutions of Phosphoric
Acid , 78
D-2. Kinematic Viscosity of Phosphoric Acid Solutions .... 78
D-3. Vapor Pressure of Phosphoric Acid Solutions ... .78
D-4. Partial Pressure of Hydrogen Fluoride Over HF-H£O
Solutions .79
D-5. Partial Pressure of Water Over HF-H2O Solutions ... .79
D-6. Vapor Pressure of Anhydrous Hydrogen Fluoride ... .79
D-7. Physical Properties of Fluorine and Silicon Compounds 80
-------
FIGURES
Flow Diagram Illustrating Wet-Process Phosphoric Acid
Plants. 12
Typical Material Balance of Fluorine in Manufacture of
Wet-Process Phosphoric Acid 14
Fluoride Emission from Gypsum Pond Water Containing
10,200 ppm Fluorine 15
Principle of Spray Cross-Flow Packed Scrubber 25
-1. Sample Box with Pitot Tube, Impingers, and Umbilical
Cord 58
-2. Meter Box Controls 58
-3. Particulate Sampling Train 60
-4. Gas Sampling Train 62
-5. Schematic of Sampling Apparatus 62
-6. Operating Nomograph 66
-7. Correction Factor C for Figure B-6 67
-8. Fluoride Distillation Apparatus 71
viii
-------
CONTENTS
SUMMARY 1
Wet-Process Phosphoric Acid Production 1
Potential Emissions from Unit Processes '
Control of Emissions 2
Emission Guidelines 3
GROWTH OF WET-PROCESS PHOSPHORIC ACID INDUSTRY. . . 5
Historical Background 5
Current Production and Uses 6
Trends in Wet-Process Phosphoric Acid Manufacture 7
WET-PROCESS PHOSPHORIC ACID MANUFACTURE 9
Process Chemistry 9
Raw Materials 9
Final Product 10
Process Description 11
Distribution of Fluorine 14
EMISSIONS FROM WET-PROCESS PHOSPHORIC ACID MANU-
FACTURE 17
General Information 17
Sources of Emissions 17
METHODS OF EMISSION CONTROL 21
Contactor Design Considerations 21
Control Devices 21
Spray Cross-Flow Packed Scrubber 25
Packed Tower 26
Venturi Scrubber 2(
Spray Tower 2'
Impingement Scrubber 2'
Performance of Control Systems 2'
Description of Control Equipment in Plants Tested by PHS ... 3(
Handling of Scrubber Water 35
-------
Current and Future Air Pollution Potential .32
UMMARY OF SAMPLING AND ANALYTICAL TECHNIQUES. . . .35
Particulate Fluorides 35
Gaseous Fluorides 35
LOSSARY OF TERMS 37
PPENDIX A. EMISSION AND OPERATING DATA FOR WET-
'ROCESS PHOSPHORIC ACID PLANTS 45
PPENDIX B. SAMPLING AND ANALYTICAL TECHNIQUES . . 57
Introduction 57
Theory of Sampling Train Design 57
Particulate-Matter Sampling Procedure 58
Discussion of Gas Sampling 61
Apparatus for Gas Sampling; 61
Considerations Common to Gaseous and Particulate Sampling . 63
Selection of Sampling Points 63
Sampling Time and Equipment Cost 63
Field Calculations 64
Sampling Cleanup 68
Analysis of Particulate Matter 69
Spadns Determination of Fluorides 70
Calculations 72
PPENDIX C. WET-PROCESS PHOSPHORIC ACID ESTABLISH-
1ENTSINU. SD 75
PPENDIX D. PHYSICAL DATA 77
JBJECT INDEX 81
EFERENCES 85
-------
ATMOSPHERIC EMISSIONS FROM
WET-PROCESS PHOSPHORIC
ACID MANUFACTURE
SUMMARY
WET-PROCESS PHOSPHORIC ACID MANUFACTURE
In 1966, the production of wet-process phosphoric acid, expressed
as P^Oc, was approximately 3. 5 million tons. MVIuch of this was pro-
duced as 54 percent ^2^5' and virtually all of it was used to produce
various phosphate fertilizers j Fertilizers are produced by treating
phosphate rock with wet-process phosphoric acid to form triple Super-
phosphate, TSP, or by reacting phosphoric acid with anhydrous
ammonia to form ammonium phosphates, especially diammonium phos
phate, DAP.
Wet-process acid is produced by treating fluorapatite 1 Ca. 10(PQ4)6
F2 or phosphate rock, with sulfur ic acid. Phosphoric acid is formed.
calcium sulfate is precipitated and filtered off, and the acid is concen-
trated from about 32 percent P2O5 to about 54 percent ?2O5 Phosphate
rock is found all over the world and varies in physical and chemical
properties. An acid plant must be designed for the rock it will process
Although sulfur ic acid of any strength will cause the desired reaction,
in practice, 98 percent acid is used.
POTENTIAL EMISSIONS FROM UNIT PROCESSES
Phosphate rock must be finely ground to react properly with sul-
furic acid, and standard control equipment is normally used to prevent
objectionable dust emissions.
The emissions of most concern are fluoride compounds liberated
from the rock by the sulfuric acid. These consist of hydrogen fluoride,
silicon tetrafluoride, and some products of reaction and decomposition
of the latter. Most phosphate rock contains 3.5 to 4 percent fluorine,
and half of this may be volatilized in the processing. This represents
a large potential source of pollution.
Fluoride emissions may occur from exposed surfaces of reaction
slurry, aqueous solutions of fluorine compounds, and any evaporation
process. Thus, reactors, open-slurry launders, flow splitter boxes,
-------
aporators, filters, and sump tanks are potential emission sources.
The quantity of gaseous fluorides generated in the digester ranges
>m 0.037 to 2.16 pounds per ton of acid produced. The level of
seous fluorides evolved from the filter ranges from 0.011 to 0.63
und per ton of acid, while as much as 0.26 pound of gaseous fluoride
r ton of acid is generated in the sump and associated vents. Total
rticulate emissions amount to approximately 0.20 pound per ton of
id for filter operations, and as much as 11 pounds per ton for digester
erations. Only a small portion, i. e, , 3 to 6 percent of the particu-
;e emissions, consists of fluorides. Fluoride emissions may occur
>m gypsum ponds, and the quantity of emissions depends on pH and
emical composition of the pond and upon temperature and wind speed.
ita for one gypsum pond given in this report indicate a possible
loride emission of 0. 4 to 1. 8 pounds of fluoride per acre per day,
pending on temperature.
»NTROL OF EMISSIONS
Because the principal atmospheric contaminants generated in the
Dcess are gaseous fluorides, vapor scrubbing is universally em-
>yed to control emissions. Specific devices used for control include
ituri scrubbers, impingement scrubbers, and various kinds of spray
vers. Fluoride removal efficiency of these devices varies widely,
d staging may be required for satisfactory control. Plugging, or
Ticulty in removing precipitates and dust, may also be experienced.
Tables in Appendix A show the results of MCA-PHS stack tests on
i wet-process phosphoric acid plants in various parts of the country.
r nine of these plants, the range of gaseous fluoride emissions from
rious types of collectors was 0. 006 to 0. 17 pound of fluoride per ton
P^Os produced. The concentration range of gaseous fluorides in the
ses from collectors was 3 to 40 parts per million, and 0.0011 to
1147 grain per standard cubic foot for eight of the ten plants. Public
alth Service stack-test data agree reasonably well with results from
int questionnaires and information from miscellaneous sources, both
which are tabulated in Appendix A.
The spray cross-flow, packed scrubber is reported to be capable
over 99 percent efficiency in the removal of pollutants. The usual
it concentration range for this type of scrubber is 0. 001 to 0. 01
lin of fluoride per standard cubic foot, according to stack samples
ten by Public Health Service personnel for this project.
Scrubber efficiency is affected substantially by the'loading of the
3 stream. Heavy loading enhances scrubber efficiency, and light
ding reduces scrubber efficiency. Therefore, scrubber-exit-gas
icentration is a better indicator of overall plant emission control
in is scrubber efficiency. The best criterion of plant performance
the weight of emission per ton of P2O5 produced.
A summary of plant tests made for this project shows the range of
icentration of gaseous fluoride emissions after the scrubber.
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
GASEOUS FLUORIDE
Control device
Company-constructed I
spray chambers
Venturi scrubbers
Cyclone spray towers!
Spray cross -flow
packed scrubber
Scrubber
efficiency, %
55 - 75
84 - 96
90 - 95
60 - 93
Scrubber exit
loading,
gr/ scf
0. 0026 - 0.090
0.010 - 0. 023
0.0016 - 0.003
0. 001 - 0. 014
Emissions,
Ib/ton P^Oc)
produced
0. 072 - 0.63
0. 027 - 0.047
0. 047 . 0. 082
0.006 - 0. 17
Performance data on the first two control devices mentioned in the
preceding list relate to the treatment of digester emissions only. Thei
performance for emissions other than jfaseous emissions was as follov
Control
device
Company-
constructed
spray
chamber
Venturi
scrubber
Pollutant
Particulate
Insoluble F
Soluble F
Particulate
Insoluble F
Soluble F
Scrubber exit
loading,
gr/scf
0. 04 - 0. 47
0. 0009 - 0. 0011
0. 0003 - 0. 014
0 0.009
none found
0.0009
Scrubber
efficiency,
% !
above 98
about 100
94 ' 97
Emissions,
Ib/ton Pz°5
produced
0.28 . 0.50
0 . 0.00!
0.0075 . 0.09'
0. -0.3
none found
0.003
EMISSION GUIDELINES
The major source of gaseous fluoride emissions in wet-process
phosphoric acid plants is the digester. Only trace quantities of partic
ulate fluorides are normally present in exit gases from digesters and
filters, and these can be removed effectively by scrubbing.
The technology for controlling gaseous fluoride emissions by wate
scrubbing has been available for many years. By proper attention to
mechanical design and good mass-transfer practice, such a unit can be
built and operated to obtain almost any desired reduction in gaseous
fluoride emissions. Such scrubbers are capable of operating with
collection efficiencies of over 99 percent. The usual exit concentratio
for this type of scrubber ranges from 0. 001 to 0. 01 grain of fluoride j
standard cubic foot or 0.006 to 0.17 pound of fluoride per ton of P^Oc
produced.
It should be practical to operate wet-process phosphoric acid plani
within the above ranges if plants are designed to prevent or collect
Summary
-------
emissions , if modern scrubbers are used, and if attention is directed
toward proper operation and maintenance of both process and emission
control equipment.
Proper attention to air pollution control would dictate that the water
scrubbers be started before process equipment and operated for a brief
period after plant shutdown.
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
GROWTH OF WET-PROCESS
PHOSPHORIC ACID INDUSTRY
HISTORICAL BACKGROUND
The agricultural benefits gained by mixing materials such as bone
and bird guano with the soil have been observed from ancient times.
Gahn, in 1669, was perhaps the first to associate the phosphorous con-
tent of such materials with their ability to fertilize the soil. Jjiebig,
in 1840, suggested solubilizing the phosphorous content of bones by
treatment witn sulfuric acid. 3 By this time, population growth had
caught up with the ability of European soil to produce, and this fur-
nished incentive for the extension of Lie big' S ideas into various ways
of treating phosphorus -bearing materials with strong acids. This
activity soon led to the idea of treating phosphate rock with phosphoric
acid instead of sulfuric acid. The phosphoric acid, it was discovered,
could be made by decomposing phosphate rock with sulfuric acid and
filtering off the resulting calcium sulfate. Thus, by 1872, wet-process
phosphoric acid was being made in Germany and used in manufacturing
triple superphosphate. 4 This work was taken up in America, and, by
1890, a triple-superphosphate plant was operating in Baltimore.
Early wet-process phosphoric acid plants were simple; they
involved batch treatment of phosphate rock with dilute sulfuric acid.
The physical chemistry involved was poorly understood, and process
controls were rudimentary. Filtration difficulties resulted in losses
of phosphate in the calcium sulfate filter cake.
The control difficulties of batch processing led to early attempts
to devise a continuous wet-acid process. The Dorr weak-acid process
was an important contribution developed before 1930. It used a con-
tinuous reaction system, but was capable of producing acid no more
concentrated than 20 to 22 percent. 5
The principle of the Dorr strong-acid process, developed about
35 years ago, is employed in the production of most wet-process acid
today. This process involves adding ground-rock feed and sulfuric
acid to a large stream of recirculated reaction slurry. Compared to
the weak-acid process, this process enhances yield and filterability
by minimizing local changes in sulfate ion concentration, by furnishing
system capacitance, and by furnishing proper sites for crystal growth.
The acid filtrate is 30 to 32 percent
Several variants of the above process are employed in modern
plants to separate calcium sulfate as gypsum crystals. In addition to
the Dorr strong-acid process, the Pray on and St. Gobain '' "~p~r o c e s s e s
are used. All produce an acid filtrate containing about 30 to 32 percent
-------
process complications prevent production of stronger acid by
hese processes. Concentration of this filtrate to about 54 percent
'jOc is accomplished by evaporation in yacuiiTn evaporators or by
iubmerged combustion. Growth of the industry is shown in Table 1.
Table 1. 'GROWTH OF WET-PROCESS PHOSPHORIC
ACID INDUSTRY IN UNITED STATESl
Year
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
Production of 100 % Pg^S' *°ns
131.000
119.000
127,000
141.000
133.000
165,000
175.000
221.000
245.000
299,000
338.000
389,000
496.000
631,000
775.000
812.000
936.000
1.033.000
1,141,000
1.325.000
1,409,000
1.577.000
1.957.000
2,275,000
2.837.000
3,533,000
lUEEENT PRODUCTION AND USES
In spite of the large absolute value of wet-process phosphoric acid
reduction, the yearly rate of increase in production is maintained
ecause of the soaring demand for concentrated or high-strength ferti-
izers, which consume most of the wet-process acid. Monoammonium
hosphate and diammonium phosphate, two important examples of this
ype of fertilizer, are produced by ammoniating wet-process phos-
horic acid with anhydrous ammonia. By adding various amounts of
>ther ammonium salts, potash, and inert extenders, a great variety
if solid and liquid fertilizers can be produced at any desired ratio of
itrogen-phosphorous-potassium content.
Because wet-process phosphoric acid contains a few impurities
such as fluoride) in significant amounts and many impurities in trace
mounts, uses of wet-process acid in other fields are limited. If the
verall economics are favorable, this acid can be used for uranium
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
recovery^ or for phosphate salt production. Phosphoric acid made
from elemental phosphorus by the thermal process is used for foods
and in applications requiring chemical purity. In 1966, thermal-pro-
cess phosphoric acid accounted for only 22 percent of the total United
States production of phosphoric acid, whereas wet-process phosphoric
acid made up the balance. ^
TRENDS IN WET-PROCESS PHOSPHORIC ACID MANUFACTURE
The current trend in wet-process phosphoric acid manufacture is
toward larger producing units with closer control of operating variables.
Two important incentives for change exist: the increasing demand for
sulfuric acid has exerted strong upward pressure on sulfur prices, and
handling, and shipping costs have increased the demand for higher-
strength phosphoric acid.
As the price of sulfuric acid increases, the relative cost of acidu-
lation of phosphate rock with nitric acid will become more attractive.
Nitric acid acidulation is presently practiced in Europe, and an
increase in the developmental activity on improvements in this process,
and probably in methods for acidulation with hydrochloric acid, can be
foreseen.
Special processes are used to concentrate 54 percent P^O^ *° ^
percent P^^c super phosphoric acid. The reduction in water content
of course reduces shipping cost. This acid is less corrosive than 54
percent P^Or acid. Because it can be supercooled without solidifying,
it can be stored in liquid form at subfreezing temperatures. Further
technological development is expected.
Another trend is toward processes that directly produce acid
filtrates of higher P^Oc content. One designer offers a process that
produces 42 percent P2^5 filtrate by a method involving two-stage
crystallizationT"" OtrTer" designers are working on processes that
achieve similar results. Solvent extraction is being investigated, and
high acid concentrations have been achieved by solvent processes on a
small scale. ^"
Growth of Industry
-------
Page Intentionally Blank
-------
WET-PROCESS
PHOSPHORIC ACID MANUFACTURE
PROCESS CHEMISTRY
The basic raw material of this process is phosphate rock, con-
taining the mineral fluorapatite and numerous impurities. Fluora-
patite, Cajn(PO4)6 F?> is a salt from which phosphorus can be
extracted as orthophosphoric acid by double decomposition with a
mineral acid. In practice, 93 or 98 percent sulfuric acid is normally
used. Calcium sulfate precipitates, and the liquid phosphoric acid is
separated by filtration. The reaction is described by the following
equation:
3 Caio(PO4)6 FZ t 30 H2SO4 t SiO2 t 58 H2O *-30 CaSOq ' 2HzO t
18 H3PO4 t H2 SiF6
In commercial practice the sulfuric acid is generally mixed and
diluted in the reaction vessel with a large excess of phosphoric acid in
the rock slurry. By proper control of the temperature and amount of
water in the slurry, crystal structure of the precipitated calcium sul-
fate can be controlled to facilitate filtration. Low temperatures and
low acid content yield calcium sulfate dihydrate, CaSC>4 2H20, or
gypsum, whereas higher temperature and higher acid strengths yield
semi-hydrate, CaSOq . 1/2 H2O, or anhydride, CaSOq. The ease of
filtration is dependent on proper growth, size, and shape of crystals.
Most modern plants are designed so that they produce the dihydrate.
Phosphate rock usually contains 6.5 to 9. 0 percent silica, which,
in the presence of acid, reacts in various ways with the fluoride in the
rock. It is probable that fluorine is released as a mixture of fluosilicic
acid, silicon tetrafluoride, and hydrogen fluoride.
