SN 16893.000
ENGINEERING AND
COST EFFECTIVENESS STUDY
OF FLUORIDE EMISSIONS CONTROL
(FINAL REPORT)
JANUARY 1972
VOLUME I
*
Prepared under Contract EHSD 71-14
for
OFFICE OF AIR PROGRAMS
ENVIRONMENTAL PROTECTION AGENCY
TRW Systems Group
7600 Colshire Drive
McLean, Virginia 22101

-------
SN 16893.000
ENGINEERING AND
COST EFFECTIVENESS STUDY
OF FLUORIDE EMISSIONS CONTROL
(FINAL REPORT)
JANUARY 1972
J.M.Robinson (Program Manager), G.I. Gruber, W.D. Lusk, and M.J. Santy
VOLUME I
Prepared under Contract EHSD 71-14
for
OFFICE OF AIR PROGRAMS
ENVIRONMENTAL PROTECTION AGENCY
Resources Research, Inc.
7600 Colshire Drive
McLean, Virginia 22101
TRW Systems Group
7600 Colshire Drive
McLean, Virginia 22101

-------
ACKNOWLEDGEMENT
This study was performed under the technical direction of
W. R. King, Project Officer, Industrial Studies Branch, Applied
Technology Division, Bureau of Stationary Sources. In addition, TRW
Systems would like to acknowledge the contributions of Dr. E. A. Largent,
Dr. R. E. Kimball, Dr. H. R. Hickey, Dr. M. L. Spealman, Dr. R. S. Ottin-
ger, 0. W. Flock, J. S. Land, J. R. Denson, F. Harpt, and C. T. Weekley.

-------
CONTENTS
Page
1.	SUMMARY			1-1
2.	INTRODUCTION		2-1
3.	INDUSTRIAL SOURCES				3-1
3.1	General			3-1
3.1.1	Economic Analyses - Discussion		3-1
3.1.2	Thermochemical Analysis Approach		3-11
3.2	Primary Alumium Smelting Industry		3-13
3.2.1	General			3-13
3.2.2	Industry. Description.			3-13
3.2.3	Production Trends 		3-16
3.2.4	Fluoride Control and Emission Summary . .	3-20
3.2.5	Process Description 		3-22
3.2.6	Economic Analysis 		3-36
3.3	Iron and Steel Industry		3-57
3.3.1	General			3-57
3.3.2	Industry Description		3-57
3.3.3	Production Trends			3-58
3.3.4	Fluoride Control and Emissions Summary. .	3-59
3.3.5	Process Description and Economics ....	3-64
3.3.6	Economic Analysis 		3-99
3.4	Coal Combustion - Electrical Power Generation. . . .	3-127
3.4.1	General		3-127
3.4.2	Industry Description		3-127
3.4.3	Production Trends 		3-128
3.4.4	Control Techniques		3-131
3.4.5	Process Description 		3-133
3.4.6	Economic Analysis 		3-135
3.5	Phosphate Rock Processing		3-137
3.5.1	General		3-137
3.5.2	Industry Description		3-137
3.5.3	Production Trends			3-139
3.5.4	Fluoride Control and Emissions Summary. .	3-143
3.5.5	Wet Process Description		3-143
3.5.6	Economic Analysis 		3-194
3.6	Glass Manufacture		3-217
3.6.1	General			3-217
3.6.2	Industry Description		3-217
3.6.3	Production Trends 		3-217
3.6.4	Fluoride Emission Control Techniques. . .	3-220
3.6.5	Fluoride Emissions			3-220
iii

-------
CONTENTS (Continued)
Page
3.6.6 Economic Analysis 		3-224
3.7	Frit Smelting			3-231
3.7.1	General 		3-231
3.7.2	Industry Description. 				3-231
3.7.3	Production Trends	3-231
3.7.4	Fluoride Control Techniques ........	3-231
3.7.5	Fluoride Emissions. ... 		3-235
3.7.6	Economic Analysis 				 .	3-240
3.8	Heavy Clay Products. 			 		3-245
3.8.1	General			3-245
3.8.2	Industry Description ...........	3-245
3.8.3	Production Trends		 . 		3-245
3.8.4	Fluoride Emission Control Techniques. . . .	3-249
3.8.5	Fluoride Emissions			3-249
3.8.6	Economic Analysis			3-249
3.9	Expanded Clay Aggregate	3-259
3.9.1	General 	 .......	3-259
3.9.2	Industry Description	. ,	3-259
3.9.3	Production Trends 				3-259
3.9.4	Fluoride Emission Control Techniques. . . .	3-262
3.9.5	Fluoride Emissions. 				3-262
3.9.6	Economic Analysis 		3-262
3.10	Cement Manufacture . . . 		3-271
3.10.1	General 		3-271
3.10.2	Industry Description. ...... 		3-271
3.10.3	Production Trends 		3-271
3.10.4	Fluoride Emission Control Technique ....	3-277
3.10.5	Fluoride Emissions. 		3-277
3.10.6	Economic Analysis . 		3-283
3.11	HF Alkylation Processes	3-287
3.11.1	General 				3-287
3.11.2	Industry Description	3-287
3.11.3	Production Trends 		3-287
3.11.4	Fluoride Emission Control Techniques. . . .	3-289
3.11.5	Fluoride Emissions. . 				3-291
3.11.6	Economic Analysis ... 	 .	3-291
3.12	HF Production. 		3-295
3.12.1	General				 .	3-295
3.12.2	Industry Description	3-295
3.12.3	Production Trends ...... 		3-295
3.12.4	Fluoride Emission Control Techniques. . . .	3-295
3.12.5	Fluoride Emissions. . . 		3-300
3.12.6	Economic Analysis . 			3-300
3.12.7	Impact of Control 	 .	3-300
iv

-------
CONTENTS (Continued)
Page
3.13	Nonferrous Metals Smelting and Refining Industry . . .	3-303
3.13.1	General 		3-303
3.13.2	Copper Smelting and Refining Industry . . .	3-304
3.13.3	Lead Smelting and Refinining Industry . . .	3-307
3.13.4	Zinc Smelting and Refining Industry ....	3-311
3.14	Other Industries 		3-318
3.14.1	Fluorine and Fluorocarbon Chemicals ....	3-318
3.14.2	Uranium Fluoride Production 		3-319
3.14.3	Aluminum Anodizing	3-321
3.14.4	Beryllium Production	3-322
4.	RESEARCH AND DEVELOPMENT PLANNING 		4-1
4.1	Summary and Priorities	4-1
4.2	Detailed Projects by Industry	4-5
4.2.1	Primary Aluminum Smelting Industry	4-5
4.2.2	Iron and Steel Manufacture	4-7
4.2.3	Coal Combustion	4-14
4.2.4	Cement, Ceramic and Glass Manufacture . . .	4-22
4.2.5	Nonferrous Metals Smelting and Refining
Industry. 		4-30
5.	ENVIRONMENTAL EFFECTS. . . . 		5-1
5.1	Vegetation Effects 		5-1
5.2	Effects on Farm Animals	5-3
5.3	Fluoride Effects in Man	5-7
5.4	Etching of Glass		5-10
5.5	Effects of Fluorides on Structures 		5-11
6.	MEASUREMENT TECHNOLOGY 		6-1
6.1	Sampling		6-1
6.1.1	Sampling Procedures 	 .....	6-1
6.1.2	Performance of Sampling Trains	6-2
6.1.3	Process Factors Affecting Sampling	6-7
6.1.4	Sampling Summary	6-8
6.2	Fluoride Separation	6-9
6.3	Analytical Methods 	 .....	6-11
6.3.1	Summary of Analytical Methods and
Recommendations 		6-12
6.3.2	' Spectrophotometry Analysis	6-15
V

-------
CONTENTS (Continued)
Page
6.3.3	Titrimetric Analysis		 . ...	6-17
6.3.4	Instrumental Methods	6-19
6.3.5	Continuous and Semi continuous Methods . . .	6-24
VOLUME II
7.	APPENDIX					7-1
8.	BIBLIOGRAPHY ..... 	 ...	8-1
vi

-------
1. Summary

-------
1. SUMMARY
This study, performed by RRI/TRW Systems for the Office of Air Programs
of the Environmental Protection Agency, inventoried fluoride emission sources
and investigated the technical and economic aspects of implementing soluble
fluoride emission controls for major industrial sources. Soluble fluorides
are defined as those fluorides with appreciably greater solubility (and
ecological impact) than calcium fluoride.
Table 1-1 presents estimates of present and projected (year 2000)
emissions of soluble fluorides from the major industrial process sources in
the United States. Indications of the relative levels of confidence in the
current estimates are included, as are projected emissions assuming that 99%
emission control has been attained.
Table 1-2 presents current production rates and those projected for
the year 2000. The evolution and emission factors for each process are
presented in Table 1-3. The evolution factor Includes all soluble fluorides
leaving the process prior to control. The emission factor corresponds to
that portion of the evolved soluble fluorides that eventually enters the
atmosphere.
The following observations can be made after consideration of the
information presented in the body of the report:
Five of the first six industries listed typically utilize no
fluoride control (the exception 1s aluminum production).
It is technically possible to control soluble fluorides with
available devices such as wet scrubbers; the immediate
problem lies in implementation of that control, including
collection of the evolved fluorides by hoods and similar
effluent capture systems for treatment in the abatement
devi ces.
Implementation of control involves a cost which reduces return
on investment by varying amounts in different industries.
Control of fluoride emissions becomes largely a matter of
economics and/or control regulations.
1-1

-------
The lesser confidence attached to about half of the current
emission estimates,including the top four processes in the
current rankings of Table 1-1, indicates the need for more
direct experimentally obtained data on both emissions and
feed stock compositions for these industries.
The fluorides currently emitted may damage economic crops,
farm animals, and materials of decoration and construction.
It should be noted, however, that the potential for the
observed ambient atmospheric levels to cause fluoride effects
in man is negligible.
The major problem in measuring the fluoride contents of
industrial effluent streams is that of obtaining representative
samples which do not change by internal reaction prior to
analysis. Current analytical techniques are satisfactory for
laboratory analysis. No satisfactory continuous sampling and
analytical system for continuous direct monitoring of process
streams and stacks has been developed.
Research and development work is required to select control
systems from those currently available which minimize the
economic impact on a given industry, thereby making
implementation of fluoride control as painless as possible.
Research and development work on fluoride contents of raw material
feed stocks, process streams, and process/plant effluents is required to
permit proper control design. Table 1-4 presents a summary of recommended
research and development projects. Approximate total contract cost for the
projects scheduled is estimated at $1.4 million.
1-2

-------
Table 1-1. Soluble Fluoride Emissions ^

Current



Year 2000 Mith Current Practice

Year 2000 Assuming 99% Control Efficiency
Ranking Process
Emission
Thousands
Of Tons F
Per Year (Yr)
Relative^
Confidence
Process
Emission
Thousands
' Of Tons F
Per Year
Process
Emnission
Thousands
Of Tons F
Per Year
. 1
Coal Burning For Power
27
(1970)
II
Primary Aluminum Manufacture
141
Primary Aluminum Manufacture -
8.1
2
Open Hearth Steelmaking
25
(1968)
II
Coal Burning For Power
Iron Ore Pelletlzing1 '
86
Coal Burning For Power
0.9
3
Iron Ore Sintering
18
(1968)
II
39
HF Production
0.7
4
Iron Ore Pelletizing'8'
18
(1968)
II
Expanded Clay Aggregate
(C)
Met Process Phosphoric Acid* ¦'
HF Alkylation^ :
25'
Electrothermal Phosphorus
0.4
5
6
Primary Aluminum Production
: Heavy Clay. Products
16
10
(1970)
(1968)
I
11
22
16
Iron Ore Pelletizing
Met Process Phosphoric Add^c'
fcl
Triple Superphosphate1 '
0.4
0.3
7
. Wet Process Phosphoric Acid^
6,4
(1970)
I
Heavy Clay Products
(r)
Triple Superphosphate1 '¦
10
0.3
8
HF Alkylat1on(D)
5.8
(1971)
' I
7.3
Expanded Clay Aggregate
fc
Defluorinated Phosphate Rock1
0.3
9
Expanded Clay Aggregate
5.3
(1968)
II
Electrothermal Phosphorus
6.6
0.2
10
Normal Superphosphate
5.0
(.1970)
1
Opal Glass Production
5,5
HF Alkylation
0.2
11
Electrothermal Phosphorus
4.1
(1968)
I
HF Production
5.3
Normal Superphosphate .
0.1
12
Triple Superphosphate'1*^
3.8
(1970)
' I
Iron Ore Sintering
(C)
Defluorinated Phosphate Rock1
4.8
Heavy Clay Products
0.1
13
Opal Glass Production
''3.3
(1968)
1 II
- 2.7
Opal Glass Production
<0.1
14
Blast Furnace
2.8
(1968)
. II .
Blast Furnace^
•2.6
Iron Ore Sintering
<0.1
15
Defluorinated Phosphate Rock'C'
. 1.8
(1970)
I
Normal Superphosphate
1.4
Blast Furnace
f CI
Arnnonium Phosphate
<0.1
16
HF Production
0.7
(1970)
I
Enamel Frit Production
1.1
<0.1
17
Enamel Frit Production
0.7
(1968)
II
Cement Manufacture
.0.8
Enamel Frit Production
<0.1
, 18
Copper Smelting and Refining
I r\
Arnnonium Phosphate1
0.6
(1967)
III
fC) '
Ammonium Phosphate* '
1.6

<0.1
19
0.3
(1970)
I
0,8
Cement Manufacture
<0.1
20
Cement Manufacture
0.3
(1964)
11
Open Hearth Steelmaking
0
Open Hearth Steelmaking
0
21
Lead Smelting and Refining
0.2
(1967)
III

2.4

<0.1
22
Z1nc Smelting and Refining
0.2
(1967)-
-III

2.4

<0.1
Total
for Processes Considered
155.3



384.3

12.0
(A)	Excludes CaF,
(B)	Assumes No Fluorspar Addition to Pellet	.
(C)	Includes Prorata Allocation of Gypsum Pond Fluoride Emissions (Estimated as 6,300 Tons for 1970 and 21,000 Tons for 2000)
(0) Assumes 25% of Production Uses L1me Pit Disposal of Acid Sludges
(E)	Assumes No Limestone Other Than That in Pellets or Sinter
(F)	Relative Confidence Levels: I is Excellent; II is Good; III is Fair to Poor * '

-------
Table 1-2. Production Trends to Year 2000
Product
Process
Present Production
Million Tons/Year (year)
Year 2000 Production
Million Tons/Year
Steel
Blast Furnace
130 (1968)
120 (2)

Iron Ore Sintering
50 (1968)(3)
13(3)

Iron Ore Pelletizing
50 (1968)(3)
107(3)

Open Hearth ¦
66 (1968)
0

Basic Oxygen
48 (1968)
135

Electric Arc
16 (1968)
35
Phosphates
Wet Process Phosphoric Acid (54%)
3.8 (1970)
13
as P205
Ammonium Phosphate
2.4 (1970)
7.0

Triple Superphosphate
1.4 (1970)
2.7

Normal Superphosphate
Electrothermal Phosphorus
Defluorination of Phosphate Rock
0.7 (1970)
1.6 (1968)(4)
0.09 (1970)(4)
0.2
2.6^
0.14(4)
Aluminum
Prebake and Soderberg
4.0 (1970)
35
Cement
Wet and Dry Process
68(1964)
200
Expanded Clay
Aggregate .

9.3 (1968)
44
Heavy Clay
Products

24 (1968)
24
Coal
Power Generation
333 (1970)
1080
HF

0.34 (1970)
2.6
Alkylate
HF Alkylation
236,000 bbl/day (1971)
7,800 tons CaF2^
Utilized (1968)
643,000 bbl/day
Enamel Frit

11,800 tons CaF„
Utilized ;
Opal Glass

34,500 tons CaF2
Utilized (1968)
57,200 tons CaF2
Utilized
. (1)„	Expressed as tons of CaF2 utilized in manufacture of product due to the varying compositions used.
(2)	Expressed as ore tonnage in blast furnace burdens.
(3)	Expressed as pre tonnage in process feed.
(4)	Phosphate rock used in process feed as P2°5-

-------
Table 1-3. Soluble Fluoride Evolution and Emission Factors by Process

Evolution Factor
Emission Factor
Industry and Process
LbF/Ton
Product
LbF/Ton
Product
Primary Aluminum Smelting




Prebaked Anode
46
(Al)
6.94
(Al)
Horizontal Stud Soderberg
46
(Al)
10.12
(Al)
Vertical Stud Soderberg
46
(Al)
9.66
(Al)
Iron and'Steel Manufacture




Iron Ore Sintering
0.73
(Ore)
0.69
(Ore)
Iron Ore Pellet1z1ng
0.73
(Ore)
0.69
(Ore)
Blast Furnace
0.088
(Ore)
. 0.065
(Ore)
Open Hearth Furnace
0.81
(Steel)
0.77
(Steel)
Basic Oxygen Furnace^
0
(Steel
0
(Steel)
Electric Arc Furnace^3)
0
(Steel)
' 0
(Steel)
Coal Combustion




Electric Power Generation
. 0.16
(Coal)
0.16
(Coal)
Cement Ceramic and Glass Mfr.



-
Opal Glass Production
21.8
(Glass)
21.8
(Glass)
Enamel Frit Production
3.15
(Dry Frit)
2.64
(Dry Frit)
Heavy (Structural) Clay Products
0.81
(Brick)
0.81
(Brick)
Expanded Clay Products
1.14
(Aggregate)
1.14
(Aggregate)
Portland Cement
0.008
(Cement)
0.008
(Cement)
Phosphate Rock Processing




Wet Process Phosphoric Acid
4.07
(P2o5)(b)
3.36
(w'b! '
Diammonium Phosphate
1.3
.(p2o5)
0.23
(p2°5) *
Triple Superphosphate
' ?1
(p2°5)
5.4'
(p2o5)(b>
Normal Superphosphate
7.1
(p2o5)
14.2
(P2°5), .
Electrothermal Phosphorus
30
(P2o5)
5.1

Non-Ferrous Metal Smelting and




Refining




Copper Smelting and Refining
0.78
(blister Cu)
0.78
(blister Cu)
Lead Smelting and Refining
0.34
(Lead) .
0.34
(Lead)
Zinc Smelting and Refining
0.46
(Zinc)
0.46
(Zinc)
HF Alkylation Processes
0.18
(bbl alkylate)
0.15
(bbl. alkylate)
HF Production
52
(HF)
4.1
(HF)
(a)	Estimated at zero soluble fluoride emission on the basis of thermodynamic equilibria
analyses and the assumed total unavailability of hydrogen for conversion of other
species to HF.
(b)	Includes pro-rata allocation of gypsum pond fluoride emissions.
1-5

-------
Table 1-4. Fluoride Emission Control - Recommended Research
and Development (Projects by Industry)
~"~--^_Project Type
Industry
Ore Fluoride Content
Characterization
Determination of Fate
of F 1n Mfg. Process
Determination of Fate
of Fluoride Evolved
Development of F
Control Techniques
Determination of Fate of
F 1n Control Process
Muminum


) Experimentally
determine F capture by
hoods and define building
control requirements and
characteristics


Iron and Steel.
1) Experimentally
determine and verify
average F contents of
Iron ore bodies and
sinter/pellet plant
charges for U.S. regional
areas
2) Experimentally
determine species and
quantities of F com-
pounds evolved and
emitted by primary
Iron/steel production
processes

3) Design, develop and
test on bench and
portable pilot plant
economic, effective F
emission control
processes for Iron/
steel

;oal Burning Steam-
ilectrlc Power
1) Experimentally
determine area F, alkali,
and alkaline earth metal
contents and variabilities
of coal
2) Experimentally
determine the F species
emitted by coal
combustion as function
of feed composition
4) Experimentally
determine effect of SO
and NO removal
processes on fluorides
In coal and effect of
fluorides on processes


3) Experimentally
determine quantities
and types of F compounds
removed by current and
projected SO- control
processes ana effect on
processes
dement, Ceramic and
jlass
1) Experimentally
determine by U.S.
regional areas the F
contents of cement
feedstocks
3)	Experimentally
determine the F content
of feedstocks and process
streams 1n frit mfr.
4)	Experimentally
determine by U.S.
regional areas the F
content of heavy clay
product and expanded clay
aggreg. feedstocks
2) Experimentally
determine F species
and quantities emitted
in cement mfr. as
function of feedstock
F content



lonferrous metals
1) Experimentally
determine the F contents
of Cu Pb Zn ores and
feedstocks by geographical
area
2) Experimentally
determine F species
and quantities
evolved and emitted
In Cu Pb Zn smelting
as functions of feed
and process parameters


3) Experimentally
determine the F species
and quantities emitted
from smelter byproduct
H2SO4 plants as
functions of process
parameters and feeds

-------
Introduction

-------
2. INTRODUCTION
This report presents the results of a study performed by Resources
Research Incorporated and the Chemistry and Chemical Engineering Laboratory
of TRW Systems Group under the direction of the Office of Air Programs,
Environmental Protection Agency. Primary emphasis has been placed upon the
determination of the engineering and cost effectiveness of control of
soluble fluoride emissions from the major industrial sources. Soluble
fluorides have been defined as those fluorine compounds which have a
substantially greater solubility in water than calcium fluoride. Calcium
fluoride emissions have to a large extent been deemphasized in the study
because of their inert character and presumed lack of significant ecological
and environmental impact. Other terms of special interest are evolution
and emission. Evolution, as used in this report, includes all soluble
fluorides evolved from a given process which would be discharged to the
atmosphere if no control systems were utilized to reduce the quantity.
Emission, as used in this report, includes only actual soluble fluoride
discharges into the atmosphere, from a process in its current or projected
state of emission control. Emission may equal evolution for those processes
having no fluoride control , or may be lower by orders of magnitude where
efficient control systems are employed.
Study tasks included inventory of fluoride emitting processes,
process modeling, assessment of the state of the art of measurement and
control technology, determination of control costs, projection of trends
to the year 2000, and recommendations for research and development (R&D)
activities to minimize soluble fluoride emissions in a cost effective
manner. Performance of these tasks required evaluation of each industry
processing significant quantities of fluoride-containing materials to
determine the magnitude of fluoride evolution and emission, emission
points, methodology available for measurement and control of emissions, and
product costs with and without control. Obviously, this was a difficult
task requiring definitive information about the specifics of various
industrial processes. For many processes and industries, definitive
information was not available. Many industries have not felt the need
2-1

-------
to concern themselves with fluoride emissions and have not allocated funds
for acquiring emission data. Some industries, specifically phosphate rock
processing and primary aluminum manufacture, have a high fluoride emission
potential and are under significant pressure to minimize emissions.
Industrial sources may or may not recognize that they have a potential
fluoride emission problem, and they may or may not make emission data
available if they exist. For each process considered in this study,
emission data were gathered from public information, RRI experience, and
results of other OAP studies. In cases where no data were available or
data were so sketchy as to be without usefulness, emission factors were
estimated based on raw material and product analyses, process chemistry,
proprietory thermochemical equilibrium programs, information from con-
sultants, and experience with similar products. In spite of the obvious
uncertainty associated with estimates of emission factors, these estimates
were necessary to prioritize emission sources and allow planning of future
R&D activities to achieve maximum impact on the fluoride emission problem.
Therefore, some assessment of the emission potential of each process under
consideration was made to allocate resources logically to minimize the
overall fluoride problem.
The following sections cover:
Industry-by-industry descriptions of each production and
control process of significance from a fluoride emission
control standpoint; discussions of production trends
extrapolated to the year 2000; process flow diagrams and
mass balances for typical current plants; estimates of
current and projected fluoride emissions; analyses of
production and control process economics.
Descriptions, cost estimates, priority assignments and
schedules for the additional research and development
programs recommended as the result of this study.
Environmental and ecological effects of the emitted fluorides.
Techniques for sampling, and measurement of fluoride
pollutants in the various effluent streams.
2-2

-------
In appendix format — a general discussion of fluoride
emission control devices; an inventory of pertinent
industrial plants and their locations; tabulations of the
physicochemical parameters of the evolved fluorides.
The study bibliography.
2-3

-------
Industrial Sources

-------
general

-------
3. INDUSTRIAL SOURCES
3.1 GENERAL
This section presents discussions of the fluoride emission control
problem on an industry-by-industry basis. For each process emitting
significant quantities of fluorides, production process models, projections,
emission estimates, economic analyses, and fluoride control process models
are described. The detail presented was sufficient to enable determination
of the engineering and cost effectiveness of the currently employed control
systems for soluble fluoride emissions - the stated objective of this study.
This objective did not include provision of a basis for construction of
facilities. The process descriptions, control systems, and economic
analyses presented are therefore deliberately limited in scope and
comprehensiveness to include only those elements necessary to estimate the
cost associated with application of currently used devices to fluoride
emission control, and its impact upon the profits obtained through sale of
the various products manufactured. For two of the industries considered as
fluoride sources, the phosphate rock processing and the primary aluminum
reduction industry, specific studies are being conducted by CHEMICO and
Singmaster and Breyer for EPA, the results of which should be available in
1972. Portions of their preliminary results are included in this report.
3.1.1 Economic Analyses - Discussion
All of the economic analyses performed for this study had the sole
objective of determining the impact of the cost burden imposed by fluoride
emission control systems upon the profitability of the basic manufacturing
processes. The index selected to determine the impact was the relative
change in return on investment (aROI) caused by the addition of emission
control systems to manufacturing processes devoid of such pollution control
devices. aROI was selected because of its relative lack of sensitivity to
minor inaccuracies in determining cost of manufacture and to differences in
accounting methods. These two points were deemed important because of the
great variation in the accounting methods employed by the many industries
3-1

-------
involved, and because of the company-to-company and location-to-location
deviations from the published manufacturing cost data for each production
process.
Manufacturing Cost Economics. Return on investment for each production
process or typical manufacturing complex was calculated from estimates of
the total capital required (both construction and working capital) for the
production system devoid of fluoride emission control devices; an assumed
~
ratio of 20 to 80 for debt to equity capital ; yearly "f.o.b. costs" which
were the sum of annual manufacturing costs plus an assumed average 2% general
and sales cost burden; annual gross income; and a 50% tax rate on profit.
Estimates of total capital required and manufacturing costs were based upon
data available from the literature and from TRW files, and upon the
assumptions listed in Table 3-1.
As an example of ROI calculations, consider a 120,000-ton per year
primary aluminum plant. The ROI is calculated from the following:
90 $MM total capital
72 $MM equity capital (80% of total capital)
51.2 $MM/year f.o.b. costs (50.2 $MM/year manufacturing costs
plus 1.0 $MM/year for general and
sales expense at 2%)
69.6 $MM/year gross income
50% tax rate
Thus, ROI is generally,
R0i _ 12 8% = 0-5 x (69.6 $MM/,year - 51.2 $MM/year) x 10^%
72 $MM
The difference between the ROI before and after requiring fluoride
pollution control equipment divided by the ROI before pollution control is
termed (for this study), as the relative percent decrease in ROI (aROI).
Capital for the pollution control equipment is added to the plant capital for
the final ROI and aROI calculations. For our example, if the aluminum plant
*
Applicable only to production equipment and plant systems. For the purposes
of this study, pollution control equipment was assumed to be completely equity
capital funded.
3-2

-------
Table 3-1. Manufacturing Process Economics Assumptions
Base years for cost data^
Phosphate Rock.Industries:	1966
Power Plant:	1969
HF Production:	1969
Aluminum:	1969
Steel:	1970
Opal Glass:	1970
Expanded Clay:	1969
Structural Clay:	1968
Portland Cement:	1970
• "Battery Limits" include the means of production usually represented by the
flow sheet.
Electric Power
Cooling Water
Boiler Feed Water
Operating Labor
Labor Overhead
Maintenance and Supplies
Depreciation
Plant Overhead
Taxes and Insurance
Supervision and Benefits
Interest (Investment and/or work
capital)
Distribution and Sales Cost
Start-Up Costs
General and Administrative Costs
Storage
Operating Rate
Offsltes
Contingency
Offsltes (where applicable)
Water supply treatment and cooling
Solids handling
Docks
Office Buildings
Yard Piping
Yard Electrical
Maintenance Equipment
Trucks and Tractors
Furniture
Tanks
$0,007 per kwh/hr
$0.03 per M gal
$0.25 per M gal
$4.00 to $5.00/man-hr
30 percent of operating labor and supervision
6 percent of investment/yr
10 percent of investment/yr
70 percent of operating labor and supervision
3 percent of investment per year
100 percent of operating labor
(supervision 20 percent,, fringe benefits 80 percent)
Not included
Not Included
Not Included
Not Included
Where specified
100 percent capacity, 330 days per year
20-60 percent of Investment
10 percent
Paving
Compressors
Waste Disposal
Piling
Concrete
Insulation
Structural
011 Storage
Boiler
Worker Housing
~Exceptions are indicated on the individual process sheets.
(1)
CE plant cost index was used to standardize costs to January 1971.
3-3

-------
experiences a 12.8% ROI before installation of fluoride control equipment
and a 9.7% ROI after installation, the aROI is 24.2%. It should be noted
that aROI is very sensitive to the installed capital. For example, if the
installed capital is decreased by using indirects and contingency of 10%
and 10% instead of 15% and 20%, the aROI reduces to 12% from 24.2%.
In general, smaller facilities experience a somewhat larger decrease in
ROI relative to larger facilities. This means that they will find it more
difficult to finance new control equipment.
Control Cost Economics. The process control economic analysis uses an
incremental cost approch. Capital and operating costs are added to the
uncontrolled process economics for those items which are necessary for
fluoride pollution control, but are not required for manufacturing. For
example, maintenance costs for the control equipment are included but
operating labor is not. The extra operating labor would be minimal and it
is assumed that the regular operating personnel would assume these duties.
Presented in the following paragraphs is a brief description of the cost
estimating procedure for the fluoride pollution control systems. The
control process economic analysis is based on the pollution control model
presented for the plant size, flows and process conditions listed under the
various "Uncontrolled Process Model" diagrams and mass balances.
The economic analysis contains the following cost elements: (1) capital
charges, (2) total operating cost and (3) total pollution control cost in
units of dollars per hour (or dollars per heat) and dollars per ton of
product.
Equipment cost (f.o.b.) of the control device is the primary input to
the economic analysis. The f.o.b. equipment cost at the required capacity
is obtained from vendor quotes, industry survey and correlations or published
cost curves. In the situations where the equipment cost quote is for a
different capacity than required, the conventional relationship is used,
3-4

-------
C2 = C1 (V2/V1)K
where
C2 = cost of equipment at the new capacity (V2)
C-| = cost of equipment at quoted capacity (V-j)
K factors for the various equipment types employed in this analysis are
presented in Table 3-2.
Auxiliary materials required to install the equipment and installation
labor costs are added to the f.o.b. equipment costs. Auxiliary materials
include foundations, electrical conduit, piping, valves, instrumentation,
and such other permanently installed materials as are required for installa-
tion of the primary control device. Installation labor includes all labor
necessary for installation. For most preliminary economic analyses, these
costs may be estimated by "field Installation factors" published in the
literature. These factors usually run between 1.5 and 3.0 times the f.o.b.
equipment costs and are presented in the economic section after each piece
of equipment.
The installation factor as presented in the control process economics is
defined by the following expression,
r-rtni. _ Installed Cost
- Equipment Cost
This definition will tend to give a low value for the factor since the
"actual installed factor" is based on the cost of the equipment constructed
of carbon steel. The special construction materials costs (usually
neoprene-1ined steel) are added to the equipment cost but do not increase
the installation costs. For example, a pressure vessel 4 feet in diameter
by 6 feet in tangent height has a base cost (carbon steel) of about $2300.
Assuming a field installation factor of 4.23 and in this case, a monel-clad
adjustment factor of 3.89, the total installed cost is given by
3-5

-------
Table 3-2. Scaling Factors (K)
Equipment	Range	K
Fan	1.5-4.0 (Mcfm)	.68
Cyclones
Heavy Duty	4-28 (Mcfm)	.82
Light Duty	2-15 (Mcfm)	.87
Multiple	1-19 (Mcfm)	.80
Multiple (In Combination
with Low Voltage Precipitator)	1-19 (Mcfm)	.83
Electrostatic Precipitator	10-2,000 (Mcfm) (Note A)	.647
Spray Scrubber	(Note B)	.60-.65
Wet Scrubber	(Note B)	.60-.65
Spray Chamber	(Note B)	.60-.65
Floating Bed Scrubber	(Note B)	.60-,65
Spray Towers	(Note B)	.60-.65
Quench Tower	(Note C)
Thickener	- „	.7
Vacuum Filter	10-2,200(ft )	.78
Venturi Scrubber	8-100 (Mcfm)	.50
Cyclonic Clarifier	100-10,000 (gpm)	.73
Waste Heat Boiler	100-10,000 (ft^)	.65
Secondary Exchanger	100-10,000 (ft^)	.65
Bag House	10-2,000 (Mcfm)	.68
Water Cooled Duct	-	.65
Spark Box	-	-.6
Dry Alumina Adsorption	Process -	.6
Spray Screen	18-96 (ft )	.97
Notes
(A) The size, installed cost, operating cost and maintenance cost for a "turn-key" single
stage industrial electrostatic precipitator were calculated from actual installed cost
correlations provided by a precipitator supplier. Installed cost includes flues,
support steel, fan, stack, low voltage wire and conduit, cooling tower, pressure and
temperature controls, dust handling system and foundations. The precipitator requires
particle sizes greater than two microns and a loading greater than 0.1 grains/ft3
(1.43 x 10-5 lbs/ft3). Electrical conductivity of particles must be between a good
conductor and a good insulator or resistivity must be between 10^ and 10^ ohm-cm.
Sizing requires the capacity (10-2,000 Mcfm), required efficiency (90-99%), a pre-
cipitation constant (between 0.05 and 0.50) and knowledge of the stream velocity. Addi-
tional required information includes the current Chemical Engineering Plant Cost Index
(C.I.), classification of the gas as corrosive or non-corrosive, and the voltage (E).
Correlations
Operating Cost ($/hr) = 0.0006 x (cfm) x 0.007
Maintenance Cost = 1300$ per year to 8000$ per year
(B)	Installed costs are estimated from graphs presented in the Guthrie article. The
primary sizing parameters are the diameter and height (costs are most sensitive to
diameter). The diameter is calculated from the gas flow (cfm) and the allowable gas
velocity. Height is calculated on an individual case (depending on scrubber type)
from data supplied by equipment manufacturers.
(C)	Quench tower cost from private sources; data is for the actual capacity required; no
scale factor was used.
3-6

-------
Equipment Cost (EC)	=	Base Cost x Material Factor
Installation Cost (IC) =	Base Cost x (installation factor - 1)
Installed Cost ($)	=	Equipment Cost + Installation Cost
or
EC = $2300 x 3.89	= $ 8,947
IC = $2300 x (4.23-1)	= $	7,429
$ = $8,947 + $7,429	= $16,376
Note that although the carbon steel installation factor is 4.23, the factor
for monel steel, is
Fartnr = installed cost _ $16,376 . ,
racxor equi-pment cost " $ 8,947 ~ ' J
Project indirect costs are added to the total pollution control system
and represent items such as sales taxes, U. S. freight, packing, insurance,
ocean freight, import duties, temporary facilities, small tools, project
engineering, fees, procurement, etc. Indirect costs vary considerably from
project to project. For this analysis, an indirect charge of 10% is assumed
for the total installed equipment cost (Reference 4386). The 10% is composed
of 3% contractor's fees, 2% engineering and 5% construction expenses. An
"on-going" company will usually have an existing engineering department which
implies additional engineering requirements should be minimal. Thus a
minimal engineering expense was selected. Contingency is taken as 10% of
capital (Reference 4386). Contingency may range between 5-15% (Reference
4386) and one reference takes it as 34% of capital (Reference 4383). Since
the control systems utilize known technology, 10% was selected and is
probably on the "High" side. Note, that an incremental approach was taken
1n this analysis, that is, only costs were included in the control economic
estimates that are in addition to normal production costs. For example,
corporate engineering is not considered since it is already taken as a part
of the overhead in the uncontrolled process model.
3-7

-------
The capital charge is taken as the fraction of capital per year which
corresponds to the straight line depreciation schedule permitted for the
industry by the U. S. Internal Revenue Service in June, 1971. Start-up
costs and interest on the investment during construction were not capitalized
in order to make control process economics consistent with uncontrolled
production economics.
Operating costs were calculated on an item by item basis. That is,
operating cost corresponding to the operation of a unit is calculated and
charged to that unit. For example, if a dry multiple cyclone cleans
60,000 cfm at 68°F with an inlet dust-loading of 5 grains/cu ft, its
operating cost is estimated by knowing its pressure drop; 4.3 in w.g.
pressure drop (Reference 4387). The following relationship gives the
required horsepower (Reference 4388).
Horsepower = °-0158 xecfm " M9
where
cfm = volume at temperature in cubic feet per minute
Wg = pressure or suction in inches water gage
e = percent fan efficiencies
or
Horsepower = 68
For most initial economic analyses, the operating costs for a limited
number of pumps may be omitted (such as in the control system case). As a
worst case example, a wet high energy venturi with a capacity of 60,000 cfm
uses an eleven horsepower pump to feed the required 480 gpm. The operating
cost for the pump is less than $.06 per hour. This is insignificant when
compared to the $2.68 per hour operating cost for the 31.5 in. w.g. pressure
drop in the gas stream.
3-8

-------
The following cost factors for maintenance were utilized in determining
operating costs.
Unit	$ Per Year
Bag House	10,000
Cyclone	1,000
Cyclonic Clarifier	1,000
Cyclonic Spray Tower	1,000
Electrostatic Precipitator	3,000
Fan	3,600
Floating Bed Scrubber	2,000
Lime Pit	3,000
Liquid Solid Separator	500
Quench Tower	1,600
Radiant Cooling Coils	800
Secondary Heat Exchanger	800
Liquid Solid Separator	400
Spray Chamber (Scrubber or Tower)	2,000
Spray Screen	2,500
Spark Box	1,000
Thickener	500
Vacuum Filter	2,600
Venturi Scrubber	1,000
Venturi Separator	1,000
Waste Heat Boiler	870
Water Cooled Duct	800
Wet Cyclone	1,000
Wet Scrubber	2,000
3-9

-------
CONTROL PROCESS ECONOMICS
FOOTNOTES
(1)	Includes piping, concrete, steel, instruments, electrical, insulation,
paint, etc., site preparation and installation labor.
(2)	Includes project engineering, process engineering, design and
drafting, procurement, temporary facilities, construction equipment,
insurance, sales and other taxes, etc.
(3)	Supplied by steelmaking furnace.
(4)	Assuming 25 prebaked anode, or horizontal stud
Soderberg cells per electrostatic precipitator.
(5)	Assuming 10 prebaked anode or horizontal stud
Soderberg cells per scrubber.
(6)	Includes the liquid-solid separator cost.
(7)	Included in the installed capital cost.
(8)	Assuming 10 prebaked anode or vertical stud Soderberg cells per
control process.
(9)	Assuming a 50,000 tpy horizontal stud Soderberg plant and 15 cells
per tower.
10)	Includes the cost of the liquid-solid separation unit.
11)	Assuming a 50,000 tpy prebaked anode or vertical stud Soderberg plant.
12)	Includes reactors, fans, alumina handling equipment, site preparation
and indirects.
13)	Includes power and maintenance and a net credit for recovery of
reusable material.
14)	Assumes a 50,000 tpy plant.
15)	Includes liquid solid separation unit.
16)	Assuming an initial cost of $50,000 and 0.5 men to maintain it
(at $6000 per year).
17)	Pollution control costs could be reduced to 3.40 + 2.00 $/ton A1 if
fluoride credit is taken and if larger electrostatic precipitators
are used.
18)	At 0.007 $/kwhr.
19)	200 tons per heat; 12 heats per day.
3-10

-------
(20) The size, installed cost, operating cost and maintenance cost for
a "turn-key" single-stage industrial electrostatic precipitator were
calculated from correlations provided by a precipitator supplier.
Installed cost includes flues, support steel, fan, stack, low voltage
wire and conduit, cooling tower, piping, pressure and temperature
controls, dust handling system and foundations. The precipitator
requires particle sizes greater than 2 microns arid a loading greater
than 0.1 grain/ft3 (1.43 x 10-5 lb/ft3). Electrical conductivity of
particles must be between a good conductor and a good insulator or
resistivity must be between 10^ and 10^0 ohm-cm. Sizing requires the
capacity (10-2,000 Mcfm), required efficiency (0 to 99%), a precipi-
tation constant (between 0.05 and 0.50) and knowledge of the stream
velocity. Additional required information includes the current
Chemical Engineering Plant Cost Index (C.I.), classification of the
gas as corrosive or noncorrosive, and the voltage (E).
Correlations
Operating Cost ($/hr) = 0.0006 x (cfm) x 0.007
Maintenance Cost	=0.63 $/hr for noncorrosive gas
1.00 $/hr for corrosive gas
(21)	At $0.02 /1000 gal.
(22)	At $0.50 /ton.
(23)	Company private design information.
(24)	200 tons per heat; 10 hours per heat, 2 heats per day.
(25)	$6.20/ton Ca(0H)2.
(26)	Furnace capacity 75 tons/heat; 4 heats per day.
3.1.2 Thermochemical Analysis Approach
As indicated in Section 1, significant uncertainty is associated with
determination of the evolution and emission rates from many of the processes
of interest. In the absence of definitive experimental data, an analytical
approach based on thermodynamic equilibrium calculations was utilized.
TRW has applied its proprietary Chemical Analysis Program (CAP) to
evaluation of pollutants from a number of high temperature processes, such
3-11

-------
as combustion of coal, manufacture of glass, and smelting of iron and steel.
The desirability of this analytical approach was fourfold. First, in cases
where experimental measurements of pollutant output were not available, the
calculated values provided a basis for evaluating the pollution potential of
the process or industry considered. Second, even when experimental data
were available for a given process output, the chemical composition of the
pollutant stream was, on occasion, not measurable, e.g., both HF and SiF^
were measured and reported as gaseous fluoride, and trace components,
possibly toxic, were not measured at all. The CAP provided a component by
component breakdown including trace components. Third, where only limited
data were available, the theoretical approach was employed to verify or
modify the conclusions on fluoride evolution. Fourth, consideration of the
formation mechanisms of the various pollutants yielded valuable information
on potential abatement and control strategies.
All equilibrium calculations were performed using TRW's proprietary
Chemical Analysis Program with thermochemical data derived from JANAF and
similar high quality sources. Because of the complexity of the chemical
systems under consideration (over 300 possible gaseous and over 100 possible
condensed species) the analyses were performed on various combinations of
the systems elements to derive the final product distribution. All analyses
performed contained, however, the basic components as defined 1n the various
input mass balances. The nitrogen was replaced in the analyses by helium
after initial calculations indicated no production of trace element nitrogen
compounds from the nitrogen in the air or coal.
Illustrative of the CAP approach to a particular problem is the analysis
of coal combustion in power plants. The primary objective was to calculate
the forms and quantities of fluorides evolved. A secondary objective was to
determine the fate of heavy metal constituents and their potential as toxic
emissions. The results of this evaluation, as well as those for opal glass
manufacture and the smelting of iron and steel, are presented in the text of
the report.
3-12

-------
3.2 THE PRIMARY ALUMINUM SMELTING INDUSTRY

-------
3.2 PRIMARY ALUMINUM SMELTING INDUSTRY
3.2.1	General
Primary aluminum production is both a present and a projected major
source for soluble fluorides emitted to the atmosphere. The fluorides
emitted are mainly gaseous hydrogen fluoride (HF), particulate cryolite
(Na3AlFg) and aluminum fluoride (AlF^). Virtually all of the soluble
fluorides emitted by the industry come from the reduction process, which
thus merits detailed examination.
This section covers those general considerations for which the primary
aluminum smelting, industry may be considered as an entity. Such considerations
include: reduction technology; current and projected levels of soluble
fluorides discharged to the atmosphere; current and projected production levels;
fluoride emission control systems; and the impact of fluoride emission control
systems on return on investment (ROI).
3.2.2	Industry Description
Aluminum is manufactured by the electrolytic reduction of alumina
(A12^3) dissolved in a molten salt bath. The molten salt "bath" is com-
posed of cryolite (Na^AlFg) and various salt additives. The electrolysis
is performed in a carbon crucible, housed in a steel shell, using the
crucible as cathode, and using carbon anodes. The reduction cells are
referred to as "pots," and there are three basic types of pot in use
today. These are the prebake (PB), the horizontal stud Soderberg (HSS),
and the vertical stud Soderberg (VSS). Figure 3-1 illustrates the three
cell types.
As indicated by Figure 3-1, the three pot types differ chiefly
in their anode configurations. The prebaked anode cell, as indicated
by its name, employs a replaceable, consumable carbon anode, formed by
baking prior to its use in the cell. Both Soderberg pots employ con-
tinuously formed consumable carbon anodes where the anode paste is baked
by the energy of the reduction cell itself.
3-13

-------
CO
I
* REMOVABLE CHANNELS
t
ALUMINA HOPPER
SOLIDIFIED CRUST OF
ELECTROLYTE AND
ALUMINA
STEEL SHELL
INSULATION*
GAS COLLECTION DUCT
GAS COLLECTION HOODS
SOLIDIFIED CRUST OF
ELECTROLYTE AND ALUMINA
ALUMINA
HOPPER
ANODE BUS bar
CARSON ANODES
GAS AND FUME EVOLVING
STEEL SHELL
INSULATION
ELECTROLYTE
MOLTEN ALUMINUM
CATHODE COLLECTOR BAR
^CARBON LINING.
\ ^ ^
BUS BAR
PREBAKED
RI5ERS
STUDS
GAS COLLECTION DUCT
ANODE CASING
ANODE PASTE
TO GAS
POT ENCLOSURE DOOR
7 7 "
ANODE
^ PASTE/
TREATMENT
PLANT
PARTIALLY BAKED
/PASTE
BURNER
ANGOE STUDS
GAS AND FUME EVOLVING

gas AND TAR
BURNING
SOLIDIFIED CRUST f
OF ELECTROLYTE
AND ALUMINA
GAS EVOLVING
FULLY BAKED ANODE
BAKEDANODE
STEEL SHELL
aEajDlYTE
MOLTEN ALUMINUM
ELECTROLYTE
MO LT E N~AL UMIN UM
CATHODE
COLLECTOR BA&
CAT HO DE
COLLECTOR BAR
CARBON LINING
x V \ \ \
CARBON LINING
\ \ V \ \
HORIZONTAL SODERBERG
THERMAL
INSULATION'
VERTICAL SODERBERG
Figure 3-1. Details of Prebaked and Soderberg Aluminum Reduction Cells

-------
Present American production practice makes extensive use of prebake
cells and horizontal stud Soderberg cells. Although the Soderberg cell was
hailed as an economic advance when developed, currently, the trend in new
plant design is to use the prebake cells because of operating difficulties
associated with condensed tars from the volatilization of Soderberg anode
paste pitch.	'
The "pot" contains both the molten electrolyte (bath) and molten
aluminum. The bath, composed, as noted above, of cryolite and additives,
is less dense than the molten aluminum. Because of this, the molten
aluminum collects as a layer of liquid in the bottom of the cell. The
function of the molten cryolite layer is threefold - to act as an elec-
trolyte, to dissolve the alumina charged to the cell, and to protect the
aluminum from the atmosphere. Bath temperature is approximately 950°C.
During production, the carbon of the anode is oxidized (about 0.45 to
0.55 pound per pound aluminum) by the oxygen released from alumina in
electrolysis, but the carbon of the cathode is not oxidized. Detailed
information describing the production of aluminum is available in
Reference 4164.
The source of fluorides emitted to the atmosphere by aluminum
reduction is the fluoride electrolyte which contains cryolite, ATF3, and
fluorspar. Cryolite can be represented as a complex of 3NaF ' A1F3. For
this molecular proportion, the weight ratio of NaF to AlF^ is 1.50. In
practice, this ratio is maintained in the range of 1.36 to 1.43 by addition
of ^CO^, NaF, or AlFg. The alumina feed material contains about 0.25%
NagO (Reference 4164) which results in a continuing requirement for A1F3
addition in normal cell operation to hold the ratio in the operating range.
Fluorspar is added to the bath to depress the freezing point of cryolite and
allow operation at lower temperatures than would otherwise be possible.
Usage of fluoride containing electrolyte ranges from 0.03 to 0.10 pound/pound
aluminum.
A portion of the feed fluorides evolves from the pot as both gaseous
and particulate material. At the same time, the pot lining absorbs fluorides
3-15

-------
which may be recovered after the useful life of the pot is expended (about
3 years). The amount of fluoride absorbed in the linings has apparently
not been accurately determined and reported. It is estimated by TRW as
approximately 50% of the total input fluoride based on conversations with
aluminum industry personnel.
Effective capture of the off-gases and fumes for subsequent clean-
ing is the most difficult single technical problem. Two approaches to this
problem have been taken. The first approach, typical of U.S. practice,
utilizes a hood over each pot to collect the fumes evolved. However, hoods
must allow access for operations in the pot such as ore addition, anode
replacement, and aluminum removal. Thus, a portion (frequently fairly
large) of the off-gases escapes into the pot-line building and then escapes
to the atmosphere through the building roof ventilators (monitors) (Refer-
ence 4288). In some cases, the monitor output is controlled in addition
to the hood output. A need exists for hood systems that will capture the
fumes and gases evolved while minimizing the ingested air to allow cleaning
at high efficiency and minimum cost.
The second approach, used to a minor extent in the U.S. and more
frequently in European practice, allows free flow of the pot fumes into
the pot room and utilizes the monitor system to capture the entire flow
of room ventilation air including the fumes. This approach requires about
ten times as much air to be processed as the hood system and sometimes
results in higher fluoride exposures for the workers.
Raw material and energy requirements for aluminum production are
presented in Table 3-3. These values vary slightly from plant to plant, but
they are generally consistent with modern practice.
3.2.3 Production Trends
The aluminum industry is a major current source of atmospheric
fluorides. The expansion of reduction facilities at planned and existing
plants will increase these emissions to a level which will make this industry
3-16

-------
the prime emitter of fluorides by the year 2000. Some aluminum reduction
processes under development would present no fluoride problem; however, there
is no indication that any proposed new process will come into wide service in
the next several decades, if at all (Reference 4288).
Table 3-3. Raw Material and Energy Requirements for Aluminum Production
Amperes through pot line
60,000 - 225,000
Voltage drop per cell
4.3 - 5.2
Current efficiency, percent
85 - 90
KWH/pound aluminum
6.0 - 8.5
Pounds Al203/pound aluminum
1.89 - 1.92
Pounds fluoride electrolyte/pound
aluminum
0.03 - 0.10
Electrode carbon,
pound/pound aluminum
0.45 - 0.55
It can be stated with reasonable certainty that U.S. primary alu-
minum production will increase at least threefold between now and 1984.
Most major producers expect to reach this projection. A recent survey
found projected growth rates to range from a high of 9.5 percent to a low
of 4.8 percent annually through 1980 (Reference 4250). Other sources
indicated rates of 6.4 percent (Reference 4289) and 7.0 percent (Refer-
ence 4290) through 1980. If an expected value of 7.4 percent (Reference
4250) is extrapolated to the year 2000, the production of aluminum will
increase from 4.0 million tons (Reference 4290) (1970) to 35 million tons.
Year 2000 projections may be seen in Table 3-4. Current and projected
production rates for each of the three basic processes are shown in
Table 3-5.
3-17

-------
Table 3-4. Estimated Future Electrolytic Aluminum Production

1947
1967

Past
20-Year
Annual /flx
Growth Rate^
2000^
Estimated/R\
Year Annual* '
Growth Rate
U.S. Population
144.1 MM^
199.1 MM

1.7%
336.2 MM
1.6%
GNP (constant 1967
dollars)
356.8 MMM
785.1 MMM

3.9%
3.15 MMMM
4.3%
Primary Aluminum
Production
0.57 MM tons
3.27 MM
tons
9.1%
34.5 MM tons
7.4%
Total Consumption
0.95 MM tons
4.15 MM
tons
7.7%
38.7 MM tons
7.0%
Per-Capita Consumption
13 lb
42 lb

5.8%
230 lb
5.2%
(A)
v 'Growth rates are compunded annually.
^Data extrapolated from 1984 projections given in Reference 4250.
^M represents 1000 — MM represents 1 ,000,000, etc.

-------
The projections in Table 3-5 are based on the assumption that the
actual percentages produced by each process will remain the same during the
period 1970 to 2000. While this may seem inconsistent with the current
trend toward installation of PBA facilities noted below, past experience in
the aluminum industry indicates that extrapolation of trends can lead to
possibly erroneous conclusions on which single technology would become
predominant. As an example, the prior strong American trend to Soderberg cells,
which still predominate in Europe, went through a reversal in 1970. In 1970,
38% of the U. S. aluminum reduction capacity employed the Soderberg electrode -
a drop from 46% in 1965 - with only 10% of the new capacity planning to use it.
To avoid erroneous conclusions, it was considered a reasonable
compromise to project the current distribution to the year 2000. Requirements
for air pollution control may strongly influence the actual distribution
practiced by the industry.
Table 3-5. Aluminum Production
Production
Process
Percent of Total
Aluminum Production
1970
Aluminum Production
(million tons/year)
(A)
Estimatedv ' Aluminum
Production in 2000
(million tons/year)
PB
61.9%
2.48
21.7
HSS
25.5%
1.02
8.9
VSS
12.6%
0.51
4.4
Total
100.0%
4.01
35.0
(A)
v 'Assumes no change in 1970 process distribution
3-19

-------
A significant factor not connected with demand that could easily
lower the projected U.S. output would be a move by U.S. producers to add
primary capacity outside of the U.S. -- near bauxite deposits, for
instance -- in order to reduce costs, and then import the resultant metal.
If world tariffs continue to decline, this would appear to become more
and more attractive. Also, constantly increasing labor costs could force
producers to move outside the U.S.
3.2.4 Fluoride Control and Emissions Summary
Both gaseous and particulate fluorides are evolved from the
electrolytic cells producing aluminum. Dry dust collection devices
(centrifugal collectors, miltitube cyclones, or electrostatic precipi-
tators) and/or wet scrubbers may be used as control devices. Individual
pot hoods are "ganged" (manifolded) to feed the control devices. When
the hood access panels are opened, the air flow into the hood system
may be insufficient to prevent escape of gas and particulates from the
cell into the pot room. Hood capture efficiencies may drop to levels
much below the normal operating range. Individual hood exhaust flow
rates range from 1000 to 5000 cubic feet per minute per cell.
Scrubbing equipment commonly employed in alutoinum manufacture includes
redwood scrubbing towers, floating bed scrubbers, and wet cyclonic scrubbers.
In addition, the Aluminum Company of America (ALCOA) has developed a dry
adsorption process for the collection and recovery of reduction cell fluorides
(Reference 323). The process exposes the alumina feed to the pots to the
evolved fumes. The fluorides adsorb on the alumina and are returned to the
pots. A baghouse is used to prevent solids loss; Efficiency is reported to
be about 97- to 99% for both gaseous and particulate fluorides. Although
ALCOA has developed and patented their particular "398" process, the use of
alumina as a dry adsorbent in other equipment configurations has been studied
previously. There is some question about the effect of tars on the dry
adsorption process, buth ALCOA reports that it is suitable for both prebake
and Soderberg cells.
3-20

-------
Roof monitor scrubbers were not considered in detail in this study
since they represent a minority usage and the trend in the industry is
expected to be toward more efficient pot hoods.
Controlled process models for each of the three processes are
presented in the individual process description sections.
Soluble fluorides evolved from the primary aluminum smelting processes
averaged 46 pounds (as F) per ton of product in 1970 (Reference 4208). This
corresponds to an annual fluoride evolution rate of 92,200 tons in 1970. The
soluble fluoride emission factor for the year 1970 for the aluminum reduction
plants averaged 8.1 pounds (as F) per ton of product aluminum (Reference 4208),
equivalent to the annual discharge of 16,230 tons (as F) of soluble fluorides
to the atmosphere. The ranges of data presented in the referenced 0AP Study
appear to be consistent with those of the other data sources analyzed and to
include a much greater portion of the American industry in the data base. It
must be noted, however, that the source of the data was an industry question-
naire which probably does not represent emissions during less than optimum
conditions.
Projections for the year 2000 indicate that, if current control
practice continues, 141,000 tons (as F) per year of soluble fluorides will
be emitted to the atmosphere by the aluminum reduction plants. This would
be by far the largest industry source of soluble fluoride pollutants. If
improvements in control practice to achieve 99 percent control are adopted
by the industry, the projected emissions of soluble fluorides for the
year 2000 would drop to 8,050 tons (as F). Table 3-6 presents the
process and industry current totals and projections..Fop fluoride emission
values from various cell types, the basic data utilized was that of
Reference 4208, an interim report dated May 10, 1971 by Singmaster and
Breyer on the aluminum industry. These data are subject to change since the
Singmaster and Breyer program is not completed.
Dust evolution from handling of raw materials is not considered in
detail in the above analysis because the magnitude is small (about 1 to 6
3-21

-------
pounds fluorine/ton aluminum), the emission control devices are relatively
efficient, and the environmental effect is low. Much of the dust may settle
in the plant and be recovered, thereby not constituting an emission to the
surroundings.
Table 3-6. Fluoride Emissions From Aluminum Production Processes
Process
1970 (A)
Soluble Fluoride
Emissions
(103 tons F /Year)
2000 (B)
Soluble Fluoride
Emissions
Current Control
Practice
(103 tons F /Year)
2000 (B)
Soluble Fluorides
Emitted with
99% Control.
Technology
(103 tons F /Year)
Total
16.23
(A)
v 'Total A1 production was 4.01 million tons annually in 1970.
^Projected A1 production	of 35 million tons annually reflecting an
expected growth rate of 7.4%.
(C)
v 'Includes prebaked anode	furnace emissions.
3.2.5 Process Description
3.2.5.1 Prebaked Anode Aluminum Production
Present and Future Production Levels. Aluminum is currently produced
by the prebaked anode process at 19 locations (including those under
construction) throughout the U. S. A detailed listing of these locations can
be found in Appendix 7.2. The prebaked anode process will grow in production
level from 2.48 million tons of aluminum in 1970 to 21.7 million tons in 2000.
3-22

-------
Process Description. As noted earlier, the important difference
between prebake and Soderberg cells is the method of replacing spent anode.
For prebaked anode cells, prior to utilization in the reduction cell, anode
paste is solidified to block form in a separate baking furnace. These
individual blocks of carbon—typically 14 to 24 per cell—attached to metal
rods serve as replaceable anodes. These burn away at a rate of about an
inch per day. Spent anode assemblies are removed on a rotating basis,
usually two at a time.
An uncontrolled process model of the prebaked anode reduction process
is shown in Figure 3-2. The mass balance is based on an 89% hood capture
efficiency (actual experience as reported in Reference 4208). The balance
of fluoride not shown is allocated to cathode absorption.
Fluoride Emission Control Techniques. The fluoride emission control
techniques employed are illustrated in the flow diagrams and mass balances,
Figures 3-3 and 3-4 for the prebaked anode cell and the prebaked anode furnace
control systems, respectively. For cell effluent control, process C is
currently most prevalent but process D shows promise of greater acceptance
in the future. For baking furnace effluent, processes A and B are generally
used in combination, in series.
Fluoride Emissions. As shown in Table 3-6, the prebake anode process
currently (1970) emits 8610 tons of fluoride annually corresponding to about
53% of the total from the industry. This emission level reflects an
industry-wide average for prebake anode process fluoride abatement of 85%.
If expected production levels arereached in the year 2000, the emission from
this source will be 75,000 tons (as F~) annually, assuming extension of
current control practice. If 99% control efficiency technology is utilized
industry-wide, the fluoride emitted from this source would be about 5000 tons
annually.
3-23

-------
TO ATMOSPHERE OR
POLLUTION CONTROL
DEVICE
PREBAKED ANODE CELL
FAN
TO ATMOSPHERE OR
POLLUTION CONTROL DEVICE
EXHAUST PLUS DILUENT AIR
HOOD
USED ELECTRODES
HOOD LOSS TO
ATMOSPHERE OR
CONTROL DEVICE
ANODE
PASTE
REDUCTION
CELL
1790°F
ELECTRICAL
POWER
FEED
MATERIALS
MOLTEN
ALUMINUM
1. HOOD EFFICIENCY IS 89%
HOOD	 -
(AVERAGE OF OLD AND NEW
PREBAKE ANODE CELLS)
BASIS - 2070 lb ALUMINUM PRODUCED PER DAY PER AVERAGE PREBAKE ANODE CELL
PROCESS STREAMS - LBS/HR.
Material
Stream Number

1
2
3
4
5*
6*
7*
,HF




t.o^M)
•09(g)
.04(g)^
NajAl F6
aif3
2.5{s)(C)
3.3(s)(C)



).73(s) (D,B)
(as F
1 .13(s)
f as F
*
Total Fluoride^
5.8





.04
Total as F^
3.6



1.77
.21
.04
A1 umi na
165



l(s>
.l(s)

Anode Material as C

50





A1umi num



86



CO 2




170(g)
17(g)

H90




Kg)
.05(g)

C




-2{s)
.01(s)

Electrical Power


660/KWH




Approx. Total
S tream
170
50
-
86
170(a)
17
.04^
* Gaseous effluent stream
(A)	Plus 215,000 SCFH of diluent air (STP = 32°F, 14.7 psia)
(B)	Plus other exhaust gases from prebake oven
(C)	Ref. 4250
(D)	Ref. 4254
(E)	Ref 4208
(F)	The balance not shown is cathode absorption of electrolyte
(£)
Soluble F Evolution Factor = 46 lb F/ton Aluminum Produced
Figure 3-2. Electrolytic Aluminum
Production--Prebaked Anode
Cell, Uncontrolled Process
Model
3-25

-------
ELECTROSTATIC PRECIPITATOR
EFF, = 98% PARTICULATE
0% GASEOUS (ASSUME NO ADSORPTION)
FLOATING BED SCRUBBER
EFF. = 72% PARTICULATE
95% GASEOUS

2I2°F
il

¦ TO STACK
WATER -
PROCESS A
i>.
TO DISPOSAL
OR RECYCLE
212°F (EST.)

- TO STACK
SPRAY SCREEN
EFF, = 83% PARTICULATE
96% GASEOUS
PROCESS
TO LIQUID - SOLID
SEPARATION UNIT
WITH SOLIDS DISPOSAL
AND LIQUID RECYCLE
DRY ALUMINA ADSORPTION

QUENCH
TOWER
•-TO STACK
WATER


PROCESS C
WATER
BAG
FILTERS
TO LIQUID - SOLID SEPARATION
UNIT WITH SOLIDS DISPOSAL
AND LIQUID RECYCLE
ai2o3 + aif3
RECYCLE TO CELL FEED
FLUIDIZED
BED
FAN
BASIS - 2070 LBS. ALUMINUM PRODUCED PER DAY PER PREBAKED ANODE CELL
PROCESS STREAMS - LBS,/HR.
Material
Stream Number
5
6*
8
9*
10
11*
12
13*
14
15
16*
HF
:Ja3Al Fg
A1F3
1.09(g)
) . 7 3 (s )
) as F
-09(g)
) •13(s )
j as F
j .715(s )
j as F
1.09(g)
) . 015(s)^
j as F
) . 5 3 (s )
j as F
-06(g)(C)
) .20(s)'C'
j as F
) ¦61(s)
I as F
- 04(g)(C)
) .12(s)(C)
J as F

| 1,75(s)^
j as F
.01(g)(C)
) .007(s )
| as F
Total as F
1.77
.21
.715
1.06
.53
.26
.61
.16

1.75
.017
A1umina
co2
h2o
C
Ks)
170(g)
Kg)
• 2(s)
•	Ks)
17(g)
•05(g)
•	01(g)
'
.98(s)
.196(s)
• 02(s)^
170(g)
Kg)
.004(s)(C'
•72(s)
•14(s)
,28(s)(C)
170(g)
400(g)
• 06(s)^
•83(s)
• 17(s)
•17(s)(C>
170(g)
400(g)
.03(s)^
20(s) (Est)
21(s) (Est)
.198(s)
.01 (s)
170(g)
Kg)
¦002(s)(C)
Approx. Total
Stream
170
8
1
170(a)
q(B)
570(a)
-| { B)
570
zo
21
1?0^A'
~Gaseous Effluent Stream
(A)	Plus 215,000 SCFH Diluent Air (STP = 32°F, 14.7 psia)
(B)	Plus Water and Soluble Fluorides
(C)	Ref. 4208

Calc. Soluble Fluoride
Emission Factor =
lb F/ton A1
Source
Process A Process B
Process C
Process D
Treated Hood Exhaust
Emission
24.65 6.04
3.72
0.40
Untreated Hood Loss to
Atmosphere
4.88 4.88
4.88
4.88
Total Emission to Atirosphere
fron Cel 1
29.53 10.92
8.60
5.28
Reported Average Overall Prebake Anode Cell Soluble Fluoride Emission Factor = 6.94 lb F/ton A1^
(Reflects the utilization of series abatement systems by a significant portion of the industry)
Figure 3-3. Electrolytic Aluminum
Production - Controlled
Process Model Prebaked
Anode Cell
3-27

-------
ELECTROSTATIC PRECIPITATOR
EFF. = 98% PARTICULATE
0% GASEOUS (ASSUME NO ADSORPTION)
SPRAY TOWER
EFF. = 70% (EST.) PARTICULATE
90% (EST.) GASEOUS

u

•TO STACK

TO DISPOSAL
PROCESS A
WATER



TO STACK
<§>
TO LIQUID-SOLID
SEPARATION UNIT
WITH SOLIDS DISPOSAL
AND LIQUID RECYCLE
PROCESS B
BASIS - 2070 LB ALUMINUM PRODUCED PER DAY PER PREBAKED ANODE CELL BAKING FURNACE
PROCESS STREAMS - LB/HR
Fluoride Balance Only
Material
Stream Number
7
8
9*
in
11*
HF
hf-xh2o
•04(g)
0
¦04(g)(Est)
.03fi(l)(Est)
as F
.n04(g)(Est)
lotai Muoride
Total as F
.04
.04
0
0
.04
.04
.036
.004
.004
Approx. Total
Stream
.04'A'
n
-------
3.2.5.2 Horizontal Stud Soderberg Aluminum Production
Present and Future Production Levels. In 1970, the horizontal stud
Soderberg process accounted for 25.5% of the total domestic aluminum
production. This corresponds to a total of 1.02 million tons annual
production from eight production sites located within the U. S. A detailed
summary of these facilities is presented in Appendix 7.2.
Projected aluminum production levels for the year 2000 for the
horizontal stud Soderberg process are 8.9 million tons of annual production.
Process Description. The horizontal stud Soderberg cell uses a
continuous anode. A mixture of pitch and carbon aggregate called "paste"
is added at the top of the superstructure periodically, and the entire
anode assembly is moved downward as it burns away. The result of baking the
paste in place is to add heavy organic material (tars) to the cell effluent
stream.
In the horizontal stud Soderberg (HSS) cell, the anode is contained
by aluminum sheeting and steel channels. The channels are perforated with
holes about 3 inches in diameter, and the "studs" or electrode connections
are inserted through these holes into the anode while it is 3 feet or so
above the molten bath and is still fairly soft. The anode is baked solid in
the region just above the molten bath by heat from the process. As the
anode is lowered, the bottom channel is removed after the lower row of studs
is pulled out and the flexible electrical connectors are moved to a higher
row. This process requires significant mechanical manipulation with the
hood door open thereby reducing hood capture efficiency.
An uncontrolled process model of the HSS cell is presented in
Figure 3-5. The mass balance is based on an 86% hood capture efficiency
3-30

-------
HORIZONTAL STUO SODERBERG CELL
TO ATMOSPHERE
OR POLLUTION CONTROL
DEVICE
PAN
EXHAUST PLUS DILUENT AIR
HOOD
HOOD LOSS TO
ATMOSPHERE
OR CONTROL
DEVICE
ANODE
PASTE
REDUCTION
CELL
!7V0°f
(HSS)
MOLTEN
ALUMINUM
1. HOOD EFFICIENCY IS 86%
BASIS - 2070 lb ALUMINUM PRODUCED PER DAY PER HSS CELL
PROCESS STREAM - lbs/lir
Material
Stream Number

1
2
3
4
5*
6*
HF




'• 1.06 (g) (C,D)
•ii (g)
Na3A1F6
A1F3
(B)
2.5 (sf
3.3 (s)



I .69 (s)(D)
(as F
(.17 (s)
( as F
Total Fluorides
Total as F
5 8(E)
•(E)
3.6



1.70
.28
Alumina
Anode Mat'l. as C
Aluminum
co2
h2o
C
Electrical Power
165
50
660 KWH
86
1 (s)
160 (q)
1 (g)
.2 (g>
.1 (s)
27 (g)
.1 (g)
.02 (g)
Approx. Total
Stream
170
50
--
86
160(A)
27
* Gaseous Effluent Stream
(A)	Plus 250,000 SCFH of diluent air
(B)	Ref. 4250
(C)	Ref. 4254
(D)	Ref. 4208
(E)	The balance not shown is cathode adsorption of electrolyte
Soluble F Evolution Factor = 46 lb F/ton Aluminum Produced'0'
Figure 3-5. Electrolytic Aluminum Production - Uncontrolled Process
Model, Horizontal Stud Soderberg Cell

-------
(actual experience reported in Reference 4208). The balance of the fluoride
not shown is allocated to cathode absorption of electrolyte.
Fluoride Emission Control Techniques. The controlled process models
and mass balances for the HSS process are presented in Figure 3-6.
Predominant usage is Process C, Principal problem areas are blockage of
equipment by tar buildup and low. fluoride collection efficiency due to the
decreased hood capture efficiency noted above.
Fluoride Emissions. Soluble fluoride emission from the horizontal
stud Soderberg was 5160 tons (as F~) in 1970 (Reference 4208). This is
equivalent to about 32% of the fluoride emitted industry-wide and is based
on an industry-wide .HSS process fluoride abatement efficiency of 78%
(Reference 4208). The emission level will increase to 45,000 tons (as F")
in the year 2000 if current projected production levels are reached and if
current control efficiency is maintained. If 99% control technology is
utilized industry-wide, the fluoride emission level from this process will
drop to 2050 tons annually.
3.2.5.3 Vertical Stud Soderberg Aluminum Production
Present and Future Production Levels. Four facilities in the U. S.
(see Appendix 7.2) currently utilize the vertical stud Soderberg process in
the production of aluminum. They accounted for 0.51 million tons--12.6% of
the total production—in 1970. Projected production levels are 4.41 million
tons of aluminum produced via the VSS process in the year 2000. These data
are presented in Table 3-5.
Process Description. The vertical stud Soderberg is similar to the
horizontal stud Soderberg with the exception that the studs are mounted
vertically instead of horizontally in the cell. The studs must be raised and
replated periodically but that is a relatively simple process. The tar
problem is alleviated as discussed below. The uncontrolled process model of
the VSS is presented in Figure 3-7. The mass balance is based on an 80% hood
3-32

-------
FLOATING BED SCRUBBER
EFF. = 72% PARTICULATE
95% GASEOUS
ELECTROSTATIC PRECIPITATOR
EFF. = 98% PARTICULATE
0%
GASEOUS (ASSUME NO ADSORPTION)
TO STACK
TO STACK
WATER
212
PROCESS B
PROCESS A
TO LIQUID - SOLID
SEPARATION UNIT
WITH SOLIDS DISPOSAL
AND LIQUID RECYCLE
TO DISPOSAL
OR RECYCLE
WATER -
SPRAY TOWER
EFF, = 70% (EST.) PARTICULATE
90% (EST.) GASEOUS
4>
212°F
(EST.)

