EPA-450/3-80-015
Source Category Survey:
Animal Feed
Defluorination Industry
Emission Standards and Engineering Division
Contract No. 68-02-3058
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
May 1980 ,
-------
This report has been reviewed by the Emission Standards and Engineering
Division, Office of Air Quality Planning and Standards, Office of Air, Noise,
and Radiation, Environmental Protection Agency, and approved for publica-
tion. Mention of company or product names does not constitute endorsement
by EPA. Copies are available free of charge to Federal employees, current
contractors and grantees, and non-profit organizations - as supplies permit
from the Library Services Office, MD-35, Environmental Protection Agency,
Research Triangle Park, NC 27711; or may be obtained, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
VA 22161.
Publication No,. £PA-450/3-80-015
ii
-------
TABLE OF CONTENTS
1. SUMMARY 1 -1
1.1 INDUSTRY DESCRIPTION. 1-1
1.2 PROCESS DESCRIPTION AND EMISSION SOURCES 1-1
1.3 CONTROL SYSTEMS 1-2
1.4 STATE AND LOCAL EMISSION REGULATIONS 1-3
1.5 SAMPLE COLLECTION AND ANALYSIS 1-3
1.6 RESULTS AND RECOMMENDATIONS 1-4
2. INTRODUCTION 2-1
3. CONCLUSIONS AND RECOMMENDATIONS .. 3-1
3.1 CONCLUSIONS 3-1
3.2 RECOMMENDATIONS 3-2
4. DESCRIPTION OF INDUSTRY 4-1
4.1 SOURCE CATEGORY 4-1
4.2 INDUSTRY PRODUCTION 4-1
4.3 PROCESS DESCRIPTION 4-5
4.4 REFERENCES. 4-8
5. AIR EMISSIONS DEVELOPED IN THE SOURCE CATEGORY 5-1
5.1 PLANT AND PROCESS EMISSIONS 5-1
5.1.1 Feed Preparation Emissions 5-1
5.1.2 Thermal Defluorination Emissions 5-4
5.1.3 Product Storage and Shipping Emissions 5-5
5.2 TOTAL NATIONAL EMISSIONS FROM SOURCE CATEGORY 5-5
5.3 REFERENCES :... 5-7
6. EMISSION CONTROL SYSTEMS 6-1
6.1 CONTROL APPROACHES .6-1
6.2 ALTERNATE CONTROL METHODS 6-2
6.3 "BEST SYSTEMS" OF EMISSION REDUCTION 6-3
6.4 REFERENCES 6-6
7. EMISSION DATA 7-1
7.1 AVAILABILITY OF DATA 7-1
7.2 SAMPLE COLLECTION AND ANALYSIS 7-1
7.3 REFERENCES 7-4
8. STATE AND LOCAL EMISSION REGULATIONS 8-1
8.1 REFERENCES 8-3
m
-------
LIST OF TABLES
Table
4-1
4-2
5-1
6-1
7-1
Summary of Animal Feed Defluorination
Plants in the United States
List of Persons with Expertise in the
Animal Feed Defluorination Industry
Estimate of Total National Emissions of
Particulates and Fluorides Developed in the
Animal Feed Defluorination Industry
Best Systems of Control...
Emission Source Test Data.
.Page
. 4-2
. 4-3
. 5-5
. 6-4
. 7-2
LIST OF FIGURES
Figures
4-1
5-1
Generalized Animal Feed Defluorination
Industry Process Flow Diagram
Generalized Flow Scheme for Florida
Phosphate Rock Operations
Page
,. 4-6
...5-2
IV
-------
1. SUMMARY
This report documents a study conducted to assess the need for
new source performance standards (NSPS) for the animal feed defluori-
nation industry. These standards would regulate airborne emissions
from new point sources involved in the manufacture of defluorinated
animal feed products.
This chapter is provided as a general overview of the study. ,
1.1 INDUSTRY DESCRIPTION
Defluorinated animal feed is a phosphate rock product that is
added as a supplement to many animal feeds. It can constitute up
to 10 percent of a given animal feed. Phosphate rock defluorination
is accomplished by heating a mixture of phosphate rock, phosphoric
acid, and caustic soda to temperatures up to 1370°C (2500°F). The
fluorine content of the phosphate rock is thereby reduced from 3.5
percent to under 0.2 percent—a level safe for consumption by animals.
Presently, there are only three animal feed defluorination
plants operating in the United States, and they are all located in
Florida. The total combined production capacity of these three plants
is 454.0 Gg/yr (500,000 ton/yr). Actual production over the past
several years has been less than capacity. No significant growth is
expected in the industry in the foreseeable future, but some expansion
of the existing plants could occur.
1.2 PROCESS DESCRIPTION AND EMISSION SOURCES •
Animal feed defluorination can be broken down into three main
processes: feed preparation, thermal defluorination, and product
storage and shipping. Feed preparation primarily involves the mixing
and drying of raw materials. During thermal defluorination the
prepared mixture is heated in order to volatilize the-fluorine contained
in the phosphate rock. The defluorinated product is then crushed to the
1-1
-------
proper size and either stored or shipped via rail or truck.