RAW MATERIALS
If phosphate rock ore were a simple calcium orthophosphate or
even pure apatite, wet-process acid manufacture would be easier and
cheaper than it is. Phosphate rock is found in workable amounts in
many countries. The composition varies from one location to another
and even within the same rock bed. The analysis in Table 2 of a high-
grade Florida land pebble illustrates the normal complexity of phos *
phate rock. In addition to the constituents listed, other elements are
usually present in traces.
-------
Table 2. COMPOSITION OF HIGH-GRADE
FLORIDA LAND PEBBLE 10
Component
P2°5
CaO
MgO
A1203
FegOs
Si02
S03
Weight, %
35.5
48.8
0.04
0.9
0.7
6.4
2.4
Component
F
Cl
C02
Organic carbon
Na2O
K30
H20(100°C)
Weight, %
4.0
0.01
1.7
0.3
0.07
0.09
1.8
Composition ranges of impurities for 15 types of phosphate rock
from seven locations of origin are summarized in Table 3.
Table 3. IMPURITIES IN PHOSPHATE ROCK*!
Component
MgO
Alg03
S03
Cl
Na20
K20
Range of weight, %
0.01 2.2
0.5 -15
0.01 3
0.001 . 0.2
0.005 * 1.5
0.1 1.0
Commercial phosphate rock usually contains 31 to 35.5 percent
Flourine content is usually in the 3.5 to 4 percent range.
Because iron and aluminum oxides form insoluble phosphates, they
are undesirable constituents. Carbonates are undesirable because
they consume sulfuric acid, thus liberating carbon dioxide, which
contributes to foaming. Some phosphate rock has a high organic con-
tent that may cause foaming and interfere with phase separations and
the desired chemical reaction(s). For these reasons, each plant should
be designed for the particular phosphate rock that it will use.
Although any strong mineral acid can be used to decompose phos-
phate rock, sulfuric acid is used for process and economic reasons.
The insoluble calcium sulfate formed when sulfuric acid is used can be
easily separated from the liquid. In order to make the strongest pos-
sible phosphoric acid and to decrease later evaporating costs, 93 or 98
percent sulfuric acid is normally used. Spent sulfuric acid can be
used, but this introduces additional impurities that may contribute to
foaming and increase corrosiveness. Any residual organic content of
the spent acid may cause an odor problem.
FINAL PRODUCT
A-modern, wet-process phosphoric acid plant produces 30 to 32
percent P205 acid which is then concentrated to about 54 percent.
Table 4 shows a typical analysis of commercial wet-process phosphoric
acid.
10
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Table 4. COMPONENTS OF TYPICAL WET-PROCESS ACID12
Component
P2°5
Ca
Fe
Al
Mg
Cr
V
HoO and other
Weight. %
53.4
0.1
1.2
0.6
0.3
0.01
0.02
37.56
Component
Na
K
F
S03
SiOg
C
solid
Weight, %
0-2
0.01
0.9
1.5
0.1
0.2
2.9
In addition to the components listed in Table 4, which may vary
considerably, other trace elements are commonly present. Because
commercial wet-process acid is a complex, corrosive material, cor-
rosion is a major problem in its manufacture. The achievement of
effective plant designs was not possible until modern construction
materials were developed.
In addition to causing process difficulties, impurities affect physi-
cal properties of the acid. Commercial, wet-process acid has a higher
viscosity than pure orthophosphoric acid of the same concentration.
This tends to increase difficulty in filtering calcium sulfate -formed
during acidulation of the phosphate rock. In general, impurities indi*
rectly affect atmospheric emissions by increasing corrosion and sub-
sequent leakage, and increasing downtime, which provides opportuni-
ties for the escape of pollutants.
PROCESS DESCRIPTION
Most current process variations for producing wet-process phos-
phoric acid depend on decomposition of phosphate rock by sulfuric acid
under conditions where gypsum (CaS04 ' ZH^O) js precipitated. Sev-
eral va.ri3.nts of this process are offered by various contractors. The
Dorr-Oliver, " gt. Gobain, ^ Prayon, and Chemico processes are
among the better known. Most of the contractors in the chemical con-
struction industry design and build these plants. Moreover, the growth
of the fertilizer business has attracted some able new contractors
during the past 5 years. In spite of the number of contractors in the
field, new plants do not seem to differ fundamentally among themselves.
In addition, several general trends are evident, such as the use of
single-tank instead of multiple-tank reactors, one or two large hon-
zontal'tilting-pan filters, large plants of 1000-ton~per~day capacity and
more, and closed systems where atmospheric emissions are minimized.
Figure 1 is a flow diagram of a modern, wet-process phosphoric acid
plant.
Finely-groundphosphate rock is metered accurately and continu-
ously 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 use the best automatic control equipment available.
Wet-Process Phosphoric Acid
11
-------
WASH .
WATER
GYPSUM
POND WATER
H2S04
GROUND
PHOSPHATE
ROCK
GYPSUM SLURRV^.
TO POND
"
SUCK ICLOTHCAKE
DRY I WASH REMOVE | DRY
I p"SK I 3 WASH | 2 WASH | 1 WASH | LIQUOR
TO VACUUM
"AND HOT WELL
TO SCRUBBER
HYCROFLUOSILIC1C ACID
-------
The single-tank reactor illustrated in Figure 1 is a circular, two-
compartment system wherein reactants are added to the annular volurr
and the central volume is used for growing gypsum crystals. Some
years ago, plants were built with several separate reaction tanks con-
nected by launders, which are channels for slurry flow. The tendency
now is to use a single tank with several compartments. In some of
these designs, the slurry flows over and under a series of baffles.
Proper crystal growth depends on maintaining sulfate ion concen-
tration 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 by causing deposition
of calcium sulfate on unreacted rock. ^"* Good reactor design will pre
vent sudden changes of sulfate ion concentration, will-maintain this
concentration and temperature near optimum, and will provide suffi-
ciently 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 habit when thej
are present in a medium where crystallization is taking place. Usuall;
this impurity effect is detrimental. Such impurities are likely to caus
crystal fragmentation, small crystal size, or a shift to needles or
other hard-to-filter forms. It is suspected that impurities in some
plants and at some times interfere with desired crystal iformation.
Concentrated sulfuric 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 H3PO4 recoveries.
Considerable heat of reaction is generated in the reactor and mus
'be removed. This is done by blowing air over the hot slurry surface
or by vacuum flash cooling part of the slurry and sending it back into
the reactor. Modern plants use vacuum flash cooling. Figure 1 illus
\trates this method of cooling.
/" The reaction slurry is held in the reactor for periods up to 8 hour
/ depending on the rock and on reactor design, and is then sent to be
i filtered. The circular, horizontal, tilting-pan vacuum filter is illuS-
\ trated in Figure 1. Older and smaller plants may use other types of
\ filters.
In washing the resultant gypsum cake on the tilting-pan filter, VV3.S
water flow is countercurrent to the rotation of the cake, and heated
fresh water is used to wash the "cleanest" cake. These filters can be
built in very large sizes, and designs are now approaching 1000-ton-
per-day P£O5 capacity.
I The 32 percent acid from the filter generally needs concentrating
for further use. Current practice is to concentrate it by evaporation
in two or three vacuum evaporators. Concentration to above 54 per-
Wet-Process Phosphoric Acid
-------
cent P^Os is not practical, because the boiling point of the acid (Table
D-l) rises sharply above this concentration, even at 27 inches Hg
vacuum. Corrosion problems alsobecome more difficult when con-
centration exceeds 54 percent. In the evaporator, illustrated in Figure
1, provision is 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.
DISTRIBUTION OF FLUORINE
Figure 2 illustrates a typical material balance for fluorine origi-
nally present in phosphate rock. It should be noted that the results in
any given plant may differ considerably from those shown in the figure,
which represents an example based upon data from several sources.,
Actually, the fluorine distribution will depend upon the type of rock
treated, process used, and kind of operation prevailing.
VOLATILIZED
BAROMETRIC CONDENSER
AND OTHER LOSSES
1 -01 Ib F
ft
32%
PHOSPHORIC
DILUTE PRODUCT
TO CONCENTRATOR
PHOSPHORIC
ACID
2-11 Ib F
CONCENTRATED
PRODUCT
I I
s« PHOSPR'C *CID
1-10 lb F
Figure 2. Typical material balance of fluorine in manufacture of wet-process phosphoric
acid.
Figure 2 indicates that 0. 93 pound of fluorine is volatilized (as HF,
SiF4, etc.) by acid attack on 100 pounds of the rock. This volatiliza-
tion varies considerably in practice. If reactor slurry is cooled by
air, the fluoride can be absorbed from the air stream by a water
scrubber. If it is cooled by vacuum flash, much of the fluoride will
be dissolved in the barometric-condenser water. This fluoride-bearing
water may be sent to a pond, where limestone or lime may be added
to raise the pH and convert fluoride to insoluble calcium fluoride.
Here, silica would be present in the soil to convert hydrogen fluoride
to fluosilicates.
The foregoing applies also to the concentration of the 32 percent
acid, in which volatile fluorides also pass to the barometric condenser,
which is part of the system used to create vacuum for the evaporator.
Fluorides may be emitted from filters and seal boxes, feed boxes, and
other points in the plant. The fluoride evolved from the acid-concen-
14
WET-PROCESSPHOSPHORIC ACID EMISSIONS
-------
tration step will be almost completely controlled if it is recovered, as
illustrated in Figure 1, by conversion to hydrofluosilicic acid.
Figure 3 shows fluoride emissions from the water in an actual
gypsum pond, as determined by personnel at plant 17. The experi-
mental method used consisted in passing air, at a known rate, over a
relatively large amount of gypsum-pond water and analyzing the exit
air for fluoride. Measurements were made at six temperatures. The
air-water interface was 20 feet long. It is doubtful that the air was
saturated after this length of travel, so Figure 2 is probably conserva-
tive; this means that fluoride evolution from this particular pond, at
the given windspeed and with reasonable vertical mixing of air above
the pond, is probably somewhat greater than indicated by the curve.
2.d
-I ''5
1
0.5
AIR VELOCITY = 5 mph
PH = 1.3
40
50
100
no
60 70 30 90
WATER TEMPERATURE, «p
Figure 3. Fluoride emission from gypsum pond water containing 10,200 ppm fluorine.
Wet-Process Phosphoric Acid
15
-------
The data in Figure 3 may be compared with those of a recent
ournal article (JAPCA 19 (1): 15-17) where a gypsum-pond fluoride
evolution of 0.16 pound per acre per day is given. In the author's
opinion, this emission factor is a minimum. It is based on a plant other
han plant 17; also the soluble fluoride concentration was about 4, 000
>arts per million and the pH was about 1. 7.
Among the ten constituents analytically determined in plant 17
?yp.SUTn-p*ond water, were Na, K, Si, NH4, and S04. This complexity of
composition may cause the volatility of fluoride to differ among gypsum
>onds, even at similar fluoride concentrations in the water and identi-
:al water temperatures.
Fluoride evolution from gypsum ponds may be made negligible by
aising the pH of the pond by liming. Table 5 shows the results of
iming the water in the pond of plant 17. Actual ponds are seldom
imed due to cost.
Table 5. EFFECT OF LIMING ON FLUORIDE EVOLUTION
FROM GYPSUM-POND WATER
Iffl
1.4
2.6
3.0
3.3
3.9
4.5
6.1
6.25
7.72
9.7
12.1
12.3
12.5
Soluble
fluoride,
ppm
8125
4000
450
106
100
106
16
Ca(OH)2,
lb/gallon
0.116
0.145
0.156
0.157
0.160
0.192
0.193
0.207
0.213
0.222
0.246
0.346
Vapor pressure
of fluoride
8 25°C. mmHg
13.8 X 10'6
6.22 X10'6
0.86 x icr6
0.45 X 10"8
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
EMISSIONS FROM WET-PROCESS
PHOSPHORIC ACID MANUFACTURE
GENERAL INFORMATION
Emissions from wet-process phosphoric acid manufacture consist
of rock dust, fluoride gases, particulate fluoride, and phosphoric acid
mist, depending on the design and condition of the plant. Fluorine
exists as various compounds in the collection equipment: as fluorides,
silico fluorides, silicon tetrafluoride, and mixtures of the latter and
hydrogen fluoride, the mol ratio of which changes in the vapor with the
concentration of fluosilicate in the liquid and with temperature. Because
of the complex chemistry, the composition of emissions is variable.
The usual practice in sampling and analysis has been to avoid determi-
nations of individual compounds such as silicon tetrafluoride and to
express the various emissions as fluorine equivalents. Little informa-
tion has been published on actual composition of emissions or on quanti-
tative values of emissions from minor sources. Data available are
mainly for emissions from digesters and filters, and are expressed as
fluorine equivalents.
SOURCES OF EMISSIONS
^
( Phosphate rock contains 3. 5 to 4 percent fluorine, and the final
distribution of this fluorine in wet-process acid manufacture varies
widely. In general, part of the fluorine goes with the gypsum, part
with the phosphoric acid product, and the rest is vaporized. The pro-
portions and amounts going to gypsum and acid depend on the nature of
the rock and on process conditions. Disposition of the volatilized
fluorine depends on the design and operation of the plant. Substantial
amounts pass off into the air unless effective scrubbers are used A
f . .
rThe reactor, where phosphate rock is decomposed by sulfuric acid,
is the main source of atmospheric contaminants .^ The heat of reaction
is considerable and must be removed to prevent an excessive tempera-
ture rise. A practicable way to remove heat is as latent heat of evap-
oration of the slurry water. The slurry is abrasive and highly corro-
sive to most materials of construction; therefore, for many years,
cooling was accomplished by blowing air over the slurry surface, there-
by removing latent heat with the water vapor evolved from the slurry
and carrying away some additional heat as an increase in the sensible
heat of the air. With better pumps and superior materials of construc-
tion, it became possible to vacuum flash cool the slurry by pumping it
in and out of a vacuum vessel. Vacuum flash cooling is the most common
method in current use. Emissions are minimized by this method because
the system is closed. There is only a small volume of inert gases to be
liandled with the water vapor and fluorine. A disadvantage is that it
is impractical to recover fluoride from the very large volumes of
17
-------
barometric-condenser water used. Theoretically, it should be possi-
ble to remove this fluoride by the use of properly designed scrubbers
ahead of the barometric condensers, but this is not normal practice
at present.
Digester cooling by air blowing requires large volumes of air in
relation to the water vapor and fluoride removed. The fluoride can be
recovered by scrubbing; but because of the large volumes of gas han-
dled, operating costs are increased substantially.
/Acid concentration by evaporation provides another source of
fluoride emissions. In this operation it has been estimated that 20 to
40 percent of the fluorine originally present in the rock vaporizes
(Figure 2),J) The acid-concentration operation is usually vacuum evap-
oration, and the fluoride is partly dissolved in the barometric-conden-
ser water. In acid concentration, good recovery of fluoride is possi-
ble by means of absorption of the vapors in water, forming hydrofluo-
silicic acid. A process has been patented ^^ to scrub these vapors with
a 15 to 25 percent hydrofluosilic solution at a temperature at which
water vapor, which would dilute the solution, is not condensed. The
water vapor itself is later condensed in the barometric condenser
ahead of the vacuum system. The scrubbers are spray towers com-
bined with vacuum evaporators to form single vessels, resulting in a
series of evaporator -scrubbers. In such a vessel, vapors from the
lower evaporator section pass to the spray chamber above, but the
resulting fluosilicic acid flows to a storage sump; it cannot flow down
into the vacuum evaporator. This arrangement is illustrated on the
evaporator shown in Figure 1.
The filter is a third source of fluoride emissions. For circular
filters, and for filters of the Georgini pan-filter type, most of the
emissions are at feed and wash points. Emissions from filters are
not large and can be controlled by the use of hoods, vents, and scrub-
bers.
In addition to these three main sources of emissions, there are
many miscellaneous minor sources. These include vents from such
sources as acid splitter boxes, sumps, and phosphoric acid tanks.
Collectively, these sources of fluoride emissions are significant, and
they are often enclosed and vented to a suitable scrubber.
Emissions from a wet-'process phosphoric acid plant, except for
rock dust, may come from: rock digesters, filters and their acces-
sories such as the feed box and seal tank, the evaporator hot well,
sumps , and acid vessels. In most plants, all of these sources are
controlled.
Table 6 shows concentrations of fluoride at various points in the
process ahead of control equipment. All units of emissions are grains
of fluorine per standard cubic foot, except that total particulateS are
expressed in grains per standard cubic foot. These results were
obtained from tests made by Public Health Service personnel.
18 WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
TableG .CONCENTRATION OF FLUORIDES FROM UNCONTROLLED PROCESS
EQUIPMENT IN WET-PROCESSPHOSPHORICACID PLANTS
Gaseous fluoride
From digester
From filter
From sump and vent
particulate fluoride
From filter
From digester
Soluble
Insoluble
Total particulates
From filter
From digester
about 0.017
0.47 - 3.73
Stack-test data show that greater quantities of fluorides are gen-
erated in digesters than in filters and sumps. Filter emissions of
gaseous fluoride were in the range of 0. Oil to 0.063 pound per ton of
P205 produced; sump and vent emissions were as high as 0.26 pound
per ton of P^05> and emissions from digesters ranged from 0.037 to
2.16 pounds per ton of P2O5 made.
Total particulate emissions directly from process equipment were
measured for one digester and for one filter, in different plants. As
much as 11 pounds of particulates per ton of 3?2O5 was produced by the
digester and approximately 0.20 pound per ton of P£O5 was released by
the filter. Only 3 to 6 percent of these particulates were fluorides A
Particulates can be removed by jet Venturis and certain other typSs of
wet collectors. If necessary, residual fluoride gases can be removed
by scrubbers already mentioned.