TO STACK
PROCESS C
TO LIQUID - SOLID
SEPARATION UNIT WITH
SOLIDS DISPOSAL AND
LIQUID RECYCLE
BASIS - 2070 lb ALUMINUM PRODUCED PER DAY PER HORIZONTAL STUD S0DERBERG CELL
PROCESS STREAMS - LB/HR
Material
Stream Number
10*
11
U*
HF
Na3A1F6
A1F„
1.06(g)
.69(s)
as F
.11
¦17(s)
as F
.68(s;
as F
1.06(g)
.01(s)^
as F
.50(s]
as F
•05(g)
•19(s)
as F
(cT
(c)
,48(s}
as F
.11(g)(Est)
.21(s)(Est)
as F
Total as F
1.70
1.02
.24
.48
,31
A1umi na
co2
h2o
Ks)
160(g)
Kg)
• 2(g)
•	Ks)
¦27(g)
•	Kg)
.02(g)
,98(s)
. 196(s)
. 02(s
160(g)
Kg)
.004(s)
,(C)
;c)
.73(s;
.15(s)
.27(s)(C)
160(g)
450(g)
.05(s)
(C)
• 7(s)
.14(s]
• 3(s)
160(g)
450(g)
.06(s)
(C)
Approx. Total
Stream
160
(A)
27
160
(A)
(B)
610
(A)
(B)
610
(A)
~Gaseous Effluent Stream
(A)	Plus	250,000 SCFH of diluent air
(B)	Plus	soluble fluorides and water
(C)	Ref.	4208	;

Calc. Fluoride
Emission Factor -
lb F/ton A!
Source
Process A
Process B
Process C
Treated Hood Exhaust
Emi ssi on
23.72
5.58
7.20
Untreated Hood Loss to
Atmosphere
6.51
6.51
6.51
Total Emission to Atmosphere
from Cell
30.23
12.09
13.71
Reported Average Overall Horizontal Stud Cell Soluble Fluoride Emission
Factor = 10.12 lb F/ton Al(w. (Reflects the utilization of series
abatement systems by a significant portion of the industry)
Figure 3-6. Electrolytic Aluminum Pro-
duction -- Controlled Pro-
cess Model, Horizontal Stud
Soderberg Cell
3-33

-------
FAN
EXHAUST PLUS DILUENT AIR
HOOD
HOOD LOSS
TO ATMOSPHERE
OR CONTROL
DEVICE
ANODE
PASTE
FLARED
EXHAUST
REDUCTION
CELL
1790°F
(VSS)
ELECTRICAL
POWER
FEED _
MATERIALS "
MOLTEN
ALUMINUM
1. HOOD EFFICIENCY IS 80%
BASIS - 2070 LB ALUMINUM PRODUCED PER DAY PER VSS CELL
PROCESS STREAMS - LB/HR
Materi al
Stream Number
1
2
3
4
5*
6*
HF
Na3AlF6
A1F3
2.5 (s)(B)
3.3 (s)(B)



1.54 (g)(C,D)
1.13 (s)
Jas F
.27 (g)
L13 (s)
| as F
Total Fluorides^'
Total as F^'
5.8
3.6



1.59
.39
Alumina
Anode Mat'1, as C
Alum1num
co2
h2o
c
Electrical Power
165
50
660. KWH
86
1 (s)
150 (g)
1 (g)
.2 (g)
.1 (s)
37 (g)
.1 (g)
.02 (g)
- Approx. Total
Stream
170
50
--
86
15o'A'
,, 	
37
•Gaseous effluent stream
(A)	Plus 30,000 SCFH of diluent air
(B)	Ref 4250
(C)	Ref 4254
(D)	Ref 4208
(E)	The balance not shown 1s fluoride absorption by the cathode
Soluble F Evolution Factor = 46 lb F/ton aluminum produced.
Figure 3-7". Electrolytic Aluminum Production - Vertical Stud Soderberg
Cell, Uncontrolled Process Model
3-35

-------
capture efficienty (actual experience, Reference 4208). The balance of the
fluoride not shown is absorbed 1n the cathode.
Fluoride Emission Control Techniques. Both solid and particulate
soluble fluoride are evolved from the VSS cell during operation. Abatement
techniques utilized for this process (which are identical to those used for
the other alumina reduction processes) are presented as controlled process
models in Figure 3-8. Typical usage is Process C.
The VSS, because of the absence of channels, allows the use of a
fume-collecting skirt around the base of the anode. The air volume required
for fume collection is significantly lower than for the HSS. The resultant
fume concentration is such that the CO and tar can be burned to reduce the
tar content of the exhaust gas, oxidize the fume tars and prevent them from
collecting in an fouling the ducting system. Maintenance of the skirts is a
problem (melting, deformation) as is maintenance of the alumina crust to form
an effective barrier to prevent evolution of fume from the pot into the rooms.
Fluoride Emissions. Soluble fluoride emission from the vertical
stud Soderberg was 2120 tons (as F) in 1970 (Reference 4208) or about
141 of the total from this industry. This corresponds to an average over-
all fluoride abatement efficiency of 791 (Reference 4208). The emis-
sion level will increase to 16,000 tons annually in the year 2000 if
production projections are correct and if current control level is
maintained. If 99% control is established, there will be 800 tons (as F)
of fluoride emitted from this process in the year 2000.
3.2.6 Economic Analysis
3.2.6.1 Basic Process
The estimated economics for the production of primary aluminum ingots
by the prebaked anode process, without the costs imposed by fluoride
emission control, are presented in Table 3-7. Process economics for vertical
stud, and horizontal stud Soderberg processes are similar, and the data of
3-36

-------
ELECTROSTATIC PRECIPITATOR
EFF. = 98% PARTICULATE
0% GASEOUS (ASSUME NO ADSORPTION)
FLOATING BED SCRUBBER
EFF, = 72% PARTICULATE
95% GASEOUS

212°F
Li
<2>
-TO STACK
WATER -
PROCESS A
TO DISPOSAL
OR RECYCLE
212°F {EST.}
TO STACK

PROCESS B
TO LIQUID - SOLID
SEPARATION UNIT
WITH SOLIDS DISPOSAL
AND LIQUID RECYCLE
SPRAY SCREEN
EFF. = 83% PARTICULATE
96% GASEOUS
DRY ALUMINA ADSORPTION

QUENCH
TOWER
— TO STACK
WATER.

TO STACK
PROCESS C
WATER
BAG
FILTERS
TO LIQUID - SOLID SEPARATION
UNIT WITH SOLIDS DISPOSAL
AND LIQUID RECYCLE
RECYCLE TO CELL FEED
FLUIDIZED
BED
BASIS - 2070 LB ALUMINUM PRODUCED PER DAY VERTICAL STUD SODERBERG CELL
PROCESS STREAMS LBS/HR
Material
Steam Number
8*
10*
11
12*
13
14
15*
HF
Na3A1F6
A1F,
1.54(g)
• i 3 (s)
as F
27(g)
13(s)
as F
. 127(s)
as F
1.54(g)
.003(s) ^
as F
¦09(s)
as F
,08(g)^
.04(s)'C^
as F
.11(s)
as F .
•06(g)(C)
.02(s)'C'
as F
1.58(s
as F
.(C)
•Q15tg)(C)
),001 (s)^
(as F
Total as F
1.59
.13
1.47
.09
.11
.11
.08
1.58
.015
Alumina
co2
H„0
1 (s)
150(g)
Kg)
¦ 2(g)
Ks)
37(g)
1(g)
02(g)
.98(s)
.196(s)
• 02(s)^C ^
150(g)
Kg)
•004(s)(C^
•72(s)
¦ 15 (s)
• 28(s
,(C)
150(g)
60(g)(Est)
. 05(s)^C ^
,83(s)
-1 ? (s)
.17(5)
150(g)
60(g)
,03(s
(C)
,(C)
20(s)(Est)
.198(s)
21(s)(Est)
.198(s)
.01(s)(C)
150(g)
Hi)
,0Q2(s)(^
Approx. Total
Stream
150
(A)
17
150
2104
,(B)
210
(A)
20
21
150
(A)
*Gaseous Effluent Stream
(A)	Plus 30,000 SCFH of diluent air.
(B)	Plus water and soluble fluorides.
(C)	Ref 4208

Calc.
Fluoride Emission
Factor -
lb F/ton A1
Source
Process
A Process B
Process
C
Process D
Treated Hood Exhaust
Emission
34.19
2.56
1.86

0.35
Untreated Hood Loss to
Atmosphere
9.07
9.07 ,
9.07

9.07
Total Emission to Atmosphere
from Cell
33.26
11.63
10.93

9.42
Reported Average Overall Vertical Stud Cell Soluble Fluoride Emission Factor = 9.66 lb F/ton Al^
(Reflects the utilization of series abatement systems)
Figure 3-6. Electrolytic Aluminum Pro-
duction -- Controlled Pro-
cess Model, Vertical Stud
Soderberg Cell
3-37

-------
Table 3-7. Estimated Economics of Aluminum Production
(Pollution Control Cost Excluded)


Plant Capacity

Total Capital Investment^'
60 M tons/yr
120 M tons/yr 1
250 M tons/yr
48 sm
90- $MM
175 $MM
Production Costs



Direct Costs



Aluminap' (1.9 tons AlgOj/Ton Al)
130.09 $/ton Al
130.09 $/ton Al
130.09 $/tOn Al
Electrode Materials (Approx. 0.6 tons C/Ton Al)
26.40
26.40
26.40
Cryolite (0.05 Tons/Ton Al)
12.60
12.60
12.60
Aluminum Fluoride (0.03 Tons AlFj/Ton Al),
10.05
10.06
10.05
Miscellaneous (Fluorspar, Soda Ash, etc.)
8.00
8.00
8.00
Electric Energy (16000 kwh/ton at 0.3S cents/kwh)
56.00
56.00
56.00
Labor (Operating, Maintenance, Supervision and Indirect)
60,00
52.60
45.00
TOTAL DIRECT COSTS
303.14
295.74 •
288.14
Indirect Cost



Depredation
80.00
75.00
70.00
Interest (at 71, 201 debt)
11.20
10.50
9.80
Taxes and Insurance
'20.00
18.50
17.00
' Plant and Labor Overhead
20.00
18.00
15.00
TOTAL INDIRECT
131.20
122.00
111.80
Manufacturing Cost (J/ton Al)
434.34
417.74
399.94
General and Sales Expenses ($/ton A))
8.69
8.35
8.00
f.o.b. Cost ($/ton Al)
443.03
426.09
407.94
Product Revenue ($/ton Al)
580.00
' 580.00
580.00
Profit After Taxes (at 501) fl/ton Al)
68.49 $/ton Al
76.96 {/ton Al
86.03 S/ton Al
Cash Flow (JMM/yr)
8.9 $NM/yr
18.2 $MM/yr
39.0 SW/yr
Return on Investment (ROI), I
10.7 t
12.8 *
17.0 t
Cost of Alumina (Including shipping)



¦'^Costs of Prebake Anode Unit: Manufacturing Costs for Soderberg
processes are similar.



Cost of alumina (Including shipping)

$/Ton A1203

^Bauxite (5.00 $/Ton FOB mine at 2.5 tons baux1te/ton A1.0, with an 18 $MM fixed
Investment)

12.50

Shipping (2.00 $/ton bauxite, Jamaica to Gulf Coast)

5.00

Bauxite to Alumina



Direct costs



Soda (0.25 tons/ton AljOj)

8.50

Coal (0.25 tons/ton Al^Oj)

2.00

Fuel 011 (0.125 tons/ton AljOj)

2.88

Line (0.0625 tons/ton AljOj)

1.14

Operating labor

2.85

Supervision and Fringe Benefits

2.85

Maintenance and Supplies

2.25

Indirect Costs



Depredation

14.50

Taxes and Insurance

0.50

Plant and Labor Overhead

1.00

Shipping ($12.SO/ton AljOj, Gulf Coast to Pacific Northwest by rail)

12.50

Manufacturing Cost of AljOj ($/ton)

$68.47/ton

3-39

-------
Table 3-7 is valid for these processes, prior to the addition of fluoride
emission control systems. The model plant was assumed to be located in the
Pacific Northwest to take advantage of the availability of inexpensive
*
power . In addition, the firm was assumed to own a bauxite mine in Jamaica
and an alumina facility in the Gulf Coast area. The general process
economic assumptions and bases for the economic analyses are those of
Section 3.1.1, Economic Analysis - Discussion.
The uncontrolled aluminum process model has a mean estimated ROI of
13.0%.
3.2.6.2. Impact of Controls
The addition of emission control equipment causes a sharp decrease in
ROI for all of the abatement processes except the dry adsorption process.
ROI's after emission control range from 7.9 to 15.1%. The aROI's go from 2%
to 26%. The AROI's versus plant capacity curves for each control process (as
applied to the individual production processes) are almost completely flat
(Figure 3-9 illustrates the aROI versus capacity curves for PBA control
processes). A major part of the variation cited above arises from differences
between the capital and operating costs of the different emission control
processes. The emission control processes are generally unitized to cover a
pot-line, so that plant capacity differences did not affect the per ton cost
of control. Since the aROI curve is rather flat, both large and small
primary aluminum producers should respond to similar added pollution control
requirements in the same way. Because most aluminum plants have some
existing pollution control facilities, the impact of added equipment to
achieve possible new standards of control would not necessarily be as great
as that shown in Figure 3-9 .
*Bonneville Power Administration states that the actual 1970 billing rate was
1.7 mills per kwh +0.3 mill per kwh for the aluminum facilities in the
Pacific Northwest. In the analysis, 3.5 mills per kwh was us«d since an
increase is expected in the next few years. In this case, 3.5 mills per kwh
is conservative.
3-40

-------
CO
I
45*
&
-10
CD
QC
<

£
Furnace - control process "b"
Furnace - contcol process "a"
CONTROL PROCESS "d"
CONTROL PROCESS V
CONTROL PROCESS "a"
CONTROL PROCESS "c"
-3D
0	100	200	300
PLm CAPACITY (1000 tons/year)
Figure 3-9. Effect of Pollution Control Cost on the Aluminum Industries Return on
Investors' Equity - Prebaked Anode Process

-------
Tables 3-8 through 3-21 present estimates of the economics of the
various pollution control processes, as applied to each of the three primary
production processes. The bases employed for the calculations, and the
assumptions on which the estimates rest, are those contained in Section 3.1.1
under the heading "Control Cost Economics."
For all the fluoride emission control systems other than the dry
alumina adsorption process, the vertical stud Soderberg enjoys an advantage
in lowered control costs due to its relatively low diluent air requirements
(about 15% of the diluent air stipulated for the PBA and HSS production
units). The dry alumina adsorption process (currently proprietary to the
Aluminum Company of America) has by far the lowest pollution control cost
for the two production systems on which it may be used - $6.44/ton for the
PBA process and $3.38/ton for the VSS process.
Table 3-21 summarizes control costs for the three production processes,
the various control processes, and a "typical" 120,000 ton per year plant.
3-42

-------
(4)
Table 3-8. Prebaked Anode Aluminum Production x ' - Estimated Economics of Control Process A
Basis - 8514 Tons A1 Per Year Per Precipitator (one 25 cell pot line)
Capital Cost Estimates ($1000)
Item
Number
Description
Floating Bed Scrubber, 2 at 12'
diameter by 12' - 6", 8 ft/sec
velocity, mass transfer coefficient
of 80 mols/(hr) (ft3) (atm), 2 in.
W.G. pressure drop, 109,000 cfm,
46 horsepower, neoprene lined steel.
Equipment
F.0B.
Cost	
117
(6)
Reference
Number
4391
4383
4392
4390
Installation
Factor
3.02
Equi pment
Installation
Cost	
353-
Capital Subtotal	353
Indirects (@15%) 53
Contingency (@ 20%) 71_
Total Capital (as of January 1971)	477
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11
derating Cost ($ /hr )
Item
Number
Power
Cost
0.24
Maintenance
Cost
0.25
Equipment
Operating
Cost
0.49
Subtotal
121)
Water
Disposal
(22)
0.49
330 gpm, 90% recycle ) 0.04
Total Operating Cost
0.53
Total Operating Cost ($/hr)	0.53
Taxes and Insurance (2%, 330 days)	1.20
Capital (7.1%, 330 working days)	4.28
Pollution Control Cost ($/hr)	5.01
Pollution Control Cost ($/ton) •	559

-------
Table 3-9,
Prebaked Anode Aluminum Production^ - Estimated Economics of Control Process B
Basis - 8514 Tons A1 Per Year Per 2 Scrubbers (one 25 cell pot line)
Capital Cost Estimates ($1000)
Description
Quench Tower, 10' diameter by
30' height, neoprene lined steel
Spray Screen, 64 ft2, 30 lb/hr
loading, 5000 gpm, 46 horsepower
energy requirement, fiberglass
reinforced polyester
Liquid Solid Separation, 25 ft
30 lb/hr loading, 5000 gpm,
50,000 gal capacity, neoprene lined
steel
tquipment
F.O.B.
Cost
42
30
18
Reference
Number
(a)
4383
4392
4398
4392
Installation
Factor
3.05
3.83
4.22
Equipment
Installation
Cost	
(a) Company private design information
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
128
T15
76
Capital Subtotal	319
Indirects (@ 15%)	48
Contingency (@ 20%)	64
Total Capital (as of January 1971)	431
Operating
Item
Number
Cost (s 1 hl" ]
Power
Cost
0.04
0.24
0.10
Maintenance
Cost
0.20
0.32
0.06
Subtotal
121)
Equipment
Operating
Cost
0.24
0.56
0.16
Water
Disposal
(22)
0.96
(5700 gpm, 90% recycle) 0.64
Total Operating Cost
1.60
Total Operating Cost ($/hr)	1.60
Taxes and Insurance (2%, 330 days)	1.09
Capital (7.1%, 330 working days)	3.86
Pollution Control Cost ($/hr)	6.55
Pollution Control Cost ($/ton)	15J9

-------
Table 3-10. Prebaked Anode Aluminum Production^ - Estimated Economics of Control Process C
Basis - 3406 Tons A1 Per Year Per Process (one 10 cell pot line)
Item
Number
Description
Capital Cost Estimates_[$10002
Equi pment"
1. Electrostatic Precipitator,
109,000 cfm, 0.9 in W.G
pressure drop-
F.O.B.
Cost
240
Reference
Number
(a)
Installation
Factor
1.69
Equipment
Installation
Cost 	
406
Capital Subtotal	406
Indirects (0 15%)	61
Contingency (@ 20%)	81
Total Capital (as of January 1971)	548
(a) SEE FOOTNOTE 20.
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
Jperatinq Cost ($ !br )
Item
Number
Power
Cost
0.46
Maintenance
Cost
0.38
Equipment
Operating
Cost
Subtotal
T21).
Water
Disposal
(22)
Total Operating Cost
0.84
0.84
0.84
Total Operating Cost ($/hr)	0.84
Taxes and Insurance (2%, 330 days)	i*38
Capital (7.1%, 330 working days)	491
Pollution Control Cost ($/hr)	77T3
Pollution.Control Cost ($/ton)	6*53

-------
Table 3-11. Prebaked Anode Aluminum Production^ ^ - Estimated Economics of Control Process D
Basis - 50,000 Tons A1 Per Year Per System
Capital Cost Estimates ($1000)
Item
Number
Description
Dry alumina adsorption
system.
Equipment
F.U.I
Cost
Reference
Number
4254
Installation
Factor
Equipment
Installation
Cost	
2,000.
Capital Subtotal	2,000.
Indirects (f 155!)	(a)
Contingency (@ 20%)	400.
Total Capital (as of January 1971)	2,400.
(a)	Included in Installed Cost, Reference 4254
(b)	See Reference 4254
All control economics footnotes are located in Section 3.1.1, pages 3-10„and 3-11,
Operating Cost ($ /h
Item
Number
Power
Cost
(a)
Maintenance
Cost
(a)
Subtotal
121)
Water
Disposal
(22)
Total Operating Cost
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 330 days)
Capital (7,11, 330 working days)
Pollution Control Cost ($/hr)
Pollution Control Cost ($/ton J
Equipment
Operating
Cost	
13.07
13.07
13.07
13.07
6.06
21.52
40.65
6.44

-------
Table 3-12. Prebaked Anode Baking Furnace^1^ - Estimated Economics of Control Process A
Basis - 50,000 Ton A1 Per Year Plant
Capital Cost Estimates ($1000)
I tem
Number
Description
Electrostatic Precipitator,
147,000 cfm 0.9 in W.G. oressure
drop, power requirement
120 horsepower, monel clad.
tquipment
F.O.B
Cost
324
Reference
Number
(a)
Installation
Factor
1.69
Equipment
Instal1ation
Cost
548
Capital Subtotal	!>48
Indirects (@ 15%)	82
Contingency (@ 20%)	110
Total Capital (as of January 1971)
Operating Cost ($ / hr
Item
Number
Power
Cost
0.63
Subtotal
T"2i;
Water
Disposal
(22)
Mai ntenance
Cost
0.38
Total Operating Cost
tqui pment
Operating
Cost
1.01
1.01
1.01
(a) See Footnote (20)
All control economics footnotes are located
in Section 3.1.1, pages 3-10 and 3-11.
Total Operating Cost ($/hr)	1.01
Taxes and Insurance (2%, 330 days)	1-87
Capital (7.1%, 330 working days)	6.63
Pollution Control Cost ($/hr).	9.51
Pollution-Control Cost ($/ton)	1.51

-------
Table 3-13. Prebaked Anode Baking Furnace^^ -- Estimated Economics of Control Process B
Basis - 50,000 Ton A1 Per Year Plant
Item
Number
Description
Ca£i_ta_l_Cost_Estima_tes_Ji^1000]i
"tquipment
Spray Tower, 2 at 10' diameter
by 30', 8 ft/sec velocity,
76,000 cfm, 5 grains/sec loading,
2 in W.G. pressure drop, required
power 73 horsepower, neoprene lined
steel.
F.U.B.
Cost
91
Reference
Number
4383
4391
4392
Installation
Factor
2.44
Equipment
Installation
Cost	
222
Capital Subtotal	222
Indirects (@ 15%)	33
Contingency (@ 20%)	44_
Total Capital (as of January 1971) 299
Operating
Item
Number

Power
Cost
0.38
Maintenance
Cost
0.25
tqui pment
Operating
Cost	
0.63
Subtotal
T21]
Water
Disposal
(22)
0.63
(3340 gpm, 90% recycle) 0.40
Total Operating Cost
1.03
All control economics footnotes are located
in Section 3.1.1, pages 3-10 and 3-11.
Total Operating Cost ($/hr)	'•
Taxes and Insurance (2%, 330 days)
Caoital (7.1%, 330 working days)
Pollution Control Cost ($/hr)	4.46
Pollution Control Cost ($/ton)	0.69

-------
Table 3-14.
(4)
Horizontal Stud Soderberg Aluminum Productionv ' - Estimated Economics of Control Process A
Basis - 8514 Tons A1 Per Year Per Precipitator (one 25 cell pot line)
Gl>
I
-p»
UD
Item
Number
_£a£rtal_Cost_EsjMmaites_Ji$]0002
Descri ption
Electrostatic Precipitator, 146,000
cfm, 0.9 in W.G. pressure drop.
Equipment
F.U.B.
Cost
324
Reference
Number
Installation
Factor
1.69
fcqulpment
Installation
Cost	
Capital Subtotal
Indirects	)
Contingency {@20% )
Total Capital (as of January 1971)
(a) See Footnote 20.
All control economics footnotes are located in Section 3.1.lpages 3-10 and 3-11.
548
548
82
110
740
Operating Cost f$ /hr. )
Item
Number
Power
Cost
0.62
Subtotal
(211
Water
Disposal
(22)
Maintenance
Cost
0.51
Total Operating Cost
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)
Pollution Control Cost ($/hr)
Pollution Control Cost ($/ton)
tqui pment
Operating
Cost
1.13
1.13
1.13
1.13
1 .87
6.63
9.63
8.96

-------
Table 3-15.
(5}
Horizontal Stud Soderberg Aluminum Productionv ' - Estimated Economics of Control Process B
Basis - 8514 Tons A1 Per Year Per 2 Scrubbers (one 25 cell pot line)
CO
I
in
o
Item
Number
_Ca£HaJ_Cost_Estima_tes_£$10002
Descri ption
Floating Bed Scrubber, 2 at 8'
diameter by 16' - 8", 8 ft/sec
velocity, mass transfer coefficient
of 80 mols/(hr)(ft3)(atm), 2 in.
W.G. pressure drop.
Equipment
F.O.B.
Cost	
139
Reference
Number
4391
4383
4392
4390
Installation
Factor
3.02
tquipment
Instal lation
Cost
420
Capital Subtotal	420
Indirects (015%)	63
Contingency (@20%)	84
Total Capital (as of January 1971)	567
All control economics footnotes are located in Section 3.1.1pages 3-10 and 3-11.
Operating
Item
Number
Cost (S /hr- )
Power
Cost
0.29
Maintenance
Cost
0.30
Equipment
Operating
Cost
Subtotal
Water
(21)
Disposal
(440 gpm, 90% recycle)
(22)
Total Operating Cost
0.59
0.59
0.05
0.64
Total Operating Cost ($/hr)	0.64
Taxes and Insurance (2%, 330 days)	1.43
Capital (7.1%, 330 working days)	5,08
Pollution Control Cost ($/hr)	77TT
Pollution Control Cost ($/ton)	g"65

-------
(g)
Table 3-16. Horizontal Stud Soderberg Aluminum Productionv ' — Estimated Economics of Control Process
Basis - 5110 Tons A1 Per Year Per Tower (one 15 cell pot line)
Item
Number
Description
Ca£UaJ_Cost_Estiimates_£$10002
"tqui pment"
Spray Tower, 12' diameter by 28'
4 in W.G. pressure drop, fan at 75%
efficiency, 87,500 cfm,
74 horsepower power requirement,
neoprene linetf steel.
F.U.B.
Cost
8o'10)
Reference
Number
4383
4391
4392
4387
Installation
Factor
3.44
Equipment
Installation
Cost	
275
Capital Subtotal	275
Indirects (@ 15%)	41
Contingency (@ 202!)	55
Total Capital (as of January 1971)	371
3£eratijTa_Cost>J$^h£__Ji
Item
Number
Power
Cost
0.39
Maintenance
Cost
0.25
Equipment
Operating
Cost
0.64
Subtotal
Water^21'
Disposal
0.64
(800 gpm, 902! recycle ) 0.10
Total Operating Cost
0.74
All control economics footnotes are located in Section 3.1.1, Dages 3-10 and 3-11.
Total Operating Cost ($/hr)
Taxes and Insurance [2%, 330 davs)
Caoital (7.1%, 330 working days)
Pollution.Control Cost ($/hr)
Pollution Control Cost ($/ton)
0.74
0.94
3.33
5.01 .
7.73

-------
Table 3-17. Vertical Stud Soderberg Aluminum Production - Estimated.Economics of Control Process A
Basis - 60,000 Tons A1 Per Year Per Precipitator (one 176 cell pot line)
Item
Number
_Ca£vtal_Co£t_Es^ima_tes_£$10002i
Description
Electrostatic Precipitator, 108,000
cfm, 0.9 in W.G. pressure drop.
fcquipment
F.U.B.
Cost
230
Reference
Number
(a)
Installation
Factor
1.69
tquipment
Installati on
Cost	
389
Capital Subtotal
Indirects (* 15?J
Contingency 20% )
Total Capital (as of January 1971)
389
58
78
525
(a) See Footnote 20.
All control economics footnotes are located in Section3.1.1 pages 3-1Oand3-11.
OgeMtina_Cost_J$^hri;_J_
Item
Number
Power
Cost
0.46
Maintenance
Cost
0.38
bquipment
Operating
Cost
0.84
Subtotal
0.84
Waterv
Disposal
(22)
Total Operating Cost
0.84
Total Operating Cost ($/hr)	0.84
Taxes and Insurance (2%, 330 days)	1.33
Capital (7.1%, 330 working days)	4.71
Pollution Control Cost ($/hr)	6.88
Pollution Control Cost ($/ton)	o.91

-------
Table 3-18.
Vertical Stud Soderberg Aluminum Production - Estimated Economics of Control Process B
Basis - 60,000 Tons A1 Per Year Per Scrubber (one 176 cell pot line)
GO
tn
OJ
Item
Number
_£a£VtaJ_Cost_Estima_tes_^10002
Descri ption
Floating Bed Scrubber, 2 at 12'
diameter by 12' - 6", 8 ft/sec
velocity, mass transfer coefficient
of 80 mols/(hr)(ft3)(atm), 2 in.
W.G. pressure drop, 108,000 cfm,
46 horsepower, neoprene lined
steel
Equipment
F.U.B.
Cost
116
Reference
Number
4391
4383
4392
4390
Installation
Factor
3.02
Equipment
Installation
Cost	
350
Capital Subtotal
Indirects (0 15#
Contingency {zq% )
Total Capital (as of January 1971)
350
53
70
473
All control economics footnotes are located in Section 3.1.1pages 3-10 and 3-11.
Operating
Item
Number
Cost ($ /hr.
Power
Cost
0.24
Maintenance
Cost
0.25
tqui pment
Operating
Cost
Subtotal
T21)
Water
Di sposal
(22)
Total Operating Cost
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)
Pollution Control Cost ($/hr)
Pollution Control Cost ($/ton)
0.49
(330 gpm,90% recycle)
0.49
0.04
0.53
0.53
1.19
4.24
5.95
0.79

-------
Table 3-19. Vertical Stud Soderberg Aluminum Production - Estimated Economics of Control Process C
Basis - 23,800 Tons A1 Per Year Per Process (one 70 cell pot line)
Item
Number
Description
tqulpment
F.O.B.
Cost
Reference
Number
Installation
Factor

tquipment
Installation
Cost
1
Quench Tower, 10' diameter by 30'
height, neoprene lined steel
42
(a)
3.05

128
2
Spray Screen, 64 ft^, 30 lb/hr
loading - 5000 gpm, 46 horsepower,
fiberglass reinforced polyester
30
4383
4392
3.83

115
3
p
Liquid solid separation, 25 ft ,
30 lb/hr loading, 5000 gpm,
50,000 gal capacity, neoprene lined
steel
18
4398
4392
4.22

76




Capital Subtotal
319




Indirects (0 15%)
48




Contingency (@20X ) 64


Total Capital (as of January 1971) 431
(a) Company private design information.
All control economics footnotes are located in Section3.1.1 pages 3-10 and 3-n
iperatinq Cost ($ /hr. )
Item
Number
Power
Maintenance
Cost
Cost
0.04
0.20
0.24
0.32
0.10
0.06
Subtotal
bquipment
Operating
Cost
0.24
0.56
0.16
0.96
Water^^(5700 gpm, 90% recycle) 0.64
Disposal
(22)
Total Operating Cost
1.60
Total Operating Cost ($/hr)	1.60
Taxes and Insurance (2%, 330 days)	1.09
Capital (7.1%, 3.30 working days)	3.86
Pollution Control Cost ($/hr)	5.55
Pollution Control Cost ($/ton)	2!18

-------
GO
en
ui
Table 3-20. Vertical Stud Soderberg Aluminum Production
Basis - 50,000 Ton A1 Per Year Per System
___^^^^^^___Ca£itail_Cost_Estimates_^$1000^__
Estimated Economics of Control Process D
Item
Number
Description
Dry alumina adsorption system
tqui pment
F.U.B.
Cost
Reference
Number
4254
Instal lation
Factor
Equipment
Installation
Cost	
600
Capital Subtotal	600
Indirects (0 15%)	(a)
Contingency (@20% )	120
Total Capital (as of January 1971)	720
(a)	Included in Installed Costs, Reference 4254
(b)	See Reference 4254
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11
Operating
Item
Number
Cost (j / hr )
Power
Cost
(b)
Maintenance
Cost
(b)
Subtotal
Water
(21)
Disposal
(22)
Total Operating Cost
Equipment
Operating
Cost
13.07
13.07
13.07
Total Operating Cost ($/hr)	13.07
Taxes and Insurance (2%, 330 days)	1-82
Capital (7.1%, 330 working days)	6.45
Pollution Control Cost ($/hr)	21.34
Pollution Control Cost ($/ton)	3.38

-------
Table 3-21. Cost of Emission Control - Typical 120,000 ton/year Plant
_Control Process*
Production Process""""**
A
B
C
D
$/Ton J mi %
$/Ton
AROI %
$/Ton J aROI %
$/Ton
&R0I %
PBA
1
6.63 ; 13
j
5.59
12
15.19
!
25
i
6.44
11
PBA Furnace
1.51 ; 3
0.75
2

; -
-

HSS
8.96 ; 18
6.65
14
7.73
15
-
-
VSS
0.91 j 2
0.79
2
2.18
4
3.38
4
(*As identified on the Controlled Process Models for each production process -
Figures 3-3 , 3-4 , 3-6 , and 3-8 )

-------
3.3 IRON AND STEEL INDUSTRY

-------
3.3 IRON AND STEEL INDUSTRY
3.3.1 General
The iron and steel industry incorporates a wide variety of fluoride
emitting processes producing products which range from pelletized iron ore
to cast, forged and welded high alloy steel shapes. This report covers
only those primary high temperature metal production processes which are
now, or are projected to be, major sources of fluoride evolution and
emission. The many plants of the industry contain almost every possible
combination of these process elements, making generalization on an industry
wide basis extremely difficult.
The fluorides evolved and emitted by the industry are gaseous
hydrogen fluoride and solid particulate calcium fluoride.
3.3.2 Industry Description
The primary high temperature iron and steel process elements which
are current major sources of fluoride evolution and emission are:
(1)	Iron ore pelletizing
(2)	Iron ore sintering
(3)	Blast furnace operation
(4)	Open hearth furnace operation
(5)	Electric arc furnace operation
(6)	Basic oxygen furnace operation
Iron ore pelletizing operations occur almost exclusively at the mine
sites; iron ore sintering is almost always performed at a plant site in
conjunction with blast furnace operations. The remaining process elements,
listed by location in Appendix 7-2, occur in every possible combination,
singly and in multiple, with and without blast furnace operations. A
specimen integrated iron and steel plant sinters iron ore to aggregate the
3-57

-------
fines; charges the sintered iron ore, coke, and limestone into a blast
furnace to reduce the iron ore to pig iron; and then converts the pig iron
to steel, using a protective molten flux (limestone and fluorspar) cover,
steel scrap and alloy additives in one or more of the three steel producing
processes (the open hearth, electric arc, and basic oxygen furnaces).
In each case, wherever the high temperature process involves a
source of hydrogen (water or fuel), gaseous hydrogen fluoride is evolved
and emitted, accompanied by solid calcium fluoride particulate dispersed
in the exit,gas stream. Where no source of hydrogen is available, calcium
fluoride is evolved and emitted as dispersoid particulate. On the basis of
thermochemical equilibria calculations, other fluoride species are not
evolved or emitted to any measurable extent.
Because of the variety of processes and diversity of process element
combinations at the various plant locations, detailed descriptions are
provided under the individual process description sections which follow;
the reader is referred to Schueneman, High and Bye^4055^ and Varqa and
(A?OZ\
Lownie	for a more exhaustive treatment.
3.3.3 Production Trends
During the past decade the U.S. steel industry has been undergoing
a multifaceted transition as a result of technological change. Evolutionary
improvements of both practice and equipment have been made to gain more
efficient production.
Beneficiation of iron ore leading to the development of pelletized
ore was a major development of the American steel industry in the period
following World War II. The result of this effort was that in 1968, of some
130 million tons of ore consumed in iron making, more than 100 million tons
was agglomerated, and consumption of pellets had increased to about 50
(4286 4287^
million tons.1 '	This development has contributed to a reduction in
coke consumption and to an increase in the output of pig iron for existing
blast furnace facilities. The evolution of cryogenic technology, permitting
3-58

-------
the low-cost production of oxygen in quantity, led to the development of
oxygen-supplemented steelmaking practice and culminated in the acceptance of
the basic oxygen process (BOF).
The BOF produced 48 million tons in 1968 at 90% of rated installed
capacity; this level is expected to almost double by 1975, to over 80 million
tons, requiring investment of about $800 million. Production of steel from
electric furnaces will grow at a similar rate in this period from 16 million
to about 28 million tons. The balance of the steel requirements will be met
by open hearth production, which will decline from 60 million tons in 1968
to about 30 million tons in 1975.^^85,4287) metai requirements are
expected to rise only gradually; the decline of open hearth production
coupled with the increased yield from continuous casting and the dramatic
increase in BOF production will satisfy the projected hot metal needs. Thus,
no increase of blast furnace capacity beyond the current estimated 100
million tons will be required before 1975, although older and smaller units
may be modernized or even replaced with more efficient plants.^4286'4287^
Existing coke oven capacity will also be adequate, although it too is
being modernized. A summary of the current and estimated production through
the year 2000 is presented in Table 3-22.
The integrated steel plant will continue to be the predominant type
of manufacturing facility. Current nominal production levels for integrated
steel plants are more than 1 million tons/year while plants currently being
designed or installed have planned production levels on the order of 2 to 3
million tons per year. This size of steel production facility will be used
for new plant design at least through the next decade.(4286,4287)
3.3.4 Fluoride Control and Emissions Summary
The only reported attempt to control soluble fluoride emissions
in the iron and steel industry is from a program pursued at the U.S. Steel
works in Geneva, Utah.^79^ High fluoride iron ore (about 3000 ppm) was
used in the process, and the local fluoride levels in the area were such
that damage suits were brought against the company.
3-59

-------
Table 3-22. Steel Production
Process
1968
Million Tons
1975
Million Tons
Estimated
Growth Rate
Extrapolated^ ^
to Year 2000
Mill ion Tons
Blast Furnace
(Ore Tonnage)
130
120
0%
120
Sintering
(Ore Tonnage)
50
35
-4.7%
To a constant,
value of 13('>
Pelletizing
(Ore Tonnage)
50
65
-4.7%^
To a constant >
level of 107"'
Open Hearth
66
30
-12%
To zero level
Basic Oxygen
48
80
7.6%
To a constant
value of ~135
Electric Arc
16
28
8.3%
To a constant
value of ~35
Total Steel
Production
130
138
0.85%
170
^ No reference data - TRW estimate.
The abatement procedure adopted by U. S. Steel involved inclusion
of limestone in the feed material to the sintering operation. This
reportedly reduced fluoride emissions by 50%. Further treatment of the
gas stream with finely ground calcium hydroxide gave further reductions
in fluoride content. Particulate matter was collected in electrostatic
precipitators.
The off-gases from the open hearth furnaces were also treated with
calcium hydroxide for fluoride adsorption. Particulates were removed using
cyclone separators and electrostatic precipitators in series. This approach
3-60

-------
is of particular interest since the conversion of gaseous to particulate
fluoride material will result in fluoride control at the same time particu-
lates are controlled. The iron and steel industry has done much more work
on control of particulates than on control of fluorides.
It cannot be considered that the optimal solution to a fluoride
abatement problem in Utah necessarily constitutes a general solution. In
particular, the shortage of water in that area might preclude potential
scrubber approaches. Other collection and transfer equipment currently
employed in the majority of the industry is aimed specifically at collecting
and transporting the economically recoverable dusts -- the ore, iron and
other metal oxides, and metallic iron — and other airborne particulate
material that forms visible effluent plumes.
The individual pollution abatement systems for the process elements
are discussed under the individual process element headings.
Unfortunately, almost no definitive information has been published
on fluoride evolution and emission from the iron and steel industry in the
U.S., with the exception of that small part of the industry using high
fluoride ore. An account of a fluoride emission reduction program at a
facility in Utah has been reported as cited above• and several
publications have been presented for European sources. None of the
published data has distinguished between soluble fluorides and CaF^, emission.
Fluorides enter the iron and steel making process as a minor com-
ponent in iron ore (concentrations range as high as 3000 ppm) and as fluor-
spar for use in fluxing. The fluoride in the iron ore is partially
volatilized at the temperatures involved where sintering or nodulizing of
the blast furnace feed material is practiced.(4179,4055) otherwise, the
volatilization occurs in the blast furnace. In the latter case, the
evolved fluoride tends to be captured (adsorbed) by the limestone that
constitutes a portion of the blast furnace charge and is collected or emitted
as a particulate rather than a gas. The usage of fluorspar in the iron and
steel industry and steel production for various furnace types, as reported
3-61

-------
by the Bureau of Mines, can be combined to yield average fluoride input
values of 1.74, 6.14, and 3.63 pound fluoride/ton steel for open hearth, BOF,
and electric furnaces, respectively.
Estimates of the fluoride emissions from the iron and steel industry
vary widely as shown by comparing the following items of information:
(1)	Singmaster and Breyer'^ report an emission total for
steelmaking furnaces-(basic oxygen, open hearth, and electric) of 1716
tons of fluorides (soluble and insoluble) per year for 1969.
(2)	Sinter plant emissions have been reported'^ as varying
from 0.3 to 5.7 parts per million at two eastern plants, with 150 parts
per million reported from western plants (438^) prjor t0 the institution
of lime abatement processes.
(3)	The total fluoride emission factors attributed to the AISI in
the report by Environmental Engineering Inc.-Herrick Associates'*^ are
as follows:
Reported Total Fluoride Emission, lb/NT
Furnace	Wet Collection	Dry Collection
Open Hearth	0.002	0.030
Basic Oxygen	0.002	0.030
Electric Arc	0.004	0.009
(4) TRW-RRI experience indicates an emission factor typical of
an Eastern open hearth plant of 0.185 pound/NT steel in terms of soluble
fluoride.
(5) According to the Annual Report of the British Alkali and Works
nspectors	.
hearth plants is evolved.
Chief Inspectorsapproximately 50% of the input fluoride to open
3-62

-------
In view of the wide variations in the few reported values, the
failure to distinguish between soluble and insoluble fluorides, and the
lack of correlation with ore fluoride analysis, TRW performed thermochemical
analyses to determine the equilibria fluoride species and concentration
values for the high temperature reactions occurring in each of the
integrated iron and steel plant primary processes. For the varying process
conditions, the thermochemical equilibrium analyses indicated from 12 to
99.3% evolution of feed fluorides as gaseous HF if a hydrogen source was
present in the feed charge (including fuel and combustion air) and
evolution of CaF2 as the only measurable species if no hydrogen source was
available in the high temperature process. The identification of the
theoretical proportions of HF and CaF^ in the effluent streams on the basis
of hydrogen availability and reaction zone temperatures is very significant.
CaF^ has low environmental impact, and can be removed as a particulate by
the control devices normally employed by the industry. Detailed results of
these equilibria analyses are given in the following sections under the
appropriate process headings.
The equilibria analyses together with the following assumptions were
used to define the species emitted and the soluble fluoride emission factors:
(1)	Based on Reference (4276), about 50% of the input fluorides exit
in the slag. Therefore, gaseous fluorides emitted from uncontrolled iron
and steel processes are estimated at 50% of the theoretical equilibrium
analyses values computed for the high temperature zones. The mechanisms
involved in this reduction are probably adsorption and chemisorption of
gaseous fluorides on solid particles, particularly limestone.
(2)	The weighted average fluoride content of iron ore was estimated
assuming (on the basis of the geographical distribution of industry capacity)
95% usage of normal fluoride content ores at 650 parts per million F
concentration*, and 5% usage of high fluoride content ores at 3000 parts per
million F concentration.
*Iron ore was.assumed to contain the same average fluoride content as the
earth's crust(4242).
3-63

-------
(3) The distribution of wet and dry control systems cited from
Reference (4225) in each of the control process model diagrams was used in
conjunction with the control system efficiencies presented in the diagrams
for the calculations.
Table 3-23 presents the resulting evolution and emission factors and
tonnages for the processes and the industry. Mass balances and schematics
for each of the processes are presented under the individual process
headings. It should be noted that the estimated open hearth emission factor
(4385)
is much higher than the reference ' values. It was assumed that there
are no process hydrogen sources in the BOF and electric arc steel making
systems; therefore, soluble fluoride evolution from these processes was
estimated as zero.
The estimated soluble fluoride evolution for the iron and steel
industry was 69,100 tons for 1968, with an emission of 64,600 tons of
soluble fluorides during that year. In 2000, soluble fluoride evolution
for the industry is projected to be 49,800 tons; soluble fluorides emitted
if current practices continue would be 46,400 tons. If control processes
capable of 99% abatement efficiency are adapted, soluble fluoride emission
by the industry would drop to 500 tons in the year 2000.
3.3.5 Process Description and Economics
3.3.5.1 Iron Ore Sintering and Pelletizing
With very few exceptions, modern blast furnaces use as charge iron
ore which has been agglomerated by sintering or pelleti zing, or sized by
screening. As noted earlier, pelletizing plants are generally located at
the mine site. Sintering plants are normally a portion of an integrated
iron and steel operation. The trend towards a straight pellet or pellet-
sinter practice will continue in essentially 100% of industry practice.
Since both sintering and pelletizing involve high temperature agglomeration
of fluoride-containing material, the processes are quite similar. Therefore,
3-64

-------
Table 3-23. Soluble Fluoride Evolution and Emission From the Iron and Steel Industry

Iron Ore
Sintering
Iron Ore
Pelletizing
Blast
Furnace
Basic Open
Hearth Furnace
Basic Oxygen
Furnace
Electric Arc
Furnace
Total Iron and
Steel Industry
1968 Production
(10® tons product/year)
S0<"1
g
130(B)
87
66
48
16
130^
2000 Production
(10® tons product/year)
13(A)
]4(B)
107(b!
113(B)
i2o!b!
80
0
135
35
170^
Soluble fluoride evolution factor'0^
(lb F/ton product)
.73
,73(C'
.088
.81
o(F)
cP
1.06(E)
Soluble fluoride emission factor'^
(lb F/ton product)
.69
.69(C)
.065
.77
0
0
0.99(e)
1968 soluble fluoride evolution
(10^ tons F/year)
19.3
19.3
3.8
26.7


69.1
2000 soluble fluoride evolution
(103 tons F/year)
5.1
41.2
3.5
0
0
. 0
49.8
1968 soluble fluoride emission
(103 tons F/year)
18.2
18.2
2.8
25.4
0
0
64.6
2000 soluble fluoride emission
with current technology
(10-3 tons F/year)
4.8
39.0
2.6
0
0
0
46.4
2000 soluble fluoride emission
with 99% control technology
(10-3 tons F/year)
.05
.4
.04
0
0
0
.5
Notes: (A) Expressed as ore tonnage in process feed.
(B)	Expressed as product tonnage.
(C)	TRW estimation.
(D)	Reflects estimated 5% usage of high fluoride content (3000 ppm) ore, 95% usage of
average fluoride content (650 ppm) ore and application of wet control processes as
noted on the control mass balances.
(E)	Expressed as per ton of steel production.
(F)	Based on assumotion that there are no process hydrogen sources in the BOF and electric arc furnaces.

-------
only sintering will be considered in detail and the assumption is made that
sintering and pelletizing are essentially interchangeable from a fluoride
evolution standpoint.
Process Description. Figures 3-10 and 3-11 present flow diagrams
and mass balances for normal and high fluoride content iron ore sintering
operations. Modern sinter plants range from 2000 to about 10,000 tons
per day.
Present and Future Production Levels. The program for the develop-
ment of beneficiation and agglomeration techniques to enhance the quality
of ore charged to blast furnaces has culminated in the treatment of ores
previously considered of too low a grade to be used in the furnaces. The
first pelletizing plant was installed in 1949. By the end of 1968, U. S.
pelletizing capacity totaled some 50 million tons (more than 35% of the
130 million tons charged to blast furnaces) and several million tons of
additional capacity was being installed.^4286>4287) jhe growth rate of the
pelletizing process is expected to be about 4J% annually to a constant
level of about 107 million tons of ore pelletized per year through 2000.
The increase in pelletizing will be accompanied by a decrease in
the ore sintering process from a level of 50 million tons in 1968 to a
constant level of approximately 13 million tons annually in the year 2000.
Future trends in ore treatment will include eventual use of pre-
reduced agglomerates, pellets, or briquettes which will permit the furnace
operator to vary capacity to satisfy the varying hot metal demands without
blowing in additional furnaces (prereduced pellets could increase capacity
by at least 50%). In addition, research can also be expected on the
possibility of an integrated charge, i.e., agglomerates containing all
ingredients - carbon, iron units, and flux (probably precalcined). Finally,
attention will be focused on the use of computers to calculate burdens
charged to the blast furnace/4286' 4287)
3-66

-------
BASIS - 2000 TONS/DAY OF SINTER PRODUCED
PROCESS STREAMS - TONS/DAY
CO
CT>
-~l
FEED
MATERIALS
0

CRUSHERS


~
350° F
[ MIXING \
I DRUM J
COMBUSTION PRODUCTS
AND DUST TO
ATMOSPHERE OR
POLLUTION CONTROL
DEVICE
SINTERING
MACHINE
0
<5
COMBUSTION
AIR
IGNITION
FUEL AND
AIR
HOT
SINTER
HEATED AIR AND DUST
TO ATMOSPHERE OR
POLLUTION CONTROL
DEVICE
<5
COOLING AIR-

SINTER
COOLER
SINTER FINES
RECYCLE
200°F(est.)
SCREENS

SIZED SINTER
TO BLAST
FURNACE
Material
HF
CaF2 equiv.
Total Fluorides
Total as F
Iron Ore?*
Limes tone^'®^
Coke
Water
Natural Gas
Air (50% R.H.)
N,
°2
CO
C02
Sintered Ore
Appro*. Total
Stream
Stream Nunfcer
3.05
3.05
1.48
1900
110
500
40
2500
16
400
400
16000
16000
4*
0.77(g)
0.02(s)'A,B,D'
0.79
0.74
19 (s)
250(g)
12500(g)
2000(g)
1000(g)
900(g)
16700
2000
2000
0.34(s)
(A.B)
0.34
0.17
21 (s)
2000(g)
2000
Gaseous Effluent Stream
(A)	Reference 4179
(B)	Reference 4053
(C)	Reference 4055
(D)	Reference 4249
(E)	Does not include CaF^ added from control device recycle streams.
Soluble Fluoride Evolution Factor = 0.73 lbF/ton Sinter
NOTE: Fluoride content of the iron ore reflects 95% usage of 650 ppm fluoride
. content ore and 5% usage of 3000 ppm fluoride content ore.
1.16(s)(£|B'C.'
1.16
0.56
-2000
2000
Figure 3-10. Iron Ore Sintering - Uncontrolled Process Model

-------
BASIS - 2000 TONS/DAY OF SINTER PRODUCED
PROCESS STREAMS - TONS/DAY
FEED
MATERIALS
0

CRUSHERS


35Q°F
MIXING
DRUM
}
COMBUSTION PRODUCTS
AND DUST TO
ATMOSPHERE OR
POLLUTION CONTROL
DEVICE
SINTERING
MACHINE
<3>
<£>
COMBUSTION
AIR
IGNITION
FUEL AND
AIR
COOLING AIR-

HOT
SINTER
HEATED AIR AND DUST
TO ATMOSPHERE OR
POLLUTION CONTROL
DEVICE
<5>
SINTER
COOLER
SINTER FINES
RECYCLE
200°F(est.
SCREENS
0
SIZED SINTER
TO BLAST
FURNACE
Material
Stream Number
1
2
3
4*
5
6*
7
HF



3.01(g)



CaF2 equiv.
11.9


0.09(s)(A,B,D)

1.35(s)(A,B)
4.55(s)(A'B'C'£>
Total Fluorides
11.9


3.10

1 .35
4.55
Total as F
5.8


2.90

0.68
2.22
Iron Ore**
1900


19(s)

21 (s)

Limestone^'6'
110






Coke
500






Water
40


250(g).



Natural Gas

16





Air (50% R.H.)

400
16000

2000
2000(g)

N2



12500(g)



°2



2000(g)



CO



1000(g)



C02



900(g)



Sintered Ore






2000
Approx. Total
Stream
2500
400
16000
16700
2000
2000
2000
Figure 3-11
Gaseous Effluent Stream
**
Fluorine Content of Ore = 0.3 wt.%.
(A)	Reference 4179
(8)	Reference 4053
(C)	Reference 4055
(D)	Reference 4249
(E)	Does not include CaF^ added from control device recycle streams
Soluble fluoride evolution factor of facilities utilizing high
fluoride content (3000 ppm) iron ore = 2.86 lbF/ton of sinter produced.
High Fluoride Content Iron Ore Sintering -
Uncontrolled Process Model

-------
Fluoride Emission Control Techniques. The emission control techniques
employed currently in conjunction with normal and high fluoride content iron
ore are presented in Figures 3-12 and 3-13. As noted in Figure 3-12,
Processes A and B are used in over 90% of the sinter plants.
Ore sinter plants are hooded fop collection and transport of dust at
the majority of particulate discharge points in the system. The ore sintering
machine is hooded to vent, under induced draft, the dust-laden waste
combustion gas to the dust collectors shown.
Fluoride Emissions. Estimated soluble fluoride emissions from
sintering and pelletizing of iron ore were each 18,200 tons in 1968. By
2000, production will have decreased for iron sintering to where soluble
fluoride emissions, using currently employed control techniques, are
estimated at 4800 tons. For pelletizing, production increases could cause
the soluble fluorides emitted to rise to 39,000 tons by 2000, if currently
employed control techniques were used. If control techniques capable
of 99% efficiency are employed, soluble fluoride emissions in 2000
would drop to 50 tons for sintering and 400 tons for pelletizing.
3,3,5,2 Blast Furnace Operations
P1g iron is produced by high temperature reduction of the iron ore
charged to the blast furnace. Other portions of the blast furnace burden
Include coke and limestone; natural gas and air are injected to furnish
part of the heat. The high temperature environment, coupled with the
availability of hydrogen from the combustion products, causes volatiliza-
tion of much of the CaF2 present in the ore feed and conversion of a
portion of the volatilized fluorides to HF.
Process Description. Figures 3-14 and 3-15 present flow diagrams
and mass balances for production of pin iron from normal and high fluor-
ide ores by the use of the blast furnace.
3-69

-------
PROCESS C
ROCESS
PARTICULATE
0% GASEOUS
)NE
RECYCLE
TO SINTER
SPRAY
WATER
5 FAN
200°F AND STACK
ELECTROSTATIC PRECIPITATOR
EPF. = 95% PARTICULATE
0% GASEOUS
SPRAY
SCRUBBER
EFF.
80% (EST.)
PARTICULATE
90% GASEOUS
WASTE
TO DISPOSAL
OR RECYCLE
TO "SINTER FINES
RECYCLE,"
FIGURE 3-11
I50°F
(EST.)
CYCLONE
EFF.
70% (EST.) PARTICULATE
0% GASEOUS
TO FAN
AND STACK
200°F
(EST
KK
TO
FAN
AND
STACK
THICKENER
ELECTROSTATIC
PRECIPITATOR
EFF. = 95% (EST.) PARTICULATE
0% GASEOUS
WASTE TO DISPOSAL
RECYCLE
TO SINTER
CAKE TO .
DISPOSAL
VACUUM
FILTER
>NE
70% (EST
PARTICULATE
0% GASEOUS
"" RECYCLE '
TO SINTER
NOTE: PROCESS A FOR TREATMENT OF EVOLVED GASES FROM SINTERING MACHINE
PROCESSES B AND C FOR TREATMENT OF EVOLVED GASES FROM SINTERING COOLER
ASSUMES NO ADSORPTION OF FLUORIDES ON PARTICULATE MATTER.
RECYCLE
WATER
~-RECYCLE
WATER
(30-50 PPM
SOLIDS)
BASIS - 2000 TONS/DAY OF SINTER PRODUCED
PROCESS STREAMS - TONS/DAY
Materi al
Stream Number
4
6
8
9
10*
11
12
13*
14
15*
HF
CaF2 equiv.
0.77(g)
0.02(s)
0.34(s)
0.02(s)(C)
0.002(s)^C)
0.77(g)
0.0003(s)(Est)
Q.24(s)(Est)
Q.IQ(sKEst)
0.005(s){Est)
0.08(s}^
0.02(s)(Est)
Total Fluorides
Total as F
0.79
0.74
0.34
0.17
0.02
0.01
0.002
0.001
0.77
0.73
0.24
0.12
0.10
0.05
0.005
0.003
0.08
0.04
0.02
0.01
Iron Ore
Water
Air (50% R.H.)
h
h
CO
C02
19(s)
250(g)
12500(g)
2000(g)
1000(g)
900(g)
21 (s)
10000(g)
17(C)
1.9(C)
0.1(s)
250(g)
lOOOO(o)
12500(g)
2000(g)
1000(g)
900(g)
14.7(s)(Est)
6.0(s)(Est)
0.3(s)
10000(g)
5.0(s)(B'
1.3(s)
10000(g)
Approx. Total
Stream
16700
10000
18
2
16700
15
6
10000
5
10000
Source
Soluble Fluoride
Emission Factor -
lb F/ton sinter
Process A
Process B
Process C
Sinter Machine
0.73
-
—
Sinter Cooler
_
-
-
Assumed Fugitive
0
0
0
Total Soluble Fluoride Emission Factor
0.73
0
0
* Gaseous Effluent Stream
(h)
Overall process soluble fluoride emission factor = 0.69 lb F/ton sinter1 '
(A)	Reflects estimated 5% usage of high fluoride (3000 ppm) content ore, 95% usage of average fluoride content (650 ppm) ore and application
of wet control processes to 6.52 (Reference 4225) of the sintering facilities.
(B)	Reference 4006
(C)	Reference 4055
Figure 3-12. Iron Ore Sintering
Controlled Process
Model
3-71

-------
ELECTROSTATIC
PRECIPITATOR
EFF. = 96% PARTICULATE
0% GASEOUS
CaCO,
800° F
CYCLONE
EFF. = 70% (EST.) PARTICULATE
0% GASEOUS
FAN
CYCLONE
EFF. = 90% PARTICULATE
0% GASEOUS
TO STACK
STEAM '
Ca(OH),
TO
DISPOSAL
TO
DISPOSAL
RECYCLE TO
"SINTER FINES RECYCLE
FIGURE 3-12
BASIS - 2000 TONS/DAY OF SINTER PRODUCED
PROCESS STREAMS - TONS/DAY
Materials
Stream Number
4
6
8
9
10
11
12
13
14*
HF
3.01(g)







0.15(g)
CaFg equlv.
0.09(s)
1.35(s)


4.9(s)(B)

1.48(s)(Est)
0.61(s)(B)
0.03(s)(Est)
Total Fluorides
3.10
1.35


4.91

1.48
0.61
0.13
Total as F
2.90
0.68


2.40

0.72
0.30
0.12
Iron Ore
19(s)
21 (s)


36(B)

2.8(Est)
1.1(B)
0.1(s)
Water
250(g)

4000(B)





4250(g)
A1r (50? RH)

10000(g)






10000(g)
N2
12500(g)







12500(g)
°2
2000(g)







2000(g)
CO
1000(g)







1000(g)
co2
900(g)







900(g)
Ca(0H)2



14.4
13.5




CaC03





46^)
32
12
1.0(s)
Approx. Total
Stream
16700
10000
4000^
14
54
46
35
13
30700
* Gaseous effluent stream
(A)	Stream
Soluble fluoride emission factor of facilities utilizing high fluoride content (3000 ppm)
Iron ore = 0.095 lb F/Ton of Sinter Produced
(B)	References 4055 and 4246
Figure 3-13. High Fluoride Content
Iron Ore Sintering -
Controlled Process
Model
3-73

-------
FEED
MATERIALS"
FUEL AND
COMBUSTION AIR"
TO AIR PREHEATERS,
POLLUTION CONTROL
DEVICE AND USE AS
FUEL


BLAST
FURNACE
~w
¦SLAG
PIG
" IRON
BASIS - DAILY FURNACE PRODUCTION OF 1000 TONS OF PIG IRON
PROCESS STREAMS - TONS/DAY .
Ma terials
Stream Number
1
2
3
4
5*
HF
CaF2
1.5 (s)

0
0.77(s)(C,D)
0.046(g)
0.67 (s)
Total Fluorides
Total as F
1.5
0.74

0
0
0.77
0.37
0.71
0.37
Sintered Iron Ore^
Screened Iron Ore^
Coke
CaCO. (Limestone)
Natural Gas
Air (502S R.H.)
Slag
Pig Iron
h2o
n2
h2
CO
C02
Fe203
1150
400
500
60
20
2150
1000
230
70(g)
1600(g)
130(g)
1000(g)
400(g)
25(s)(C)
Approx. Total
Stream
2100
2200
1000
230
3300
* Gaseous effluent stream	(C) Reference 4006
(A)	Contains 0.9 tons of CaF^ .	(U) Reference 4179
(B)	Contains 0.6 tons of CaF2 equiv.
Soluable fluoride evolution factor = 0.088 lb F/ton pig iron
Mote; Fluoride content of the iron ore and sinter reflects 95'* usage of 650 ppm
fluoride content ore and 5S usage of 3000 ppm fluoride content ore.
Figure 3-14. Pig Iron Production - Uncontrolled Process Model
3-75

-------
FEED
MATERIALS"
FUEL AND
COMBUSTION AIR'
TO AIR PREHEATERS,
POLLUTION CONTROL
DEVICE AND USE AS
FUEL


BLAST
FURNACE


-SLAG
PIG
IRON
BASIS - UAILY FURHACE PRODUCTION OF 1U00 TONS OF PIG IRON
PROCESS STREAMS - TOHS/UAY
Materials
Stream Number

1
2
3
4
5*
HF



3.0 (s)
0.18(g)
CaF2
6.0(s)

0
2.60(s)
Total Fluorides
6.0

0
3.0
2.78
Total as F
2.9

0
1.46
1.44
(A)
Sintered Ir:n Grev
1150




Screened Iron Ore'"'
400




Coke
500




CaCO^ (Limestone)
60




llatural Gas

20



Air (MS R.H.)

2150



Slag



230

Pig Iron


1000


h2o




70(g)





1600(g)
H2




130(g)
CO




1000(g)
C02
Fe2°3




400(g)
25(s)(C)
Approx. Total
Stream
2100
2200
1000
230
3300
* Gaseous effluent stream	{C) Reference 4006
(A)	Contains 3.6 tons of CaFj	(0) Reference 4179
(B)	Contains 2.4 tons of CaF,, equiv.
Soluble fluoride evolution factor of facilities utilizing high fluoride content
(3000 ppm) iron ore = 0.34 lb F/ton pig iron
gure 3-15. Pig Iron Production From High Fluoride Content
Iron Ore - Uncontrolled Process Model
3-76

-------
Present and Future Production Levels. Currently, pig iron is pro-
duced at the locations shown in Appendix 7.2. As noted earlier, the
industry practice will shift to pellet or pellet-sinter feed completely.
In 1968, the U.S. steel industry utilized 130 million tons of ore,
both treated and untreated, in blast furnace burdens. The total amount
of pellets charged to the blast furnaces for piq iron production was
50.7 million tons. The total amount of sinter product was 49.8 million tons,
down from 51.6 million tons in 1967 and 54,7 million tons in 1966.
It is expected that the amount of ore tonnage consumed in blast
furnace burdens will decrease slightly between now and the year 2000 to a
level of about 120 million tons annually. This is because of expected
higher furnace efficiencies and greater use of reusable scrap for hot
metal.
Fluoride Emission Control Technigues. Flow diagrams and mass bal-
ances for the current method employed for fluoride control on normal and
high fluoride content ores are presented in Figures 3-16 and 3-17.
Fluoride Emissions. Much of the gaseous HF evolved in the high
temperature reduction zones is adsorbed in the upper, cooler zones of the
burden, or on the heavy load of particulate suspensoid carried by the
blast furnace effluent gases. The portion of the cooled HF adsorbed on
the particulate suspensoid is removed along with the dust in the dust
control eguipment. The remainder of the cooled gaseous HF is emitted
when the blast furnace product gas is burned for energy recovery, or
flared. Additional fluoride emission occurs when the pressure spikes
developed as the result of "slips" (dropping charge breaks) are released
to the atmosphere by the collection system "bleeders" (over-pressure relief
valves).
Soluble fluoride emissions from blast furnace operations were 2800
tons during 1968, and will drop to 2600 tons in 2000 if current control
practices are continued. If controls effective at the 99% level are
adopted, soluble fluoride would drop to 40 tons.
3-77

-------
BASIS - DAILY FURNACE PRODUCTION OF 1000 TONS OF PIG IRON
PROCESS STREAMS - TONS/DAY
ELECTROSTATIC
PRECIPITATOR
EFF. = 90% PARTICULATE
WET	0% GASEOUS
00
I
00
WATER
TO COMBUSTION
* AND STACK
CYCLONE
EFF. = 60% PARTICULATE
0% GASEOUS
TO DISPOSAL
700°F
(EST.)
— RECYCLE WATER
THICKENER
CAKE
TO DISPOSAL
VACUUM
FILTER
RECYCLE
TO SINTER
PLANT
SINTER FINES
RECYCLE
RECYCLE
WATER
A A
RECYCLE
WATER
(30-50 PPM SOLIDS)
Material
Stream Nunber
5
6
7
8
9*
HF
0.046(g)
0.40(s)^
0.041
0.024(s)^
0.0046
CaF2 equiv.
0.67 (s)
0.24 (S)
0.0026
Total Fluorides
0.71
0.40
0.28
0.024
0.007
Total as F
0.37
0.19
0.16
0.012
0.006
h2o
. 70(g)



70(g)
N2
1600(g)



1600(g)
HZ
130(g)



130(g)
CO
1000(g)



1000(g)
C02
Fe203
400(g)
25(s)
15(s)(B)
9(s)
CO

-------
BASIS - DAILY FURNACE PRODUCTION OF 1000 TONS OF PIG IRON
ELECTROSTATIC
PRECIPITATOR
EFF. = 90% PARTICULATE
PROCESS STREAMS - TONS/DAY
-J
<-o

WET
SCRUBBER
EFF. = 90% PARTICULATE
90% GASEOUS
SPRAY
WATER
TO COMBUSTION
AND STACK
CYCLONE
EFF. = 60% PARTICULATE
0% GASEOUS
650°F
(EST.)
TO DISPOSAL
70O°F
(EST.)
RECYLCE WATER
THICKENER
CAKE
TO DISPOSAL
VACUUM
FILTER
RECYCLE
TO SINTER
PLANT
RECYCLE
WATER
A A
RECYCLE
WATER
(30-50 PPM SOLIDS)
NOTE:
ASSUMES NO ADSORPTION OF FLUORIDES
ON PARTICULATE MATTER.
Material
Stream Number
5
6
7
8
9*
HF
CaF2 equiv.
0.18(g)
2.60(s)
1.56(s)(A'
0.16
0.94(s)(A)
0.094(s)(A'
0.018(g)
0.010(s)
Total Fluorides
2.78
1.56
1.10
0.094
0.028
Total as F
1.44
0.76
0.61
0.046
0.022
h2o
70(g)



70(g)
N2
1600(g)



1600(g)
"z
130(g)



130(g)
CO
1000(g)



1000(g)
C02
Fe203
400(g)
25 (s)
15(s)(A)
9(s)(A)
9(s)(A)
400(g)
0.1(s)
Approx. Total
Stream
3300
17
10
9
3300
* Gaseous Effluent Stream
Soluble fluoride emission factor of facilities utilizing high fluoride content
(3000 ppm) iron ore - 0.034 lb.F/ton pig iron
(A) Reference 4055
Figure 3-17.
Pig Iron Production From High Fluoride Content
Iron Ore - Controlled Process Model

-------
3.3.5.3 Open Hearth Furnace Operations
Open hearth steelmaking in the U.S. dates back to 1870 and is cur-
rently declining in importance as the basic oxygen furnace process increases
in capacity. The open hearth is actually a shallow hearth inside a rec-
tangular furnace. The furnace charge, composed of molten pig iron ("hot
metal"), scrap and flux, is heated by mixtures of natural qas, tar, and
oil. High temperature oxidation of the carbon, silicon and manganese con-
tained in the hot metal converts the charge to steel. The flux forms a slao
with the oxidized silicon and manganese, and with portions of the sulfur and
phosphorus impurities. The oxygen necessary for the oxidation comes, in the
older furnaces, from the air and iron oxide portions of the charge; in the
newer furnaces, this is supplemented with gaseous oxygen introduced through a
water-cooled cover.
The flux employed is limestone, with fluorspar added. The fluorspar
serves as the source for the majority of the fluorides evolved from the
molten charge. Half of the evolved fluorides are converted to gaseous hydro-
gen fluoride and emitted from the furnace.
Process Description. Figures 3-18 and 3-19 present the process models
and mass balances for the manufacture of steel (using both normal and high
fluoride content iron ore) via the open hearth process without oxygen sources.
Production Trends. Open hearth operations accounted for 90% of steel
production in the period after World War II. Open hearth output peaked at
105 million tons in 1955, and declined to 66 million tons in 1968 and about
60 million tons in 1969. (^286,4287) ^ es-timated that open hearth oro-
duction will decline at a rate of about 12%. annually to a near zero level,
and will remain at that level through the year 2000.
Fluoride Emission Control Techniques'. Figures 3-20 through 3-22
present process models and mass balances for the processes currently
employed for control of emissions from open hearth operations. Normally,
the covered open hearth furnaces are vented through checkerwork regenerators
before passing the gases on to the waste heat boilers and dust abatement
3-80

-------
NON-OXYGEN LANCED
TO ATMOSPHERE
OR POLLUTION
CONTROL DEVICE
EXHAUST LEAVING
REGENERATIVE CHAMBERS
ATAPPROX. 1300°F
FEED
MATERIAL
-SLAG
OPEN
HEARTH
FURNACE
LOW QUALITY STEEL
(SCRAP) FOR RECYCLE
2900°F
COMBUSTION
FUEL AND AIR
- STEEL
NOTE: (1) IF A WASTE HEAT BOILER IS
INSTALLED, THE TEMPERATURE
OF THE GASES LEAVING THIS
UNIT WOULD BE ABOUT 500°F.
BASIS - FURNACE CAPACITY OF 200 TONS/HEAT - N0N OXYGEN LANCED
(2 HEATS PER DAY 0 APPR0X. 10 HOURS PER HEAT)
PROCESS STREAMS - TONS/HEAT
Materi al
Stream Number
1
2
3
4
5*
6
HF
CaF2
0.26(s)^


0.13 (s)
0.066(g)

Total Fluorides
Total as F
0.26
0.13


0.13
0.063
0.066
0.063 -

Pig Iron (Hot)
Scrap
Iron Ore^
CaC03 (Limestone)
Raw Steel
Low Quality Steel
for Recycle
Slag
N2
°2
h2o
co2
so2
Fe2°3
Air (50% R.H.)
Fuel Oil
100
70
15
15
155
5
18
i
180(g)
5(g)
12(g)
75(g)
0.3(g)
l(s)(B)
t
235
15
Approx. Total
Stream
200
155
5
18
270
250
* Gaseous effluent stream
(A)	Contains 0.02 tons of CaF2 equivalent
(B)	Reference 4006
(C)	Reference 889
Soluble fluoride evolution factor = 0.81 lb F/ton steel
Note: Fluoride content of the iron ore reflects 95% usage of 650 ppm fluoride
content ore and 5% usage of 3000 ppm fluoride content ore.
Figure 3-18. Basic Open Hearth Steelmaking -
Uncontrolled Process Model
3-81

-------
NON-OXYGEN LANCED
TO ATMOSPHERE
OR POLLUTION
CONTROL DEVICE
FEED
MATERIAL
COMBUSTION
FUEL AND AIR"



EXHAUST LEAVING
REGENERATIVE CHAMBERS
ATAPPROX. I300°F
OPEN
HEARTH
FURNACE
2900°F

-»-SLAG
^ LOW QUALITY STEEL
ISCRAP) FOR RECYCLE

«- STEEL
NOTE: (1) IF A WASTE HEAT BOILER IS
INSTALLED, THE TEMPERATURE
OF THE GASES LEAVING THIS
UNIT WOULD BE ABOUT 500°F.
BASIS - FURNACE CAPACITY OF 200 TONS/HEAT - NON OXYGEN LANCED
(2 HEATS PER DAY @ APPROX. 10 HOURS PER HEAT)
PROCESS STREAMS - TONS/HEAT
Material
Stream Number
1
2
3
4
5*
6
HF
0.33(s)^


0.17 (s)(C)
0.084(g)

CaF2




Total Fluorides
0.33


0.17
0.084

Total as F
0.16


0.080
0.080 *

Pig Iron (Hot)
100





Scrap
Iron Ore^
70
15





CaC03 (Limestone)
15





Raw Steel

155




Low Quality Steel
for Recycle


5



Slag



18


n2




180(g)

°2




5(g)

h2o




12(g)

co2




75(g)

so2
Fe2°3
Air (50% R.H.)