All sources of emissions directly involved with animal feed
defluorination are considered in this study. Included are mixing,
drying, thermal defluorination, and storage and shipping. Similar
processes related to other parts of the phosphate rock industry
(such as the phosphate fertilizer industry) are not considered since
they are dealt with in a separate NSPS study.
Particulate matter is emitted to some degree from all of the
processes mentioned above. Fluorides are emitted during the thermal
defluorization process and, to a much lesser degree, during drying.
The only emission data available are for well-controlled sources, and
then only for mixing, drying, and'thermal defluorination. Controlled
particulate and fluoride emission rates for the combined mixing and
drying processes are 0.11 kg/Mg P205 (0.22 Ib/ton P20g) and 0.0065 kg
F/Hg P205 (0.013 Ib F/ton P205)» respectively. Average particulate and
fluoride emission rates for thermal defluorination are 0.65 kg/Mg P20g
(1.29 Ib/ton P205) and 0.13 kg F/Mg (0.26 Ib F/ton P205), respectively.
Total national emissions from the quantifiable processes are 138 Mg/yr
(142 ton/yr) of particulates, and 24.8 Mg/yr (27.3 ton/yr) of
fluorides.
1.3 CONTROL SYSTEMS
Emissions of particulates and fluorides from all sources are
controlled to a high degree. Baghouses or wet scrubbers control fugitive
particulate emissions resulting from the handling of raw materials and
product. Wet scrubbers are used to control particulate and fluoride
emissions from the mixing and drying operations. The major source of
emissions from animal feed defluorination plants, the thermal defluorina-
tion process, is controlled most commonly by a cyclonic entrainment
separator followed by a spray crossflow packed scrubber. Particulates
are removed and recycled by the cyclone, while fluorides and additional
particulates are removed by the scrubber.
Best demonstrated emission control technology, with one minor
exception, is essentially the same as described above: baghouses or
1-2
-------
scrubbers for raw material preparation and product handling, and cyclone/
wet scrubber systems for thermal defluorination. The different operating
characteristics of the fluidized-bed thermal defluorination reactor (as
opposed to the rotary kiln), however, may require one additional piece
of control equipment. An ionizing wet scrubber (IWS) may be needed to
control fine particulate emissions. Therefore, a cyclone/wet scrubber/IWS
system constitutes best demonstrated control technology for a fluidized-bed
reactor, while the IWS is not required when a rotary kiln is used.
The storage and shipping operations produce fugitive particulates
during the loading of railcars and trucks. No emission data are available
for this operation, but the best control system for this operation
utilizes local exhaust hooding and ventilation ducted to baghouses.
Visible observations of this operation indicate that this system effectively
controls fugitive emissions.
1.4 STATE AND LOCAL EMISSION REGULATIONS
All three of the operating animal feed defluorination plants are located
in Florida and are regulated by the Florida Department of Environmental
Regulation (DER). The DER has promulgated regulations for both particulate
and fluoride emissions from this industry. These regulations specify
limits which require that the best demonstrated control technology be
installed.
Allowable fluoride emissions are limited to 0.37 pounds of.fluoride
per ton of P20g input to the process. Based on a 3.5 percent fluoride
content and a 35 percent P205 content for the feed material, this limit
requires a control efficiency of 99.8 percent or greater.
Particulate emissions are regulated on the basis of a process weight
table. No control efficiencies can be determined due to the lack of data
on input particulates to the control devices. However, the most commonly
used particulate control device, baghouses, typically achieve 99 percent
or greater control.
1.5 SAMPLE COLLECTION AND ANALYSIS
There are EPA reference methods for the pollutants of concern from
this industry, particulates and fluorides. EPA Method 5 is used for the
1-3
-------
determination of-particulate matter. EPA Methods ISA and 13B are used
for the determination of total fluorides.
1.6 RESULTS AND RECOMMENDATIONS
The results of this study indicate that there is no need for the
development of NSPS for the animal feed defluorination industry. This
recommendation is based on two important conclusions. First, all of
the operating plants in this industry are located in the State of
Florida and are well regulated. These plants have installed the best
demonstrated technology to comply with state regulations. Second, there
is limited potential for growth in this industry, and any growth or
expansion which does occur will most likely occur at the present sites.
Since these .sites are effectively controlling the sources to comply with
state regulations, any new sources at these sites would also be effectively
controlled.
1-4
-------
2. INTRODUCTION
The objective of this study is to determine the need for new source
performance standards (NSPS) for the animal feed defluorination industry.
The Clean Air Act (CAA), as amended in 1977, provides authority for
the U.S. Environmental Protection Agency (EPA) to control discharges
of airborne pollutants. The CAA contains several regulatory and enforce-
ment options for control of airborne emissions from stationary sources.