High fluoride concentration and low pH of the scrubbing water will
tend to evolve fluorides from the scrubber and from the gypsum pond,
which stores the scrubbing water. Surveillance is advisable.
Small amounts of SO2 are sometimes evolved from a digester;
the origin of this gas is not clear. Odors sometimes develop from
Organic material in the phosphate rock or in the sulfuric acid used, if
the latter is spent acid.
Emissions
19
-------
Page Intentionally Blank
-------
METHODS OF EMISSION CONTROL
CONTACTOR DESIGN CONSIDERATIONS
Reactions involved in phosphate rock attack by sulfuric acid are
complex and subject to debate, but may be generally representated by
the following equation:
3 Ca)0(PO4)6 F2 t 30 H2SO4 t SiO2 t 58
30 Ca SO4 . 2 H20 t 18 H3PO4 t
Under the existing conditions of temperature and acidity, the
fluosilicic acid decomposes as follows:
H2 SiF6 > SiF4 t 2HF
HF
Actually, the mol ratio c:p changes with conditions, such as concen-
tration, and is not usually equal to 2. The SiF4 and HF constitute the
gaseous emissions to be controlled. When SiF4 contacts water, the
following reaction occurs:
3 SiF^ t 4H2O j_ 2 HzSiFfc t Si (OH)4
Hydrated silica in the wet and newly formed state sticks to control
equipment surfaces and plugs gas flow channels. Furthermore, it
absorbs additional SiF4.
All wet-process 'phosphoric acid plants emit SiF^ and probably HF
to a lesser extent. Designers for control recognize this fact and send
the various streams to scrubbers adapted to handle each stream. The
tendency today is toward one scrubber combining the above functions
and having at least two entrances to accommodate the different kinds
of gases.
In general, control of HF by absorption is straightforward. Hydro-
gen fluoride can be absorbed by several kinds of scrubbers, including
conventional packed towers and irrigated packed sections. Because
of the tendency of SiF4 to decompose and cause plugging due to the
deposition of silica, high SiF4 gas loadings are best reduced by spray
towers or other devices that are less susceptible to plugging. After
the SiF4 loadings are substantially reduced, the residual SiF4 can be
handled without complications. A good design should include provision
for removal of any silica that does form, especially if packing or grids
are used.
CONTROL DEVICES
A control device should capture all of the emissions from proces-
sing without any leaks or losses. The device should then be able to
21
-------
absorb substantially all of the fluoride without equipment stoppage or
failure due to plugup by solids, precipitated or otherwise. Ideally,
the device should be inexpensive and the pressure drop and maintenance
costs should be low. The following discussion covers control devices
that are commonly used; not all of them are well adapted to general use
in wet-process phosphoric acid plants. Most of them are adaptations
of equipment pieces originally developed for use in other industries:
such as, ordinary packed towers, which absorb some gases well, but
which must be used with caution in wet-process phosphoric acid plants.
A description of control equipment in the ten plants tested by Public
Health Service personnel will be given later. Detailed information on
emissions and the performance of emissions control systems is in-
cluded in the Appendix, Tables A-l through A-3. Tables 7, and 8
summarize these data.
Table 7. SUMMARY OF EMISSION DATA ON PERFORMANCE OF CONTROL.
EQUIPMENT IN WET-PROCESS PHOSPHORIC ACID PLANTS'*
Plant number
Collector type
Gaseous fluoride entering
collector per ton of PpOc;
produced. Ib
Gaseous fluoride emitted from collec-
tor per ton 1*2^5 Pro(luced, Ib
Collection efficiency. %
Concentration of gaseous fluoride
emitted from collector.
grain/scf
ppm
Particulate emitted from collector
per ton PoOt produced. Ib
Total participates
Efficiency. %
Insoluble particulate fluorides
Efficiency, %
Soluble particulate fluorides
Efficiency. %
1
Rectangular
spray
chamber
1.265-2.16
0.52-0.63
67-72.
0.075-0.090
802-243
0.28-0.60
.0006-0.0081
1.050-0.094
2
Square
horizontal
spray duct
Not
determined
0.072-0.101
0.0026-0.0035
7.0-9.4
0.36-0.47
0-0.0013
0.0075-0.036
3
Venturi
scrubber.
Water-
actuated
0.21-0.31
0.027-0.047
84.2-87.0
0.0104-0.0147
28-40
0-0.029
98.5-100
none found
100
0.0023-0.0029
94.0-97.0
4
Venturi
scrubber.
Water-
actuated
0.49-0.67
0.028-0.03 8
92-96
0.018-0.023
49-62
5
Spray
cross-flow
packed
scrubber
0.078-0.087
0.006-0.018
80-92.4
0.0011-0.0032
B . 0 - 8 . 61
aPlants 1-10 were tested by NAPCA.
22
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Table 7 (continued). SUMMARY OF EMISSION DATA ON PERFORMANCE OF CONTROL
EQUIPMENT IN WET-PROCESS PHOSPHORIC ACID PLANTS*
Plant number
Collector type
Gaseous fluoride entering
collector per ton of P^Og
produced, Ib
Gaseous fluoride emitted from collec-
tor per ton PgOj produced, Ib
Collection efficiency, %
Concentration of gaseous fluoride
emitted from collector.
grain/set
ppm
Pamtttoiiitste aniita^d from cotlraetor
per to" tPgC^ -produced. Ib
Total particulates
Efficiency, %
Insoluble participate fluorides
Efficiency, %
Soluble participate fluorides \
Efficiency, %
6
TWO
impingement .
scrubbers
in series
0.013-0.016
0.006-0.011
15-62
0.0020-0.003;
5.4-10.0
7
Spray
cf oss-flow
packed
scrubber
1.20-1.48
e,J«KOH7
8
Spray
cross-flow
packed
scrubber
0.05-0.06
0.0170-0.02;;
86-93 56.7-68.4
.0054-0.0088
15-24
0.0022-0.002^
5.9-7.8
9
Cyclone
spray
tower
0.85-1.00
0.047-0.082
90.4-95.3
0.00 16-0 .0025
4.3-7.8
10
Spray
cross-flow
packed
scrubber
Not
determined
1.135-0. 157
.1.0120-0.014
32-38
0.29-0.36
0.006-0.09
0.070-0.14
"Plants 1-10 wet6 tested by NAPCA.
Methods of Emission Control
23
-------
§
o
I
13
I
Table 8. SUMMARY OF EMISSION DATA ON PERFORMANCE OF CONTROL EQUIPMENT
IN WET-PROCESS PHOSPHORIC ACID PLANTS*
>lant number
Jollector type
Gaseous and water-soluble paniculate
fluoride entering collector per ton of
PgOg produced, Ib
Gaseous and water-soluble particulate
fluoride emitted from collector per ton
of PgOg produced, Ib
Efficiency, %
Concentration of gaseous and water-
soluble particulate fluoride emitted
from collector.
grain/scf
ppm
11
Venturi scrubber,
water-actuated
2.0
0.26
0.058
167
12
Cyclonic
spray
1.23
84
0.031
87
13
Spray cross-flow
packed scrubber
0.53
0.044
92
0.0032
9
14
Spray cross-now
packed scrubber
77
0-038
99-9
0.0019
5
^Information on plants 11 through 13 acquired through private communication.
O
to
-------
Table 8 (continued). SUMMARY OF EMISSION DATA ON PERFORMANCE
OF CONTROL EQUIPMENT IN WET-PROCESS
PHOSPHORIC ACID PLANTS1
Plant number
Collector type
Total fluoride emitted from collec-
tor per ton of PyQ^ produced, Ib
Concentration of total fluoride
emitted from collector,
"grain/scf
ppm
15
Impingement
0.037
0.0087
24
16
packed tower, two-
stage cyclonic
scrubber, in parallel
0.0073
0.00035
1
alnformation on plants 15 and 16 acquired through questionnaire.
SPRAY CROSS-FLOW PACKED SCRUBBER
Figure 4 illustrates a spray cross-flow packed scrubber. In
theory *" and in practice, the spray cross-flow packed scrubber is the
most satisfactory control device presently available for general use in
wet-process phosphoric acid manufacture, and many new plants and
capital replacements employ this scrubbing principle. This type of
scrubber is used in plants 5, 7, 8, and 10 summarized in Table 7, and
described in the section, "Description of Control Equipment in Plants
Tested by Public Health Service. " The gas streams of a particular
plant can be treated in the spray cross-flow packed scrubber. Those
gas streams that precipitate solids go'into the spray section, and
streams containing mostly hydrogen fluoride go to the packed section,
as does all gas leaving the spray section. The packed section is seldom
more than 3 or 4 feet thick and it is usually set up on edge, with gas
passing through it horizontally. Wash water is poured over the top of
the packing and runs down at right angles to the motion of the gas.
CLEAN GAS
TO STACK
IRRIGATED
PACKING
WATER
Figure 4. Principle of the spray cross-flow packed scrubber.
Methods Ot Emission Control
-------
This tends to wash away any solids that escape precipitation in the
empty spray section of the device and settle on the packing. The
packing itself is usually a light polyethylene structure having many
liquid redistribution points and causing only low pressure drop. Plant-
test results show a 1 - to 8-inch water-pressure drop. It is possible to
irrigate the packing with a high water rate for the first few inches
after the entrance of the gas into the packing to wash away particulates
and precipitates.
Because of its design, this collector tends to operate free from
plugging, and high degrees of fluoride removal can be achieved by its
use. If necessary, the packing can be easily washed or replaced.
Table 8 indicates relatively high performance for this type of collector,
showing that gaseous fluoride emissions from the collector are in the
range of 0. 001 to 0. 014 grains per standard cubic foot and collector
efficiency is 57 to 99.9 percent. It should be noted, however, that the
99. 9 percent collection efficiency was obtained for an extremely high
inlet fluoride loading (i.e. , 3. 9 grains per standard cubic foot of gas).
PACKED TOWER
This device can be designed for any degree of hydrogen fluoride
removal; unfortunately, it is subject to plugging due to precipitation
from some compounds of fluorine, such as the solid reaction products
of SiF4 and water. Development of self-cleaning packing has not yet
been achieved. Plant 16, as shown in Table A-3, had a 5-pound-per»
day fluoride emission rate from this type of scrubber. This datum
illustrates that good scrubbing can sometimes be accomplished with a
packed tower.
VENTURI SCRUBBER
For economic reasons, the manufacturers of wet-process phos *
phoric acid prefer the water-actuated venturi or jet venturi scrubber
rather than the gas-actuated type. The jet venturi is primarily a
device for removal of particulates from gas streams by impaction,
yet it can be effective on soluble gases through absorption in the motive
water. * /
An important reason for using Venturis in wet-process acid ser-
vice, is that they are self-cleaning because of the great force of the
motive water. Thus, they are able to handle fluoride particulates and
gases, other than hydrogen fluoride, in spite of the formation of pre-
cipitates.
The flow of water should be continuous while emissions are enter-
ing the device so that the spray nozzle will not be plugged. Tables 7 and 8
shows that the efficiency range of the jet venturi scrubber is 84 to 96
percent. Gaseous fluoride emissions from this scrubber are in the
range of 0. 0104 to 0.023 grain per standard cubic foot.
26 ' WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
SPRAY TOWER
Spray towers are relatively inexpensive to build and are not a ;
source of much trouble from plugging, if the sprays are carefully
designed. Towers, however, do not usually have enough transfer units
to remove fluorine effectively. The performance of cyclonic spray
towers (one of several types of spray towers), is indicated by data
presented for plants 9 and 12, in Tables 7 and 8. Collection efficien-
cies of this device for gaseous fluorides range from 90 to 95 percent.
Emissions of gaseous fluorides from plant 9 are in the range of 0. 0016
to 0. 0029 grain per standard cubic foot. The overall removal efficiency
of this device for gaseous and water-soluble particulate fluorides for
plant 12 was 84 percent. The concentration of fluoride emission was
0. 031 grain per standard cubic foot.
Some devices that are not true spray towers were tested, (as
reported for plants 1 and 2, in Table A-l). These are rather crude
devices that have mediocre performance resulting from bypassing,
defective water distribution, poor spray drop size, and other factors.
Particulate removal was notably poor in these devices. Efficiency of
gaseous fluoride removal was 57 to 72 percent for the one tower where
both inlet and outlet could be sampled, and the range of gaseous fluoride
emissions was 0. 075 to 0. 090 grain per standard cubic foot.
IMPINGEMENT SCRUBBER
There are several types of scrubbers in this classification, but
the impingement type most commonly used in the fertilizer industry is
the Doyle Scrubber. Results of operating a scrubber of this type are
reported for plant 6 in Table 7.
Gas to be treated contacts the surface of a pool of water at high
velocity, undergoing a reversal in direction. Solids impinge on the
water and are retained, and absorption of fluoride gases is promoted
by the turbulence and by the droplets generated by impact.
Theoretically, one would not expect high absorption efficiency for
gases in this scrubber; however, a better efficiency range than the 15
to 62 percent range indicated for plant 6 should normally be possible.
This particularly poor efficiency was probably due to the abnormally
low inlet fluoride concentrations during these runs. In addition, this
scrubber normally serves a nearby triple-superphosphate plant and
this connection was blanked off during the stack tests of the acid plant.
The resulting gas flow was substantially below design, which would be
expected to contribute to low efficiency.
PERFORMANCE OF CONTROL SYSTEMS
Ten plants were sampled by Public Health Service personnel and
the gases were analyzed for gaseous fluoride. In a few cases, concen-
trations of participates were also determined. Tables A-l and A-2
summarize the control-equipment performances calculated and the
operating data taken. Control-equipment efficiencies have been deter-
mined wherever sampling of both inlet and outlet gases was possible.
Methods of Emission Control 27
-------
Tables 7 and 8 have been developed from the primary dcLtcL in
Tables A-l, A-2, and A-3, and summarize the performance of the
collectors used. The Public Health Service tests were mainly concerned
with collector performance and the efficiency of fluoride removal. For
a more detailed description of the fluoride scrubbers tested, refer to the
'section entitled "Description of Control Equipment in Plants Tested by
Public Health Service. If
The more efficient types of scrubbers are represented in Table 7
by those installed in plants 3 through 10. Particulates are reported
only for plants 1, 2, 3, and 10 because so little particulate matter was
found that testing for it was discontinued. For example, the venturi
scrubber in plant 3 is a highly efficient type and chemical analysis of
various samples of the exit gas from this scrubber varied from
no detectable particulates to 0. 029 pound per ton of P^Ot produced, or
about 0. 009 grain per standard cubic foot (Table A-l). Analyses of
scrubber exit gases in plants 4 through 9 gave similar negligible values
in every case sampled. A few of these scrubbers were not adapted to
particulate sampling, because isokinetic sampling was made impos
sible by equipment geometry and piping arrangement. Therefore, no
particulate fluoride results can be given for plants 4 through 9. Because
plant 10 showed 40 percent opacity of the stack plume, read at time
of the sampling, particulate samples were taken from the stack.
Table 7 verifies the visual evidence given by the stack-plume opacity
by showing a range of 0.29 to 0. 36 pound of total particulates in the
plume per ton P2^5 produced, or a concentration of 0. 025 to 0. 031
grain per standard cubic foot.
Data in Table 7 show a large variation in insoluble particulate
fluorides, in the range of 0 to 0.09 pound of fluoride per ton of P^Oe,
produced. This is because the analytical chemical methods determine
total particulate fluoride and soluble particulate fluoride directly;
insoluble particulate fluoride is then calculated by difference. Because
minuend and subtrahend happen to be nearly equal, subtraction gives
a small result and the variation shown for insoluble particulate fluoride
values is due to small differences between two relatively large numbers,
In plant 3, the weak phosphoric acid from the digester is concen-
trated by direct contact evaporation or submerged combustion using
hot combustion gases produced by burning hydrocarbon off-gases. The
concentrator off-gases are fed through a spray chamber, two impinge-
ment scrubbers in series, two venturi fume scrubbers in series, and
a cyclonic spray scrubber before being discharged into the stack. None
of these control devices are noted in the tables. Originally, only the
spray chamber and venturi scrubbers were installed, but these had so
little effect on the acid fog formed in the concentrator that the other
items were added. In spite of'the presence of these several evapora-
tor emission abatement scrubbers, company stack tests reported
emissions of 250 pounds of P-Oc per day and 1400 pounds of fluoride
per day. Even assuming reasonable sampling and analytical errors,
it seems clear that direct contact-combustion-gas evaporation of phos-
phoric acid produces stubborn fogs that pass through most scrubbing
WET-PROCESS PHOSPHORIC ACID EMSSIONS
-------
equipment essentially unaltered. This is one of the reasons that such
evaporation systems are seldom built today. Instead, closed vacuum
evaporators in two or more stages are commonly used to concentrate
phosphoric acid. The Public Health Service stack-gas tests made at
plant 3 were done at the digester-off-gas scrubber, and not at the
evaporator stack.
At plant 6, concentration of fluoride in the inlet gases was low
(about 0.005 grain per standard cubic foot) making the scrubber appear
inefficient. The same comment applies to plant 8; the type of scrubber
used in this plant can achieve efficiencies of over 99 percent; *"
Concentrations of emissions from various types of scrubbers in
the ten plants tested are given in Table 7 and Appendix Tables A-l, and
A-2. Particulate emissions from scrubbers were 0. 0 to 0.50 pound
per ton of P^Oj produced. Gaseous fluoride emissions from most of
these ten plants were in the range of 0.006 to 0. 17 pound per ton of
P,0e produced.