0.3(g)
Us)(B>
235
Fuel Oi 1





15
Approx. Total
Stream
200
155
5
18
270
250
* Gaseous effluent stream
(A)	High fluoride content (3000 ppm) ore con-
taining 0.09 tons of CaF2 equiv.
(B)	Reference 4006
(C)	Reference 889
Soluble fluoride evolution factor = 1.03 lb. F/ton steel
Figure 3-19. Basic Open Hearth Steelmaking
Utilizing High Fluoride Con-
tent Iron Ore - Uncontrolled
Process Model
3-83

-------
ELECTROSTATIC
PRECIPITATOR
EFF. = 98% PARTICULATE
WASTE HEAT
BOILER
0% GASEOUS
TO FAN
AND STACK
1300°F
500°F
PROCESS A
TO DISPOSAL
BAG HOUSE
WASTE HEAT	SECONDARY EFF. = 99.9% PARTICULATE
BOILER
GASEOUS
EXCHANGER
TO FAN
"AND STACK
500°F
220° F
PROCESS B
TO DISPOSAL
NOTE: ASSUMES NO ADSORPTION OF FLUORIDES
ON PARTICULATE MATTER.
BASIS - FURNACE CAPACITY OF 200 TONS/HEAT - NON-OXYGEN LANCED
PROCESS STREAMS - TONS/HEAT
Material
Stream Number
5
7
8*
9
1"*
HF
0.066(g)

0.066(g)

0.066(g)
Total Fluorides
Total as F
0.066
0.063

0.066
0.063

0.066
0.063
n2
°2
h2o
co2
S°2
Fe2°3
180(g)
5(g)
12(g)
75(g)
0.3(g)
Kg)
0.98(s)'A*
180(g)
5(g)
12(g)
75(g)
0.3(g)
0.02(s)
0.999(s)'A'
180(g)
5(g)
12(g)
75(g)
0.3(g)
O.OOI(s)
Approx. Total
Stream
270
1
270
1
270
* Gaseous Effluent Stream
(A) References 4006 and 4055
Figure 3-20. Basic Open Hearth Steelmaking - Controlled
Process Model (Processes A and B)
3-85

-------
SECONDARY
EXCHANGER
WASTE HEAT
BOILER /
VENTURI SCRUBBER
EFF. = 95% PARTICULATE
90% GASEOUS
TO FAN
AND STACK
250° F
SEPARATOR
PROCESS C
VACUUM
FILTER
MAKE-UP
WATER
CAKE
TO DISPOSAL
THICKENER
WATER FOR
RECYCLE TO
SEPARATOR OR
VENTURI
NOTE: ASSUMES NO ADSORPTION OF FLUORIDE ON	SOLIDS)
PARTICULATE MATTER.
BASIS - FURNACE CAPACITY OF 200 TONS/HEAT - NON-OXYGEN LANCED
PROCESS STREAMS - TONS/HEAT
Material
Stream Number
5
11
12*
HF
CaF2
0.066(g)
0.059(1)
0.007(g)
Total Fluorides
Total as F
0.066
0.063
0.059
0.056
0.007
0.0067
n2
°2
H20
C02
S02
Fe2°3
Ca(0H)2
180(g)
5(g)
12(g)
75(g)
0.3(g)
Ks)
0.95(s)^
180(g)
5(g)
6(g)
75(g)
0.03(g)(Est)
0.05(s)
Approx. Total
Stream
270
1
270
Source
Soluble Fluoride Emission Factor - lb F/ton steel
Process A
Process B
Process C
Furnace
Ass lined Fugitive
0.81
0
0.81
0
0.086
0
Total Soluble Fluoride
Emission Factor
0.81
0.81
0.086
* Gaseous Effluent Stream
(A)
Overall process soluble fluoride emission factor = 0.77 lb F/ton steelv '
(A)	Reflects estimated 5% usage of high fluoride content (3000 ppm) ore,
95% usage of average fluoride content (650 ppm). ore and application of
Process C to 6.1% (Reference 4225) of operating facilities.
(B)	References 4006 and 4055 '
Figure 3-21. Basic Open Hearth Steelmaking - Controlled
Process Model (Process C)
3-86

-------
ELECTROSTATIC
PRECIPITATOR
EFF. = 95% PARTICULATE
0% GASEOUS
Ca(OH)2
INJECTION _
WATER
. INJECTION
CYCLONE
EFF. = 60% PARTICULATE
0% GASEOUS
WASTE HEAT
BOILER
TO FAN
AND STACK
TO
DISPOSAL
PROCESS D
TO
DISPOSAL
BASIS - FURNACE CAPACITY OF 200 TONS/HEAT - HON-OXYGEN LANCED
PROCESS STREAMS - TONS/HEAT
Material
Stream Number
5
13
14
15*
16
17
HF
CaF2
0.084
0.094(s)(B)
0.059(s)(A)
0.004(g)(A)
0.002(s)(A)


Total Fluorides
0.084
0.094
0.059
0.006


Total as F
0.080
0.046
0.029
0.005


N2
180(g)


180(g)


°2
h2o
5(g)
12(g)


5(g)
42(g)(A)

30(g)(A)
C02
75(g)


75(g)


S02
Fe2°3
Ca(0H)2
0.3(g)
Ks)
0.60(s)^
1.3(s)(B)
0.76(s)(A)
0.03(g)(Est)
0.04(s)(A)
2.1(s)(A)

Approx. Total
Stream
270
2
1
270
2
30
* Gaseous Effluent Stream
Soluble fluoride emission factor of facilities utilizing high fluoride content
(3000 ppm) iron ore = 0.052 lb F/ton steel
(A)	Reference 4053
(B)	Reference 4055
Figure 3-22. Basic Open Hearth Steelmaking Utilizing High
Fluoride Content Iron Ore - Controlled
Process Model
3-87

-------
systems. It should be noted that Processes C and D are the only systems
currently in use which are capable of abating soluble fluoride emissions from
the open hearth furnace. Process C is used to a minor extent only (6.1%
of open hearth emission control installations4225), and Process D is cur-
rently employed only in conjunction with Utah open hearth facilities which
charge high fluoride content iron ore to the furnace.
Fluoride Emissions. Soluble fluoride emissions in 1968 from open
hearth operations are estimated at 25,400 tons. Fluoride emissions will
decline to 0 in 2000 because of the phase-out of open-hearth furnace steel
making.
3.3.5.4 Basic Oxygen Furnace Operations
The major materials employed in the basic oxygen steelmaking process
(B0F) are pig iron (hot metal), scrap, flux, and gaseous oxygen. The flux
is composed of burnt lime (90%) and fluorspar (101). No external heat is
supplied - the heat produced by the reactions between the gaseous oxygen
blown into the molten charge and the metals of the charge is sufficient to
produce steel. While large guantities of calcium fluoride are volatilized,
the absence of any hydrogen source prevents conversion of the volatilized
fluoride to gaseous HF. Thus, based on thermochemical equilibrium calcula-
tions no measurable quantities of soluble fluorides are emitted by the B0F
process. This conclusion should be verified experimentally.
Process Description. Figure 3-23 presents a flow diagram and mass
balance for a basic oxygen steel making furnace rated with a capacity of
200 tons per heat.
Production Trends. In integrated steel plants, the open hearth fur-
nace has been the predominant steelmaking process, however, the basic oxygen
furnace has become increasingly important and in 1970 surpassed the open
hearth. (In August 1969 the monthly output of BOFs actually exceeded that
of onen hearths for the first time.) In nonintegrated plants the open
hearth has essentially been displaced by the electric arc furnace.
3-88

-------
BASIS - FURNACE CAPACITY OF 200 TONS/HEAT
(12 HEATS PER DAY)
PROCESS STREAMS - TONS/HEAT
TO ATMOSPHERE
OR POLLUTION
CONTROL DEVICE
$
HOOD
MATERIALS
BAS C
OXYGEN
FURNACE
2900°F
OXYGEN
NOTE:
(1) ASSUMED HOOD EFFICIENCY
OF 90%.
HOOD LOSS
TO ATMOSPHERE OR
CONTROL DEVICE
SLAG
LOW QUALITY STEEL
(SCRAP) FOR RECYCLE
RAW STEEL
Materials
Stream Number
1
2
3
4
5
6
7*
8*
CaF2
1.04(s)(C'D)



0.51(s)
0.51(s)
0.46(s)
0.05 (s)
Total Fluoride
1.04



0.51
0.51
0.46
0.05
Total as F
. .51



0.25^
0.25^
0.22
0.025
Pig Iron (Hot)
130







Scrap
60







CaO (Burnt Lime)
10







°2

11



9(g)
9(g)
Kg)
Low Quality Steel
for Recycle



6




Slag




20



Raw Steel


165





co2





3(g)
3(g)
0.3(g)
Fe2°3





3.3(s)(B)
3.3(s)^
0.3(s)
Approx. Total
Stream
200
11
165
6
20
16
16(A)
2
* Gaseous effluent stream
(A)	Plus 350,000 CFM^of dilution air
Soluble Fluoride Evolution Factor = 0 lb F/ton steel produced
(B)	Reference 4006
(C)	Reference 4246
(D)	Reference 4248
(E)	Reference 4055
(F)	Reference 889
Figure 3-23. Basic Oxygen Steelmaking - Uncontrolled Process Model

-------
In 1954, the first BOF unit was installed in the United States. From
levels of only a few million tons in 1960, oxygen steelmakina capacity beqan
to approach installation rates on the order of 10 million tons a year in the
mid-1960s, and this nominal rate continues. In 1968, BOF shop accounted for
about 37% of U.S. steel production, and their share exceeded 50% in 1970.
Many BOF operations have been installed adjacent to comparatively larne open
hearth shops, to replace or augment open hearth capacity. The BOF process
is ideally suited for low-carbon grades, and by the end of 1969 the majority
of this steel was produced by the B0F.^4286» 4287^ It is expected that BOF
production will increase at a rate of 7.6% annually to a constant value of
approximately 135 million tons through the year 2000.
Fluoride Emission Control Techniques. Figure 3-24 presents process
models and mass balances for currently employed dust and fume control systems.
Basic oxygen steel furnaces are hooded by "dust reclaimers" for
collection and transport of pollutants to dust control equipment. Gas
temperatures may reach 3000° at the collection point.
Fluoride Emissions. Soluble fluoride pollutants from this source are
essentially zero since the fluorides emitted are in the form of CaF2, a
relatively harmless insoluble solid. Bases for this conclusion are discussed
above.
3.3.5.5 Electric Arc Furnace Operations
Electric arc furnace steelmaking, with two exceptions, does not use
hot metal as a part of the charge. The charge materials in the majority of
cases, are composed of solid steel scrap of high quality, iron oxide (ore or
mill scale), burnt lime, fluorspar, carbon, and gaseous oxygen. The heat
necessary is supplied as electrical energy.
Considerable amounts of calcium fluoride are evolved. Because of the
absence of a hydrogen source, there is no conversion of any of the evolved
calcium fluoride to hydrogen fluoride. Therefore, on the basis of
3-90

-------
SPRAY
ELECTROSTATIC
PRECIPITATOR
EFF< = 99% PARTICULATE
0% GASEOUS
) PARTICULATE
.) GASEOUS
TO FAN
AND STACK
500° F
(20-30% H90)
450° F
SPRAY
WATER
RECYCLE
WATER
(30-50 PPM
SOLIDS)
750° F
TO
DISPOSAL
OR SINTER
PLANT
WATER
THICKENER
VACUUM
FILTER
CAKE
TO DISPOSAL
PROCESS A
7\ A
WATER
<3>
QUENCH
TOWER
VENTURI
SCRUBBER
EFF. = 99% PARTICULATE
90%(EST.) GASEOUS



250° F »

L_*~
n Ti

<2>



750° F


t>T
WATER
TO FAN
AND STACK
CYCLONIC
CLARIFIER
EFF. = 50% (EST.) PARTICULATE
VACUUM
FILTER

PROCESS B
NOTE: ASSUMES NO ADSORPTION OF FLUORIDES
ON PARTICULATE MATTER.
TO
DISPOSAL
CAKE
TO DISPOSAL
RECYCLE
WATER
(30-50 PPM
SOLIDS)
BASIS - FURNACE CAPACITY OF 200 TONS/HEAT (12 HEATS PER DAY)
PROCESS STREAMS - TONS/HEAT
Materi al
Stream Number
7
9
10
n*
12
13
14
15
16*
CaF2
0.46(s)

0.13(s)(C)
0.002(s)(C)
0.32(s)
0.23(s)

0.23(s)
0.005(s)^D)
Total Fluorides
Total as F
0.46
0.22

0.13
0.06
0.002
0.001
0.32
0.16
0.23
0.12

0.23
0.12
0.005
0.0025
°2
co2
Fe203
h2q
9(9)
3(g)
3.3(s)
250(1)(B)
0.99(s)(C)
9(g)
3(g)
0.01{s)(C)
250(g)
2,3{s)(Est)
1.7(s) (Est)
1.7(1)(Est)
80(1) (Est)
1,6(s)(Est)
9(g)
3(g)
0.03(s)(D)
80(g)(Est)
Approx. Total
Stream
i6{a)
250^
1

-------
thermochemical calculations, calcium fluoride is the only fluoride emitted
by the electric arc furnace in measurable quantity.
Process Description. Figure 3-25 presents the process model and
mass balances for electric arc steelmaking.
Production Trends. The combination of low capital investment and
flexibility in use of scrap has led in the past several years to the increased
installation of electric furnace steelmaking shops in areas remote from
major steelmaking centers and even more recently in the heart of the steel-
making districts themselves. By locating in areas remote from the major
steelmaking centers, a small semi-integrated or nonintegrated shop has sev-
eral economic advantages. The price of scrap in the local area is lover
than in the major steelmaking districts because with no transportation costs
it is, in effect, discounted from the standard price; likewise, the products
of the local nonintegrated plant also bear no transportation costs and thus
can be priced competitively with products shipped to the area from a major
steel producing plant. At an integrated steel plant, electric furnace facil-
ities offer the flexibility of easily started peak shaving capacity and the
benefits of using in-plant generated scrap unsuitable for charging to
the BOF. The electric furnace has great flexibility to produce a wide variety
of steels, ranging from low-grade to high-quality steels.
The production of steel by electric arc furnace is expected to grow
at a rate of 8.3% (4286, 4287) from a 1968 level of 16 million tons to a con-
stant 35 million tons annually through 2000.
Fluoride Emission Control Techniques. Figure 3-26 presents the
process models and mass balances for current processes for electric arc.
furnace emission control.
Electric arc melting furnaces are hooded in one of three general
types: (1) canopy hoods, (2) enclosing or roof-ring hoods, and (3) direct
furnace taps. Canopy hoods are located above the crane-way but require large
volumes of indraft air to capture furnace effluent efficiently. Roof-ring
3-93

-------
WITH OXYGEN INJECTION TO FAN AND STACK OR
wimuA^tiN	CONTROL SYSTEM
FEED
MATERIALS
OXYGEN
INJECTION'
ELECTRIC
POWER


ELECTRIC
ARC
FURNACE
2900°F
HOOD

<5>

HOOD LOSS (2)
•-TO ATMOSPHERE OR
CONTROL DEVICE
-SLAG
LOW QUALITY STEEL
"(SCRAP) FOR RECYCLE
- RAW STEEL
NOTES:
(1)	FURNACE EQUIPPED WITH
AN INTEGRAL HOOD.
(2)	ASSUMED HOOD EFFICIENCY
OF 95%.
BASIS - FURNACE CAPACITY OF 75 TONS/HEAT WITH OXYGEN INJECTION
(4 HEATS PER DAY)
PROCESS STREAMS - TONS/HEAT

Stream Number
Materials
1
2
3
4
5
6
7
8*
9*'
CaF2
0.3(s)(C'D)
v



0.14(s)
0.13(s)
0.12(s)
0.01(s)
Total Fluorides
0.3




0.14
0.13
0.12
0.01
Total as F
0.15




0.07
0.07^
0.06^
o.offr
Steel Scrap
70








Mill Scale
0.5








Carbon (Used
Electrodes)
0.3








CaO (Burnt Lime)
3








°2

1




0.8(g)
0.8(g)
0.08(q)(Est.)
Slag





5



Low Quality Steel
for Recycle




2




co2






i.i(g)
i.Kg)
0.1(g)(Est.)
Fe2°3






0.5(s)
0.5(s)(B)
0.05(s)(Est.)
Raw Steel



67





Electrical Power


160.000KW






Approx. Total
Stream
75
1
-
67
2
5
2.5
2.5(A)
0.25
* Gaseous effluent stream
(A)	Plus 150,000 CFM of dilution air
Soluble fluoride evolution factor = 0.16 lb F/ton steel produced
(B)	Reference 4006
(C)	Reference 4246
(D)	Reference 4248
(E)	Reference 889
Figure 3-25. Electric Arc Steelmaking -
Uncontrolled Process Model
3-95

-------
<*>
WATER COOLED
DUCT
BAG HOUSE
EFF. = 99% PARTICULATE
0% GASEOUS
250°F
PROCESS A

TO FAN
AND STACK
<3>
TO
DISPOSAL
<£>
VENTURI
SCRUBBER
WATER COOLED EFF. = 95% (EST.) PARTICULATE
DUCT		85% (EST.) GASEOUS
1 250° F
MAKE-UP
WATER
PROCESS B
WATER
RECYCLE (30-50 PPM SOLIDS)
CAKE
TO DISPOSAL
RECYCLE
WATER

CYCLONIC
SEPARATOR
THICKENER
VACUUM
FILTER
ELECTROSTATIC
PRECIPITATOR
EFF. = 95% PARTICULATE


500° !=¦
PROCESS C
TO
NOTE: ASSUMES NO ADSORPTION OF FLUORIDES DISPOSAL
ON PARTICULATE MATTER.
SPARK
BOX
TO FAN
AND STACK
TO FAN
AND STACK
BASIS - FURNACE CAPACITY OF 75 TONS/HEAT WITH OXYGEN INJECTION
PROCESS STREAMS - POUNDS/HEAT
Materials
Stream Number
8
10
11*
12*
13
14 '
15*
CaF2
240(s)
237(s)^
2(s)(B)
12(s)(B*
228(s)(B*
228(s)^
*12(s)(B)
Total Fluorides
240
237
2
12
228
228
12
Total as F
118
115
1
6
111
111
6
°2
1600(g)

1600(g)
1600(g)


1600(g)
co2
Fe2°3
2200(g)
1000(s)
990(s)(B)
2200(g)
10(g)
2200(g)
50(S)^
950(s)^
950(s)(B*
2200(g)
50(s)(B)
h2o



100( g) (Est)



Approx. Total
Stream
4900^
1200
3800^
3900^
1200
1200
3900^
Source
Soluble Fluoride
Emission Factor
- lb F/ton steel
Process A
Process B
Process C
Treated Hood Effluent
0
0
0
Hood Loss to Atmosphere
0
0
0
Total Soluble Fluoride Emission
0
0
0
* Gaseous Effluent Stream
Overall soluble fluoride emission factor = 0 lb F/ton steel
(A)	Plus 150,000 cfm of dilution air
(B)	References 4006 and 4055
Figure 3-26. Electric Arc Steelmaking -
Controlled Process Model
3-97

-------
hoods and direct furnace tap hoods require the smallest volumes of indraft
air for efficient collection. The hot furnace exit gas is cooled via
evaporative cooling or radiation coolers and then transported by induced
draft to the dust abatement systems shown in the process models.
Fluoride Emissions. No measurable amounts of soluble fluorides are
emitted by electric arc furnace operations, based on thermochemical
equilibrium calculations.
3.3.6 Economic Analysis
3.3.6.1 Basic Processes
Three specimen cases have been selected for economic analysis as
representatives of some of the existing combinations of process elements
which occur in the iron and steel industry. These specimen cases are
presented in Table 3-24. The steel scrap prices used vary as a function of
the types and quantities of scrap required by the process being considered.
Case A, the simplest, is the economic model for an isolated 500,000
ton per year electric furnace steel plant, using steel scrap and finishing
additions as the metal charge. Return on investment (equity) for this
four-furnace plant, with no fluoride control process, is estimated at 15.3%.
After the addition of pollution control equipment currently employed by the
industry, return on investment drops to 14.4%.
Case B is the economic model for an integrated iron and steel
"division" which purchases pelletized taconite ore, is equipped with coke
ovens, produces pig iron in its own blast furnaces, and makes 2 million tons
of steel per year in a basic oxygen furnace line. Return on investment prior
to use of fluoride control processes is estimated at 7.5%; subsequent to
adaption of pollution controls currently employed by the industry, return on
investment is estimated at 7.3%.
3-99

-------
Table 3-24. Estimated Economics of Electric Furnace Steel Production,^ Case A
(Pollution Control Cost Excluded)
Plant Capacity/.x
0.5 MM Tons/Year^
Total Capital Investment
60.9 $MM
Capital Charges

Depreciation (at 5% per year)
Interest (at 7%)
Local Taxes and Insurance (at 1.5%)
Total Capital Costs ($MM/year)
3.1 $MM/year
0.9
0.9
4.9
Operating Costs for Electric Furnace Facility

Steel Scrap (1.068 tons/NTS at 36.00 $/ton(c)
Finishing Additions (0.007 tons/NTS at $210 (Ave)/ton)
Electric Power (480 kwh/NTS at $0.007/kwh)
Electrodes (9.5 lb/NTS at 0.291 $/ton)
Burnt Lime (0.04 tons/NTS at 16.00 $/ton)
Rpf rs rtnn pc
Oxygen (11 lb/NTS at 12.00 $/NTS)
Repairs and Maintenance
Labor (0.8 Man-Hrs/NTS at 5.00 $/Man Hrs)
38.45 $/NTS
1.47
3.36
2.76
0.64
1.60
0.07
1.80
4.00

-------
Table 3-24. Estimated Economics of Electric Furnace Steel Production,^ Case A
(Pollution Control Cost Excluded) (Continued)
Plant Capacity
0.5 MM Tons/Year
(b)
Utilities	0.45
Yard Switching and Slag Disposal	0.40
Miscellaneous	0.35
General Overhead	0.50
Total Operating Cost for E. F. Facility	55.85 $/NTS
Operating Costs for Continuous Casting
Billets
Gross Metallics (1.042 tons molten steel/NTP at 55.85 $/NTS)	58.20 $/NTP
Scrap Credits (0.042 tons scrap/NTP)	(1.14)
Other Costs	7.00
Total Operating Cost for Billets	64.06 $/NTP
Wire Rod
Continuously Cast Billets (1.075 tons/ton	rod at 64.06 $/NTS) 68.86 $/NTP
Scrap Credit (0.065 tons/NTP)	(1.61)
Other Costs	14.75
Total Operating Cost for Wire Rods	82.00 $/NTP
Merchant Bar
Continuously Cast Billets (1.11 tons/NTP	at 64.06 $/NTS) 71.11 $/NTP
Scrap Credit (0.10 tons/NTP)	(2.50)
Other Costs	12.25	
Total Operating Cost for Merchant Bar	80.86 $/NTP

-------
Table 3-24. Estimated Economics of Electric Furnace Steel Production/3^ Case A
(Pollution Control Cost Excluded) (Continued)
Plant Capacity/.v
		0.5 MM Tons/Year^ '
Total Weighted Average Operating Cost^ ($/ton)
Capital Charges Per Net Ton of Final Product
(at 0.875 product tons/ton molten steel)
Total Average Manufacturing Cost
General and Sales Expenses ($/NTP)
F.o.b. cost ($/NTP)
Average Product Revenue^
Profit After Taxes
Cash Flow
ROI^
NTS = Net tons (short) of molten steel
NTP = Net tons product
(a)
Semi-integrated plant-four 37.4 (Net tons) electric furnaces, casting machines and mills
^Capacity based on molten steel produced
^Price may range between 40.00 to 50.00 $/ton
^Based on 38% (weight) wire rod and 62% merchant bar production
^Return on investors equity (assumed to be 80% of installed capital)
81.29 $/NTP
11.20
92.49
1.85
94.34
128.50
18.01 $/NTP
7.88 $MM/year
15.3%

-------
Table 3-24. Estimated Economics of BOF Integrated Steel Production,^ Case B
(Pollution Control Cost Excluded)
Plant Capacity
2MM Tons/Year^
Total Capital Investment	720 $MM
Capital Charges
Depreciation (at 5%/year)	36.0 $MM/year
Interest (at 7.0%)	10.1
Local Taxes and Insurance (at 1.5%)	10.8	
Total Capital Costs ($MM/year)	56.9
Operating Costs for Blast Furnace
Taconite Pellets (1.6 tons/NTI at 13.70/ton)	21.92 $/NTI
Coke(c) (0.6 tons/NTI at $15.70/ton)	9.42
Limestone (0.26 tons/NTI at 2.05 $/ton) < 0.53
Oil (0.04 tons/NTI at 24.88 $/ton)	1.00
Gas Credit	(0.71)
Dust and Sludge Credit	(0.06)
Labor (0.1 Man-Hrs/NTI at 5.00 $/Man-Hr)	0.50
Utilities	1.00
Refractories	0.15
Reline Costs	0.60
Maintenance and Repair	0.75
Miscellaneous Supplies	0.78
General Overhead	0.45
Total Blast Furnace Operating Cost	36.33 $/NTI

-------
Table 3-24. Estimated Economics of BOF Integrated Steel ProductionCase B
	(Pollution Control Cost Excluded) (Continued)	
Plant Capacity
2MM Tons/Year^
Operating Costs for a Basic Oxygen Facility
Hot Metal (0.8 tons/NTS at 36.33 $/ton)	29.06 $/NTS
Steel Scrap (0.34 tons/NTS at 26.50 $/ton)	9.01
Finishing Additions (0.007 tons/NTS at APX 210 $/ton)	1.47
Burnt Lime (130 lb/NTS at $16.00/ton)	1.04
Refractories	1.40
Oxygen (148 lb/NTS at 12.00 $/ton)	0.89
Labor (0.6 Man-Hrs at 5.00 $/Man-Hr)	3.00
Repairs and Maintenance	2.88
Utilities	0.40
Yard Switching and Slag Disposal	0.45
Miscellaneous Supplies and Services	0.40
General Overhead	0.50	
Total Operating Cost for BOF Facility	50.50 $/NTS
Operating Costs for Continuous Casting
Slabs
Gross Metallics (1.042 tons molten steel/NTP at $50.50/NTS)	52.60 $/NTP
Scrap Credits (0.042 tons scrap/NTP)	(1.00)$/NTP
Other Costs	6.00	
Total Operating Cost for Slabs	57.60
Hot-Rolled Sheets
Slabs (1.11 tons/NTP at 57.60 $/NTS)	63.93/NTP
Scrap (0.087 tons/NTP)	(2.15)
Other Costs	11.09	
Total Operating Cost for Hot-Rolled Sheets	72.87 $/NTP

-------
Table 3-24. Estimated Economics of BOF Integrated Steel Production/3^ Case B
(Pollution Control Cost Excluded) (Continued)
Plant Capacity
2MM Tons/Year^b)
Cold-Rolled Coils (Annealed and Tempered)
Slabs (1.142 tons/NTP at 57.60 $/NTS
Scrap (0.035 tons/NTP)
Other Costs
Total Operating Cost for Hot Rolled Sheets
Cold-Rolled Sheets
Slabs (1.202 tons/NTP at 57.60 $/NTS)
Scrap Credit (0.06 tons/NTP)
Other Costs
Total Operating Cost for Cold-Rolled Sheets
Total Weighted Average Operating Cost ^ ($/ton)
GO
Capital Charges Per Net Ton of Final Product
o	(0.837 product tons/ton molten steel)
CJ1
65.76
$/NTP
(3.40)

20.63

82.99
$/NTP
69.29
$/NTP
(4.91)

25.67

90.05
$/NTP
81.85
$/NTP
33.99
$/NTP
Total Average Manufacturing Cost ^	115.84
General and Sales Expenses ($/NTP)	2.32
F.o.b. Cost ($/NTP)	118.16
(d)
Average Product Revenue	169.91
Profit After Taxes (Tax at 50%)	25.88 $/NTP
Cash Flow	79.3 $MM/year
R0I^e)	7J5%	
NTI = net tons (short) iron
[fllntegrated plant - coke ovens, blast furnace, steelmaking furnaces (BOF), casting machines and mills
/ I Based on molten steel production
/.^Assume: 1.44 tons low S coal/ton (coke), $4.73/ton credit and $ 6.00/ton	operating costs
*	'Based on 28% (weight) hot rolled sheets, 48% annealed and tempered cold rolled coils and 24% cold rolled
/ xsheets produced
*	Return on investors equity (assuming equity is 80% of installed capital)

-------
Table 3-24. Estimated Economics of OH Integrated Steel Production,^ Case C
	(Pollution Control Cost.Excluded)	
Plant Capacity
2MM Tons/Year^
Total Capital Investment	742 $MM
Capital Charges
Depreciation (at 5%/year)	37.1
Interest (at 7%)	10.4
Local Taxes and Insurance (at 1.5%)	11.1
Total Capital Cost ($MM/year)	58.6
Operating Costs for Sintering
Mesabi Iron Ore (0.95 tons/ton si-nter at $10.60/ton)	10.07 $/ton
sinter
Limestone (0.06/ton sinter at $2.05/ton)	0.13
Coke(c) (0.25 tons/ton sinter at $15.70/ton)	3.93
Water (5.0 gal/ton sinter at $0.08/1000 gal)	0.01(d)
Natural Gas (400 scf/ton at $0.35/1000 scf)	0.14
Dust Credit	(0.30)
Labor (0.1 man-hr/ton sinter at $5.00 /Man-Hr)	0.50
Utilities	0.60
Maintenance	1.00
Miscellaneous Supplies	.0.40
General Overhead	0.40
Total Sintering Operating Cost	16.88 $/ton
sinter
Operating Costs for Blast Furnace
Sinter (1.15 tons/NTI at $16.88/ton)
Screened Ore (0.4 tons/NTI at $10.60/ton)
19.44 $/NTI
4.19

-------
Table 3-24. Estimated Economics of OH Integrated Steel Production/3^ Case C
	(Pollution Control Cost Excluded) (Continued)	
Plant Capacity
2MM Tons/Year^
Coke^ (0.52 tons/NTI at $15.70/ton)	8.05
Limestone (0.06 tons/NTI at $2.05/ton)	0.13
Natural Gas (1 scf/NTI at $0.35/1000 scf)	0.01(d)
Gas Credit (4850 pounds)	(0.71)
Dust and Sludge Credit	(0.06)
Labor (0.1 Man-Hr/NTI at $5.00/Man-Hr)	0.50
Utilities	1.00
Refractories	0.15
Reline Costs	0.60
Maintenance and Repair	0.75
Miscellaneous Supplies	0.78
General Overhead	0.45
Total Blast Furnace Operating Cost	35.28 $/NTI
Operating Costs for an Open-Hearth Facility
Hot Metal (0.66 tons/NTS at $35.28/NTI)	23.20 S/NTS
Scrap Steel (0.44 tons/NTS at $36.00/ton)	15.77
Ferro Alloys (0.00 7 tons/NTS at Apx $210/ton)	1.47
Flux (0.09 ton CaF2/NTS at $65.00/ton)	5.85
Iron Ore (0.1 ton ore/NTS at $10.60/ton)	1.06
Fuel Oil (0.09 ton oil/NTS at $24.88/ton)	2.24
Labor (0.5 Man-Hr at $5.00/Man-Hr)	2.50

-------
Table 3-24. Estimated Economics of OH Integrated Steel Production,^ Case C
(Pollution Control Cost Excluded) (Continued) 	
Plant Capacity
2MM Tons/Year^
Maintenance
Utilities
Switching and Slag Disposal
Miscellaneous Supplies and Services
General Overhead
Total Operating Cost for OH Facility
3.00
0.40
0.50
0.40
0.50
56.89
Slabs

Gross Metal lies (1.042 tons molten steel/NTP at 56.89/NTS)
Scrap Credits (0.042 tons scrap/NTP)
Other Costs
Total Operating Cost for Slabs
59.28 $/NTP
(1.00)$/NTP
6.00
64.28
Hot-Rolled Sheets

Slabs (1.11 tons/NTP at 64.28 $/NTS)
Scrap (0.087 tons/NPT)
Other Costs
Total Operating Cost for Hot-Rolled Sheets
71.35 NTP
(2.15)
11.09
80.29 $/NTP
Cold-Rolled Coils (Annealed and Tempered)

Slabs (1.142 tons/NTP at 64.28 $/NTS)
Scrap (0.035 tons/NTP)
Other Costs
Total Operating Cost for Hot Rolled Sheets
73.41 $/NTP
(3.40)
20.63
90.64	$/NTP

-------
Table 3-24. Estimated Economics of OH Integrated Steel ProductionCase C
(Pollution Control Cost Excluded) (Continued)
Plant Capacity
2MM Tons/Year^
Cold-Rolled Sheets
Slabs (1.202 tons/NTP at 64.28 $/NTS)	77.26 $/NTP
Scrap Credit (0.06 tons/NTP)	(4.91)
Other Costs	25.67
Total Operating Cost for Cold Rolled Sheets	98.74 $/NTP
Total Weighted Average Operating Cost(e) ($/ton)	89.69 $/NTP
Capital Charges Per Net Ton Final Product	35.01
(at 0.837 tons product/ton molten steel)
.(e)
Total Average Manufacturing Costv '	124.70
Average Product Revenue^	169.91
General and Sales Expenses ($/NTP)	2.49
F.o.b. Cost ($/NTP)	127.19
Profit After Taxes (Tax at 505S)	21.36 $/NTP
Cash Flow	72.9 $MM/year
R0I(f) 6.0%
^Integrated plant - coke ovens, sintering, blast furnace, steelmaking furnaces (OH), casting machines
and mills
^Based on molten steel production
(c)
'Assume: 1.44 tons low S coal/ton coke, $4.73/ton credit and $6.00/ton operating costs
^Conservative estimate
/e)
'Based on 28% hot rolled sheets, 48% annealed and tempered cold rolled coils and 24% cold rolled sheets
^Return on investors equity (taken at 80% of installed capital)
NTI = Net tons iron produced
NTS = Net tons steel produced

-------
Case C covers economic analyses of an integrated iron and steel
division, producing 2 million tons per year of steel from open hearth
furnaces. Operating facilities in addition to the open hearth furnaces
at the division include coke ovens, an ore-sintering plant and blast
furnaces for hot metal (pig iron) production. Return on investment, without
addition of fluoride control processes, is estimated at 6.0%. After
pollution control equipment currently employed by the industry is added,
return on investment drops to 5.8%.
In actual practice, a single iron and steel facility usually contains
a coke plant, a sintering plant, blast furnaces and a number of open hearth,
electric arc, and basic oxygen steel making furnaces. Return on investment
for a facility of this type, before addition of fluoride control process
costs, is estimated at 7.7%. After addition of pollution control equipment
currently employed by the industry, the facility is estimated to return 7.5%
on investment.
The relative reduction on return of investment for the four cases
discussed above is presented in Figure 3-27. Plant and division locations,
for modeling purposes, are assumed to be in the Great Lakes area of Ohio.
The general process economic assumptions contained in Section 3.1.1 were
used in developing the estimates for the iron and steel industry.
The industry, or for that matter, an individual company's return
on investment is difficult to establish clearly. Net income (after taxes)
for any one firm may vary substantially from year to year as the result of
a number of related factors. Heavy "start-up" costs for major investments,
close economic ties to a single industry (e.g., automobiles), imported
steel products, stockpiling of steel in strike years, production rates,
accounting practices (inventory accounting systems or change from accelerated
to straight-line depreciation schedules) and obsolete plant equipment
carried on the "books" may all substantially change the return on invest-
ment calculation.
3-110

-------
CASE
A = ELECTRIC
B= BOF
C = OH	;	
D = INTEGRATED
Figure 3-27. Relative Reduction on Return on
Investment ( ROI) for Iron and
Steel Industry
3,3.6.2 Impact of Controls
The individual process element pollution control cost analyses sheets
for the systems currently employed by the industry, which are presented
below, may be summarized as follows for the three specimen cases:
Case A: The annual cost of systems currently employed for control
of emissions from the 500,000 ton per year electric arc
furnace plant is estimated at $280,000 per year, equivalent
to $0.64 per ton of steel product.
Case B: The annual cost of systems currently employed for control
of emissions from the integrated iron and steel "division"
using purchased pelletized ore and producing steel in
basic oxygen furnaces is estimated at $885,000 per year,
equivalent to $0.56 per ton of steel.
Case C: The annual cost of systems currently employed for control
of emissions from the integrated iron and steel "division"
sintering its ore and producing steel in open hearth
furnaces is estimated as $1,390,000 per year, equivalent
to $0.83 per ton.of steel.
Most of the systems currently employed for control of emissions were
installed solely for control of particulate discharges, and control soluble
3-111

-------
fluoride emissions incidentally, if at all. The dry collection devices
remove none of the gaseous fluorides evolved. The costs of emission control
given above involve control of soluble fluorides only to the extents noted
in the controlled process diagrams; no soluble fluorides are evolved or
emitted by the basic oxygen and electric arc furnace processes.
The iron and steel industry is generally assumed to be subject to
oligopolistic competition; that is, prices are usually set by the largest
firms with smaller firms following the pattern thus established. For
this type of competition, prices tend to hold without much change for
considerable periods even in the face of fluctuating demand and increasing
cost pressures, such as the current labor contract settlement. This suggests
that different firms will be required to absorb a greater share of the cost
of fluoride pollution control than others and that the cost differential
will not be reflected in price competition.
Because of the extensive anti-inflationary pressure on the iron and
steel industry, the basic unit sale price must be assumed to be constant.
Thus, the total burden of the control costs will probably rest on the
stockholders of the firms. This would mean that the full reduction in
return on investment would be felt and the firms' cash generating capability
decreased.
Iron Ore Sintering. Economic analyses of the control processes for
normal fluoride content ore are presented in Tables 3-25 through 3-27. Costs
range from $0.13 to $0.21 per ton for normal fluoride content ore; costs for
high fluoride content ore are $0.93 per ton of sinter.
Blast Furnace.
Control Economics. Table 3-28 covers analysis of the costs of
current fluoride control processes for normal and high fluoride content
iron ore. Estimated cost for "normal" ore is $0.32 per ton of pig "iron.
3-112

-------
Table 3-25. Iron Ore Sintering - Estimated Economics of Control Process A
Basis - 2000 Tons/Day of Sinter Produced
CO
I
Item
Number
_£a£i_taj_Cost_Estimates_^|1000]i
Description
Cyclone, 460,000 cfm, 4.9 in. Wg
pressure drop, low alloy steel,
26.4 lb/min loading, 350°F, 475 hp
Electrostatic Precipitator,
375,000 cfm, 0.9 in. WG pressure
drop, low allow
tquipment
F.O.B
Cost
186
286
Reference
Number
4387
(a)
Installation
Factor
2.82
1.69
Equipment
Installation
Cost	
(a) See footnote 20.
All control economics footnotes are located in Section 3.1.1 pages 3-TQ and 3-11
524
483
Capital Subtotal	1,007
Indirects (0 15%)	151
Contingency (@ 20%) 201
Total Capital (as of January 1971) 1,359
)£ej^ijiq_Cost_J$J/_hri___L
Item
Number
Power
Cost
2.48
1.58
Maintenance
Cost
0.13
0.38
Subtotal
Water
(21)
Disposal
(22)
Total Operating Cost
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)
Pollution Control Cost (S/hr)
Pollution Control Cost ($/ton)
tquipment
Operating
Cost
2.61
1.96
4.57
4.57
4.57
3.43
9.60
17.60
0.21

-------
Table 3-26. Iron Ore Sintering - Estimated Economics of Control Process B
Basis - 2000 Tons/Day of Sinter Produced
Capital Cost Esti
mates_£$10002
Item
Number
Description
Cyclone, 225,000 cfm, 4.9 in. WG
pressure drop, low alloy steel,
232 hp.
Electrostatic Precipitator
210,000 cfm, 0.9 in. WG pressure
droD
Equipment
F.U.I
Cost
103
193
Reference
Number
4387
4390
4392
(a)
Installation
Factor
2.82
1.69
Equipment
Installation
Cost	
(a) See footnote 20.
290
326
Capital Subtotal	616
Indirects (9 15%)	92
Contingency (@ 20%)	123
Total Capital (as of January 1971)	831
Operating Cost (S /^r )
Item
Number
Power
Cost
1.21
0.88
Subtotal
(21]
Water
Disposal
(22)
Maintenance
Cost
0.13
0.38
Total Operating Cost
2.60
2.60
All control economics footnotes are located
in Section 3.1.1 pages 3-10 and 3-11
Total Operating Cost ($/hr)	2.60
Taxes and Insurance (2%, 330 days)	2.10
Capital (7.1%, 330 working days)	5.88
Pollution Control Cost ($/hr)	10.58
Pollution Control Cost ($/ton)	0.13

-------
Table 3-27. Iron Ore Sintering - Estimated Economics of Control Process C
Basis - 2000 Tons/Day of Sinter Produced
Capital Cost Estimates ($1000)
Item
Number
Description
tquipment
F.U.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
Cyclone, 225,000 cfm, 4.9 in. WG
pressure drop, low alloy steel,
232 hp.
77
4387
4390
4392
2.66

205
2
Spray Scrubber, 2 at 14 ft by 30 ft,
8 ft/sec allowable velocity,
148,000 cfm, 80 moles/(hr) (ft3)
(atm) mass transfer coefficient,
2 in. WG pressure drop, neoprene
1ined steel , 62 hp.
138
4387
4388
4390
4391
2.58

356
3
Thickener, 5000 gal per min, 31 lb/
min loading, 100,000 gal capacity,
neoprene lined steel
42
4383
4392
2.36

99
4
Vacuum Filter, 100 ft2 area, 31 lb/
min, 5000 gal/min, 38 hp required.
78
4383
4392
2.70

168




Capital Subtotal
828




Indi rects
(e 15%)
124




Contingency (@ 20%) irk


Total Capital (as of January 1971^
1 ,118
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
Operating Cost ($ /hr )
Item
Number
Power
Cost
1.21
0.32
0.20
Mai ntenance
Cost
0.13
0.25
0.06
0.33
Subtotal
Equi pment
Operating
Cost
121)
Water
Disposal
(22)
Total Operating Cost
1.44
0.74
0.06
0.53
(1700 gpm, 90% recycle )
2.77
0.57
3.34
Total Operating Cost ($/hr)	3.34
Taxes and Insurance (2%, 330 days)	2.82
Capital (7.1%, 330 working days)	7.91
Pollution Control Cost ($/hr)	14.07
Pollution Control Cost ($/ton)	0.17

-------
Table 3-28. Pig Iron Production - Estimated Economics of Control Process
Basis - 1000 Tons Pig Iron Per Day
Capital Cost Estimates ($1000)
Item
Number
Description
tquipment
F.U.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
Cyclone, 140,000 cfm, 4.9 in. WG
pressure drop, 145 hp required,
low alloy steel.
49
4387
4390
2.80

137
2
Wet Scrubbers, 2 at 14 ft by 24 ft,
T.6 in. WG pressure drop, 3040 gpm
liquid rate, 8 ft/sec allowable
velocity, 80 moles/(hr) (ft^) (atm)
mass transfer coefficient,
neoprene lined steel, 94 hp.
95
4387
4388
4390
4391
2.53

240
3
Thickener, 3100 gpm, 12.5 lb/min
loading, 62,000 gal capacity,
neoprene lined steel.
28
4383
4392
2.36

66
4
Vacuum Filter, 60 ft2, 3100 gpm,
12.5 lb/min loading, 63 hp,
neoprene lined steel.
52
4383
4392
2.15

' 112
5
Electrostatic Precipitator,
105,000 cfm, 0.9 in. WG pressure
drop.
116
(a)
1.69

196




Capital Subtotal
751




Indirects (@ 15%)
113




Contingency (@ 20%) 150


Total Capital (as of January 1971) 1,014
(a) See footnote 20.
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
Qperati nq
Cost ($ /hr )
Item
Number
Power
Cost
Maintenance
Cost

tqui pment
Operating
Cost
1
0.76
0.13

0.89
2
0.49
0.50

0.99
3
-
0.06

0.06
4
0.33
0.14

0.47
5
0.57
0.16

0.73
Subtotal
Water<21> (
Disposal ^22'
3.14
3100 gpm, 90% recycle) 0.37
Total Operating Cost
3.51
Total Operating Cost ($/hr)	3-51
Taxes and Insurance (2%, 330 days)	2.56
Capital (7.1%, 330 working days)	7.17
Pollution Control Cost ($/hr)	13.24
Pollution Control Cost ($/ton)	0.32

-------
Open Hearth Furnace
Control Economics. Tables 3-29 through 3-32 present analyses of
costs for the processes currently employed for control of emissions from
open hearth operations. Implementation of new control systems may be hard
to justify, due to the proposed phase-out of the open hearth.
Basic Oxygen Furnace
Control Economics. Tables 3-33 and 3-34 present current emission
control process costs for BOF steel working.
Electric Arc Furnace
Control Economics. Tables 3-35, 3-36, and 3-37 present the cost of
control of emissions from electric arc furnaces.
3-117

-------
Table 3-29. Basic Open Hearth Steelmaking^2^ - Estimated Economics of Control Process A
Basis - Furnace Capacity of 200 Tons/Heat (2 Heats Per Day @ Approx. 10 Hours Per Heat)
Item
Number
Capital Cost Estimates ($1000)
Description
Waste Heat Boiler, 2000 ft surface
area, assumed Atm = 241° and UAt =
4,800 Btu/hr/ftS 33,000 cfm, t. =
1300°F, tout = 500°F, 20 in. WGln
pressure drop, low alloy steel,
340 hp.
Electrostatic Precipitator, 20,000
cfm, 0.9 in. WG pressure drop.
tqulpment
F.O.B.
Cost
33
56
Reference
Number
4392
Installation
Factor
3.09
1.69
Equipment
Installation
Cost	
(a) See footnote 20.
All control economics footnotes are located in Section 3-1.1 pages 3-10 and 3-11
102
95
Capital Subtotal	197
Indirects (9 15%)	30
Contingency (@ 20%)	39_
Total Capital (as of January 1971)	266
Operating
Item
Number
Cost ($ /heat )
Power
Cost
10.26
0.84
Maintenance
Cost
1 .32
4.56
Subtotal
(2i:
Water
DiSDOsal
(22)
(1 ton/heat of
Fe2°3)
Total Operating Cost
Equipment
Operating
Cost
11.58
5.40
16.98
0.50
17.48
Total Operating Cost ($/heatj	17.48
Taxes and Insurance (2%, 330 days) 8.06
Caoital (7.1%, 330 working days)	22.57
Pollution Control Cost ($/heat)	48.31
Pollution Control Cost ($/ton)	0.24

-------
Table 3-30. Basic Open Hearth Steel making^24^ - Estimated Economics of Control Process B
Basis - Furnace Capacity of 200 Tons/Heat (2 Heats Per Day @ Approx. 10 Hours Per Heat)
_Ca£HaJ_Cost_Estimates_£$1000]i
Item
Number
Description
Equipment
F.U.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
Waste Heat Boiler, 2000 ft2 surface
area, Atm = 241° and UAtm = 4800 Btuy
hr/ft2, 33,000 cfm tin = 1300°F,
t0ut = 500°F,1 ow alloy steel, 20 in.
WG pressure drop, 340 hp.
33
4383
4392
3.09

102
2
Secondary Exchanger, 1650 ft2
surface area, V = 25 Btu/hr/ft2-°F,
UAtm = 1944 Btu/hr/ft2, 20,000 cfm,
tin = 500°F, tout = 220°F,
Steel, 20 in. WG pressure drop, 84
hp.
27
4383
4392
3.07

83
3
Baghouse, 13000 cfm, 3.34 lb/min
loading, 2.5 in. WG pressure drop,
fabric felter-shaker, 7 hp.
13
4387
4383
4.13

54




Capital Subtotal
239




Indirects (9 15%)
36




Contingency (@ 20%).


Total Capital (as of January 1971) 333
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
Operating Cost (S / heat
Item
Number
Power
Cost
10.26
4.40
0.36
Maintenance
Cost
1.32
1.20
15.12
Subtotal
121)
Water
Disposal
tquipment
Operating
Cost
11.58
5.60
15.48
(22)
(1 ton Fe20g/heat)
Total Operating Cost
Total Operating Cost ($/heat)
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)
Pollution Control Cost ($/heat)
Pollution Control Cost ($/ton)
32.66
0.50
33.16
33.16
9.79
27.41
70.36
0.35

-------
Table 3-31. Basic Open Hearth Steelmaking^24^ - Estimated Economics of Control Process C
Basic - Furnace Capacity of 200 Tons/He^t (2 Heats Per Day @ Approx. 10 Hours Per Heat)
Item
Number
Description
hquipment
F.U.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
Waste Heat Boiler, 2000 ft^ surface
area, Atm = 241° and UAtm = 4800
Btu/hr/ft2, 33,000 cfm, tjn =
1300°F, tout = 500°F, 20 in. WG
pressure drop, 340 hp, low alloy
steel.
33
4383
4392
3.09

102
2
Secondary Exchanger, 1650 ft ~
surface area, V = 25 Btu/hr/ft -°F,
20,000 cfm, tin = 500°F, tout =
220°F, 20 in. Wg pressure drop,
84 hp, carbon steel.
27
4383
4392
3.07

83
3
Venturi Scrubber, 130,000 cfm,
monel clad, 3.34 lb/min loading,
110 gpm, neoprene lined steel, 31.5
in. WG pressure drop, 86 hp.
32
4383
4390
4391
1.75

56
4
Separator, 1000 gal capacity,
neoprene lined steel.
6

3.5

21
5
6
Thickener, 10,000 gal capacity,
3.17 lb/min loading.
Vacuum Filter, 3.2 lb/min loading,
_neoorene lined steel. lOOfl?.. 38hn
17
78
4383
4383
2.35
2.17

40
169
Capital Subtotal	471
Indirects (9 15%)	71
Contingency (@ 20%)	94
Total Capital (as of January 1971)	636
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
Operating Cost (S / heat )
Item
Number
Power
Cost
Maintenance
Cost
10.26
1.32
4.40
1.2
4.50
1.56
0.12
0.60
1.98
0.76
Subtotal
Equipment
Operating
Cost
11.58
5.60
6.06
0.72
2.74
(21)
Water
Disposal
(22)
26.70
(110 gpm, 90% recycle) 0.13
(1 ton Fe20.j/heat ) 0.50
Total Operating Cost
27.33
Total Operating Cost ($/heat)
Taxes and Insurance (2%, 330 days)
Cam'tal (7.1%, 330 working days)
Pollution Control Cost ($/heat)
Pollution Control Cost ($/ton)
27.33
100!56
0.50

-------
Table 3-32. Basic Open Hearth Steelmaking^24^ - Estimated Economics of Control Process D
Basis - Furnace Capacity of 200 Tons/Heat (2 Heats Per Day @ Approx. 10 Hours Per Heat)
Item
Number
Description
tquipment
F.O.B.
Cost
Reference
Number
Installation
Factor

tquipment
Installation
Cost
1
Waste Heat Boiler, 2000 ft^ surface
area ,Atm = 241°, mtm = 4800 Btu/hr/
ftz, 33,000 cfm, tin = 1 300°F, tout
= 500°F, 20 in. WG pressure drop,
340 hp, low alloy steel.
33
4383
4392
3.09

102
2
Cyclone, 22,000 cfm, 4.9 in. WG
pressure drop, low alloy steel.
15
4390
4387
2.8

42
3
Electrostatic Precipitator, 21,000
cfm, 0.9 in. WG pressure drop.
56
(a)
1.69

95




Capital Subtotal
239




Indirects (P 15%)
36




Contingency (@ 20%) ' 48


Total Capital (as of January 1971
323
(a) See footnote 20.
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
Qperatim
Item
Number
Cost (S /heat )
Power
Cost
10.26
1.19
0.88
Maintenance
Cost
1.32
1.52
4.55
Subtotal
Water
T2i;
,,(22)
( 15 gpm
Equipment
Operating
Cost
11.58
2.71
5.43
)
Disposal (3 tons CafOHK,
,ocfe203 and CaF?/heat)
Slaked Lime'") ((j.21 tons Ca(0H)2/
heat)
Total Operating Cost
19.72
0.17
1.50
1.30
22.69
Total Operating Cost ($/heat)
Taxes and Insurance (2%, 330 days)
Caoital (7.1%, 330 working days)
Pollution Control Cost (S/heat)
Pollution Control Cost ($/ton)
22.69
9.79
77 41
59.89
0.30

-------
Table 3-33- Basic Oxygen Steelmaking^^ - Estimated Economics of Control Process A
Basis - Furnace Capacity of 200 Tons/Heat (12 Heats Per Day)
Item
Number
Description
tqui pment
F.O.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
Spray Chambers, 5 at 14 ft dia by
26 ft, 358,000 cfm, 8 ft/sec
allowable velocity, 2 in. WG
pressure drop, 151 hp, mass transfer
coefficient of 80 moles/(hr) (ft3)
(atm), carbon steel
125
4386
4388
4390
4391
3.73

466
2
Thickener, 7800 gpm, 34 lb/min.
160,000 gal capacity, carbon steel.
28
4383
3.57

100
3
Vacuum Filter, 7800 gal/min,
160 ft^, 34 lb/min loading, carbon
steel,61 hp.
59
4383
3.22

190
4
Electrostatic Precipitator, 270,000
cfm, 0.9 in. WG pressure drop.
220
(a)
1.69

372




Capital Subtotal
1 ,128




Indirect;
(e 15%)
159




Contingency (@ 20%) 226


Total Capital (as of January 1971
1 ,523
(a) See footnote 20.
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
Operating
Item
Number
Cost ($ / heat )
Power
Cost
1.58
0.64
2.27
Maintenance
Cost
2.50
0.12
0.66
0.76
Equi pment
Operating
Cost
Subtotal
121)
Water
Disposal
(22)
Total Operating Cost
4.08
0.12
1 .30
3.03
7770 gpm, 90% recycle)
8.53
0.94
9.47
Total Operating Cost ($/heat)
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)
Pollution Control Cost ($/heat)
Pollution Control Cost ($/ton)
9.47
7.69
11
40.27
0.20

-------
(19)
Table 3-34. Basic Oxygen Steelmakingv ' - Estimated Economics of Control Process B
Basis - Furnace Capacity of 200 Tons/Heat (12 Heats Per Day)
CO
J
ro
CO
Item
Number
Description
Equipment
F.U.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
Quench Towers, 4 at 14 ft dia by
20 ft, 358,000 cfm, 63 lb/min
loading, 917 ft/sec allowable
velocity, 2.0 sec residence time,
170 gal per min, carbon steel.
86
(a)
4.37

376
2
Venturi Scrubber, 1500 gpm,
240,000 cfm, 55 lb/min loading,
carbon steel, 31.5 in. WG pressure
drop, 1600 hp.
128
4383
4390
4391
2.40

307
3
Cyclonic Clarifier, 1500 gpm,
55 lb/min loading, carbon steel.
5
4383
3.60

18
4
y
Vacuum Filter, 1500 gpm, 35 ft ,
carbon steel, 35 hp.
18
4383
3.22

58




Capital Subtotal
759




Indirects (® 15%)
114




Contingency (@ 20%) 152


Total Capital (as of January 1971
1,025
(a) See footnote 23.
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
Item
Number
Power
Cost
Maintenance
Cost

Equi pment
Operating
Cost
1
0.40
1.60

2.00
2
16.63
0.25

16.88
3
-
0.25

0.25
4
0.36
0.66

1.02
Subtotal 20.15
Water(21^ ( 1500 gpm, 90% recycled 0.36
Disposal^^( 2 tonsper heat,
Fe20j, CaF2) 1.00
Total Operating Cost
21.51
Total Operating Cost ($./heat)
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)
Pollution Control Cost ($/heat)
Pollution Control Cost ($/ton)
21.51
5.18
41.18
0.21

-------
Table 3-35. Electric-Arc Steelmaking^26^ - Estimated Economics of Control Process A
Basis - Furnace Capacity of 75 Tons/Heat (4 Heats Per Day)
Item
Number
CO
I
PO
-p>
Capital Cost Estimates ($1000)
Description
Water-Cooled Duct, 12,000 ft ,
tjn = 1000°F, tout = 250°F, carbon
steel,5 in. WG pressure drop,
151,000 cfm, 160 hp.
Baghouse, fabric filter-shaker,
73,000 cfm, 250°F, 2.5 in. WG
pressure drop, 38 hp.
tquipment
F.O.B.
Cost
78
44
Reference
Number
4383
4387
4383
Installation
Factor
1.80
4.13
Equi pment
Installation
Cost	
140
182
Capital Subtotal	322
Indirects (@ 15%)	48
Contingency (@ 20%) _64_
Total Capital (as of January 1971)	434
All control economics footnotes are located in Section3.1.1 pages 3-10 and 3-11
derating Cost ($ /hpat )
Item
Number
Power
Cost
4.98
1.19
Maintenance
Cost
0.92
11.37
Subtotal
121)
Water
Disposal
(22)
tqui pment
Operating
Cost
5.90
12.56
( 17 gal/min
( 0.62 tons CaF2 and
Fe203/heat)
18.46
) 0.12
0.31
Total Operating Cost
18.89
Total Operating Cost ($/heat)
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)
Pollution Control Cost ($/heat)
Pollution Control Cost ($/ton)
18.89
6.58
18-40
43.87
0.58

-------
Table 3-3-36. Electric-Arc Steelmaking^26^ - Estimated Economics of Control Process B
Basis - Furnace Capacity of 75 Tons/Heat (4 Heats Per Day)
Item
Number
Description
tquipment
F.O.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
Water-Cooled Duct, 12,000 ft^ t-jn =
1000°F, tout = 250°F, carbon steel,
5 in. WG pressure drop, 151,000 '
cfm, 160 hp.
78
4383
1.80-

140
2
Venturi Scrubber, 73,000 cfm, 250°F
31.5 in. WG pressure drop, carbon
steel, 485 hp.
53
4383
4390
4391
2.40

127
3
Cyclonic Separator, 620 gpm, 3.44
lb/min loading, carbon steel.
3
4383
4.00

12
4
Thickener, 620 gpm, 3.44 lb/min
loading, carbon steel, 12,000 gal
capacity.
10
4392
3.50

35
5
Vacuum Filter, 30 ft^, 3.44 lb/min
loading, carbon steel, 12 hp.
16
4383
4392
3.19

51




Capital Subtotal
365




Indirects
(e 15%)
55




Contingency (@ 20%) 73


Total Capital (as of January 1971'
493
All control economics footnotes are located in Section 3.1.1 pages 3-10 and 3-11
0£ej^j_nqCost_J$d/_he^^^
Item
Number
Power
Cost
4.98
15.17
0.38
Maintenance
Cost
0.92
0.78
0.78
0.38
1.97
tqui pment
Operating
Cost
5.90
15.95
0.78
0.38
2.35
Subtotal
T21)
Water
Disposal
(22)
( 670 gpm, 90% recycle)
( 0.6 tons CaF, and )
Fe20j/heat)
Total Operating Cost
25.36
0.48
0.30
26.14
Total Operating Cost ($/heat)	26.14
Taxes and Insurance (2%, 330 days)	4.97
Capital (7.1%, 330 working days)	1 ^
Pollution Control Cost ($/heat)	45.05
Pollution Control Cost ($/ton)	0.60

-------
Table 3-37.
t2c\
Electric-Arc Steelmaking - Estimated Economics of Control Process C
Basis - Furnace Capacity of 75 Tons/Heat (4 Heats Per Day)
00
¦
PO
Item
Number
Description
tquipment
F.U.B.
Cost
Reference
Number
Installation
Factor

tquipment
Installation
Cost
1
Spark Box, 151,000 cfm, 1000°F,
assume 150,000 volts and 0.1 amp,
80 ft/sec, 16' x 16' x 12', carbon
steel.
15
4383
3.0

45
2
Water-Cooled Duct, 8100 ft2,
tin = 1000°F, t0ut = 500°F, carbon
steel, 150,000 cfm, 3 in. WG pressure
drop, 95 hp.
53
4383
1.80

95
3
Electrostatic Precipitator,
0.9 in. WG pressure drop,
99,000 cfm,carbon steel
99
(a)
1.69

167




Capital Subtotal
307




Indirects
(0 15%)
46




Contingency (@ 20%). 61


Total Capital (as of January 1971) 414

Item
Number
Power
Cost
0.63
2.97
1.66
Subtotal
Maintenance
Cost
0.78
0.61
2.27
Equipment
Operating
Cost
CaF2/ton)
Total Operating Cost
1.41
3.58
3.93
Water^21' (20 gpm	)
Disposal ^22^( 0.6 tons Fe203 and
8.92
0.10
0.30
9.32
(a) See footnote 20.
All control economics footnotes are located
in Section 3.1.1, pages 3-10 and 3-11
Total Operating Cost ($/heat)	9.32
Taxes and Insurance (2%, 330 days)
Capital (7.1%, 330 working days)	17-56
Pollution Control Cost (fyheat)	33.15
Pollution Control Cost ($/tbn)	0.44

-------
3.4 COAL COMBUSTION - ELECTRICAL POWER GENERATION

-------
3.4 COAL COMBUSTION - ELECTRICAL POWER GENERATION
3.4.1	General
Coal is known to contain low concentrations of combined fluorine,
generally present as calcium fluoride. The concentrations vary widely with
an average of 0.01 weight percent for samples from diverse sources.(^302)
Because of the enormous quantity of coal consumed annually (421 million tons
in 1970), the quantity of fluoride potentially available for emission to the
atmosphere is very large. Most of the coal sold in the country is burned;
of the total, the electrical utility industry burns about 60%. Unfortunately,
both number and representative character of the experimental determinations
of fluoride content of gaseous and solid coal combustion products are
extremely limited. To circumvent the paucity of data, the thermochemical
approach discussed in Section 3.1.2 was used, in conjunction with the limited
information available, to estimate the quantities and types of fluorides
evolved. Coal burning for power, based on this approach, emits large quantities
of gaseous hydrogen fluoride (HF) in l.ow concentrations, in the power plant
stack gas.
For purposes of simplicity, the major industry involving coal burning,
electric power generation by public utilities, has been selected for
analysis. The other coal-burning processes are virtually identical in
effluent characteristics; they have not been included in overall emission
calculations, because of uncertainties in consumption data.
3.4.2	Industry Description
Steam-electric utility plants in the United States consumed coal oil
and gas equivalent to 518 million tons of coal during 1969 for the generation
of electric power. Of this total, the utilities burned bituminous coal,
lignite, and anthracite equivalent to more than 320 million tons total.(4283)
Bituminous coal use was over 98% of the total.
3-127

-------
3.4.3 Production Trends
Since the turn of the century, the demand for coal in the U.S. has
fluctuated between 300 and 620 million tons. The low point occurred in
the early years of the depression, and the peak was in World War II, with
an all-time high of 620 million. Over the past 65 years, sales to large
traditional customers - the railroads and space heating - have either
disappeared or have become greatly reduced. Electric utilities are now
coal's largest customers. In 1965, they accounted for 243 million tons,
or 53% of total coal demand; in 1955, by contrast, they took 33%. In the
early 1960's, this rapid increase in electric utility consumption reversed
the industry's 15-year downward trend from the postwar peaks.
Total demand for U.S. coal will amount to between 1060 million tons
and 1300 million tons by the year 2000.(4281,4282) Annua-| growth will
range between 2.4% and 4.0%, as opposed to a rate of about 5% for the past
10years.<428,>4282>
Electric utilities will consume between 780 million and 1080 million
tons in the year 2000, accounting for most of the domestic use.(4281,4282,4283)
Because of competition from nuclear power, coal's rate of growth in this
market will decrease to between 3.4% and 4.0%, compared to 5.6% over the
last 10 years. Table 3-38 presents a summary of data found in literaure
and obtained from private sources.
Since the first nuclear plant went into operation in 1957 at
Shippingport, Pennsylvania, nuclear power has risen rapidly. In some cases,
the choice of nuclear power was made strictly on the basis of cost, but
increasingly stringent regulations on air, water, and land pollution have
influenced a decision not to use coal. Under these circumstances and
because of favorable costs, electric utilities have moved to nuclear power.
An example is the recent selection of nuclear over coal in one of the
latter's traditional strongholds, the TVA region, where coal costs to the
system are the lowest of any utility in the nation. Such competition,
spurred also by air pollution factors, will increase; and, as a result,
coal's share of the electric generation market will drop from its current
3-128

-------
Table 3-38. Coal Usage

1965
(millions
of tons)
1970
(mil 1 ions
of tons)
1980
(mil lions
of tons)
Expected
Growth Rate
(Percent)
Extrapolated to
Year 2000
millions of tons)
Total coal
usage
459<4281>
421(4281)
660(4281>
2.4
1060





4.0
1300
Electrical
utility


400(4281)
3 4(4281)
780
coal usage

32o(42S2)

3.9
iooo(4282>


333*(4283)

3 0(4285)
1080*
*Values used in this study.

-------
level of 54 to 37% in 198o/4284^ However, because of the overall growth
in power generation, the volume of coal used by utilities will increase
substantially, as noted above.
3.4.4 Control Techniques
Currently, with only four known exceptions in this country, coal
burning power plant pollution control equipment is limited to fly ash
abatement. The exceptions are facilities using alkaline injection and/or
wet scrubbing systems, including the limestone injection-wet scrubbing
processes currently under investigation by the Office of Air Programs for
abatement of SO2 generated by coal burning steam-electric plants. These
processes should also remove substantial portions of the HF content of the
flue gases. In theory, due to the high solubility of HF in water and the
extreme insolubility of CaF2» HF should be absorbed very rapidly by the
wet scrubber solutions. Practically, the actual effectiveness of fluoride
emission control via the limestone injection-wet scrubbing processes will
require experimental verification because of the relatively low concentrations
of hydrogen fluoride in the flue gases. These concentrations are of the
order of 20 to 100 parts per million (volume).
Coal contains minor quantitites of inorganic fluorides, with a
nominal concentration of approximately 0.01 weight percent fluorine.(4302)
(This value is considerably lower than the 400 ppm value reported as maximum
from another source.){4285) ^ ^ temperatures associated with combustion,
near complete volatilization and conversion to HF of the fluoride content is
probable; however, a portion of the evolved fluoride may adsorb or react
with fly ash or other solid surfaces in cooler parts of the process. A
recent study by Orning, et al/4231 ^ indicates the following distribution
of fluoride from coal combustion:
Retention in ash overhead
Exit as gas
Unaccounted
10%
70%
20%
100%
3-130

-------
TRW-RRI data files indicate that about 10% of the fluoride contained
in the coal fed to the burner is removed with the fly ash in the precipitator
while 80% is emitted to the atmosphere in the gas phase as HF; the remaining
10% is emitted as suspensoid particulate (micron size uncollected fly ash).
Thermochemical analyses of the coal/air system were performed using a
proprietary chemical analysis program assuming 20% excess air in the
combustion process. Calculations were performed at one atmosphere total
pressure and at temperatures ranging from the adiabatic flame temperature
(approximately 4000°F under these conditions) to below 1300°F, the probable
lower limit for gaseous phase attainment of species equilibrium at finite
rates. The results of these calculations are presented in Figure 3-28,
which indicates the percent conversion to HF (gaseous) at equilibrium of the
*
fluoride species contained in coal as a function of temperature.
At high temperatures (above 2000°F), HF (g) is almost the sole
equilibrium species containing fluorine and the time required for attainment
of equilibrium is virtually microseconds. As temperatures decrease, the
rate of attainment of equilibrium via gas phase reaction decreases exponen-
tially. Thus, even though equilibrium conditions below 1480°F shift to
favor increased formation of CaF£, the rate of gas phase reaction plus the
adsorption/chemisorption process rates is so low that overall reversion to
CaF£ of the HF formed at higher temperatures does not take place to an
extent greater than 20% in the time available in the power plant. It should
be noted that a probable major mechanism for reversion to CaF£ at the lower
temperatures (below 1300°F) is the absorption/chemisorption process that
involves reaction between the suspensoid solid CaO of the fly ash and
gaseous HF.
A soluble fluoride emission factor of 0.16 lb F/ton coal burned was
used for the emission determinations. The current and future estimations
of soluble fluoride (reflecting no soluble fluoride abatement) emissions
'ic
The percent conversion is almost independent of CaF2 content for
bituminous coals, in the range of 0.005 to 0.100% (wt)-
3 -131

-------
1300 1400 1500 1600 1700 1800 1900 2000
TEMPERATURE, °F
Figure 3-28. Percent Conversion of Fluorine Species in Coal
to HF at Equilibrium
3-132

-------
are 26,600 tons (as F) in 1970 and 86,400 tons (as F) in 2000. If 99%
fluoride control is applied, the emission level will drop to 860 tons (as F)
in 2000v
3.4.5 Process Description
Host coal burning power plants air inject pulverized coal (ground to
90% through 200 mesh) through special burners into a combustion chamber. As
indicated by the process-flow diagram in Figure 3-29, the combustion chamber
has a formed draft feed "of secondary air. All of the burning is completed in
the combustion chamber at flame temperatures over 4000°F for a 201 excess of
air. Reaction between the metallic fluorides contained in the coal and the
hydrogen from water or other sources is almost quantitative and instantaneous
at this .temperature. Hydrogen fluoride evolution takes place almost
completely in the combustion chamber.
From the combustion chamber, the hot gas and suspensoid fly ash flow
to the boiler, where heat exchange takes place in the water jacket boiler
tubes, generating steam and dropping the gas temperature to levels where
adsorption or reaction of the evolved HF by the fly ash starts. After
further temperature reduction in the superheaters, reheaters and air
preheaters, the combustion gas stream ("flue gas") is stripped of the major
portion of the suspensoid fly ash in the dust collectors and is discharged
through the stack to the atmosphere.
All currently operating coal burning electric utility plants use
dust collection devices for removal of fly ash. With only four known
exceptions, the dust collection devices are dry systems, which remove only
particulate material from the flue gas. Gaseous pollutants, other than
the small amounts removed by adsorption or reaction with the particulate
material, pass through dry collection systems unhindered and uncontrolled.
The process flow diagram and mass balances presented in Figure 3-29
give data for a typical power plant burning coal of average fluoride content.
3-133

-------
STEAM
FLUE GAS


COAL

m n
¦	i i i
i • i i
¦	i i i
LJ LJ
SUPERHEATERS
AND
REHEATERS
BOILER
r-j
LJ
i—i
i	j

air-

s
FORCED
DRAFT FAN

•WATER
ECONOMIZER
JAIR preheater
DUST
COLLECTOR


FLY ASH
INDUCED
DRAFT
FAN
STACK
BASIS: 1100 MW POWER PLANT
PROCESS STREAMS - TONS/DAY

Stream Number
Material
1
2
3
4
5
6*
HF (Equiv.)
CaF2
2.85(s)(b)



0.29(s)(b)
1.17(g)(b)
0.29(s)(b)
Total Fluorides
Total as F
2.85
1.39



0.29
0.14
1.46
1.25
Coal
n2
°2
h2o
co2
so2
N0X
Fly Ash
13,900(s)(a)
97,000(1)(Est)
25,800(g)(Est)
800(g)(Est)
80.000(g)
(Est)
80.000(g)
(Est)
850(s)(Est)
97,000(g)'3'
2,700(g)'a)
6,000(g)'a)
29,000(g)'3'
8QQ(g)(a)
50(g)(a)
200(s)(Est)
Approximate
Total Stream
13,900
123,600
(c)
80,000
80,000
850
136,700
~Gaseous Effluent Stream
Soluble Fluoride Evolution Factor = 0.16 lbF/ton coal
Soluble Fluoride Emission Factor = 0.16 lbF/ton coal
(a)	Ref. 4384
(b)	Ref. 4266
(c)	Includes makeup water for boiler blow down
Figure 3-29. Coal-Fired Plant (1100 mw) - Uncontrolled Process Model
3-134

-------
3.4.6 Economic Analysis
3.4.6.1	Basic Process
The economics estimated for coal fired power plants are summarized in
Table 3-39 as a basis for future use in determining the economic impact of
control of soluble fluoride emissions.
3.4.6.2	Impact of Controls
The absence of any current, clearly defined techniques for controlling
fluoride emissions from coal burning power plants prevents attempting an
economic analysis of control costs.
3-135

-------
Table. 3-39. Estimated Economics of Coal Firing Power Plants
(Excluding Pollution Control Cost)
CAPACITY

100 mw
400 mw
700 mw
Installed Capital^ Cost
31.5 $MM
74.0 $MM
125.4 $MM
(2)
Operating Costs



Direct Costs



Coal (9000 Btu/kw hr at 22.3*/MM Btu FOB Mine)
2.01 mills/kwh
2.01 mills/kwh
2.01 mills/kwh
Coal Transportation (4.7(t/MM Btu)
0.42
0.42
0.42
(3)
Operation and Maintenance^
0.49
0.44
0.43
Total Direct Costs
2.92
2.87
2.86
Indirect Costs



Depreciation (3 1/3% per year)
1.71
1 .01
0.97
Interest Charges (7% of ave debt at debt/equity
0.90
0.53
0.15
of 0.5)


0.06
Insurance (0.2% of capital)
0.10
0.06
Local Taxes (1.2% of capital)
0.60
0.36
0.35
Total Indirect Costs
3.31
1.96
1.89
Total Cost (mills/kwh)
6.23 mills/kwh
4.83 mills/kwh
4.75 mills/kwh
Total Annual Cost ($MM/yr)
3.82 $MM/yr
11.85 $MM/yr
20.39 $MM/yr
Product Revenue ($MM/yr; based on a 7% return on
investment)
7.34
20.13
34.43


7.02
Profit After Taxes (at 50%; $MM/yr)
1.76
4.14
Cash Flow ($MM/yr)
2.81 $MM/yr
6.61 $MM/yr
11.20 $MM/yr
Return on Investment
7%
7%
7%
(1) Includes transmission capital cost from generator to distribution area
(2)	Based on a net plant factor of 70%
(3)	Primarily labor for plant and transmission
(4)	Equity plus debt estimated at 1.3 x capital less depreciation for return on investment calculation.