Section 111 of the CAA calls for issuance of standards of performance
for new, modified or reconstructed sources which may contribute signi-
ficantly to air pollution. The standards must be based on the best
demonstrated control technology. Economic, energy and non-air environmental
impacts of best control systems must be considered in the development of
standards.
To determine which processes and pollutants, if any, should be
regulated by national NSPS, the following information has been provided
in this survey:
•description of facilities included in source category,
•number and location of facilities,
•past and current volumes of production and sales, products and
product uses,
•past and future growth trends in the industry,
•description of the processing operations and indentification
of emission sources,
•characterization of emissions from source category,
•identification and description of control techniques currently
used in the industry, including the degree of control achieved,
•identification of "best systems" of control,
•effectiveness of state regulations for new and modified facilities
in the source category, and
•preferred methods of sampling and analyzing the pollutants.
Information in the above categories was gathered from several sources.
First, a telephone survey of the industry trade association, National
2-1
-------
Feed Ingredients Association, and other non-industry organizations was
conducted to collect background information on the animal feed defluorina-
tion industry. A literature search was completed to obtain published
information. Producers of defluorinated feed product, regional offices
of EPA,'and state air pollution control agencies were then contacted via
phone or letter to obtain additional information. Two feed defluorination
plants and three Florida air pollution control agencies were visited.
This study focuses on all emission sources of air pollutants which
are not regulated by other NSPS or which are not associated with phosphate
mining operations. Standards for this industry should not be confused
with those for the phosphate fertilizer and phosphate rock processing
industries. A discussion on emissions specific to the phosphate rock
defluorination industry is given in Chapter 5.
2-2
-------
3. CONCLUSIONS AND RECOMMENDATIONS
3.1 CONCLUSIONS
The following conclusions have been made based on the results of
this study:
•Growth of Industry—the animal feed defluorination industry will
continue to grow at its historic rate of 3 to 6 percent per year.
This growth translates into possibly one more unit, either a rotary
kiln or a fluidized-bed reactor, being installed in the next 3 to
5 years. This expansion will most likely occur at the present
defluorination plant sites, all of which are located in Florida.
•Available Control Technology—control technology is presently
available to control all operations in the animal feed defluorination
industry to levels required by state regulations. The control
technology used in this industry is considered best demonstrated
technology.
•Identification of Significant Emission Sources—the only two signifi-
cant emissions sources are the effluents from feed preparation and
thermal defluorination. Emissions from feed preparation consist
primarily of particulates while thermal defluorination emits both
particulates and fluorides.
•Availability of Emission Data—relatively few emission measurements
are available for this industry. Most of the available data are for
the thermal defluorination operation. Very limited data are
available for the feed preparation operation. The available data
do not include inlet measurements, but include only outlet emission
levels. Therefore, determination of control efficiency for most
operations involves estimates of inlet amounts. The data do indicate
compliance with state regulations.
3-1
-------
•Applicable Methods for Sample and Analysis--EPA reference methods
are available and applicable for both pollutants of concern. EPA
Method 5 is used for particulate determination and EPA Methods ISA
and 13B determine total fluoride emissions.
3.2 RECOMMENDATIONS
The recommendation of this source category survey report is that new
source performance standards not be developed for the animal feed
defluorination industry. This recommendation is based on two important
conclusions reached during this study. The first of these conclusions
is that all of the operating plants are in the state of Florida. The
Florida Department of Environmental Regulations (DER) has promulgated
regulations controlling the emissions of particulates and fluorides from
the processes in this industry. These regulations specify emission
limitations which require that the best demonstrated control technology
be installed to control particulates and fluoride emissions from animal
feed defluorination. Available emission data indicate that the control
technology presently used can achieve these levels of emissions.
The second conclusion is that there is a small potential for growth
in this industry and that any expansion would take place at the present
locations. Therefore, any expansion would require installation of
best demonstrated control technology to meet the Florida State regulations.
3-2
-------
4. DESCRIPTION OF INDUSTRY
1
4.1 SOURCE CATEGORY
At present, the animal feed defluorination industry consists of
just three functional plants. The Borden Chemical Company operates a
facility at Plant City, Florida; International Minerals and Chemical
Corporation operates one at Mulberry, Florida; and Occidental Chemical
Company operates one at White Springs, Florida. Other phosphate rock
defluorination plants have been in operation at one time or another in
the past but have been shut down. Rocky Mountain Phosphate ran a small
operation at Garrison, Montana, but was forced to close it down because
of financial difficulties. According to one source with the State of
Montana Air Pollution Control Agency, the plant was not well run and
would require significant capital investment to make it operable again
There appear to be no plans to reopen the facility. Similarly, 01 in
Chemical Company's 68.T Gg/year (7F,000 ton per year) defluorination
plant in Pasadena, Texas was closed in January of 1978. The plant's
location, requiring phosphate rock to be shipped in over considerable
2
distances, caused it to be unprofitable. Table 4-1 summarizes the
animal feed defluorination industry.