Sulfur dioxide was detected in gases from the reactor, and from
filters, and vents of plant 7. Three analyses were made of the stack
gas after scrubbing and a concentration of about 13 parts per million
was found. For plant 8, stack concentrations were 6 to 18 parts per
million and for plant 9 the range of stack concentrations was 1 to 2
parts per million. These correspond to average SO2 emissions of 87
and 6 pounds per day for plants 8 and 9, respectively. The origin of
this sulfur dioxide is not clear; perhaps it comes from reduction of the
sulfuric acid by organic material of the phosphate rock, or even from
dissolved SO2 in the sulfuric acid itself.
Table A-3 gives some fluoride-scrubber performance data from
additional sources. Information for plants 11 through 14 was obtained
through private communication. Plants 13 and 14 indicate the degree
to which the rate of untreated emissions depends on processing. Thus,
plants 13 and 14 use different processes and generate respectively 240
and 34, 600 pounds per day of particulate fluorides to send to their
scrubbers. Also, they show again that high scrubber efficiency is more
likely to be obtained if the inlet fluoride concentration is high. Both
scrubbers were designed to reduce the fluoride emissions to 0. 01 ton
per day. They have been quite successful in meeting design specifica-
tions , as have several other designs of this type.
Data for plants 15 and 16 were obtained in response to a question-
naire. Reported concentrations were determined by sampling and
analysis by plant staffs.
Plant 15 was sampled downstream from the scrubbers. Fluoride
analyses were by the Willard and Winter method. Circular, horizontal,
tilting-pan filters are used in this plant. Samples from the filter were
taken using an experimental hood over the slurry charge area on one
filter and extrapolating data to the total plant filter area. A baghouse
is used to control rock dust to a design value of 0.002 grain per stand-
ard cubic foot but no measurements are available on performance.
Methods of Emission Control 29
-------
Plant 16 has two trains: one served by a cyclonic scrubber, and
one by a packed tower. The tilting-pan filter is hooded and vented to
the system served by the cyclonic scrubber and both scrubbers are
vented to a common 120-foot stack from which the gas samples were
taken. The stack was sampled at the top and the modified Willard and
Winter distillation method was u'sed to determine fluorine content.
Table A-4 is derived from a paper by Huffstutler and Starnes.
Certain of their data are presented, but differences between plant
capacity and production rate at times of their tests are not stated. The
given emission values, in pounds per hour, are based on tests by the
'Florida State Board of Health and by the companies or their consulting
firms. Company samples and those of the State Board were not taken
at the same time.
The emission rates presented in Table A-4 agree reasonably well
in magnitude with the results of plants tested in this MCA-PHS study,
summarized in Table 7. Actually, many of the control units that were
measured to make Table A-4 have been improved or replaced, and
current emissions from the same plants probably now contain less
fluorine.
DESCRIPTION OF CONTROL EQUIPMENT IN PLANTS TESTED BY
PUBLIC HEALTH" SERVICE
Plant No. 1 - The digester collector used in this plant is a water-
spray chamber 4 feet long, 3 feet wide, and 5 feet high. A center
baffle extends from the top to within 6 inches of the bottom of the cham-
ber. Gases enter the top of the first compartment, pass under the
baffle and exit through the side of the second compartment. Water
sprays are provided on 3 sides of the inlet compartment and on top of
the discharge compartment. The filters have a hood which collects
vapors and discharges them to the atmosphere through the roof of the
building.
Plant No. 2 The collector used here treats fluoride-containing
gases from the digesters. The collector itself is a duct, 100 feet long
and 4 feet square. Gases enter this duct at several points from 3 of 4
digesters. Eight 3/4-inch water-spray nozzles are provided in the top
of the collector duct and five in the sides. Water-flow rate could not
be determined.
Plant No. 3 The off-gases from two digesters pass to a water-
actuated venturi scrubber and then to the stack. The scrubber dis-
charge chamber is 6 feet . 6 inches in diameter and 6 feet in height.
The venturi scrubber is actuated by a pump rated at 400 gallons per
minute, 46 pounds per square inch gauge, and 25 horsepower. This
plant uses only spent acid for digesting the phosphate rock.
Plant No. 4 This plant is a one-reactor, Prayon unit. Gases
from the digester pass to a water-actuated venturi scrubber. This
discharges to a closed tank, then to a stack via a fan rated at appTOXl-
3'0 WET-PROCESSPHOSPHORIC ACID EMISSIONS
-------
mately 1,600 cubic feet per minute. The weak-acid holding tank is also
vented to the duct leading to the scrubber. The collector is a 3-fOOt~
by-14-foot venturi eductor scrubber, designed for a pressure drop of
2.35 inches of water at a water rate of 475 gallons per minute at 90
pounds per square inch gauge.
Plant No. 5 - The fluoride scrubber is a spray cross-flow, packed
unit 9 feet wide, 10 feet high, and 30 feet long. There are three cham-
bers of sprays with wood baffling and a 7-foot-lo-inch section of
polyethylene ring packing. Gases from the digesters enter the spray
section through a 30-inch plastic duct. Gases also enter the packed
section separately, from the filter, filter feed box, filter flash column
seal tank, filter seal pump, the 22-percent-acid feed box, and the sump.
These are combined for entry through a 26-inch diameter duct. Design
water rate is 1000 gallons per minute for a gas pressure drop of 5. 8
inches of water at 380 feet per minute superficial gas velocity. The
blower is rated at 40, 000 cubic feet per minute and the stack has an
inside diameter of 4 feet.
Plant No. 6 - There are two reaction lines. Discharge from each
reactor is fed to the scrubbing system by a 26-inch diameter duct. The
scrubbing system consists of two impingement scrubbers in series.
Each scrubber is 10-1/2 feet wide, 13 feet long, and 12 feet high.
Pressure drop averages about 9 inches of water for the first scrubber
and 7 inches of water for the second. Effluent from the scrubber
passes to a stack 4 feet 2 inches in diameter.
The above scrubbers also serve part of a nearby triple-SUperpllOS -
phate plant. This connection was closed off during the test and sampling
program.
Plant No. 7 Fluoride-containing gases are collected in a spray
cross-flow, packed scrubber 9 feet wide, 9 feet high, and 42 feet long.
The packed section contains 3 feet of polyethylene rings. Pond water
is recirculated to the scrubber at a rate of about 800 gallons per minute
at 60 pounds per square inch gauge. Gases are collected from digesters
filters, and sump tanks.
Plant No. 8 - Fluoride emissions are collected from the digesters
and from the filter. The collector is a spray cross-flow, packed
scrubber, 10 feet high, 11 feet wide, and 33 feet 7 inches long. The
packed section contains 6 feet of polyethylene rings. About 1000
gallons per minute of pond water is recirculated at about 70 pounds per
square inch gauge. Pressure drop through the scrubber is approxi-
mately 8 inches of water for a gas flow of approximately 40, 000 cubic
feet per minute at a superficial velocity of 360 feet per minute.
Plant No. 9 « Gases from the digesters, filter, evaporator hot
well, clarification tanks, and 54-percent-acid storage tank are sent to
a spray tower scrubber that is 40 feet tall and has an inside diameter
of 10 feet. The gas enters the top of the tower tangentially, passes
downward past three banks of water spray nozzles, and then through
Methods of Emission Control 31
-------
about 8 inches of polyethylene packing at the tower bottom, to agglo-
merate small drops of water. The scrubber pressure drop is about
8 inches of water at a pond-water recirculation rate of 900 gallons per
minute at 60 pounds per square inch gauge. Gas rate is around 30,000
standard cubic feet per minute and the exhaust stack from scrubber is
33 inches in diameter. Gas leaving the bottom of the scrubber passes
tangentially into the base of the stack where some additional drops of
water are removed.
Plant No. 10 - This collector is a spray cross.-flow, packed scrubber,
6 feet wide, 7 feet high, and 31 feet long. Gases from the digester
are fed to the spray section and combined emissions from filter feed
tank, filter, and 22-percent-acid mixing box go to the packed section.
A third line vents the sump tank, vacuum-scrubber seal tank, hot well,
and filter seal tank indirectly to the filter. There is approximately
174 cubic feet of polyethylene-ring packing and the scrubber pressure
drop is somewhat greater than 1 inch of water at a water rate of 420
gallons per minute and superficial gas velocity of 280 feet per minute.
Because of the tight piping arrangement, isokinetic sampling before
the scrubber was impossible and tests and samples were run only on
the 30-inch-diameter, scrubber-outlet stack.
HANDLING OF SCRUBBER WATER
New plants can be designed to control fluoride emissions by
incorporating closed process procedures where possible and by scrub-
bing the gas streams from those points where emissions occur.
Because effective scrubbing of the large volumes of gas usually involved
requires substantial quantities of water, it is common practice to use
a large storage pond from which water is recycled to the scrubbers.
Washed gypsum from the filters is also sent to this pond. Residual
phosphoric acid in the gypsum slurry tends to reduce the pH, as does
hydrogen fluoride, (including that produced by the decomposition of
silicon tetrafluoride) picked up by the scrubbers. While a low pH can
be expected to increase the vapor pressure of hydrogen fluoride,
thereby promoting the release of this gas to the atmosphere, this
tendency is opposed by the presence of soluble calcium in the gypsum
pond which will react with the fluoride to form the highly insoluble
calcium fluoride. Because calcium sulfate is many times more soluble
than calcium fluoride and becomes even more soluble as the pH is
lowered, there is always an excess of soluble calcium in the gypsum
pond. The vapor pressure of residual hydrogen fluoride can be further
reduced by raising pH, for example, by adding hydrated lime or lime-
stone, however, this procedure is not generally practiced in the wet-
process acid industry.
CURRENT AND FUTURE AIR POLLUTION POTENTIAL
Over 400, 000 tons of fluorine was present in the phosphate rock
consumed in making wet-process acid in 1966. Theoretically, the
wet-process acid production of 1966 could have released about 200,000
tons of fluoride into the atmosphere of the United States. Actually,
32 WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
because of the extensive use of scrubbers, the amount released was
substantially smaller.
Much attention is now being given to closed designs, better collec-
tion systems, and improved mass-transfer design, both for existing
plants and for new construction. Emission sources varying greatly
in fluoride concentration are treated by different scrubbers: dilute
fluoride concentrations are sent to scrubbers with low liquid concentra-
tions and large numbers of transfer units; higher fluoride concentrations
are sent to scrubbers with high liquid concentrations and one or two
transfer units.
Methods of Emission Control 33
-------
Page Intentionally Blank
-------
SUMMARY OF SAMPLING
AND ANALYTICAL TECHNIQUES
Stack-sampling and analytical work for the joint MCA-PHS study
of wet-process phosphoric acid manufacture was done by the Public
Health Service field test team with Public Health Service laboratory
support. Detailed descriptions of these sampling and analytical tech-
niques are presented in Appendix B.
PARTICULATE FLUORIDES
Particulate matter, as collected by the Public Health Service
sampling train, includes any materials that are solid or liquid at 250°F.
This temperature level is necessary to cause complete reaction between
the hydrogen fluoride in the sample and the silicon dioxide in the glass
probe. It should be noted that particulate matter also includes any
residue left from liquid evaporated at this temperature. Soluble
fluoride particulate matter is that part of the total particulate matter
collected that will dissolve in water. Total particulate fluoride is
determined by acidifying the sample and then distilling the slurry. The
amount of insoluble particulate fluoride is the difference between total
and soluble particulate matter.
At each point sampled for particulate fluoride, pitot tube traverses
were made to determine the velocity profile of the gases in the duct.
Sampling was performed isokinetically at a number of traverse points.
The stack gases were drawn through a sampling train consisting of a
glass probe, cyclone, and glass-fiber filter collection system, heated
to preclude condensation from the sample gas stream and enhance the
reaction of the fluoride gas with the glass. The cyclone collected
particles larger than 5 microns. Particles smaller than 5 microns
were collected on the fine, glass-fiber filter. Particulate fluoride
analysis was done by the Spadns Determination of fluorides.
GASEOUS FLUORIDES
Gaseous fluorides were collected in two different ways. The first
method uses the particulate-matter train as described above, plus four
Greenburg-Smith impingers in an ice bath. Deionized water is used
in the impingers to collect the gas sample. This gas sample will con-
tain any fluorides driven off by evaporation in the heated portion of this
train.
Gas samples were also collected nonisokinetically in a gas train.
This is a much simpler train than the one previously described. A
measured gas volume is pulled through a heated glass probe, a heated
pressure filtration funnel, and then through four midget impingers
using deionized water as the absorbing agent. The midget impingers
35
-------
re used because the sensitivity of the analytical method requires only
small sample. Gaseous fluorides were also analyzed by the Spadns
etermination of fluorides.
WET-PROCESSPHOSPHORIC ACID EMISSIONS
-------
GLOSSARY OF -TERMS
"Be Degrees Baumd (unit of specific gravity)
9 C Degrees centigrade
cfrn Cubic feet per minute
Al' Pressure drop
ft Feet
*F Degrees Fahrenheit
fpni Feet per minute
gprn Gallons per minute
gr 1 grain (7,000 grains = 1 pound)
HP Horsepower
ID Inside diameter
Ib Pound
ppm Parts per million
psia Pounds per square inch absolute
scf Standard c.ubic feet measured
Scfm Standard cubic feet per minute, 60" F and
29. 92 inches Hg
T Short ton (2,000 pounds)
37
-------
CHEMICAL SYMBOLS
Al
A1203
C
Ca
CaO
Ca10{P04)6F2
as
M
ri
'-PROCESS
PHOSPHOE1
o
a
D
S
8
CO
CaS04
CaS04 -
CaSO4 .
Cl
CO2
01-
F
Fe
Fe O
HF
H20
H2S04
H2SiF6
1/2' H20
ZHzO
Aluminum
Aluminum oxide
Carbon
Calcium
Calcium oxide
Apatite (Fluorapatite)
Calcium sulfate, anhydrous
Calcium sulfate, semi-hydrate
Calcium sulfate, dihydrate,
gypsum
Chlorine
Carbon dioxide
Chromium
Fluorine
Iron
Iron oxide, ferric oxide
Hydrogen fluoride
Water
Sulfuric acid
Hydrofluosilicic acid
H3PO4
K2O
Mg
MgO
Na
Na20
NaOH
SiF4
Si(OH)4
SiO2
so2
SOS
Orthophosphoric acid
Potassium oxide
Potassium fluositieate
Magnesium
Magnesium oxide
Sodium
Sodium oxide
Sodium hydroxide
Sodium fluosilicate
Phosphorous pentoxide
Silicon tetrafluoride
Hydrated silica
Silica or silicon dioxide
Sulfur dioxide
Sulfur trioxide
-------
DEFINITIONS
Air contaminant
Apatite
Attack tank
Barometric condenser
Centistokes
Collector
Control device
Crystal nuclei
DAP
Digester
Effluent
Emission
Evaporator
Filter
Dust, fumes, gas, mist, smoke,
vapor, odor, or particulate matter
or any combination thereof present
in the atmosphere.
CaJo(PO4)6 F 2 - The main phos-
phorous bearing component of
phosphate rock.
See reactor.
Device used to condense steam from
vacuum jet. Uses direct contact of
steam with cold water, and 34 foot
water leg to balance atmospheric
pressure and allow water to escape
by gravity.
Centipoise/specific gravity (Table
D-2).
See control device.
One or more pieces of process
equipment used to remove air
pollutants from gas stream.
Small crystals in a reactor, which
furnish sites on which additional
material of the same kind can
deposit. See reactor.
Diammonium phosphate, made by
reacting anhydrous ammonia with
wet-process phosphoric acid.
See reactor.
Waste-gas stream that enters the
atmosphere from the process.
Any gas stream emitted to the
atmosphere.
Unit which concentrates 32 percent
^2^5 a°id, by vacuum evaporation,
submerged combustion or otherwise.
Device to remove calcium sulfate
from dilute phosphoric acid by
forcing the slurry through a cloth
or screen.
Glossary of Terms
39
-------
luor apatite
luorine
ypsum
ypsum pond
npingement scrubber
aunder
SP
j, 32 percent
cn 54 percent
hosphate rock
hosphoric acid
See apatite.
Generic term referring to fluorine
content of any material in a wet-
process phosphoric plant.
Small liquid particles which form
relatively stable aerosols and are
notoriously difficult to collect,
Typical size range, 1 to 100 microns.
Common name for CaS04 . ZH^O,
calcium sulfate dihydrate.
A large pond, commonly unlined,
and used to dispose of gypsum from
the wet-process phosphoric acid
filters. The pond also acts as a
surge for fluoride scrubber water,
which is commonly recycled to the
scrubber s.
A device which impinges a gas at
high velocity onto a liquid surface,
followed by a 180" reversal on exit,
such as the Doyle scrubber.
A channel, usually rectangular, for
gravity conveying slurry from one
reactor to another.
Normal superphosphate, made by
reacting phosphate rock with con-
centrated sulfuric acid. NSP con-
tains about 20 percent
The usual product of the filter in a
wet-process phosphoric acid plant.
This concentration limit is set by
the process used and by economics.
The normal limit of concentration
by evaporation of B2°4 PSOc* Set by
boiling point elevation and economics.
See super phosphoric acid.
The only commercial ore of phosphorus,
widely distributed over the world
and containing many trace impurities.
See apatite.
Hj PO 4., orthophosphoric acid, the
main phosphorus bearing component
of wet-process acid.
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Reactor
Scrubber
Spent acid
Spray tower
Spray cross -flow-
packed scrubber
Stack test
Submerged combustion
Superficial gas
velocity
Superphosphoric acid
Transfer unit
One or more tanks or vessels in
\vhich the reaction between phos
phate rock and sulfuric acid is
carried out to make wet-process
phosphoric acid.
Generic term for any device in
which contaminants are removed
from a gas by contacting with an
absorbing liquid, usually water or
solutions of base or acid.
Sulfuric acid which has been used
for another purpose, but is still
reasonably high in concentration
such as. used nitration acid.