-------
PHOSPHATE ROCK PROCESSING

-------
3.5 PHOSPHATE ROCK PROCESSING
3.5.1	General
Phosphate rock, which is predominantly insoluble fluorapatite
[3Ca3(P04)2 • CaF2] and hydroxyapatite [3Ca3(P04)2 • Ca(0H)2], is the raw
material from which the water and citrate soluble phosphates and polyphos-
phates sold to the ultimate consumer are made. Each of the end products
requires thermal and/or chemical processing of the phosphate rock as an early
process step, with concomitant release and partial evolution of the fluoride
content of the rock as gaseous, soluble fluorldea-silicon tetrafluoride
(SiF^) and/or hydrogen -fluoride (HF). Many of the intermediate and final
process steps necessary for obtaining saleable commodities cause additional
evolution of SiF^ and/or HF.
3.5.2	Industry Description
The following subsections will cover only the fluoride evolving manu-
facturing processes for obtaining saleable commodities from phosphate rock.
These are the manufacture of wet process phosphoric acid; diammonium
phosphate; triple superphosphate; normal superphosphate; electrothermal
process phosphorus; and defluorinated phosphate rock.
One gauge of the magnitude of the problem handled by the phosphate
industries is the volume of rock produced and processed,, viewed in the con-
text of a fluoride content which ranges from 3 to 4 weight percent
(nominal = 3.5%). Tables 3-40, 3-41, and 3-42 present production
and usage distribution for 1969 for phosphate rock mined in the United
States. The figures given are those of the Bureau of Mines, adjusted to
include only direct usage of rock for triple superphosphate manufacture,
and to reflect all rock used for wet-process phosphoric acid manufacture.
The state of the art of control technology in the phosphate rock
processing industry is quite good. Systems for fluoride emission control
have been in service in the phosphate industry for many years. Concentration
*Solubility in neutral ammonium citrate solution is used as the means of
measuring availability as nutrient of the P2O5 content of processed
phosphates. All water soluble phosphates are citrate-soluble; only a
part of the citrate soluble phosphates are also water soluble.
3-137

-------
Table 3-40. Production of Phosphate Rock in the
United States (4262) (Thousand Short
Tons)


Florida^





Total


Tennessee
Western
States
United States
Use
Rock
P2°5
Content
Rock
P2°5
Content
Rock
P2°5
Content
Rock
P2°5
Content
Domestic:








Agricul tural
17,501
5,629

...
1,039
328
18,540
5,958
Industrial
553
166
3,193
851
3,107
772
6,853
1,789
Total
18,054
5,795
3,193
851
4,146
1,100
25,393
7,747
Exports
10,811
3,519

...
525
166
11 ,336
3,685
Total
28,865
9,313
3,193
851
4,672
1,266
36,730
11,431
^Includes North Carolina
^Includes Alabama (1969)
3
Data may not add to totals shown because of independent rounding
Table 3-41. Phosphate Rock Sold or Used by Producers,
by Uses and States (4262) in 1969
(Thousand Short Tons)
State
Mine
Production
Mine Production
Used Directly
Washer
Production
Marketable
Production
Rock
P2°5
Content
Rock
P2°5
Content
Rock
P2°5
Content
Rock
P2°5
Content
1969








Florida^
111 ,178
15,711
92
28
29,838
9,575
29,930
9,603
4
Tennessee
5,648
1,080
533
128
2,741
730
3,274
858
2
Western States
4,886
1,253
3,539
905
982
318
4,521
1 ,223
Total''
121,712
18,044
4,164
1,061
33,561
10,623
37,725
11,684
^Includes North Carolina
2
Includes California, Idaho, Montana, Utah, and Wyoming
3
Data may not add to totals shown because of independent rounding
4
Includes Alabama
3-138

-------
Table 3-42. Direct Uses of Phosphate Rock in the United States
in 1969 (Thousands of Short Tons)
Use
Rock
P205 Equivalent
Wet Process Phosphoric Acid
12,688
4028
Electric Furnace Phosphorus
6,758
1759
Triple Superphosphate
1,085
374
Normal Superphosphate
3,524
1150
Other
1,338
436

25,393
7747
of phosphate rock processing plants in particular geographical areas such
as Florida has led to increased surveillance by environmental control
authorities. This is due both to concern for the environment and to the
legal actions brought by plant neighbors due to alleged damages to/
cattle, plants, or fruit because of the effects of fluoride emissions.
This attention has caused the companies concerned to institute effective
monitoring and control of fluoride emissions at many of the phosphate
processing plants. In this context, control means prevention of emission
of fluorides into the ambient environment. While it is true that fluorine
and fluorine compounds are valuable industrial commodities and that the
fluorine liberated from phosphate rock processing annually could constitute
a major supply of fluorine, fluoride released from processing of rock is
considered by the industry a waste product to be disposed of in the most
expeditious manner. Emphasis on more restrictive and expensive fluoride
control requirements as well as an increased focus on the value of the
waste fluoride is adding emphasis to efforts to capture the emitted fluorides
in a saleable form.
3.5.3 Production Trends
The principal solid fertilizer products in the future will continue to be
ammonium phosphates, triple superphosphates, normal superphosphates, solid
and liquid NPK mixtures, and nitrophosphates. Phosphoric acid will continue
to be the key intermediate in the production of triple superphosphate, the
3-139

-------
ammonium phosphates and superphosphoric acid, and will be used as now in
mixtures and direct applications. It is doubtful that any new product will
appear within the next thirty years that is not related to one or another
of these materials or that will be independent of the use of phosphoric acid,
sulfuric acid or nitric acid as an intermediate. (^88). Estimated future
production levels in the United States of the primary phosphate containing
materials are listed in Tables 3-43, 3-44, and 3-45.
These projections reflect an outlook for a continually declining
market for normal phosphate and a stable market attended by some increase
for triple superphosphate; the market for ammonium phosphate is expected to
expand rapidly until 1980 and then drop down somewhat by 2000. Super-
phosphoric acid, which is now a main supplier to the liquid market, will
increase its share continually for the foreseeable future. Merchant phos-
phoric and superphosphoric acids are expected to account for about 25% of
the P90r liquid fertilizer by the year 2000. Nitrophosphates appear to have
L J
a long term outlook for slow growth.
Table 3-43. U.S. Agricultural Phosphate - Recent
and Estimated Future Production(4226)
(all tonnages as P2^5^





Estimated


1950
1960
1970
1980

2000

MM
Tons
t
Mkt.
MM
Tons
%
Mkt.
MM
Tons
t
Mkt.
MM
Tons
%
Mkt.
MM
Tons
t
Mkt.
Normal Superphosphate
1.7
85.0
1.3
46.4
0.7
13.5
0.3
i.l
0.2
1.5
Triple Superphosphate
0.3
15.0
r.o
35.7
1.4
26.9
1.5
20.2
2.7
19.5
Ammonium Phosphates
-
-
0.4
14.3
2.4
46.1
4.3
58.1
7.0
50.7
Superphosphoric Acid
-
-
-
-
0.4
7.6
0.8
10.8
2.8
20.3
Merchant Phosphoric Acid
(mixtures, direct
application and animal
feed phosphate)


0.1
3.6
0.2
3.8
0.3
4.1
0.7
5.1
Nitrophosphates
-
-
-
-
0.1
1.9
0.2
2.7
0.4
2.9
Total Agricultural
Po0r Production
L 0
2.0
100.0
2.8
100.0
5.2
100.0
7.4
100.0
13.8
100.0
Total Agricultural
PgOg Consumption
2.0

2.6

4.6

7.2

13.7

Total Wet Process
h3p°4
-
-
-
-
3.8
-
-
-
13.1
-
3-140

-------
Table 3-44. Defluorination of Phosphate Rock

1970
2000
Phosphate Rock
Used (1000 tons
as P2°5)
87.7(4264)
141.2
U.S. Population.
(Millions)
208.8
336.2
Per-Capita
Consumption
(lb/person)
0.84
0.84
Table 3-45.
Electric Furnace Phosphorus Production

1968
2000
Phosphate Rock
Used (1000 tons
as P205)
1570^4393^
2620
U.S. Population
(mill ions)
202.3
336.2
Per-Capita
Consumption
(lb/person)
15.56
15.56
3-141

-------
Table 3-46. Phosphate Industries Fluoride Emissions
Wet Process	Diairmonium	Triple	Normal	Electrothermal Defluorination	Industry
Phosphoric Acid	Phosphate Superphosphate Superphosphate Phosphorus	of Phosphate Rock	Totals
1970 Production
(10® tons/yr)
3.8^
2-4(b)
1.4
o.7(b>
1
0.09^' '
10.4^
Projected 2000
Production
(70® tons/yr)
13
7.0
2.7
0.2^
2.6^
0.14^
26.3
Soluble Fluoride
Evolution Factor
4.07^
lb F/ton P^Oj.
1.31 ^
lb F/ton P^Og
21.2'C)
lb F/ton P20,j
71
lb F/ton P205
30
lb F/ton P205
243^
lb F/ton P205
16 to 9.3^C,E'
lb F/ton P20g

in Product
in Product
in Product
in Product
In Phosphate
Rock Feed
ir, Phosphate
Rock Feed
Equiv. in
Products
Soluble Fluoride
Emission Factor
with Current
Control
3.36^
lb F/ton P205
in Product
0.23^C^
lb F/ton P205
in Product
5.4^
lb F/ton P205
in Product
14.2
lb F/ton P20j
in Product
5.1
lb F/ton P205
in Phosphate
Rock Feed
3g(c)
lb F/ton P205
Phosphate
Rock Feed
4.1 to 3.l(C,E*
lb F/ton P205
Equiv. in Products
Soluble Fluoride
Emission Factor
with 99% Control
0^04^
lb F/ton P205
0.013^
lb F/ton P20g
0.21 ^
lb F/ton P205
0.71
lb F/ton P205
0.30
lb F/ton P205
2.4
lb F/ton PjO,.
0.09^
lb F/ton P205

in Product
in Product
in Product
in Product
in Phosphate
Rock Feed
in Phosphate
Rock Feed
Equiv. in Products
Soluble Fluoride
Evolved in 1970
(103 Tons F/year)
7.73(C)
1.57
14.8
24.9
24.0
10.9
83.9
Soluble Fluoride
Evolved in 2000
(103 tons/year)
26.4(C)
4.59
28.6
7.10
39.0
17.0
123
Soluble Fluoride
Emission in 1970
6.38^
0.28^
3.78^
4.97
4.08
1.76^C)
21.3
(103 tons F/year)







Soluble Fluoride
Emission in 2000
with Current Control
21.8^
0.81^C ^
7.29^
1.42
6.63
2.73^
40.7
(103 tons F/year)







Soluble Fluoride
Emission in 2000
with 99J Control
0.26^
0.046^
0.28^
0.071
0.39
0.17(C)
1.22
(103 tons F/year)







(A) Expressed as P,0,- equivalent in phosphate rock feed	(C) Includes gypsum pond emissions
.	.	.	(B) 1968 Data
(I) Expressed as P^Og equivalent in product	(E) Reflects change due to shifts in production trends thru 2000

-------
3.5.4 Fluoride Control and Emissions Summary
Fluoride control systems for the phosphate rock processing industry
typically utilize water scrubbers to absorb the gaseous fluoride effluents
from the individual processes. The Individual control systems appropriate
for each fertilizer product are described in the following section.
Soluble fluoride evolution and emission associated with the current
and projected production levels are tabulated in Table 3-46. It should be
noted that the emission factors of the individual processes have been
allocated their share of the soluble fluorides volatilized from gypsum ponds
so that each emission factor is a composite of appropriate manufacturing
process emissions and gypsum pond emissions.
3.5.5 Process Description
3.5.5.1 Wet Process Phosphoric Acid
Process Description. Wet process phosphoric acid is produced by
the digestion of phosphate rock with sulfuric acid. The wet process acid
contains impurities that preclude its use in pharmaceuticals and similar
products so that acid made from elemental phosphorus (furnace grade or
"white acid") is used for these products. Wet process acid is used almost
exclusively in the manufacture of concentrated phosphate fertilizers. The
reactants in wet process phosphoric acid manufacture are phosphate rock,
which contains 30 to 35% ?20g and 3 to 4% fluoride, and sulfuric acid at
93 to 99% concentration. The primary products of the reaction are impure,
dilute, phosphoric acid containing 28 to 32% P205 and impure gypsum
(calcium sulfate dihydrate).
Various concentrations of wet process phosphoric acid are used for
the different fertilizer products. Although some dilute (roughly 32% P2O5)
acid is used directly for manufacture of a few products and acid concen-
trated to approximately 38 to 40% is used for some of the ammonium phosphates,
the great majority of the acid produced is concentrated to about 54% P2°5-
3-143

-------
Essentially the same basic processing steps are involved in making wet
process phosphoric acid at any acid manufacturing facility although the
process hardware and flow sheet will vary from plant to plant. A typical
modern wet process phosphoric acid plant flow sheet is shown in Figure 3-30.
Individual plants differ in the number and arrangement of reactors, recycle
flows, types of evaporators, etc. The similarity of requirements at all
plants has caused a few given hardware items to become "the accepted devices"
for given applications. For instance, most modern plants now use tilting
pan filters for gypsum filtration. This phenomenon leads to some measure
of uniformity in equipment selection, plant size, and emission factors for
wet acid plants.
Production Trends. As shown in Table 3-46, production of wet
process phosphoric acid is more than twice the production of the next largest
phosphate rock product. The major uses of the acids are in the production
of ammonium phosphates, triple superphosphate, and dicalcium phosphate with
about 75 percent of production going into ammonium phosphates and about 20%
into triple superphosphate. Concentration of the acid to produce super-
phosphoric acid (approximately 70% P^Og content) has promise of becoming
increasingly significant. Superphosphoric acid is a concentrated, convenient
form of PgOg for shipping or further processing.
The production of wet process phosphoric acid is expected to grow
from 3.8 million tons per year in 1970 (as P?0r) to 13 million tons per
year in 2000.^4226)
Fluoride Emission Control Techniques. Standard practice in emission
control from wet process phosphoric acid plants involves the universal appli-
cation of wet scrubbing systems. Specific types of wet scrubbing systems
currently employed include the liquid ejector venturi scrubber, liquid
impingement control systems, and the spray tower. These processes are pre-
sented as control process flow diagrams and mass balances in Figure 3-31.
3-144

-------
ATM.'
HEATED WASH WATER
FAN
GYPSUM FOND WATER
SUCK
DRY
CLOTH
WASH
CAKE
REMOVE
SUCK
DRY
WASH
LIQUOR
WA5H
WASH
TILTING PAN FILTER
VACUUM'
WATER STEAM
TO VACUUM
ATMOSPHERE
FAN
ATM.'
ATM.
STEAM
CONDENSER
SUMP
ATM'
COOLER
SUMP
ATM.
HOT WELL
TO GYPSUM POND
FEED —J-
MATE RIALS
SINGLE TANK
REACTOR
FILTRATE SEAL TANK
ATM'
(LAUNDER)
CONCENTRATED
X H3P04
® (54% P,0,)
ATM"
32%
GYPSUM SLURY TO POND
SURGE
TANK
EVAPORATOR
NOTES:
OF ALL GASEOUS FLUORIDE POLLUTANTS FOUND IN THIS STREAM.
, APPROXIMATELY
2. THOSE STREAMS WILL CONTAIN SMALL AMOUNTS OF FLUORIDE POLLUTANTS AND MAY
BE SCRUBBED PRIOR TO VENTING.
3. POSSIBLE FUGITIVE EMISSIONS FROM LAUNDERS (SEMICIRCULAR TROUGHS)
BASIS - 1000 TONS/DAY PHOSPHORIC ACID (P205) PRODUCTION
PROCESS STREAMS - TONS/DAY

Stream Number
Material
1
2
3**
4*
5*
6**
7**
8**
g**
10**
n
H2SiF6

23(1)^

44(1)(Est)
107(2){Est 3





67(1)
HF
sif4
CaF2 equiv.
280(G)

Unknown

4.2(s)(Est)
Q.5(g)(E>
Q,2(s)^
0.03(g)*E*
0.03(g)^
0.01(g)(E)
0.03(g)(E)

Total Fluorides
Total as F
280
140
23
18
Unknown
'44
35
m
86.5
0.7
0.4
0.03
0.02
0.03
0.02
0.01
0.007
0.03
0.02
67
53
Phosphate Rock
3300



50(Est)



0.06(s)(Ej


Sulfuric Acid
(96%)
2900










Phosphoric Acid
(54»205)

1850









Phosphoric Acid
(32%P205)










3120
h2o



1270(g)(Est)







Gypsum




5000






Approx. Total
Stream
6400
1900

1300(A)
5000^
3
-------
®
WET SPRAY SCRUBBER
EFF. =60% (EST.) PARTICULATE
65% GASEOUS
TO FAN
AND STACK

SPRAY
WATER
WATER AND
SOLIDS
GYPSUM
i POND
1.	ALL OR PART OF	COMPLETELY
UNCONTROLLED IN MANY PLANTS.
2.	A FEW PLANTS LIME THEIR GYPSUM PONDS
TOpH4 TO DECREASE HF EVOLUTION
®
VENTURI SCRUBBER
EFF. = 99% PARTICULATE
90% GASEOUS

TO FAN
AND STACK
WATER
©
SPRAY CROSS-FLOW, PACKED
SCRUBBER
EFF. = 95% (EST.) PARTICULATE
79% GASEOUS
WATER AND
SOLIDS
TO ATM
: GYPSUM
POND
TO FAN
AND STACK
SPRAY
WATER
SPRAY
WATER
WATER AND
SOLIDS
TO
ATM.
GYPSUM
POND Sk
BASIS-1000 TONS/DAY PHOSPHORIC ACID (54% P20g) PRODUCTION
PROCESS STREAMS - TONS/DAY
Material
Stream Number

6
7
8
9
10
12*
13
14*
15
16*
17
18*
19*
20*
Sip4
HF
CaF2
0.5(g)(A)
; 0.2(s) ^
0.03(g)(A)
0.03(g)(A)
0.01(g)(A)
0.03(g)^
0.21(g)(A)
(A)
0.08(s)
0.39(aq)
0.12(aq)
0.06(g)(A'
(A)
0.002(s)
0.54(aq)
0.2(aq)
0.13(g)(A'
(A)
0.01(s)
0.47(g)
0.19(aq
(A)
1.65(9)
1.65(g)
%
(A)
1.65(g)
Total
Fluorides
Total as F
0.7
0.4
0.03
0.02
0.03
0.02
0.01
0.007
0.03
0.02
0.29
0.19
0.51
0.34
0.062
0.044
0.74
0.49
0.14
0.10
0.66
0.44
1.65
1.57
1.65
1.57
1.65
1.57
Phosphate
Rock
3(s)(A)


0.06 (s)'A'

1.22(s)
1.84(aq)
0.03
2.97
0.18
2.91



Approx.
Total
Stream
3(B)
0.03'C'
0.03^
0.1(D'
0.03^C'
i .si (b»c»d:
2.35(E'
0.092'B,C'D
>3.71
-------
The state of the art with regard to pollutant collection and removal
from wet phosphoric acid manufacture must be rated as good since almost any
desired collection efficiency can be achieved by the design and application
of appropriate control devices. Liquid collection systems using either
fresh or recirculated water have proven efficient in the collection of
silicon tetrafluoride and hydrogen fluoride. Adequate particulate collection
can also be achieved with the ejector venturi and cross-flow packed scrubber
being the most efficient collectors in this regard. The major single con-
sideration in sustaining the high collection efficiencies which new equip-
ment is capable of providing is proper maintenance of control process
equipment. Without such maintenance on a continuous basis, collection
efficiencies and removal efficiencies drop at an exponential rate, due to
plugging, corrosion, and erosion.
The principle soluble fluoride emission sources in phosphoric acid
manufacture are the acid and rock mixing points, reactor (digester) tanks,
and the liquid filtering units. Reactor tanks are generally closed systems
with the free space maintained at pressures slightly below ambient atmos-
phere. The other emission sources are also hooded. Gaseous effluents from
the emission sources are transported through large ducts at moderate linear
velocities to the abatement devices and prime movers. The transfer system
ducts are frequently rectangular in cross section with removable lids or
plates to permit frequent wash out for removal of silica hydrate gel. Duct
system linear velocities in plants equipped with vacuum flash coolers
range from 5 to 50 feet per second. Slurry transport launders, which recycle
material to the digester tanks, normally vent to the free space of the
digester tanks. Tilting pan filter hood systems normally collect fune
evolution from only the "32% acid" and "first wash" sections of the filter,
using sheet rubber flaps to form make-and-break flexible seals with the
filter pans as they rotated. "Thirty-two percent acid" and "first wash"
filtrate receivers are normally the only filtrate receivers vented to
fluoride emission control devices, and in many facilities they are vented
directly to the atmosphere. The digestion chambers and filtering units
are the primary sources of atmospheric contaminants. Additional sources
associated with phosphoric acid manufacture are the evaporators, which are
used to concentrate the product acid, and the waste water ponds used for
3-149

-------
the storage of gypsum and scrubbing water. Evolutions from the acid con-
centration (evaporation) units are absorbed in the cooling water used in
the vacuum system barometric condensors. The primary fluoride emissions
are silicon tetrafluoride and hydrogen fluoride. In addition to the major
sources of fluoride, emissions, there are many miscellaneous minor sources.
Process vents from such sources as the phosphoric acid tanks and transfer
facilities release exhaust gases to the atmosphere.
One of the primary unsolved problems relating to fluoride emission
in the wet phosphoric acid manufacturing industry, and in the entire phos-
phate rock processing industry, is the ultimate disposal of the fluoride
collected by the wet scrubbing systems used to reduce atmospheric pollution.
Volatilization of fluorides from gypsum/scrubber recirculation ponds repre-
sents a significant portion of the overall emissions. The solution-liming
of the gypsum ponds—is employed by only a few of the processors.
Fluoride Emissions. The fluorides evolved from wet process acid
manufacture consist primarily of gaseous Siand HF emanating from the
digestors, filters, and various sumps and vents. Since the entire process
involves liquid streams, particulate evolution is quite low with only minor
quantities of rock emitted from grinding, handling, and process streams. In
terms of fluoride evolution, an anomalous situation exists for the acid con-
centration step. Approximately 70% of the fluoride in weak acid is
volatilized in the concentration process. However, the water vapor is recon-
densed in the barometric condensers in the vacuum system which results in
collection of almost all of the fluoride. Since this collection is accom-
plished as an intrinsic part of the process and no specific collection device
is required, this volatilization and recapture of fluoride will be consid-
ered internal to the process and not a fluoride evolution. This distinction
could not be made for acid concentration by submerged combustion; however,
submerged combustion is no longer used significantly, partially because of
extensive fluoride evolution. A parallel situation exists for cooling of
the digestor slurry where vacuum cooling has displaced air cooling. In
both cases, the quantity of soluble fluorides volatilized from the gypsum
pond due to barometric scrubber discharges is allocated to phosphoric acid.
3-150

-------
Fluoride evolution and emission data have been published for various
facilities and circumstances. The values reported typically do not include
adequate information to allow calculation of emission factors and a material
balance. Public Health Service document AP-57, "Atmospheric Emissions from
Wet Process Phosphoric Acid Manufacture,"(4262^presents the most comprehensive
data available since it includes plant type, capacity, production rate,
equipment descriptions, fluoride evolution and emission factors, concentra-
tions, etc. The evolution and emission factors and effluent concentrations
reported in AP-57 are consistent with RRI experience for similar processes
and with data from other sources such as Huffstutler and Starnes; therefore,
AP-57 will be considered the prime data source for evaluation of evolution
and emissions from wet phosphoric acid manufacture. A summary of the
reported evolution factors is presented in Table 3-47. Concentration
levels for soluble fluorides as fluorine were 3 to 40 ppm or 0.0011 - 0.0147
gr/scfm(^®^ It is apparent from examination of this table that the range of
possible evolution factors and concentrations is very large. It should also
be considered that these data were likely to have been obtained under the
best conditions for minimum evolution. The actual evolution factors in a
given case are dependent on such parameters as feed material analysis,
acidulation ratio, operating temperature, and quality of processing equip-
ment and maintenance. High range evolution factors appear to be on the
order of 2 to 3 pounds of fluoride per ton of PgO^ produced. Evolution
factors of this magnitude are believed to result when well-designed facil-
ities are properly operated and maintained. Much higher evolution rates
may be observed in some cases when improper maintenance or operation of the
facility creates operating conditions which preclude operation at low fluoride
outputs. If monitoring of fluoride concentration indicates evolution factors
significantly above the 2 to 3 pounds per ton of P205 level, the need for
corrective action should be recognized. In a similar manner, well-designed
scrubber installations should run above 85% efficiency and in some cases in
the high 90% efficiency range. If a scrubber of good basic design drops
below this range of values, some operational problem probably exists.
Typically, such problems result from deposition of hydrated silica within
the scrubber packing or water nozzles so that the liquid-vapor contact is
affected.
3-151

-------
Table 3-47. Summary of Evolution Factors, and Emission Factors
from Reference 4262
Process Element
Evolution Factor
(pounds soluble fluorides
as fluorine/ton P2O5)
Emission Factor
(pounds soluble fluorides
as fluorine/ton P2O5)
Digester/Reactor
Filter
Sump and Vents
TOTAL
0.037 - 2.16
0.011 - 0.063
Up to 0.26
0.006 - 0.17*
*9 out of 10 plants
The material balance is important in determination of the disposition
of fluoride in the various streams associated with phosphoric acid production.
In this connection, the material balance presented on page 14 of AP-57 was
evaluated and found to be inconsistent with the experimental data presented
in the same document. Since the reported material balance data were there-
fore questionable, the AP-57 experimental data for the digestor evolution
were combined with the TVA data to serve as a basis for the proportioning
of fluoride distribution in other streams. The resulting fluoride distri-
bution is shown in Table 3-48. The maximum total fluoride evolved in the
three gaseous streams is approximately 2.5 pounds fluoride per ton P2°5«
Emission factors from wet process acid plants range from 0.003 to
0.3 pound of soluble fluorides expressed as fluoride per ton P205 according
to RRI experience over a range of equipment types and sizes. The equivalent
range from AP-57 is 0.006 to 0.6 pound per ton.
The soluble fluorides emitted by wet process phosphates and plants
are estimated at 6380 tons for 1970, and are projected to reach 21,800 tons
per year in 2000 if current practices are continued. If technology capable
of 99% effectiveness is adopted and properly maintained, emissions in 2000
will drop to 260 tons per year.
3-152

-------
Table 3-48. Fluoride Distribution for Wet Process
Phosphoric Acid
Process Stream
Factor
(pound fluorine/ton P2O5)
Reactor/Digester Gaseous Effluent
0.04 to 2.2
Filter Gaseous Effluent
0.01 to 0.063
Sump and Vent Gaseous Effluent
Up to 0.26
Gypsum
97.5
Evaporation
79.2
Acid
32.6
3.5.5.2 Ammonium Phosphates
Process Description. The term "ammonium phosphates" describes a large
number of products prepared by combining phosphoric acid, sulfuric acid,
ammonia, TSP, urea, and potash to make various nitrogen-phosphorus-potassium
(N-P-K) materials. About three-fourths of ammonium phosphate production is
accounted for by diammonium phosphate (DAP) which is typically 18-46-0 material
(18% nitrogen, 46% available P2°5» and K20.
DAP will be considered representative for the purpose of evaluating
evolution and emission of fluorides from all ammonium phosphates since it is
by far the largest individual product and process data for DAP are readily
available. Figure 3-32 presents a process schematic for DAP.
Production Trends. Ammonium phosphate production will increase from
2.4 million tons (as PgOg) in 1970 to 7 million tons in 2000!4264^
Fluoride Emission Control Techniques. The wet process phosphoric
acid used in the manufacture of diammonium phosphate is the source of fluoride
emissions from the process. Wet process acid of the concentration employed
3-153

-------
TO SCRUBBER
TO SCRUBBERi PHOSPHORIC
„ NH, f ACID
TO SCRUBBER
RECYCLE FINES
WATER
AMMONIATOR
COOLER
DRYER
COARSE
PRENEUTRALIZER
MILL
FINES
TO SALE
SCREENS
BULK
STORAGE
SHIPPING
BASIS: 500 TONS/DAY 18-46-0 DIAMMONIUM PHOSPHATE
PROCESS STREAMS TONS/DAYS
Ma ten al
1
2
3
4*
5*
6*
7*
8
NH4F Equiv.
l^SiFg Equiv.
CaF2 Equiv.


7.8(1)W
0.08(g)(Est)
0.067{g)(C^
0.067{g)(C)
0.02(g)(Est)
%
12.5(s)
Total Fluoride
Total As F


7.8
6.2
0.08
0.06
0.067
0.034
0.067
0.034
0.02
0.02
12.5
6.1
nh3
Phosphoric
Acid(4-5% P205)
h2o
18-46-0
Diararnonium
Phosphate
(1% wt Moist)
105(g)
-------
generally contains around 2% fluoride. Less fluoride is evolved in the
manufacture of either normal or triple superphosphate fertilizer. The
primary emission sources for fluoride are the reactor/granulator system and
the dryer. Two different scrubber systems are generally used by diammonium
phosphate plants. The first system is used to recover ammonia escaping from
the reactor and to control gaseous fluorides evolved in the reactor and
ammoniator/granulator. The ammonia scrubber and preneutralizer used in the
TVA diammonium phosphate process are closed vessels which are vented via
large ducts. Evolved gases are transferred from the preneutralizer under
induced draft to the ammonia scrubber and fluoride control equipment. The
granulator is similarly vented. The second control system is used to control
dryer evolution and the dust generated by the screening operations. Liquid
scrubbing systems are typically used for the control of both exhaust gas
streams. The scrubbing liquid in the first system is the wet process phosphoric
acid used in the process. This liquid, which is recycled, has a high fluoride
content (about 8000 ppm) and is acid (pH from 2 to 4.5).
Control devices for diammonium phosphate production facilities are
generally chosen at the time of plant construction. Devices in general con-
sist of cyclonic spray towers, venturi scrubbers, and impingement scrubbers.
Each of these device types have proven efficient in reducing fluoride
emissions to acceptable levels. Generally, design is such that maintenance
procedures are simple and effective enough to allow the control device to
be maintained at the original design operating efficiency. Optimum condi-
tions for the recovery of ammonia, that is, pH of 4.5 or lower and absorp-
tion system temperatures of about 160°F, may also cause fluoride to be
stripped from the ammonia scrubber liquid and re-enter the exhaust gas
stream.
The wet scrubbing devices used in the second scrubbing system must
provide a satisfactory collection efficiency for gaseous fluorides at low
fluoride concentrations in the exhaust gas stream and efficiently remove
particulate material containing 1 to 2% fluoride. In this case, the exhaust
gas effluent is generated by the cooling, drying, and sizing sections of the
diammonium phosphate processing plant. A controlled process model is pre-
sented as Figure 3-33.
3-157

-------
NOTE:
SEVERAL TYPES OF SCRUBBERS ARE IN
GENERAL USE, INCL. CYCLONIC- TO ATM.
SPRAY TOWERS, VENTURI SCRUBBERS,
IMPINGEMENT SCRUBBERS
SPRAY
WATER EFF> = 79% xv
'—~ T A II ClAC N/
J<3>
PLANT
STACK
H3P°4
TAIL GAS
SCRUBBER

.	» JV*(\UDDC(\
I <8>^t%
SCRUBBER
TO GYPSUM POND
EFF. = 79%
0
-
TO
PRENEUTRALIZER
MAKE-UP
H3P°4
H3P°4(aq)
pH2
<£>—
DUST
SCRUBBER
RECYCLE
TANK
TO
RECYCLE
FINES
TO ATMOSPHERE
(1) GYPSUM POND
(I) A FEW PLANTS LIME THEIR GYPSUM PONDS TO PH 4 TO DECREASE FLUORIDE
EVOLUTION
BASIS - 500 TONS/DAY
PROCESS STREAMS - TONS/DAY
Materials
Stream Number
4
5
6
7
8*
9*
10
11
12
13
14
15
NH^F equiv.
0.08(g)
0.067(g)
0.067(g)
0.02(g)
0.002(g)(Est)
0.049(g)^
0.147(g)
0.031(g)
0.018(g)
0.069(g)
0.116(1)

Total Fluorides
Total as F
0.08
0.041
0.067
0.0345
0.067
0.0345
0.02
0.010
0.002
0.001
0.049
0.025^
0.147
0.075^°'
0.031
0.016^
0.018
0.009^
0.069
0.035^
0.116
0.060^

NH3
h2o
2(g)
50(1)
Kg)
81(g)
Kg)
42(g)
5(g)





1(1)
47(1)

*
3(1)
131(D
Approx. Total
Stream
52
82
43
5
—
0.049
0.147
0.031
0.018
<
CO
0.116(B)
134^
* Gaseous Effluent Stream
(A)	Plus recycling H^PO^ containing 8000 ppm F
(B)	Plus scrubbing water
(C)	Plus ammoniated H3P04
(D)	Reference 4267
(E)	Assumes no line addition to gypsum ponds
Source
Soluble Fluoride Emission Factor - lb F/ton P^Og in product
Stack Emission
DAP Gypsum Pond
Emission
0.22
0.009
Total Soluble
Fluoride Emission
0.23
Overall soluble fluoride emission factor for diammonium phosphate production (including gypsum ponds)
0.23 lb F/ton P20g in product
Figure 3-33. Diammonium Phosphate
Production — Controlled
Process Model
3-159

-------
Fluoride Emissions. Fluoride input to the process is from the wet process
phosphoric acid. The acid is typically used at a concentration of about
40 percent P^. Such acid has a fluoride content of about 93 pounds
fluoride/ton P205- This residual fluoride content has been retained through
the acid production process and is relatively stable under the conditions
associated with DAP production. Evolution occurs at the reactor/granulator
and the dryer and screens. Table 3-49 presents the fluoride evolution
and emission from manufacture of DAP.^^ Based on Table 3-49 values,
the annual fluoride evolution from ammonium phosphate production is 966 tons
fluoride per year. The annual soluble fluoride emission is 161 tons fluoride
per year. These values were generated utilizing a 3,222,000 tons P205 per
year production rate for ammonium phosphate as reported in preliminary form
under the Engineering and Cost Study of Emissions Control in the Phosphate
Industry (CPA 70-156).
Table 3-49. Evolution and Emission of Fluorides from
Production of Ammonium Phosphates
Process Evolution
Source
Evolution (pounds
soluble fluorides as
fluorine/ton P2O5)
Emission (pounds
soluble fluorides as
fluorine/ton P2°5)
Preneutralizer/
0.3
0.05
Ammoniator


Dryer/Cooler/Screens
0.3
0.05
TOTAL
0.6
0.1
3.5.5.3 Tr1pie Superphosphate
Process Description. Triple superphosphate (TSP) is made by acidulating
phosphate rock with wet process phosphoric acid. The product may be in either
pulverized or granular form; the P205 content is approximately 46%. TSP is
relatively concentrated, which minimizes shipping costs, and the phosphorus
3-161

-------
content is almost entirely in plant available form. Detailed information on
chemistry, processes, and equipment is available in References 4242, 4263,
4264, and 4265.
Figure 3-34 is a typical process diagram for production of granulated
triple superphosphate. Production of granular triple superphosphate will be
considered as typical of the problems since "run-of-pile". production is
decreasing rapidly.
Production Levels. Triple superphosphate will increase from a 1970
level of 1.4 million tons to 2.7 million tons in 2000^26^.
Fluoride Emission Control Techniques. Run-of-pile (ROP) triple
superphosphate setting belts and disintegrators are hooded and maintained
at slightly below ambient pressure. Gaseous effluents are transported at
moderate linear velocities through large ducts which are generally rec-
tangular in cross-section. Conveyor belts transporting the fresh ROP triple
superphosphate to the storage building are, to an indeterminate extent,
hooded and vented to the storage building. Most triple superphosphate
storage buildings are maintained slightly below ambient pressure and are
vented through roof outlets to fluoride control devices and prime movers.
Screens and mills for processing cured triple superphosphate are hooded
and vented through circular ducts to the cyclones and then to prime movers.
The acidulators employed in granular triple superphosphate manufacture
are closed vessels maintained slightly below ambient pressure and vented to
fluoride control equipment. The launders between tanks vent to the free
space of the acidulators. The granulator and dryer are hooded as are the
screens and crusher. Transfer ducts for dust and gaseous pollutant laden
effluent gas streams are provided with clean-out plates in the form of
removable ports or lids. The controlled process model is presented in
Figure 3-35.
3-162

-------
BASIS - 200 TONS PRODUCT/DAY (48% P205)
PROCESS STREAMS (TONS/DAY)
WET-PROCESS
PHOSPHORIC
ACID
PHOSPHATE
ROCK
CONSTANT
WEIGHT
FEEDER
ACID
CONTROL
RECYCLE FINES
WATER
JXEI
TO ATMOSPHERE
.GRANULATOR
DEN
AIR AND GAS
TO ATMOSPHERE
OR SCRUBBER
ROTARY
DRYER
OVERSIZED^
SCREENS
COOLER
¦~FINES TO RECYCLE <8
AIR
PRODUCT TO
STORAGE OR
SHIPMENT
HAMMERMILL




Stream
Number



Material
1
2
3*
4
5*
6*
7
8
H2SiF6 (equiv.)
SiF4
CaF2 (equiv.)
7.4(s)(B)
3.0(1)^
0.64(g)(C)
12.9(s)
0.73(g)^
0.07(g)(C)
10.16 (s)
1.5(s)
Total Fluoride
Total as F
7.4(s)
3.6 (s)
2.4
0.64
0.47
12.9
6.28
0.73
0.53
0.07
0.05
10.16
4.95
1.5
* 0.75
Triple Super-
phosphate
48% P205



217(s)


190.0(s)
28.0(s)
Wet Process
Phosphoric Acid
54%

129(1)^






Phosphate Rock
32.3% P205
83(s)(B)







h2o


15(g)

5.0(g)
2.0(g)


Approx. Total
Stream
90
132
16
230
60
2
200
30
F Evolution Factor = 21 lb F/ton P^g
* Gaseous effluent stream
(A)	Plus combustion products and diluent air
(B)	Reference 4264
(C)	Reference 506
Soluble fluoride evolution factor = 21 lb F/ton P205 produced.
Figure 3-34. Granulated Triple Super
Phosphate Production —
Uncontrolled Process
Model
3-163

-------
BASIS - 200 TONS/DAY (48X P^)
PROCESS STREAMS-TONS/DAY
W
I
CT)
tn
TO FAN
AND STACK
SPRAY
WATER
TO FAN
AND STACK
CYCLONIC I
SPRAY
TOWER11"
EFF. - 95%
r-rjr
WET; SPRAY
SCRUBBER^
TO ATM
13
EFF. = 95%
GYPSUM
POND
(SEVERAL IN
PARALLEL OR
SERIES)
NOTE:
(1)	SOME PLANTS LIME THEIR GYPSUM PONDS
TO p,H>4 TO REDUCE HF EVOLUTION
(2)	REFLECTS OPTIMUM OPERATING AND
MAINTENANCE CONDITIONS
Materials
Stream Number
3
5
6
9
10*
11*
12
13*
H2SiF6
SiF4
HF
0.64(g)
0.73(g)
0.07(g)
0.62(1)
0.03(g)(0)
0.03(g)(0)
0.63(1)<0>
0.01(g)(Est)
Total Fluorides
Total as F
0.64
0.47
0.73
0.53
0.07
0.05
0.62
0.49
0.03
0.03
0.03
0.03
0.63
0.50
0.01
0.01
H2°
15(g)
5.0(g)
2.0(g)
17(1)


5.0(1)

Approx. Total
Stream
15.6
5.73
2.07
17.6(A>
0.03
0.03
0.01
* Gaseous Effluent Streams
(A)	Plus 67,000 gal recycled scrubbing water
(B)	Plus recycled scrubbing water
(C)	Plus 33 x 10^ scf air
(D)	Reference 4267
(E)	Estimate average industry abatement efficiency of 75% caused by inproper maintainance and operation
of equipment and fugitive sources.
Source
Soluble Fluoride Emission
Factor - lb F/ton P2O5 Produced
(assuming optimum conditions)
Soluble Fluoride Emission
Factor - lb F/ton P2O5 Produced
(assuming average industry operation)
Control Oevice Emissions
1.2
5.2
TSP Gypsum Ponds
0.2
0.2
Total Soluble
Fluoride Emission
1.4
5.4
Overall soluble fluoride emission factor from triple superphosphate plants (including gypsum ponds) s	(E)
5.4 lb F/ton P2O5 produced.
Figure 3-35. Manufacture of Granular Triple Superphosphate - Controlled Process Model

-------
Fluoride Emissions. Fluorides enter the TSP process in the phosphate
rock and in the wet process acid. They are volatilized and evolved during
acidulation, digestion, and curing. In contrast to the wet acid and ele-
mental phosphorus processes, manufacture of TSP does not include a direct
contact condensation step as part of the basic process and does involve
transport of solids on conveyor belts as opposed to transport of liquids in
closed systems. The former characteristic means that any fluoride captured
to reduce emissions is caused by installation of control equipment. The
latter characteristic means that the processing system is not sealed as a
matter of course; conveyor belts, transfer points, etc., are all potential
emission sources if they are not hooded in some appropriate manner. After
acidulation and completion of the initial reaction, TSP is stored and cured
in large buildings for several weeks to complete the reactions converting
the phosphorus present to an available form. During this time period,
fluorides continue to evolve at a much lower rate than for the initial
j reaction steps. This long-term, low-level evolution, if uncontrolled, may
constitute the major portion of the emissions from a TSP manufacturing
facility.
I The process presented in Figure 3-34 is one of several currently used
for the production of granular TSP. An alternative, widely used process is
the Dorr-Oliver process. Data for the run-of-pile, and Dorr-Oliver
granulated TSP processes are presented in Table 3-50, covering the evolution
and emission factors for various source points in the TSP processes. The
rate of evolution of fluorides from the manufacture of TSP is dependent on
many factors, e.g., rock composition, acid/rock ratio, and temperature.
The reported values are based on RRI experience with several processes.
The emissions factors of Table 3-50 are based on the assumption that the
emissions from mixers, dens, reactors, and granulators, are controlled while
conveyors and cure buildings may or may not be. It is apparent that the
fluoride emitted can be reduced to low levels by application of appropriate
fume capture and control equipment (hooding and wet scrubbers). However, any
uncontrolled emission, even from a low level source such as the cure building,
becomes the dominant emission value and determines the process emission
3-166

-------

Table 3-50.
Evolution and Emission Factors for
Triple Superphosphate Manufacture

Product
Emission Source Points*
Evolution Factor
(pound soluble fluorides as
fluorine/ton P2O5)
Emission Factor
(pound soluble fluorides as
fluorine/ton P20g)
Run-of-Pile
Mixer Den Conveyor
3
0.1

Cure Building
3
O
CO
1
CO

TOTAL
6
0.4 - 3.1
Granulated
Reactor-Granulator
9
0.01

Dryer
12
0.3

Cure Building
3
0.3 - 3

TOTAL
24
0.6 - 3.3
*Evolved material from various points combined and routed to control device.

-------
factor. The higher evolution factor for granulated TSP is partially due to
utilization of less concentrated acid which contains more fluoride and to
higher temperature processing with application of heat for drying.
It is estimated that 3780 tons of soluble fluorides (expressed as F)
were emitted in 1970, and that, on the same basis, assuming current control
efficiencies, 7290 tons will be emitted in 2000. If 99% efficient control
techniques are employed, only 280 tons will be emitted in 2000.
3.5.5.4 Normal Superphosphate
Process Description. Normal superphosphate (NSP) is produced by
acidulating phosphate rock with sulfuric acid. The product contains calcium
acid phosphates and calcium sulfate with an available PgO,. content of about
20 percent. The production rate of NSP has been decreasing yearly as dif-
ficulty in competing with more concentrated fertilizers increases. NSP was
the first phosphatic fertilizer in major use and many of the production
facilities are outmoded. The cost of shipping a 20% available P2O5
product and the cost of updating facilities for efficient operation with
acceptable air pollution control is accelerating a trend toward reduced
NSP production.
NSP can be produced either batchwise or continuously as illustrated
in the process schematics in Figures 3-36 and 3-37. In both cases, there
are three processing steps involved. These are: mixing of the acid and
rock, temporary holding while reacting to form a solid (denning), and
storage for completion of reactions (curing). The time scale of the latter
two processes are about 1 to 2 hours for denning and about 4 to 6 weeks for
curing.
Production Trends. Production of normal superphosphate will drop
from 0.7 million tons in 1970 to 0.2 million tons in 2000^26^.
3-168

-------
PHOSPHATE ROCK	PHOSPHATE DUST
SURGE
SILO
FUMES TO SCRUBBER
AND INDUCED DRAFT FAN
SULFURIC ACID
TANK
CURING
~ FUMES
FUGITIVE
EMISSIONS
DEN
"JDOOR
T HOIST
DUST
SEAL
ACID
SCALE
DUMP
CAR
PRODUCT
ELEVATOR
MILL
CURING
PILE
PAN MIXER
CUTTER
NORMAL
SUPERPHOSPHATE
TO SALE
FUGITIVE EMISSIONS
DEN
DOOR
DEN CAR
BATCH DEN
BASIS: BATCH PRODUCTION OF NORMAL SUPERPHOSPHATE (0-20-0) AT
40 TONS PER BATCH ( 1 BATCH PER HOUR)
PROCESS STREAMS
Material
Stream Process
Tons/Batch
1
2
3
4*
5*
SiF4 (Equiv.)
CaF2 (Equiv.)
(C)
1.87(s)

, (C)
1.27(s)
0.36(gP
0.04(g)(C)
Total Fluorides
1.87

1.27
0.36
0.04
Total as f
0.91

0.62
0.26
0.03
Phosphate Rock
(75 b.p.l)
Sulfuric Acid
(54°Be)
25(s) ^
(A)
22(1)



Normal Super-
phosphate
(0-20-0)


/ N(C)
40(s)
(B)

Denning Fumes
Curing Fumes



13000(g)**
( L) -A--*--*-
0.03 (g)
Approx. Total
Stream
25
22
40
13,000**
0.03***
F Evolution factor = 73 lb F/ton Pg05	(A) Reference 4263
* Gaseous effluent stream	(B) Reference 4267
** SCFM	(C) Reference 4264
*** Fluorine component only
Soluble fluoride evolution factor = 73 lb	F/ton P205 produced
Figure 3-36. Normal Superphosphate
Batch Production - Uncon-
trolled Process Model
3-169

-------
DILUTION
WATER
SULFURIC ACID
STORAGE
SECONDARY
' COOLER
DILUTION
AND
COOLING
TANK
FUGITIVE EMISSIONS
PHOSPHATE ROCK
SILO
FUMES TO SCRUBBER
AND INDUCED DRAFT FAN
MIXER
CURING
FUMES
CUTTER
FUGITIVE
EMISSIONS
CONVEYOR
CURING PILE
AND STORAGE
WEIGH
FEEDER
MILL
CONTINUOUS
DEN
CONVEYOR
NORMAL
SUPERPHOSPHATE
TO SALE
BASIS: CONTINUOUS PRODUCTION OF NORMAL SUPERPHOSPHATE (0-20-0)
AT 200 TONS PER DAY
PROCESS STREAMS
Materials


Stream
Number


Tons/Day
1
2
3
4
5*
6*
SiF4 (equiv.)
Cat 2 (equiv.)
8.91(s)(A)


6.10(s)(A*
1.68(g)(A)
0.19?g)(A)
Total Fluorides
8.91


6.10
1.68
0.19
Total as F
4.34


2.97
1.23
0.14
Phosphate Rock
(75 b.p.l.)
120(s;(A)





Sulfuric Acid
(66 Be1)

75(1)(A)
30(s)(A)



Water





Normal Super- •
Phosphate
(0-20-0)
Denning Fumes



200(s)^
2600(gj
(A) ***
0.14(g)
Curing Fumes





Approx. Total
Stream
120
75
30
200
2600**
0.14***
* Gaseous effluent stream	(A) Reference 4263
** SCFM	(B) Reference 4267
*** Fluorine component only
Soluble fluoride evolution factor = 69 lb F/ton P2OJ- produced
Figure 3-37. Normal Superphosphate Con-
tinuous Production —
Uncontrolled Process Model
3-171

-------
Fluoride Emission Control Techniques. In both continuous and batch
processes for normal superphosphate manufacture, there are several points of
fluoride emission. Particulate fluorides are produced by the grinding and
drying operations performed on the phosphate rock prior to its acidulation
and by product handling. Gaseous fluorides, primarily silicon tetrafluoride,
are produced and emitted during acidulation, denning, transport, and curing
processes. The major portion of fluoride evolution occurs during acidula-
tion and denning with only minor evolution during the cutting and bulk
storage cure of the product. Effluent control systems have been designed
and applied to collect all significant process emissions associated with
normal superphosphate manufacture; however, typical practice may include
control of only the mixer and den due to economic constraints. In that
case, uncontrolled emissions from transfer and curing may become the
dominant emission.
The pan mixer and batch den used in batch manufacture of normal
superphosphate are vented through extremely large ducts ("tunnels") to the
fluoride control equipment and prime movers. The batch den is a closed
vessel during the period of time required for the single superphosphate to
"set up" except for the vent duct. When the NSP has "set up," a side of
the den is dropped. The cutter is hooded and ducts are provided with clean-
out doors. Continuous process superphosphate mixers, dens, and cutters
are hooded with the den and other equipment-maintained below atmospheric
pressure. The air swept into the system and evolved particulate matter
and fluoride pollutants are vented through extremely large ducts to
fluoride control equipment. The ducts are provided with drains and clean-
out doors.
The volume of exhaust gases evolved during the manufacture of
normal superphosphate varies greatly from facility to facility. Plant
design, tightness of the collection equipment, and the number of pieces of
equipment being vented all are factors affecting the exhaust gas volume.
The measured exhaust gas flow rates on existing normal superphosphate
plants ranges from 3,000 to 35,000 actual cubic feet per minute (ACFM). The
range in production capacity associated with these exhaust gas volumes is
Q
3-173 ,

-------
from 6 to 40 tons per hour superphosphate. Continuous plants typically
produce a larger volume of exhaust gas than do batch operations. However,
the exhaust gas stream from continuous operations generally has a lower
fluoride content.
The control of particulate fluoride emissions produced during the
rock grinding and drying operations represent a classical dust control
problem. Certain dust control devices such as cyclone collectors or
baghouses are generally considered to be part of the grinding process equip-
ment. The use of high-efficiency, multi-cyclone collectors and/or fabric
filters may produce efficiencies of from 99 to 99.5%. All of the particu-
late material collected is directly returned to process and so there is no
waste disposal problem associated with the particulate fluoride emissions.
The gaseous fluoride emissions produced by phosphate rock acidulation
both in the mixer and the curing den are generally controlled by wet scrub-
bing units. As stated earlier, the principal fluoride emission from this
operation is silicon tetrafluoride with some hydrogen fluoride. Spray
tower scrubbers are often used to absorb the fluoride compounds. Effi-
ciencies in the range of 90 to 991 are reported for new or well maintained
units depending upon the number of scrubbing stages used.
Recently designed normal superphosphate plants often utilize com-
merical scrubbing equipment of the wet cyclone or ejector venturi type.
The gaseous fluoride removal efficiency of these types of equipment is
generally in the 95 to 99% range. Recirculation is the universal practice
in ejector venturi installations. The control process mass balances are
presented as Figures 3-38 and 3-39.
Fluoride Emissions. Fluorides are evolved during all three proces-
sing steps in NSP manufacture. Table 3-51 shows typical evolution and
emission factors as indicated by RRI experience. The NSP process is simi-
lar to the TSP process in that fluorides are evolved while forming and
handling a solid product. Collection and removal of this fluoride in an
efficient, economical manner is especially difficult since the process
3-174

-------
1ST STAGE
EJECTOR
VENTURI
WATER TO ATM
SUPPLY
300 GPM
DUCT
SPRAY
SEPARATOR
TANK
OVER-
FLOW
700 GPM
700 GPM
RECIRCULATING PUMPS
OVERALL
EFF. = 97%
2ND STAGE
EJECTOR
VENTURI
PACKED
DEMISTER
DUCT
SPRAY
WATER
SUPPLY
140 GPM
2ND STAGE
CYCLONIC
1ST STAGE
CYCLONIC
110 GPM
30 GPM
SEAL TANK
rTZsi	 	1
X«7i SEAL TANK |
LIQUID EFFLUENT 300 GPM ^
PROCESS A
(1) REFLECTS OPTIMUM OPERATING AND MAINTENANCE CONDITIONS
LIQUID EFFLUENT 140 GPM
PROCESS B

BASIS - 40 TONS/BATCH (1 BATCH/HR) (20% PgOg)
PROCESS STREAMS-TONS/BATCH
TO ATM
<£>
-i
FAN
MATERIAL
Stream Number
4
5
6
7*
8
. 9*
SiF4(Equiv.)
HF
H2SiFg{Equiv.)
0.36(g)
0.4(g)
0.355(aq)
0.009(gfD)
0.363(aq)
0.003(gf0)
Total Fluorides
Total as F
0.36
0.26
0.04
0.03
0.355
0.281
0.009
0.009
0.363
0.287
0.003
0.003
*
Denning Fumes
Misc. Curing Fumes
13000(g)**
0.003(g)

13000(g)**
0.003(g) .

13000(g)**
0.003(g)
Approx. Total
Stream
13000**
o.oi"
0.355^)
13000** ^
0.363 (C)
13000*^A)
OVERALL(1)
EFF. = 99.25%
* Gaseous effluent streams	** SCFM
(A) Plus 14,000 cfm air carrier gas	(B) Plus 300 gpm scrubbing water
(C) Plus 140 gpm scrubbing water	(0) Reference 4267
(E) Estimate average device efficiency to be 90% with current maintainance and operation
techniques and fugitive emissions equal to 10% of evolution.
Overall soluble fluoride emission factor = 14.6 lb F/ton PgOg produced
(E)
Source
Soluble Fluoride Emission
Factor - lb F/ton P2O5 Produced
(Assuming optimum conditions)
Soluble Fluoride Emission
Factor - lb F/ton P2O5 Produced
(Assuming average industry operation)

Process A Process B
Process A Process B
Control Device Emission
2.3 0.8
CO
CO
Assumed Fugitive
Emission
0 0
7.3 7.3
Total Soluble
Fluoride Emission
2.3 0.8
14.6 14.6
Figure 3-38. Normal Superphosphate
Batch Production — Con-
trolled Process Model
3-175

-------
WATER
SUPPLY
140 GPM
TO ATM
WATER TO ATM
SUPPLY
300 GPM
1 ST STAGE
EJECTOR
VENTURi
DUCT
SPRAY
OVERALL v
EFF, = 97%
FAN
2ND STAGE
EJECTOR
VENTURI
PACKED
DEMISTER
2ND STAGE
CYCLONIC
1ST STAGE
CYCLONIC
t "m\
SEPARATOR
TANK
OVERALL
EFF. = 99.25%
DUCT
SPRAY
OVER-
FLOW
10 GPM
30 GPM
700 GPM
700 GPM
SEAL TANK
SEAL TANK
RECIRCULATING PUMPS
LIQUID EFFLUENT 140 GPM
PROCESS B
LIQUID EFFLUENT 300 GPM
PROCESS A
(1) REFLECTS OPTIMUM OPERATING AND MAINTENANCE CONDITIONS
BASIS - 200 TONS/DAY (20% PgOg)
PROCESS STREAMS-TONS/DAY

Stream Number
MATERIAL
5
6
7
8*
9
10*
SiF^ (Equiv.)
HF
1.68(g)
0.19(g)

0.04(g)^D)

0.01 (g)^
HgSiFg (Equiv.)


1.68(aq)

1.72(aq)

Total Fluorides
1.68
0.19
1.68
0.04
1.72
0.01
Total As F
1.23
0.14
1.33
0.04
1.36
0.01
Denning Fumes
2600(g)**


2600(g)**

2600(g)**
Misc. Curing Fumes

0.14

0.14

0.14
Approx.
Total Stream
2600**
0.33^
1.68(b)
2600(g)**(fi)
1.72(C)
13000**(A)
* Gaseous Effluent Streams	** SCFM
(A) Plus 14,000 cfm air carrier gas	(B) Plus 300 gpm scrubbing water
(C) Plus 140 gpm scrubbing water	(D) Reference 4267
(E) Estimate average device efficiency to be 90% with current maintainance and operation
techniques and fugitive emissions equal to 10% of evolution.
Source
Soluble Fluoride Emission
Factor - lb F/ton P2O5 Produced
(Assuming optimum conditions)
Soluble Fluoride Emission
Factor - lb F/ton P2O5 Produced
(Assuming average industry operation)

Process A Process B
Process A
Process B
Control Device Emission
2.0 0.5
6.9 .
6.9
Assumed Fugitive
Emissions
0 0
6.9
6.9
Total Soluble Fluoride
Emission
2.0 0.5
13.8
13.8
(E)
Overall soluble fluoride emission factor = 13.8 lb F/ton PgOg produced
Figure 3-39. Normal Superphosphate
Continuous Production -
Controlled Process Model
3-177

-------
Table 3-51. Evolution and Emission Factors for Manufacturing
Normal Superphosphate

Evolution
Emission

(pounds soluble fluorides as
(pounds soluble fluoride as
Source Point
fluorine/ton P205)
fluorine/ton P205
Mixer-Den
68
2
Storage
3
0.3 - 3
Total
71
2.3 - 5
involves many mechanical operations which are difficult to seal effectively.
These include den cutting, material transport, and product curing. Even
where fume collection and scrubbing is utilized, comparatively low collec-
tion efficiency and high air flow requirements can be anticipated if the
system leaks or is opened for appreciable periods of time. No information
has been found that provides a good basis for estimating leakage losses from
denning. The emission factors presented in Table 3-51 do not include such
losses. It is clear that improperly sealed or poorly maintained units might
involve multiples of the emission factors listed. Soluble fluoride emis-,
sions from NSP manufacture are estimated as equivalent to 5000 tons of F in
1970. For the year 2000, if current control practices continue, emissions
will be about 1400 tons. If 99% control efficiency is utilized emissions
in 2000 would drop to less than 100 tons (expressed as F).
3-5.5.5 Defluorination of Phosphate Rock
Process Description. Phosphate rock can be used as an animal or
poultry feed supplement if the fluoride content is reduced to an acceptable
level to prevent adverse biological effects. The product must be reduced
from about 3.5% fluorine to less than 0.2% fluorine. This reduction is
typically accomplished by thermal and/or chemical processing. The heating
processes utilize rotary kiln or fluidized bed heaters in which the rock is
mixed with additives such as phosphoric acid, silica, etc. to aid in
defluorination and heated to 2500°to 2900°F to drive off the fluorides.
3-179


-------
This process is shown in Figure 3-40. An alternate defluorination process
involves reacting hot phosphoric acid with limestone thereby precipitating
and volatilizing fluorides. No published descriptions of this process
have been found.
Production Trends. The amount of phosphate rock utilized as a feed
to the defluorination process in 1970 was 87,700 tons (as P2O5)^^. This
corresponded to a per capita consumption of 0.84 pound (as P2O5). If this
consumption level remains constant to the year 2000, there will be 141,200
tons of rock (as P205) utilized in the defluorination process.
Fluoride Emission Control Techniques. Since defluorination removes
more than 951 of the incoming fluorine, the exhaust gas stream from the
kiln or fluid bed reactor is quite concentrated in fluoride content. With
the majority of fluorides present as hydrogen fluoride and silicon tetra-
fluoride.this gas stream is vented via induced draft to a wet scrubbing
system. The high fluoride concentration results in highly corrosive condi-
tions in the exhaust gas ducting and collection equipment. The evolved
gaseous fluorides are removed by water scrubbing in multi-pass spray cham-
bers and spray towers in series with high energy wet scrubbers. Efficiencies
average 99.5% for the overall scrubbing system. Attainment of this high
efficiency level requires good basic system design, and prompt periodic
maintenance. A controlled process model is presented as Figure 3-41.
Fluoride Emissions. Fluoride evolution and emission values are
shown in Table 3-52. In the processes used for defluorination of rock, essen-
tially all of the input fluoride is volatilized. Highly effective collection/
scrubbing systems are employed to minimize emissions and mitigate an obvious
potential problem. It is also obvious that a small percentage decrease in
scrubber efficiency would result in multiplication of the emission factors.
Evolution from the production of dicalcium phosphate is poorly
defined and no firm data has been reported. Based on RRI experience, it is
estimated that 1760 tons of soluble fluorides were emitted by phosphate rock
3-180

-------
 o
PHOSPHORIC ACID +NaCI
FLUE GAS
PRODUCT TO
- STORAGE
OR SALE
MILL
CLINKER
STORAGE
GRINDING
KILN
RAW
MATERIAL
STORAGE
FUEL OIL
AND AIR
BASIS: 100 TONS/DAY
PRODUCT (39.9% P20g)
PROCESS STREAMS-TONS/DAY

STREAM NUMBER
MATERIAL
1
2
3*
4
HF


4.4(g)(C)

CaF2 (Equiv)
8.77(s)


,16(s)(B)
Total Fluoride
8.77

4.41
.16
Total As F
4.27

4.19
.08
Phosphate Rock
38% P205
92(s)
33(1)(A)


Phosphoric Acid
(15% P205)



NaCl

3(1)(A)


Product
P2°5



39.9(s)(B'
CaO



50.2(s)
Na20



4.6(s)
Insolubles



5.2(s)
h2o


33(g)

Combustion
Products


92( g)(es t ]

(15% excess air)




APPR0X. TOTAL
STREAM
101
36
129
100
* Gaseous effluent stream	'
(A)	Reference 5
(B)	Reference 506
(C)	Reference 4299
Soluble fluoride evolution factor = 210 lb F/ton P.0,. in product
(240 lb F/ton P^Oj in phosphate rock fed.)
Figure 3-40. Defluorination of Phosphate Rock -
Uncontrolled Process Model
3-181

-------
BASIS - 100 TONS/DAY OF PRODUCT (39.9% P^)
PROCESS STREAMS - TONS/DAY
TO STACK
SPRAY
WATER
WET
SPRAY
SCRUBBER
(2 STAGE)
EFF. = 98%
TO ATM.
(1)	A FEW PLANTS LIME THEIR GYPSUM PONDS
JO pH 4TO DECREASE HF EVOLUTION.
(2)	REFLECTS OPTIMUM OPERATING AND
MAINTENANCE CONDITIONS
Material -

Stream
Number

3
5 .
6*
7*
HF
4.41(g)'8'
4.32(1)
0.09(g)(C'
0.05(g)(Est)
Total Fluoride
4.41
4.32
0.09
0.05
Total as F
4.19
4.10
0.09
0.05
h2o
33(g)
33(1)


Hydrocarbon
Combustion
Products and
N2
92(g)(est)

92(g)

Approx. Total
Stream
129
37
92
0.05
* Gaseous Effluent Streams
(A)	Plus 1.8 x 10° gal scrubbing water
(B)	Reference 4299
(C)	Reference 4267.
(D)	Estimate average scrubber efficiency to be 90% with current
maintainance and operation techniques and fugitive emissions equal
to 5% of evolution.
Source
Soluble Fluoride Emission
Factor lb F/Ton P2O5 in
Product (assuming
Optimum Conditions)
Soluble Fluoride Emission
Factor lb F/Ton P2O5 in
Product (assuming
average Industry Operation)
Scrubber
4.5

21.0
Gypsum
Pond Emission
.2.5

2.5
Assumed Fugitive
Emission
0

10.5
Total Soluble
Fluoride
Emission
7.0

34.0
Overall soluble fluoride emission factor for defluorination	(D)
of phosphate rock (including gypsum ponds) = 34 lb F/ton P205
in product (39 lb F/ton P205 in phosphate rock fed)
Figure 3-41. Defluorination of Phosphate Rock — Controlled Process Model

-------
defluorination 1n 1970, gnd that 2730 tons of soluble fluorides (as F) will
be emitted in 2000. If 99% efficient control systems are employed, emission
for 2000 will drop to 170 tons.
Table 3-52. Evolution and Emission of Fluoride From
Defluorination of Phosphate Rock
Defluorination
Evolution
Emission
Processor
(pound fluoride/ton P2O5)
(pound fluoride/ton P2O5)
Thermal


Kiln
210
0.6
Fluid Bed
210
0.3
Chemical


DiCal Reactor*
~ ~
0.04
*defluorination of phosphoric acid
**not reported
3.5.5.6 Elemental Phosphorus Production
Industry Discussion. Elemental phosphorus is produced by electric
furnace smelting of phosphate rock with silica and coke. A flow chart for
a typical plant is shown in Figure 3-42. The phosphate rock is agglomerated
by sintering or nodulizing and, with the proper proportions of silica and
coke, fed to an electric furnace for smelting. The phosphorus produced is
volatilized and evolved from the furnace in a gas stream consisting mainly
of carbon monoxide. Particulates are removed using an electrostatic pre-
cipitator and recycled to the nodulizer, and the product phosphorus is
condensed in a direct contact scrubber/condensor. Iron present in nearly
all phosphatic rock is reduced to elemental iron which alloys with the
phosphorus generated and constitutes a ferrophosphorus byproduct. The
mineral constituents of the feed (calcium, silicon, etc.) form a molten
slag layer. Details of the process are available in Reference 4244.
3-183

-------
OFF GAS TO SCRUBBER
~
POSSIBLE
FUGITIVE
EMISSIONS
PHOSPHATE ROCK

•SILICA
AND COKE
NODULIZING
KILN
FUEL AND AIR
CARBON MONOXIDE
A
OFF-GAS STREAM
POSSIBLE
FUGITIVE
EMISSIONS STORAGE AND
T	FEED BINS

ELECTRICAL
POWER


TO STACK AND FLARE
A
ELECTRIC
FURNACE
SLAG_!_ i j
TAPPING N/
EMISSION
TO ATM 1
TO
NODULIZING
KILN
~ SLAG
mm
9 z
X Ui
a- q
 =7
go
± U

$
WATER
SPRAY
A
FERRO-
' PHOSPHORUS
LIQUID
PHOSPHORUS
TO STORAGE
"PHOSSY" WATER
TO RECYCLE
NOTE: ALTERNATE METHOD OF PROCESS OPERATION SHOWN AS DOTTED LINES (-
BASIS: ELECTRIC FURNACE PRODUCTION OF 30 TONS PHOSPHORUS PER DAY
PROCESS STREAMS (Tons/Day)
Materials
Stream Number
1
2
3
4
5
6
7*
8 >
9*
SiF4 (Equiv.)
CaF2 (Equiv.)
\
18.3(s)(B)

16.2(1)(B)


0.003(g)(A)
J.4(g)(B)

0.03(g)(A)
%
Total Fluoride
Total as F
18.3
8.9

16.2
7.9


0.003
0.002
1.4
1 .0

0.03
0.02
Phosphate Rock
Silica (Gravel)
Coke
Electrical Power (KW)
Slag
Ferro-Phosphorus
CO
co2
n2
Phosphorus
210(s)*D)
70(s)(D)
35(s)(D)
270(1)^
10(1)(B)
30(1)(C)
80(g)(B)
125(g)Est"
80(g)Est *
15,000KW(C)

Approximate Total
Stream
210
105
270
10
30
80
205
—
0.03
* Gaseous Effluent Stream
Soluble fluoride evolution factor = 68 lb F/ton P produced (30 lb F/Ton P2O5 equiv. in phosphate rock fed)
(A)	Reference 4299
(B)	Reference 4264
(C)	Reference 4244
(d) Reference 4263	Figure 3-42. Electrothermal Phosphorus
Production — Uncontrolled
Process Model
3-185

-------
Production. In 1968, 1.57 million tons of phosphate rock (expressed
		{4394)
as P205) were utilized in the production of electrothermal phosphorusv
for a per-capita consumption of 15.6 pounds per person. If the per-capita
consumption is assumed constant to the year 2000, the annual utilization of
phosphate rock in this process will be 2.62 million tons (as P205)• The
current emphasis on low and no-phosphate detergents could materially alter
these projections.
Fluoride Emission Control Techniques. Emphasis is placed on control
of F emissions from the feed preparation, condensor off-gas, and slag tapping
operations. The exhaust gas from feed preparation (nodulizing, etc.) contains
the major portion of fluorides evolved from phosphorus production. Substantial
emphasis has been placed on controlling this source of pollutant. The rock
pretreatment kilns are hooded and vented under induced draft to dust and
fluoride control equipment. Cooling prior to fluoride control is mandatory
due to the high temperature (2200° to 2600°F) of the pretreatment kiln.
Spray towers or wet cyclone collectors are the control devices most
commonly applied to the kiln gas stream. A relatively high concentration of
particulate material may also be present in the exhaust gas stream. This
particulate material is often collected using an electrostatic precipitator
prior to the application of gas scrubbers and returned to the feed preparation
process. Efficiencies of 96 to 991 have been achieved using the wet scrubbers.
The process of'condensing and removing the product phosphorus from
the furnace gas also absorbs most of the fluoride evolved in the furnace
operation. The water used in the condensing system is recycled and clarified
to remove dissolved phosphorus. A small amount of fresh water is added to
prevent the build-up of dissolved material.
The gases released during the slag tapping and cooling operation are
usually collected using a water cooled hood and ducted to wet scrubbers.
Fluoride emissions from this source depend upon the method used to quench the
slag after it is removed from the furnace. Water quenching generally pro-
duces approximately 6 times the amount of fluoride emissions as does air
3-187

-------
quenching of slag. In either case, the gas stream contains a low concen-
tration of gaseous fluoride. The control devices predominantly used are
simple water sprays or low pressure drop wet cyclones.
The attention given to the water pollution potential of the fluorides
collected in scrubbing liquids has given rise recently to consideration of
the use of fluorides recovered as by-product material from elemental phos-
phorus production. Recovery processes to produce such saleable by-products
as cryolite and aluminum fluoride from scrubbing liquid fluorides have
recently been described^6). Figure 3-43 presents a controlled process
mass balance.
Fluoride Emissions. The fluorides introduced in the phosphate rock
feed are evolved from three points in the process. These are feed prepara-
tion, condensor off-gas, and slag tapping. Volatilization of fluorides in
the electric furnace itself is not considered to be a major emission source
since the direct contact condensor will recapture most of the fluorides
evolved. Table 3-53 shows the quantities of fluorides evolved and emitted
from the three source points.
Feed preparation - drying, calcining, nodulizing, and cooling — is
clearly the major element of total fluoride emission. Sintering and nod-
ulizing of the rock feed requires heating to 2200 to 2600°F which results
in evolution of fluorides. The emissions from feed preparation are listed
as 5 to 59 pounds fluoride/ton of phosphorus since the degree of emission
control and capture efficiency for emissions from feed preparation vary
widely. Some facilities have no control and emissions will correspond to
evolution. Reasonably efficient scrubbing should reduce fluoride emission
by at least 90%, but feed preparation will still be the dominant emission
component.
The off-gas from the phosphorus condensor has been processed to
remove most of the particulate matter in the electrostatic precipitator
and most of the phosphorus and gaseous fluoride in the condensor. The
particulate matter is recycled to feed preparation. The condensor water is
3-188

-------
BASIS - 30 T0NS/0AY PHOSPHORUS PROOUCTION
PROCESS STREAMS - TONS/OAY
TO STACK
SPRAY

WATER .
A l\ <\
<2>,
WET
SPRAY
SCRUBBER
(2 STAGE)
EFF. = 99%

WATER AND 
H2S!F6 JV
TO TREATMENT AND DISPOSAL
(1) REFLECTS OPTIMUM OPERATING AND
MAINTENANCE CONDITIONS
Material
Stream Number
7
g.(E)
10
11*
SiF^ Equiv.
i.4(g)(c) .
0.03(g)(B)
1.39«"
O.Ol'0'
Total Fluorides
Total as F
1.4
1 .0
0.03
0.02
1.39
0.99
0.01
0.01
co2
N2
125(g)
80(g) .