A number of persons have been helpful in providing information
regarding the animal feed defluorination industry. Included are persons
involved in the industry at both the corporate and plant level, as
well as those engaged in a regulatory capacity at the state level.
Table 4-2 lists these people, their affiliations, and their addresses
and phone numbers.
4.2 INDUSTRY PRODUCTION
Production figures for the animal feed defluorination industry
proved to be difficult to ascertain. As of 1971, production capacity
was 485.8 gigagrams per year (535,000 tons per year), according to the
3
literature. That figure would be accounted for (approximately) by the
combined capacities of the Borden, Occidental, 01 in, and Rocky Mountain
4-1
-------
Uu OO
UJ UJ
01-
UJ OO
UJ
3J£
: uj
>-»-«
ce:
CO
CO
UJ
UJ
o£ >-
UJ O
CJ «
o t—
o &.
CD
D:
o_
Q.
O
CO
r~-
en
to
oo
LO
•=c •—
CO
OJ
r~--
OJ
CO
en
CO r—
O CO
CTl CO
its s^-
-o o
•r- «O r—
S- T3 U-
O •{-
i— S- •>
u_ o (/>
o
O)
CL
oo
O)
(O
X
O)
-------
UJ
jE
z:
i— i
UJ >-
CO OH
I-t h-
I— CO
UJ CD
r* ZT
X •-»
UJ
zr-
71"" CD
l—i-*
co »-«
z: oi
CD C3
CO =3
UJ U—
O_ UJ.
o
u.
CD a
UJ
r- UJ
CO U_
H- 1
—I —1
•=c
• 1— 1
CM z:
4^t
UJ
*
CQ
1—
UJ
CQ
§
^"
UJ
z:
o
0.
UJ
—1
1 i i
CO
CO
UJ
ac
o.
a.
«=c
z:
o
>"~ 1
r-
^^
1— 1
*
1 — 1
u_
U.
1 I 1 1 I 1 1 1 1 II 1
CM to cr> to in co CM to **• men CM
co *s3" oo in ^•^ cr^ ^-* in *d~ Is*-* co CM
i i i i i i i i i ii i
in •*'5i-CMoooococMto oo"* in
«* co cri co co co< co oo «* co cr» m
• • as
t- as as as t — •
*r- *p— r— " • C • U- OS OS
i— c: U- • as * as as r—* S
fCJ.l— r— i — • 1 — -M l — «>U_O
O O5«i — U- -» i — «*-+o CJ)«
A «r— r— •» * U ^" •* E CO *
v< ^» !«_ « ^>k |0 «^ ^^ fr— QJ ^
£- T-C4-5I — •»£!« -t-> S_ CO CU
••— t- u_ >•, .i— c i- a. to c:
D_ -^ C O) C_3 S- OJ O O €/> 1C «r-
r— O i — * S- r^ «/} JC O
o o GO ai -t-> « ai cu T- +j co i —
C S- O (=: — CO33CO i — jCIO'CO
r"*
as *
JE o o
O .p- O
I- r-!> P— P— £
CO O O T- O J= -O «C
in -i- * «p- * as H- <_3 * CO
co £C£ E o; +-> £ cgcotn
O« CO UJ CO UJ C
CO ^* ^** c~ f^ 23 CO -^" as ^^ £
-O+j<_) CJ O4->O4-> P— CO
&-=3 (8 (8 SL OS C«8i8-p-
o ••-> • c 13 c-o JE c a> -o -c *o ;
M- M- >' CL 0) "O i- O CD
+JC 01 — s Oi — s:^ o^r o 01 — «c
co i-< CQ u_ t-t CQ u_ »— i i— i cc a. CQ o u. z: i— i
C >> C - CO T3
as as T- CO c
•p- 3-0 c «e
+J O-US-fcOlsl p-S-cfl
•p-coi — i — a> i- s- i — -p-co-4-J'p-
CO S-p-jQ<8,S--p-S-CO S Z: CO 5>
Z5 J>3 O J3 CD t- CD <8 ^ CO O
r-cjcj-p- «e n: _i E: s- c
CQ OCOCD>> >> 0)1-
S=E O S-^>fS-- ^i O) 4J >
O >>OCO&-OCO(8ia
1— — JI"OOCQCQ1"DI"33C *~3CDSS
c
o
•4-J
as
p— •
i cr>
CO
or
1 *
E
CO
£
o
s-
•p-
E
UJ
o
E
£
ifr.^
s-
<8
CL
CO
CD
C£.
UJ
o
*
4-3
-------
plants (the IMC plant did not start operating until 1978). Actual
production figures for that year are unavailable.