A scrubber for contacting gas with
a spray of water inside a tower.
May have straight line motion or
tangential motion.
A scrubber providing twTo or more
sections to-treat plant gas streams
according to their composition. See
description of control equipment,
for plants 5, 7, 8, 10, and Fig. 4.
Sampling and analysis of any gaseous
effluent which may also contain
participate s.
Actual contact of flame with liquid
by total submergence of the burner.
Used in concentration of wet-process
phosphoric acid, but not common.
Gas velocity in an equipment piece
(such as a packed tower) calculated
as if the equipment piece wTere
empty.
A product of about 70 percent P205,
containing polyphosphoric acids.
Made by burning elemental phos «
phorus in the presence of water or
by evaporating wefrprocess acid in
evaporators of special design.
A number expressing the difficulty
in absorbing a solute from a gas.
It increases with the required degree
.of reduction in solute concentration
and with reduction in the driving
force for absorption. Applied to
particulates, it is the numerical
value of the natural logarithm of
Glossary of Terras
41
-------
the reciprocal of the fraction pass-
ing through the scrubber,
Triple superphosphate, made by
treating phosphate rock with Wet
process phosphoric acid containing
40*-49 percent
Venturi scrubber The jet venturi of this report is a
device furnishing: scrubbing water
as a high velocity jet along the axis
of a ventures throat. This action
causes gas to be drawn into the
venturi, where particulateS are
removed by impact ion and soluble
gases by absorption in the water
droplets.
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
APPENDICES
A. EMISSION AND OPERATING DATA FOR WET-PROCESS
PHOSPHORIC ACID PLANTS
B. SAMPLING AND ANALYTICAL TECHNIQUES
C. WET-PHOSPHORIC ACID ESTABLISHMENTS IN U.S.
D. PHYSICAL DATA
43
-------
APPENDIX A.
EMISSION AND OPERATING DATA FOR
WET-PROCESS PHOSPHORIC ACID PLANTS
The data of Tables A-l and A-2 represent emission data and analyt-
ical results from actual stack samples representing approximately 10
percent of the current number of establishments in the United States.
The stack sampling program was carried out by Public Health Service
personnel. Data for plants 11 through 14 of Table A-3 are from pri-
vate communications. Data for plants 15 and 16 were obtained in
response to a questionnaire submitted to producing plants. Table A-4
is derived from a paper by Huffstutler and Starnes. 17
45
-------
Table A-l. PERFORMANCE OF EMISSION CONTROL EQUIPMENTIN WET-PROCESSPHOSPHORIC ACID PLANTS-
GASEOUSANDPARTICULATEFLUORIDEEMISSIONDATA
Plant number
Plant type
Related capacity, tons/day P^Oc,
Production, tons/day PpOc
Gas scrubber type
'Scrubber water, gpm
Emission source
Test location
Gas temperature. 0 F
Dry gas tatBa, Ecfm
Gaseous fluoride.
Ib/day
Ib/ton Pg^S produced
grain/ scf
Efficiency, %
Particulates (total).
Ib/day
Ib/ton PgOs produced
grain/scf
Efficiency, %
Soluble fluoride particulates.
Ib/day
Ib/to" PgO5 produced
grain/ scf
Efficiency, %
Insoluble fluoride particulates.
Ib/day
Ib/ton PgOs produced
grain/ scf
Efficiency, %
1 1
Chemico
100
107.5
Rectangular
spray chamber
Digester
Spray c
Inlet
136
1.265
0.190
57.
lamber
Ontlet
160
3,675
58.3
0.54
0.082
45.8
0.430
0.064
10.07
0.094
0.014
0.643
0.0060
0.00090
1 1
Chemico
100
107.5
Rectangular
spray chamber
Digester
Spray c
Inlet
202
1.87
0.270
72
lamber
Outlet
160
3.843
56.2
0.52
0.075
;
53.7
0.500
0.072
9.67
0.090
0.0133
0.068
0.0063
0.000091
1
Chemico
100
107.5
Rectangular
spray chamber
Digester
Spray c
Inlet
232
2.16
0.306
amber
Outlet
160
3,877
68
0.63
0.090
70.7
30.0
0.280
0.040
5.34
0.050
0.0071
1
0.841
0.0078
0.0011
1
Chemico
loo
107.5
None
Filter
Filter hood
Ontlet
5,970
5.1
0.047
0.0044
Is
0.170
0.015
0.760
0.0071
0.00065
0.094
0.00087
0.00008-
so
6,604
5.6
0.053
0.0044
6.507
5.7
0.053
0.0045
21.2
0.200
0.017
0.867
0.0081
0.00067
0.027
0.00025
0.000021
22.1
0.210
0.017
0.973
0.0091
0.00077
0.378
0.00350
0.0003
8
Dorr-Ohver
150
156
Square horizontal
spray duct
3 digesters
Spray duct
22.870
15.2
0.097
0.0033
56.8
0.360
5.71
0.036
0.0013
none
Ontlet
76
22.840
15.8
0.101
0.0035
57.8
0.370
0.013
5.37
0.034
0.0012
0.0013
22,070
10.2
0.072
0.0026
73.3
0.470
0.470
1.17
0.0075
0.00027
0.062
0.00040
M
H
§
B
g
g
a
£3
o
55
52
i
en
a60 o F and 29.92 in. Hg.
-------
o.
H'
Table A-l (continued). PERFORMANCE OF EMISSION CON'
'GASEOUS AND PARTICU
Plant number
Plant type
Rated capacity, tons-''day PgOs
Production, tons/day PgOg
Gas scrubber type
Scrubber water, gprri
Emission source
1 est location
Gas temperature. = F
Dry fas iate,a scfm
Gaseous fluoride,
Ib/day
Ib/ton PgOg produced
gradn/scf
Efficiency, %
Partieulates (total),
Ib/day
IMon PgO§ produced
grain/sci1
Efficiency, %
Soluble fluoride particulates.
Ih/day
Ib/ton PgOg produced
graiti/gcf
Efficiency, %
Insoluble fluoride particulates.
today
3b/ton PgO5 produced
grain/ scf
Efficiency. %
Prayon
100
;3:
Venturi, water-actuated
400
Digester
Venturi
Inlet
130
1,987
40,3
0,31
0.104
Outlet
130
1,751
593
4.550
1.550
IOC
10. 1
0.078
0.026
none
detected
tl
0,30
0.0023
0.00088-
97.0
40.6
0.310
0.105
none
detected
100,b
TJOL EOpIPMENf IN WET-PROCESS PHOSPHORIC ACID PLANTS-
r.AIE PLUOBIDI EMISSION DATA
3
Prayon
100
130
Venturi, water-actuated
400
Digester
Venturi
Inlet
138
2,688
39,0
0.30
0.077
Outlet
130
8,146
6.16
0.047
0.0147
84.8
240
1.860
0.478
3,7
0.089
0,0088
98.5
8.0
0,068
0.016
0.37
0.0029
0.00088
95.4
8.85
0.068
0.0175
none
detected
lOpb
3
Prayon
100
130
Vetitnri, water-actuated
400
Digester
VerttUFi
Inlet
135
1,968
27.4
0.211
0.071
Outlet
135.
1,770
3.58
0.0275
0.0104
87.0
1,430
11
3.730
0,306
0.0024
0.00089
99.97
5.12
0.039
0.013
0.307
0.0024
0.00089
94.0
10.6
0.082
0.028
none
detected
100 b
10
Prayon
150
ISO
Spray cross-flow packed
420
Digester, , -filter, --accsssori^
Scrubber
Outlet
108
8,nS7
23.6
0.157
0,014
ins
8,536
23,2
0.155
0.014
102
8,478
£0,8
0.135
0,012
54.8
0.364
0.0312
44.8
0.264
0.0262
44,2
0.294
o.oesg
18.5
0.123
0.011
21.4
0.148
0,013
11.0
0.073
0,0067
8.85
0.059
0.0054
0.84
0.0056
13.7
0.091
0.0083
a60 BF and 29.92 in. Hg.
Essentially complete removal.
-------
Table A-2. GASEOUS FLUORIDE EMISSION DATA FROM WET-PROCESS
PHOSPHORIC ACID PLANTS
Plant number
Plant type
Rated capacity, tons/day PgOc
Production, tons/day Pp^S
Gas scrubber type
Scrubber water, gpm
Emission SOUTC6
Test location
Gas temperature. " J1
Dry gas rate. SCfltl
Gaseous fluoride
Ib/day
Ib/ton PgO^ produced
grain/ ecf
Efficiency, %
4
Prayon one
reactor
450
639
Venturi.
water-actuated
475
Digester
Venturi
Inlet
5.281
398
0.62
0.388
Outlet
84
5.281
20.9
0.033
0.020
94.75
4
Prayon one
reactor
450
639
Venturi.
water-actuated
475
Digester
Venturi
Inlet
5,281
25
0.67
0.414
Outlet
84
5.281
18.2
0.028
0.018
95.7
4
Prayon one
reactor
Venturi,
water- actuated
Digester
Venturi
Inlet
5.281
113
0.49
0.304
Outlel
84
5.281
24.1
0.038
0.023
92.3
L8
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Table A-2 (continued). GASEOUS FLUORIDE EMISSION DATA FROM WET-PROCESS PHOSPHORIC ACID PLANTS
X
>
Plant mimbeff
Plant, type
Rated capacity, tons/day PgO%
Prodlietioxi, tons/day PgOK
Gas scrubber type
Scrubber water, gpm
Emission source
Test, location
Gas temperature, s F
Dry gas rate. Seta
Gaseous fluoride
Ib/day
lb/t.on Pg05 produced
grain/scf
Efficiency, %
5
Prayon
680
100
Separate gas feeds,
spray cross-flow packed
800
Digester
Scrubber
Inlet
145
7,500
41.5
0.089
0.028
Filter
Scrubber r
Inlet
12,231
19.4
0.028
0.0081
Combined
digester
and Ota
Inlet
19,731
60,9
0.087
0.016
Scrubber
Serubbsi
Outlet
90
12.3
0.018
0.009: !
79.6
5
Pmyon
660
700
Separate gas feeds,
spray cross-flow packed
800
Digester
Sarnhhw
Inlnt
145
7,500
38.8
0.056
0,027
Filter
S^mhhfir
Inlet
12,231
16.0
0.023
O.CC67
8S
Combined
Sigester
and Kite*
Inlet.
19,731
54.8
0.078
0.014
^crabber
Scrubber
Outlet
90
5.9
0.0086
0,0015
.3
5
Pmyon
660
700
Separate gas feeds,
spray cross-flow packed
8C0
Olgestar
Scrubber
Inlet
us
7,500
46.6
0.067
0.032
Filter
Scmbber
Inlet
18,231
8,0
0.011
0.0033
9
Combined
digester
and filter
Inlet
19,731
64.6
0.078
0.014
Serubber
&rubber
Outlet
90
416
0.0060
o.ooy
.4
-------
Table A-2 (continued). GASEOUS FLUORIDE EMISSION DATA FROM WET-PROCESS PHOSPHORIC ACID PLANTS
H
0
M
CO
g
3
o
S
o
>
a
3
I
CO
§
Plant number
Plant type
Rated capacity, tons/ day PgOs
Production, tons/ day PgOg
Gas scrubber type
Emission source
Test location
Gas temperature, °F
Dry gas rate, scfm
Gaseous fluoride
Ib/day
Ib/ton PsO5 produced
grain/scf
Efficiency, %
6
Dorr-Oliver
1,200
two lines @ 600
1.080
Doyle
(two in series)
Acid
line
No. 1
Scrubber
Inlet
82
8.300
3.6
0.0022
Acid
line
,No. 2
Scrubber
Inlet
100
8,370
10.8
0.0066
Combined
lines
1 and 2
Inlet
16.670
144
0.0133
0.0044
Scrubber
Scrubber
Outlet
16.670
12.2
ib.ona
0.0037
IB
6
Don-Oliver
1.200
1.080
Acid
line
NO. 1
Scrubber
Inlet
82
8.300
5.4
0.0033
Doyle
(two in series)
Acid
line
No. 2
Scrubber
Inlet
100
8.370
11.6
0.0071
Combined
lines
1 and 2
16.670
17.0
0.0158
0.0052
Slcrubbei
'Scrubber
Outlet
16,670
6.5
0.0060
0.0020
62
6
Dorr-Oliver
1.200
1.080
Doyle
(two in series)
Acid
line
No. 1
S.crubber
82
8.300
3.5
0.0022
Acid
line
No. 2
Scrubber
Inlet
100
8.370
11.6
0.0071
Combined
lines
1 and 2
Inlet
16.670
15.1
0.0140
0.0047
icrubber
Scrubber
Outlet
16.670
7.8
0.0072
0.0024
48
-------
o
1
o.
x°
Table A-2 (continued). GASEOUS FLUORIDE EMISSION DATA FROM WET-PROCESS PHOSPHORIC ACID PLANTS
Plant number
Plant type
Rated capacity, tons/day P£O;J
Production, tons/day PgOs
Gas scrubber type
Scrubber water, gpm
Emission source
Test location
Gas temperature. °F
Dry gas rate, scfrrr1
Gaseous fluoride
Ib/day
Ib/ton Po^S produced
grain/scf
Efficiency, %
Prayon
400
Separate gas feeds,
spray cross-flow packed
Sump
and
vent
Scrubber
Inlet
70
14.100
10.6
0.026
0.0038
800
Digester
Scrubber
Inlet
140
16.200
460
1.15
0.145
Filter
Scrubber
Inlet
Scrubber
Outlet
85
9.700 40,000b
10.9
0.027
.0.0058.
68
0.170
.0.0088,
85.9
7
Prayon
400
400
Separate gas feeds,
spray cross-flow packed
800
Sump
and
vent
Scrubber
Inlet
70
14.100
9.7
0.024
0.0035
;Digester
; Scrubber
Inlet
140
16.200
571
1.43
0.180
89
Filter
Scrubber
Inlet
9.700
12.1
0.030
0.0064
4
Scrubber
Outlet
85
40,000 b
63
0.157
0.0081
Prayon
400
400
Separate gas feeds,
spray cross-flow packed
800
Sump
and
vent
Scrubber
Inlet
70
14.100
16.9
0.042
0.0062
;Digester
i Scrubber
Inlet
140
16,200
548
1.37
0.173
92
Filter
Scrubberr
Inlet
-
9.700
9.6
0.024
0.0051
.7
Scrubber
=^^^=r-
Outlet
^=
85
40,000b
42
0.105
0.0054
^
a60°F and 29.92 in. Hg.
Contains approximately 13 ppm SOg.
-------
Table A-2 (continued). GASEOUS FLUORIDE EMSSIONI DATA FROM WET-PROCESS PHOSPHORIC ACID PLANTS
K
H
'O
a
o
o
O
en
"0
O
2
O
Q
a
w
£
o
z
en
Plant number
Plant type
Rated Capacity, tons/day PpCk
Production, tons/day PgC^
Gas scrubber type
Scrubber water, f pm
Emission source
Test location
Gas temperature, QF
Dry gas rate, scfma
Gaseous fluoride
Ib/day
Ib/ton Pg^5 produced
gram/scf
Efficiency, %
8
Prayon
750
Separate gas feeds.
spray cross-flow packed
960 1,800
Filter
Scrubber
Inlet
95
19,600
8.2
"frTdii
0.0021
Digester
Scrubber
Inlet
150
10,000
30.2
0.040
0.0154
,56.
Combined
digester
and filter
Inlet
38.4
0.051
0.0067
7
Scrubber
Scrubber
Outlet0
95
29.600
16.6
0.022
0.0029
8
Prayon
750
745
Separate gas feeds.
spray cross-flow packed
860 * 1,203
Filter
Scrubber
Inlet
95
19,600
141
0.019
0.0037
Diges ter
Scrubber
Inlet
150
10.000
27.6
0.037
0.014
Combined
1 (digester
and filter'
41.7
0.056
0.0072
Scrubber
Scrubber
Outlet"
95
29,600
13.2
0.0177
0.0023
68.4
8
Prayon
745
Separate gas feeds,
spray cross-flow packed
960 1,200
Filter
Scrubber
Inlet
95
19,600
6.3
0.011
0.0022
Digester
Scrubber
Inlet
150
10.000
28.4
0.038
0.0145
Combined
digester
and filter
Inlet
36.7
0.049
0.0063
Scrubber
Scrubber
Outlet0
95
29.600
12.7
0.0170
0.0022
65.4
a60 = F and 29.92 in. Hg.
"Contains 6-18 pprn SOg.
-------
Table A-2 (continued). GASEOUS FLUORIDE EMISSION DATA FROM WET-PROCESS PHOSPHORIC ACID
&
Plant number
Plant type
Rated capacity, tons/day, ?gO5
Production, tons/day, PjjO^
Gas scrubber type
Scrubber water, gpm
Emission source
Test location
Gas temperature, °F
Dry gas rate, scfm
Gaseous fluoride
Ib/day
Ib/ton ?2°5 produced
grain/set
Efficiency, %
9
Prayon
76. 5
103
Cyclonic spray tower
900
Sump
and
vent
Inlet
88
5,730
25.4 '
0.247 I
0.023
Digester
Tow
Inlet
6.130
59
0.673
0.049
Filter
er
Inlet
90
3.440
3.5
0.034
0.0062
Tower
Outlet1
90
15,300
6.4
0.082
0.0029
so. 4
i
9
Prayon
76.5
103
Cyclonic spray tower
900
Sump
and
vent
Digester Filter Tower
Tower
Inlet
88
5,730
25.0
0.243
0.022
Inlet
6.130
74
0.718
0.062
Inlet
so
3.440
4.3
0.042
0.0064
Outlet1
90
15.300
6.8
0.066
0.0023
93.4
Cyclo
Sump
and
vent
Inlet
88
5.730
26.6
0.258
0.024
Dige
Inl
6.13
70.3
0.68;
0.0:
Ol
Ul
-------
Table A-3. GASEOUS AND TOTAL FLUORIDE EMISSIONSFROM WET-PROCESS
-PHOSPHORIC ACID PLANTS
Plant number
Plant type
Rated capacity, tons/day PgO^
Production, tons/day PgC^
Gas scrubber type
Scrubber water, gpm
Scrubber AP. inches H20
Emission source
Test location
Gas temperature. °F
Wet-gas flow rate, scfm
Total fluoride.
lb/day
lb/ton PgO^
grain/scf
ppm
Gaseous and water-soluble parti Clli
fluoride.