125(g)
80(g)
Approx. Total
Stream
205
0.03
1.39(A)
205
* Gaseous Effluent Stream
(A)	Plus 128,000 gal scrubbing water
(B)	Reference 4299
(C)	Reference 4264
(0) Reference 4267
(E)	Slag Tapping Emission
(F)	Estimate average scrubber efficiency to be 90% with current
. maintainance and operation techniques and fugitive emissions
equal to 5% of evolution.
Source
Soluble Fluoride Emission
Factor - lb F/Ton P
Produced (assuming
optimum conditions)
Soluble Fluoride Emission
Factor - lb F/Ton P
Produced {assuming
average Industry Operation)
Scrubber
Effluent
0.7
6.8
Tapping
Emission
1.3
1.3
Assumed Fugitive
Emissions
0
3.4
Total Soluble
Fluoride
Emitted
2.0
11.5
Overall soluble fluoride emission factor c 11.5 lb F/Ton P produced^
(5.1 lb F/Ton P£05 equiv in phosphate rock fed)
Figure 3-43. Electrothermal Phosphorus Production —
Controlled Process Model
3-189

-------
Table 3-53.
Fluoride Evolution and Emission
From

Elemental Phosphorus


Evolution Factor
Emission Factor

(pound fluorine/
(pound fluorine/
Process Element
ton phosphorus)
ton phosphorus)
Feed Preparation
66.6
7 - 66.6
Condensor Off-Gas
0.1
0.1
Slag Tapping
_J.3
1.3
Total
68
00
00
also recycled arid may be treated to either recover fluoride values or prepare
it for disposal. Pond emissions are discussed in Section 3.5.5.7. The
off-gas is primarily carbon monoxide; it is typically either flared for
disposal or burned for its heating value in the kiln. The fluoride evolution
and emission factors for this gas are approximately 0.1 pounds fluorine/ton
phosphorus.
Most of the fluoride entering the furnace exits in the slag. Slag is
tapped periodically, and a small quantity of fluoride is evolved (about 1.3
pounds fluorine/ton phosphorus). A small minority of plants use a control
system comprised of a hood over the tapping port and a scrubber on the duct.
Using the evolution factor of 68 pounds of fluoride per ton of
phosphorus,(4157) anc| ta|
-------
3.5.5.7 Gypsum Pond Emissions
In the manufacture of wet-process phosphoric acid, the washed gypsum
filter cake with its entrapped acids and fluorides is res lurried with recy-
cled water and transported via a flume to a gypsum pond for disposal. The
various plant fluoride emission control scrubbers, which to a large extent
use recycled water, similarly discharge to the gypsum pond. In an integrated
plant, the wet emission control systems — for wet process H3PO4, triple
superphosphate, ammonium phosphates, defluorinated phosphate rock — typically
discharge to the gypsum pond. The pond is used for cooling, and for separa-
tion by settling of the solids, prior to reuse of the supernatant clarified
overflow water. Some plants may use a separate pond for cooling and
recirculation of process water.
Gypsum ponds have been observed to remain at a nearly constant fluoride
content of 5000 to 10,000 ppm fluoride for months without dumping of water
to disposal. It appears that an equilibrium exists between fluoride inputs
from the processes and fluoride outputs by volatilization, precipitation,
etc. The volatilization, which is of principal concern for this study, has
(3231
been measured by Cross and Rossv ' to be a minimum of 0.16 pounds/acre-day
for a 160-acre gypsum pond in a fertilizer complex producing sulfuric acid,
DAP, TSP, and phosphoric acid (500 tons P20,-/day). The water system also
included a 100-acre cooling pond, but emissions from it were not measured.
(43001
An investigation pursuing similar goals was performed by Taterav ' who
measured and correlated emissions from a model pond. For a water temperature
of 85°F and a wind velocity of 5 miles per hour (7.3 feet/second), Tatera's
emission values were 7 and 4.5 pound/acre-day for gypsum pond and process
water ponds, respectively. Based on these factors and assuming the same
ratio of pond area to P205 production as reported by Cross and Ross, the
total annual fluoride emission from gypsum ponds is 4520 tons fluoride/year
and from process water ponds is 1810 tons fluoride/year. The magnitude of
the emission factor estimated from evaporation rates^**^ is consistent
with the Tatera data, and the sizes and chemistry of the ponds for the
reported Florida complex should be representative.
3-191

-------
Emission control at or above the 99% level is readily attainable by
"liming" the ponds - using sufficient lime or limestone to react with the
soluble fluorides and acid. Currently, this represents the only practical
process for pond emission control.
Table 3-54 presents estimates of soluble fluoride emissions for the
phosphate rock processing industry. Total for 1970 is estimated at 6250 tons
(as F); if current practice continues, the total for 2000 would drop to 210
tons.
It should be noted that an appropriate portion of the total soluble
fluoride emissions from ponds has been allocated to each of the various
processes utilizing the pond as shown in Table 3-54,
3.5.5.8 Other Phosphate Processes
Furnace Grade Phosphoric Acid. Furnace grade phosphoric acid is
made from phosphate rock through the intermediate production of elemental
phosphorous. The fluoride content of the rock is evolved in the
manufacture of the phosphorus rather than in the manufacture of the acid.
Evolution of fluoride in the manufacture of phosphorus is discussed in
Section 3.5.5.6 and no appreciable further evolution of fluoride will
occur in making furnace acid.
Superphosphoric Acid. Manufacturing processes for production of
superphosphoric acid from wet phosphoric acid are essentially extensions of
vacuum concentration process techniques to the 70 percent P205 level, and
evolved fluorides are collected in barometric leg condenser water. Evolu-
tion of fluoride will occur from sumps and vents in the same manner as in
the other vacuum concentration processes. Based on RRI experience, the
evolution from sumps, vents, and tail gas in the superphosphoric acid
process will amount to approximately 1 pound of soluble fluoride (as F) per
ton of P205 produced. Scrubber efficiencies are at least 90%. Based on
these values and the production rate of 600,000 tons P205 per year, the
3-192

-------
Table 3-54. Gypsum Pond Emissions

Wet Process
Phosphoric Acid
Granulated
Triple Superphosphate
Diammonium
Phosphate
Defluorination of
Phosphate Rock
Industry^
Total
Soluble Fluoride
Emission Factor(C)
3.14
lb F/ton
0.2
lb F/ton P20g
0.009
lb F/ton P20g
2.9
lb F/ton P20g
1.6 to 1.8
lb F/ton PgOg

in Product
in Product
in Product
in Phosphate
Rock Feed
in Product
1970 Production
(106 tons/year)
3.8^
1.4
2>4(b)
0.09^
7.7
Projected 2000
Production
i3(b)
2.4^
7.0^
0.14^
22.9
(10® tons/year)





Soluble Fluoride
Emitted Currently
5.97
0.14
0.011
0.13
6.25
(103 tons F/year)





Soluble Fluoride
Emitted in 2000 with
Currently Used Controls
20.4
0.27
0.032
0.20
20.9^
(103 tons F/year)

-



Soluble Fluoride
Emitted in 2000 with
99% control
0.2
0.003
0.0003
0.002
0.21^
(103 tons F/year)





(A)	Expressed as equivalent in phosphate rock feed
(B)	Expressed as P20g equivalent in product
(C)	Assumes no lime addition to ponds and an emission equivalent to 0.013 lb F per lb. F feed to ponds
(D)	Reflects change due to shifts in production trends through 2000
(E)	Includes only processes which utilize ponds for disposal

-------
evolution and emission of fluoride from superphosphoric acid manufacture
are 310 tons of soluble fluorides as fluorine per year and 31 tons per year,
respectively.
Insoluble Dust Emissions* Evolution and emission of fluoride as
insoluble dust is of less importance than gaseous emissions because of the
lesser environmental effect produced. This difference results from the
relatively inert character of the insoluble solid fluorides, and the
tendency not to be transported out of the plant boundaries because of
particle fallout. The majority of dust emissions can be expected to be
phosphate rock since all rock undergoes size reduction, drying, and handling
in fine particulate form while many products do not. Available data^30^
indicate emission rates of about 1 pound dust/ton of product for grinding
and drying of rock. That is equivalent to 0.1 pound fluoride/ton P^Og
or a total annual emission of 410 tons fluoride/year.
3.5.6 Economic Analysis
3.5.6.1 Met Process Phosphoric Acid
Basic Processes. The economic analyses for wet process phosphoric acid
and the requisite captive sulfuric acid production are presented in Tables
3-55 and 3-56, respectively, for four typical plant sizes. If the equity
funding requirements for the captive sulfuric acid plants are included in
total equity funding requirements, return on investment for wet process
phosphoric acid plants varies between 4.5% (for the 100 ton per day plant)
and 44.0% (for the 1000 ton per day plant), assuming that all of the acid
produced 1s sold as merchant 75% phosphoric acid. This is almost never the
case. Most wet process HgPO^ plants sell only a small portion of their
output as merchang H^PO^; the bulk of the H^PO^ produced is employed as
captive acid in the manufactureof ammonium phosphates and simple superphos-
phate.
3-194

-------
Table 3.55.,
Estimated Economics of Wet Process Phosphoric Acid Production {54% P~CU
(Excluding Pollution Control Costs)	c 5


Plant Capacity (Tons P205)
Capital Investment
Installed Capital
Off Sites
(A)
Total Capital Investment
Total Capital Investment
Production Costs
(A)
;d)
Direct Costs
Phosphate Rock (3.48 tons of 66, bpl/ton P^Oj)
Sulfuric Acid

-------
Table 3-56. Estimated Economics of Sulfuric Acid Production
(Excluding Pollution Control Cost)
Plant Capacity (Tons 100% H^SO^)
Capital Investment
200 Tons/Day
600 Tons/Day
1200 Tons/Day
2000 Tons/Day
Installed Capital
0.85$MM
1.90$MM
3.06$MM
4.15$MM
Off Sites
0.26
0.57
0.92
1.25
Total Capital Investment
1.11$MM
2.47$MM
3.98$MM
5.40$MM
Production Costs




Direct Costs




Sulfur (0.3 long tons s/ton 100% H?S0.
S at $21/1ong ton) '
Water (6 M-gallons/ton 100% H?S0.)
Electric Power (8kwh/ton 100% ^§0^
Operating Labor (2 men/shift)
Supervision and Fringe Benefits
Maintenance and Supplies
6.36$/ton
0J8 ioo% h2so4
0.06
0.96
0.96
1.01
6.36$/ton
0.18 100% H2S04
0.06
0.32
0.32
0.75
6.36$/ton
0.18 100% H2S04
0.06
0.16
0.16
0.60
6.36$/ton
0.18 100% H2
0.06
0.10
0.10
0.49
Total Direct Cost (B)
9.53
7.99
7.52
7.29
Indirect Costs




Depreciation
1.68
1.25
1.01
0.82
Local Taxes and Insurance
°-50
0.37
0.30
0.25
Plant and Labor Overhead
1.15
0.38
0.19
0.12
Total Indirect Cost
3.33
2.00
1.50
1.19
Manufacturing Cost ($/Ton 100% f^SO^)
12.86
9.99
9.02
8.48
(A)	Total H^SO^ production assumed to be captive to wet process HgPO^ usage in-house.
(B)	Steam credit not included (valued at 1.1 tons of high pressure steam/ton 100% ^sO^).

-------
Impact of Control. Analyses of the cost of fluoride emission control
using each of the three processes currently employed by the Industry are
presented in Tables 3-57, 3-58, and 3-59. The decrease in ROI is nominal
in each case.
3.5.6.2	Piammonium Phosphate
Basic Processes. Table 3-60 presents economic analyses of two typical
sizes of DAP plant. ROI varies from 15.5% at the smaller plant size to
19.1% for the 1000 ton per day unit.
Impact of Control. Table 3-61 presents an analysis of the cost of
fluoride emission control for DAP plants. aROI's are about -2%.
3.5.6.3	Triple Superphosphate
Basic Process. Table 3-62 presents the economics of production for
granular triple superphosphate at the indicated sales price. ROI's for the
four plant sizes chosen ranges from 6.0 to 18.0%.
Impact of Control. Table 3-63 presents the analyses of the cost of
control of fluoride emissions from TSP manufacture, using current techniques.
Change in Return on Investmend due to emission control for a 600 ton per day
TSP plant is approximately 14%.
3.5.6.4	Normal Superphosphate
Basic Process. Table 3-64 presents an economic analysis of production
costs and ROI's for three typical plant si zes. It should be noted that for
plant sizes below 200 tons/day (and many of the older plants are in this
range), operations fall below the break-even point and are profitable only
if equipment has been fully depreciated.
Impact of Control. Tables 3-65 through 3-68 present emission control
economics for continuous and batch NSP plants.
3-197

-------
Table 3-57. Wet Process Phosphoric Acid - Estimated Economics of Control Process A
Basis - 1000 Tons Per Day of Phosphoric Acid (P2O5 basis) Produced.
Item
Number
BCagUaJ>
-------
Table 3-58. Wet Process Phosphoric Acid - Estimated Economics of Control Process B
Basis - 1000 Tons Per Day of Phosphoric Acid (P2°5 basis) Produced
Ogerati£q_Cost_J$/h£
Item
Number
Description
tquipment
F.U.B.
Cost
Reference
Number
Installation
Factor

tquipment
Installation
Cost

Item
Number
Power
Cost
Maintenance
Cost

tquipment
Operating
. Cost
1
VENTURI SCRUBBER, 85,000 cfm,
714 gpm, neoprene lined steel,
31.5 in W.G. pressure drop, 564
horsepower
112
4383
4390
4391
1.73

194

1
2
2.94
0.13
0.38

3.07
0.38
2
GYPSUM POND^17)




50














Subtota


3.45








Water^21' (714 gpm, 90% recycle)
0.09




Capital Subtotal 244
Indirects (9 15%) 37
Contingency (@ 20%) 4q

Disposal

-


Total Capital (as of January 1971) 330

Total Operating Cost
3.54
Total Operating Cost ($/hr)	3.54
Taxes and Insurance (2%, 330 days)	0.83
aii * i	ft*	i * j • <- *• i i ,	Capital (9.1%, 330 working days)	-3-r»
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.	Pollution Control Cost ($/hr)	8.16
Pollution Control Cost ($/ton)	0120

-------
Table 3-59. Wet Process Phosphoric Acid - Estimated Economics of Control Process C
Basis - 1000 Tons Per Day of Phosphoric Acid (P2O5 basis) Produced
Item
Number
CO
1
no
o
o
_Ca£UaJ_Cost_Es^imates__[£10002i
Description
SPRAY CROSS - FLOW PACKED SCRUBBER,
2 at 10'-6" X 10'x12', 2 in W.G.
pressure drop, 85,000 cfm, neoprene
lined steel, 8 ft/sec velocity, 30 hp
GYPSUM POND^17)
Equipment
F.O.B.
Cost
39
Reference
Number
4383
4391
4392
Installation
Factor
2.56
"TquTpmenT"-
Installation
Cost	
100
50
Capital Subtotal	150
Indirects (0 15%)	23
Contingency (@ 20%')	30_
Total Capital (as of January 1971)	203
0£eratuiq_Cost_|$^Jii^^_
Item
Number
Power
Cost
0.16
Maintenance
Cost
0.50
0.38
Subtotal
Equipment
Operating
Cost
0.66
0.38
T21)
Water
Disposal
(22)
1 .04
(3000 gpm, 90% recycle) 0.36
Total Operating Cost
1.40
All control economics footnotes are located
in Section 3.1.1 , pages 3-10 and 3-11.
Total Operating Cost ($/hr)	1-40
Taxes and Insurance (2%, 330 days)	0.51
Capital 19.1%, 330 working days)	2.33
Pollution Control Cost ($/hr)	4.24
Pollution Control Cost ($/ton)	0.10

-------
Table 3-60. Estimated Economics of Diammonium Phosphate (18-46-0) Production
(Excluding Pollution Control Cost)

Plant Capacity

800 tons/day 1000 tons/day
Installed Capital
1.52 $MM
1.67 $MM
Off Sites
0.30
0.33
Total Invested Capital^
1.82
2.00
Production Costs


Direct Costs
Phosphoric Acid (0.87 tons 54% P,0c/ton)
(o \ ^ 3
Ammoniav (0.23 Tons/ton)
Electric Power (20 kwh/ton)
Fuel Oil (3 gallon/ton)
Operating Labor (2 men/shift)
Supervision and Fringe Benefits
Maintenance and Supplies
31.40 $/Ton
30.39 $/Ton
13.80
0.14
0.24
0.25
0.25-
0.41
13.80
0.14
0.24
0.20
0.20
0.36
Total Direct Cost
46.49
45.33
Indirect Costs


Depreciation
0.69
0.61
Interest (at 7%, 20% debt)
0.10
0.08
Taxes and Insurance
0 .21
0.81
Plant and Labor Overhead
O
CO
o
0.24
Total Indirect Cost
1 .30
1 .11
Manufacturing Cost
47.79
46.44
General and Sales Expenses
0 .96
0.93
F.o.b Cost ($/ton)
48.75
47.37
Product Revenue ($/ton 18-46-0)
60.00
60.00
Profit After Taxes (at 50%, $/ton)
5 .63
6.32
Cash Flow ($MM/yiear)
1 .67 $MM/yr
2.28 $MM/yr
Return on Investment (C)
15.5%
19.1%
(A) Wet Process Phosphoric Acid Plant co-located.
B]	Assumes 95% recovery.
C)	Assumes 80% equity funding capital outlay for Wet Process Phosphorus acid plant not
shown but allocated for calculating ROI.
3-201

-------
Table 3-61. Diammonium Phosphate Production - Estimated Economics of Control Process
Basis - 500 Tons Per Day of 18-46-0 Diammonium Phosphate Produced
Item
Number
Description
tquipment
F.O.B.
Cost
Reference
Number
Installation
Factor

tquipment
Installation
Cost
1
AMMONIA SCRUBBER, 3' diameter by
8', 2000 cfm, 2 in. W.G. pressure
drop, neoprene lined steel
4
4387
4390
4391
2.50

10
2
TAIL GAS SPRAY SCRUBBER, (Cross
flow), 6' by 2' by 4', 2000 cfm,
W.G. pressure drop,neoprene lined
3
4387
4390
4391
3.00

9
3
DUST SCRUBBER, 3' diameter by 8',
1000 cfm, 2 in. W.G. pressure drop,
neoprene lined steel
4
4387
4390
4391
2.50

10
4
RECYCLE TANK, 10,000 gal, neoprene
lined steel
6
4383
2.50

15




Capital Subtotal
44




Indirects (0 15*)
7




Contingency (@ 20i ) 9


Total Capital (as of January 1971) gg

All control economics footnotes are
located in Section
3.1.1, pages
3-10 and 3
-11.
Item
Number
Power
Cost
Maintenance
Cost

tquipment
Operating
Cost
1
0.02
0.25

0.27
2
0.02
0.25

0.27
3
0.01
0.25

0.26
4

0.06

0.06
Subtotal
Water'21^ (
Disposal'22^
170 gpm, 90% recycle
0.86
) 0.02
Total Operating Cost
0.88
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 330 days)
Caoital (9.12, 330 working days)
Pollution Control Cost ($/hr)
Pollution Control Cost ($/ton)
0.88
0.15
0.69
1.72
0.08

-------
Table 3-62. Estimated Economics of Granular Triple Superphosphate Production (0-46-0)
(Excluding Pollution Control Cost)
Plant Capacity

300 Tons/Day
600 Tons/Day
800 Tons/Day
1000 Tons/Day
Installed Capital
Off Site
1.19 $MM
0.24
1.66 $MM
0.33
1.90 $MM
0.38
2.14 $MM
0.43
Total Capital Investment
1.43
1.99
2.28
2.57
Production Costs
Direct Costs
Phosphoric Acid
(0.648 tons 542 P20c/ton (0-46-0)
Phosphate Rock (75 bpl at
0.393 tons/ton (0-46-0)
Electrical Power (7 Kwh/ton at 0.007 $/Kwh)
Fuel Oil (3 gal/ton)
Rocking Grinding and Handling
Operating Labor (2 men per shift)
Maintenance and Supplies
22.83 $/Ton(D)
4.01
0.05
0.18
0.68
0.68
0.87
22.83 $/Ton(D)
4.01
0.05
0.18
0.68
0.34
0.60
21.17 $/Ton^E)
4.01
0.05
0.18
0.68
0.25
0.52
21.17 $/Ton(E)
4.01
0.05
0.18
0.68
0.20
0.47
Total Direct Costs
29.98
29.03
27.11
26.96
Indirect Costs




Depreciation
Interest (at 7%, 20% debt) .
Taxes and Insurance
Plant and Labor Overhead
1.44
0.20
0.44
0.82
1.01
0.14
0.30
0.41
0.86
0.12
0.26
0.30
0.78
0.11
0.24
0.24
Total Indirect
2.90
1.86
1.54
1.37
Manufacturing Cost ^
32.88
30.89
28.65
28.33
General and Sales Expenses
0.66
0.62
50.57
0.57
F.o.b. Cost
33.54
31.51
29.22
28.90
Product Revenue
38.00
38.00
38.00
38.00
Profit after Taxes (at 50%)
2.23 $/Ton
3.25 $/Ton
4.39 $/Ton
4.55 $/Ton
Cash Flow
0.36 $MM/Yr.
0.84 $MM/Yr.
1.39 $MM/Yr.
1.76 $MM/Yr.
Return on Investment^'^(%)
6%
10%
16.6%
17.8%
CO
i
ro
o
OJ
(A) A pulverized superphosphate plate (for use in amnoniated fertilizers) requires an initial capital investment approximately equal to that of a
Granular triple superphosphate plant of equal capacity. This cost differential (approximately 2.50 $/Ton) is considerably reduced when local
air pollution regulations require an extensive gas scrubbing system due to the increased cost of "cleaning up" a granular plant.
Includes capital for wet process phosphoric acid facility-co located.
(B)
(C)	Assumes 80% equity funding.
(D)	"
(E)
Based on H,P0. manufacturing cost for 400 NT (P205)/day wet process H^PO. plant.
Based on H^PO^ manufacturing cost for 700 NT ^OgJ/day wet process H^PO^ plant.

-------
CjO
r>o
o
4^
Table 3-63. Granulated Triple Super Phosphate - Estimated Economics of Control Process
Basis - 200 Tons Per Day of Triple Superphosphate (48% P2O5) Produced
Item
Number
Description
Equipment
F.O.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost
1
CYCLONIC SPRAY TOWER, 23,000 cfm,
8 ft/sec allowable velocity, 4 ft
diameter by 10 ft, neoprene lined
steel, 3 in W.G. pressure drop, 15 hp
20
4386
4383
4387
2.15

43
2
GYPSUM POND'16)
-
-
-

50
3
SPRAY SCRUBBER, 20,000 cfm, 8 ft/sec
allowable velocity, 2 at 4 ft
diameter by 8 ft, neoprene lined
steel, 2 in W.G. pressure drop, 17 hp
18
4387
4390
4391
2.55

46




Capital Subtotal
139
-



Indirects (P15%)
21




Contingency (@20%) 28


Total Capital (as of January 1971) 188
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
Operating Cost ($ / hr )
Item
Number
Power
Maintenance
Cost
Cost
0.08
0.13
-
0.38
0.09
0.25
Subtotal
Equipment
Operating
Cost	
0.21
0.38
0.34
121)
Water
Disposal
(22)
0.93
( 645 gpm, 90% recycle )0.08
Total Operating Cost
1.01
Total Operating Cost ($/hr)	J-01
Taxes and Insurance (2%, 330 days)	0.47
Capital (9.1%, 330 working days)	2.16
Pollution Control Cost ($/hr)	3.64
Pollution Control Cost ($/ton)	0.44

-------
Table 3-64. Estimated Economics of Normal Superphosphate (0-20-0) - Continuous
Production (Excluding Pollution Control Cost)
Plant Capacity
E00 Tons/Day
350 Tons/Day
500 Tons/Day
u>
i
r\j
o
tn
Capital Investments
Installed Capital
Off Sites
Total Capital Investment
Production Costs
Direct Cost
Phosphate rock^ (75 bpl, 0.60 tons/ton product)
Sulfuric Acid (0.357 tons 100% H2S04/ton product)
Electric Power (20.8 kwh/ton product)
Operating Labor (2 men/shift)
Supervision and fringe benefits
Maintenance and Supplies
Total Direct
Indirect Costs
Depredation (7.5 year, straight line)
Interest (at 7%, 20S debt)
Taxes and Insurance
Plant and Labor Overhead
Total Indirect
Manufacturing Cost ($/ton product)
General and Sales Expenses ($/ton product)
F.o.b. Cost (S/ton product)
Product Revenue ($/ton product)
Profit after Taxes (Taxes at 501)
Cash Flow ($/year)
Return on Investment
(b)
(a)
(b)
Includes $0.54/ton grinding
Assumes 30% equity funding
0.48 $MM
0.14
0.62
5.94 $/Ton Product
4.60
0.15
0.96
0.96
0.74
13.35
1.24
0.13
0.37
1.15
2.89
16.24
0.32
16.56
17.00
0.22
0.10
0.68 $MM
0.20
0.88
5.94 $/Ton Product
4.60
0.15
0.55
0.55
0.60
12.39
1.01
o.n
0.30
0.66
2.08
14.47
0.29
14.76
17.00
1.12
0.25
18.4*
0.87 $MM
0.26
1.13
5.94 $/Ton Product
4.60
0.15
0.38
0.38
0.54
11.99
0.91
0.10
0.27
0.46
1.74
13.73
0.27
14.00
17.00
1.50
0.40 SMM
27.4S

-------
Table 3-65.
Continuous Normal Super Phosphate Production - Estimated Economics of Control Process A
Basis - 200 Tons Per Day of Normal Superphosphate (20% P2°5) Produced
00
I
ro
o
CD
Item
Number
_Ca£i^aJ_Cost_Es^imates_[£1000]i
Description
VENTURI SCRUBBER -
SEPARATOR TANK, 2 at 35,500 cfm,
300 gpm, neoprene lined steel, 33 1n.
W.G. pressure drop.
tquipment
F.O.B.
Cost
142
Reference
Number
4386
4387
4390
4383
Installation
Factor
1.72
tquipment
Installation
Cost	
244
Capital Subtotal	244
Indirects (0 15%)	37
Contingency (0 20%)
Total Capital (as of January 1971)	330
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
Operating Cost (Sl/hr )
Item
Number
Power
Cost
2.58
Maintenance
Cost
0.25
tquipment
Operating
Cost
2.83
Subtotal
121)
Water
Di sposal
(22)
2.83
(300 gpm, 90% recycle ) 0.04
Total Operating Cost
2.87
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 330 days)
Capital (9.1%» 330 working days)
Pollution Control Cost ($/hr)
Pollution Control Cost ($/ton)
2.87
0.83
3.79
7749
0.90

-------
Table 3-66. Continuous Normal Superphosphate Production - Estimated Economics of Control Process B
Basis - 200 Tons Per Day of Normal Superphosphate (20% P2°5) Produced
OgenitijTg_CostiJ£]JQ]£i
Item
Number
Description
Ca£HaJ_Cost_Estimates_JJ10002
"tqulpment"
CYCLONIC SEPARATORS,
2 at 140 gpm, 35,500 cfm, 5 in.
W.G. pressure drop,neoprene lined
steel
F.U.B.
Cost
58
Reference
Number
4383
Installation
Factor
2.60
Equipment
Installation
Cost	
151
Capital Subtotal	151
Indirects (0 15%)	23
Contingency (0 20%)	30_
Total Capital (as of January 1971)	204
Item
Number

Power
Cost
0.40
Maintenance
Cost
0.25
Subtotal
Water
Disposal
tqulpment
Operating
Cost
0.65
i	0.65
(140 gpm, 90% recycle ) 0.02
.,,(22)
Total Operating Cost
0.67
All control economics footnotes are located
in Section 3.1.1, pages 3-10 and 3-11.
Total Operating Cost ($/hr)	0-67
Taxes and Insurance (2%, 330 days)	0-52
Capital (9.1%, 330 working days)	—2.34
Pollution Control Cost ($/hr)	3.51
Pollution Control Cost ($/ton)	0.42

-------
Table 3-67. Batch Normal Superphosphate Production - Estimated Economics of Control Process A
Basis - 40 Tons Per Batch (1 batch per hour) of Normal Superphosphate (20% P0O5)
Produced
_Ca£itail_Cos^_Estimates_J^1000]i
Item
Number
Description
VENTURI SCRUBBER -
SEPARATOR TANK, 2 at 70,000 cfm,
300 gpm, neoprene lined steel, 33 in.
W.G. pressure drop.
EquTpmenT
F.O.B.
Cost
238
Reference
Number
4386
4387
4390
4383
Installation
Factor
1.73
fcqui pment
Installation
Cost	
412
Capital Subtotal	412
Indirects (P15%)	63
Contingency {0 20% ).	82
Total Capital (as of January 1971)	557
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
Operating Cost
Item
Number

JL
Power
Cost
5.15
Maintenance
Cost
0.25
tquipmenF
Operating
Cost
5.40
Subtotal
(21)
Water
Disposal
(22)
(300 gpm, 90% recycle
5.40
) 0.04
Total Operating Cost
5.44
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 330 days)
Capital (9.1%, 330 working days)
Pollution Control Cost ($/hr)
Pollution Control Cost ($/ton)
5.44
1.41
6.40
13.25
0.33

-------
Table 3-68. Batch Normal Superphosphate Production - Estimated Economics of Control process B
Basis - 40 Tons Per Batch (1 batch per hour) of Normal Superphosphate (20% P2°s) Produced
Item
Number
to
i
ro
o
"vo
Description
Ca£Ha2_Cost_Estimates_£J>10002
"fcquipment'
F.O.B.
CYCLONIC SEPARATORS,
2 at 140 gpm, 70,000 cfm, 5 in. W.G.
pressure drop, neoprene lined steel.
Cost
88
Reference
Number
4383
Installation
Factor
2.42
"TqiiTpmenT"™
Installation
Cost	
213
Capital Subtotal	213
Indirects (0 15%) 32
Contingency (@ 20% ) 43_
Total Capital (as of January 1971)	288
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
)£eratuiqCost_|$J^£____2
Item
Number
Power
Cost
0.78
Maintenance
Cost
0.25
tqui pment
Operating
Cost
1.03
Subtotal
121)
Water
Disposal
(22)
1.03
(140 gpm, 90% recycle ) 0.02
Total Operating Cost
1.05
Total Operating Cost ($/hr)	1.05
Taxes and Insurance (2%, 330 days)	0.73
Capital (9.1%, 330 working days)	3.31
Pollution Control Cost ($/hr)	5"iw^
Pollution Control Cost ($/ton)	q 13

-------
3.5.6.5 Defluorination of Phosphate Rock
Basic Process. Table 3-69 presents the analysis of production costs
and investment requirements for the defluorination of phosphate rock. Return
on investment is estimated as 24.0%.
Impact of Control. Table 3-70 presents an analysis of the costs of
currently employed fluoride emission control techniques. Impact on ROI
(aROI) is estimated as -7%.
3.5.6.6 Elemental Phosphorus
Basic Process. Table 3-71 presents the analyses of the costs and
returns on investment (excluding pollution control costs) for integrated
"white" phosphoric acid - elemental phosphorus plants at four levels of
production - 100, 250, 450 and 600 tons of P20g per day. Return on investment
ranges from 18.6 to 50.7%.
Impact of Control. Costs of control of emissions from elemental
phosphorus plants are presented in Table 3-72. Impact on ROI is minimal.
3-210

-------
Table 3-69. Estimated Economics of Defluorlnated Phosphate Rock Production
(Pollution Control Cost Excluded)

PLANT CAPACITY

100 Tons Product/Day
Total Capital Investment
2.2
$MM
Production Costs
Direct Costs
Phosphate Rock (75 b.p.l., 0.92 tons/ton product 9 $9.20/ton)
Phosphoric Acid (0.33 tons 15% P20c/ton product @ $15.00/ton)
Sodium Chloride (0.03 tons/ton product @ $28.60/ton)
Fuel Oil (8 gallon/ton product @ $0.10/gallon)
Labor (4 positions @ $4.00/hr)
Fringe Benefits and Supervision
Maintenance and supplies @ 6%)
8.46
4.95
0.86
0.80
3.84
3.84
4.00
$/net ton product
Total Direct Costs
26.75

Indirect Costs
Depreciation (P10%)
Interest (at 7%, 20% debt)
Local Taxes and Insurance (@3%)
Plant and Labor Overhead
6.67
0.93
2.00
4.61

Total Indirect Costs
14.21

Total Average Costs ($/net ton product)
40.96
$/net ton
General and Sales Expenses ($/net ton product)
0.82
$/net ton
F.o.b. Cost ($/net ton product)
41.78
$/net ton
Average Product Revenue ($/net ton product)
70.00
$/net ton
Average Profit after Taxes (@ 50%)
14.11
$/net ton
Cash Flow ($MM/yr)
0.64
MM$/.yr
Return on Investment (%)^
24.0%

^Assumes 80% equity funding

-------
Table 3-70. Defluorination of Phosphate Rock - Estimated Economics of Control Process
Basis ^ 100.Tons Per Day of Product (39% P2O5)
to
IN3
rv5
Item
Number
_Ca£HaJ_Cos^_Est1mates_^10002
Description
SPRAY SCRUBBER, 2 STAGES
4 ft diameter by 10 ft,
monel clad, 10,000 cfm,
8 ft/sec allowable velocity,
6 lb HF(g)/min, 2 in. W.G.
pressure drop, 4 horsepower
GYPSUM POND*17)
Equipment
F.O.B.
Cost
13.
Reference
Number
4383
4391
4392
Installation
Factor
1.77
"TqliTpmenT™""
Installation
Cost	
23.
50.
Capital Subtotal	73
Indirects (P 15%)	11
Contingency (@20%)	15_
Total Capital (as of January 1971)	99
All control economics footnotes are located in Section 3.1.1 , pages 3-10 and 3-11.

Item
Number
1.
2.
Power
Cost
0.02
Maintenance
Cost
0.25
0.38
Subtotal
121]
TquTpmen^
Operating
Cost
0.27
0.38
Water
Disposal
(22)
217 gram, 90% recycle)
Total Operating Cost
0.65
0.03
0.68
Total Operating Cost ($/hr)
Taxes and Insurance (2%, 300 days)
Capital (6.7%, 300 working days)
Pollution Control Cost ($/hr)
Pollution Control Cost ($/ton)
0.68
0.28
0.92
1.88
0.45

-------
Table 3-71. Estimated Economics of Phosphoric Acid Production by Electric Furnace
(Pollution Control Costs Excluded)
100 Tons/Day
Plant Capacity (Tons P2O5)
250 Tons/Day
450 Tons/Day
600 Tons/Day
Capital Investment
Installed Capital^
Off Sites
Total Capital Investment
Production Costs
4.70 $MM
0.94
5.64
11.04 $MM
2.21
13.25
18.21 $MM
3.64
21.85
23.50 $MM
4.70
28.20
Direct Costs
CO
1
PO
00
Phosphate Rock (66 bpl 3.82^ ^tons/ton P?0,-)
Silica (Gravel at 1.27 tons/ton P2O5)
Coke(t) (0.60 tons/ton P2O5)
Lime (0.0133 tons/tor P2O5)
Water (31 M-gal/ton P9O5)
Electric Power (6220 Rwh/ton P2O5)
Steam (2.75 M-lbs/ton P2O5)
Electrodes (13 lbs/ton P2O5)
Fuel Oil (50 gal/ton P2O5)
Operating Labor (22 men per shift)
Maintenance and Supplies
Supervision and Fringe Benefits
24.83 $/Ton P-0,-
3.81	c 3
15.00
0.23
1.55
43.54
1.90
3.25
3.00
21.12
10.25
21.12
24.83 $/Ton P-0,
3.81	* 5
15.00
0.23
1.55
43.54
1.90
3.25
3.00
8.45
9.64-
8.45
24.83 $/Ton P,0,
3.81	* 3
15.00
0.23
1.55
43.54
1.90
3.25
3.00
4.69
8.83
4.69
24.83 $/Ton P-0,
3.81	i 5
15.00
0.23
1.55
43.54
1.90
3.25
3.00
3.52
8.55
3.52
Total Direct Costs
149.60
123.65
115.32
112.68

-------
Table 3-71. Estimated Economics of Phosphoric Acid Production by Electric Furnace (Continued)
Plant Capacity (Tons P2O5)
100 Tons/Day	250 Tons/Day	450 Tons/Day	600 Tons/Day
Indirect Costs
Depreciation
Interest (at 7%, 20% debt)
Taxes and Insurance
Plant and Labor Overhead
Total Indirect
Manufacturing Cost ($/ton P2O5)
General and Sales Expenses ($/ton P2O5)
F.o.b. Cost ($/ton P205)
Product Pvevenue {$/ton P20,j)^
Profit After Taxes (taxes at 50%, $/ton P2O5)
Cash Flow ($MM/yr)
Return on Investment^
(A)	100 Tons PgOs/Day plant requires a single 25,000 kva furnace; the 250, 450 and 600 tons P-Oc/Day plants require multiple furnace units
Of 50,000 Eva.
(B)	Assuming 90% overall recovery
(C)	Assumes 86% fixed carbon, at $15.00 per ton FOB, plus $10.00 per ton freight.
(D)	Revenue calculated by assuming a value of 139 $/ton 75% H.jP04 commercial grade acid.
(E)	Assumes 80% equity funding
17.09
2.39
5.13
26.88
51.44
201.09
4.02
205.11
256.00
25.45	$/ton P205
1.40	$MM/yr
18.6%
16.06
2.25
4.82
10.74
33.87
157.52
3.15
160.67
256.00
47.67 $/ton P205
5.26	$MM/yr
37.1%
14.71
2.06
4.42
5.98
27.17
142.49
2.85
145.34
256.00
55.33 $/ton P20g
10.40 $MM/yr
47.1%
14.24
1.99
4.28
4.48
24.99
137.67
2.75
140.42
256.00
57.79 $/ton P^
14.26 $MM/yr
50.7%

-------
Table 3-72. Electrothermal Phosphorus Production - Estimated Economics of Control Process
Basis - 30 Tons of Phosphorus Produced Per Day
_£a£HaJ_Cost_£stinTates_^10002
Item
Number
Description
SPRAY SCRUBBER, 5'6" diameter by
13', 90 gpm water, monel clad, 5000
cfm, 2 in. W.G. pressure drop
Equipment
F.O.B.
Cost
20
Reference
Number
4387
4390
4391
Installation
Factor
1.77
Equipment
Instal lation
Cost	
36
Capital Subtotal
Indirects (0 15%)
Contingency (@ 20%)
Total Capital (as of January 1971)
36
5
48
0£era£uiqCost_J$j/iiijj£
Item
Number
Power
Cost
0.02
Maintenance
Cost
0.25
tqulpment
Operating
Cost
0.27
Subtotal	o.?7
T21)
Waterv	(90 gal min, 90% recycle) 0.02
Disposal	_
Total Operating Cost
0.29
All control economics footnotes are located
in Section3.1.1,pages 3-10 and 3-11.
Total Operating Cost ($/hr)	0-29
Taxes and Insurance (2%, 330 days)	0.12
Capital (9.1%, 330 working days)	—0.55
Pollution Control Cost ($/hr)	o.96
Pollution Control Cost ($/ton)	o177

-------
3.6 GLASS MANUFACTURE

-------
3.6 GLASS MANUFACTURE
¦3.6.1 General
The glasses are manufactured by melting sand, limestone, soda ash
and cullet (broken glass scrap) in a furnace. Minor constituents are
added such as fining, oxidizing, coloring, and bleaching agents. Variations
in the feed materials and compositions enable production of hundreds of
product glasses, tailored to specific uses. One variety--soda-lime glass-
accounts for 90% of U.S. production.
Because of the opacity in product glass caused by the presence of
fluorspar (CaF,,) as a component or impurity in the furnace feed, CaFg has
been minimized as a component in soda-lime glass for containers and in
most other glass compositions. Only one glass composition—opal glass-
makes deliberate use of fluorspar as a feed material in order to produce
a translucent glass. The manufacture of opal glass is the major glass
industry source of fluoride evolution and emission in the U.S.
3.6.2	Industry Description
Large direct fired furnaces are used for the continuous production
of glass. Figure 3-43 contains a flow diagram and mass balance for a typical
60-ton-per-day, continuous-production opal glass furnace. Since fluoride
evolution takes place only while the glass is molten, the portion of the
plant involved in the manufacture of finished glass products is not shown.
Fluorides are evolved from the glass furnace as both gaseous and
particulate species. Over half of the fluoride charge is volatilized as
soluble fluorides, i.e., gaseous HF, solid and gaseous NaF, and gaseous
BOP.
3.6.3	Production Trends
There are currently in the U.S. five major furnace installations
producing opal glass; their individual capacities range from 60 to 70 tons
per day. No estimates of future production are currently available. Since
there are numerous types of opal glass with varying fluoride contents, pre-
dictions are based on the amount of fluorspar consumed for manufacture of
opal glass.
3-217

-------
BASIS - 60 TONS/DAY OF OPAL GLASS PR30UCTI0N
PROCESS STREAMS - LB/HR
TO ATMOSPHERE
DUST
LOST TO ATMOSPHERE
DURING TRANSPORT
OF MATERIAL
GAS TEMPERATURE = ?00°F
30 FEET UP THE STACK f
STACK
BATCH
STORAGE
BIN
RAW MATERIALS
FUEL AIR
MOLTEN GLASS
WEIGH
HOPPER
GLASS
MELTING
FURNACE
2800-F
rotary bucket
MIXER	ELEVATOR

Stream Numbers
Material
1
2*
3
4*
5
6
HF
CaF2
560(s)(fl,B)
3 (s)(C)
447(1}(B)
54 (g)!C)


NaF



0.6(s)^C'


Total Fluorides
560
3
447
55


Total as F
. 273
1-5
218
54


Si02
2850
15
2840



A12°3
450
2.4
450



B2°3
45
0.2
45



Na^COj
525
2.9

4.5(s)(0)


K?0
90
0.5
90



CaCOj
450
2.3




ZnO
420
2.3
420



h2o
90


1350 (g)


CaO


240



Ha20


320



n2



10800 (g)


o2



1200 (g)


C02



1950 (g)


NO



30 (g)


Natural Gas




530

Air (501 R,H. )"





14300
Approx Total
5480
29
4950
15000
530
14300 '
•Gaseous effluent Stream
Soluble fluoride evolution factor = 21.8 lb F/ton opal qlass (193 lb F/ton CaF, fed)
(A)	Reference 4251
(B)	Reference 4255
(C)	Reference 889
(0) Reference 4244
Figure 3-43. Opal Glass Production — Uncontrolled Process Model

-------
The amount of fluorspar consumed in opal glass production will
(42Q1}
increase from about 35,000 tons in 1968v 'to about 57,000 tons in 2000
(assuming a constant per-capita consumption of 0.34 pound--l968 value).
Table 3-73 gives current and projected estimates of the amount of fluorspar
used in opal glass production.
Table 3-73. Opal Glass Production

1967
1968
Estimated^
to Year 2000
Fluorspar Used in
Glass(A) Production
(tons)
U.S. Population (million)
Per-Capita Consumption of
Fluorspar Used in GlassW
(pounds)
31,800(4278)
199.1
0.32
34,500(^278)
202.3
0.34
57,200
336.2(B)
0.34
(A)
' Assumed to be opal glass.
^ Based on estimated 1.61 annual growth rate.
^ Based on per-capita consumption.
3-219

-------
3.6.4	Fluoride Emission Control Techniques
As noted earlier, baghouse control techniques suitable for
abatement of particulate dispersoid and fume from melting operations are
the control devices currently employed in the opal glass industry. These
devices are designed for use with the large, direct-fired continuous pro-
duction glass furnaces.
A summary of information on glass manufacturing and pollution con-
trol is contained in Reference 5156.
Figure 3-44 presents flow diagrams and mass balances for the three
control processes currently employed, applied to abatement of effluent from
a 60-ton-per-day opal glass plant. It should be noted that only Process A
(the wet cyclone process) affords any control of soluble fluoride emissions,
and that this process is used to a very minor extent—estimated at less than
1% of industry capacity.
3.6.5	Fluoride Emissions
Relatively few data points on fluoride emissions from glass plants
are available in the literature. Weyl^^ reports that about 20% of the
feed fluorspar is emitted. Substantiating this value, TRW-RRI experience
indicates emissions which range as high as 17 pounds of fluorine per ton
of glass produced, equivalent to 25% to 30% of the feed fluorspar for the
glass compositions tested.
To verify the above data prior to the estimation of evolution and
emission factors for opal glass manufacture, the proprietary thermochemical
analysis program described in Section 3.1.2 was employed on a typical opal
glass charge composition. At a system temperature of 2800°F (the normal
glass furnace temperature), if equilibrium conditions were attained in the
gas phase, 51.67% of the charged fluoride would be volatilized (Table 3-74).
Almost all of the volatized fluorides would be soluble fluorides. Gaseous
HF would be formed in major quantities from the reaction which takes place
at high temperatures between gaseous CaF2 and any gaseous source of hydro-
gen (including the water formed by combustion).
3-220

-------
WATER
^ COOLING DUCT
1500°F "E
PROCESS A
(FROM FURNACE)
-TO FAN
AND STACK
WET CYCLONE
EFF. = 52% PARTICULATE
80% (EST.) GASEOUS
TO DISPOSAL
OR RECYCLE
AFTER LIQUID-SOLID
SEPARATION
BAG HOUSE
EFF. = 99% PARTICULATE
0% GASEOUS
COOLING DUCT
1500°F "l
PROCESS B
(FROM FURNACE)
400°F
TO FAN
AND STACK
TO DISPOSAL
OR RECYCLE
BAG HOUSE
EFF. = 99% PARTICULATE
0% GASEOUS
:
TO FAN
'12> AND STACK
PROCESS C
(FROM ELEVATORS)
TO RECYCLE
NOTE: ASSUMES NO FLUORIDE ADSORPTION
ON PARTICULATE MATTER
BASIS - 60 TONS/DAY OF OPAL &ASS PRODUCTION PROCESS STREAMS - LB/HR
PROCESS STREAMS - LB/HR

Stream Number
Material
2
4
7
8*
9
10*
11
12*
HF
CaF2
NaF
3(s)
54(g)
0.6(s)
43(1)
0.3(s)(°)
ii(g)(D)
0.3(s)^
0.6(s)(D)
54(g)(D)
0.006(s)
3(s)(D)
0.03(s)(D)
Total Fluorides
Total as F
3
1.5
55
54
43
41
11
11
0.6
0.3
54
51
3
1.5
0.03
0.015
Si02
Al2°3
B2°3
Na2C03
k2o
CaC03
ZnO
h2o
n2
°2
co2
NO
15(s)
2.4 (s)
0.2 (s)
2.9(s)
0.5 (s)
2.3(s)
2.3 (s)
4.5(s)
1350(g)
10800(g)
1200(g)
1950(g)
30(g)
2.4(s)
900(1)
2.1(s)(D)
4502(g)
10800(g)
1200(g)
1950(g)
30(g)
4.5(s)
0.04(s)(°)
1350(g)
10800(g)
1200(g)
1950(g)
30(g)
15(s)
2.4(s)
0.1(s)
2.8 (s)
0.5(s)
2.2 (s)
2.2(s)
0.15(s)
0.03(s)
0.002(s)
^.03(s)
0.05(s)
0.03(s)
0.03(s)
Approx. Total
Stream
29(b)
15000
900
14100
5
15000
30
0.3^
~Gaseous Effluent Stream
(A)
Plus scrubbing water and recycled soluble fluorides.
(B)Plus	dilution air.
^Utilization of soluble fluoride controlling processes (Process A) estimated to be less than 156.
^Reference 5156
Soluble Fluoride Emission Factor lb F/ton glass
Process A Process B Process C
4.4 21.8
-	-		0
Total Soluble
Fluoride Emission	4.4	21.8	0
Overall soluble fluoride emission factor for the industry = 21.8 lb F/ton opal glass (193 lb F/ton CaF2 fed)
Figure 3-44. Opal Glass Production —
Controlled Process Model
3-221
Source
Glass Furnace
Bucket Elevator

-------
Table 3-74. Opal Glass Equilibrium Analysis Charge Composition
(Includes Combustion Gas Charge)
Weight, %
Mole, %
Si02
A1203
B2°3
Na2C03
k2o
CaCOg
ZnO
CaF0
CH,
C2H6
n2
co2
°2
H-0
Product
Compound
CaF2
NaF
HF
NaF
BOF
CaF0
• 9.75
1.62
0.17
2.60
0.32
1.46
1.46
1.615
2.20
0.74
59.30
0.04
18.02
0.70
Fluoride Distribution
at 2800F
Concentration
in Gas
Condensed (liquid)
Condensed (liquid)
9442.8 ppm
451.6 ppm
88.4 ppm
38.0 ppm
5.148
0.504
0.076
0.777
0.108
0.463
0.568
0.913
4.361
0.783
67.172
0.029
17.860
1.234
% of Total
Fluoride Charged
47.58
0.74
48.51
2.32
0.45
0.39
3-223

-------
On the basis of the thermochemical analyses program results and
private and open literature reports, an emission factor of 21.8 pounds of
soluble fluorides per ton of opal glass product was chosen as a typical
value. This emission factor corresponds to .the emission of 20% of the
charged fluorspar, in the form of soluble fluorides.
The opal glass industry emitted 3300 tons of soluble fluorides in
1968. Baghouse control used for dust abatement affords essentially no
control of soluble fluoride emissions. If the baghouse control techniques
currently employed are maintained through 2000, soluble fluoride emissions
will rise to 5500 tons per year. If wet scrubbing techniques are employed
to effect 99% abatement, soluble fluoride emissions in 2000 would be 55
tons per year. Table 3-75 summarizes soluble fluoride emission data for
opal glass.
3.6.6 Economic Analysis
Basic Process
Table 3-76 presents the estimated economics for producing opal
glass at two plant sizes characteristic of current production practice.
Labor rates and material costs are typical of Gulf Coast data. Return on
investment prior to use of fluoride control processes is estimated to be
14.5% and 46.7% for the 20 ton and 60 ton per day plants.
Impact of Control
Tables 3-77 through 3-79 show the estimated capital outly and
operating costs for three current process approaches to dust and fume
abatement in the opal glass industry. Because of emissions control,
aROI is estimated between -2% and -3%.
3-224

-------
Table 3-75.
Soluble Fluoride Emissions From
Opal Glass Production
1968	2000
CaFp utilized in Opal Glass	34.5	57,2
uai-? utilized in upai bias:
Production (103 tons/year)
Soluble Fluoride Evolution Factor	193	193
(lb F/ton CaF2 fed)
Soluble Fluoride Emission Factor	193	193
with Current Control (lb F/ton CaFg fed)
Soluble Fluoride Emission Factor with	-	1.93
991 Control (lb F/ton CaF^ fed)
Soluble Fluoride Evolution	3.32	5.51
(103 tons F/year)
Soluble Fluoride Emission with	3.32	5.51
Current Control (103 tons F/year)
Soluble Fluoride Emission with	-	0.055
991 Control (103 tons F/year)
3-225

-------
Table 3-76. Estimated Economics of Opal Glass Production
(Pollution Control Cost Excluded)

PLANT
CAPACITY
20 Tons/Day
60 Tons/Day
TOTAL CAPITAL INVESTMENT^
3.2 $MM
5.8 $MM
PRODUCTION COSTS


DIRECT COSTS


GLASS SAND (.8 Tons/Net Ton @ 12.75 $/Ton)
10.20 $/Net Ton
10.20 $/Net Ton
SODA ASH (.15 Tons/Net Ton @ 34.00 $/Ton)
5.10
5.10
FLUORSPAR (.16 Tons/Net Ton @ 65.00 $/Ton)
10.40
10.40
BORAX (.01 Tons/Net Ton @ 50.25 $/Ton)
.50
.50
FELDSPAR (.25 Tons/Net Ton @ 20.00 $/Ton)
5.00
5.00
LIMESTONE (.12 Tons/Net Ton @ 4.00 $/Ton)
.48
.48
NATURAL GAS (4.6 MM BTU/Ton X .40 $/MM BTU)
1.84
1.84
LABOR (7 Positions @ 4.00 $/hr)
33.60
11.20
FURNACE REPAIRS (1.2%)
5.82
3.52
FRINGE BENEFITS AND SUPERVISION
33.60
11.20
GENERAL MAINTENANCE AND SUPPLIES (1.25%)
6.08
3.67
TOTAL DIRECT COSTS
112.62
63.11
INDIRECT COSTS


DEPRECIATION(@ 10*)
48.48
29.29
INTEREST (AT 7%, 2ti% DEBT)
6.74
4.10
LOCAL TAXES AND INSURANCE (@ 3%)
14.54
8.79
. PLANT AND LABOR OVERHEAD
40.32
13.44
TOTAL INDIRECT COSTS
110.13
55.62
TOTAL AVERAGE COST*2) ($/Net Ton)
222.75
118.73
GENERAL AND SALES EXPENSES($/Ton)
4.46 $/Net Ton
2.37 $/Net Ton
F.O.B. COST ($/Net Ton)
227.21 $/Net Ton
121.10 $/Net Ton
AVERAGE PRODUCT REVENUE ($/Ton)
340.00 $/Ton
340.00 $/Net Ton
AVERAGE PROFIT AFTER TAXES (@ 502)
56.40 $/Ton
109.45 $/Ton
CASH FLOW ($MM/YR)
0.69 $MM/YR
2.75 $MM/YR
RETURN ON INVESTMENT (%)
14.5%
46.7%
^^Small tank furnace facility
(2)
'Transporatlon charges alter costs substantially
3-226

-------
Table 3-77. Opal Glass Production - Estimated Economics of Control Process A
Basis - 60 Tons Per Day of Opal Glass Production
Item
Number
1.
2.
_Ca£j_taJ__Cost_Estimates_££l0002
Description
COOLING DUCT, tins1500°F,
tout = 400°F, carbon steel, 22,500 cfm
500 ft2 surface area, air cooled,
2 in. W.G. pressure drop
WET CYCLONE, 100 gal/min,
22,500 cfm, neoprene lined steel,
2 in. W.G. pressure drop
tquipment
F.O.B.
Cost
50
Reference
Number
4383
4387
4383
Installation
Factor
1.67
1.55
tquipment
Installation
Cost	
78
Capital Subtotal	83.
Indirects (9 15%)	12.
Contingency (@.20k)	17_.
Total Capital (as of January 1971)	112.
All control economics foot notes are located in Section 3.1.1, pages 3-10 and 3-11.
D£ei^tj_rmCosJ^$^hr____2
Item
Number
Power
Cost
0.05
0.05
Maintenance
Cost
0.10
0.13
Subtotal
Water
121)
tquipment
Operating
Cost
0.15
0.18
Disposal
0.33
(100 gpm, 90% recycle ) 0.02
Total Operating Cost
0.35
Total Operating Cost ($/hr )
Taxes and Insurance (2%, 330 days)
Capital (7.1%,330 working days)
Pollution Control Cost ($/hr )
Pollution Control Cost ($/ton )
0.35
0.28
1.00"
1.63
0.65

-------
Table 3-78. Opal Glass Production - Estimated Economics of Control Process B
Basis - 60 Tons Per Day of Opal Glass Production
Item
Number
_Cajjita1_Cost_Estimates_J£10002
Description
COOLING DUCT, tin = 1500°F,
tout = 400°F,carbon steel, 22,500 cfm,
500 ft^ surface area, air cooled,
2 in. W.G. pressure drop
BAGHOUSE, 22,500 cfm, 2.5 in. W.G.
pressure drop, 0.3 lbs solid loading
per hour, fabric
rquTpmen?
F.O.B.
Cost
20
Reference
Number
4383
4383
Installation
Factor
1.67
4.13
Equipment
Installation
Cost	
83
Capital Subtotal	88
Indirects (P 152)	13
Contingency (@ 20% )	|8_
Total Capital (as of January 1971)	119
All control economics footnotes are located in Section 3.1.1, page 3-10 and 3-11.
Jperatinq Cost (S / hr
Item
Number
Power
Cost
.05
.06
Maintenance
Cost
0.10
1.26
tquipment
Operating
Cost
Subtotal
121)
Water
Disposal
(22)
Total Operating Cost
0.15
1.32
1.47
1.47
Total Operating Cost ($/hr )	1-47
Taxes and Insurance {2%, 330 days)	0.30
¦Capital (7.1%, 330 working days)	1-07
Pollution Control Cost ($/hr )	2.84
Pollution Control Cost ($/ton )	1.14

-------
Table 3-79. Opal Glass Production - Estimated Economics of Control Process C
Basis - 60 Tons Per Day of Opal Glass Production
Item
Number
CO
i
ro
ro
10
_Ca£UaJ_Cost_Es^mates_^10002
Description
BAGHOUSE, 7500 cfm, 2 lb solid
loading per hour, fabric
Equipment
F.O.B.
Cost
10.
Reference
Number
4383
Installation
Factor
4.00
"TquTpmenT"""
Installation
Cost	
40.
Capital Subtotal	40
Indirects (P152)	6
Contingency (@ 20% )	8_
Total Capital (as of January 1971)	54
Operating Cost (j /hr )
Item
Number
Power
Cost
0.01
Maintenance
Cost
1.26
tquipment
Operating
Cost
1.27
Subtotal
121]
1.27
Water
Disposal
(22)
Total Operating Cost
1.27
All control economics footnotes are located
in Section 3.1.1, pages 3-10 and 3.11
Total Operating Cost ($/hr )	1.27
Taxes and Insurance (2%, 330 days)	0.14
Capital (7.1%, 330 working days)	0148
Pollution Control Cost ($/hr)	1.89
Pollution Control Cost ($/ton)	0.76

-------
3.7 FRIT SMELTING

-------
3.7 FRIT SMELTING
3.7.1 General
Ceramic coatings are applied to metal, glass, and pottery as protective
and decorative coatings. These coatings are applied as mixtures of glassy
particles and clay in a water suspension called slip. The glassy particu-
late, or frit, is produced by melting refractories and flux materials in
a furnace followed by quenching and grinding to produce a finely ground,
fusable material. The primary elements of frit are refractories and fluxes
with minor amounts of colors, opacifiers, and other additives. The fluxes
include soda ash, borax, cryolite and fluorspar. Opacifiers of the devitri-
fication type include cryolite and fluorspar; the insoluble opacifiers do
not.
3.7.2	Industry Description
The production of enamel glass ("frit") 1s generally a batch process.
After the raw material charge is mixed, the batch is fed to a large hearth
smelter and held at temperatures which range to 2700°F until uniformly melted.
The melt is poured into a quenching tank of cold water and shatters into the
friable particles which constitute the frit. A typical integrated enamel-
frit- smelter, sheet-steel-enameling plant is shown in Figure 3-45.
3.7.3	Production Trends
No information is available on the total production of enamel frit.
This is due to the fact that a major amount of frit is consumed by cast-iron
plumbing fixture manufacturers. These manufacturers do not belong to a
representative organization and details on their individual production levels
of frit are not available to the public. For this reason, predictions of
future production are based on the amount of fluorspar consumed in the manu-
facture of enamel frit. Predictions of the year 2000 were made on a per-
capita consumption basis.
The amount of fluorspar used in frit production will increase from
the level of 7800 tons in 1968 to 11,800 tons in 2000 (assuming a per-
capita consumption of 0.07 pound). These data are presented in Table 3-80.
3.7.4	Fluoride Control Techniques
Smelting of the frit volatilizes gaseous and particulate fluorides and
other very fine particulate matter. Emphasis has been on removal of the
. 3 -231

-------
RAW MATERIALS
INPUT
TO ATMOSPHERE]
o

WEIGH
HOPPER


COMBUSTION
FUEL AND AIR
SHEET STEEL ENAMELING
FINISHED
PRODUCT
600° F
STACK
HEARTH
SMELTER
MELT
MIXER
WATER
RAW MATERIALS
INPUT
QUENCH
TROUGH
CATCH
BASKET
TO ATMOSPHERE
(POTENTIAL
FLUORIDE EMISSION)
WATER RECYCLE
WATER
MATERIAL TO BE
ENAMELED
O
FRIT SLURRY
FURNACE
NOTES:
1.
COMBUSTION
FUEL AND AIR
IMMERSION TANK
SMELTER OPERATES AT ~ 2700°F
SMELTER SIZE ~ 30 LB. OF BATCH
PER SQ. FOOT HEARTH.
THE EFFLUENT STEAM FROM THE QUENCH
TROUGH MAY CONTAIN FLUORIDE COMPOUNDS,
LITERATURE SEARCH REVEALED NO DATA.
2.
PARTICLE SIZE OF SOLIDS IN
THE FRIT SLURRY ARE:
GROUND COAT (1ST COAT) = 95% THRU 200 MESH
COVER COAT (2ND COAT) = 98% THRU 325 MESH
SHEET STEEL COATINGS USUALLY 5 TO 8 MILS THICK.
4.	FIRING TEMPERATURE = 1500° F FOR 5 MIN.
5.	MATERIALS BEING ENAMELED MUST
BE RECYCLED THROUGH A SECOND
COVER COAT AFTER THE INITIAL
GROUND COATING.
BASIS - 1000 LBS/HR FEED (STREAMS 1 & 4), SHEET STEEL ENAMELING
PROCESS STREAMS - LBS/HR
Material
Stream Numbers

1
2*
3
4
5
6
HF

0.74(g)(Est.)




NaF

1.55(s)(Est.)




CaF2
13 (s)(B'C)



10 (s)(C)

Total Fluorides
13
2.3


10
l
Total as F
6.3
1.4(D)


4.9
%
Feldspar
600





Borax
300





Si02
40





Na2C0g
25
2




NaN0?
10





N2

2700 (g)




°2

320 (g)




CO 2

380 (g)




h2o

330 (g)


100
100
Air (50% R.H.)


3500



Natural Gas


130



CoO



1


MnO



4


NiO



1


Frit




890

Approximate Total
990
3700
3600
6
1000

Stream






~Gaseous effluent stream
(A)	Plus make-up H?0 for steam loss
(B)	Reference 4251
(C)	Reference 4257
(D)	Reference 4258
Soluble fluoride evolution factor = 3.15 lb F/ton dry frit (215 lb F/ton CaF2 fed)
Figure 3-45. Enamel Fritting — Uncon-
trolled Process Model
3-233

-------
particulate material, arid the most frequently used devices are baghouses
and venturi scrubbers.	The venturi scrubber approach will remove
gaseous fluorides. Although central processes are currently applied to the
smelter effluent, it may become necessary to control the quench trough and
baking furnace effluents also. Mass balances and process flow diagrams are
presented in Figure 3-46 for two currently employed control processes.
Table 3-80. Enamel Frit Production

1967
1968
Estimated^
to year 2000
Fluorspar used in Frit
Production (tons)
4900(4278)
7800(4278)
11,800
U.S. Population
(mil lion)
199.1.
202.3
336.2^
Per-Capi ta.Consumpti on
of Fluorspar used in Frit
(pounds)
0.05
0.08
0.07
3.7.5 Fluoride Emissions
Soluble fluoride emissions from the enamel frit industry are estimated
at 700 tons in 1968, and are forecast at 1060 tons in 2000, assuming contin-
uation of the use of venturi and wet scrubber control at the current level.
It is estimated that only 20% of the operational facilities currently utilize
wet scrubbers. If wet scrubber devices are applied throughout the industry
and provide abatement at the 99% efficiency level, the soluble fluoride
tonnage emitted by the frit industry would drop to about 13 tons in 2000.
Table 3-81 presents a summary of these data.
Emission of soluble fluorides^from the hearth smelter used for frit
production, on the basis of the proprietary thermochemical analyses program,
follows a like mechanism to that involved in evolution and emission of
soluble fluorides from opal glass furnaces. At the high temperatures
present in the hearth, the volatilized fluorides react with water vapor so
that at equilibrium, roughly equal molal concentrations of gaseous HF and
NaF are formed. When the smelter gas passes into the stack and cools off,
the NaF forms a particulate dispersoid fume.
3-235

-------
VENTURI SCRUBBER
EFF. = 65% PARTICULATE
94% GASEOUS	PROCESS A
TO STACK
WATER
TO LIQUID - SOLID SEPARATION UNIT
WITH SOLIDS DISPOSAL AND LIQUID RECYCLE
RADIANT COOLING
COLUMNS
BAG HOUSE
EFF. =98% PARTICULATE
0% GASEOUS
r~\
<£>
3>
. v —TO FAN
10> AND STACK
PROCESS B
TO DISPOSAL
NOTE: ASSUMES NO ADSORPTION OF FLUORIDES OR RECYCLE
ON PARTICULATE MATTER.
BASIS - 1000 LBS/HR FEED, SHEET STEEL ENAMELING PROCESS STREAMS - LBS/HR

Stream Number
Material
2
7
8*
9
10*
HF
NaF
0.74(g)
1.55(s)
0.70 (1)
1.01 (s)
0.04(g)(C)
0.54(s)^
1.52(s)
0.74(g)
0.03(s)
Total Fluorides
Total as F
2.3
1.4
1.71
1.12
0.58
0.28
1.52
0.69
0*77
0.71
Na2C03
n2
°2
co2
h2o
2(s)
2700(g)
320(g)
380(g)
330(g)
1.3(s)(C)
230(1)
0.7(s)(C)
2700(g)
329(g)
380(g)
100(g)(Est.)
1.96(s)(C)
2
0.04(s)^
2700(g)
320(g)
380(g)
330(g)
Approx. Total
Stream
3700
230(a)
3500
2
3700
*Gaseous Effluent Stream
(A)	Plus scrubbing water and recycled soluble fluorides.
(B)	Utilization of wet control processes
(C)	Reference 5156
(D)	Control Processes estimated to be utilized by 20% of the industry;
remaining 80% uncontrolled.		