For the years 1976 and 1977, the amounts of defluorinated animal
feed "sold or used" in the U.S. were 235.2 and 270.6 gigagrams (259,000
and 298,000 tons), respectively.4'5 It must be assumed that these figures
would closely approximate actual production figures. Production capacity
for those years would have been about 431.3 Gg/yr (475,000 tons per
year) (from Borden, Occidental, and Olin).
Production capacity for the past year (1979) was 454.0 gigagrams
(500,000 tons) which is less than the 1971 production capacity. The
three operating plants in 1979 were Borden, Occidental, and IMC.
Again, actual production figures were hard to determine. According to
sources at the three plants, Occidental was producing at or near capacity.
IMC, in only its first full year of operation, had achieved a production
rate of nearly 63.6 Gg/yr (70,000 tons per year) . Borden produced an
unknown amount of animal feed, which was reportedly less than capacity.
All of the above figures are estimates.
As to the future growth in the industry, several sources have
910
placed the annual growth rate at between two and six percent. ' However,
it was also learned that the demand for defluorinated animal feed products
is cyclic in nature and changes from year to year. The major factor
effecting this demand is the price of other feed supplements, such as
fish meal, which can be substituted for defluorinated phosphate rock.
Expansion plans are also somewhat indefinite. Borden plans to
build a more thermally efficient rotary kiln at the present location to
I O
replace an older unit which would then be used to handle peak loads.
The date for this expansion was not available. IMC is currently conducting
a market survey to determine growth and expansion potential. Sources at
Occidental indicated that expansion of the defluorination facilities
within the next few years did not appear too likely.
The likelihood of thermal defluorination facilities locating outside
of Florida seems small. Due to economics, operations generally have to
4-4
-------
be mine mouth and the phosphate rock has to be high quality. Texasgulf
operates a phosphate rock plant at Aurora, North Carolina, which has
supplies of animal feed grade phosphate rock. However, this rock evidently
has some different characteristics which preclude the use of current
technology for defluorination. Therefore, Texasgulf presently has no
plans to enter the defluorination industry. Should a technological
breakthrough occur which would make processing of this particular rock
technically feasible, Texasgulf indicated they would seriously investigate
the possibilities of entering the industry.
4.3 PROCESS DESCRIPTION
The animal feed defluorination process considered in this study
consists of three operations: 1) feed preparation, 2) thermal defluorina-
tion, and 3) product storage and shipping. Figure 4-1 presents a
general flow sheet for this process.
The primary raw materials for this process are phosphate rock,
phosphoric acid, and caustic soda. These materials are combined in a
mixer, typically a pug mill or a rotary drum mixer, to produce a homo-
genous mixture. This mixture is then dried to the proper moisture
content as required for the thermal defluorination operations and is
stored in hoppers. Emissions from the feed preparation operation include
particulates from the mixing operation and possibly gaseous fluorides
and particulates from the dryer. The moist nature of the feed materials
minimizes the amount of particulates produced during the mixing operation.
The dryer does not operate at high enough temperatures to release significant
amounts of fluoride as compared to the defluorination operation.
Thermal defluorination of phosphate rock is accomplished by heating
the feed mixture up to approximately 1370°C (2500°F). The fluorine is
driven off primarily as hydrogen fluoride (HF). This operation is
carried out in two types of equipment, rotary kilns and fluidized-bed
reactors. In the rotary kiln operation, the feed mixture is introduced
at one end of a long slightly inclined cylindrical kiln. The kiln
rotates and is fired from the end opposite the feed, resulting in counter
current flow. Product material, defluorinated rock» is discharged at
the firing end, while the process gases exit at the feed end.
4-5
-------
n:
o
O Di
C3
O -
CE: o:
UJ I—
•Z. C/5
uj rs
i.
o-a
j=-i-
O.U
in «=C
O
O)
4->
CO
O. O
(/•> O
Q.
CD
-------
The fluidized-bed reactor operation differs in that the reactor
remains stationary while a stream of air and fuel fluidize the feed
mixture or bed within the reactor. The process of fluidizing converts a
bed of solid particles into an expanded, suspended mass that resembles a
boiling liquid. Space must be provided for vertical expansion of the
solids and for disengaging entrained material. The usual shape is a
vertical cylinder. The product material overflows at the top of the bed
and the process gases exit at the top of the reactor.
Both the rotary kiln and the fluidized-bed reactor are usually
fired with natural gas, with LPG or fuel oil used during curtailment.
There is very limited potential for use of coal in these, processes due
to the purity requirements of the feed supplements.
The emissions produced by these operations are also very similar,
consisting of particulates and gaseous fluorides. The parti oilates are
composed primarily of product and are recovered for economic reasons as
well as emission regulations. The fluoride emissions, primarily in the
form of HF, are the major pollutant which must be controlled. The
defluorinated rock product is cooled and sent to product shipping and
storage. The product may be shipped in bags or loose in railcars or by
trucks. Loading of rail cars and trucks with loose product produces
fugitive particulate emissions consisting of product material.