Ib/day
Jb/ton P205
grain/ suf
ppm
Effioiencv. %
ii"
460
Water-actuated venturii
1,040
1
Digester and filter
Scrubber
Inlet
10.000
900
2.0
0.45
1,280
Outlet
120
0.26
0.058
167
87
12"
550
Cyclonic spray
470-550
3-
5
Digester and filter
Scrubber
Inlet
108.000
4.200
7.6
0.19
540
Outlet
680
1.23
0.03
87
84
13"
4BOC
Spray cross-flow packed
690
6-10
Digester and filter
Scrubber
Inlet
30,000
240
0.53
0.039
108
Outlet
20
0.044
0.0032
'
92
. received in private communication.
Data received by questionnaire.
cTwo separate plants, using different processes.
54
WET-PROCESSPHOSPHORIC ACIDEMISSIONS
-------
Table A-3 (continued). GASEOUS AND TOTAL FLUORIDE EMISSIONS FROM WET-PROCESS
PHOSPHORIC ACID PLANTS
^lant number
3lant type
lated capacity, tons/ day pQQ,
'roducti on. tons/day p Q,
}as scrubber type
Scrubber water, gpm
icrubber AP . inches HoO
mission source
'est location
}as temperature, o f
Vet-gas flow rate, scftf)
^otal fluoride.
Ib/day
Ib/to" P2°S
grain/ scf
ppm
jaseous and water-soluble
articulate fluoride,
Ib/day
Ib/ton ?205
grain/ set'
ppm
Efficiency. %
14"
450C
pray cross-flow packec
800
6-10
Digester and filter
Scrubber
Inlet
43.000
34.600
77
3.9
10,000
Outlet
n
0.038
0.0019
5
99.9
«b
Dorr-Oliver multi-tank
900
1,035
npingement None
Digester and filter
Scrubber
Outlet
96
1.170
38
0.037
0.008:?
24
Outlet
95
29,000
80
0.077
0.01
:38
18b
Dorr-Oliver single tank
Train 1 packed tower
Train 2 t two-stage
cyclonic scrubber
Digester
Combined scrubber
Outlet
96
66,000
5
0.0073
0.00035
-
:Data received m private communication.
''Data received by questionnaire.
"'TWO separate plants, using different processes.
Appendix A
55
-------
Table A-4. WET-PROCESS PHOSPHORIC ACID PLANT FLUORIDE
EMISSIONS AFTER CONTROL UNITSl?
Plant
production
capacity, PgOg
tons/day
leactors and filters'5
500
400
300
170
leactors only"
260
200
175
Fluoride efflfssionsa
COHpany reoocted
Ib/hr
0.407
1.000
0.685
0.145
0.740
0.750
0.133
Ib/day
9.8
24.0
16.5
3.5
17.8
18.0
3.2
Ib/ton
PgOs
'V
0.020
0.060
0.055
0.020
0.055
0.090
0.018
Florida State Board
ofHealth reported
Ib/hr
1.750
2.660
0.040
1.090
0.055
0.234
1.700
Ib/day
42.0
64.0
0.96
26.0
1.33
5.6
40.8
Ib/ton
P205
0.084 ,
0.135
0.0032
0.153
0.0051
', 0,028 '
;0.23
'Gaseous fluorides and water-soluble particulate fluoride only.
'Controlled sources.
56
WET-PROCESSPHOSPHORIC ACID EMISSIONS
-------
APPENDIX B.
SAMPLING AND ANALYTICAL TECHNIQUES
INTRODUCTION
The sampling equipment for this study was constructed by the Public
Health Service for the specific task of measuring air pollutant emissions
at their sources. The following description of the apparatus is general.
The reader is referred to APCA Journal, 18| 1 );1Z ""14 for a more com-
plete discussion of stack-gas testing. The bag ic measurements per-
formed during this sampling were of effluent flow, and gaseous and
particulate fluoride concentrations.
The overall, source-testing procedure may be divided into three
major phases: preliminary survey, field sampling, and laboratory
analysis.
Preliminary surveys were done well in advance of the actual tests.
The purpose was to determine which type of pollutant to measure and
to arrange the logistics involved in conducting a source test. The
source test itself was composed of several components, including set-
up for operation, and sample clean-up. Since the sampling trains
differ, gaseous and particulate sampling will be discussed separately.
THEORY OF SAMPLING TRAIN DESIGN
There are two types of sampling trains - gaseous and particulate
and each contains a heated and a cooled section. The glass probes and
filtering elements of both trains are electrically heated to 250" F.
Both trains have gas impingement systems that are cooled by ice bath.
These two systems are similar for distinct chemical reasons.
The glass probe is heated to 250" F to cause reaction to occur
between the hydrogen fluoride in the sample-gas stream and the silicon
dioxide in the glass walls of the probe. The reaction is:
4HF t S1Q2 > ^iF4 (g) + 2H20(g)
Because further reaction occurs "in cool water, the filtering element
is also heated to 250" F in order to prevent water condensing on the
filter and clogging the pores of the paper. The heated probe and filter
arrangement also prevent the hydrogen fluoride ga s from reacting with
the filter or filtered media.
Both trains use gas impingers with a collecting medium of chilled
water. Water is used as the absorbing agent because fluorides are
highly soluble in water. The gases pass into this ice-bath-cooled
section and the silicon tetrafluoride gas hydrolyzes in the water to the
-------
final stable products, soluble fluosilicic acid and slightly soluble ortho-
silicic acid. The reaction is:
SiF4 t 41^0-2 H2 SiFfct Si (OH) 4
H20 t SiO2
The formation of silicon dioxide, a gelantinous precipitate, makes it
necessary to remove the nozzle tip from the first impinger. If this
were not done, the nozzle tip could become clogged after a few minutes
of sampling.
PARTICULATE-MA'M'ER SAMPLING PROCEDURE
The sampling tests are performed isokinetically along a represent-
ative traverse of the stack. The equipment, shown in Figures B-l,
B-2, and B-3, is assembled as shown with the heated box, ice bath,
and glassware designed to move with the probe. The probe is kept
Figure B-1. Sample box with pilot tube, impingers, and umbilical cord.
Figure B-2. Meter boxcontrols.
58
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
sufficiently hot to avoid condensation and to cause the fluoride-silica
reaction to occur. The remainder of the equipment is placed at some
convenient location and connected by the rubber umbilical hose. A
probe tip is selected so that isokinetic sampling will be maintained at
approximately 0. 75 cubic foot per minute. This flow rate approximates
the design flow rate through cyclone and Greenburg-Smith impingers.
Thus, a flow rate of 0. 75 cubic foot per minute provides an efficient
separation of gaseous and particulate pollutants.
Adjustment to isokinetic conditions is accomplished by use of a
needle valve and a bypass gate valve. At any instant, the dry-gas
sampling rate may be determined from the inclined-vertical manometer,
which is connected across a calibrated orifice. A conversion from the
dry-gas sampling rate to the total-gas sampling rate is made, by
correcting for moisture condensed and absorbed from the stack effluent
by the impingers and silica gel; or generally from the preliminary wet-
and dry-bulb measurements. In the case of wet-process phosphoric
acid manufacture, fluorine compounds found in the stack gas are not a
significant part of the gas volume sampled and need not be considered
in calculating total volume.
The thermometer in the cap of the outlet of the fourth impinger
indicates the ice-bath efficiency. This temperature is important, in
that if it goes above 70" F the ice bath is no longer serving its function
and all of the moisture may not be removed. The volume of gas sam-
pled at each point of the traverse is determined by reference to the
indicated values on the dry-gas meter, to assure that the calibrated
orifice is operating properly. The temperature of the dry gas in the
meter is obtained by averaging the meter inlet and outlet thermometer
temperatures. This temperature is necessary to calculate the gclS -
sample volume at standard conditions of temperature and pressure.
Particulate-Matter Sampling Apparatus
The particulate-matter sampling apparatus (shown in Figure B-3)
consists of a probe, a cyclone, a filter, four Greenburg-Smith impingers
a flowmeter, a manometer, a dry-gas meter, and an air pump. The
stainless steel, button-hook-type probe tip {!)* is drawn to 5/S inch so
that it will connect, by a stainless steel coupling (2) with a Viton "0"
ring bushing, to the probe (3). The probe (3) is fabricated of 5/8 inch,
medium-wall, Pyrex glass tube with a 28/12 ball joint on one end. The
glass probe is wound with 25 feet of 26-guage Nichrome wire. The
Nichrome-wound glass tube is wrapped with a fiber-glass tape, and
during the sampling the Nichrome wire is connected to a variable auto-
transformer so that the amount of heat transmitted to the probe can be
controlled. The wire-wound probe is encased in a 1-inch stainless
steel tube. The front end of the tube has a nut welded to it for COUnGC-
tion to the stainless steel coupling and nozzle tip. The probe connects
to a cyclone and flask (4). The cyclone is described in detail in
Reference 5, except for the 28/12 female ball joint on the arm. The
in parentheses refer to numbered parts of Figure B-3.
Appendix B 59
-------
| ^f_ _L J 9 10 U_
Figure B-3. Particulate sampling train.
cyclone is designed to provide a particulate separation with a size cut
of 5 microns. It is connected to a fritted'glelSS filter (5) which holds
a 2-1/2 inch, Number 41 Whatman filter paper. 'The cyclone, flask,
and filter are contained in an electrically heated enclosed box. The
flow of sample gas leaves the heated-particulate-filtering system and
passes into the ice-bath-cooled, gas-impingment section. The first
impinger (8) of this system is of the Greenburg-Smith design, modified
by replacing the orifice plate tip with a 1/Z-inch I D glass tube extending
to one-half inch from the flask bottom.
This impinger is filled with 250 milliliters of deionized water. The
second impinger (9) is a standard Greenburg-Smith impinger filled with
150 milliliters of deionized water. The third impinger (10) is a Green-
burg-Smith impinger modified like the first. This impinger is left
dry to collect any entrainment. The fourth impinger (11) is also a
Greenburg-Smith impinger modified like the first. This impinger con-
tains approximately 175 grams of accurately-weighed dry silica gel.
From the fourth impinger (11), the effluent stream flows through a
check valve (13) to a flexible rubber vacuum tubing (14). The sample
gas goes through a needle valve (16) and then a vacuum pump (17),
rated at 4 cubic feet per minute at 0 inches of mercury gauge pressure,
which is in parallel with a bypass gate valve (18); a dry-gas test meter
(lo), with a scale of 1 cubic foot per revolution is used to record the
volume sampled. The three thermometers (12) are dial type with a
range from 25" to 125" F and having a 5"inch stem. The vacuum gauge
(15) is calibrated from 0 through 30 inches of mercury. The
60
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
meters (21) across the calibrated orifice (20) and pitotmeter (22) are
the inclined-vertical type, graduated in hundreths of an inch of water
from 0 to 1. 0 inch, and in tenths from 1 to 10 inches.
DISCUSSION OF GAS SAMPLING
The gas sampling train, since it does not sample isokinetically is
much simpler in operation and theory. There is no concentration
gradient of gaseous fluorides in. a well-mixed stack gas, so the probe
may be held stationary. The fact that no traversing is required allows
the probe impingers and ice bath to be securely fastened to the wall of
the stack. This is done with an ell-shaped platform which is designed
to be strapped to the flue. The sample-gas rate is metered at another,
more convenient location. This gas flow is carried in a vacuum-hose
umbilical cord which also carries the wires that supply electricity to
heat the probe and filtering elements. The midget impingers that are
-used were" designed to absorb gases most efficiently at a sampling rate
of 0. 1 cubic foot per minute. The sample flow rate is accordingly
maintained near'this value by checking a small air rotameter, and
making adjustments as needed. Emission rates are generally expressed
on a dry basis. This is done to prevent changes in scrubber-gas mois-
ture content from altering the control efficiency of the unit. The stack
moisture is determined from wet- and dry-bulb temperature measure-
ments which are made during the preliminary test. This is also the
time at which the velocity traverse of the stack is made. The traverse
is done with the most accurate methods available using the technique
described in Los Angeles Source Testing Manual.
In the actual procedure for gas sampling, the amount of sample gas
withdrawn from the flue is not critical. The major considerations are
that an accurately measured amount of fluoride be absorbed in the
impinger water and that there be enough time to provide an average
response in the process. Hence, the samples are generally taken over
a 15-minute period.
Apparatus for Gas Sampling
The apparatus for gaseous fluoride sampling (Figures B-4, B-5) is
considerably simpler and more portable than the particulate train.
The gas train is composed of a probe (1)*, filter (4), impingers (6, 7,
8, 9), pump (13), rotometer (15), and dry-gas test meter (17). More
specifically, the sample is first drawn in through a two-foot medium-
wall Pyrex glass probe (1) which is wound with 18 feet of 26-guage
Nichrome heating wire (2). This wire is connected to a variable tra.HS -
former (3) which allows the voltage and hence the heat input to be con-
trolled. The wire is covered with a fiber glass insulation tape and
placed in a stainless steel probe sheath.
A Gelman #4300 pressure filtration funnel (4) is next in line. It
is wire-wrapped, as the probe is, and contains a 1-inch, WhS-tlTL&n
#41 filter paper. It is in this manner that the sample gas is heated so
*Number in parentheses refer to numbered parts of Figure B-5.
Appendix B 61
-------
Figure B-4. Gas sampling train.
16
it
n
i
n
7
n
B
^
i
15
Figure B-5. Schematic of sampling apparatus.
62
WET-PROCESS PHOSPHORIC ACID EMISSIONS'
-------
that moisture will not clog the filter and the gaseous fluorides will
react with the glass of the probe to form a water-soluble compound
-(silicon tetrafluoride). A three-way valve (5) is used to purge the line
before the actual sampling begins. This valve also allows the ice bath
(10) and impinger system to be sealed off and removed after the run is
completed. The heated sample gas now enters the impinger system
where the soluble fluorides are scrubbed out. Due to the sensitivity of
the analytical method only a small sample is required, therefore, four
midget impinger s are used. The first three impinger s (6, 7, 8) con-
tain approximately 15 milliliters of distilled water with the nozzle tip
of the first impinger removed to prevent silicon dioxide (a byproduct
of the glass-fluoride reaction) from clogging the opening. The fourth
impinger (9) is left dry and is used to collect any entrained water. The
scrubbed gas flows from the collecting media through an umbilical cord
to a silica-gel-packed drying tube (12) which removes the moisture
from the sample gas. A diaphragm pump (13) is used to pull the gas
through this system. From the pump the gases pass through a grOSS -
flow control rotometer (15) with the needle valve (14) of the rotcmeter
normally set to maintain a sampling rate at 0. 1 cubic foot per minute.
The actual sample volume is read from adry-gas meter (17) with an
accuracy of plus and minus 1 percent. The dry-gas meter is fitted
with thermometers (16) on the meter inlet and outlet sides. These
thermometers give the gas temperature inside the meter, allowing
correction of the sample volume to standard conditions.
CONSIDERATIONS COMMON TO GASEOUS AND PARTICIPATE
SAMPLING
Selection of Sampling Points
The locations and number of sampling points are based on size and
shape of the duct, uniformity of gas flow, availability of sampling port,
and space required to set up sampling equipment. Straight vertical
ducts with no flow obstructions for at least eight diameters upstream
and two diameters downstream of the sampling point are preferred.
To insure a representative sample of stack gas, the duct should
be divided into a number of equal areas and sampled at the center of
each of these areas. The number of areas depends on the size of the
stack. It is also desirable to sample across the largest dimension of
the stack. Horizontal flues should be sampled in the vertical direction
to prevent erroneous results due to stratification of the particulates in
the duct. The number of areas into which the duct area was divided
for the sampling was decided on the basic of criteria discussed in
Western Precipitation Company's Bulletin WP-50.
Sampling Time and Equipment Cost
The time necessary to perform the series of triplicate tests at
each point is determined largely by engineering ingenuity. However,
through the use of packaged sampling equipment, designed on the basis
of a thorough preliminary survey, the time required for the field
assembly of sampling equipment can be greatly reduced. Thus, it
Appendix B 63
-------
should rarely take over two man-days for a sample point to be tested
from start to finish.
The cost of specific particulate-matter sampling equipment, enough
material and apparatus for three replicate tests, would be approximately
S3,000. But, for approximately another SI,000, additional equipment
could be purchased so that virtually all types of particulate matter and
acid mists could be collected.
The gas sampling train is considerably less expensive. Necessary
train components and glassware for three runs should not cost more
than S400.
Field Calculations
The mathematical development of the field calculations necessary
for the proper operation of this isokinetic, particulate sampler will be
discussed in this section. The gaseous train requires no field calcula-
tions as no adjustments have to be made once the flow is adjusted to
0. 1 cubic foot per minute. The particulate train, on the other hand,
must be able to sample at various volumetric rates depending on
changes in the stack-gas velocity. The following material explains how
and why these rate changes are made.