Soluble Fluoride Emission Factor -
lb F/ton dry Frit
Source
Process A
Process B
Smelter Emission
0.63
1.60
Assumed Fugitive Emission
0
0
Total Emission
0.63
1.60
Overall soluble fluoride emission factor = 2.64 lb F/ton dry frit (180 lb F/ton CaFg fed)^ ^
Figure 3-46. Enamel Fritting — Control-
led Process Model
3-237

-------
Table 3-81. Soluble Fluoride Emissions from Enamel Frit Production

1968
2000
CaF2 utilized in Enamel Frit
Production (103 tons/year)
7.8
11.8
Soluble Fluoride Evolution Factor
(lb F/ton CaF2 fed)
215
215
Soluble Fluoride Emission Factor
with Current Control
(lb F/ton CaF2 fed)
180
180
Soluble Fluoride Emission Factor
with 99% Control
(lb F/ton CaF2 fed)

2.15
Soluble Fluoride Evolved
(103 tons F/year)
0.84
1.27
Soluble Fluoride Emission with
Current Control
(103 tons/year)
0.70
1.06
Soluble Fluoride Emission with
99% Control
(103 tons/year)

0.013
3-239

-------
3.7.6 Economic Analysis
Basic Process. The economics of a characteristic frit production
plant are summarized in Table 3-82. While there is a moderate volume of
"merchant" frit production for sale, a considerable portion of the frit
produced is for captive consumption. Return on investment without fluoride
control processes, 1s estimated at 18.8% (Table 3-82).
Impact of Control. Tables 3-83 and 3-84 present economic analyses
of control processes currently employed for abatement of fumes from frit
manufacture- AROI because of emissions control is estimated as 12%,
equivalent to a reduction in ROI to 16.5%, due to the added costs of
pollution control.
3-240

-------
Table 3-82. Estimated Economics of Enamel Frit Production
(Pollution Control Cost Excluded)
Basis - 890 lb/hr Frit Production
Total Capital Investment	1.9 $MM
Production Costs
Direct Costs
Fluorspar (0.015 Tons/Net Ton at $65.00/Ton)	.98 $/Net Ton
Feldspar (0.67 Tons/Net Ton at $20.00/Ton)	13.40
Borax (0.34 Tons/Net Ton at $50.25/Ton)	17.09
Silica (0.045 Tons/Net Ton at $12.75/Ton)	.57
Soda Ash (0.028 Tons/Net Ton at $34.00/Ton)	.95
Soda Niter (0.011 Tons/Net Ton at $213./Ton)	2.34
Cobalt Oxide (0.0011 Tons/Net Ton at $4400./Ton)	4.84
Manganese Oxide (0.0045 Tons/Net Ton at $440./Ton)	1.98
Nickel Oxide (0.0011 Tons/Net Ton at $2700./Ton)	2.97
Natural Gas (5300 SCF/Ton at $.40/1000 SCF)	2.12
Water (190 Gal/Ton at $.20/1000 Gal)	.04
Labor (Four Positions at 4.00 $/hr)	35.96
Smelter Repairs (1.2%)	6.47
Fringe Benefits and Supervision	35.96
General Maintenance and Supplies (1.25%)	6.74
Total Direct Costs	132.41
Indirect Costs
Depreciation (at 7.1%)	38.28
Interest (at 7%, 20% Debt)	3.36
Local Taxes and Insurance (at 3%)	16.17
1 Plant and Labor Overhead	43.15
Total Indirect Costs	100.96
Total Average Costs ($/Net Ton)	233.37 $/Net ton
General and Sales Expenses ($/Net Ton)	4.67 $/Net ton
F.O.B. Cost ($/Net Ton)	238.04 $/Net ton
Average Product Revenue ($/Net Ton)	400.00 $/Net ton
Average Profit After Taxes (at 50%)	80.98 $/Net ton
Cash Flow ($MM/yr)	0.42 $MM/yr
Return on Investment (%)	18.8%
3-241

-------
Table 3-83. Enamel Fritting - Estimated Economics of Control Process A
Basis - 10 Tons of Frit Produced Per Day
i
ro
ro
Item
Number
Description
Ca£Vtal_Cost_Es_Mmates_^10002
"Equipment"
F.O.B.
VENTURI SCRUBBER, tjn = 600°F,
0.04 lbs solid/min loading, monel
clad, 2300 cfm, 31.5 in W.G.
pressure drop
Cost
22
Reference
Number
4383
4390
4391
Instal lation
Factor
1.63
Equipment
Installation
Cost
36
Capital Subtotal	36
Indirects (9 15%)	5
Contingency (I? 20% )	_7_
Total Capital (as of January 1971)	48

Qperatin'
Item
Number
q_Cost_J$^_hr_^_
Power
Cost
0.9
Maintenance
Cost
0.13
Subtotal
121)
Water
Di sposal
(22)
20 gpm
Total Operating-Cost

tquipment
Operating
Cost
0.22
0.22
0.01
0.23
All control economics footnotes are located
in Section 3.1.1, pages 3-10 and 3-11.
Total Operating Cost ($/hr )	0.23
Taxes and Insurance (2%, 330 days)	0-™
Capital (7.1%, 330 working days)	0.34
Pollution Control Cost ($/ hr )	u.by
Pollution Control Cost ($/ton )	1.66

-------
CO
I
ro
CO
Item
Number
Table 3-84.
Enamel Fritting - Estimated Economics of Control Process B
Basis - 10 Tons of Frit Produced Per Day
Description
Ca£Hal_Cost_Est1mates_^10002
"Equipment"
RADIANT COOLING COLUMNS, 1300 ft',
2300 cfm, 2 in W.G. pressure drop,
carbon steel t^n=300°F, t„11+=l50°F
out
BAGH0USE, fabric filter, 2300 cfm,
2.5 in W.G. pressure drop,
1.2 horsepower
F.O.B.
Cost
10
20
Reference
Number
4383
4383
4387
Installation
Factor
1.80
4.13
Equipment
Installation
Cost	
18
83
Capital Subtotal	101
Indirects (3 15%)	15
Contingency (@20%)	. 20_
Total Capital (as of January 1971)	136
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
Operating Cost ($ / hr
Item
Number
Power
Cost
0.01
Maintenance
Cost
0.10
1.26
Subtotal
121)
Water
Disposal
(22)
Total Operating Cost
tquipment
Operating
Cost
0.10
1.27
1.37
1.37
Total Operating Cost ($/hr)1.37
Taxes and Insurance (2%,330 days)	0.34
Capital (7.1%, 330 working days)	1.22
Pollution Control Cost ($/ hr )	2.93
Pollution Control Cost ($/ton )	703

-------
3.8 HEAVY CLAY PRODUCTS

-------
3.8 HEAVY CLAY PRODUCTS
3.8.1 General
Heavy clay products include the structural elements produced by
firing shapes fabricated from common clay with or without glazing, and
the refractories and pottery produced by firing shapes fabricated from
specialty clays and additives. The structural heavy clay products—brick,
pipe, and tile—use over 20 million tons of clay per year. This raw
material is reported to contain about 500 parts per million of combined
fluorine.(889»4297) jhe process temperatures (2000-3000°F) required to
produce the fired clay products are sufficient to volatilize much of the
fluoride content of the feed.
3.8.2 Industry Description
Figure 3-47 presents a flow diagram and mass balance for a typical
stiff-mud continuous production brick plant. Soluble fluoride evolution
occurs only in the high temperature portions of the tunnel kiln. All of the
evolved fluorides are converted to gaseous HF as the result of the high
temperature reaction with water vapor.
3-8.3 Production Trends
The use of clay for the production of heavy clay products (mainly
brick and tile) has remained relatively constant over the last 15 years.
This trend is not expected to change for at least the next 10 years and for
purposes of these projections is expected to be unchanged through the year
2000. This would keep the tonnage of clay used for heavy product production
at about 23 million tons annually, (^91) as shown in Table 3-85.
Table 3-85. Heavy Clay Product Projections

1954
1959
1963
1968
Estimated to
Year 2000
Clay Usage(4280^
(million tons)
23.4
23.3
22.6
23.7
23.5
3-245

-------
CLAY
FEED
~
STORAGE
HOPPER
DUST TO ATMOSPHERE
OR CONTROL DEVICE
DUST TO ATMOSPHERE
OR CONTROL DEVICE
STIFF-MUD PROCESS
SCREENS
DUST TO
ATMOSPHERE
OR CONTROL
DEVICE
yv" return
<2> TO SUPPLY
CRUSHER

DUST TO
-ATMOSPHERE
OR CONTROL
DEVICE
STORAGE
HOPPER
DUST TO -/k
ATMOSPHERE
OR CONTROL
DEVICE
CRUSHET^
SCREENS
FINES
RECYCLE
\
<3>
STORAGE
HOPPER
WATER
_w
N/S^W
EXTRUDER
CUTTER
TO ATMOSPHERE
OR POLLUTION
CONTROL DEVICE
FUEL


BRICKS TO
STORAGE
COOLING
SECTION

<3>
FIRING
SECTION
180d°F
HEATING
SECTION
TUNNEL
KILN
COMBUSTION
AIR
BASIS: 50 Tons per day Bfick Production
PROCESS STREAMS - LB/HR

Stream Number
Material
1
2
3
4
5
6
7*
8*, 9*, 10*
11*, 12*
HF






1.79(g)


CaF2 (Equiv.)
(A,B)
5.0(s)
, 0.5
-------
3.8.4	Fluoride Emission Control Techniques
Figure 3-48 presents process flow diagrams and mass balances for two
control processes currently employed in conjunction with structural clay
production. Note should be taken that Process B, involving a spray scrubber,
is estimated to be used in less than 1 percent of the heavy clay plants.
3.8.5	Fluoride Emissions
No definitive data exist in the open literature covering the quantities
and types of fluorides emitted by the heavy clay industry. Evolution is
reported as ranging from 30% to 95% of the feed fluoride content. ^9)
Because of the absence of definitive data in the literature, manu-
facturing processes for bricks were examined using a proprietary thermo-
chemical analyses program to determine the equilibrium distribution of
fluorides in the process outlet streams. If the condensed phases (solid
and liquid clay) are in complete chemical equilibrium with the gas phase,
all feed fluorine is found as HF in the exhaust gases, with no fluorine
remaining as condensed material. Table 3-86 summarizes the input charge
composition, the temperature at which the analysis was run, and the resulting
concentrations of HF produced in the effluent gas. At the higher tempera-
tures of the actual process, equilibria favor HF even more markedly.
For this study, it was assumed that the feed clay contained 500 ppm
of fluoride;(889,4297)	was qq^ evo-|ution (based on partial
attainment of thermochemlcal equilibrium); and that no abatement devices
suitable for collection of soluble fluorides were used. On these bases,
soluble fluoride emissions will remain constant at the 1970 level of 9700
tons (as F) annually through 2000 'if current control policies (which are
essentially no fluoride control) are continued. If control processes with
99% efficiency are applied on an industry wide basis, the heavy clay
industry-would emit 97 tons of fluoride in 2000. Table 3-87 is a summary
of the above data.
3.8.6	Economic Analysis
Basic Process. Table 3-88 presents an analysis of the economics of
structural clay production for a typical 50 ton per day brick plant. Return
on investment, prior to use of soluble fluoride control processes, is
estimated as 11.3%.
3-249

-------
BAG HOUSE
EFF. = 99% PARTICULATE
0% GASEOUS
CYCLONE
EFF. = 60% PARTICULATE
0% GASEOUS
300-400°F

TC
TO STACK
PROCESS A
RECYCLE
TO FEED
RECYCLE
TO FEED
SPRAY SCRUBBER
EFF. = 80% PARTICULATE
90% (EST.) GASEOUS
WATER-

A*

17> TO STACK
PROCESS B

TO LIQUID - SOLID SEPARATION
UNIT WITH SOLIDS DISPOSAL
BASIS - 50 TONS PER DAY BRICK PRODUCTION
PROCESS STREAMS - LB/HR.
Material
Stream Number

7
8, 9. 10, 11, 12
13
14
15*
15
17*
HF
1.79(g)




1.61^
0.18(g)(E)
CaF2 equiv.
-
0.23(s)
0.14(s)(E)
0.089(s)^
(E)
0.001(s)


Total Fluorides
1.79
0.23
0.14
0.089
0.001
1.61
0.18
Total as F
1.7
0.12
0.07
0.044
0.0005
1.53^^
0.17
Clay (dust)

230(s)
140(s)(E)
(E)
89 (s)
(E)
1 (s)

Ik
h2o
1900(g)




1500(1)(Est)
400(g)(est)
n2
10800(g)





10800(g)
°2
650(g)





650(g)
co2
1850(g)





1850(g)
Approx. Total
Stream
15200
(C)
230
140
90
(C)
1
(A)
1500
15200
*Gaseous Effluent Stream
(A)	Plus water and soluble fluorides
(B)	Water soluble fluorides
(C)	Plus dilution air
(D)	Utilization of soluble fluoride controlling processes (Process B) estimated to be less than 1 percent
(E)	Reference 4261
Source
Soluble Fluoride Emission Factor -
lb F/ton Product
Process A
Process B
Kiln Effluent
0.081
-
Dust Emissions
-
0
Assumed Fugitive
0
0
Total Soluble Fluoride Emission
0.081
0
Overall Soluble fluoride emission factor = 0.81 lb F/ton Product^
Figure 3-48. Structural Clay Production —
Controlled Process Model
3-251

-------
Table 3-86. Brick Production Process Charge Composition

Weights
Mole %
Si02
15.43
7.912
A12°3
3.88
1.173
Fe2°3
1.24
0.239
MgO
0.35
0.268
CaO
0.35
0.192
Na20
0.20
0.099
k2o
0.75
0.245
CaF2
0.015
0.006
h2o
5.45
9.333
n2
52.70
58.019
°2
15.91
15.326
ch4
3.73
7.186

Fluoride Distribution


at 1700F

Product
Concentration
% of Total
.Compound
In Gas
Fluoride Charged
HF
128.3 ppm
100.
3 -253

-------
Table 3-87. Soluble Fluoride Emissions from Structural Clay Production

1968
2000
Heavy clay production
(106 tons/year)
24
24
Soluble Fluoride Evolution Factor
(lb F/ton product)
0.81
0.81
Soluble Fluoride Emission Factor with Current Practice
(lb F/ton product)
0.81
0.81
Soluble Fluoride Emission Factor with 99% Control
(lb F/ton product)
-
0.0081
Soluble Fluoride Evolved
(103 tons/year)
9.72
9.72
Soluble Fluoride Emissions with Current Practice
(103 tons/year)
9.72
9.72
Soluble Fluoride Emissions with 99% Control
(103 tons/year)
-
0.097

-------
(A)
Table 3-88. Estimated Economics of Structural Clay Production*
(Pollution Control Cost Excluded)
Plant Capacity
50 tons/day
Installed Capital Investment	2.3 $MM
Operating Costs
Direct Costs
Clay (1.2 tons/ton product at $2.50/ton)	3.00$/ton
Water (30 gal/ton product at $0.03/1000 gal)	0.01
Natural Gas (6800 scf/ton at $0.35/1000 scf)	2.38
Electric Energy (98 kwh/ton at $0.007/kwh)	0.69
Operating Labor (4.00 $/hr)	3.84
Supervision and Fringe Benefits	3.84
Maintenance and Fringe Benefits (at 5% of	7.67
Capital/Year)
Total Direct Costs	21.43
Indirect Costs
Depreciation (at 4%/yr)	6.14
Interest (at 7%, 20% Debt)	2.15
Insurance and Local Taxes	4.60
Overhead	4.61
Total Indirect Costs	17.50
Total Manufacturing Cost ($/ton)	38.93
General and Sales Expenses ($/ton) 0.78
F.O.B. Cost ($/ton)	39.71
Average Product Revenue ($/ton)	67.45
Profit After Taxes (at 50%)	13.87 $/ton
Cash Flow ($MM/yr) 0.30 $MM/yr
Return on Investment^	11.3%
3-255

-------
Impact of Control. Tables 3-89 and 3-90 present analyses of the
control systems currently employed 1n conjunction with heavy clay products.
aROI due to emission control 1s estimated as -2 to -4%, decreasing ROI to
about 11%.
3-256

-------
Table 3-89. Structural Clay Production - Estimated Economics of Control Process A
Basis - 50 Tons of Brick Produced Per Day
CO
I
ro
<_n
Item
Number
_Ca£i^aj_Cos^_Est^nates_^10002
Description
CYCLONE 1500 cfm, carbon steel
400°F, 4.9 in W.G. pressure drop,
2 horsepower requirement
BAGH0USE, 1500 cfm, 2.5 in W.G.
pressure drop, fabric filter
Lquipment
F.O.B.
Cost
3.0
4.0
Reference
Number
4387
4390
4392
4383
Installation
Factor
2.00
4.13
Equipment
Installation
Cost	
6.0
16.0
Total Capital
Capital Subtotal	22.0
Indirects (0 15%)	3.3
Contingency (@ 20% )	4.4
(as of January 1971)	29.7
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
0£e«tjjiq_Cost_J$^_h£__J_
Item
Number
Power
Cost
0.01
0.01
Maintenance
Cost
0.13
1.26
Subtotal
(21)
Water
Disposal
(22)
Equipment
Operating
Cost
0.14
1.27
1.41
Total Operating Cost	1.41
Total Operating Cost (,5/hr )	1.41
Taxes and Insurance (2^,300 days)	0.08
Capital (6.7%, 300 working days)	0.28
Pollution Control Cost ($/ hr )	1.77
Pollution Control Cost ($/ton )	0.85

-------
Table 3-90.
Structural Clay Production - Estimated Economics of Control Process B
Basis - 50 Tons of Brick Produced Per Day
CO
r\s
ui
00
Item
Number
Capital Cost Estimates ($1000)
Description
SPRAY SCRUBBER, 6000 cfm
4 ft diameter by 8 ft, 2 in W.G.
pressure drop, neoorene lined steel
0.03 lbs KF(g)/min, 2.5 horse-
power, 130 gal/min
Equipment
F.O.B.
Cost
5.4
Reference
Number
4387
4391
4383
Installation
Factor
2.67
Equipment
Installation
Cost	
14.4
Capital Subtotal	14.4
Indirects (#158)	2.2
Contingency (0 20*)	2.9
Total Capital (as of January 1971)	19.5
All control economics footnotes are located In Section 3.1.1, pages 3-10 and 3-11.
O^eratijo^os^JS^jT^
Item
Number
Power
Cost
0.02
Maintenance
Cost
0.25
¦paapMHpk
Equipment
Operating
Cost
Subtotal
J21)
Water
Disposal
£22)
Total Operating Cost
Total Operating Cost- ($/hr
T
Taxes and Insurance (2%,300 days)
Capital (6.7%, 300 working days)
Pollution Control Cost ($/ hr )
Pollution Control Cost ($/ton )
0.27
0.27
( 130 gpm, 90% recycle) 0.02
0.29
0.29
0.05
0.18
0.52
0.25

-------
3.9 EXPANDED CLAY AGGREGATE

-------
3.9	EXPANDED CLAY AGGREGATE
3.9.1	General
Expanded clay aggregate manufacture produces fired, low-density
high-crush-strength pellets from clay for use in high strength concrete.
The bulk of the production is employed by the construction industry as a
light-weight aggregate replacement for gravel in structural concrete.
3.9.2	Industry Description
Figure 3-49 presents a flow diagram and mass balance for a
400-ton per day continuous production, expanded clay aggregate plant.
(Production is normally rated in cubic yards, with 1 cubic yard roughly
equivalent to 1 ton.)
Small pellets (1/4 x 1/2 inch) are formed in the pelletizer from
clay material, with added small amounts of oil and water. The pelletized
clay is fired in a sintering machine (generally a rotary kiln) to product
temperatures ranging between 1850° and 2100°F. The fired clay pebbles are
cooled and conveyed to storage after screening. Gaseous HF is evolved in
the sintering machine as the result of the high temperature reaction
between volatilized CaFg and water vapor.
3.9.3	Production Trends.
Current (1968) use of expanded clay aggregate is estimated at
9 million tons. Very little information is available on the projected use
of this product. The Bureau of Mines reports that the production of clay
and shale for lightweight aggregates will probably continue to increase at
a rate above that of GNP but not as rapidly as it has during the past 10
years.^91) por pUrp0ses 0f this projection the GNP is estimated to in-
crease at a rate of 4.3%^^ annually. The estimated increase in pro-
duction was assumed to be 5% compounded annually. This would put the
usage of clay for expanded clay aggregate at 44 million tons in 2000. As
Indicated by recent practice, new plants will be larger and more efficient.
Table 3-91 presents the past, current and projected production tonnages
for expanded clay aggregate.
3-259

-------
(a>
I
ro
cn
O
CLAY FEED
COMBUSTION PRODUCTS AND
DUST TO ATMOSPHERE OR
POLLUTION CONTROL DEVICE
@1400°F
BASIS - 400 TONS/DAY OF EXPANDED CLAY AGGREGATE PRODUCED
PROCESS STREAMS - LB/HR.
PELLETIZER
FUEL AND
COMBUSTION
AIR
PRODUCT
HOPPER
DUST TO ATMOSPHERE
OR CONTROL DEVICE DUST TO ATMOSPHERE
OR CONTROL DEVICE
DUST TO
ATMOSPHERE
OR CONTROL
DEVICE
CRUSHER
SCREENS
DUST TO ATMOSPHERE
OR CONTROL DEVICE
OVERSIZE
CRUSHER
DUST TO
ATMOSPHERE
OR CONTROL
DEVICE
DUST TO
ATMOSPHERE
OR CONTROL
DEVICE
SCREENS
SINTERING
MACHINE
2I00°F

Stream Number
Materials
1
2,3,4
5
6*
7*,8*.9*.10*.11*.12*
HF



20.0(g)

CaF2 Equiv.
(A.B.D)
46(s)
(B,C)
6.8(s)


(C)
0.2(s)
Total Fluorides
46
6.3

20.0
0.2
Total as F
22.4
3.3

,9.0(8,0
0.1
Clay
44,400


200(s)(C)
200(s)(C)
Natural Gas


2,700(Est)


A1r(50i R. H.)


50,800
-------
Table 3-91. Expanded Clay Aggregate Projections

1958
1968
Past 10 Yr Aooual
Growth Rate(°)
Estimated GNP^4250^
Growth Rate to 2000
Estimated^Usage
in Year 2000
Clay Usage
(Million tons)
4.46<4278'
9.28<4278'
7.6 %
4.3%
44.2
(A)	Growth rate assumed to be 5% annually.
(B)	Growth rates compounded annually.

-------
3.9.4	Fluoride Emission Control Techniques
The application of processes capable of control 1ing soluble
fluoride emissions to the production of expanded clay aggregate is esti-
mated as covering less than 1% of the currently used production capacity.
Figure 3-50 presents flow diagrams and mass balances for the two currently
employed emission control systems; only Process A is capable of abatement
of the gaseous HF emitted by the sintering machine (kiln).
3.9.5	Fluoride Emissions
There is the same lack of published information on the fluoride
species and quantities emitted in the manufacture of expanded clay aggre-
gate as was noted for the heavy clay products industry. Because of the
physicochemical similarity of the two processes, the assumptions and
thermochemical models used to estimate heavy clay fluoride emissions were
used to estimate soluble fluoride emissions from the expanded clay aggre-
gate industry. Charge composition and theoretical equilibrium evolved gas
composition are shown in Table 3-92.
Based on the logic discussed in Section 3.8, soluble fluoride
emissions will increase from 5300 tons annually in 1968 to about 25,100
tons annually in 2000, if the current lack of control continues. If high
efficiency (99%) control technology is employed, estimated soluble fluor-
ide emissions will drop to 250 tons annually in 2000. These data are
summarized in Table 3-93.
3.9.6	Economic Analysis
Basic Process
Table 3-94 summarizes the economic analysis of expanded clay
aggregate production for a 400 ton per day plant. Return on investment,
without control processes, is estimated at 31.7%.
Impact of Control
Tables 3-95 and 3-96 summarize the cost of fluoride pollution
control, using Processes A and B as presented in Figure 3-50, Controlled
Process Model.
3-262

-------
RECYCLE
OR DISPOSAL
TO PELL1TIZER
CYCLONE
EFF. = 60% PARTICULATE
0% GASEOUS
<7> <8> <9> 
1
DISPOSAL
OR RECYCLE
TO PELLETIZER
SPRAY SCRUBBER
EFF. = 80% PARTICULATE

CYCLONE
EFF. = 60% PARTICULATE
0% GASEOUS
- TO STACK
WATER
PROCESS A
TO LIQUID - SOLID
SEPARATION UNIT
WITH SOLIDS
DISPOSAL
BAG HOUSE
EFF. = 99% PARTICULATE
0% GASEOUS
n
TO STACK

DISPOSAL
OR RECYCLE
TO PELL1T1ZER
PROCESS B
NOTE: STREAMS 11 AND 12
ARE NOT CONTROLLED.
BASIS - 400 TONS/DAY OF EXPANDED CLAY AGGREGATE PRODUCED
PROCESS STREAMS - LB/HR
Material
7,8,9,10
11*.12*
13
Stream Number
14
15*
16
17
18*
HF
20.0(g)
18.0
2.0(g)
CE)
CaFg Equiv.
0.1(s)(Est.)
0.1(s)(Est.)
0.06(s)('fE)
0.039(s)

0.001(s)
(E)
Total Fluorides 20.0
Total as F
19.0
0.1
0.05
0.1
0.05
18.0
17.1
2.0
1.9
0.06
0.03
0.039
0.019
0.001
0.005
Clay (dust)
n2
°2
co2
h2o
200(s)
39100(g)
2000(g)
12700(g)
10400(g)
100(s)(Est.)
100(s) (Est.)
120(s)
(E)
64(s)
CE)
8800(1)
16(s)
39100(g)
2000(g)
12700(g)
1600(g)(Est)
60(s)
(E)
39(S)(£)
1 (s)
(E)
Approx. Total
Stream
64400
100
(A,B)
100
120
9000
(C)
53000
60
40
~Gaseous Effluent Stream
(A)	Plus dilution air.
(B)	Assumes a 50/50 split of the total amounts from streams 7, 8, 9, 10, 11, 12.
(C)	Plus scrubber water and recycled soluble fluorides.
(D)	Utilization of soluble fluoride controlling processes (Process A) estimated to be
less than 1 percent.
(E)	Reference 4262.
Soluble Fluoride Emission Factor lb F/ton Aggregate
Source Process A Process B
Sintering Machine
0.11
-
Dust Evolution
-
0
Assumed Fugitive Emission
0
0
Total Soluble Fluoride Emission
0.11
0
Overall Soluble Fluoride Emission Factor = 1.14 lb F/ton^ Aggregate.
Figure 3-50. Expanded Clay Products —
Controlled Process Model
3-263

-------
Table 3-92. Expanded Clay Process Charge Composition

Weight %
Mole %
Si Og
28.17
16.529
A12°3
7.27
2.514
Fe2°3
2.25
0.497
MgO
0.67
0.586
CaO
0.67
0.421
Na20
0.41
0.233
k2o
1.33
0.498
CaF2
0.03
0.012
h2o
4.92
9.636
n2
39.53
49.772
°2
11.98
13.199
ch4
2.77
Fluoride Distribution
at 1900F
6.104
Product
Compound
Concentration
in Gas
% of Total
Fluoride Charged
HF
279.9 ppm
0
0
3-265

-------
Table 3-93. Soluble'Fluoride Emissions from
Expanded Clay Aggregate Production

1968
2000
Clay Aggregate Produced
(10° tons/year)
9.3
44
Soluble Fluoride Evolution Factor
(lb F/ton Aggregate)
1.14
1.14
Soluble Fluoride Emission Factor with
Current Practice(A)
(lb F/ton Aggregate)
1.14
1.14
Soluble Fluoride Emission Factor with
99% Control
(lb F/ton Aggregate)
—
0.011
Soluble Fluoride Evolved
(103 tons F/year)
5.30
25.1
Soluble Fluoride Emissions with Current
Practice
(103 tons F/year)
5.30
25.1
Soluble Fluoride Emissions with 99% Control
(103 tons F/year)
-
0.25
(A) No fluoride emission control.
3-266

-------
Table 3-94. Estimated Economics of Expanded Clay Products^
{Pollution Control Cost Excluded)
Plant Capacity
400 tons/day
Installed Capital Investment
0.45 $MM
Operating Costs

Direct Costs
Clay (1,4 tons/ton product at $2.50)
Natural Gas (3800 scf/ton at 0.35 $/l000 scf)
Electric Energy (62 kwh/ton at 0.007$/kwh)
Operating Labor (at 4.00 $/hr)
Supervision and Fringe Benefits
Maintenance and Supplies (at 4$ of investment/yr)
3.50 $/ton
1.35
0.43
0.96
0.96
0.15 ,
Total Direct Costs
\
7.35
Indirect Costs
Depreciation (at 10% per yr)
Interest (at 7%, 20% Debt)
Insurance and Local Taxes
Plant and Labor Overhead
0.38
0.05
o.n
1.15
Total Indirect
1.69
Total Manufacturing Cost
General and Sales Expenses ($/ton)
F.O.B. Cost ($/ton)
Average Product Revenue
9.04
0.18
9.17
11.06
Profit After Taxes (taxed at 501)
0.95 $/ton
Cash Flow ($MM/yr)
0.16 $MM/yr
(B)
Return on Investmentv '
31.7%
(k\
v 'Assumes 300 operating days per year.

^Assumes 80% equity funding.

3-267

-------
Table 3-95. Expanded Clay - Estimated Economics of Control Process A
Basis - 400 Tons Per Day of Expanded Clay Aggregate Production
00
1
ro
en
CO
_Ca£jitaJ_Cost_Es^imates__y^0002i
Item
Number
Description
Equipment
F.U.B.
Cost
Reference
Number
Installation
Factor

tquipment
Installation
Cost
1
CYCLONE, 16,000 cfm, 3.33 lbs
solid/min, neoprene lined steel,
4.9 in W.G. pressure drop, 16.5
horsepower
15.0
4387
4390
4392
2.07

31 .0
2
SPRAY SCRUBBER, 16,000 cfm,
5 ft diameter by 12 ft, neoprene
lined steel, 360 gal/min, 2 in W.G.
pressure drop, 8 horsepower
20.0
4386
4387
4391
2.40

24.0
3
LIQUID-SOLID SEPARATOR, 360 gal/
min, 3.3 lbs/min loading,
22,000 gal capacity, neoprene
lined steel
19.0
4398
4392
4.26

81.0




Capital Subtotal
136.0




Indirects (P 15%)
20.0




Contingency ) 27.0


Total Capital (as of January 1971) 183.0
All control economics footnotes are located in Section 3.1.1 pages3-10 and3-ll
Ogeratinj
Item
Number
Cost (S / hr )
Power
Cost
0.9
0.4
Maintenance
Cost
0.13
0.25
0.06
Subtotal
T21)
Equipment
Operating
Cost	
0.22
0.29
0.06
Water
Disposal
(22)
0.57
(360 gpm, 90 % recycle) 0.04
Total Operating Cost
0.61
Total Operating Cost ($/hr)	o.61
Taxes and Insurance (2%, 300 days) 0.51
Capital ( 6.7%, 30(V
-------
Table 3-96. Expanded Clay Products - Estimated Economics of Control Process B
Basis - 400 Tons Per Day of Expanded Clay Aggregate Production
Capital Cost Estimates ($1000)
Item
Number
CO
IN)
cn
10
Description
CYCLONE, 5,000 cfm, 4.9 in W.G.
pressure drop, carbon steel,
5 horsepower
BAGH0USE, 5,000 cfm, 4.9 in W.G.
pressure drop, fabric filter-
shaker, 3 horsepower
TquTpmeri?
F.O.B.
Cost
3.0
9.0
Reference
Number
4387
4390
4392
4387
4383
Installation
Factor
3.0
4.13
tquipment
Installation
Cost	
9.0
36.0
Capital Subtotal	45.0
Indirects (0 15%)	6.8
Contingency (0 20%)	9.0
Total Capital (as of January 1971)	60.8
All control economics footnotes are located in Section 3.1.1, page 3-10 and 3-11.
0£e^tjjTqCost_J$^_h^
Item
Number
Power
Cost
0.04
0.02
Maintenance
Cost
0.13
1.26
Subtotal
T21)
Water
Disposal
(22)
Total Operating Cost
tqulpment
Operating
Cost
0.15
1.28
1.43
1.43
Total Operating Cost ($/hr )	1-43
Taxes and Insurance (2%,300 days)	0.17
Capital (6.7%, 300 working days)	0-57
Pol 1 ution. Control Cost ($/hr.)	2.17
Pollution Control Cost ($/ton)	0.13

-------
3.10 CEMENT MANUFACTURE

-------
3.10 CEMENT MANUFACTURE
3.10.1	General
Portland cement is manufactured in this country at the rate of
almost 100 million tons per year by grinding the clinkers resulting from
the calcination of mixtures of clay and limestone in rotary kilns. Blast
furnace slag, by product calcium carbonate, gypsum, sand, waste bauxite
and iron ore are frequently used in varying quantities, in addition to
the clay and limestone. The major raw materials contain calcium fluoride;
during the direct-fired kiln calcination process, the calcium fluoride
serves as a source material for the evolution of gaseous HF.
3.10.2	Industry Description
Process models covering the wet and dry processes for the manu-
facture of portland cement are presented in Figures 3-51 and 3-52. The
plant sizes selected were 10,000 barrels per day. In both wet and dry
processes, current practice charges the materials in the necessary pro-
portions to the closed circuit grinding system used to obtain the physical
size and intimate contact required for the chemical conversions which take
place 1n the kiln. The heat required for the chemical conversions is pro-
vided by direct-firing the kiln with oil, gas or pulverized coal. The
kilns are the major points of fluoride evolution in the cement processes.
3.10.3	Production Trends
The cement industry will continue to expand both in response to
population growth and to growing need in highway, industrial and housing
construction. Estimates of the future demand for this material are based
on long-term per-capita consumption rates and historical market trends.
Production has increased at an average annual rate of 3.7% (200 MM bbls
in 1949 to 360 MM bbls in 1964) since 1949.*4292)
On a per-capita basis, consumption of portland cement can be
expected to rise to almost two barrels per person by 1975. In the past
(4292)
decade the per-capita average has been about 1.8 barrels. ' Some
increase over the present level is expected because much of the net increase
in population in the next 10 years will occur in states where per-capita
use of cement is considerably above the national average. Consumption of
3-271

-------
STORAGE AND
PROPORTIONATING EQUIPMENT
LIMESTONE
FROM
QUARRY
CRUSHER V
MILL FEEDER
HAMMER
MILL
PRIMARY MILL
CLAY SLURRY
STORAGE BASIN
WATER
KILN
FEED
TANK
CLASSIFIER
SECONDARY
MILL
SLURRY MIXING AND
BLENDING TANKS
ROTARY KILN
GYPSUM
. FUEL AND
^ COMBUSTION
FLUE GAS
TO STACK
CLINKER
STORAGE
GYPSUM
STORAGE
CLINKER
COOLER
(POTENTIAL FLUORIDE
EVOLUTION POINT)
CEMENT
STORAGE
SILOS
FINISH MILL
PORTLAND CEMENT TO PACKING
BASIS: WET PROCESS MANUFACTURE OF PORTLAND CEMENT
AT 10,000 BBL PER DAY
PROCESS STREAMS - TONS/DAY

Stream Nuntoers
Materials
1
2
3
4
5
6*
r
7 '
HF





0.008(g)^D)

CaF2 (Equiv.)
3.35(Est)
0.82(Est)



4.17 (s)(D)

Total Fluoride
3.35
0.82



4.17

Total as F
1.63(Est)
0.40(Est)



2.03

Limestone
(E)
2500(s)





\
Shale (Clay)

620(s)(E)





Water
Gypsum
Fuel**
Coal
Oil***
Natural Gas****
Flue Gas****
Flue Partic-
ulates


1600(1 )(A)
80(s)(E)
530(s)^
92,000(1)^
8,800(g)^
270,000(g)(C)
190(s)(C)
~ ~~~~
(B)
1880 (s)
Portland Cement






Approx. Total
Stream
2500
620
1600
80
-
-
1880
* Gaseous Evolution stream
** Coal, oil and natural gas fuel options are presented
*** Gal per day
**** SCFM (1 atm, 60°F)
***** Equivalent to 10,000 bbI per day
Soluble fluoride evolution factor = 0.008 lb F/ton of cement (16 lb F/10^ bbl of cement)(
(A)	Reference 2220
(B)	Reference 2027
(C)	Reference 4298
(D)	Reference 4266
(E)	Reference 4244
Figure 3-51. Wet Process Portland Cement
Production - Uncontrolled
Process Model
3 -273

-------
SHALE (CLAY)
LIMESTONE
FROM
QUARRY
STORAGE BINS
CRUSHER
HAMMER
MILL
FLUE GAS TO STACK
(POTENTIAL FLUORIDE
EVOLUTION POINT)
PROPORTIONATING
EQUIPMENT
(POTENTIAL
FLUORIDE
EVOLUTION
POINT)
FUEL AND
COMBUSTION
AIR 	
DRY MIXING
AND
BLENDING
SILOS
CLAY AND ROCK
DRYER
VERTICAL
MILL
TUBE
MILL
FLUE GAS TO STACK
ROTARY KILN
GYPSUM
FUEL AND
COMBUSTION
AIR
KILN
FEED /
STORAGE
GYPSUM
STORAGE
CLINKER
STORAGE
CLINKER
COOLER
(POTENTIAL FLUORIDE
EVOLUTION POINT)
CEMENT
STORAGE
SILOS
FINISH MILL
PORTLAND CEMENT TO PACKING
BASIS: DRY PROCESS MANUFACTURE OF PORTLAND CEMENT
AT 10,000 BBL PER DAY
PROCESS STREAMS - TONS/DAY
Material
Stream Numbers
1
2
3
4
5
6*
7
HF
CaF2 (Equiv.)
3.35(Est)
0.82 (Est)

Amount
Unknown
II

0.008(g)(C)
4.17(s)(C)

Total Fluoride
3.35
0.82



4.17

Total as F
1.63
0.40



2.03

Limestone
2500(s)^





%
Shale (clay)

620(s)(°)





Gypsum


80(s)




Fuel**







Coal
Oil***
****
Natural Gas
Flue Gas****



(E)
190.000(g)
470(s)(D(A)
80,000(1 )
7,800(g)(D)
(B)
270,000(g)

Flue
partic-
ulates



300(s)(E)

(5)
215(s)

Portland
Cement:






*****
(A)
1880(s)
Approx. Total
Stream
2500
620
80
-
-
-
1880
* Gaseous Evolution Stream
** Coal, oil and natural gas fuel options are presented
*** Gal per day
**** SCFM (1 atm 60°F)
***** Equivalent to 10,000 bbl per day
Soluble fluoride evolution factor = 0.008 lb F/ton of cement (16 lb F/104 bbl of cement)
(A)	Reference 2027
(B)	Reference 4298
(C)	Reference 4266
(D)	Reference 4244
(E)	Reference 2096
Figure 3-52. Dry Process Portland Cement
Production — Uncontrolled
Process Model
3-275

-------
Portland cement should reach 450 to 500 million barrels by 1975.(^"292,4293)
This would represent an increase of about 140 to 190 million barrels over
the 1960 level, for an average annual increase of 2.5% to 3.0%. If these
rates continued to the year 2000, the annual production would be 830 to
1050 million barrels (156 to 197 million tons). The 3% rate was assumed
for the fluoride emission determination. Table 3-94 summarizes expected
cement production levels.
3.10.4	Fluoride Emission Control Techniques
The particulate emission control technique currently employed
in the cement industry is presented in Figure 3-53. This technique is
effective in removal of particulates and the soluble fluorides absorbed
thereon.
3.10.5	Fluoride Emissions
Cement production is of special interest since it involves evo-
lution of fluorides in the presence of limestone. Information concerning
the fate of the fluoride in this circumstance may be used to infer con-
clusions for similar situations, e.g., iron and steel processes or dry
1imestone process SOg control.
As noted earlier, the kilns are the major points of fluoride
evolution in the cement process. At equilibrium, 100% of the charge
fluoride would be evolved as gaseous HF at 2700°F (see Tables 3-95 and
3-96). At the same time as the gaseous HF is evolved, very large quanti-
ties of limestone and high free-lime-content particulate material are
dispersed into the combustion product stream. The active alkaline
surface area thus made available for adsorption and reaction with the
evolved HF is enormous, and much of the evolved gaseous HF is removed
from the gas stream.
Unfortunately, normal operating data are not available defining
fluoride emission factors for cement production. The cement industry has
concentrated on the particulate problem. Limited experience with fluor-
ides added to the feed^^^ does indicate that: (1) 70 to 80% of the
evolved fluoride can be collected in an electrostatic precipitator, and
(2) gaseous and water soluble fluorides were only about 10% of the total
3 -277

-------
Table 3-94. Future Cement Production

,»2)
Estimated
1975
Annual Growth
Rate
Extrapolated
to Year 2000
Cement Production




Million Barrels
360
450(4292)
2.5%
830
(million tons)
(67.7)
(84.6)

(156)

360
500^293J
3.0%
1047

(67.7)
(94.0)

; (197)

360
646(4293)
5.5^
2463

(67.7)
(121.4)

(463)
^ Reference (4293) indicates that this rate is probably far too optimistic.

-------
BASIS - WET AMD DRY PROCESS MANUFACTURE OF PORTLAND CEMENT AT 10,000 B8L PER DAY
PROCESS STREAMS - TON/OAY
- BAG HOUSE
EFF. = 99% PARTICULATE
0% GASEOUS
CYCLONE
EFF. = 50% (EST.) PARTICULATE
0% GASEOUS
TO STACK
WASTE HEAT
BOILER
BLENDING AND CEMENT
STORAGE SILOS ¦ *
BLENDING AND CEMENT
STORAGE SILOS

Stream number
Material
6
8
9
10*
HF
0.008(g)


0,003(q)
CaF2(equiv.)
4.17(s)
2.09(s)(B,C)
2.06(s)(B'C)
0.02(s)(B,C'
Total Fluorides
4.17
2.09(s)
2.06
0.03
Total As F
2.03
1.02
1.01
-0.018
Flue Gas**
270.000(g)


270.000(g)
Flue Particulates
202(s)***
101(s)
-------
Table 3-95.	Dry Cement Process Charge Composition
Weight %	Mole %
Si 02	0.175	0.082
A1203	0.045	0.012
Fe203	0.014	0.002
MgO	0.005	0.003
CaO	0.005	0.002
Na20	0.003	0.001
K20	0.009	0.003
CaC03	1.127	0.317
CaF2	0.005	0.002
CH4	3.515	6.187
N2	72.100	72.519
02	22.081	19.433
H20	0.917	1.435
Fluoride Distribution
at 2700F
Product	Concentration	% of Total
Compound	in Gas	Fluoride Charged
HF	40.0 ppm	100.
3-280

-------
Table 3-96. Wet Cement Process Charge Composition

Weight %
Mole %
CO
o
ro
0.154
0.072
A12°3
0.040
0.011
Fe2°3
0.012
0.002
MgO
0.004
0.003
CaO
0.004
0.002
Na20
0.002
0.001
o
CM
0.008
0.002
CaC03
0.989
0.277
CaF£
0.005
0.002
ch4
3.483
6.102
n2
72.028
72.115
°2
21.767
19.069
o
C\J
H!
1.504
Fluoride Distribution
at 2700F
2.342
Product
Compound
Concentration
in Gas
% of Total
Fluoride Charged
HF
40.0 ppm
100.
3-281

-------
fluoride emissions. This would appear to verify the effectiveness of
limestone, and possibly other particulate matter, in adsorbing or reacting
with gaseous fluorides. The result is that alleviation of the particulate
emission problem, which is being actively pursued, will also alleviate
fluoride emission problems.
Assuming average fluoride content of limestone and shale (650
ppm F) with no fluorspar addition to feed, soluble fluoride emissions will
grow from 270 tons (as F~) in 1964 to 800 tons in 2000 if current control
levels (no gaseous fluoride control) are maintained. If 99% efficient
control systems are utilized, the emissions will decrease to less than
10 tons (as F") in 2000. Table 3-97 summarizes the emission data.
Table 3-97. Soluble Fluoride Emitted From The Cement Industry

1964
2000
Cement Production
(10° ton/year)
68
200
Soluble Fluoride Evolution Factor
(lb F/ton cement)
0.008
0.008
Soluble Fluoride Emission Factor
With Current Practice
(lb F/ton cement)
0.008
0.008
Soluble Fluoride Emission Factor
With 99% Control (lb F/ton cement)
—
0.00008
Soluble Fluoride Evolution
(103 ton F/yr)
0.27
0.80
Soluble Fluoride Emission With
Current Practice
(103 ton F/yr)
0.27
0.80
Soluble Fluoride Emission With
99%-Control
(10 ton F/yr)

0.008
3-282

-------
3.10.6 Economic Analysis
Basic Process
Table 3-98 presents the estimated economics for the production of
Portland cement by either wet or dry processes for three plant sizes.
The extreme sensitivity of ROI to plant size is of interest—the estimated
ROI's for the three plant sizes before installation of emission control
are:
Plant Size
Million bbls/year	ROI, %
3.1
18.8
33.5
Impact of Control
Table 3-99 indicates the estimated costs for the current emission
control process used in both wet and dry methods of manufacture of portland
cement. Impact on ROI for the 4 million barrel per year plant is approxi-
mately 18%.
1
4
8
3-283

-------
(A)
Table 3-98. Estimated Economics of Portland Cement Manufacturev '
(Pollution Control Cost Excluded)
l.MM bbl/yr 4.MM bbl/yr 8.MM bbl/yr
Installed Capital Investment
8.0 $MM
25.0 $MM
35.2 $MM
Operating Costs





Direct Costs





Limestone (0.249 tons/bbl at 1.35$/ton)
0.34
$/bbl
0.34 $/bbl
0.34 $/bbl
Shale (0.062 tons/bbl at 2.00 $/ton)
0.12

0.12

0.12
Gypsum (0.008 tons/bbl at 2.00 $/ton)
0.02

0.02

0.02
Fuel (1.3 mm Btu/bbl at 0.35$/mm Btu)
0.46

0.46

0.46
Electric Energy (23 kwh/bbl at 0.007$/kwh)
0.16

0.16

0.16
Water (70 gal/bbl at 0.08$/gal)
0.01

0.01

0.01
Operating Labor(B)
0.58

0.20

0.13
Supervision and Fringe Benefits
0.58

0.20

0.13
Maintenance and Supplies (at 4% of invest./year)
0.32

0.25

0.18
Total Direct Costs
2.59

1.76

1.55
Indirect Costs





Depreciation (at 5% per year)
0.40

0.31

0.22
Interest (at 7%, 20% Debt)
0.11

0.09

0.06
Insurance and Local Taxes
0.24

0.19

0.13
Overhead
0.69

0.24

0.15
Total Indirect Costs
1.44

0.83

0.56
Total Manufacturing Cost
4.03
$/bbl
2.59
$/bbl
2.11 $/bbl
General and Sales Expenses
0.08

0.05

0.04
F.o.b. Cost
4.11
$/bbl
2.64
$/bbl
2.15 $/bbl
Average Product Revenue
4.51
$/bbl
4.51
$/bbl
4.51 $/bbl
Profit After Taxes (taxed at 50%)
0.20
$/bbl
0.94
$/bbl
1.18 $/bbl
Cash Flow ($MM/year)
(r\
0.60
$MM/yr
5.00
$MM/yr
11.2 $MM/yr
Return on Investment^ '
3.1%

18.8%

33.5%
(A) Wet process plant; dry process costs are similar to wet process costs. Assumes 300 working days
per year.
(B)	Assuming 20, 28, 35 men/shift for the 1, 4 and 8 MM bbl/year plants, respectively.
(C)	Assumes 80% equity funding.

-------
Table 3-99. Portland Cement Production - Estimated Economics of Control Process
Basis - 10,000 bbl Per Day of Portland Cement Produced
t hr
Item
Number
Description
tquipment
F.U.B.
Cost
Reference
Number
Installation
Factor

Equipment
Installation
Cost

Item
Number
Power
Cost
Maintenance
Cost

tquipment
Operating
Cost
1
WASTE HEAT BOILER, 58,000 ft2
tin = 1500°F' Vt = 55°°F'
1 ,250,000 cfm, UAt = 4800 Btu/hr/
170-0
4383
3.64

619.0

1
27.49
0.11

27.60







2
2.82
0.13

2.95

ft2, low alloy steel, Q = 288 MM
Btu/hr, 20 in W.G. pressure drop.






3.
1.10
1 .26

2.36
2
CYCLONE, 524,000 cfm, 4.9 in W.G.
pressure drop, carbon steel.
127.0
4387
4390
4392
2.98

379.0






3
BAGH0USE, 498,000 cfm, 500°F,
fabric filter, 2 in W.G. pressure
drop.
168.0
4383
4.13

696.0














Subtota


32.91








Water(21>






Capital Subtotal
Indirects (G 15%)
1 ,694.0
254.0

Disposal<22>

-




Contingency (020%). 339.0







Total Capital (as of January 1971) 2,287.0

Total Operating Cost
32.91
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11.
Total Operating Cost ($/hr)	32.91
Taxes and Insurance (2%, 300 days)	6.35
Capital (6.7%, 300 working days)	21 .28
Pollution Control Cost ($/hr)	60.54
Pollution Control Cost ($/bbl)	0.15

-------
3.11 HF ALKYLATION PROCESS

-------
3.11 HF ALKYLATION PROCESSES
3.11.1	General
Alkylation is a process for reacting isoparaffins with olefins or
olefin mixtures to form highly branched, high octane, paraffinic products.
The major currently used alkylation processes utilize either sulfuric
acid or hydrofluoric acid as a catalyst. Of the total U.S. alkylate
production (775,000 bbl/day in 1970), about 30% (236,000 bbl/day) was
produced in 62 units utilizing the HF process.(4238) The units varied in
capacity from 500 to 14,500 bbl/day with the average plant being
3800 bbl/day.^4238^
3.11.2	Industry Description
A simplified flow diagram and fluoride mass balance of the HF
alkylation process is presented in Figure 3-54. It is estimated that 75%
of the producing refineries burn the acid residue (stream 3) while the
remaining 25% treat this stream as discussed later in this section. This
estimation is based on the assumption that only those units located in
high density urban areas utilize the control process.
3.11.3	Production Trends .	-
Demand for petroleum products has been steadily increasing from
10.6 million bpd in 1963 to 13.0 million bpd in 1970. It is expected that
(4294)
demand will reach 15 million bpd in 1975. ' This is equivalent to a
growth rate of about 2.7% per year in contrast to the 3% experienced since
1955 and the 5.5% rate from 1945 to 1955. Rapid growth in passenger
vehicle and freight mileage over the next 10 or 15 years will create a
large absolute demand for gasoline and probably boost the annual rate of
increase to about 3.5%. In contrast to the high rates of growth
experienced in the past, however, the outlook is for an increase in demand
for gasoline and for total refined products more closely geared to the
rate of growth of U.S. economic activity.
3-287

-------
FEED
STREAM
MAKEUP
HF STREAM
	PRODUCT
" STREAM
DEFLUORINATORS
(ALUMINA)
HF REMOVED >>
AS SOLID AIF0 <2
/ v
ACID REGENERATOR
~BURN FOR FUEL VALUE,
FLARE TO ATMOSPHERE,
OR SEND TO SEPARATOR
FOR WASTE DISPOSAL
HF
ALKYLATION
UNIT
NOTE:
*THISlTAR STREAM IS VERY SMALL
(-0.4 BBl/pAY)'FOR AN AVERAGE UNIT.(4271)
Basis: 3800 bpd Production Unit (Fluoride Balance Only)
(A)
Process Streams - Tons/Yearv '

Stream Number
Material
1
2
~
3
HF
125^

122(b)
HF Equivalent

2.5*B)

Total Fluoride
125
2.5
122
Total as F
119
2.4
116
*Gaseous effluent stream
Soluble fluoride evolution factor = 0.18 IbF/bbl alkylate
(A)
v 'Assumes 330 operating days/year
^References 4239, 4240, 4241
Figure 3-54. Simplified HF Alkylation - Uncontrolled Process Model
3-288

-------
Total alkylate production in March 1971 was 775,000 bpd of which
236,000 bpd (30.4%) was produced by the HF alkylation process.^295^ If
a 2.5% growth rate Is assumed, the production rate (via HF alkylation)
will be 643,000 bpd in 2000. These data are presented in Table 3-100.
Table 3-100. HF Alkylation Projection

(42951 Estimated
197r ; Rate of Increase*
Estimated
Year 2000
Production
Alkylate produced
Utilizing HF
Catalyzed Process
(103 bpd)
236 2.5%
643
~Estimation is highly questionable because of possible changes in
energy sources for future transportation and octane requirements
if the source of energy is gasoline.
3.11.4 Fluoride Emission Control Techniques
The HF content of product streams 1s removed by passing those streams
through defluorinating towers packed with alumina. The fluoride is then
removed from the system and disposed of as AlF3,(4240,4247) a so1id that js
readily recovered, and easily sold. The bottoms product from the acid re-
generation unit which contains the bulk of the fluoride (98%) is sent to a
separator. The aqueous phase, containing about 70% of the HF (-16 tons/day
on an Industry basis) 1s generally pumped to a 1ime pit where the fluoride
is converted to CaF2 and disposed of in that form. ^ The remainder of
the HF in the acid regenerator bottoms product leaves the system in the or-
ganic phase from the separator.(4241) yh1s orgarr|c 11 quid containing approx-
imately 7 tons/day fluoride (on an industry basis) is either used for fuel
value in reboilers or furnaces at the refinery or blended into various pro-
duct streams.(42<*0,4241) jn either case, this fluoride (from the organic
phase) 1s eventually emitted to the atmosphere. A process flow diagram and
mass balance are presented in Figure 3-55. Since no definitive data could
be found regarding the degree of control utilized throughout the industry,
it was estimated that 25% of the production utilized lime pit disposal.
3 -289

-------
LIME
SEPARATOR
AQUEOUS
PHASE
LIME PIT
~ TO PLANT FOR
2??« HEAT CONTENT
PHASE OR BLENDING.
EVENTUALLY BURNED
AND EMITTED TO
ATMOSPHERE
SOLID CaF2
TO DISPOSAL
Basis: 3800 bpd Production Unit (Fluoride Balance Only)
(k)
Process Streams - Tons/Year^ ;
Material
Stream Number
3
4
5*
HF
122(c)
86^
36
Total as F
116
82
34
~Gaseous effluent stream
(A)
'Assumes 330 operating days/year
^Reference 4241
^References 4239,4240,4241
^Estimate usage of control process on 251 of
production (based on facilities located in high
density urban areas)

Soluble Fluoride

Emission Factor
Source
IbF/bbl Alkylate
Control Process
0.05
Assumed Fugitive
0
Total Emission
0.05
Overall soluble fluoride emission
factor(O) = 0.15 IbF/bbl Alkylate
Figure 3-55. HF A1kylation - Controlled Process Model
3-290

-------
3.11.5: Fluoride Emissions
Various sources of information indicate the levels of HF
(4239
consumption or loss to be in the range of 0.05 to 0.8 pound acid x '
4240 4241)
' ' per barrel of alkylate produced. It was found that most
producers assume 0.21 pound acid per barrel of alkylate to be the correct
(4241 )
value for most currently operated large units. ¦ This corresponds
to a total industry HF consumption (or loss) of 23.6 tons/day. Industry
(4240 4241)
sources ' indicate that approximately 2% of this amount
(0.6 ton/day) is lost to the product stream while the remaining 98%
(23.0 tons/day) exits the processes as a bottoms product from the HF
acid regeneration units. This bottoms product is either flared directly,
burned for fuel value or sent to a separator. In the latter case the organic
phase is recycled, and the aqueous acid phase sent to lime pits for disposal.
The fluorides contained in the product gasoline (200 tons per year) are dis-
charged, as the result of combustion, in the form of HF.
Current industry practice produced soluble fluoride emissions at
the refineries of 5800 tons in 1971 and will produce 15,900 tons (as F ) in
2000 if current abatement procedures are followed.* If 99% efficiency
control devices are utilized industry-wide, the 2000 projection is 190 tons.
These data are presented in Table 3-101.
3.11.6 Economic Analysis
Basic Process
Table 3-102 presents estimates of the current economics of production
of alkylate by the use of HF. Return on investment is estimated as ranging
from a negative value to 8.3%. The systems operate at a loss except for the
5000 bpd facility.
Impact of Control
Table 3-103 presents estimates of the cost of control of lfuoride
emissions from HF alkylation.
~Assumes that 25% of production utilizes lime pits for partial control
acid wastes.
3 -291

-------
Table 3-101. Soluble Fluoride Emissions from HF Alkylation
1971	2000 ,
HF Catalyzed Alkylate Production	236	643
(103 bpd)
Soluble Fluoride Evolution Factor	, 0.18	0.i8
(lbF/bbl Alkylate)
Soluble Fluoride Emission Factor with	0.15	0.15
Current Control (lbF/bbl Alkylate)*
Soluble Fluoride Emission Factor with	—	0.0018
99% Control (lbF/bbl Alkylate)
Soluble Fluoride Evolved	7.0	19.0.
(103 tons F/year)
Soluble Fluoride Emissions with Current	5.8	15.9
Practice (103 tons F/year)
Soluble Fluoride Emissions with 99%	—	0.19
Control (103 tons F/year)
~Estimated usage of control process is 25% of production (based on
facilities located in high population density urban areas).
3-292

-------
Table 3-102. Estimated Economics of HF A1kylation Units
(Pollution Control Cost Excluded)
	Plant Capacity (bpd)	
500	3800	5000
Total Capital Investment	0.5 $MM	1.3 $MM	1.5 $MM
Operating Costs
Direct Costs
Olefin Feed (1.1 bbl olefin/bbl Alkylate 9 $3.00/bbl)
Hydrogen Fluoride (0.2 lbs H,F,/bbl Alkylate 0 $.40/1 b)
Alumina (0.004 lbs Al20,/bbrAfkylate @ $0.10/1b)
Steam (700 lbs/bbl Alkylate at $0.45/1000 lbs)
Electricity (5.8 kwh/bbl 9 $0.007/kwh)
Labor (0.14, 0.1 and 0.09 man-hr/bbl i $5.00/man-hr)
Supervision and Fringe Benefits
Maintenance and Supplies
Total Direct Costs
3.30 $/bbl Alk.	3.30 $/bbl Alk.	3.30 $/bbl Alk.
0.08	0.08	0.08
(a)	(a)	(a)
0.32	0.32	0.32
0.04	0.04	0.04
0.70	0.50	0.45
0.70	0.50	0.45
0.18	, 0.06	0.05
5.32	4.80	4.69
Indirect Costs
Depreciation
Interest (at 7%, 201 Debt)
Taxes and Insurance
Plant and Labor Overhead
Total Indirect Costs
N-Butane Credit (8 lbs/bbl Alkylate at 0.01 $/lb)
Manufacturing Cost
General and Sales Expenses
FOB Cost
Product Revenue
Profit After Tax (@ 50% tax)
Cash Flow ($1000/year)
ROI {%)
0.30	0.10	0.09
0.02	0.01	0.01
0.04	0.03	0.03
0.84	0.60	0.54
1.20	0.74	0.67
(0.08)	(0.08)	(0.08)
6.44	5.46	5.28
0.13	0.11	0.11
6.57	5.57	5.39
5.50	5.50	5.50
(1.07)	(0.07)	0.06
(127)	38	239
8.3
(a)	$1100/year or 0.0009 $/bbl
(b)	Assumes 80% equity Funding

-------
Table 3-103. HF Alkylation - Estimated Economics of Control Process
Basis - 3800 bpd of alkylate produced.
GJ
I
ro
UD
Item
Number
Capital Cost Estimates ($1000)
Description
SEPARATOR, 0.5 lb HF/min, 8 bpd,
0.1 lb hydrocarbons/min, 100 gal
capacity,neoprene lined steel
LIME PIT, 5000 gal capacity,
0.36 lb HF/minineoprene lined steel
tquipment
F.U.B.
Cost
10
Reference
Number
4383
4383
Installation
Factor
1.50
1.50
Equipment
Installation
Cost	
15
Capital Subtotal	18
Indirects (0 15%) 3
Contingency (@ 20%) 4_
Total Capital (as of January 1971)	25
Operatinc
Item
Number
Cost It / hr)
Power
Cost
0.05
0.05
Maintenance
Cost
0.15
0.38
Subtotal
Water^ ^
Disposal(22)(
CaC03 (100 lb/hr
Total Operating Cost
TquTpmenF
Operating
Cost
0.20
0.43
1 bbl tar/day
$3.00 ton)
0.63
) 0.08
0.15
0.86
All control economics footnotes are located in
Section 3.1.1, pages 3-10 and 3-11.
Total Operating Cost ($/hr)	0.86
Taxes and Insurance (2%, 330 days)	0.06
Capital (9.0%, 330 working days)	o.28
Pollution Control Cost ($/hr)	1.20
Pollution Control Cost ($/bbl)	Q.01

-------
3.12 HF PRODUCTION

-------
3.12 HF PRODUCTION
3.12.1	General
Manufacturing processes for fluorine, hydrogen fluoride, and
derivative chemicals differ in two important respects from all other
processes discussed in this report. First, the fluorine involved is a
portion of both raw material and product. This creates an economic
incentive to minimize fluoride losses. Second, the quantity and concen-
tration of toxic fluorides constitutes a potential hazard which requires
treatment to preclude adverse legal and regulatory action.
3.12.2	Industry Description
Figure 3-56 presents a flow schematic^^'^^'^^'^^ and
mass balance for the production of HF at the rate of 25 tons per day (50%
anhydrous HF, 25% each of 50% and 80% HF).
3.12.3	Production Trends
The historical growth of HF production has been 7.8% annually for
(4296)
the period 1959 to 1969.v	HF consumption for aluminum fluoride and
synthetic cryolite production (used in manufacture of aluminum) plus the
growing fluorocarbon market will tend to keep HF demand high. Expected
growth rates between 5% and 7% are seen during the next few years. (^96)
If these rates are extrapolated to the year 2000, HF production (as
anhydrous HF) will increase from 337,000 tons in 1970 to between 1,430,000
and 2,565,000 tons.
3.12.4	Fluoride Emission Control Techniques
Hydrofluoric acid plant stack gas control systems are normally
integral with the manufacturing process. Collection and transport systems
are in-line extensions of the production system to the fluoride effluent
control system. The spent charge from the kiln must be treated properly to
prevent evolution of residual HF, but collection systems are not typically
used for control of spent charge emissions.
Figure 3-57 presents a typical control installation using water
scrubbing for a 25 ton per day plant. Either wet or caustic scrubbers
are added as a final plant stage to act as a final HF removal step.
3-295

-------
SULFURIC
ACID (96%\
ASSUMES 50% ANHYDROUS HF,
25% EACH OF 50% AND 80% HF
* THE ADSORPTION SYSTEM USUALLY
CONTAINS 7 TOWERS; 5 WEAK ACID
AND 2 STRONG ACID ADSORBERS.
** FRACTIONATING COLUMN HAS
CONVENTIONAL KETTLE-TYPE
RE BOILER OPERATING AT
APPROXIMATELY 240°F.
*** TYPICAL PARTICLE SIZE:
1% ON 100 MESH
12% ON 200 MESH
30% ON 250 MESH
45% THRU 325 MESH
H2S04 RECYCLE
DUST SEPARATOR
h2so4
300° F
ROTARY
ILN
150°F
75% H2F2
ACID GRADE
FLUORSPAR***
^ * 30 TO 60
MIN AT 400°F
FUEL IN KILN
DUST TO ATM. DURiNG	CALCIUM
MOVEMENT OF POWDERED	SULFATE
FLUORSPAR	RESIDUE
50% H2F2 RECYCLE
*
0
"Z eo
Oy q
tS) Q
<
30° F
0£
LU
zO
so
CO U
Q *
—
Uui

HYDROFLUORIC
ACID (80%)
240° F
7 *
OOz
i—z 5
up: zd
u
50°F
LU
I 1 I -Jt
zO
O
u
ANHYDROUS
-HYDROGEN
FLUORIDE
40% H2F2
BASIS - 25 TONS/DAY HF PRODUCTION (ASSUMES 50% ANHYDROUS HF, 25% EACH OF 50% AND 801 HF)
PROCESS STREAMS -LBS/HR
Material
Stream Numbers
1
2
3
4
g(B)
6
7
8
*
9
10
HF
SiF4
CaFg
4100(s)^

20(s)^

40(s)(C)
510 (1)
510 (1)

26 (g)(D)
34 (g)(D)
1010(C)
Total Fluorides
4100

20

40
510
510

60
1010
Total as F
2000

10

20
480
480

50
960
CaCOg
SiOg
S
40^
40^
2(a)



20





H2S04 (98%)

5400








CaS04




7100





co2








16(g)

so2
h2o
Fuel
4
-------
WATER-

150°F

-TO STACK
SPRAY SCRUBBER
EFF. = 90% GASEOUS

TO NEUTRALIZATION
AND DISPOSAL
BASIS - 25 TONS/DAY HF PRODUCTION
(ASSUMES 1/2 ANHYDROUS HF, 1/4 50% HF AND 1/4 80% HF PRODUCED)
PROCESS STREAMS - LB/HR

Stream Number
Materials
9
11
12*
HF
26(g)
24(g)
2(g) (Est.)
sif4
34(g)
31(g)
3(g) (Est.)
Total Fluorides
60
55
5
Total as F
50
46
4
co2
16(g)

16(g)
h2o
3(g)
2.5 (al)(A)
0.5(g)
Approx. Total
Stream
80
60(A)
20
*Gaseous Effluent Stream
(A)	Plus scrubbing water.
(B)	Assumes 100% usage of scrubbers on all facilities.
Source
Soluble Fluoride Emission
Factor - lb F/ton HF
Scrubber
4.1
Assumed Fugitive
0.0
Total Emission
4.1
Overall soluble fluoride emission = 4.1 lb F/ton HF^
Figure 3-57. HF Production - Controlled Process Model
3-299

-------
3.12.5 Fluoride Emissions
Soluble fluoride emissions will increase from 700 tons (as F ) in
1970 to about 5330 tons in 2000 at current abatement levels. If 99%
control devices are employed, the fluoride emission will drop to 680 tons
in the year 2000. These data are summarized,in Table 3-104.
^•12.6 Economic Analysis
Table 3-105 presents the estimated economics of HF production at
three plant capacities (5, 25, and 80 tons per day). Returns on invest-
ment for the three plants prior to the use of fluoride controls are
"0.5%, 32,5% and 58.1%, respectively.
3.12.7 Impact of Control
Table 3-106 indicates the estimated costs of fluoride pollution
control for a 25 ton per day plant. Impact on R0I is estimated as a aROI
of about -0.5% for the 25 ton per day plant size.
Table 3-104. Soluble Fluoride Emissions from HF Production

1970
2000
HF Production
(106 tons/year)
0.34
2.60
Soluble Fluoride
Evolution Factor
• (lb F/ton HF)
52
52
Soluble Fluoride
Emission Factor
with Current Practice
(lb F/Ton HF)
4.1
4.1
Soluble Fluoride
Emission Factor with
99% Control
(lb F/ton HF)
—
0.52
Soluble Fluroide
Evolution
(103 tons F/year)
8.84
67.6
Soluble Fluoride
Emission with
Current Practice
(103 ton F/year)
0.70
5.33
Soluble Fluoride
Emission with
99% Control
(10J ton F/year)
—
0.68
3-300

-------
Table 3-105. Estimated Economics of Hydrofluoric Acid Production (excluding pollution control cos
5 Tons/Day
Plant Capacity
25 Tons/Day
80 Tons/Day
U)
CO
o
Capital Investment
Installed Capital^
Off Sites
Total Capital Investment
Production Costs
Direct Costs
Fluorspar (acid grade: 2.02 tons/ton 1002 H2F2 at ^6.39)
Sulfuric Acfd'6' (2.0 tons 100%/ton 100% HjFj at $12.86 $/ton)
Fuel (6720 std. cu ft/ton 1001 HjFj at $0.35yiOOD std. ft3)
Water (2200 gal/ton 100% H?F2 at S0.40/103 gal)
Electric Power (450 kwh/ton 100% HgFj at 0.007 $/kwh) '
Operating Labor (4 men/shift)
Supervision and Fringe Benefits
Maintenance and Supplies
Total Direct Cost
Indirect Costs
Depreciation (at 102)
Interest (at 7%, 201 Debt)
Local Taxes and Insurance
Plant and Labor Overhead
Total Indirect Costs
Total Manufacturing Cost ($/ton 1001 H^F^)
General and Sales Expenses ($/Ton 1002 Hg^)
FOB Cost ($/Ton 100% H^)
Product Revenue ($/ton 100% H^F^)
Profit After Taxes (at"50%)
Cash Flow ($MH/yr)
Return on Investment
(C)
1.4 $MM
0.6
2.0
113.91 $/ton H2F2
25.72
2.35
0.88
3.15
76.80
38.40
18.32
279.53
121.21
16.97
30.30
92.16
260.64
540.17
10.80
550.97
560.00
4.52 $/ton H2F2
0.21 $MM/yr
0.5%
3.2	%m
1.3
4.5
25.72
2.35
0.88
3.15
15.36
7.68
7.93
176.38
54.55
7.64
13.64
18.43
94.26
270.64
5.41
276.05
560.00
141.98 $/ton H2F2
1.6 $MM/yr
32.5%
6.9 $MH
2.8
9.7
113.91 $/ton H2
25.72
2.35
0.88
3.15
4.80
2.40
4.58
157.07
36.74
5.14
9.19
5.76
56.83
213.89
4.28
218.17
560.00
170.92 $/ton H2F2
5.5 $MM/yr
58.1%
(A) Capital for sulfuric acid plant not included.
(6) Sulfuric acid plant collocated.
(C) Assumes 80% equity funding.

-------
Table 3-106. Hydrofluoric Acid Production-Estimated Economics of Control Process
Basis - 25 tons per day of HF (assumes 50% anhydrous HF and 25% each
of 50% and 80% HF produced.
CO
I
CO
o
IN}
Item
Number
_Ca£UaJ_Cos^_Es^1mates_yn0002
Description
SPRAY SCRUBBER, 1 ft - 6 in.
diameter by 8 ft, monel clad,
50 ft^/min, 8 ft/sec max velocity,
2 gal/min, 2 in W.G.
Equipment
F.O.B.
Cost
Reference
Number
4383
4391
4392
Installation
Factor
1.77
Equipment
Installation
Cost	
Capital Subtotal
Indirects (9 15%)
Contingency (0 20%)
Total Capital (as of January 1971)
11
(a) $/ton HF
All control economics footnotes are located in Section 3.1.1, pages 3-10 and 3-11
Operating
Item
Number
Cost
4
/hr)
Power
Cost
0.01
Maintenance
Cost
0.25
Subtotal
121)
Water
Di sposal
(22)
I 2 gpm, 0 recycle
Total Operating Cost
Equipment
Operating
Cost
0.26
0.26
) 0.01
0.27
Total Operating Cost ($/hr)	0.27
Taxes and Insurance (2%, 330 days)	0.03
Capital (9.U, 330 working d'ays)	0-13
Pollution Control Cost {$/hr)	o.43
Pollution Control Cost ($/ton)	'0.41

-------
3.13 NONFERROUS METALS SMELTING AND REFINING INDUSTRY

-------
3.13 NONFERROUS METALS SMELTING AND REFINING INDUSTRY
3.13.1 General
The nonferrous metals smelting and refining industry produces
copper, lead and zinc by thermal processing of the metal ores. Most of
the ores processed in the U.S. are sulfide mineral concentrates,
separated by various roughing and flotation techniques from a wide range
of gangue minerals. The separation is of necessity incomplete, and a
portion of the gangue minerals accompanies the sulfide minerals through
the thermal processing. The gangue minerals frequently contain inor-
ganic fluorides (Table 3-107*). These fluorides are evolved as gaseous HF
in the high temperature zones - the copper reverberatory furnaces, lead
refining kettles, and zinc sintering machines and roasting furnaces - where
temperatures range from 1400° to 2400°F, and more than sufficient combined
hydrogen to satisfy the stoichiometry of the reaction is present. There
is currently no information available in the open literature on the
fluoride contents of the various ores and concentrates, and no data has
been published on fluoride emissions from U.S. smelters.
Sixteen of the 37 American copper, lead and zinc smelters have
(42711
by-product sulfuric acid plantsv ' abating sulfur oxide emissions
on portions of the smelter effluent stack gases. Each of these plants
has a humidifying tower, cooling tower and mist precipitator, where the
hot smelter exit gases are treated to lower the gas temperature and
remove the excess moisture and mist load. Theoretically, most of the
fluorides fed to the acid plant should be removed in these devices.
Practically, the presence of weak ^SO^ and SO^ in the solution will
raise the partial pressure of the HF sufficiently so that less than
maximum removal will take place.
Because of the uncertainties associated with fluoride emission
quantities in the nonferrous metal smelting and refining industry, no
economic analyses have been made of production costs for the industry.
Since there are no current processes used for fluoride emission control,
no analyses have been made of control costs.
*Table 3-107 appears at the end of Section 3.13
3-303

-------
3.13.2 Copper Smelting and Refining Industry
Copper bearing ore bodies are associated with a wide range of
fluoride containing minerals. The fluoride containing minerals vary
widely in F content and in distribution through the ore body. Concen-
tration by flotation does not separate the copper sulfide ore completely
from any associated fluoride minerals. Detailed analyses of the various
ores and concentrates produced in the U.S. are needed; these data are
not available in the literature.
Industry Description. The primary copper minerals are chalcocite
(CugS - 79.81 Cu), cha 1 copyrite (CuFe$2 - 34.61 Cu), bornite (Cu3FeS3 -
55.5% Cu) and covellite (CuS - 66.51 Cu). These occur in copper-bearing
ore bodies containing varied gangue minerals - typical ore bodies are
copper-bearing sulphides chiefly chalcopyrite or lean, copper-bearing
pyrite in igneous rocks; irregular masses of copper-bearing sulphides in
contract zones, associated with lime silicates; veins along faults, with
greater or less replacement and impregnation of the walls; lenticular or
pod-shaped bodies of pyrite or pyrrhotite, with chalcopyrite.
The industry typically concentrates as-mined ores by crushing,
grinding and flotation. The grind is kept at 60 mesh with rougher and
cleaning flotation. Carryover of fluoride-bearing minerals would depend
on their flotation characteristics as compared to the ore.
Ore concentrates may be fed either to a roaster, or as in
Figure 3-58, directly to a reverberatory furnace for smelting. Figure 3-
gives the process model and mass balances for a typical 230-ton (blister
copper) per day plant.
If roasting is employed, this first stage of the smelting process
operates at temperatures of about 1200°F. Little fluoride is volatilized
in this stage.
The reverberatory furnace melts the metal-bearing change and forms
the matte and slag. Typical operating conditions for the "reverbs"
.(4271 )
Furnace bath temperature
Dust load in offgas
SOg in offgas
2400°F
2-5 grains/scf
0.5-3.5$
3-304

-------
DUST RECYCLE
SILICA FLUX
BLISTER
COPPER
FEED
MATERIAL
MATTE
24,800 SCFM
DILUTION AIR
(THREE TIMES
OFFGAS FLOW)
COMBUSTION
FUEL AND AIR
SLAG
SLAG
640° F
750°F
COOLING
AIR 21400
SCFM
DUST RECYCLE TO
REVERBERATORY
FURNACE
TO STACK
OR SULFURIC
ACID PLANT
DILUTION AIR 20900 SCFM
450° F
DUST RECYCLE TO
REVERBERATORY FURNACE
TO STACK
FLUES AND DUST
COLLECTION
WASTE HEAT
BOILERS
CONVERTER
2200°F
FLUES AND DUST
COLLECTION
REVERBERATORY
FURNACE 2400°F
NOTE:
COPPER SMELTING PROCESS MODEL B (4271) WHICH INCLUDES A "ROASTER" AHEAD
OF THE "REVERBERATORY FURNACE," HAS ALMOST ALL FLUORIDE EVOLUTION
FROM THE REVERBERATORY FURNACE,* AND THE SOLUBLE F EMISSION FACTOR
IS IDENTICAL WITH THAT OF PROCESS MODEL B.
* TRW SYSTEMS ESTIMATE
Basis: Smelter Capacity of 230 Tons/Day of Blister Copper
Process Stream - Tons/24-Hour Day

Stream Number
Material
1
2
3
4
5
6
7
8*
HF







0.09^^
CaFg
0,18^^



0.00^

0.00^B)
-
Total fluorides
0.18



0.00(B^

0.00^
0,09^ ^
Total as F
0.09



0.00^

0.00^B)
0,09^^
Cu
234

23
0.7
253
0.7^
230
0.4(s)
Fe
228

227
15
212
0.7(s)

0.4(s>
S
258

4.4
0.2
163
159^
0.2
85^
Si02
92

190
185
6.4
0.9^

0,5^
CaO
45

5.6
4.3
1.3
-

0.1(s)
A12°3
30

8.7
8.7
-
-


Other
79

72
3.3
3.8
2.5(s^

1.5
Water
91

_





Air

36200^



64800(A)

' 90800^
Natural Gas

2800^






Approx. Total Stream
1056
39000^
530
217
640
64800(A5
230
90800^
Soluble F emission factor = 0.78 lb/ton blister copper produced
(A)	SCFM (32°F, 1 atm)
(B)	TRW estimate, in the absence of any reported data
*Gaseous effluent stream
Figure 3-58. Copper Smelting -Uncontrolled
Process Model A^71)
3-305

-------
The fluoride contained in the charge is evolved as HF under these
conditions, at equilibrium. The molten matte produced in the reverberatory
furnace is transferred to the converters whose function is to oxidize and
and separate the iron and sulfur from the matte. The oxidation reaction
is sufficient (when air blown) to maintain the converter at approximately
2250°F. S02 is emitted with the other flue gases. Silica flux is added
to combine with the iron oxide to form a fluid iron silicate slag.
Production Trends. Current (1967) copper metal production is
1.628 million tons per year. Assuming a 3% annual increase in production,
copper metal production in 2000 would be 4.2 million tons.
Fluoride Emissions. Because of the total lack of data in the
1iterature, it was necessary to estimate first the "average" fluoride
content of the concentrate fed to U.S. copper smelters; then, from this
information, the fluoride evolved in the reverberatory furnaces; and
finally the portion of the evolved fluorides captured in the by-product
sulfuric acid plant. The average copper mineral content (Cu, S and Fe)
of ore concentrates was estimated at 83.5%. The remaining 16.5% of the
concentrates is gangue minerals, with an estimated 650 parts per million
of fluoride* Based on these assumptions, and those noted above, soluble
fluoride emissions are estimated at 634 tons annually for 1967, and
projected to be 1638 tons in 2000 if current lack of control continues.
Fluoride Emission Control Techniques. There are no fluoride
emission control techniques currently employed in the industry.
3-13.3. Lead Smelting and Refining Industry
The major lead ore bodies are associated with a wide range of
fluoride containing minerals which include fluorite. The fluoride con-
taining minerals vary widely in distribution through the ore bodies.
Concentration by roughing and cleaning flotation does not entirely separate
the economic mineral from the associated gangue minerals, including the
fluoride minerals. As with copper, detailed analyses of the various ores
and concentrates produced in the U.S. are needed and are not available
in the literature.
*Based on the average fluoride content of the earth's crust (Ref. 4242)
3-307

-------
Industry Description. The primary economic lead mineral is
galena (PbS - 86.6% Pb). Other lead minerals, of secondary importance,
are cerussite (PbCO^ - 77.5% Pb); anglesite (PbSO^ - 68.3%); and
pyromorphite (3 Pb^PO^ 'PbClg - 76.2% Pb).
Galena comprises the vast majority of the ore mined. There are
three general classes of lead ore: (a) those containing lead alone as
an economic metal, (b) lead-zinc ores, (c) lead-silver ores; calcite,
dolomite and pyrite are the common gangue minerals of the first two
classes, quartz of the third class.
Concentration is normally accomplished by crushing, grinding (a
gravity separation is sometimes used at this stage) and, finally, by a
series of roughing and cleaning flotation steps. The concentrate consists
generally of the following range of compositions:
Component	Quantity, %
Pb	55 to 70
Zn	Up to 6.5
Cu	0.5 to 4.0
S	13 to 18.5
Fe	Up to 5
Silica,	lime, cadmium Minor amounts
silver, gold, arsenic,
fluorine
The first operation in lead smelting is sintering which roasts the
ore to remove the sulfur and at the same time produces a strong porous
mass suitable for the blast furnace. Silica and limestone are added as
fluxes which control the proportion of the sulfur in the mix and thus
keep the temperature below 1400°F. Some 85% of the sulfur is eliminated
in this step, 14% remains in the slag and other solid by-products and 1%
is eliminated by the blast and the dross furnaces.
3-308

-------
The purpose of the blast furnace is to reduce the lead oxide to
lead by carbon monoxide produced from coke added to the furnace. The
furnace temperature ranges from 1000° to 1200°F. Dust from this operation is
collected in a baghouse, and SO2 in the flue gas ranges from 0.01 to 0.25%.
Some flux (silica or limestone) may be added to the furnace charge. The
products of the blast furnace are:
•	Lead metal
•	Matte	44-62% copper 10-20% lead
13% sulfur ^2% zinc, iron and silica
•	Speiss	55-65% copper, 8-18% lead, sulfur,
arsenic, zinc, iron and silica
•	Slag	Siliceous with 10-20% zinc ^2% lead
and ^3% sulfur and some iron and
sulfur
The fluorides contained in the sinter feed are not decomposed
extensively in the sintering furnace or blast furnace. They are, however,
decomposed at dross reverberatory furnace temperatures to a greater
extent. The dross reverberatory furnace and refining kettles, which
operate at gas temperatures from 1400° to 1800°F, constitute the lead
refining process. The fluorides contained in the charge, would, based
on thermochemical equilibria studies, be evolved quantitatively as
gaseous HF at the upper temperature.
Figure 3-59 presents a process model and partial mass balance for
a 271 ton (lead bullion) per day lead smelting plant.
Production Trends. Current (1967) production of lead bullion in
the U.S. is 1.24 million tons. Assuming a 2% annual increase in produc-
tion, lead bullion production in 2000 would be 2.4 million tons.
3-309

-------
©	CD
ORE AND LEAD	ZINC PLANT
RESIDUE
COKE BREEZE
CONCENTRATE
AIR AND
NATURAL
GAS
NATURAL AIR
GAS
NATURAL GAS
AIR —|
i-COKE
r FLUX
AIR —|
MATTE
•AND
SPERSS
RECYCLE
SLAG
LEAD©
DUST
COLLECTION
DUST
COLLECTION
DUST
COLLECTION
SINTERING
MACHINES
1200°F
DROSS REVERB
1600°-1800°F
TO ATMOSPHERE	TO ATMOSPHERE
OR SULFURIC ACID PLANT
Basis: Smelter capacity of 271 tons/day lead bullion
process streams - tons/24 hour day


Stream
Number


Material
1
2
3
4
5
6
7
8
9*
HF








0.049
CaF2
9.096








Total
0.096







0.049
Fluorides









Total
0.047







0.047
as F









Pb
269
2

4

4
271
0.8

Cu







1.8

As







0.2

Sb







0.1

Fe







0.4

Slag



128

161



Other
91
46





0.7

Flux


60






Coke




66



72,000^)
Offgas








Approx.
360
48
60
132
66
165
271
4
	
Stream









Total









Soluble flouride emission factor = 0.34 lb/ton lead produced
(A) SCFM at 32°F, 1 atm
Figure 3-59. Lead Smelting - Uncontrolled Process^^
3-310

-------
Fluoride Emissions. As with copper, the total lack of data in the
literature made 1t necessary to estimate "average" fluoride content of
concentrate charge to American lead smelters, fluoride evolved in the
kettles and dross "reverbs," and overall fluoride emission factors. The
average lead mineral (Pb, Zn and S) content of lead ore was estimated at
80%. The remaining 20% of the concentrates is assumed to be gangue
minerals, with a fluoride content of 650 parts per million. Based on
these assumptions and the others noted above, soluble fluoride emissions
are estimated at 210 tons for 1967, and projected to be 408 tons in
2000 if current lack of control continues.
Fluoride Emission Control Techniques. There are no fluoride
emission control techniques currently in use in the U.S.
3.13.4 Zinc Smelting and Refining Industry
The major primary ores of zinc are associated with a wide and
varying range of fluoride containing minerals. As noted with copper and
lead, the fluoride containing minerals vary widely in distribution through
the ore body. The normal concentration steps employed for production of
zinc concentrate do not separate the zinc sulfide ore completely from the
other minerals. Detailed analyses of the various ores and concentrates
produced in the U.S. are needed and are not available in the literature.
Industry Description. The primary zinc minerals are sphalerite
(ZnS - 67% Zn), hemimorphite (2 Zn O'HgO'SiOg - 54.2% Zn) and smithsonite
(ZnCO^ - 52.1% Zn). Other minerals such as willemite, zincite and
franklinite form a separate group found at Franklin Furnace, New Jersey.
Neither lead nor zinc deposits have been found in immediate association
with igneous rocks; they reach their places of precipitation in solution.
The fluoride-bearing gangue minerals associated with the zinc bearing ore
bodies are listed in Table 3-107.
Concentration of as-mined ore follows a typical sequence of
crushing, gravity concentration, grinding and flotation. Assays of zinc
concentrate (mainly from flotation systems) cluster in the fifties, with
the majority below 55% Zn.
3-311

-------
Zinc smelting follows an extraction process of roasting, sintering
and calcining. Some plants both roast and sinter zinc sulfide concen-
trates before extraction. Calcining is performed only on oxide ores or on
material that has previously been oxidized by roasting.
Zinc sulfide concentrates are usually converted by a roasting
process to zinc oxide by any one of a number of types of roasters which
typically remove 93 to 97% of the sulfur at temperatures that range
between 1200° and 1900°F. At the higher temperatures, much of the
fluorides present are driven off as gaseous HF. The Table 3-108 shows the
temperature ranges of typical zinc roasting operations.
Table 3-108. Typical Zinc Roasting Operations (4271)
Operating
Type of Roaster	Temperature, °F
Multihearth	1,200-1,350
Multihearth ^	1,600-1,650
Ropp^	1,200
Fluid bed^
(Dorr-01iver)	1,640
Fluid bed^
(Dorr-Oliver)	1,650
Fluid bed
(Lurgi)	1,700
Suspension	1,800
F1 uid column	1,900
^ ^Dead roast except where noted otherwise.
(2)
'First stage is a partial roast in multihearth, second
stage is a dry-feed dead roast in Dorr-Oliver fluid bed
^Partial roast
^Slurry feed
3-312

-------
Sintering is used mainly to agglomerate a roaster calcine for
subsequent processing. Operating temperature is typically 1900°F. Feed
for the sintering operation is a mixture of calcine or concentrates,
recycled ground sinter, and the required amount of carbonaceous fuel of
proper particle size and moisture content for pelletizing. Those
fluorides not volatilized in the roasting furnace are at least partially
evolved in the sintering machine.
Calcining is a heat-treating process that is used for oxidized
materials such as oxide ore concentrates or material from roasting of
sulfide ore concentrates. It may be called nodulizing, since hard nodules
of random sizes are produced when the calcining is done in a rotary kiln.
The nodulized kiln product is subsequently treated for zinc extraction.
Roasting, sintering, and calcining are preliminary steps to one of
the extraction methods; pyroreduction or leaching and electrolysis.
Pyroreduction distillation or retorting of the sinter or calcine
is performed in horizontal or vertical retorts, electrothermal open or
submerged arc furnaces, or blast furnaces. Horizontal retorts are small
ceramic cylinders that are mounted horizontally in racks that hold several
rows of retorts mounted one over the other. They are fed with coal and
sinter and produce liquid zinc metal as do the larger and more modern
vertical retorts.
Figure 3-60 presents the process model and mass balances for a
423 ton per day (zinc metal basis) smelter.
Production Trends. In 1968, primary slab zinc plants were operated
in 14 locations with a capacity of 1.3 million tons and a production of
1.07 million tons. At the 2,5% rate of increase estimated for zinc
production, U.S. zinc production in 2000 would be 2.4 million tons.
Fluoride Emissions. The procedure employed to develop estimates
of fluoride emission for zinc smelting and refining is similar to
those used for copper and lead production-associated fluoride emissions.
The average zinc mineral content of zinc ore concentrates was estimated
at 88%. The remaining 12% gangue minerals were assumed to have a
fluoride content of 650 parts per million. On the bases of these
assumptions and the others noted above, soluble fluoride emissions are
3-313

-------
estimated at 246 tons per year for 1967, and projected to be 550 tons
per year in 2000.
Fluoride Emission Control Techniques. There are no fluoride
emission control techniques currently employed in the industry.
WATER
AND COAL
NATURAL GAS
AND AIR
AIR
AIR
COAL
BINDER
COKE
DRYV
FEED'
. ZINC
PRODUCT
RESIDUE •
TO ATMOSPHERE
TO ATMOSPHERE
TO ATMOSPHERE
TO ATMOSPHERE
ROASTING
FURNACE
1800°F
DUST
COLLECTION
COKING
FURNACE
SINTERING
MACHINE
1900°F
DUST
COLLECTION
RETORT
OR SULFURIC ACID PLANT
Basis: Smelter capacity of 423 tons/day of zinc metal
process streams - tons/24 hour day




Stream
Number



Material
1
2
3*
4
5
6
7
8
9*
10
11
HF


0.065





0.037


CaF2
0.12



0.009


0.060



Total Fluorides
0.12

0.065

0.009


0.060
0.037


Total as F
0.062

0.062^ .