The amounts of emissions produced by these processes and the types
of control equipment used to control these emissions are discussed in
Chapters 5 and 6, respectively.
4-7
-------
4.4 REFERENCES
1. Telecon. Meling, J.L. Radian Corp. with Harry Keltz, Montana
Air Pollution Control Agency. December 4, 1979.
2. Telecon. Meling, J.L. Radian Corp. with Tom Blue, Stanford
Research Institute. November 30, 1979.
3. "Fluidized Bed Process for Defluorination of Phosphate Rock",
in Minerals Processing, March 1972. p. 16.
4. Pit and Quarry, January 1978. p. 102.
5. Pit and Quarry, January 1979. p. 99.
6. Meling, J.L. Trip Report: Occidental Chemical Company, White
Springs, Florida. Radian Corp. Durham, N.C. December 26, 1979.
7. Telecon. Meling, J.L. Radian Corp. with Jack Harris, International
Minerals and Chemical Corp. January 15, 1980.
8. Hoover, J.R. Trip Report: Borden Chemical Company, Plant City,
Florida. Radian Corp. Durham, N.C. January 4, 1980.
9. Reference 2.
10. Reference 8.
11. Reference 8.
12. Reference 8.
13. Reference 8.
14. International Minerals and Chemical Corp., Response to Section 114
request for information, January 14, 1980.
15. Reference 6.
16. Telecon. Meling, J.L. Radian Corp. with Frank Robinson, Texasgulf.
January 21, 1980.
17. Reference 8.
4-8
-------
5. AIR EMISSIONS DEVELOPED IN THE SOURCE CATEGORY
This chapter deals with the types and quantities of emissions
resulting from the source category. Included are particulates and
fluorides emitted during feed preparation, thermal defluorination, and
product storage and shipping. Sources considered in this source category
are different than those covered by proposed NSPS regulations for the
phosphate rock industry, even though the source types are similar.
Common emissions source types, such as drying, grinding, and screening,
that are not directly associated with the defluorination process are not
considered. Figure 5-1 illustrates the general flow scheme for the
Florida phosphate rock industry. The defluorination of phosphate '
rock involves mixing the rock with phosphoric acid and caustic soda, a
process unique to the rest of the phosphate rock industry. Only emission
sources related to this and other defluorination operations are covered
here, as is shown in Figure 4-T.
5.1 PLANT AND PROCESS EMISSIONS
This section is concerned with the development of emission factors
for the pollutants and sources identified above. The term emission
factor, as used in this report, refers to a number which quantifies the
emission per unit of product passing through a process. Both emission
and process units are usually given in mass units. Thus, emission
factor units typically take the form of kg/Mg (Ib/ton).
Emission factors were developed exclusively using the limited
amount of industry stack test data. Since the stacks were tested while
emission controls were engaged» no actual data for uncontrolled emissions
are available. Also, EPA document AP-42 contained no information on
emission factors for animal feed defluorination.
5.1.1 Feed Preparation Emissions
Feed preparation involves the mixing of raw materials, as well as
the drying of the resulting mixture. Only particulates are emitted
during the mixing of raw materials (phosphate rock, phosphoric acid, and
5-1
-------
ID
-O
O)
(L)
(O
S-
O
oo
o
UJ
a.
o
CJ)
§
D-
to
O
Q_
g
I—H
a:
LO
t- Q)
-------
caustic soda), while participates and a small amount of fluorides are
emitted during the drying process. No emissions data exist for uncontrolled
emissions from the feed preparation area, but a test was performed on
the stack from controlled mixing and drying processes. The measurement
was taken after a wet scrubber and indicated "particulate emissions of
0.11 kg/Mg P205 (0.22 Ib/ton P20r) and fluoride emissions of 0.0065 kg
F/Mg P20g (0.013 Ib F/ton P20g). No data exist to make a similar
estimation for an uncontrolled plant.
5.1.2 Thermal Defluorination Emissions
Both particulates and fluorides are emitted during the thermal
defluorination process (rotary kiln or fluidized-bed reactor). Available
data for well-controlled plants (using both rotary kilns and fluidized-bed
reactors) suggest particulate emission rates of 0.65 kg/Mg P20 * (1.29
Ib/ton P205). Fluoride emission rates average 0.13 kg F/Mg P205** (0.26
Ib F/ton P205), as opposed to the Florida DER emission limitation of 0.185 kg
F/Mg P205 (0.37 Ib F/ton P205). The limited amount of data indicates no
significant difference between the total weight of emissions from rotary
kilns and fluidized-bed reactors.