Isokinetic sampling requires that the sampling velocity through the
nozzle be equal to the effluent velocity in the stack. The nozzle velocity
is determined from the volumetric sampling rate. Both stack velocity
(measured by the pitometer) and volumetric sampling rate (measured
by the calibrated-orifice) are indicated by manometer pressure differences.
For isokinetic sampling, the calibrated-orifice manometer is made
dependent on the pitometer-manometer by combining the pitometer
equation;
v - r
vp - cp
with an equation relating volumetric sampling rate to effluent velocity
through the nozzle:
0= ^D2 Vp .Mc (2)
where: C " pitometer calibration factor, dimensionless
g^ = gravitational constant, 32. 17 (lb,,,,) (ft)/(lbforce) (sec2)
AP = pitometer pressure differential, (Ibf nrce)/ (^ )
R = gas-law constant, 1545 (ft) (lbforce)/( °R) (Ib-mole)
T = stack gas temperature, °R
Ps = stack gas pressure, {Ibf Orce)/(ft )
M s = effluent molecular weight, (lbmas 3 ) / (l"b ~rnole )
Orn = volumetric flow rate, f t-^ / S e C
D = nozzle diameter, ft
Vp = stack velocity at traverse point, ft/sec
Tm = effluent temperature at the meter, R
6 4 WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Pm = meter pressure,
Mc = mole fraction of dry gas in stack effluent, dimensionles:
When a calibrated orifice is used to measure
V5
TmAH2gc R
JA
Pm Mm
where: J -l orifice coe f f i cient, dimensionless
A - orifice area, ft
AH - orifice pressure di f f e rential, ( Ibf Q
Mm = dry effluent molecular weight, (lt>mas s) I (lb-mole) .
Combining and rearranging Equations, 1, 2, and 3 give the dependency
of the calibrated-orifice manometer on the pitometer-manometer,
A H = K (AP)D4 Ts -1 (4)
Mm Ps
where: K - -- Tm (5)
Stack-gas velocities not only change between different traverse
points, but also vary at a given point because of variable flow condi-
tions t The calculation of Equation 4 presents an undesirable time lag
between changes in stack velocity and sampling rate. In addition,
frequent errors are made when calculations are attempted under the
stress of field sampling conditions. The net result is deviation from
isokinetic sampling.
A three-independent-variable nomograph (Figure B-6) was con-
structed to represent Equation 5 and reduces calculation time to a few
seconds. The K term in Equation 4 is usually a constant during sam-
pling, but may change for different sampling locations or processes.
A four-independent-variable nomograph could be used; but, because
this would make the nomograph that much larger and unwieldy and
because K does not frequently vary, K is incorporated into the T scale.
This is done by making the T scale movable and Setting its position
with a C scale. The C scale is a ratio of the true value of K to an
assumed va 1 u e . If the values for K vary from the assumed values, then
a new value for C is obtained from a second nomograph (Figure B-7).
The nomograph of Figure B-6 is based on the assumption that the
stack gas is 5 percent water, but that no water passes through the
orifice. It assumes a dry-gas molecular weight of 29, atmospheric
and stack preSSUTGS of 29. 92 inches of mercury, a meter temperature
of 70" and a ^Ho (the orifice pressure differential that gives 0. 75
cubic foot per minute for dry air at 70" F and 29. 92 in. Hg) of 1.84.
The nomograph of Figure B-7 corrects for different stack and atmos-
pheric pressures, different stack- ga s moisture contents, and di f f e rent
gas temperatures at the orifice. Figure B-7 does not correct for
moisture in the gas passing through the orifice nor molecular weight
Appendix B 65
-------
ORIFICE READING
AH
10
9 -
8
7
6
5
4
3 =
2'
I1 a
I.t-Ij
0.8 |
1.7 |
I.I I
~^
1.5 |
I
=
.3'^
1.2
.1 __
REIF.
-2.0
I c
.1.5
I CORRECT/ON
FACTOR
1.0
-0.9
l|- 0.8
07
, I
REF .
STACK
- 0.5 TEMPERA TOfti
2500
2000
1500
1000
800
600
500
400
300
200
100
0
E T,
~
-
;
1
r
i-
=-
=
0.001
K FACTOR MIOr READ|N
-------
a
a
I
R
Ed
3.0
2.0
1.5
REF 1
. H20
1.0
100
50-
-50-
REF 2
10-
20-
30
EXAMPLE:
H20
= 2.7 in. H20
= 0-F
= 30
= 1.1
40
FIND C
1.0
0.9
0.8
DRAW LINE FROMAHg TO Tm TO OBTAIN POINT A ON REF. 1.
DRAW LINE FROM POINT A TO % H20 AND READ B ON REF. 2
DRAW LINE FROM POINT B TO Ps/Pm, AND OBTAIN ANSWER OF 0.85 FOR C.
-------
anges other than those due to water in the stack gas.
Directions for the use of the nomograph (Figure B-7) are as follows:
Prior to sampling
1. Obtain C from Figure B-7, and set the T scale.
2. Make a rough preliminary pitot traverse, and determine the
minimum, average, and maximum AP.
3. Measure approximately the stack temperature, T.
4. Align T and the AP's from step 1, and choose a convenient
nozzle diameter, D.
5. Align T and D to obtain a AP.
6. Align the P from step 5 and the reference point on the H line
to obtain a K factor setting.
7. Keep this K factor setting as a pivot.
During sampling
8. Determine AH for the AP's of the pitot traverse.
9. If T changes, repeat steps 3 through 8. *
The nomograph calculates isokinetic conditions for an average AH
1. 84. This A Ha should correspond to a flow rate of about 0. 75 cubic
)t per minute if the orifice plate is 0. 18 inch in diameter ir( a 0. 5-
;h I.D. tube with pressure taps 1 inch on either side of the orifice
ites.
For the details of constructing nomographs the reader should con-
It other references, such as "Chemical Engineers Handbook, " New
irk: McGraw-Hill Book Co., Inc., 1950.
tripling Cleanup
This section discusses the step by step method employed to trans-
the sample from the trains to storage containers. It is written in
3 parts, one covering the particulate-matter train, the other the
ieous train.
rticulate-matter train cleanup - It is necessary that proper care be
jrcised in moving the collection train from the test site to the clean-
area so that none of the collected sample is lost and so that no out-
e particulate matter enters the train, contaminating the samples.
Samples are placed in plastic containers as follows:
is not necessary to change the probe tip diameter, merely adjust
le new temperature through the Original probe tip diameter to obtain
P.
WET-PRCCESS PHOSPHORIC ACID EMSSIONS
-------
Container No. 1 carefully remove the filter from the filter holder,
place in the container, and seal with tape.
Container No. 2 contains any loose particulate and acetone washings
from the probe, cyclone, and cyclone flash. The
inside of the cyclone and cyclone flash are brushed
with a Nylon brush and the inside of the probe is
brushed with a Nylon brush fitted on a stainless
steel rod, to loosen adhering particles.
Container No. 3 - contains any loose particulate and acetone washings
of the front half of the filter holder. The inside of
this part is brushed with a Nylon brush.
Container No. 4 - the 1^0 from the first three Greenburg-Smith
impingers is measured to within ± 5ml and placed ii
the container. H?O rinsings from the back half of
the filter holder, the fritted glass support, all
connectors, and the first three Greenburg-Smith
impingers are also to be placed in this container
and the container sealed with tape.
Container No. 5 - the spent silica gel is weighed to the nearest 0. 1
gram and then returned to a container and sealed
with tape.
Gaseous train cleanup ~ The cleanup of the gaseous train is a simple
one-step operation because only the impingers and connectors are
washed out. The contents of the impingers are poured carefully into
the container. The impingers and connectors are then rinsed out three
times using approximately 20 milliliters of distilled water for ^ach
wash. This wash is combined with the impinger solution. The containi
is sealed and the top wrapped with tape.
Analysis of Particulate Matter
The following section discusses the procedure used by the labora-
tory in particulate fluoride analysis.
<, I " transfer the filter and any loose particulate matter
from the sample container to a tared glass weighing
dish and condition for 24 hours in a desiccator or
constant humidity chamber containing a saturated
solution of calcium chloride or its equivalent. Dry
to a constant weight and record the results to the
nearest 0. 1 milligram.
Container No. 2
Container No. 3
transfer the acetone washings from the probe, cycle
and cyclone flask, to a tared beaker, and evaporate t<
dryness at ambient temperature and pressure.
Desiccate for 24 hours and dry to a constant weight.
Record the results to the nearest 0. 1 milligram.
transfer the acetone washings of the front half of
the filter holder to a tared beaker and evaporate to
dryness at ambient temperature and pressure.
Appendix 3.
69
-------
Desiccate for 24 hours and dry to a constant weight.
Record the results to the nearest 0. 1 milligram.
Transfer all particulate samples (1, 2, 3) to a 250-ml graduated
glass-stoppered cylinder, and dilute to 250 milliliters with distilled
water. Shred the filter with forceps before transfer. Mix well, and
transfer the total contents of the graduated cylinder to a 300-milliliter
Erlenmeyer flask.
To estimate the appropriate aliquot size to be used in the distilla-
tion procedure, take a 25-milliliter aliquot of each type of sample
(impinger, water-soluble particulate, total particulate) and apply the
spectrophotometric procedure found in the following section on chemical
fluoride analysis. From the amount of fluoride found in the undistilled
aliquot, calculate the sample aliquot needed to yield 0. 5 milligram of
fluoride in the distilled sample.
Distillation is used to remove any interfering substances. Any
chloride interferences are removed by addition of Ag^SO^ to the dis-
tillation mixture.
Spadns Determination of Fluorides
Introduction » This method for determining fluoride concentration is
used for both particulate and gaseous fluorides. It also includes the
distillation necessary to separate the fluorides from the particulate-
malter samples collected. There are other methods available, but it
was the one employed by the Public Health Service laboratory in
analyzing the samples collected in this study.
Reagents All chemicals used must be ACS analytical reagent grade.
Spadns s olution » Dissolve 0. 959 gram of 4, 5-dihydroxy-3(p-sulfophen~
ylazo)-2,7 -napthalene disulfonic acid, and trisodium salt (Spadns), at
room temperature, if protected from sunlight. 24
Zirconyl chloride octahydrate solution» Dissolve 0. 133 gram of ZrOCL^
8H2O in 25 milliliters of HZO. Add 350 milliliters of concentrated
HC1, and dilute to 500 milliliters with distilled water. This solution
is stable at room temperature for at least three months.
Spadns reagent » Combine equal parts of the Spadns solution and ZrOCLp-
8H2O solution, and mix thoroughly. This reagent is stable for at
least 2 years.
Reference solution Dilute 7 milliliters of concentrated HC1 to 10
milliliters with distilled water. Add 10 milliliters of Spadns solution to
100 milliliters of distilled water and add the HC1 solution. Mix well.
This solution is used to set the spectrophotometer zero point and is
stable indefinitely.
Standard fluoride solution «. Dissolve 2. 2105 grams of dry NaF, and
dilute to 1 liter with distilled water. Dilute 1 milliliter of this solution
to 1 liter. This final solution contains 1. 0 micrograms per milliliter
of fluoride ,
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Distillation procedure « Place 400 milliliters of water, 200 milliliters
concentrated sulfuric acid, and about two dozen Carborundum chips into
the boiling flask and swirl to mix. Caution- the sulfuric acid water
solution should be mixed thoroughly before heat is applied to prevent'
splattering. Connect the apparatus as shown in Figure B-8. Begin
heating slowly at first, then rapidly until a temperature of 180" C has
been reached. ^-~> The connection between the boiling flask and condenser
must be separated immediately after the heat is removed to prevent
SUckback of the sample and for safety reasons. About 300 milliliters
of water should have been distilled over in about 45 minutes. At this
point, the apparatus has been flushed free of fluoride and the acid-
ratio has been adjusted. When the flask has cooled to 120" C, the
apparatus is ready for the sample.
CONNECTING TUBE
12-mm I.D.
THERMOMETER
WITH J 10/30
J 24/40
I-LITER
BOILING
FLASK
CONNECTING TUBE J 24/40
24/40
CONDENSER
BURNER
I I 500-ml
/ \ ERLENMEYER
/ \ FLASK
LJ
Figure B-8. Fluoride distillation apparatus.
Add 300 milliliters of distilled water containing an aliquot of the
impinger sample, corresponding to 0. 5 to 0. 9 milligram of fluoride
to the boiling flask, swirl to mix, and connect the apparatus and distill
as before, until the distillation temperature reaches 180" C. For
distillation of water-soluble particulate fluorides, take a suitable aliquot
of the supernatent liquid of the particulate sample, dilute to 300 milli-
liters with distilled water, and add to the distillation flask. For dis-
tillation of total particulate fluorides, use a suitable aliquot of the
Appendix B
71
-------
water-soluble-plus water-insoluble sample. To obtain a representative
sample, withdraw an aliquot using a calibrated, sawed-off pipette,
immediately after intimate mixing of the samplk. In no case should the
aliquot contain more than 0. 9 mg of fluoride. Distill the sample, as
before, until a temperature of 180" C has been reached. Fluoride
content of phosphate rock or fertilizer may be determined using these
same procedures, provided the approximate percentage weight of fluo-
ride in the sample is known so that the still is not overcharged. Weigh
out a sample to the nearest 0.1 milligram, corresponding to about 0.5
milligram of fluoride, dilute to 300 milliliters with distilled water and
distill as before until a temperature of 180" C has been reached. Pipet
a suitable aliquot (containing 10 to 40 micrograms of fluoride) from the
distillate and dilute to 50 milliliters. Add 10 milliliters of Spadns
reagent, mix thoroughly, and read the absorbance. If the absorbance
falls beyond the calibration curve range, repeat the procedure using a
smaller aliquot.
Discussion of procedure -The estimated error for the combined
sampling and analytical procedure is ± 15 percent. The error of the
analytical method is ±4 percent. The spectrophotometric measurements
should be reported to the nearest 0.5 microgram.
Aluminum, calcium, chloride, ferric, manganese, magnesium,
phosphate, and sulfate ions interfere positively in the Spadns method. 1
These interferences are removed during the distillation of the sample.
Chloride interference can be eliminated when present in high concentra-
tions by the addition of 5-milligrams silver sulfate per milligram of
chloride. Addition of a few crystals of Ag£SO^ to a small portion of the
sample should be performed before distillation to determine if chloride
ions are present.
The determination of fluorides using this procedure may be
carried out at any temperature within the range of 15° to 30" C. The
important consideration here is that standards and sample should be
at nearly identical temperature, because an error of 0.01 milligram
per liter of fluoride is caused by each degree difference in temperature. "
Color, after the initial 15 -minute period, is stable for about 2 hours.
When the fluoride content of the aliquot is above 0. 9 milligram, the
distillation apparatus should be purged with 300 milliliters of distilled
water , so that there will be no residual fluoride carried over when the
next sample is distilled. Keeping the fluoride content around 0. 5
milligram eliminates the necessity of purging the distillation apparatus
between samples. The acid need not be replaced until the accumulation
of ions causes carry-over of interferences or retards fluoride recovery.
An occasional recovery check with standard fluoride samples will indi-
cate when the acid should be replaced.
Calculations
, . , 44.82 (C) (F)
ppm tluoride = 77
vs
7Z WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
53jrR^
492 "R X
1 J 1
mole
22.4 'iig^, lo6
mole
Y 103 mg v 2 8 32
g
liters
cu ft
where:
C = concentration fluoride in aliquot, milligram
F = dilution factor
Vg = volume of gas sample at 70" F and 29. 92 in. Hg, scf
44.82 =
Preparation of calibration curve Pipet exactly 0. 0, 10. 0, 20. 0,
30. 0, 40. 0, and 50. 0 milliliters of standard NaF solution into separate
100-inilliliter beakers. Add 50.0, 40. 0, 30.0, 20.0, 10.0, and 0.0
milliliters of distilled J^O respectively, to the beakers. Add 10
milliliters of Zirconyl-Spadns reagent to each beaker. Mix thoroughly
and let stand for 15 minutes at room temperature. Set the instrument
to zero absorbance using distilled water. Determine the absorptivity
of the reference solution. The absorptivity of the reference solution
should be in the range of 0.82 to 0. 85, using 0. 5-inch cells. Then,
set the instrument to zero absorbance using the reference solution.
Plot concentration versus absorbance on rectilinear graph paper.
Appendix B / 73
-------
APPENDIX C.
WET-PROCESS PHOSPHORIC ACID
ESTABLISHMENTS IN UNITED STATES
The purpose of this tabulation of wet-process phosphoric acid
manufacturing establishments (Table C-l) is to indicate the distribution
and principal areas of concentration of this industry. The industry
tends to be concentrated near the supply of phosphate rock; rock deposits
are located in Florida, Tennessee, and the Idaho-Utah area.
Information was drawn from various sources and is believed to
represent the operable installations existing as of May 1967. As a
result of sale, merger or lease, some company identifications may
differ from those presently in use, but this listing should serve the
intended purpose of general identification.
Table C-l. WET-PROCESS PHOSPHORIC ACID ESTABLISHMENTS IN UNITED STATES
(as of May 1967)
State
Arkansas
California
California
California
California
California
Delaware
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Idaho
Idaho
Idaho
Illinois
Illinoi s
Illinois
City
Helena
Bena
Dominquez
Helm
Lathrop
Tcona
North Claymont
Bonnie
Bartow
Bactow
Brewster
Fort Meade
Hamilton
Mulberry
Nichols
Pierce
; Pinev Point
Plant City
Plant City
Ridgewopd
South Pierce
Tamp a
Green Bay
White Springs
Conda
Kellogg
Pooatello
Dupue
_E. St. Louis
iJoiiet1
Company
Ackla Chemical Corporation
AFC Inc.
Westecn States Corporation
Valley Nitrogen Products, Inc.