0.005


0.030
0.035


Zinc
432
432








423
Sul fur
230
9









Other
58
96









Coal




49


304



Coke







122



Binder







30



so2


443








Dust


0.8
72

53
0.6




Water
1



51






Air (B)


44,300(A)



73,62o'A'

58,520(A)
44,48o'A'
Natural Gas











Approx. Total Stream
721
537
-
72
110
53
-
456

-
423
Soluble F emission factor = 0.46 lb/ton of zinc produced
(A)	SCFH at 32°F, 1 atm
(B)	Air defined as nitrogen, oxygen and carbon dioxide
(C)	Ultimate fate of HF in acid plant undetermined
Figure 3-60. Zinc Smelting - Uncontrolled Process^^
3-314

-------
Table 3-107. Gangue Minerals
Note: The 78 minerals listed, which vary from common to extremely
rare, are found in association with various ore bodies containing
copper, lead and zinc ores of significance. The content of individual
minerals contained in ore bodies, as well as fluoride content of any
individual mineral, varies widely from undetectably low levels to
major contamination. There is a variation in mineral content
associated with ore body geology, as well as the variation in mineral
content association with differences in ore bodies and ore types.
This list was taken from Orsino C. Smith, Identification and
Qualitative Chemical Analysis of Minerals, D. Van Nostrand Company,
New York City, New York 1946 (Reference 4272).
Common	Uncommon
Pyrochlore Na, Ca, Cb9 CL.F	Cordylite Fluocarbonate of Ce
Metals and Ba
Topaz fll203 '(OH.FJ'SIOj	Cardylite BaF2"Ce203'C02
Chondrodite 4Mg 0'2S102 'Mg (F,OH)2 chalcolamprite Na4(CaF)2 Cb2
Fluorapatite 9Ca0'3P20g'CaF2	Si 0g
Apatite 3Ca3(P04)2*Ca(F,Cl)2
Fluorite CaF2
Cryolite 3NaF'AlF3
Zunyite A1203* Si02 *A1(0H,F,C1)3
Lepidolite (K,Li)20 * A12O3"3Si 02
with F
Uncommon
Matlockite PbF, CL
Bastnaesite (Ce,La,Di)F.C02
Marignacite Variety of Pyrochlore
Durangite NaF.AlAsO^
Parisite 2(Ce,La,Di ,Th) 0F'Ca0'3C03
Triplite (Fe,Mn) FP04 with Ca
and Mg
Metajarlite NaSr^Al^F-jg
Magnesium - Orthrite
7 [(Mg,Fe,Ca)0 + (Fe,Al,Ce,
Cb,La)203], 6Si02'H20 +F
Fersmannite
8(Ca,Na2)(0,F2) 4Ti02'3Si02
Svabite 9 Ca 0'3 (As205'P205)
Ca(F"0H)2
Fermorite (Ca,Si)0*(P,As)205*
CafOH.F^
3-315

-------
Table 3-107. Garigue f
Uncommon
Yttrofluorite (Ca3Y2)Fg
Metatriplite 6 MnO'SP^O^'
2 (Mn,Ca)F2'4H20
Montebrasite A12O3" P2(-)5" (OH,F)
Ephesi te (Na,Ca,Li4 2^10
(o,oh,f)2
Norbergite 3Mg0"Si02'H20 +F
Edenite 8Ca0'2Na20'18MgO"
4Al203*26Si02'H20'3F2
Meli phani te 2CaO'2BeO'3Si 02'NaF
Sellaite M F,
9 2
Herderite Ca(F,0H)2'Ca0'2Be0*P205
Tilasite 2Ca0'Mg0"As203'MgF2
Reddingite 4P205'3H20 +F
Hamlinite PO^ of A1 and Ba with
H20 and F
Quercyi te 6CaO'2P205 * 2CaO"2C03'CaF2
Francoli te 1OCaO* 3P205 * CaF2 *C02 * H20
Lecroixite 2(Na,F,0H)* 2(Mn,Ca)0*A12i
•p2°5'H2°
Pachnolite NaF, CaF2, A1F3"H20
3-3
nerals (Continued)
Uncommon
Zinnwaldite K20"Li20"FeO"F2'
2Al203*6Si02*H20
Thomsenoli te NaF'CaF2'A1F3'H20
Ellestadite CaO, S03, Si02>
P205V C02, CI, F
Manganapatite 9(Ca,Mn)0"3P205"
Ca(OH,F)2
Magnophorite Ca,Na,K,Mg ,Fe,Ti,
Mn,Si ,A1,Ti,0,
0H,F
Cuspidine 3Ca0'CaF2'2Si02
Wagnerite Mg3(P0^)'MgF2
Weberite Na^gAlF^
Villiaumite NaF
Fluoborite 6Mg0'B203'3(H20,F2)
Creedite Ca0*2Al(F,0H)3*2CaF2*
S03*2H20
Pachnolite NaF'CaF2'AlF3*H20
Zeophyl1i te 3CaO'CaF2'3Si 02"H20
Custerite 3Ca0'CaF2'2Si02'H20
3 Prosopite CaF2'2Al2(0H,F)3
Bulfonteinite CaSi02*(0H,F)^

-------
Table 3-107. Gangue
Uncommon
Jezekite Ca0"Al203'2(Na,Li)F*
P205*2(Na,Li)(0H)
Leucophanite NaF'Ca0'B203'2Si02
Morinite 3Al203'2Na20*4P205"
6CaF2'18H20
Chiolite 5NaF"3AlF3
Cryolithionite 3NaF * 3Li F * 2A1F^ *
2K20'10(Mg,Fe)0
Si 1icomagnesiofluorite
^Ca^Mg^Si2^7^10
Polylithionite (Na, K)3Li&A12
Si8°22F8
Gearksutite CaF2'Al(F,0H)3'H20
Kurskite 2Ca3(P04)2'CaF2'CaC03
Nocerite 2 MgO'MgF2"CaF2
Hieratite 2KF"Si F^
Minerals (Continued)
Uncommon
Cryophyllite 3(Li ,K)20'2Fe0"
4A1203 *20Si02•
3 H20'8(Li,K)F
Malladrite 2 NaF'SiF^
Leifite Na20'Al203'9Si02'2NaF
Ralstonite (Mg ,Na2)F2'3Al(F,0H)3*
2H20
Sulphohalite 2Na2S04*2NaCl*NaF
Schairerite Na2S0^'Na(F ,C1)
Minyulite 2K(0H,F)*2A1203*
2P205'7H20
Ferruccite NaBF^
Avogadrite KBF^ + 10% CsBF^
Fluellite A1F3'H20
Cryptohalite 2NH^F'SiF^
3-317

-------
3.14 OTHER INDUSTRIES

-------
3.14 OTHER INDUSTRIES
A number of miscellaneous processes which emit relatively small
quantities of soluble fluorides are covered in this section. These are
fluorine production, fluorocarbon chemical production, uranium fluoride
production, aluminum anodyzing, and beryllium production. Because of
the small quantities of fluorides emitted, no process control models, mass
balances or economic analyses were developed.
3.14.1 Fluorine and Fluorocarbon Chemicals
General Discussion. Manufacturing processes for fluorine and
derivative chemicals as noted earlier are different in two important
respects from all other processes discussed in this report. First, the
fluorine involved is a portion of both raw material and product. This
creates an economic incentive to minimize fluoride losses. Second, the
quantity and concentration of toxic fluorides constitute a potential
hazard that requires special consideration to preclude adverse legal or
regulatory action.
A literature survey indicates that the amount of fluoride emitted
to the atmosphere from the fluorine and fluorocarbon manufacturing
(4242 4244)
processes is very small.^ ' ' Because of the highly toxic and
corrosive nature of the feed or product materials (HF, F^)» extreme
care is taken to control spills and leakage. Furthermore, gaseous
effluent streams are generally scrubbed to remove all but trace quantities
of fluoride compounds.
In fluorocarbon production, there are no gaseous effluent streams
and all product streams are scrubbed to meet purity specifications. In
these processes, the unreacted HF is removed as solid CaF? and disposed
of in that fo™.<4242-4243)
Although definitive emission data have not been found, it is
probable that the fluorine chemical industry does not present a significant
fluoride emission problem, because of the high abatement efficiencies
obtained by the use of the caustic scrubbers.
3-318

-------
Figures 3-61 and 3-62 present flow diagrams for fluorocarbon
production and for fluorine production, with control systems included as
integral process elements.
3.14.2 Uranium Fluoride Production
General Discussion. UF^, a solid below 969°C at 1 atm, is formed
by reacting UO2 with an excess of HF at 550°C. The solid UF^ is then
reacted at 250°C with F2 produced on site to form UFg (sublimation point
56°C at 1 atm) which is then fractionated from any residual HF and F2.
Production and Fluoride Emissions. Virtually all the UF^ and UFg
produced in the U.S. is manufactured in one plant operated by the Allied
Chemical Company, Metropolis, 111., on the Ohio River. This plant, which
resumed production in 1968 after a 3-year shutdown, has a capacity of
100,000 tons UFg/yr. (714,4233) /\nnuai hF production in the U.S. is
approximately 300,000 tons, of which the fraction used in atomic energy
is under 10% and decreasing, so that the maximum consumption at Metropolis
is 30,000 tons HF/yr.(4233»4234)i
Although no information was found that applied directly to the Metro-
polis plant, information was found on two UFg conversion plants which are
now closed - the National Lead Company facility at Fernald, Ohio, on the Miami
River, and the Union Carbide Corporation facility at Paducah, Kentucky, up-
stream from Metropolis, 111. on the Ohio River.(714) The National Lead facil-
ity recovered 85% of its HF effluent as 70% hydrous solution, which it sold
in bulk. Union Carbide claimed a 95% recovery as the 70% hydrous solution.
National Lead tried scrubbing the remaining 15% with Ca(OH)2 or KOH. The
KF was sold and the CaF2 was buried. The Ca(OH)^ scrubbing proved to be
less expensive.
Unprecipitated fluoride which passed through the Fernald scrubber
was stored and released into the Miami River on a schedule such that the
concentration in the river water never exceeded 0.8 ppm.^7^ For the
first 10 months of 1955, the plant discharge into the river averaged
of 16.1 tons F"/month, which extrapolates to 194 tons F~/year. Assuming
that the Fernald and Paducah plants shared the production equally and that
3-319

-------
hydrogen fluoride -
CARBON TETRACHLORIDE-
DISTILLATION
COLUMN
HYDROGEN
CHLORIDE •
ABSORBER
WATER
M
REACTOR
SbC#.
CATALYST
DICHLOROD1FIUORO METHANE <
MONOFLUOROTRICHLOROMETHANE «

DRYERS
CAUSTIC
SCRUBBERS	ACID
(2)	SCRUBBER
SODIUM	SULFURIC
HYDROXIDE ACID
HYDROCHLORIC SPENT WASH, SPENT WASH,
ACID, TO DISPOSAL 'TO DISPOSAL TO DISPOSAL

5
DISTILLATION COLUMNS COMPRESSOR
Figure 3-61. Fluorocarbon Production
VAPORIZER
ELECTROLYTIC	TO SCRUBBER
CELLS	AND FLARE CONDENSORS SURGE TANK
HYDROGEN
FLUORIDE 1
FLUORINE TO
LIQUEFACTION
AND STORAGE
TO ATMOSPHERE
CAUSTIC
AIR 	*
FUEL
DILUTION AIR
TO DISPOSAL
TO DISPOSAL
SCRUBBER
BURNER
NaF ABSORPTION
AND DESORPTION TOWERS
Figure 3-62.
Fluorine Production
3-320

-------
the Metropolis plant now operates under the same conditions, it is
currently dumping approximately 400 tons F~/year into the Ohio River if
it is as efficient as the Fernald plant or 133 tons F"/year if it is as
efficient at the Paducah plant.
Furthermore, the Fernald plant buried 2500 tons HF/yr as CaF2,^71^
which would translate to 5000 tons/year for the Metropolis plant operating
at the same recovery rate. Assuming a 99% efficiency for the Ca(0H)2
scrubber, the II being vented through the stack amounts to 55 tons HF/year.
For a 99.51 scrubber efficiency, 27 tons HF/year is being vented. If the
Metropolis plant operates with the recovery rate of the Paducah plant, the
atmospheric emissions are 18 tons HF/year and 9 tons HF/year for scrubber
efficiencies of 991 and 99.5%, respectively.
3.14.3 Aluminum Anodizing
General Discussion. Mixtures of HF and HNO^ are used in one process
for cleaning A1 alloys prior to anodizing or conversion coating with
phosphate or chromate. The pretreatment typically consists of dipping
batches of parts in a tank of solution containing 50 to 751 HNOj and 2 to
20% HF for 1 to 5 minutes at room temperature. If the A1 alloy is
particularly high in Si, a 3:1 HN03:HF solution might be used for 30 to
(42371
60 seconds at room temperature.v '
Dragout of the acid is considered economically undesirable for
two reasons: (1) it wastes acid, and (2) it ruins the subsequent solutions
and the final product. For this reason, all parts are carefully rinsed
with cold water, and the wash is recycled to the acid bath. Baths are
used over and over again until flocculation becomes excessive.(4236,4237)
For conversion coating, baths containing only 0.6% F ion are used.
(4236 42371
Again, as with cleaning baths, they are used over and over again.v ' '
Production and Fluoride Emissions. The anodizing industry is very
fragmented and consists of many small job shops, 58 of which are in the
Los Angeles area alone. A spot telephone check of five shops chosen at ,
random, showed an average consumption of 0.285 ton HF/year/plant with a
maximum 1.27 tons.The largest user employed an alkaline scrubber
mounted on the fume hood; the others had no controls for atmospheric
emission. Since the Los Angeles Air Pollution Control District is one of
3-321

-------
the strongest pollution control agencies in the country, it may be assumed
that these represent the most stringent control conditions.
An extrapolation on a population basis indicates that the total
national use of HF in the anodizing industry amounts to 668 tons per
year. The worst possible case from a pollution standpoint would be total
volatilization as HF and uncontrolled venting of the HF. This would
correspond to an annual national evolution of 668 tons HF (635 tons as
soluble fluorides) by 2320 plants, or an average 0.29 ton HF/year/plant.
3.14.4 Beryllium Production
General Discussion. There are two beryllium producers in the U.S.,
Brush Beryllium and Kawecki-Berylco (formerly the Beryllium Corporation
of America). The production is in three stages: (1) the naturally occur-
ring beryl (BeO.A1203.6Si02) is converted to Be(0H)2, (2) the Be(0H)2 is
converted to BeF«, and (3) the BeF9 is reduced to Be metal. The total
/	*(4154)
annual production is 90 tons (as Be metal). '
There are two processes currently in use for the production of
Be(0H)9 - the fluoride process used by Kawecki-Berylco and the sulfate
(4154)
extraction process used by Brush Beryllium.v ' Only the fluoride
process involves fluoride and consequent fluoride emissions. The reactions
are:
(1)
(2)
(3)
2Na3FeF6+3Be0.Al203.6Si02		 3Na2BeF4
+Fe203 + +3A1203 + +18Si02 *(750°C, dry)
3Na2BeF4+6Na0H——- 3Be(0H)2 * +12NaF
12NaF+Fa2(S04)3	- 2Na3FeF6 * +3Na2S04
The sodium fluoferrate is largely recovered and recycled. Sane is lost,
however, by volatilization as FeF3, A1F3 and SiF^ in reaction (1), by
entrainment in the discarded precipitated oxides resulting from the
leaching of the products of reaction (1) with water, by entrainment with
the Be(0H)2 precipitate of reaction (2), and by solution in the discarded
Na2S0^ filtrate of reaction (3). If an overall 80% recovery efficiency is
assumed for the Na3FeFg, and if Be(0H)2 production by the fluoride process
accounts for half the total, the maximum amount of fluoride released by
3-322

-------
uncontrolled emission to the atmosphere, in precipitates and in effluent
streams would amount to 93 tons per year as soluble F.
The Be(0H)2 obtained by the fluoride process is further treated by
dissolving in sulfuric acid, adding organic chelating agents such as the
Na salt of ethylene diamine tetraacetic acid (EDTA), and neutralizing with
NH^OH to reprecipitate a highly purified Be(0H)2. The Be(0H)2 filter eake
containing 50% free water is reacted with 35% excess NH^HFg to yield BeFg
after dehydration on a drum dryer. The 35% excess fluoride is volatilized.
(4154)
' If this volatilization is uncontrolled, the maximum total atmos-
pheric emission is 74 tons/year as soluble fluoride
The BeF2 is converted to Be metal by reduction with Mg.^4154^ By this
stage, the toxicity of the materials because of their Be content far over-
shadows the fluoride problem, so that adequate effluent control is provided.
The maximum fluoride emission that can be expected, therefore,
from the beryllium industry is 167 tons per year as F, assuming 20% of
the Na^FeFg input is lost to recovery and recycling and the volatilization
of the ammonium fluoride species in the BeF2 production is uncontrolled.
3-323

-------
4. Research and
Development' Planning

-------
4. RESEARCH AND DEVELOPMENT PLANNING
4.1 SUMMARY AND PRIORITIES
Section 3 discussed industrial sources of fluoride emissions to the
atmosphere. In many of the industries discussed, definitive information
related to emission levels is not available, i.e., input fluoride concen-
trations, output quantities, release mechanisms, etc. Research and develop-
ment (R&D) projects have been planned and prioritized which will expand the
informational base and provide the knowledge and methodology required to
improve fluoride emission control.
Table 4-1 presents a summary of the recommended projects by industry
and objective. Table 4-2 presents the relative priority and time phasing
of each project together with the approximate rate of expenditure required
to fund all programs. Table 4-3 summarizes the manpower, cost, and calen-
dar time requirements of the programs in the same format as Table 4-1.
An examination of Table 4-1 shows projects which are primarily
oriented toward collecting information on fluoride inputs, and improving the
level of understanding of the process mechanisms releasing and collecting
fluorides. The single exception is development of control processes for the
ion and steel industry. This emphasis resulted from the current lack of
information of this type for the indicated industries. Once the process
characteristics are known, existing control device technology will probably
be applicable. The recommended projects have been planned in reasonable
detail and are presented in the following section. Additional R&D work will
be required for each industry, based on the results obtained from the recom-
mended projects, to apply the knowledge gained to development and economic
evaluation of applicable control processes. Since the specifics of the addi-
tional work will be determined by the results of the recommended projects,
detailed plans have not been formulated. It is anticipated that in each
case a project will be required that is roughly similar in approach, resources,
and time to that recommended for development of control processes for the
!iron and steel industry.
The prioritization shown in Table 4-2 is based on the ouanitity of
soluble fluorides emitted from a particular industry7~both-current and
projected, and the state of knowledge of fluoride emission in the industry.
4-1

-------
Table 4-1. Fluoride Emission Control - Recommended Research and Development Projects by Industry
J^ecOypT"
Industry
Feed/Ore Fluoride
Content Characterization
Determination of Fate
of F in Mfg. Process
Determination of Fate
of Fluoride Evolved
Development of F
Control Techniques
Determination of Fate of F
in Control Process
Aluminum
^Experimentally determine F
capture by hoods and define
building control requirements
and characteristics
Iron and Steel ^Experimentally determine and
2)
Experimentally determine
gj "¦ —
'Design, develop and test on
verify average F contents of
species and quantities of F
bench and portable pilot plant
iron ore bodies and sinter/
cpds evolved and emitted by
economic, effective F emission
pellet plant charges for U.S.
primary iron/steel production
control processes for iron/steel
regional areas
processes

Coal Burning
Steam-
Electric
Power
I
ro
^Experimentally determine by U.S.
regional area F, alkali, and
alkaline earth metal contents
and variabilities of coal
'Experimentally determine
the F species emitted by
coal combustion as function
of feed composition
4)
Experimentally determine
effect of S&N purification
processes on fluorides in
coal and effect of fluorides
on processes
^Experimentally determine quantities
and types of F cpds removed by current
and projected S02 control processes
and effect on processes
Cement,
Ceramic and
Glass
Experimentally determine by U.S.
regional areas the F contents of
cement feedstocks
^Experimentally determine the F
content of feedstocks and process
streams in frit mfr.
4)
Experimentally determine by U.S.
regional areas the F content of
heavy clay product and expanded
clay aggreg. feedstocks
2~)
'Experimentally determine
F species and quantities
emitted in cement mfr. as
function of feedstock F
content
^Experimentally determine
F species and quantities
emitted in heavy and expanded
clay products mfr. as
function of feed
Non-ferrous
^Experimentally determine the
21 . 	 ......
'Experimentally determine
3)
'Experimentally determine the F species
metals
F contents of Cu Pb Zn ores and
F species and quantities
and quantities emitted from smelter by-

feedstocks by geographical area
evolved and emitted in Cu Pb
product H«S0- plants as functions of
process parameters and feeds

Zn smelting as functions of
feed and process parameters

-------
Table 4^2. Fluoride Emission Control - Recommended Research and Development
Priority and Time Sequence
I
oj
Project Priority	YEAR NO. 1 		YEAR NO. 2 		YEAR NO. 3 		YEAR NO. 4
Industry	No.	 I	II	III	IV	I II	III	IV I	II	III	IV	I	II
Iron & Steel
1
1

2
2

3
3
Coal
1
4

2
5

3
6

4
7
Aluminum
1
8
Cement
4
.9
Cerarni c
5
10
& Glass
3
11

1
12

2
13
Non-Ferrous
1
14
Metal
2
15
-o
	o
Approx. Expenditure
Per Quarter $M	48 96	94 92 153 151 148 133 123 . 132 116 130 23 15

-------
-p»
I
-p>
Table 4-3. Fluoride Emission Control - Recommended Research and Development
Tvoes and Costs by Industry
PROJECT
INDUSTRY
A1 uminum
ORE FLUORIDE CONTENT
CHARACTERIZATION
DETERMINATION OF FATE
OF F IN MFG. PROCESS
DETERMINATION OF FATE
OF FLUORIDE EVOLVED
Prof. Man Hrs. - 1720
Non-Prof. Man Hrs. - 1620
Total Cost - $70,000
Time Span - 9 Months
	in	
DEVELOPMENT OF F
CONTROL TECHNIOUES
DETERMINATION OF FATE OF
FIN CONTROL PROCESS
Iron S Prof. Man Hrs. - 1200
Steel Non-Prof. Man Hrs. - 1500
Total Cost - $60,000
Time Span - 7 Months
(1)
Prof. Man Hrs. - 1600
Non-Prof. Man Hrs. - 2000
Total Cost - $85,000
Time Span - 12 Months
(2)
Prof. Man Hrs. -
12400
Non-Prof. Man Hrs. -
12000
Total Cost $580,000
Time Snan - 24 Months
(3)	
Coal Prof. Man Hrs. - 1360
Burning Non-Prof. Han
Steam - Hrs. - 2410
Electric Total Cost - $75,000
Power Time Span - 9 Months
	(1)
Prof. Man Hrs. - 1160
Non-Prof. Man Hrs. - 1680
Total Cost - $62,000
Time Snan - 8 Months
	(2)	
Prof. Man Hrs. - 1750
Non-Prof. Man
Hrs. - 1330
Total Cost - $65,000
Time SDan - 9 Months
(3)
Prof. Man Hrs. - 1400
Non-Prof. Man Hrs. - 1080
Total Cost - $55,000
Time .SDan - 9 Months
(4)	
Cement, Prof. Man Hrs. - 760
Ceramic Non-Prof. Han
S Glass Hrs. - 900
Total Cost - $45,000
Time Span - 6 Months
	 (1)
Prof. Man Hrs. - 870
Non-Prof. Man Hrs. - 1030
Total Cost - $43,000
Time Span - 7 Months
	(2)
Prof. Man Hrs. - 880
Non-Prof. Man
Hrs. - 1040
Total Cost - $48,000
Time Span - 7 Months
	(3)
Prof. Man Hrs. - 1080
Non-Prof. Man Hrs. - 1080
Total Cost - $61,000
Time Span - 8 Months
	(5)	
Prof. Man Hrs. - 1000
Non-Prof. Man
Hrs. - 1240
Total Cost - $53,000
Time SDan - 7 Months
(4)
Non-	Prof. Man Hrs. - 1700
Ferrous Non Prof. Man
Metals Hrs. - 2500
Total Cost - $93,000
Time Span - 12 Months
(1)
Prof. Man Hrs - 1200
Non- Prof. Man
Hrs. - 600
Total Cost - $61,000
Time Snan - P Months
	(2) 	
Prof. Man Hrs. - 1200
Non-Prof. Man
Hrs. - 1200
Total Cost - S64,000
Time SDan - 9 Months
(3)

-------
4.2 DETAILED PROJECTS BY INDUSTRY
The following R&D projects are recommended.
4.2.1 Primary Aluminum Smelting Industry
Project 1. Determination of the Fluoride Capture Efficiency of
Pot-Line Hoods
Introduction
Primary aluminum production is both a present and a projected
major source for soluble fluorides emitted to the atmosphere. Virtually all
of the soluble fluorides emitted by the industry come from the reduction
process pot-lines. Total emissions of soluble fluorides during 1970 for
the industry are estimated at 16,200 tons. Over 70% of this total is esti-
mated to be accounted for by fluoride evolution which escapes capture by
the pot-line hood system.
In view of the quantity of soluble fluorides involved, hood capture
efficiency must be known accurately. The current estimates' of hood capture
efficiency are based upon industry responses to an OAP sponsored question-
naire. Direct experimental verification of the industry-provided values is
required.
Objectives
1.	Determination of the effectiveness and efficiencies
under normal and "sick-pot" operating conditions of
each of the different types of pot-line hoods.
2.	Definition of the requirements for and characteristics
of pot-line building effluent capture and abatement
devi ces.
Approach
1. Based on statistically designed experiments and
sampling plans, experimentally determine the amount
of fluorine-containing materials captured by the fume
collecting devices for each of the three basic smelter
types, under normal and (insofar as possible) "sick-
pot" operating conditions.
4-5

-------
2.	Experimentally determine the effect that different
types of fume collecting equipment and different col-
lection techniques have on the effectiveness of
collection of the fluoride effluents from the pots.
3.	Determine experimentally the efficiency of the building
fume collection hoods and the conditions under which
they operate.
4.	Estimate, from experimental data derived around
operating smelters, the actual overall efficiencies
of the fume collection devices.
5.	Define the operating characteristics for building
collection systems and abatement devices.
Tasks
1.	Design a statistically based experiment to obtain
the required information on the fume capture effi-
ciencies of the hoods used with the three pot types.
2.	Experimentally determine by flow measurement,
sampling and analyses the quantity of fluorine
compounds (fluorides) collected by the pot-line fume
collection devices for each of the three basic
smelter types for a 1 day period.
3.	Experimentally determine by flow measurement,
sampling and analyses the quantity of fluorides
emitted through the pot-line building ventilation
devices.
4.	Determine experimentally the effects of variations in
individual pot fume collection devices on fluoride
capture efficiency as per (2) and (3) above.
5.	Estimate, the actual efficiency of the pot fume
collection hoods, when in operation, either singly
or manifolded.
6.	Estimate the overall pot-line fume collection
efficiencies for each smelter type.
7.	Define the flow characteristics for pot-line build-
ing ventilation systems, to provide allowable
building fluoride concentration levels, and capture
of fluorides which escape the pot hoods.
8.	Define the characteristics of building effluents
that abatement systems will be required to handle
in terms of flow, gaseous and particulate fluoride
concentrations, other contaminant concentrations,
gas composition, temperature, and pressure.
4-6

-------
Costs
: Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
Other Direct Costs
Total Cost
! Time Span
1,720
1,620
4
$5,000
$70,000
9 Months
4.2.2. Iron and Steel Manufacture
Introduction
The manufacture of iron and steel produced an estimated 64,600 tons
of soluble fluoride emissions in 1968, and, if current practices continue,
will produce 46,400 tons of soluble fluoride emissions in 2000. Almost all
of the emissions occur in ecologically sensitive urban areas. Relatively
little has been reported in the open literature on the fluoride content of
the iron ores employed in this country. With the exception of scanty
reports on the emission of soluble fluorides associated with the use of
high fluoride ore at one plant location, almost.no data exist on the fate of
the fluorides fed into the various iron and steel production processes as
contaminants and fluxes. Finally, only one control process has been
employed, at one plant location only, for the specific purpose of control-
ling fluoride emissions.
The highest priorities have been assigned to research and develop-
ment projects designed to eliminate the lack of factual experimental data
resulting from the above deficiencies. The three projects are: (1) quanti-
tative determination of fluoride contents of iron ores by geographical area,
(2) determination of the fate of fluorine compounds in the current iron and
steel industry, and (3) development of a cost-effective fluoride control
technique utilizing current state-of-the-art techniques.
The research and development plans presented in the following discus-
sion cover the above three projects, which are designed to operate with a
considerable degree of parallelism, thereby significantly diminishing the
potential elapsed time required. Descriptions of the individual projects
together with schedule, level of effort and cost estimates are presented.
4-7

-------
Project 1. Determination of Fluoride Content of Iron Ores by Geographical
Area
The main purpose of this project is to determine the quantities of
fluorides present in the different iron ores used in the United States.
Very little data now exist on the total amount of fluorine present in the
iron ores from the various domestic and foreign sources used in current iron
and steel manufacturing operations. The determination of fluoride concen-
trations is necessary for planning operations over the next several years,
for abatement and also for process development.
Objectives
Approach
1.	Determine the fluoride content of the iron ore used
as raw material for the production of iron and steel
in each of the major industry areas in the United
States.
2.	Determine the average fluoride content of iron ore
representative of each of the major ore bodies
employed as raw material sources.
1.	Utilizing statistical experimental design techniques,
determine the quantity of iron ore samples from
each given location and the number of locations
required to provide valid estimates of the mean
fluoride content and its variance.
2.	Collect iron ore samples from the various active
mines supplying the iron and steel industry in each
of the selected areas.
3.	Collect iron ore samples from feed to the ore
sintering plants in each of the selected areas.
4.	Utilizing accepted quantitative analytical chemical
techniques, determine the fluoride concentrations
in the ore samples. This is to include an investi-
gation of the various fluoride analytical techniques,
selection of the best methods and experimental
design of the testing procedure to provide valid
statistical information.
5.	Map the United States into regional areas with
respect to mean fluoride distribution for both
sinter plant raw material feed and ore body source.
4-8

-------
6. Verify the above raw material fluoride content values
by using the blending ratios employed as the bases
for calculating plant feed fluoride content from ore
body fluoride concentrations.
1.	Design a statistically based experiment to yield
the number of samples required from each location.
2.	Design a statistically based experiment to give
the number of locations to be sampled to provide
the desired information.
3.	Collect the number and type of samples required from
each location.
4.	Perform an evaluation of the analytical methods to
be used in the project.
5.	Analyze the samples collected as per Task 3.
6.	Obtain the blending ratio employed at the plant
locations sampled.
7.	Reduce the analytical data to provide the appro-
priate statistics.
8.	Map the various statistical data to provide the
functions of geographical distribution required.
9.	Cross check the sinter-plant feed analytical values
by comparison with those derived from blending
ratios and ore-source fluoride analyses.
Tasks
Costs
(Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
iOther Direct Costs
Total Cost
Time Span
1,200
1,500
8
$4,000
$60,000
7 Months
For the purposes of this project sinter plant and pelletizing plants
are synonymous.
4-9

-------
Project 2/ Determination of the Fate of Fluoride Compounds in the Iron
and Steel Industry
The main purpose of this project is to determine experimentally
the species and quantities of fluoride compounds evolved and emitted by
the primary iron and steel production processes. The data will be used
to verify the values estimated by use of thermochemical analysis, and
will serve as one basis for selection of control processes. The primary
iron and steel production processes to be investigated include iron ore
sintering and pelletizing operations.
Ojectives
1.	Determination of the quantity, type and distribution
of fluoride compounds evolved during the primary iron
and steel production processes.
2.	Determination of the quantity and form of the fluoride
species emitted from iron and steel operations.
Approach
1.	Utilizing statistical experimental design techniques,
determine the sampling plan required to provide valid
experimental data on the fluoride species and mean
fluoride compound concentration levels in the iron
and steel fabricating process feed streams, product
streams, by-product streams and stack effluents.
2.	Examine the various sampling and analytical chemical
techniques and select the best methods of analysis.
3.	Utilizing the selected sampling and analytical chemi-
cal techniques, determine the fluoride and cofactor
values of the process streams and effluents.
4.	Incorporate the above information into a correlation
model which can be utilized to predict both fluoride
amounts and fluoride species found in the process
effluent streams as a function of the input ore
composition and process fluoride additives.
5.	Utilizing the correlation model in 4 above, tabulate
the fluoride amounts and species for the process
effluent streams.
4-10

-------
Tasks
1.	Design the statistically based sampling plan required
to obtain valid experimental data on the fluoride
species and individual fluoride compound concentra-
tions in the feed streams, product streams, and other
process effluents for each of the four basic iron and
steel producing processes (iron ore sintering and
pelletizing, and blast and open hearth furnace
operations).
2.	Evaluate the various available sampling and chemical
analytical methods, and select those to be used in the
project.
3.	Collect the numbers and types of samples required by
the sampling techniques selected.
4.	Determine the fluoride species and fluoride and
cofactor values for each of the process streams by
the chemical analytical methods selected.
5.	Utilize the above information to synthesize a correla-
tion model for prediction of fluoride species evolution
and emission as functions of ore and flux fluoride
charge quantities.
6.	Tabulate the fluoride emission values appropriate to
the various ores, additives and process conditions in
common use.
Project 3. Development of Cost Effective Fluoride Control Techniques
The main purpose of this project is to develop, through fundamental
studies, engineering evaluations and pilot plant studies, a cost effective
fluoride control technique for each of four primary iron and steel pro-
duction processes. Only one (very expensive) fluoride control facility is
now in operation and a more cost effective system is necessary, especially
Costs
Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
Other Direct Costs
j Total Cost
I Time Span
1,600
2,000
16
$8,000
$85,000
12 Months
4-11

-------
where there are low concentrations of fluoride compounds in the process
effluents. The control of soluble fluorine compound emissions represents
a major abatement process development effort in the iron and steel indus-
try. Utilizing the approach specified below, cost effective techniques
can be developed for the control of soluble fluoride emissions in the iron
and steel industry.
Objectives
Approach
1.	Determine the specific applicability of current
fluoride control techniques (both wet and dry) to
iron and steel operations.
2.	Determine the fluoride control costs as a function
of equipment, plant size and efficiency.
3.	Provide new, more economical (compared to the U.S.
Steel processes) control through current control
process modifications.
1.	Through a literature search and direct contact with
equipment manufacturers, verify prior findings on the
materials of construction, capital costs and operating
costs of processes and equipment potentially suitable
for fluoride control in the iron and steel industry.
2.	Perform detailed parametric analyses (utilizing com-
puter simulation) of the pollutant control economics
with input fluoride concentration, species and
control efficiency as the variable parameters.
3.	Perform engineering studies on theoretically modified
current control techniques for removal of fluorides
from the effluents of the iron and steel industry.
Evaluate t+ie effects these changes may have on process
economics and efficiencies.
4.	Under laboratory conditions simulate the most effec-
tive modification made to control processes and verify
the effect on fluoride emission of these changes.
5.	Evaluate the most promising processes on a pilot
plant scale at representative iron and steel
processing facilities.
6.	Provide detailed recommendations of economically opti-
mal fluoride pollutant control systems for various
plant sizes and configurations.
4-12

-------
Tasks
1.	Perform a literature search to verify materials of
construction, capital costs and operating costs of
processes and equipment potentially suitable for
fluorine control in the iron and steel industry.
2.	Initiate direct contact with equipment manufacturers
to verify prior findings on the materials of con-
struction, capital costs and operating costs of
processes and equipment potentially suitable for
fluorine control in the iron and steel industry.
3.	Prepare a mathematical model to simulate pollution
control economics.
4.	Perform parametric analyses (utilizing computer
simulation) of the pollution control economics with
input fluoride concentration, species and control
efficiency as the variable parameters.
5.	Perform engineering studies on theoretically modified
control techniques for removal of fluorides from the
iron and steel effluents.
6.	Evaluate the effects the changes to the basic pro-
cesses have on the economics and efficiencies of the
iron and steel operations.
7.	Simulate under laboratory conditions the most effec-
tive modifications to control processes.
8.	Utilizing laboratory scale mini-piants, verify the
effect modifications to control processes have on
fluoride emissions.
9.	Construct a portable pilot plant of the most promis-
ing process.
10.	Operate the portable pilot plant at representative
iron and steel facilities.
11.	Evaluate the analytical results of the pilot plant
testing.
12.	Provide detailed recommendations of economically
optimal fluoride pollutant control systems for
various plant sizes and configurations.
4-13

-------
Cost
Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
Other Direct Costs
Capital Equipment Cost
Total Cost
Time Span
12,000
12,000
40
20
50
$580,000
24 Months
4.2.3 Coal Combustion
Introduction
The combustion of coal represents a major probable source of hydro-
gen fluoride emission and, therefore, requires considerable emphasis on
the investigation and development of techniques for the abatement of pollu-
tion from this source. The specific deficiencies identified in the course
of the Fluoride Emissions Control Study will be eliminated by: (1) quanti-
tative statistical assessment of the means and variances of the fluoride
content of coals from the various actively mined beds; (2) experimental
determination of the fate of the fluorine in the combustion, heat transfer
and various effluent streams of coal-fired systems; (3) determination of
the effect of various control processes designed to remove sulfur dioxide
from flue gases on the types and quantities of fluoride produced as well as
the effects of fluoride contamination on these processes and ultimate by-
products recovery; and (4) determination of the effects of the various coal
pretreatment processes on fluoride content in the final fuel form and the
effect of the fluorides on these processes and ultimate recovery.
The research and development plans presented in the following dis-
cussion consist of four projects, each addressed to solve a specific
deficiency, designed to operate with a considerable degree of parallelism
thereby significantly diminishing the potential elapsed time required.
Descriptions of the individual projects together with time requirements,
level of effort and cost estimates are presented below.
Project 1. Determination of the Fluoride Content of Coal
The quantities and types of fluoride containing pollutants emitted
by coal-fired combustion systems depend first on the quantity of fluorine
contained in the coal and, second, on the types and quantities of other
4-14

-------
elements whose combustion products might react with or adsorb fluorine
compounds. Since the limitation of fluoride emissions might require
restrictions on the quantity of fluoride contained in the coal and since
the forms in which the fluoride is evolved could effect the design of
potential abatement processes, a determination is required of the averages
and variances associated with the content of fluorine, alkali and alkaline
earth metals, and possible adsorbents.
Objectives
Approach
1.	To determine the species and expected weight percent
ranges for the fluoride, alkali and alkaline earth
metals and possible adsorbents in coal.
2.	To determine the geographical distribution of the
fluoride and other specific components in coal.
1.	Utilizing statistical experimental design techniques,
determine the quantity of coal samples from a given
location and the number of locations required to pro-
vide valid estimates of the mean fluoride content and
its variance. Since the vast majority of coal con-
sumed for power plant usage comes from the area of the
U. S. east of the Mississippi River, the Eastern
Interior and Appalachian Regions, initial efforts
should be concentrated in this area. Typical major
producing beds of interest are: the Pittsburgh, Upper
and Lower Kitanning, Freeport and the Illinois No. 5
and 6 which correspond to neighboring areas in Indiana
and Western Kentucky. Obtain required coal cofactors
(heat values, ash content and metals content) from
1i terature/suppli ers.
2.	Collect coal samples from the various active mines in
the selected areas which supply power plants.
3.	Utilizing qualitative and quantititive chemical
analytical techniques, determine the fluoride and other
specified species concentrations. This is to include
an investigation of the various available analytical
techniques, selection of the best techniques, and
experimental design of the testing procedures to
provide valid statistical information.
4.	Map the major U. S. regional areas with respect to
mean fluoride and other specified element distributions.
4-15

-------
Tasks
1.	Determine from historical information available from
the Bureau of Mines the variability of ash content,
sulfur concentration, and other factors and use these
data to assess the number of samples required to pro-
vide accurate statistics.
2.	Perform a literature survey to identify candidate
techniques for the determination of the various ele-
ments of interest.
3.	Based on Bureau of Mines and ASTM coal sampling tech-
niques and other information derived from 1 and 2 above,
determine the types, sizes and sources of samples to be
taken from the various coal beds.
4.	Test the various chemical analytical techniques for
accuracy, sensitivity and reproducibility; select the
best procedures, and develop new methods where necessary.
5.	Collect the indicated number, type, and size samples
from the selected sources.
6.	Perform the appropriate analyses on the samples.
7.	Reduce the analytical data to provide the appropriate
statistics.
8.	Map the various statistics as a function of geographical
distribution.
Costs
Professional Man-Hours	1,360
Nonprofessional Man-Hours	2,410
Computer Units	8
Other Direct Costs	$4,000
Total Cost	$74,000
Time Span	9 Months
Project 2. Determination of the Fate of Fluorides in the Power Plant
Combustion Process
Direct experimental data on the fate of the fluoride compounds in
the coal burned in steam-electric power plants is essential to proper
design of abatement processes, and for corroboration of the high priority
currently assigned to this source because of the estimated present and
future magnitude of soluble fluoride emissions. These are, for 1970, an
4-16

-------
annual emission rate of 27,000 tons of HF; for 2000, a projected emission
rate of 86,000 tons of HF.
Objectives
1.	Determination of the quantity, form and distribution of
fluoride species in the products of coal combustion
in the power plant environment.
2.	Determination of the effect of other coal composition
factors, e.g., quantity and types of metals, ash con-
tent, etc., on the fluoride distribution.
3.	Development of a quantitative chemical model of a
typical power plant for predicting the fate of fluorides
in the combustion products as a function of input coal
types.
Approach
1.	Utilizing statistical experimental design techniques,
determine the sampling plan required to provide valid
estimates of the mean fluoride compound concentration
levels in power plant feed streams, ash residues,
purification solutions and stack effluents.
2.	Utilizing qualitative and quantitative analytical
chemical techniques, determine the fluoride and cofactor
values of the various power plant streams. This is to
include in-depth analysis of the various analyti-
cal techniques and the selection of the best techniques.
3.	Incorporate information into a correlation model which
will be utilized to predict both fluoride amounts and
species found in the power plant effluent streams as a
function of the input coal composition.
Tasks
1.	Design a statistically based experiment to determine
the sampling plan for feed streams, ash residues,
purification solutions and stack effluents to provide
valid estimates of the mean fluoride concentration
levels.
2.	Perform a literature survey to determine the best
possible analytical procedures for fluorine compounds.
3.	Evaluate the available qualitative and quantitative
chemical analytical methods.
4-17

-------
4.	Utilizing the coal composition information obtained
in Project 1, a thermodynamic equilibrium program
and a kinetic rate program, determine the probable
chemical products formed and the quantities expected
from the combustion of coal.
5.	Sample representative power plant streams as per the
sampling plan.
6.	Analyze the samples by the selected methods for
fluoride species and concentrations.
7.	Perform a statistical evaluation of the data.
8.	Design a correlation model utilizing the above infor-
mation, to predict both the fluoride levels and
species to be found in the power plant effluent
streams as a function of input coal composition.
Costs
Professional Man-Hours	1,160
Nonprofessional Man-Hours	1,680
Computer Units	10
Other Direct Costs	$4,000
Total Cost	$62,000
Time Span	8 Months
Project 3. Concurrent Removal of SO2 and Fluoride Combustion Products
from Power Plant Stack Gas
Several processes have been proposed for the removal of sulfur
dioxide (SO2) from coal burning power plant stack gas. An evaluation of
the effect of these processes on the fluoride compounds in the stack gas,
and of the fluoride compounds on the scrubbing process chemistry and equip
ment is desirable to allow design of units which will remove detrimental
fluoride compounds from the stack gas concurrently with the S02» and to
insure that the fluoride compounds will not adversely effect the SO2
scrubbing process chemistry or cause premature equipment failure.
Objectives
1.	Determination of the quantities and types of fluoride
compounds (evolved from combustion of coal and present
in power plant stack gas) removed in selected current
and projected SO2 scrubbing processes.
2.	Determination of the chemical products formed by the
reacted and/or absorbed fluoride, and the effects of
these fluorides on process equipment in terms of
corrosion, scaling, etc.
4-18

-------
Approach
3. Determination of the changes in the scrubbing system
product/by-product make resulting from the fluoride
compound uptake.
1.	Utilizing the correlation model developed in Project 2,
determine the fluoride compound product distribution
expected in coal-fired power plant flue gas.
2.	Utilizing existing literature data, estimate the
effect of expected fluoride compounds on current and
projected catalytic SC^ removal processes.
3.	Utilizing existing literature data, estimate the
effect of expected fluoride compounds on wet scrubbing
process chemistry.
4.	Utilizing existing literature data,	estimate the
effect of expected fluoride compound reaction products
on process equipment, both in terms	of scaling and
corrosion.
5.	Conduct a market study to determine	the impact of
fluoride reaction product inclusion	in SCL pollution
abatement process by-products.
6.	Propose a laboratory or pilot plant experimentation
study designed to enhance available	data to allow a
full assessment of the SCL/fluoride concurrent removal
process design problems.
Tasks
1.	Making use of the data already collected in Project 1,
organize the input data available on fluoride concen-
trations and compounds existing in several typical
boiler fuel coals.
2.	Utilizing the correlation model developed in Project 2,
determine the probable chemical products formed from
combustion of the several typical boiler fuel coals.
3.	Identify the current and projected most promising flue
gas SO2 abatement processes.
4.	Conduct a literature search, and determine the effects
of the expected fluoride compounds on current and
projected catalytic SOg abatement processes.
5.	Identify the current and projected most promising wet
scrubbing processes for SOg removal from flue gas.
4-19

-------
6.	Utilizing literature information, determine the effect
of the expected fluoride reaction products on the wet
scrubbing chemistry, including potential side reactions.
7.	Determine the scaling, erosion, and corrosion potential
of fluoride compounds removed by and formed in the wet
scrubbing circuit utilizing literature data and infor-
mation.
8.	Utilizing the information generated in Tasks 4 and 6,
above, conduct a user-oriented market survey to deter-
mine the effect of fluoride content in potentially
salable products from SC^ pollution abatement process.
9.	Identify the gaps in knowledge required to completely
assess the effect of S0« abatement processes on fluoride
emissions and of fluoriae content of stack gas or S02
abatement processes.
10. Propose an experimental program to develop required
data to satisfy gaps in the published data identified
in Task 9, above.
Project 4. Determine the Effect of Current and Projected Coal
Purification (Sulfur Removal) on Captive Fluorides
The purpose of this project is to determine the effect of coal
purification by the removal of sulfur compounds on the fluoride compound
levels of the coal. The reactions of the fluorine species in the coal to
the different solvent systems, to organic sulfur compounds removal and to
inorganic sulfur compound removal processes will be investigated. Since
sulfur removal from coal is in its infancy, the main emphasis of the
fluoride program will be to determine potential process impact.
Objecti ves
1. Determination of the portion of the captive fluorides
in the various coals removed by chemical reaction
or leaching during the treatment (both current and
projected processes) of coal for sulfur removal.
Costs
Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
Other Direct Costs
Total Cost
Time Span
1,600
1,300
6
$1,000
$65,000
9 Months
4-20

-------
2. Determination of the form of the reacted fluorides
and their impact on process equipment and by-products.
Approach
Tasks
Determine the probable reaction products and confirm
the probable product distribution expected under
process reaction conditions.
Utilizing
undertake
results.
statistical experimental design techniques,
a laboratory study to confirm the theoretical
3.	Knowing the form of the fluoride reaction products,
estimate their effects (corrosiveness and material
compatibility) on current and projected process equip-
ment design.
4.	Through a market study, determine what impact the
fluoride reaction products will have on the market-
ability of by-products (mainly sulfur compounds).
1.	Utilizing thermodynamic equilibrium and kinetic rate
data, determine the fluoride compound product distri-
bution expected under different process reaction con-
dition, i.e., residence time, temperature and pressure.
2.	Using statistical experiment techniques, design a
laboratory study to validate the theoretical results.
3.	Conduct a literature search and determine the effects
of the expected fluoride compounds on current and
projected sulfur removal processes.
4.	Identify the current and projected most promising sulfur
removal processes.
5.	Utilizing the literature information, determine the
effect of the expected fluoride reaction products on the
sulfur removal chemistry, including side reactions.
6.	Conduct a laboratory scale verification to confirm
theoretical compounds.
7.	Estimate the effects of the fluorine compounds on pro-
cess design for the sulfur removal process.
8.	Perform a literature search on the effect the fluoride
compounds will have on materials of construction and on
potential corrosion of process equipment for the current
and projected desulfurization processes.
4-21

-------
9. Conduct a user-oriented market survey to determine the
effect of fluoride content on potentially soluble
products from desulfurization of coal.
Costs
Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
Other Direct Costs
Total Cost
Time Span
1,340
1,000
8
$500
$55,000
9 Months
4.2.4 Cement, Ceramic and Glass Manufacture
The cement, ceramic and glass industries (glass manufacture, frit
smelting, heavy clay product, expanded clay aggregate, and cement) involve
high temperature production of vitreous or refractory shapes from siliceous
raw materials. In all of the industries, fluorides are a part of the raw
material charge - either accidentally, as contaminants, or deliberately, as
additions.
There is almost no data available on the fluoride content of the raw
material charges, or on the fluoride species and quantities evolved and
emitted. It is estimated, however, on the basis of the sparse information
available, that these "silicate" industries currently emit 19,300 tons of
soluble fluorides per year, and will emit 42,200 tons per year by 2000.
Knowledge of the fluoride content of the feed materials, and the fluorides
evolved, is mandatory for adequate control design. As a first step toward
this goal, research and development programs are proposed in the following
sections to assess the amounts of fluorine compounds in various feedstocks
and to determine their fate in the actual production processes.
Project 1. Determination of Fluoride Concentrations in Production
Feedstocks by Geographical Areas
As a prerequisite to accurate definition of the problem of fluoride
pollution, and to adequate design of an appropriate control program, the
exact chemical nature and quantity of the pollutants must be determined.
For cement production, the fluoride emission originates from the feedstock
components which contain fluorine compounds. Hence, it is pertinent to
know the quantity of fluorides present in the various feedstocks to the
cement production process.
4-22

-------
Objectives
Approach
Tasks
1.	Determination of the fluorine content of the various
feedstocks to the cement production process.
2.	Establishment of the geographical distribution of the
various feedstocks with respect to their total fluoride
contents.
1.	Based on statistical principles for the design of
experiments, develop a sampling plan which will produce
valid and representative estimates of the mean fluorine
content and its variance at each chosen geographical
area.
2.	Collect the samples according to the sampling plan from
the various sources supplying cement plants in the
selected areas.
3.	Utilizing the best qualitative and quantitative chemical
analysis techniques, determine the concentrations of
. fluorine in the samples.
4.	Relate the mean fluorine content of each cement feedstock
component to the U.S. regional area where it originates.
This will establish the geographical distribution of
fluorine-containing feedstock components in the U.S.
1.	Utilizing statistical techniques for experimental design,
set up a sampling plan which determines the size and
number of the samples from a given location, and the
number of locations required from each geographical area.
2.	Collect the required samples according to the plan
determined in Task 1.
3.	Assess the various available methods for quantitative
chemical analysis for fluorine and select the best
method for the feedstocks. Modify the existing methods
or develop a new method if necessary.
4.	Perform the chemical analyses required for the samples
taken.
5.	Reduce the analytical data obtained in Task 4 to yield
the desired statistics.
6.	Map the various statistics as a function of geographical
distribution.
4-23

-------
Cost
Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
Other Diurect Costs
Total Cost
Time Span
760
900
4
$10,000
$46,000
6 Months
Project 2. Determination of Fluorides Emitted by the Cement Production
Processes.
Before any work can be planned and undertaken to develop an economi-
cally feasible process to control or reduce the fluoride emission from the
cement production process, it is necessary to know the chemical identities
and quantities of the fluorine compounds evolved and emitted. This
research and development program is proposed to obtain the needed
information.
Objectives
1.	Identify and determine the quantities of all fluorine
compounds emitted by the cement production processes.
2.	Develop a correlation model to relate the emitted fluoride
species and their amounts to the feedstock fluoride
content.
1.	Based on statistical methods for experimental design,
develop a sampling plan to provide valid estimates of
the mean fluorine compound concentration levels in the
feed streams and the stack effluent streams for cement
production plants.
2.	Evaluate the available sampling, qualitative and quanti-
tative analytical methods to determine the best methods
to obtain and analyze the samples taken.
3.	Sample and perform the analyses.
4.	Use the analytical data to develop a mathematical cor-
relation which can be utilized to predict both the
fluoride species and their amounts in the effluent
streams as a function of the feedstock fluoride
content.
5.	Apply the correlation to obtain the fluorine species
and contents of stack effluents from cement plants.
Approach
4-24

-------
Tasks
1.	Utilizing statistical methods for the.design of experi-
ments, establish a sampling plan which will determine
the size and number of samples to be taken at each given
location and the number of locations required from each
geographical area.
2.	Evaluate the various available methods for sampling
and qualitative and quantitative chemical analysis
for fluorine. Select the best methods for the various
streams. Make any modifications necessary or develop
a completely new method if necessary.
3.	Collect the required samples according to the above
sampling plan and methods.
4.	Perform the chemical analyses for all samples taken.
5.	Use the test results to develop a mathematical correla-
tion relating the types of fluorine compounds and
quantities discharged in the effluent streams to the
feedstock component composition.
6.	Utilize the correlation developed in Task 5 to compute
the fluorine contents of the stack effluents for the
various fluoride ranges found in Project 1.
Project 3. Determination of Fluorides in the Effluents of Opal Glass
and Enamel Frit Production Processes
Fluorine compounds such as fluorspar, cryolite, etc. are among the
many components used in opal glass and enamel frit productions. As a
result, some 3,300 tons of hazardous fluorides are emitted from these
sources every year. In order to develop the technology needed to reduce
the fluoride emissions from these industries, it is necessary to determine
the species and quantities of the fluorine compounds present in the effluent
streams under normal operation conditions, and how are they effected by
minor variations in the feed compositions. It is the purpose of this pro-
posed project to obtain the needed information.
Costs
Professional Man-Hours
Nonprofessional -Man-Hours
Computer Units
Other Direct Costs
Total Cost
Time Span
870
1030
8
$5,000
$48,000
7 Months
4-25

-------
Objectives
Approach
1.	Identify and determine the quantities of all fluorine
compounds found in the effluent and production streams
of the opal glass and enamel frit production processes.
2.	Determine the effect of variations in the fluoride content
of the feed components on the fluoride content and distri-
bution in the effluent and production streams.
3.	Develop a mathematical model to relate the fluorides input
in the feed to the fluorides in the effluent streams.
1.	Based on statistical principles, establish a sampling
plan for taking samples from the feed streams, the product
streams, and the effluent streams from opal glass and
enamel frit production processes.
2.	Utilizing the best techniques for qualitative and quanti-
tative chemical analysis, determine the fluorine content
of various process streams.
3.	Correlate the experimental data to form mathematical
models which will be utilized to perdict the fluoride
species and quantities in the effluent streams as a
function of the input fluoride.
4.	Apply the mathematical models to compute the stack effluent
fluoride concentration level for each fluoride species.
Tasks
1.	Utilizing statistical methods for designing experiments,
set up a sampling plan which specifies the optimum size
and number of samples to be taken from a given location
and the number of locations for each geographical area.
2.	Collect the samples according to the sampling plan of
Task 1.
3.	Investigate various techniques for qualitative and quanti-
tative chemical analysis for fluorine compounds. Select
the best analytical method. Develop a completely new
method if necessary.
4.	Perform the chemical analyses on all samples taken.
(Note: Feed stream samples are to be analyzed quanti-
tatively only for fluoride content. No qualitative
analyses need be performed on feed streams.)
4-26

-------
5.	Analyze and correlate the test data to yield correla-
tion models which may be used to predict the fluoride
species and amounts in the effluent streams from the
fluoride contents of the feed.
6.	Utilize the correlation models developed in Task 5 to
compute the stack effluent concentrations for each
fluoride species for the various fluoride ranges found
for the feed streams.
Project 4. Determination of Fluoride Concentrations of Clay Product
Feedstocks from Various Geographical Areas
Heavy clay products and expanded clay aggregates contribute an esti-
mated 15,000 tons per year of soluble fluorides to current atmospheric pol-
lution. To verify this estimate, and to enable proper control planning,
knowledge of the fluoride contents of the various production feedstocks is
required.
Objectives
1.	To determine the fluorine content of the various feed-
stocks to the heavy clay and expanded clay aggregate
plants.
2.	To establish the geographical distribution of the various
feedstocks with respect to their total fluoride contents.
1.	Compile all available data on chemical compositions of
feedstocks for clay products from various geographical
locations.
2.	Based on statistical principles, set up the sampling
and experimental plan to characterize the feedstocks.
•3. Collect samples from various geographical areas.
4. Utilizing best available chemical analysis techniques,
determine the fluoride contents of all samples taken.
Cost
Professional Man-Hours
Nonprofessional Main-Hours
Computer Units
Other Direct Cost
Total Cost
.Time Span
880
1040
8
$4,000
$48,000
7 Months
Approach
4-27

-------
5.
Map the variations in the fluoride content of the feed-
stocks as a function of geographical areas.
Tasks
1.	From a literature search and other available sources,
compile data on chemical composition of feedstocks used
for producing various clay products. It is necessary to
have data covering a wide range of geographical areas.
2.	Utilizing a statistical method for experimental design,
establish a sampling plan which specifies the optimum
size and number of samples to be taken from a given
location and the number of locations required.
3.	Collect the samples according to the plan set up in
Task 2.
4.	Investigate various techniques for chemical analysis of
fluorine. Select the best analytical method. Develop
a completely new method if necessary.
5.	Perform the chemical analyses of all samples taken.
6.	Correlate the results of chemical analyses and the litera-
ture data with the locations of samples to establish the
fluorine content in the feedstocks as a function of the
geographical areas.
Cost
iProfessional Man-Hours	1,Q00
Nonprofessional Man-Hours	1,240
Computer Units	6
Other Direct Costs	$5,000
;Total Cost	$53,000
Time Span	7 Months
Project 5. Determination of Fluorides Emitted by the Heavy Clay and
Expanded Clay Aggregate Processes
Accurate knowledge of the fluoride species and quantities evolved
and emitted by the heavy clay and expanded clay aggregate production proc-
esses as functions of the feedstock fluoride contents is required for
optimum design of control processes. This project is designed to obtain
that knowledge experimentally over a wide range of feedstocks and plants.
Objectives
1. To identify and determine the quantities of all fluorides
emitted by the heavy clay and expanded clay aggregate
processes.
4-28

-------
Approach
2. To develop correlation models to relate the emitted
fluoride species arid their amounts to the feedstock
fluoride content.
1 . Based on statistical methods for experimental design,
develop a sampling plan to provide valid estimates of
the mean fluorine compound concentration levels in the
feed streams and the stack effluent streams for heavy
clay and expanded clay aggregate production plants.
2.	Evaluate the available sampling and qualitative and
quantitative analytical methods to determine the best
methods to be used to obtain and analyze the samples
taken.
3.	Sample and perform the analyses.
4.	Use the analytical data to develop mathematical cor-
relations which can be utilized to predict both the
fluoride species and their amounts in the effluent
streams as a function of the feedstock fluoride content.
5.	Apply the correlations to obtain the fluorine species
and contents of stack effluents from heavy clay product
and expanded clay aggregate plants.
Tasks
1.	Utilizing statistical methods for the design of experi-
ments, establish a sampling plan which will determine
the size and number of samples to be taken at each
given location and the number of locations required
from each geographical area.
2.	Evaluate the various available methods for sampling,
and qualitative and quantitative chemical analysis
for fluorine. Select the best methods for the various
streams. Make any modifications necessary or develop
a completely new method if necessary.
3.	Collect the required samples according to the sampling
plan and methods selected.
4.	Perform the chemical analyses for all samples taken.
5.	Use the test results to develop a mathematical correla-
tion relating the types of fluorine compounds and the.
quantities emitted in the effluent streams to the
feedstock fluoride content, for each process.
4-29

-------
6. Utilize the correlations developed in Task 5 to compute
the fluorine contents of the stack effluents for the
various fluoride ranges found in Project 4.
Costs
Professional Man-Hours	1080
Nonprofessional Man-Hours	1080
Computer Units	18
Other Direct Costs	$4,500
Total Cost	$61,000
Time Span	8 Months
4.2.5 Nonferrous Metals Smelting and Refining Industry
Copper, lead and zinc sulfide ores are smelted and refined to yield
the three nonferrous metals, and an unknown amount of soluble fluoride
emission. The nonferrous metals industry is a potentially major source of
soluble fluoride compounds. Little is known or available on the fluoride
contents of the various sulfide ores. Considerable sampling and analytical
effort will have to be expended to provide description of the potential
problem with sufficient adequacy and accuracy for the definition of control
requirements and processes. The various methods and processes used to con-
centrate and smelt the ores, and to refine the metals will have to be
investigated, in conjunction with investigation of by-product recovery
processes, to determine experimentally the effects of the process steps
and parameters on fluoride emission.
Project 1. Determination of Fluoride Content of Copper, Lead and Zinc
Ores by Geographical Location
At the present time very little is known about the fluoride content
of the various copper, lead and zinc ores. This investigation will charac-
terize by geographical location and ore type the amounts of fluorine
present. This information, used as data base for a thermochemical equilib-
rium program, will yield a reasonable approximation of the types and
quantities of fluorides that can potentially be liberated in the ore
smelting and refining processes.
4-30

-------
Objectives
Approach
Tasks
1.	Determination of the mean fluoride concentrations and
variability of fluoride content in feedstock ores.
2.	Determination of the effect of geographical location on
the distribution of fluoride content within common
feedstocks.
1.	Locate and identify current mining areas in the U.S.
2.	Develop statistically valid experimental sampling plans
and quantitative analytical approaches for determining
mean concentrations and variability of fluoride content
¦ of concentrates as a function of geographical locations
and concentration processes.
3.	Collect ore and concentrate samples.
4.	Utilizing the selected quantitative chemical analysis
techniques, determine the required fluoride information.
5.	Map U.S. regional areas with respect to fluoride
distributions.
6.	Summarize the effects of identified concentrating
processes on ore fluoride content.
7.	Determine the effect of process parameters in concen-
trating methods on fluoride content.
1.	Perform a literature search to identify and locate
current copper, lead and zinc ore mining areas in the
U.S.
2.	Perform a literature search to identify and locate the
current copper, lead and zinc smelting and refining
areas in the U.S.
3.	Design a statistically based experiment for the deter-
mination of the average fluoride content of the various
ores and concentrates, and for the determination of the
variabilities of the fluoride content.
4.	Experimentally develop quantitative analytical approaches
and procedures for the determination of the fluoride
contents of the ores and concentrates.
4-31

-------
5.
Collect samples from all geographical locations where
mining and smelting operations are being conducted.
6.	Experimentally determine the required analytical data on
the samples collected in (5) above.
7.	Prepare a map of the United States showing by areas, the
different fluoride distributions based upon "as mined"
and "concentrate" fluoride levels and variabilities.
8.	Analyze the different processes for concentrating ores
and summarize the effects of each step on the fluoride
content of the processed material.
9.	Identify and point out the impact of process parameters,
including equipment, on the change in concentration of
fluoride containing materials in the process.
Project 2. Determination of the Fates of the Fluorides in the Non-
ferrous Metal Smelters and Refineries
In each of the sulfide ore smelting and refining processes, the
species and quantities of fluorides evolved and emitted will be deter-
mined, as functions of feed ore fluoride concentrations and production
process parameters. This information is not available in either private
or open literature. The project will provide this information, to an
extent sufficient for emission control process planning and design.
Objectives
1.	Determine the chemical species and quantities of the
fluorides evolved and emitted by the nonferrous metal
smelting and refining processes.
2.	Determine valid functional relationships between
fluoride content of feed ores, process parameters
and evolved and emitted fluorides
Costs
Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
Other Direct Costs
Total Costs
Time Span
1,700
2,500
4
$10,000
$86,000
12 Months
4-32

-------
Approach
Develop correlation models of the smelting processes,
to relate the evolved and emitted fluoride species
and quantities to fluoride ore feed content and process
parameters.
Utilizing statistical experimental design techniques,
determine the sampling plan required to characterize
the process stream fluoride levels and process param-
eters for each of the smelter types and each of the
metals.
Select appropriate sampling and chemical analytical
techniques for the determination of fluoride species
and concentrations in the various process streams.
Use the selected sampling plan and sampling and ana-
lytical techniques to obtain and analyze the various
process stream samples.
Incorporate ore analytical process and mass balance
information into correlation models for predicting
fluoride effluents as functions of ore fluoride con-
tent and process parameters.
Tasks
1.	Design the statistically based sampling plans required
to characterize the process parameters and process
stream fluoride species and fluoride levels for each
smelter type and each metal.
2.	Select appropriate sampling and fluoride identification
and determination procedures.
3.	Collect the samples and analyze them for fluoride
species and content, as per the plans and procedures
selected.
4. Develop correlation models to predict fluoride species
and quantities evolved and emitted as functions of ore
fluoride content and process parameters.
Costs
Professional Man-Hours	1,200
Nonprofessional Man-Hours	600
Computer Units	16
Other Direct Cost	$10,000
Total Cost	$61,000
Time Span	8 Months
4-33