No test data are available for fluorides from an uncontrolled
facility. However, the fluoride emission rate can be estimated through a
material balance. The defluorination process reduces the weight percent
fluorine in the phosphate rock from 3.5 percent to 0.2 percent2'3, resulting
in uncontrolled emissions of 95 kg F/Mg P205 (190 Ib F/ton PgOg.). (This
is in close agreement with an estimate of 105 kg F/Mg P^Q,. (210 Ib F/ton
A
P20g) found in the literature. )
* Average value. Actual test data ranged from 0.26 to 1.09 kg/Mg PJD,-
(0.53 to 2.18 Ib/ton P205).5'6'7
**Average value. Actual test data ranged from 0.02 to 0.25 kg F/Mg P-O,-
(0.04 to 0.49 Ib F/ton P205) with only one value (0.49) over the 0.37
Ib F/ton P205 limitation.8'9'10
5-3
-------
5.1.3 Product Storage and Shipping Emissions
Product storage and shipping includes the crushing, screening,
conveying, and loading out of defluorinated animal feed. Emissions from
these sources are limited to fugitive particulates. No estimate of
emission rates can be made.
5.2 TOTAL NATIONAL EMISSIONS FROM SOURCE CATEGORY
An estimate of total national emissions from the three operating
feed defluorination plants can be made using production information
presented in Chapter 4 and emission factor information presented above.
The estimate is based on plant production capacities, which would be the
worst case. Total national emissions of both particulates and fluorides
can be calculated knowing that emissions (Mg/yr or ton/yr) are equal to
the total national production rate (Mg product/yr or ton product/yr)
multiplied by emission factors (kg/Mg product or Ib/ton product)* and the
proper conversion factors. Total national emissions are presented in
Table 5-1.
*Emission factors are converted to kg/Mg (Ib/ton) product knowing that
defluorinated animal feed is approximately 40 percent P°-
5-4
-------
TABLE 5-1. ESTIMATE OF TOTAL NATIONAL EMISSIONS
OF PARTICULATES AND FLUORIDES DEVELOPED IN THE
ANIMAL FEED DEFLUORINATION INDUSTRY
Particulates*
Fluorides
138.0 Mg/yr
(142.0 ton/yr)
24.8 Mg F/yr
(27.3 ton F/yr)
Not including potential fugitive emissions from crushing, screening,
conveying, and product storage and shipping.
5-5
-------
5.3 REFERENCES
1. Borden Chemical Company, Plant City, Florida. Stack Compliance
Test Results. October 2, 1978 through October 26, 1979.
2. Hoover, J.R. Trip Report: Borden Chemical Company, Plant City,
Florida. Radian Corp. Durham, N.C. January 4, 1980.
3. Resources Research, Inc. Engineering and Cost Effectiveness
Study of Fluoride Emissions Control (Final Report). January,
1972. p. 3-179.
4. Reference 3. p. 183.
5. Reference 1.
6. International Minerals and Chemical Corporation, Mulberry,
Florida. Emission Test Results. September, 1979.
7. Occidental Chemical Company, White Springs, Florida. Point
Source Test Results. February 16, 1977 through August 10, 1978.
8. Reference 1.
9. Reference 5.
10. Reference 6.
11. Reference 3. p. 181.
5-6
-------
6. EMISSION CONTROL SYSTEMS
The various types of air pollution-control equipment in use in the
animal feed defluorination industry to control airborne emissions are
briefly reviewed in this section.
The major emissions of concern from animal feed defluorination are
particulates and fluorides. The other pollutants resulting from drying
and thermal defluorination are primarily combustion products. The use
of clean burning fuels, such as natural gas, and use of good combustion
practices result in negligible levels of carbon monoxide and sulfur
oxides.
The various manufacturing operations used in this industry are
discussed in Chapter 4. The important control systems presently in use
are discussed in the following sections of this chapter.
6.1 CONTROL APPROACHES
The major source of emissions from animal feed defluorination
plants is the gas stream from the thermal defluorination operation.
This gas stream must be treated to eliminate particulates and gaseous
fluoride prior to discharge into the atmosphere.
The most common method of control for this stream consists of a
123
cyclonic entrainment separator followed by a wet scrubber system.1' » The
cyclone achieves bulk removal of the particulates. The wet scrubber
system is used to increase particulate removal and to control gaseous
fluoride emissions. This wet scrubber system is usually a spray cross-
flow packed scrubber, a design which incorporates the features of both
spray towers and packed towers. The fluoride removed by the wet
scrubbing system can be disposed of as a waste or recovered as a
byproduct. The most common method of removal as a waste involves
precipitation of the fluorine as calcium fluoride by the addition of
lime to the scrubbing liquor.4 This reaction occurs in settling ponds
and the pond overflow is recycled^to the scrubber.
6-1
-------
The other method of fluoride disposal involves reaction of the fluoride
with chemicals to produce a usable, saleable byproduct. One plant
recovers at least 60 percent of the fluoride as byproducts, such as
potassium fluoroborate.
In addition to the wet scrubber system, one plant has installed
ionizing wet scrubbers (IWS) to control fine particulates from their
fluidized-bed reactors. Each IWS system is a double stage unit. The
IWS system consists of a high voltage ionizing section and a wet scrubbing
section. As the gas passes through the IWS, submicron particulate
matter as well as soluble and reactive gases are removed. This system
is designed to remove solid and/or liquid particulates down to 0.05
micron size.