The Best Fertilizers Corporation
American Potash and Chemical Corporation
Allied Chemical Corporation
International Minerals and Chemicals Corp.
Armour Agricultural Chemical Company
Swift and Company
American Cyanamid Company
Armour Agricultural Chemical Company
Occidental Petroleum Company
F. S. Roystec Guano Company
Mobil Chemical
Consumers Cooperative Association
Borden Chemical Company
Borden Chemical Company
Central Phosphates
H. E. Grace Company
American Agricultural Chemical Co.
Tennessee Corporation,
U. S. Phosphoric Products Division,
Cities Service
Farmland Industries
Occidental Agricultural
El Paso Products Company
The Bunker Hill Company
J. E. Simplot Company
New Jersey Zinc Company
Allied Chemical
Olin Mathieson Chemical Corporation
Capacity,
tons/yr (PgOs)
20,000
12.000
60.000
20,000
5.000
33.000
495.000
272.000
90.000
200,000
165.000
550,000
30,000
230,000
75,000-
100.000
140,000
100,000
165,000
228,000
340.000
110,000
250,000
90,000
33,000
270.000
130,000
35.000
125.000
75
-------
Table C-l (continued), TOT-PROCESS PHOSPHORIC ACID
IN UNITED STATES' (as of May 1967)
ESTABLISHMENTS
State
Illinois
Illinois
Illinois
Illinois
Indiana
Louisiana
Louisiana
Louisiana
Minnesota
Mississippi
Missouri
Missouri
Missouri
Missouri
New Jersey
North Carolina
Oklahoma
Texas
Texas
Texas
Texas
Utah
City
Marseilles
M orris
Stieator
Tuscola
Gary
Convent
Geismac
Hahnville
Pine Bend
Pascaugoula
Joplin
Joplin
Joplin
Joplin
Paulsboro
Aurora
Tulsa
Houston
Pasadena
Pasadena
Texas City
Garfield
Company
National Phosphates (Hooker Chemical Corp.)
Des Plaines Chemical (StaufFer Company)
Borden Chemical Company
U. S. Industrial Chemicals Corporation
SoCOIiy Mobil Oil Company
FteepOrt Chemical
Allied Chemical Corporation
Hooker Chemical Corporation
Northwest Cooperative Mills, Inc.
Coastal Chemical Corporation. Inc.
Consumers Cooperative Association
Farmers Chemical Company
W. R. Grace Company
W. R, Grace Company
Dixon Chemical Industries (not operating)
Texas Gulf Sulphur
Nipak, Inc.
Phosphates Chemical Inc. (Stauffer)
Olin Mathieson Chemical Corporation
Phosphate Chemicals Inc. (Stauffer)
Borden Chemical Company
Western Phosphates Inc. (Stauffer)
Capacity,
WES/yr (PgOg)
200,000
90,000
33, (MO
30.000
40,000
600.000
180.000
100,000-
120.000
54:ooo
50,000
53.000
50,000
50,000
33.000
40,000
375,000
30,000
100.000
200.000
80.000-
100.000
40,000
100. 00
76
.WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
APPENDIX D
PHYSICAL DATA ON PROPERTIES OF
CHEMICALS, AND SOLUTIONS RELATED TO'
WET-PROCESS PHOSPHORIC ACID MANUFACTURE
77
-------
Table D-l. PHYSICAL PROPERTIES OF AQUEOUS SOLUTIONS
OF PHOSPHORIC ACID 29
Concentration, %
HSP04
0
5
10
20
30
50
75
85
100
115
Pa05
0
3.62
7.24
14.49
21.73
36.22
54.32
61.57
72.43
83.29
Density,
e/cc
0.997
1.024
1.052
1.113
1.180
1.332
1.574
1.685
1.864
2.044
i
jBoiiing;
point,
'°c
100.0
100.1
100.2
100.8
101.8
108
135
158
261
Specific;
heat,
cal/ga
0.973
0.939
0.871
0.798
0.656
0.542
0.493
Specific
electrical conductivity,
at 18 °C, mho
0.0566
0.1129
0.1654
0.2073
0.1209
0.0780
"Average value from 20 " to 120°C.
Table D-2. KINEMATIC VISCOSITY OF PHOSPHORIC ACID SOLUTIONS29
(centistokes)
Concentration, ,
% H3P04
0
5
10
20
30
50
75
85
100
115
Temperature. °C
20
1.0
1.1
1.2
1.6
2.2
4.3
15
28
140
30
0.80
0.89
0.99
1.3
1.7
3.3
10
19
81
40
0.66
0.74
0.83
1.1
1.4
2.6
7.8
14
53
60
0.48
0.54
0.61
0.78
1.0
1.8
4.8
8.1
25
1500
80
0.37
0.42
0.47
0.60
0-79
1.4
3.3
5.1
14
600
100
0.30
0.33
0.38
0.48
0.62
1.1
2.4
3.8
9.2
250
140
-
-
-
-
-
-
-
2.2
4.5
68
180
.
-
-
-
-
-
-
-
2.9
28
Table D-3. VAPOR PRESSURE OF PHOSPHORIC ACID SOLUTIONS29
(nunHg)
Concentration,
Temperature. ° Q
% H3P04
0
5
10
20
30
50
75
85
100
20
17.6
17.5
17.3
17.0
16.3
13.0
5.65
2.16
0.0285
30
31.8
31.5
31.0
30.0
28.9
23.1
10.0
3.95
0.0595
40
55.3
54.5
54.2
53.0
50.5
40.3
17.5
6.95
0.120
60
150
147
146
141
136
108
47.0
19.7
0.430
80
355
352
350
341
327
257
111
48.8
1.33
100
760
755
753
735
705
575
240
111
3.65
110
1075
1068
1066
1040
996
814
340
160
5.80
140
.
4
895
445
20.3
78
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Table D-4. PARTIAL PRESSUREOF HYDROGEN FLUORIDE
OVER HF-H20 SOLUTIONS36
(mm Hg)
Hydrogen fluoride,
wt %
0
10
20
30
50
70
100
Temperature, C
0
0
0.03
0.09
0.30
3.66
41.2
364
20
0
0.14
0.41
1.27
12.4
118
773
40 60
0
0.51
1.51
4.46
35.8
295
1516
0
1.62
4.75
13.4
91.4
662
2778
80
0
4.50
13.2
35.7
209
1355
4801
100
0
11.2
32.7
85.5
440
2570
7891
Table D-5. PARTIAL PRESSURE OF WATER OVER HF-H5>0 SOLUTIONS30
(mmHg)
Temperature, "C
Wt %
0
10
20
30
50
70
100
0
4.58
4.46
3.63
2.72
0.76
ml
0
20
17.54
16.0
13.1
9.25
2.98
0.1
0
40
55.32
48.9
40.3
30.6
9.86
0
60
149.38
131
108
82.6
28.2
0
80
355.1
312
259
199
71.9
0
100
760.0
679
566
436
165
0
Table D-6. VAPOR PRESSURE OF ANHYDROUS
HYDROGEN FLUORIDE31
Temperature, °C
-10
0
10
20
30
40
50
60
70
Vapor pressure,
4.65
7.00
10.3
15.0
21.2
29.5
39.8
53.8
71.0
psia
Appendix D
79
-------
Table D-7. PHYSICAL PROPERTIES OF FLUORINE AND SILICON COMPOUNDS
SifV1
Boiling point
Density @ 0 °C, 1 atm
cp
Heat of formation
AHV @ 183 °K, 1320 mm
i 1 fl 1 TO loOiC-o i-ir
H2Si F6
Density of water solutions
% HgSJFg 6
d 1.0491
pH of industrial aqueous
wt % HgSiFg 1.0
pH 1.4
-95 °C
4.69 g/liter
18.2 cal/mol °C
-370 K cal/g mo)
4.46 K cal/g mo]
>or pressure, mpi Hg
, g/cc
14 22
1.1190 1.1941
HgSiFg solutions23
0.1 0.01
2.2 3.0
30
1.2742
0.001
3.8
SiF4 + 2HF XH20 = H2SiF6 + XH20
AH = , 67K Cal/g mol
SiF4 + aH20 = 2 HgSiFg (aq) + Si02
AH = -556.2 K cal/g
Azeotcopes33
wt%
HF HgSiFg
38.26
10 36
41
HgO P. mm Hg
61.74 750.2
54 759.7
59 760
B P,
112.
116.
111.
°C
Oa
i
5a
^ A Z e otr Op 6 S estimated from ternary phase diagram.
80
WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
SUBJECT INDEX
Abatement equipment (see scrubbers) 25-27
Air pollution potential 1-2, 32-33
Ammonium phosphates 1, 6
Analytical techniques 69-73
Apatite (see fluorapatite) 39
Azeotropes 79
Calcium sulfate
Anhydrous 9
Crystallization 9, 13
Dihydrate (see gypsum) 5, 9, 11
Gypsum 9
Semihydrate 9
Concentration (see evaporation) 13, 14, 28
Control equipment (see scrubbers)
Control methods 2, 3, 4, 21, 25-27
Definitions 39-42
Digester (see reactor) 13
Effluents (see emissions) 39
Emissions 22-25, 46-56
Abatement equipment for 25-27
Concentrations 3, 22-25
Data 46-56
Filter 46, 49, 51-53
Fluoride 2, 3
Gaseous fluoride 2, 3, 15
Insoluble fluoride 3, 22, 23, 46, 47
Origins '1, 2, 14, 15, 17-19
Particulate 2, 3, 22, 23
Ranges 2, 3, 22, 23
Rates 3, 46-56
81
-------
Reactor 46-49, 51-53, 55-56
Soluble fluoride 3, 46, 47
Sulfur dioxide 19, 29
Sump and vent 5 1, 53
Uncontrolled 19, 46-56
Evaporation 12, 13, 14, 28, 40
Filtration 5, 12, 13, 40
Fluorapatite 1, 9, 10, 40
Fluorine 40
Analytical determination 35, 36, 70-73
Fluorine compounds
Distribution in wet-process acid manufacture 14, 15
Emissions 2, 3, 15, 22-25, 46-56
Phosphate rock, content of 10, 17
Physical properties 79
Points of emission 1, 2, 14, 15, 17-19
Recovery 14, 18
Wet-process phosphoric acid, content in 11
With silicon 21
Fluosilicates 9, 21
Fluosilicic acid 9, 21, 79
Origin in wet-process phosphoric acid manufacture 9, 18, 21
Fog 28, 29, 40
Fume abatement 28, 29
Glossary 37
Gypsum 9, 11, 13, 40
Gypsum pond 2, 15, 16, 32, 40
Hydrogen fluoride 21
Physical properties 78
Hydrofluosilicic acid (see fluosilicic acid) 38
Isokinetic sampling 64
Operating nomographs 66, 67
Normal superphosphate 40
Odor 19
Phase equilibria 78,79
82 WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
Phosphate rock 41
Apatite 39
Chemical composition 10
Occurrence 1, 9
Mineral acid, reaction with 1, 21
Phosphoric acid (see also evaporation) 40
Physical properties 77
Physical properties 77-79
Pollutants (see emissions) 22-25, 46-56
Pond 2, 15, 16, 32
Process flow diagram 12
Raw materials 9, 10
Reactor (see digester) 13, 40
Sampling techniques 57-69
Scrubbers 25-27, 30-32, 41
Data 30-32, 46-56
Design 21
Efficiency 2, 3, 22-24
Impingement (as Doyle) 27, 31, 40, 50
Operation 30, 32
Packed tower 26
Performance 27-30
Rectangular chamber 30
Spray cross-flow packed 25, 26, 31, 32, 41
Spray tower 27, 41
cyclonic 27
Square duct 30
Staging 2
Venturi 26, 30, 42
Silicon tetrafluoride
Reactions 21
Stack testing vii, 57-69
Submerged combustion 28, 41
Sulfur dioxide 19, 29
Superphosphoric acid 7, 41
Subject Index 83
-------
Thermodynamic properties 79
Transfer unit 27, 41
Triple superphosphate 1, 5, 42
Vapor pressure 16, 77-79
Vapor-liquid equilibria 78, 79
Wet-process phosphoric acid
Chemical composition 11
Chemistry of manufacture 9, 21
Establishments 74, 75
Evaporation 6, 10, 13, 14, 18
Flow diagram for manufacture 12
History 5
Manufacturers 74, 75
PZ05 content 1, 5, 7, 10
Processes for manufacture 5, 7, 11-14
Production 6
Statis tics 6
Uses 1, 6, 7
84 WET-PROCESS PHOSPHORIC ACID EMISSIONS
-------
REFERENCES
1. Chem. and Eng. News, Annual Facts and Figures. Bureau of
the Census, U. S. Dept. of Commerce Industrial Statistics,
Section on Industrial Inorganic and Organic Chemicals.
2. Sauchelli, V. Chemistry and Technology of Fertilizers. Rein-
hold Publishing Corporation, 1960.
3. Parrish P. and A. Ogilvic. "Calcium Superphosphate and
Compound Fertilizers. " London Hutchinson & Co., 30, 1946.
4. Porter, J. J. and J. Frisken. Fertilizer Soc. (England)
Proc. 21, 3 (1953).
5. Larison, E. L. I. &E.Chem. 21, 1172 (1929).
6. Heinrich, C. Chem. Zeitung, pg. 71, (Sept. 6, 1953).
7. Chem. Eng. , 22nd, 23rd, and 24thAnnual Inventories of New
Processes and Technology.
8. Allgood, H. G. , F. E. Lancaster, Jr., J. A. McCollum, J.
P. Simpson. A High-Temperature Superphosphoric Acid Plant.
I. k E. Chem. _59, 6, 18-28 (1967).
9. Chelminski, R. and R. L. Somerville. Chem. Eng. Prog. 62,
(5) 108-111 (1966).
10. Van Waser, J. R. Phosphorus and Its Compounds. Interscience
N. Y., (1961).
11. Kirk-Othmer. Encylopedia of Chemical Technology, 2nd
edition, Interscience Pub.
12. Hignett , T. G. Manufacture of Wet-Process Phosphoric Acid.
TVA (1961).
13. Stewart, R. F. and F. W. Edwards. Development of Phosphoric
Acid by the Wet Process. The Industrial Chemist 36, 397-399
(1960).
14. Greek, B. F. , F. W. Bless, Jr., and R. S. Sibley. St. Gobain
Processes. I.& E. Chem. 52^ 8, 638-644. (1960).
15. Sanders, M. D., and W. C. Weber. A New Fluorine Recovery
Process. Dorr-Oliver Preprint No. P-2302.
16. Grant, H. 0. Pollution Control in a Phosphoric Acid Plant.
Chem. Eng. Prog. 60, (l):53-55 (1964).
17. Huffstutler, K. K. and W. E. Starnes. Sources and Quantities
of Fluorides Evolved from the Manufacture of Fertilizer and
Related Products. Journal APCA 16, 12, 682-4 (1966).
18. Teller, A. J. Control of Gaseous Fluoride Emissions. Chem.
Eng. Prog. 63, 3, 75-79 (1967).
85
-------
19. Badger, W. L. and J. T. Banchero. Introduction to Chemical
Engineering. McGraw-Hill, N. Y., (1955).
20. Harris, L. S. Fume Scrubbing with the Ejector Venturi System.
Chem. Eng. Prog. 6Z, 4, 55-59 (1966).
21. Chem. & Eng. News, Sept. 5, 1966.
22. Ballack, Ervin. Simplified Fluoride Distillation Method, JAWWA.
50, 530-6, (1958).
23. Bellack, Ervin and P. J. Schouboe. Rapid Photometric
Determination of Fluoride in Water. ANAL. CHEM. 30:2032
(1958).
24. Standard Methods for the Examination of Water and Waste
Water. APHA, AWWA, and WPCF, 12th Edition, 144-146,
(1965).
25. Bellack, Ervin. Automatic Fluoride Distillation, JAWWA, 53,
99-100, (1961).
26. Patton, W. F., and J. A, Brink. New Equipment and Techni-
ques for Sampling Chemical Process Gases. Presented at the
55th Annual Meeting of APCA, Sheraton . Chicago Hotel,
Chicago, Illinois, May 20-24, 1962.
27. Perry, J. H. "Chemical Engineers' Handbook" McGraw Hill
Book Co., New York, (1950).
28. Western Precipitation Company Bulletin WP-50.
29. Monsanto Chemical Co. Phosphoric Acid Technical Bulletin
pg. 26, (1946).
30. I. & E. Chem. 41_, 1504-1508, (1949).
31. Z. Elektrochem 6\_, (3), 348-57 (1957).
32. Sherwin, K. A. Chem. & Ind. , 1274-1281, Oct. 8, 1955.
33. Munter, P. A. 0. T. Aepli, and Ruth A. Kossatz. HF-H;>0
and HF-H2SiF6-H2O. I.&E. Chem. _39, 3, 427-431 (1947).
34. Specht, R. C. and R. R. Calaceto. Gaseous Fluoride Pollu-
tants from Stationary Sources. Chem. Eng. Prog. 63, 5,
78-84 (1967).
35. Piombiro, A. J. C. W. Report, Phosphates, Chemical Week,
109-131, Oct. 24, 1964.
36. Bixby, D. W. , D. L. Rucker, and S. L. Tisdale. Technical
Bulletin No. 8 of the Sulphur Insitutue. Feb. 1964.
37. Wilde, W. D. Wet Phosphoric Acid. Chem. Eng. Prog. 58,
4, 92-94 (1962).
38. Leonard, R. B. Bidding to Balk Corrosion in Phos-Acid Con-
centration. Chem. Eng. (158-162). June 5, 1967.
39. James, G. Russell. Filtration in the Manufacture of Phos-
phoric Acid. Chem. Eng. (137-140). Aug. 15, 1966.
36 References
------- |