-------
Project 3. Determination of Fluoride Compound Fate During
By-Product Sulfuric Acid Production Processes
Many nonferrous metal sulfide ore smelters have an associated
sulfuric acid plant, to utilize the by-product sulfur dioxide formed in the
smelting process. The main purpose of this program is to determine the fate
of the fluorine compounds evolved and emitted in the smelting processes.
Objective
Approach
Determination of the extent of fluoride compound
removal during the by-product sulfuric acid production
process.
1.	Identify current sulfur dioxide emission control
processes associated with the smelting of copper
lead and zinc sulfide ores.
2.	Develop and execute statistically valid experimental
plans to sample the fluoride collection efficiencies
and product and effluent fluoride contents character-
istic of the processes.
3.	Determine the effect of different ores, different
processes and process parameter variations on fluoride
collection characteristics and fluoride emissions.
Tasks
1.	Perform a literature search to identify the current
nonferrous metal smelter sulfur dioxide control
processes.
2.	Design a statistical sampling plan to determine the
fate of smelter emitted fluoride in the processes.
3.	Perform sampling operations as per Task (2) each of the
selected processes.
4.	Select appropriate qualitative and quantitative analyti-
cal techniques.
5.	Analyze collected samples as per the techniques selected.
4-34

-------
6. Develop correlation models to predict the fluoride
species emitted from the by-product sulfuric acid
plants, as functions of metal species, ore fluoride
content, smelter process parameters, and sulfuric
acid plant process parameters.
Costs
Professional Man-Hours
Nonprofessional Man-Hours
Computer Units
iOther Direct Cost
Total Cost
Time Span
$8,000
$64,000
9 Months
1,200
1,200
10
4-35

-------
5- Environmental Effects

-------
5. ENVIRONMENTAL EFFECTS
The effects of fluorides are discussed in terms of the lowest concen-
tration, less than one part per billion (ppb), that causes detectable
changes in vegetation and then the effects of increasingly higher concentra-
tions of gaseous fluorides are outlined. Included are: tip-burn observed
on the leaves of sensitive vegetation, other agricultural effects, etching
of glass, effects on man and some effects of process streams on building
structures.
5.1 VEGETATION EFFECTS
Claims of damage from fluorides are usually related to biologic
effects, and many studies have been performed. Five recent literature
-reviews have been published on fluoride effects. (4355>4158>4356,4159,4357)
The largest single listing of published reports related to the biological
effects of fluorides is an annotated fluoride bibliography.^^
(4355)
McCunex ' lists four types of fluoride effects on vegetation
considered important in developing air quality criteria for fluorides:
(a) visible effects such as necrosis (injured portions of leaves die and
become discolored, also called tip-burn), (b) a diminution in the growth
or in theiyield of fruit or seeds caused by fluoride injury, (c) changes
in physiological activities, metabolic activities and cellular structure
with or without visible injury, and (d) deposit and accumulation of
fluoride in the plant with increasingly higher fluoride concentrations in
its tissues.
The lowest concentration of hydrogen fluoride (HF) reported to
cause damage was 0.5 ppb when gladiolus plants were exposed continuously for
12.days. The leaf damage was well below the 10% of the leaf area considered
necessary to cause damage in terms of either growth or yield. Thus, no
economic loss would be attributed to this minimum detectable fluoride
effect.
5-1

-------
Examples Of exposures that do relate to leaf damage and potential
crop reduction are shown in Table 5-1. The spread of concentrations and
exposure durations is quite broad: from 0.7 ppb for a period of 15 days
for sorghum to 700 ppb over a period of 10 days for alfalfa. Other data
state the lower limits for visible damage to citrus trees, other fruit
trees, and certain evergreen trees (conifers); but for these plants, the
data are not yet adequate to predict reduction in crop values.^55^ Addi-
tional data presented for alfalfa show the relationship between levels of
HF exposure and the fluoride accumulated in the plant leaves. These data
are important considering the potential use of this crop as animal feed
and will be discussed further in the next section.
The accumulation of fluoride in plant tissues can increase
gradually and cause no injury to the plant even though the level of
accumulated fluoride may have exceeded the level that is safe for ingestion
by animals. Attempts to relate fluoride levels in plant tissues to injury,
if any, to the plant tissues have not been useful. Fluoride levels in
plant tissues alone can be misleading and have not been shown to be a
dependable basis for judging injury to plants. Hence, they will not be
considered further
here.
Table 5-1
Examples of HF Concentrations (ppb) and
Exposure Durations Reported to Cause
Leaf Damage and Potential Reduction in
Crop Values(4355)
Plant
Concentration and Time
Sorghum
0.7 ppb for 15 days (most sensitive
varieties)--15 ppb for 3 days (most
resistant varieties)
Corn
2 ppb for 10 days--800 ppb for 4 hours
Tomato
10 ppb for 100 days—700 ppb for 6 days
Alfalfa
100 ppb for 120 days~700 ppb for 10 days
5-2

-------
A large collection of experimental data is available relating
effects of gaseous fluorides to vegetation damage. Comparatively little
information is available for relating particulate fluoride levels to
vegetation damage. The most that can be said is that fluoride dusts are
far less injurious to vegetation than are the gaseous fluorides.
Another aspect of fluoride effects on vegetation concerns exposures
to dissolved fluorides. A fluoride solution such as sodium fluoride
administered to plants receiving all nutrients by solution feeding through
their roots has produced typical fluoride injury. Tip-burn from fluoride
solutions resembles the damage seen in leaves exposed to HF in air. This
suggests the possibility of damage to vegetation through fluoride con-
tamination of the soil. This has not been observed in the field,
probably because of the presence of sufficient calcium and aluminum in
soils to inactivate the fluoride. Some fluoride can be absorbed from the
soil but at relatively slow rates. Hence, reports of fluoride damage and
related crop losses have been attributed to fluorides entering the plants
from the air through their leaves.
The most susceptible plants show evidence of damage when exposed
to HF concentrations in the range of 0.5 to 1.2 ppb provided there is
continuous exposure for periods of several days. This group includes
gladiolus, sorghum, conifers, corn, citrus trees and certain other fruit
trees. Much higher levels of HF, 300 to 1000 ppb for periods of exposure
as short as 7 hours, will damage a wide variety of plants. Data on very
high levels of exposure are mostly limited to tests in experimental
fumigation chambers.
5.2 EFFECTS ON FARM ANIMALS
Fluorides may accumulate on and in plant tissues and raise the
fluoride level high enough to make that vegetation hazardous if eaten by
farm animals. This is particularly important in connection with fluoride
accumulations in pasture grass, hay crops, and silage since these foods
are so widely fed to livestock. Fluoride levels in these materials at
levels of 40-50 parts per million (ppm) on a dry weight basis may cause
injury in some farm animals if consumed continuously over long periods of
time (years) even though the accumulation of fluoride in the vegetation has
induced no detectable plant damages. When there is reason to refer
5-3

-------
to fluoride absorbed by plants or deposited on the surfaces of the leaves,
the fluoride units are usually 1000 times larger, parts per million, then
the units used above for describing the exposures of plants to gaseous
fluorides, parts per billion A unique group of plants in the tea family
may accumulate quite high fluoride levels, 1000 to 2000 ppm, and show no
evidence of fluoride damage. Since these plants are not used as food for
farm animals, no agricultural problems have been reported; however, these
fluoride levels have been given some attention in terms of the use of tea
as a beverage. It should be noted that fluoride accumulates in leaves but
appears in only negligible amounts in seeds or fruit produced by plants
having high fluoride levels in the leaves.
The fluorides in industrial emissions inhaled by farm animals have
not been observed to have an injurious effect; however, an important
aspect of fluoride accumulation by vegetation arises from accumulation of
fluoride dusts on leaves of plants and blades of pasture grass. The dusts
may be noninjurious to the vegetation but contain hazardous amounts of
fluoride in terms of forage for farm animals. Ingestion of fluoride has
been investigated extensively in many species of animals including man.
Phosphate rock is an example of a dust that seemingly has not injured
plants but is injurious to farm animals. This was made evident in the
1930's when an attempt was made to use the calcium and phosphate in this
mineral as a dietary supplement. Fluoride injury quickly became apparent.
Other data on phosphate rock in relation to animal feeding are included
below.
In overwhelming dosages fluorides cause toxic effects in several
vital systems of animals. Vegetation containing well above 5000 ppm of
fluoride would be needed to cause such effects in farms animals; and
since these high levels are not observed in pastures and fields, acute
effects will not be considered further.
Experimental feeding of cattle with feed containing fluorides at
levels of 600-1200 ppm was started but interrupted after 18 days because
of the rapidly diminished food intake. The fluoride was then administered
separately from the food to maintain more uniform dosage levels. The
experiments were terminated after about four months for the highest
fluoride level and about six months for the lowest fluoride level. Loss
5-4

-------
of appetite caused food consumption to decrease more than 50%, and the
outstanding effect was that of starvation. (^361) since concentrations of
fluoride in pasture grass as high as 2000 ppm were reported in 1946^^
but are no longer being observed near industrial emission sources, these
very high levels need not be considered further.
Levels of fluoride ingestion resulting from contamination of forage
with industrial fluoride emissions are usually well below those noted above
and acute fluoride effects are not likely to be found today. Considerable
experimental data have been published describing the effects of ingesting
foods containing 20-200 ppm of fluorides. Ingestion at these levels by
farm animals causes fluoride effects primarily related to the site where
fluorides accumulate to the greatest extent, the bones and teeth of the
exposed animals. These effects, listed in order of appearance in exposed
animals, are:
(a)	Dental lesions (primarily damage to incisor teeth)
(b)	Hyperostosis (overgrowth of the bones)
(c)	Lameness
(d)	Loss of appetite
(e)	Decreased milk production
(f)	Diminution in reproduction^4158,4356^
Items (e) and (f) result from loss of appetite caused by fluoride intake.
This secondary effect makes (e) and (f) somewhat contraversial.
The dietary fluoride intake need only be at the level of 20 to .
30 ppm in the food in order for detectable dental changes to develop among
cattle. However, these earliest changes amount only to white spots in the
enamel of the teeth and are not harmful. Continuous ingestion of food
containing 40 to 50 ppm causes dental changes in cattle that are severe
enough to cause the eventual destruction of the incisor teeth through
excessive attrition. This leads to loss of ability to graze normally,
a reduced intake of food and a series of problems that result in economic
losses amoung herds exposed continuously to food containing levels of
fluoride in the range of 40 to 50 ppm.
5-5

-------
Generally, the ill effects of fluoride pollutants among exposed
farm animals develop slowly. If forage levels do not exceed 50 ppm
fluorides, exposures continuing 5 years or more may be required for the
maximum economic losses to develop. As the fluoride levels are gradually
increased, the time required for the ill effects to appear is progressively
reduced.
Another factor governing the severity of the effects of fluoride
ingestion is the age of the animals when fluoride exposure is started. If
the animals are mature, all teeth will have erupted and no injury to the
incisor teeth may be expected. Even so, excessive bone growth leading to
lameness can still develop. Except for fluoride effects on bones and
teeth, other potentially harmful effects caused by higher levels of
fluoride, 100 ppm and above, are secondary to the effects related to loss
of appetite and the resulting decreased food intake.
Not all types of fluorides deposited on forage are equally hazardous.
Also, not all animal species are as susceptible as cattle to ingested
fluorides. Both of these factors are shown in Table 5-2 in terms of
related dosages and related toxicities.(4359)
Table 5-2. Safe Average Levels of Fluorides in the
Total Ration of Farm Animals^359)

F1uoride
Source
Species
Soluble Fluoride Such
As Sodium Fluoride
(F, ppm)
Phosphate Rock or
Phosphatic Limestone
(F, ppm)
Dairy Cattle
30-50
60-100
Beef Cattle
40-50
65-100
Sheep
70-100
100-200
Chickens
150-300
300-400
Turkeys
300-400
no data
5-6

-------
Very little information is available relating the concentrations of
fluorides in industrial process emissions to the fluoride concentrations
in pastures or alfalfa fields. Information about fluoride concentrations
adjacent to sources of the fluoride emissions is frequently obtained
through monitoring programs which involve periodic ambient air sampling
and fluoride analysis of vegetation obtained from the pastures, fields, or
orchards. However, corresponding information on source emission rates of
fluorides and their relationships with the ambient concentrations have not
been reported. Some experimental laboratory data exist that suggest that
fluoride will accumulate above the safe levels for cattle if alfalfa is
continuously exposed to a level of HF at 1 ppb for as long as 100 days.^4355^
No similar data have been published relating ambient dust concentrations to
levels of fluoride accumulated in or on forage. It is apparent that, in
general, source emissions cannot currently be definitively related to
environmental effects. The series of events including emission dispersion,
fallout, chemical reaction, assimilation, and effect contains too many un-
certainties to be accurately described.
Determination of the magnitude of fluoride ingestion by livestock
can be accomplished by measurement of fluoride excreted in urine and for
fluoride accumulated in sections of rib or tail bones (removed by biopsies).
These quantities will increase in proportion to the levels of exposure
(4359)
of the animals. ' Neither type of measurement has been related
directly to either ambient air fluoride levels or to rates of emissions of
fluorides from industrial processes.
5.3 FLUORIDE EFFECTS IN MAN
Information describing the effects of fluorides in man has come from
observations concerning both inhalation and ingestion of fluorides. Inhala-
tion data were collected in connection with potential exposure of workers
to atmospheric fluorides. Data related to ingestion of fluorides were
obtained for several reasons, but the greatest amount of information came
from investigations concerned with fluorides in drinking water supplies.
Together, these data provide usable guidelines for judging fluoride effects
5-7

-------
such as: (a) levels of urinary fluoride excretion as a function of the
level of exposure, (b) fluoride retention in the bones of exposed persons,
(c) radiologic evidence of excessive fluoride deposition in bones. Changes
in tooth enamel are also useful for judging exposure to fluorides; however,
since this effect does not occur after the teeth erupt, its usefulness is
limited to exposure occurring during childhood. Fluoride treatments given
to reduce bone loss (osteoporosis) in some elderly patients have also
provided useful information related to effects of fluoride in man. Data
related to very large doses of fluoride have been obtained from cases of
accidental poisoning or attempted suicide. These aspects of fluoride
(4159)
effects are included in the review by Hodge and Smith/
The teeth of children have shown fluoride effects related to the
fluoride in their drinking water. These dental effects, which appear clearly
when the water contains fluoride at the level of 2 ppm, are more pronounced
at 3 ppm and are described as severe when 4 ppm is reached or exceeded.
For persons whose teeth have erupted, no dental effects occur and no other
skeletal or organic ill effects from these levels have been found in the
United States. The effects of fluorides among children have been investi-
gated in areas where the potential for industrial emissions of fluoride
- (4159)
were believed to exist. Hodge and Smith v ' cite two reports from Russia
alleging dental effects among children. In contrast to this, a survey made
among children residing near aluminum plants in Vancouver, Washington, and
Ft. William, Scotland, revealed no effects attributable to exposure to
(4363,4364)
fluorides.v
Because of a nearby aluminum plant, total atmospheric fluoride levels
in and around Ft. William, Scotland, were measured and reported to range
between 59 and 130 ppb. The gaseous fraction was reported to be 7 to 120
(4364)
ppb.v	The investigation was made primarily to judge the fluoride
effects among factory employees and farm animals; but other groups of resi-
dents were also surveyed. A group of adults consisting of 26 men and 51
women not employed at the factory were given clinical examination, and all
but two of that group were also given X-ray examinations. A group of 113
children were also given the same examinations. No significant fluoride
effects were observed among these residents. Dental changes among a few
of the children were not considered to be solely the result of fluorides
5-8

-------
in the air since the past history of these children could not be established
clearly. The extent to which high fluorides in their drinking water was
related to these dental changes could not be determined. Some of the
children had lived elsewhere before moving to Ft. William and could have
been exposed to fluorides from some other source of drinking water of un-
known fluoride content. This emphasizes the potential ambiguity of data
that may depend on changes in teeth without documentation of the source
and amount of fluoride responsible for the dental changes.
Additional data showing fluoride effects in man were collected in
industrial work areas and in laboratory investigations where urinary
fluoride levels were studied. From these investigations, guidelines for
allowable levels of urinary fluorides have been determined. Levels of
urinary fluoride up to about 4 ppm are acceptable regardless of the source
of the exposure. So far there has been only one report relating fluorides
in the ambient air in a community to indicate the occurrence of related
urinary fluoride levels as high as 4 ppm.
The kind of fluoride-induced bone changes that are revealed by X-ray
examinations (increased radioopacity) have been used as a guide for fluoride
effects in surveying the residents of several communities. Whenever this
type of fluoride-induced change was found in such surveys, the related
fluoride exposures were found to have come from (a) excessive fluoride
levels in drinking water, or (b) occupational fluoride exposures. Exposure
to an outcropping of phosphate rock in North Africa represents an exception
to the above. Dust from this natural source caused extensive contamination
of drinking water and foods in the homes of nearby residents and resulted
in some bone changes.
Background information on fluorides in ambient air indicates the
normal level to be less than 0.24 ppb or 0.2 microgram per cubic meter
o	(4370)
(l^g/m ); the fluoride measured was probably all particulate.	In some
cities values of 2.3 ppb or 1.9 i-ig/m have been observed and reported as
particulate fluoride.(4371) Gaseous fluoride must have been a small fraction
of the fluoride in the air since no damage to vegetation was found. In any
5-9

-------
case, the potential for these levels to cause fluoride effects in man is
negligible. In a recently published, totally independent study by the
National Academy of Sciences, the same conclusion was reached.*
The highest ambient fluoride concentration reported was from testing
of rocket engines using fluorine as an oxidizer. In connection with one
such testing program at the NASA-Lewis Research Center, a fluoride scrubber
was designed to capture the reaction products, primarily HF, that result
from static engine tests. From preliminary tests for the scrubber, it was
predicted that "where full advantage is taken of the prevailing atmospheric
conditions and stack dispersion of the gases, concentrations of hydrogen
fluoride as high as 10,000 ppm may be safely tolerated.11 (^367) 5^^ tests
are usually very brief but, even so, 10,000 ppm is not really safe unless
the dilution in the atmosphere is very rapid. No data have been found
revealing the total quantity of fluorine used in this kind of testing, but
it is anticipated to be relatively small.
5.4 ETCHING OF GLASS
Many publications refer to etched or frosted windows of buildings
in areas adjoining processes suspected of excessive releases of fluorides.
In virtually all instances, the etching was completed before it was given
any attention. In these cases, etching is the result of a condition that
has occurred at some previous time; and direct investigation of the cause
is no longer possible. Some experiments have been performed in which
levels of HF were maintained in fumigation chambers primarily to test groups
of flowers and small trees. As a secondary experiment, some panes of glass
were also included.(^360) eXperjmentai results were: (a) definite
etching resulted from an exposure totaling 9 hours at a level of 590 ppb,
and (b) pronounced etching resulted from an exposure totaling 14.5 hours at
a level of 790 ppb. These levels at which glass was etched by HF are-
high enough to have caused extensive damage to many species of vegetation
~"Fluorides" National Academy of Sciences, Washington, D.C. 1971, Committee
on Biologic Effects of Atmospheric Pollutants. "Current Knowledge indicates
that airborne fluoride presents no direct hazard to man, except in industri-
al exposure. However, through the commercial, esthetic and ecologic functions
of plants, fluoride in the environment may indirectly influence man's health
and well being."
5-10

-------
if those levels prevailed in the fields around buildings where fluoride
etching of glass has been observed.(^360)
5.5 EFFECTS OF FLUORIDES ON STRUCTURES
In the absence of water, HF forms a passive coating on steel. Highly
concentrated HF solutions (above 60%) and anhydrous HF are handled in
steel lines and containers. It is recommended that steel not be used
when concentrations of HF are below 48% in aqueous solutions. These
considerations are limited to process streams in which HF is being manu-
factured. Many other process streams have comparatively low concentrations
of fluorides, and the materials of construction will most likely be deter-
mined by some other constituent in the process stream such as sulfuric and
phosphoric acids.
At relatively low concentrations of fluorides in emissions from
fluoride processes, 1000 ppm or less, the damage caused by fluorides is
limited mostly to glass and brick. Etching of windows has been discussed
above. Occasionally, damage to the interior brick lining of a stack has
been attributed to fluorides in the emissions from an industrial process.
In the furnaces used for baking carbon anodes for aluminum reduction cells,
fluoride damage occurs to the high-silica brick used in the furnace walls.
5-11

-------
6. Measurement'
Technol ogy

-------
6. MEASUREMENT TECHNOLOGY
An inventory and evaluation of the technology for measurement of the
fluoride content of process streams has been performed. The results are
discussed in the following sections under the categories of sampling,
separation of fluoride from interfering ions, and analytical methods.
6.1 SAMPLING
6.1.1 Sampling Procedures
Selection of a sampling technique for measuring the fluoride con-
tent of effluents from process sources is dictated by the effluent stream
composition and the pollutants to be determined. For sources that emit
both particulate fluorides and gaseous silicon tetrafluoride and hydrogen
fluoride, chemical reactivity presents a major sampling problem. Such
sources include the industrial plants manufacturing phosphate fertilizer,
producing pig iron, processing iron and steel, reducing aluminum ore, and
manufacturing glass and ceramics. For accurate sampling of effluent and
differentiation between particulate and gaseous pollutants from such opera-
tions, the sampling technique must prevent interaction of the gaseous and
particulate fluorides.
Sampling procedures for use in the measurement of fluorides in the
atmosphere have been developed to prevent, to some extent, the interaction
in the collection train of gaseous and particulate fluoride. Unfortunately,
except for work carried out for the Office of Air Programs (formerly the
National Center for Air Pollution Control) by Dorsey and Kemnitz
Elfers and Decker,and the Manufacturing Chemists Association, no de-
tailed methodology (other than APCO Procedure H-7, Reference 304) is available
in the open literature covering stack sampling for fluorides. The devel-
oped techniques involve the sampling of stack effluents with a hot glass
probe followed by a heated train consisting of a cyclone, filter and a
Greenburg-Smith impinger containing distilled water. Particulate fluorides
are collected using a high-efficiency cyclone followed by a Whatman No. 41
filter. Active gaseous fluorides, such as HF and F2» react with the heated
glass probe to form gaseous silicon tetrafluoride which, after passing
through the heated cyclone and filter, hydrolyzes in the water of the
6-1

-------
Greenburg-Smith impinger to form fluosilicic acid and insoluble
orthosilicic acid. The water-soluble particulate fluorides, total particu-
late fluorides, and soluble gaseous fluorides can thus be determined
separately.
Some of the procedures used for sampling fluorides in ambient air
may be adaptable for sampling fluoride emissions from cyclones, baghouses,
electrostatic precipitators, or other so-called dry collection equipment.
At least the gaseous portion of the fluoride emissions from these collectors
might be adequately measured. The procedure using sampling tubes with
alkaline coatings could be used if a suitable dilution technique were
employed. However, the emissions from scrubbers using aqueous scrubbing
liquids require other measurement methods. Water vapor, droplets of
entrained scrubber liquid, and uncaptured fluoride particulates could all
be present at the scrubber exit. Considering these problems, even the
sampling of particulate was viewed with concern by Lunde^^ who stated,
"Adequate data are not available to evaluate the performance of the equip-
ment installed for the collection of particulate fluorides." His comment
refers to scrubbing devices using liquids to capture fluorides.
The most important constituent, the gaseous fluoride emission from
the scrubber, is the constituent most difficult to separate from such a
mixture. Total fluorides could be analyzed very efficiently, but the
ambiguity concerning the proportion of gaseous and particulate fluoride in
the emission would remain.
' 6.1.2 Performance of Sampling Trains
Mixtures of fluorides are usually evolved by industrial processes.
If there is a need to separate the particulate and gaseous fluoride compo-
nents, the sample train shown in Figure 6-1 has frequently been used for
this purpose. The particulate filter shown is a porous thimble. A variety
of filters and filter holders have been used. Some portion of the sample
may deposit on the inner surfaces beginning at the probe; therefore, to
minimize the reaction of HF with the sample train, stainless steel parts
have been used. As particulates collect on the filter surface, the dust
layer tends to become a collector for gaseous fluorides. Many dusts will
absorb or adsorb HF to some degree, and two patents extol the effectiveness
6-2

-------
1
a\
OJ
The components are: (1) sampling probe*, (2) dry filter; (3) impinqer
(dust concentration sampler); (4) ice bath container; (5) thermometers;
(6) mercury manometer; (7) Sprague dry qas meter; (8) vacuum pump; and
(9) hose clamp to control gas flow rate.
Figure 6-1. Schematic Diagram of Sampling Trairr
for Dry Particulate Matter

-------
of aluminum oxide for retaining HF.^^' ^168) Limestone dust -js wen
known for its ability to remove HF from air; however, little has been
published on the reactivity of HF with dusts such as fly ash from coal
burning, borates in glass making, and clays or other mineral dusts present
in industrial processes. It seems clear that in a sample train such as that
shown in Figure 6-1, the filtering section would collect a particulate sam-
ple with an indeterminate portion of the gaseous fluorides either reacted or
adsorbed. The aqueous collectors would retain the remaining gaseous fluoride.
Citric acid-treated filter paper allows gaseous fluoride to pass
through to be collected in a following section of the fluoride sampler.
The effectiveness of this arrangement for stack sampling would be com-
pletely dependent on very small dust loading of the filter; therefore, it
is not a promising method for sampling most industrial effluent gas streams
which contain appreciable amounts of dust.
Other sampling trains have included insertion of a small cyclone
collector ahead of the filter to trap dust larger than 25 microns, reducing
the amount of dust deposited on the surface of the filter. The filter has,
in some cases, been placed in the train following the aqueous collectors.
Where the quantity of particulate matter in an effluent stream is
large, the separation of gaseous and particulate fluorides is difficult.
However, control techniques are frequently concerned only with the determi-
nation of total fluoride content. The Greenburg-Smith impinger can be con-
sidered as the standard for collecting total fluoride though other col-
lectors are sometimes used. As reported in a review by Farrah,^^ the
Greenburg-Smith impinger is fairly rugged and has collection efficiencies
ranging from 90 to 98% when operated properly, at flow rates of 1.5 to
2.0 cfm.* Pack, et al,^^ and Farrah^^ report that pure water is as
good a collector as caustic solution for fluoride contaminants. The
impinger collection solution is usually diluted to a constant volume and an
aliquot taken from this solution for determination of the fluoride content
by the separation and analytical method selected.
*	(4308}
Keenan and Fairhallv ' found that lead fume particulate collection
efficiencies improved when a flow rate of 1.6 cfm was used with a standard
impinger designed for use at a flow of 1 cfm.
6-4

-------
Using the hot glass probe sampling technique, the particulate
contaminant (free of gaseous fluorides) is transferred quantitatively from
the cyclone and probe by washing with acetone after which the particulate
material is dried and weighed. The particulate contaminant collected on
the filter paper is combined with the cyclone-collected particulate mate-
rial; the filter paper is shredded; and the contents are diluted to a con-
stant volume. If, upon acidifying, the particulate material is not dis-
solved, caustic fusion as described by Pack, et al^^ is required for
complete recovery of fluoride. Aliquots are taken to yield the desired
quantity of fluoride for the analytical method selected.
If the reactivity of gaseous fluorides could be diminished by some
mechanism, difficulties in separating them from dusts could be reduced and
fluoride sample collection simplified. Since Siis less reactive than
HF and since HF can be converted to SiF^ through contact with heated glass,
this principle was employed in designing the sample train shown in Fig-
ure 6-2. This sampling train, developed by Dorsey and Kemnitz,^^ modi-
fled by Elfers and Decker^^ and described in APCO Procedure H-7, pro-
vides, with some limitations, detailed methodology for handling the total
range of fluoride contaminants in most process effluent streams and for
differentiation between particulate and gaseous fluorides in stack gases.
Potential problems with this sampling train still await solution.
Gelatinous silica hydrate is formed by the SiF- hydrolyzed in the
(340)
impingerv ' solution. A similar problem with gelatinous silica was
(4157)
solved by using ammonium compounds. ' Whether SiF^ would react with
Iron oxide dust on the surface of the filter may need to be tested; iron
(4310)
oxide is reported to react readily with SiF^. ' In some process gas
streams, the heated probe could become coated with dust, carbon, or tarry
materials to such an extent that the desired reaction of HF with glass
could not occur.
The technique developed by Pack, et al involving use of a glass
fiber filter for collection of suspensoid particulate contaminants can be
used instead of the cyclone and heated glass probe. The glass fiber or
paper filter (which separates and collects 98% of the suspensoid particulate
material) can be washed to remove soluble particulate fluorides. Filter
discs treated with alkaline reagents, used instead of a complex sampling
6-5

-------
STEEL NOZZLE
TEFLON FERRULE
STEEL FERRULE
WELD
PROBE SUPPORT
AND HEATER
GLASS INSERT
STEEL FERRULES
COMPRESSION FITTING
ABSORBERS
FILTER
PROBE NOZZLE
COUPLING
PROBE
CYCLONE
HEATED ZONE
/
DRY
SILICA GEL
200 ML
O.IN No OH
ICE BATH
DRY GAS METER
MANOMETER
I / noicirp
o
BY-PASS VALVE
PUMP
Figure 6-2. EPA Sampling Train
6-6

-------
train, will also collect total fluorides very satisfactorily. However,
some limits on the size of the fluoride sample collected may have to be
observed to avoid exceeding the capacity for absorption of the filter while
collecting a relatively large sample compared to the intrinsic fluoride
content of the glass fiber. Micropore-type filters may be used to collect
most, if not all, submicron-size particulate fluorides.
6.1.3 Process Factors Affecting Samp!ing
Some factors contribute to the sampling problems and are described
on an industry-by-industry basis.
The phosphate industry uses phosphate rock as raw material which can
cause problems because the rock does not have a fixed composition but varies
from mine to mine and even from area to area within the same mine. Some
phosphate rock behaves as though much of the fluoride was present as
fluosilicate and some as fluorspar. Since the raw materials are treated
differently in different processes, the form of the fluoride may be very
(42651
important. ' Heat may be added as in nodulation, with Sireleased
from the fluosilicate present; or acidulation may be used as in fertilizer
manufacturing with the Siescaping but the HF derived from CaF£ staying
in the slurry to react with some of the calcium carbonate; or heat, acid,
and silica may be added to the raw material as in the manufacture of
defluorinated rock with nearly all of the fluoride volatilized, probably as
a mixture of SiF^ and HF. In each instance, some pulverized phosphate rock
may be entrained in the effluent gas stream along with the volatilized
fluoride and water vapor released by the process reactions. Each of these
mixtures of fluorides may react differently as it is drawn into and through
the sampling train.
Little has been published describing fluoride effluent gas streams
related to iron and steel manufacturing. However, it is reasonable to
assume that: (a) fluorides added to the slag in steel furnaces may react
with, moisture to release HF, (b) fluorides may be converted to fluosilicates
in the slag and then thermally decomposed to release SiF^, and (c) fluorides
may sublime as iron fluorides since the sublimation temperature (1800° to
2110°F) is well below the pouring temperature of steel.(^310) comp-|ex_
ity of these reactions in the presence of dusts and moisture in the effluent
6-7

-------
gases could make sample collection very complicated and the analytical
results difficult to interpret. As little as half of the fluoride added in
steel furnaces is recovered in the slag from the steel processing.(4276)
The nature of fluoride evolution in aluminum reduction is complex
with HF, cryolite, alumina, aluminum fluoride, chiolite, and possibly heavy
hydrocarbons in the effluent gas from the aluminum reduction processes.
The evolution mechanisms for these materials have not been completely
described but fragments of the chemistry have been reported.	®®^' ^163)
Aluminum fluoride dispersed as a fume in air reacts with moisture to form
HF and aluminum oxide, but the rates of reaction are dependent upon vapor
pressure(^163) ancj temperature. ^®®^ Hence, as aluminum fluoride leaves an
aluminum reduction cell, hydrolysis starts as soon as it encounters atmo-
spheric moisture but diminishes rapidly as the fume cools. Cooling may
occur rapidly enough from some or much of the aluminum fluoride to remain
dispersed in the effluent gas stream as an unhydrolyzed fume. Sublimed
chiolite may rearrange into other solids as it condenses but in cooling
probably gives rise to a fine fume that reacts slowly or not at all with
moist air at ambient temperatures. Over the range of water vapor pressure
and temperatures that prevail in the fluoride collection systems used in
aluminum plants, there has been no really complete description of the
chemical and physical states of the fluoride to be sampled at various points
in emissions control systems.
6.1.4 Sampling Summary
Several of the devices for collecting fluorides from effluent
streams have been discussed. Some of them performed very well in sampling
fluorides dispersed in ambient air and separate gaseous and particulate
fluorides. For sampling industrial gas streams, too little testing has
been done to demonstrate the usefulness of these devices for fluoride
levels that may be far higher than those found in ambient air.
The types of sampling trains frequently used for stack sampling have
been discussed in relation to industrial fluoride effluent gas streams.
Since these effluent streams are usually mixtures of gaseous and solid
fluorides, separation of the two phases causes problems in sample collec-
tion. Particulates deposit on the interior surfaces of probes and
6-8

-------
sampling tubes. This dust and that collected on the surface of the filter
in the sampling train may absorb or adsorb significant amounts of gaseous
fluorides. Reactivity of gaseous fluorides with the sampling train compo-
nents may further interfere with the separation of the fluorides into
gaseous and particulate samples.
Inaccuracies related to fluoride sampling and analysis of process
streams are primarily caused by procedures used for collecting samples.
The materials that are collected can be analyzed relatively accurately for
fluoride content.
6.2 FLUORIDE SEPARATION
Before determining fluoride in particulate and gas fractions col-
lected from effluents, interfering ions must be removed if any of the well-
established analytical methods are to be used. Only aliquots providing the
quantity of fluoride for the analytical method selected should be used in
order that fluoride isolation can be performed with a minimum of work.
The separation of the fluoride ions from ions interfering in fluoride
+3 -3 - -2
analyses such as A1 , PO^ , CI , SO^ is accomplished by (a) distilla-
tion, (b) ion exchange, or (c) diffusion. The most widely used of the
(4311)
separation methods is the Willard-Winterv ' distillation. This method
on the macro-scale is considered the standard by which newer methods are
evaluated. Fluorine is separated as fluosilicic acid from interfering ions
by steam distillation from solutions containing perchloric,
sulfuric,(4312,4313) Qr phosphor acids.(^314) The f-]uosi 1 icic acid is
swept out of the distillation flask with water vapor, the boiling point of
the solution being held at a constant temperature by addition of steam or
water and by regulating the heat applied to the solution. The addition of
steam rather than water reduces the time required for the distillation and
eliminates bumping of solution.(^315)	^ or-jg-jnai sample is rela-
tively free of interfering materials and the fluoride is in a form easily
liberated, a single distillation from perchloric acid is carried out at
135°C. Samples containing appreciable amounts of aluminum, boron, or
silica require a higher temperature and larger volume of distillate for
separation. In this case a preliminary distillation from sulfuric acid at
165°C is commonly used. Large amounts of chloride are separated by
6-9

-------
precipitation with silver as an intermediate step. Small amounts of
chloride are held back in the second distillation from perchloric acid by
addition of silver perchlorate to the distilling flask.(2015) distil-
lation method requires considerable operator time and results in a large
volume of distillate for quantitative recovery (250-375 ml for samples
containing up to 100 mg of fluoride).
Isolation by an ion exchange resin allows recovery of fluoride in a
more concentrated form free of interfering ions. Nielsen and
Dangerfield^^ separated microgram quantities of fluoride on a quar-
ternary ammonium styrene resin with recoveries approaching 95% for amounts
of 20p.g or less from mixtures including hydrofluoric acid, sodium fluoride,
fluosilicic acid and calcium fluoride. The technique was used to concentrate
Willard and Winter distillate, and was also used directly on impinger-
captured atmospheric fluorides. Newman^17) removed interfering anions as
well as cations on a single exchange resin (Di-Acidite FF). Funasaki,
et al,	removed interfering ions P0^~^, AsO^"^, S0^~^, and COg"^ by
means of Amberlite IRA-400. Elution was affected with 10% NaCl. Dowex anion
exchange resin was used by Ziphin, et al,^^ to separate fluorides from
_3
PO^ with gradient elution of the fluoride from the resin by sodium
hydroxide. Nielsen^^ separated fluoride from Fe+3, Al+3, PO*-3, and
-2
SO^ on the resin and removed the ions by stepwise elution with sodium
acetate.
The ion exchange columns permit separations of 1 n-g to 0.1 g of
fluoride from interferences when the sample is in a few milliliters up to a
liter of solution. The elution volumes usually are about 50 ml.
Diffusion methods for separating fluoride from interferences before
determinations are simple and show great promise. They involve collection
of fluoride in volumes ranging from a few milliliters to a liter of alkaline
solution, the liberation of fluoride by treating with mineral acids, dif-
fusion through a short distance and absorption of the fluorides in approxi-
mately 5 milliliters of alkaline solutions. These methods are generally
applicable to quantities of fluorides in the 0.05 ^g to 1 mg range. Singer
and Armstrong(^320) anc| |_|al 1 (^321) SUggested the use of polyethylene bottles
for diffusion vessels which were sealed with stoppers. Alcock^3^)
prepared a satisfactory diffusion cell of Teflon that was used at 55°C;
6-10

-------
higher temperatures released fluoride from Teflon. Taves^^ found that
fluoride passes into trapping solutions in the form of methylfluorosilane
if silicone grease is used for sealing the diffusion cell. In the presence
of the simplest silicone, hexamethyldisiloxane, the separation of fluoride
is much more rapid. A faster diffusion method for the separation of
(4324^
fluoride was proposed-/ ' fluoride was liberated in the presence of
hexamethyldisiloxane in 6N hydrochloric acid. The separation was carried
out at 25°C for 2 to 6 hours, depending on the volume of sample analyzed.
Otherwise, the separation by diffusion takes place for at least 24 hours at
much higher temperatures (usually 60°C). Tusl^2^ established a rapid
diffusion technique using polyethylene diffusion cells to which were added
a purified high vacuum silicone grease that was a homogeneous mixture of
methyl si 1icone fluid and aerogel of silica. Following the diffusion
separation, fluoride was determined by the zirconium - SPADNS colorimetric
method. Stuart^26^ followed the diffusion separation with fluoride
determination with the fluoride specific ion electrode. He isolated
0.05 (j.g 200 (j.g from a large volume of solution to a 5-milliliter solution.
Because of its wide acceptance and ability to effect satisfactory
separations of fluoride with a minimum of equipment, the Willard-Winter
distillation technique is recommended for separating interfering ions in
the wide weight range from 0.1 ^g - 1 g of fluoride collected from plant
gaseous effluents. The distillation procedure described in Procedure H-7
appears satisfactory for most applications. Though the ion exchange isola-
tion of 0.1 p.g - 1 mg fluoride from samples collected from the atmosphere
is useful, there appears little need for this technique for use with
samples from plant effluents because of the larger quantities of fluoride
in the samples. For handling a large number of samples, the diffusion
separation techniques are capable of isolating fluoride from interferences
and concentrating it into 5 milliliters with the possibility of labor
savings.
6.3 ANALYTICAL METHODS
Analytical methods are discussed in several sections as indicated
below:
•	Spectrophotometry
•	Titrimetric
6-11

-------
•	Instrumental
•	Continuous and semicontinuous
As previously noted, aliquots from the samples collected should be
subjected to some process for separation of fluoride from interfering ions.
The aliquots to be taken for most efficient separations should be only as
large as those required by the selected analytical method. For the most
accurate analyses, the aliquot should provide a mid-range fluoride concen-
tration for that method. Examples of the fluoride concentrations found in
various stack effluents are listed in Table 6-1. Table 6-2 gives the
applicable concentration ranges for the various analytical techniques
described. The concentration ranges in the table and the following discus-
sions are for solutions containing fluoride ions after separation from
interfering ions.
6.3.1 Summary of Analytical Methods and Recommendations
The spectrophotometric methods have been developed to the point
where several are accepted as standards. After the separation of soluble
fluorides from interfering ions, spectrophotometric methods can generally
be used to determine fluoride with a relative precision of 5 to 10% for
solutions containing 0.01 M-g to 0.2 mg of fluoride per milliliter. The
claim in some publications of better precision is largely unsubstantiated.
The accuracy, except on standard solutions containing NaF, has not been
established but should be about the same as the relative precision for
solutions that do not contain any interfering ions. Little or no data
exist concerning total system accuracy, i.e., sampling, collection,
removal of interfering ions and spectrophotometric analysis.
Titration methods using indicators to detect the end-point are all
difficult to perform with a high degree of precision and have been super-
ceded, to a major extent, by the use of the specific ion electrode to
determine fluoride ion content either directly for relatively dilute solu-
tions, or by the use of titrimetric methods employing a specific ion
electrode for end-point determination.
6-12

-------
Table 6-1. Concentration of Fluoride Found in Various Effluents
(At standard temperature and pressure)

3
Grains/ft
w/v
mg/M->
v/v
ppm
NORMAL SUPERPHOSPHATE



Den Scrubber Emissions
nomi na1
Building Scrubber Emissions
0.08-0.30
-0.15
-0.00035
183-686
343
0.80
220-824
412
0.96
DI-AMMONIUM PHOSPHATE



Granulator Exhaust
Dryer Duct
Dry Screens Exhaust
0.0093
0.1100
0.0025
21.3
250
5.7
25
300
6.8
WET PROCESS PHOSPHORIC ACID



AP-57
Digester-Filters-Tanks
Scrubber Exhaust "Big Plant"
Scrubber Exhaust "Medium Plant"
0.0011-0.0147
0.001-0.03
0.0048
2.5-33.6
2.5-68.6
10
3-40
3-82.3
12
TRIPLE SUPERPHOSPHATE



Scrubber Inlet
Scrubber Outlet
Den Scrubber Inlet
Den Scrubber Outlet
Reactor and Granulator Scrubber
Exhaust
Dryer
Dryer Exhaust
Granulator Scrubber Inlet
Granulator Scrubber Outlet
0.55
0.016
0.10
0.008
0.0021
1.3
0.0025
0.48
0.030
1258
36.6
229
18.3
4.8
2970
5.7
1098
68.6
1500
43.9
275
22.0
5.8
3560
6.8
1320
82.3
DEFLUORINATED PHOSPHATE ROCK



Kiln Scrubber Exhaust
Fluosolids Scrubber Exhaust
Prep, Feed to Kiln
Di-Gal (from acid and limestone)
0.00056
0.0048
0.00095
0.00020
I.3
II,0
2.2
0.5
1.6
13.2
2.6
0.6
ELEMENTAL PHOSPHOROUS



Water Sol. F(updraft dryer)
Emissions Particulate (updraft dryer)
Furnace Exhaust Gas
0.0313
0.0099
0.0031
71.6
22.9
7.1
85,9
27.5
8.5
ALUMINUM PREBAKE ANODE



Primary Control Process (average)
Secondary Control Process
0.033
0.00006-0.00042
75.5
0.13-0.96
90.6
0.16-1.16
ALUMINUM VERTICAL STUD SOOERBERG



Primary (average)
Secondary Loading
0.43
0.00049
982
1.12
1180
1.34
ALUMINUM HORIZONTAL STUD SOOERBERG



Primary (average)
Secondary Loading
0.01
0.00026-0.00042
22.9
0.59-0.96
27.5
0.71-1.15
IRON AND STEEL



SINTER PLANT



Normal Conditions, water sol. F
Normal Conditions, particulate F
Special Conditions, water sol. F
Special Conditions, particulate F
Blast Furnace Stoves
Boiler House
Coke Ovens
Open Hearth, water sol. F
Open Hearth, particulate F
0.0023
0.0011
0.0042
0.00075
0.0027
0.00014
0.00068
0.0157
0.00004
5.3
2.5
9.6
1.7
6.1
0.32 .
1.6
35.9
0.09
6.4
3.0
11.5
2.0
7.3
0.38
1.9
43.1
0.11
6-13

-------
Table 6-2. Applicable Concentration Range of Analytical
Methods
Technique
Applied to AHquots
of Fluoride Concen-
tration Range
Precision
Interferences**
Coronents
FLUOROMETRIC




Morin or quercetin
0.5-20(ig
±5%
Fe+3 ,C204"Z.cr,Mn,N03
Better accuracy is claimed for these methods than for
visual end-point techniques for concentration above 2mg.
Specific 1on electrode
0.02^g/m]-20 mg/sil
(5 ml minimum sample)
±1%
OH"*
total Ionic strength
Preferred method of end-point detection in most cases.
Precision is better than other titr1«*tric procedures.
INSTRUMENTAL




Specific ton
electrode
0.03^g/ml to 30 mg/ml
SS standard
deviation at
low range, 2%
at high range
OH",Al+3,Fe+3,pH
adjusted
Ease of use and equal precision justifies use in most
cases in place of spectrophotometry methods.
Kinetic method
0.0004 ng-0.4 mg/ml
Not established
research method
so4-2,cr,Ait3,po4"3
Research method.
Atomic Absorption
G.0Q5 fig/ml-4 mg/ml
-51 standard
deviation
*>4"*. «4~3
Can be used over a wide concentration range. Useful
when a large number of samples are to be analyzed.
X-Ray of LaF.j
ljig-apx lOoig
-St standard
deviation
None
Research method that can be developed into a rapid
method.
Radio-release of
zirconium salt
10^-100 g
~5J relative
precision
P04-3,Fet3,Al+3
Research method.
Amperometrlc
0.5^/ml-lO p,g/ml
Not established
None
Research method; could be used to detect titration
end-points.
Photo-activation
Q.01S-5S in mg size
samples (dried)
-51 relative
precision
Cl,8r,S
Useful for small samples.
Mass spectrometry
0.1 molS-100 mol%
for Hf,S1F4>CF4,C2Fg
-5$ relative
precision
¦ None
Determination of HF difficult, useful for determining
organic bound flourine.
Electrochemical




Null point measurement
of cerium (IV) to (111)
lOpjg/ml -1 mg/ml
None given
Al*3.Fet3,P04"3
Research method.
Couloinetric
0.001 ng/ml -100 jig/ml
of F2
None given
None
Specific method for F^.
Gas Chromatography
lmg/cc-100* as HP or
SiF^ in gas sample
-5S relative
precision
None
Could be developed into an automatic method.
SPECTROPHOTOMETRY*



All the spectrophotometry methods suffer from interfer-
ences from ions that form more insoluble compounds with
the metal of the complex than the fluoride itself.
pK changes also effect most of these methods.
Lanthanum-Ali zari n
Complexone
0.01-0.4^/ml
t5S
p04"3,ai ,cr,re,K0,".
c2o4-i
Nitrate and phosphate interfere only when in excess.
Titanium-Chromatropic Add
2*ig-Q,2mg/ml
±5%
P04"3,AI,Fe,C204"2
,{S04*2 In excess)
Sulfate does not interfere.
K+,Na+,NH/,cr and NO," do not interfere in small
amounts.
Iron-Ferron
0,01-0.2 mg/ml
±51
P04"3,Al,Fe,oxaUte.
S04"2
Useful at higher fluoride concentration levels.
Thorium-Alizarin
0.0Vg-0.2 mg/ml
±5X
P04"3.S04"2,A1,Fe,
oxalate
Calibration not linear at.higher fluoride concentrations.
Zi rconium-SPADNS
0.01|xg-0.2 mg/ml
iSt
P04"3,cr,*l,Fe,oxjlate
Calibration not linear at higher fluoride concentrations.
Amadac-F
0.5-4pg/Bil
±st
P04*3,C1",A1,Fb,oxalate
Affected by high acid or alkali content, pH change, and
total ionic strength.
21 rconi uai-Er iochrom-
cyanln R
0.01|^-0.2 mg/ml
i5%
po4"3,C!".AI,Fe,oxalate
(S04"Z in excess)
Calibration 1s linear from 0,Ol-2jjg/m1.
TtTftUtCTlUC



Thorium nitrate is usually preferred as titrant.
Lanthanum is also used in some cases.
Visual




Purpurin sulfonate
1i±5X
H)4-3,S04-2Jil,Fe,C204-2

Alizarin Red S
lng-10 mg
>±5$
P04-3,S04-2Jll,Fe,C204-2
Listed in order of preference as indicators.
Er1ochromcy4n1r» R
lpg-10 eg
>45%
Fe,oxalate,CI",Nn.NO,,
etc. *

Photometric




(Same metal'dye com-
plexes can be used)
lpg-100 g

Same
Technique requires a colorimeter and is slower than
visual end-point; however, operator error is reduced.
*
Accuracy is not known for most methods except for standard solutions where the accuracy is the same as the precision.
Interferences usually removed by distillation, ion exchange or diffusion.

6-14

-------
6.3.2 Spectrophotometric Analysis
Interaction of fluoride ion with a metal dye complex generally forms
the basis for the colorimetric-type methods. The metals of the complex are
from the group Th, Zr, La, Ce, Y, Bi, Fe, and A1. This group is capable of
forming insoluble or slightly ionized fluorides and also insoluble
phosphates which is a well-known interferent. Some of the more common
dyes used for this purpose are Alizarin Red S, Eriochrome cyanine R,
arsenazo, Ferron, and SPADNS. Many of these dyes function as acid base
indicators and, therefore, require close control of pH in fluoride
determi nati on. ® ^
Many semi-quantitative and qualitative techniques have been used for
estimation of fluoride; while these are not spectrophotometric, they are
colorimetric and a typical example is discussed here. Mavrodineauu^^
describes a color complex for fluoride ion sample on dry zirconium or
thorium nitrate and a lake-forming dye (sodium alizarin) absorbed on filter
paper. No interference was noted for other halogens, but sulfate and
phosphate interfered. Semi-quantitative results could be achieved by acid
treatment and color intensity comparison.
Many color complex systems for the determination of fluoride spectro-
photometrically have been described in the literature. Generally, spectro-
photometric methods provide a means for measuring a 20,000 fold range of
fluoride concentration directly with very good sensitivity. Ranges for two
common systems are reported as follows: Iron-Ferron, 0.01 - 0.2 mg/ml
(1 cm cell) and 0.01 - 0.4 pig/ml for Lanthanum-Alizarin "Complexone"
reagents. Both these systems have visible spectrum absorptions. Belcher
and West^2^ report 200% increases in sensitivity by working in the ultra-
violet region of the electromagnetic spectrum.
Decolorization of Titanium-Chromotropic acid by fluoride ion with a
detection level of 2 pig/ml (total range 2 ^g to 0.2 mg per ml) was proposed
by Babko and Khodulina.(4328) No interference was observed from sulfate,
but phosphate must be removed. Sensitivity to pH is a problem common to
this technique. If a dye is added to the system, the resultant color change
can increase sensitivity to <0.5 ^g/ml. Skanavi^*^ applied this method to
micro quantities of fluoride with a sensitivity in the range 0.3 to 17 |j.g
6-15

-------
per ml; however, accuracy was not too good in this range. Phosphate
«|»	«|»	«|»	S	<¦	a
interferes, but K , Na , NH^ , S04 , CI , and NOg in small amounts do not
cause problems.
Mal'kov and Kosareva^^ outlined a method using thorium-alizarin
to form a colored complex with fluoride. Range of the method is 0.01 (lg
to 0.2 mg fluoride per milliliter.
Amadac-F sold by Budick and Jackson Laboratories(^330) n-s a mixture
of alizarin complexan, lanthanum nitrate, acetic acid, partially hydrated
sodium acetate and stabilizers useful for quantitative determination of
fluoride in the range 15 - 50 (Jig per milliliter. A color change is
observed in this reaction complex which is affected by high acid or alkali
content, pH change, and total ionic strength.
Green iron-Ferron complexes with fluoride to produce a color change
useful for fluoride measurement in the range 0.01 - 0.2 mg/ml. Adams(^331)
in discussing this work proposed the use of this method for stack monitor-
ing with removal of sulfur dioxide, an interferent, by sodium tetrachloro-
mercurate absorber solution. A prior reference^6^ utilized airconium-
Eriochrome cyanine R for the determination.
A recent spectrophotometry technique described by West, Lyles, and
Miller^^ analyzes fluoride by complexing with alizarin complexan and
lathanum buffer. Determinations in the range 0.01 - 0.4 i^g/ml can be done
if metals, nitrates, and phosphates are removed. That is, concentrations
of'<4 fig/ml nitrate and <3 ng/ml phosphate in 0.4 i^g/ml of fluoride are
tolerable.
Because of the large volume of literature and numerous possible
combinations of metal-dye-fluoride complexes, the above summary must be
considered only as a few typical recognized procedures which reflect the
possibilities of the spectrophotometry technique. The methods described
here generally can be considered as new techniques or the latest modifica-
tion of older techniques. It is very difficult to make a statement of
preference for any of these methods unless dynamic range of applicability
of specific interference are the judgment criteria. Sensitivity and preci-
sion are nearly the same for each method.
6-16

-------
6.3.3 Titrimetric Analyis
The most commonly used titrants for the volumetric or titrimetric
determination of fluoride in aqueous systems are thorium and lanthanum
nitrate. However, because of the large variety of end-point detection
procedures the classification of the titrimetric methods will be based on
the detection technique utilized. End-point detection can be generally
broken down into the following types: visual, photometric, fluorometric,
specific ion electrode, and electrometric. While specific ion electrodes
may be classed under electrometric, their relative importance dictates a
separate class for this discussion. The first four types will be evaluated
in this section, where titration of the total sample distillate, thus
preventing dilution error, is possible.
The visual end-point detection procedures have generally been sup-
planted by other means for end-point detection and by spectrophotometry
methods. Photometric, fluorometric, electrometric and specific ion elec-
trode end-point detection have largely eliminated the operator perception
and dilution errors present in the visual methods.
6.3.3.1 Visual
The greatest difficulty in the quantitative utilization of the color
indicator end-point techniques for fluoride is that it depends on the color
perception and experience of the operator. Many indicators have been
suggested for improving the subtle color change; however, there still
remains much to be done. Much work has been done by Willard and
' Horton^3^ on these as well as other systems with the following colori-
metric indicators being recommended: Purpurin sulfonate, Alizarin Red S
Eriochrome cyanin R, dicyano-quinizarin, and Chrome Azurol S.
In visual procedures the sample of fluoride is titrated with thorium
or lanthanum nitrate to the end-point as indicated by one of the.suggested
complex colorimetric (visual) dyes. Generally the methods using visual
indicators to detect the end-point are used in the fluoride concentration
range of 1 n^g to 10 mg in an aliquot from 10 to 200 ml. Analysis in this
range is described in ASTM Method D1606-60.
6-17

-------
Not only do these titrations suffer from the above mentioned operator
error, but other problems exist depending on the composition of the sample
to be analyzed. When large amounts (above 1 mg) of fluoride are titrated,
interference may result from semi-colloidal thorium nitrate, and medium-
to-high concentrations of metals, nitrates, and phosphates interfere in
most cases.
Allison^^ in his work, compared the determination of fluoride by
both volumetric visual end-point detection and a colorimetric (spectro-
photometric) method with the conclusion that the latter technique was more
sensitive and should be used in the 0.5 to 50 n-g/ml range, while the
volumetric was faster and more useful for concentrations above 50 (ig/ml.
6.3.3.2	Photometric
The real advantage in using a photometer to determine the end-point
in a fluoride determination lies in the elimination of the variable of
operator perception differences. All the other parameters remain essen-
tially the same as for the visual indicator method above.
6.3.3.3	Fluorometric
Here the dyes recommended for use are different from the visual indi-
cator dyes because of the requirement for measuring fluorescence changes to
detect the end-point. Willard and Horton^^ recommend two; pure
sublimed morin and quercetin. The titration is again carried out using
thorium nitrate, while the end-point is observed by the fluorescence change.
Better accuracy is claimed for this method than for the color end-point
method for fluoride concentrations greater than 2 mg. Many variables again
need to be controlled, such as pH and interfering ions.
Willard and Horton^^ also describe a fluorometric technique for
the determination of trace amounts of fluoride using aluminum-oxine or
aluminum-morin systems. In these systems the fluoride complex with
aluminum decreases the aluminum-oxine or morin complex. The resultant
change in fluorescence of the system is measured. The range of sensitivity
to fluoride is around 0.5 to 20 (ig total sample. Many variables must be
controlled and standards should be run with each set of unknowns. Ions that
react with aluminum or oxine or which precipitate with fluoride at pH 4.7
must be removed.
6-18

-------
6.3.3.4 Specific Ion Electrode
The use of the fluoride specific ion electrode (lanthanum fluoride
membrane electrode) for end-point detection is a recent innovation and is
discussed by Lingane^4334) and Frant and Ross Jr.^4335^ The conclusions
reached by these and other investigators point out the usefulness of this
technique. Sensitivity to fluoride over a concentration range of five
orders of magnitude is easily achieved while ultimate sensitivity is down
to 10"7 M fluoride. The electrode is very selective to fluoride, but pH
and total ionic strength are very important considerations in the analysis.
In typical titrations of fluoride with thorium and lanthanum nitrate, the
latter yielded the best potential break with precision to ±1 mv. Far
better end-point accuracy and precision were achieved using the electrode
than could be achieved using color indicators for detection. The useful
range of this end-point detection method is for solutions in the concentra-
tion range of 0.1 (ig to 20 mg/ml.
Schultz^433^ points out that large errors can result from poten-
tiometric titrations employing ion-selective electrodes. The error
increases as the sample ion concentration decreases and as the interfering
ion concentration, solubility product constant, and dilution factor increase.
Of the above mentioned end-point detection methods the fluoride
electrode technique is the most precise (interference removed) and generally
the easiest to apply.
6.3.4 Instrumental Methods
Nearly every analytical instrument has been investigated for direct
determination of fluoride. Many of these instruments have been previously
discussed as detectors for titrimetric end-points, but in this section
instrumental methods are discussed as they apply to direct determination of
fluoride either as collected or after separation from interferences common
to most analytical methods. The various instrumental techniques are dis-
cussed below.
6.3.4.1 Specific Ion Electrode
The accepted dynamic range for the new fluoride specific electrodes
is from 50 mg/ml down to 0.10 |xg in the minimum usable volume of 5 ml.
Preliminary work with this electrode has shown promise of making fluoride
6-19

-------
ion determinations virtually as simple, rapid, and precise as hydrogen ion
activity measurements with the glass pH electrode. It must be remembered
that fluoride activity is measured and concentration is dependent on total
ionic strength as well as other factors.
(327)
Harriss and Williams ' discuss the direct measurement of fluoride
with the specific ion electrode and noted the speed and low cost of this
(4337)
analysis. Baumannv ' describes the interference from hydroxyl ion and
_5
its elimination for accurate fluoride analysis. As little as 10 M
fluoride (1 ^.g in 5 ml) could be analyzed with a relative error of =10% and
standard deviation of <5%. He suggested that interfering ions be complexed
before fluoride analysis. Electrode response time was less than one minute
in these experiments. Durst and Taylor^^ describe microchemical analy-
sis techniques for fluoride using the electrode.
A comparison of the specific ion electrode to the Spadus-Zirconium
(4265)
method by Elfers and Deckerv ' showed good agreement between the two,
but the electrode technique was much faster. Their reported detection
limit was 0.2 fig fluoride in a 5 ml aliquot.
Because of the importance of the total ionic strength on the fluoride
concentration measurement with the electrode and the effect of acidic or
basic media on the values of fluoride, it is necessary to control or eval-
uate these parameters. Vanderborgh^^"^9^ used a lanthanum fluoride mem-
brane electrode in his study of response in an acidic solution with varying
ionic strength. A recent article by Bruton^^ for the useful range of
the electrode points out that the known addition technique can be used for
the simple and accurate analysis of fluoride. The millivolt readout for the
electrode is adjusted to zero in the sample, an addition of standard
fluoride is made, and the change in potential is related to fluoride con-
centration. The activity coefficient must remain constant for accurate
measurement; where necessary this can be accomplished by the addition of a
noninterfering salt.
Frant and Ross^^ adjusted the total ionic strength, the pH, and
complexed ferric iron or aluminum (citrate used) by using a buffer in a
1/1 ratio with the samples and standards. Fluoride could be determined
accurately over the entire useful electrode concentration range using a
single calibration curve for a wide range of samples.
6-20:

-------
6.3.4.2	Kinetic Method
Because a kinetic method employs unusual instrumentation, the kinetic
(43411
method is included under the instrumental section. Klorkow, et ar '
developed a kinetic method for the determination of traces of fluoride
(3.8 x 10 - 3.8 fig/ml) based on strong inhibiting action. Fluorides act
as a negative catalyst in the zirconium-catalyzed reaction between
perborate and iodine. Kinetic measurements are accomplished by an auto-
matic potentiostatic technique. Only small quantities of extraneous ions
can be present.
6.3.4.3	Atomic Absorption
Bond and O'Donnell^^ applied the depression of absorption of the
magnesium line at 285.2 m^ into an atomic absorption method for fluoride
in the range of 0.005 n-g/ml - 2000 mg/1 using an air-coal gas flame. Both
-2	-3
S0^ and P0^ ions must be absent. A somewhat less sensitive method
(5-500 ng/1) was also established based on the enhancement of zirconium
absorption by fluoride in the nitrous oxide-acetylene flame. They also
established an even less sensitive method (400 - 4000 mg/1) without inter-
ference based on the enhancement of titanium absorption. These methods
offer the advantage of being rapid, and little handling of collected
samples is required.
6.3.4.4	X-Ray Spectrography
An X-ray spectrograph!'c method^4342^ was established for measuring
fluoride collected by nearly any of the previously described sampling
methods, adjusting the pH of the solution containing the fluoride, and
collecting the fluoride as LaF^ on a Mi Hi pore disc with a pore size of
2(x. The disc is washed, dried and submitted to X-ray spectrograph!c
measurement with a tungsten target and a lithium fluoride analyzing crystal.
Fluoride can be detected in the range of 1 p.g to about 10 mg without
interference.
6.3.4.5	Polarographic
MacNulty, et al,^535^ applied to fluoride determination the reduction
of the polarographic half-wave potential at 0.3V versus saturated-calomel
electrode (pH 4.6 in acetate buffer) for the sodium salt of 5-sulfo-2
6-21

-------
hydroxy- a benzene - azo-2 naphthol in the presence of aluminum. Fluoride
complexes aluminum and reduces the half-wave potential. The method can
detect 0.2 jig/ml but the method precision and the maximum concentration of
fluoride that could be detected was not investigated.
6.3.4.6	Radio Release
Carmichael and Whitley^4343^ established a radio-release method for
determination of fluoride (20 to 100 ng). The fluoride is converted to a
zirconium salt, placed in a neutron flux, and the radioactivity release
-3+3	+3
measured. The relative precision is about 5%; P04 , Fe and A1
interfere.
6.3.4.7	Amperometric
A patent^4344) was issued for an amperometric method for fluorides
from an air sample that was collected in 0.5 M nitric acid. The fluoride
is determined by a platinum wire or plate electrode and a zirconium wire
electrode rotating at 300 to 1600 rpm and maintained between -1 and +1 volt
with respect to an S.C.E. The current passing between the electrodes was
measured. The method can measure 0.05 jig - 1 fig/ml in the collecting solu-
tion. This method was not applied to detecting titration end-points, but
could be considered.
6.3.4.8	Photo Activation
Kosta and Slunecko^4345^ demonstrated the use of photo activation for
determining fluoride in the concentration range of 0.01 to 51 on as little
as 1160 ng of sample. The method has not been applied to gas stream
fluorides, but could be used to determine the fluoride content of particu-
lates collected on a filter. Interferences from elements such as chlorine,
bromine, and sulfur can be eliminated or reduced to a minimum by adjusting
irradiation time, waiting period, and energy of the primary electron beam.
Results obtained were in good agreement with those obtained by distillation-
titration methods.
6-22

-------
6.3.4.9	Mass Spectrometric
Mass spectrophotometry analysis of anode gases from aluminum
reduction cells was accomplished by Henry and Holliday^892^ for HF, SiF^,
CF^ and CgFg. The method determined the substances from 0.1 mol % to
100 mol %.
6.3.4.10	Electrochemical
(4346)
Curran and Fletcher1 ' determined fluoride by precipitating
fluoride ions with lanthanum ion electrochemically generated from lanthanum
hexafluoride anode. The end-point was detected with a fluoride specific
ion electrode.
Fluoride was determined by null point potent!ometric measurement of
the cerium (IV) cerium (III) reduction potential^4347^ for solution con-
taining greater than 14 fig of fluoride per ml of solution. The method can
be applied to Willard and Winter distillates or ion exchange eluates.
A coulometric specific method was established by Kaye and Griggs^260^
for free fluorine in a gas stream. In this method gas is aspirated a.t
constant flow rates between 100 and 300 ml/min through 0.2 MLiCl. The
fluorine oxidizes the CI" with one mole of fluorine corresponding to two
atoms of CI". The quantity of fluorine is determined coulometrically using
a silver anode and a platinum gauze cathode. The method determines 0.1 ppm
up to about 100 ppm of fluorine.
6.3.4.11	Gas Chromatography
The analysis of various fluorine containing compounds was investigated
by Pappas and Million.^4348^ They found the high affinity of HF toward
almost any surface to be a problem. By the use of Teflon columns prepared
with Teflon-6 support coated with fluorocarbon oil and carrier gas spiked
(4349)
with HF as proposed by Knight/ ' they found that HF along with other
fluorine compounds in the concentration range of 0.1 mol % to 100 mol %.
could be separated and measured with a gas density balance. Air was used
as the carrier gas, but greater sensitivity than the ppm level they
observed could be achieved with a carrier gas such as sulfur hexafluoride.
More sensitive detectors have not been investigated.
6-23

-------
6.3.4.12	Infrared Spectrometry and Infrared Lasers
Hydrogen fluoride has an absorption band at 3961.6 cm~\ but this
absorption band has not been used for determinations in the stack gases
because of interactions with water vapor. SiF- has absorption bands at
1010 and 800 cm . A recent review of infrared lasers for monitoring air
(4350)
pollution by Hanstv ' proposed a method using a Kr laser and the infrared
absorption line for HF.
6.3.4.13	Instrumented Methods Summary
With the exception of the specific ion electrode, the instrumental
methods presented are useful for only special cases. Because the specific
ion electrode is accurate when properly used and easy to use, the specific
ion electrode is recommended for fluoride determination whenever possible.
6.3.5 Continuous and Semi continuous Methods
There are at present no continuous or semi continuous methods in use
for the determination of the fluoride content of the various gaseous
effluents from manufacturing processes and the abatement systems employed
in connection with the processes. Most continuous or semi continuous
methods were developed for measuring fluoride content of the ambient atmo-
sphere. These methods are summarized here because they can be considered
as candidates for continuous monitoring of plant effluents.
The analysis of air for detecting fluorine compounds in the parts per
billion concentration range is usually done by aspirating a large volume of
air through distilled water or dilute alkali, concentrating the fluoride by
distillation, ion exchange or diffusion (as discussed under titration and
colorimetric methods), and then determining the quantities of fluorides by
titrimetry or colorimetry. The distillation or ion-exchange step can be
omitted only in special cases. Collection of enough fluoride for analysis
may take several hours to one or two days, thus giving long term average
concentrations. A method for continuous determination of fluoride content
of process stream effluents is needed. The various approaches that have
been developed are:
•	Mini-Adak Colorimetric Analyzer
•	Fluorescence-Quenching Method (SRI Fluoride Recorder)
6-24

-------
•	Billion-Aire Ionization Detector
•	Current Flow Method
•	Specific Ion Electrode Method
6.3.5.1	Mini-Adak Analyzer
In 1956, Adams, Darra and Koppe^^ reported on a prototype
photometric fluoride analyzer for use with a liquid reagent. In 1959,
(691)
Adams and Koppe^ ' studied this instrument extensively and established
that there was an excellent correlation between the instrument and conven-
tional sampling and analytical procedures for total soluble ion-producing
fluoride pollutants. Basically, the instrument may be characterized as a
recording flow colorimeter in which the flow forms an integral part of the
air-reagent absorption system. As a fluoride analyzer, it photometrically
measures and records the rate of reaction of zirconium-Eriochrome Cyanine R
reagent with concentration of soluble fluorides in a sampled air stream
throughout a given sampling period. High sensitivity is achieved by an
unusual absorber which permits the contact of a small volume of liquid with
a large volume of air (1 cu ft per min). The efficiency of hydrogen
fluoride absorption is reported to be 951. The volume of liquid is kept
constant by automatic addition of water to replace evaporation losses. The
liquid is periodically discarded and replaced by a measured volume of fresh
solution. The addition of fluoride ion to the zirconium-Eriochrome Cyanine
R reagent shifts the absorption maximum to 550 m^.. The color is measured
continuously by a recording colorimeter. Response of the recorder to
fluoride is nearly linear at 1 scale division per ng of fluoride ion per
15 ml solution until 20 fig are added. In the range from 0.75 to 35 ppb
hydrogen fluoride, the standard error of estimate was about 0.8 ppb.
6.3.5.2	Fluorescence-Quenching Methods (SRI Fluoride Recorder)
Chaikin and Associates at Stanford Research Institute developed a
fluoride recorder^^9,4353)	was further modified by Thomas, St. John,
and Chaikin^^'^25^ to provide an instrument that could operate under
field conditions. The method consists of drawing parallel air streams into
the analyzer through warmed glass tubes; one tube coated with NaHCO^ and
the other clean. The NaHCOg absorbs hydrogen fluoride, but the clean tube
allows it to pass. The air streams are drawn through adjacent spots on
6-25

-------
sensitized paper tape made by dipping chromatrography paper in a methanol
solution of 8-hydroxyquinoline and magnesium acetate. The resulting
magnesium salt of 8-hydroxyquinoline fluoresces when illuminated with ultra-
violet light. The visible fluorescence is quenched by hydrogen fluoride,
thus providing a quantitative measure of fluoride. The difference in
emitted light from the two areas of paper type is monitored by reflecting
the two beams of light onto balanced photomultiplier tubes. Differential
output from the tubes is recorded on a strip chart recorder. This instru-
ment is 50 to 100 times more sensitive than the Adak recorder and deter-
mines only hydrogen fluoride, not total fluoride. The instrument can
detect hydrogen fluoride in the range 0.2 to 10 ppb and appears free of
interferences by common air pollutants. However, the instrument requires
additional field testing.
6.3.5.3 Billion-Aire Ionization Detector
The Billion-Aire Ionization Detector manufactured by the Mine Safety
Appliance Company^5^ lends itself to the detection of hydrogen fluoride
and fluorine in air. When a gas is ionized in the detector between two
oppositely charged electrodes, a current is conducted depending primarily
on the strength of the ionizing source, the applied voltage, and the
composition and pressure of the gas. With air in the detection chamber,
most gaseous additives in the concentration range of several thousand parts
per million will cause only small changes in ion current. However, very
small concentrations of finely divided particulate matter produce a pro-
nounced decrease in current. The action of particles is to promote effec-
tive recombination through third body collisions and to decrease mobility
through attachment. By converting a gas to particulate matter by a suitable
reaction and measuring the decrease in ion current due to the presence of
the particles in an ion chamber, many contaminants can be detected in the
concentration of ppb. For example, HF, HC1 and N02 can be converted to
particulate aerosol by reaction with ammonia. Though the instrument is not
specific for hydrogen fluoride, it provides instantaneous response for
hydrogen fluoride and fluorine concentrations of 1 to 100 ppb.
6-26

-------
6.3.5.4	Current Flow Method
Howard, et	developed a portable fluoride analyzer based on
the fact that the current from an aluminum-platinum internal electrolysis
cell is a function of the fluoride content of an acetic acid electrolyte,
after the sampled air has been scrubbed with the electrolyte. The
analyzer responds to all substances which form fluoride ion in aqueous
solution and is specific for fluoride in the presence of common contaminants.
The method is capable of detecting from 5 to 100 ppb for a 2-liter sample.
A second electroanalytical instrument that detects fluorine but not
fluorides was developed by Kaye and Griggs.in this instrument, air
containing fluorine exidizes CI" ion in a buffered LiCl solution in a solu-
tion containing platinum and silver electrodes. The chlorine formed is
reduced at the silver cathode. Insoluble AgCl is produced on the cathode
so that chloride is removed from the solution. For every molecule of
fluorine, two electrons flow through the coulometric circuit and two atoms
of chlorine are transformed from solution to cathode. By using a pump to
deliver a constant flow of air to the instrument one can determine fluorine
concentrations between 5 and 1000 ppm without interferences from other
atmospheric oxidants.
6.3.5.5	Specific Ion Electrode Method
Light^^ discussed the adaptation of the fluoride ion specific
electrode to the continuous monitoring of gas streams. By simply scrubbing
the gaseous constituents with a suitable reagent and measuring the quantity
of gas, reagent solution, and concentration of the resulting solution with
the electrode automatic monitoring can be achieved. Direct application of
this technique to effluent gas analysis as yet has not been reported, but
recently has been applied by Mori, et al^3*^ to the determination of
hydrogen fluoride in the atmosphere. The hydrogen fluoride is collected by
absorption on dry sodium carbonate coated glass tubes. The sodium carbonate
and collected fluorides are washed to a collection container, the solution
buffered, and the fluoride concentration determined with the specific ion
electrode. Automatic cycling of the apparatus provides a continuous
recording of the hydrogen fluoride concentration of air.
6-27

-------