Emissions from the feed preparation operation, primarily particulates
with small amounts of gaseous fluorides from the drying operation, are
Q
controlled by fabric filters and/or wet scrubbers. The equipment used
for feed preparation, pug mills, rotary mills, and dryers of various
designs, are equipped with conventional hooding and/or exhaust ventilation
to capture emissions from those operations. These hoods are vented to
control equipment for particulate and, to a much lesser extent, fluoride
control. The most common control technique for particulates and fluorides
from the mixing and drying operations is wet scrubbing. One plant uses
a baghouse to control particulates from the mixing and drying operations.10
Emissions from the product storage and shipping operations consist
of particulates from the loading of trucks and rail cars with loose
product. These fugitive emissions are controlled at one plant by a
system of hooding and local exhaust ventilation ducted to baghouses.
Enclosure and ventilation of this operation is being investigated by
another company. Fugitive dust problems exist at other transfer points
but the plants were in the process of designing controls for these
areas.
6.2 ALTERNATE CONTROL METHODS
The control systems now in use are capable of meeting the control
11
6-2
-------
requirements of each of the operations described in the previous section.
Other methods of control may be introduced as conditions warrant, but there
is no apparent need for developing any new control technology at the
present time.
6.3 "BEST SYSTEMS" OF EMISSION REDUCTION
This section discusses the most effective systems of emission
control for the two operations in the animal feed defluorination industry
which are the most significant sources of emissions, feed preparation
and thermal defluorination. Table 6-1 summarizes these controls and shows
the location of the plant and the name of the plant contact.
The feed preparation operation includes mixing of raw materials and
drying, if necessary to adjust moisture content. Two systems are in
use to control emissions, essentially all particulates, from this operation:
1 p
wet scrubbers and fabric filters. Based on emissions testing data which
are available, both of these systems can achieve about the same levels
of particulate control. Both of these systems are capable of meeting
the applicable state regulations.
Two types of processes are used to thermally defluorinate phosphate
rock for use as an animal feed supplement. There are rotary kilns and
fluidized-bed reactors.
The best control system for the rotary kiln process consists of a
cyclonic entrainment separator for bulk particulate removal followed by
a spray crossflow packed scrubber for fluoride removal and final particulate
control. The overall efficiency of this system is 99.8 percent or
greater for fluorides.13 Actual control efficiency is difficult to determine
for particulates but should be approximately the same as for fluorides.
The best control system for the fluidized-bed reactor process is
the same system as for the rotary kiln with the possible addition of a
fine particulate control device. The fluidized-bed reactor process may
produce fine particulate matter which is not always efficiently controlled
by the first system and may require fine particulate control. The best
system for control of this fine particulate is an ionizing wet scrubber
(IWS). The combination of this unit with the previous controls achieves
an overall control efficiency for particulates and fluorides of 99.9 percent.14
6-3
-------
o
1—
CD
CJ
U.
O
CO
1—
CO
CO
h-
co
UJ
CO
1
CO
UJ
r 1
CO
«=c
J—
'
1—
CJ
^^
h~*
^^
CD
CJ
!S^
CD
P
8
^j
O
fV^
i—
CD
- CJ
co
•z.
O
i— <
CO
CO
1— <
%
UJ
o
^~~i
CD
CO
CD
r™* ^"^
T- CVI
CU CO
o r**»
^> CTi
CO
CU 1
o
CJ
r— •
re re
•"IT
cu »
.c: to
o en
c
i— •! —
re s-
4-3 CX
ceo
cu
•o cu
•r- 4-3
O ••-
o .c:
03
cu
to
3
o
^
en
re
CO
to
cu
4-3
re
3
O
•r—
4-3
*
C
O
•1—
4-3
£
re
CL
cu
CL.
-a
cu
cu
Li-
re
o vo
,--. r»
l__ r—
re i
CD CVI
ID
CU Is-.
O 1
3 CO
CO CO
re
•o
re o
•"LZ
^ i5?
CJ •!-
CJ
E:
CU 4-3
•o c
s- re
O r—
CQ O-
s_
cu
JO
3
S-
o
co
4.)
CU
3
oJS
to
cu
re to
r— CU
3 -O
O 'I—
•i- S-
"re!
re
o Jo
r— r~~
^_ ^-.
re i
CD CVJ
ID
cu r-.
U 1
3 CO
CO CO
*
o re
o -a
"re o
•"LZ
!*
CJ -r-
CJ
C
CU 4-3
-0 E
s- re
O i—
CQ D-
^
O
t^-
to s-
to CU
O -Q
S- JD
CJ 3
S-
T3 0
cu co
V*"
0 4-3
re cu
Q- 3
„
to
cu
re to
r— CU
3 -a
O T-
•r- i-
c2LZ
C
O
•r~
•4-^
c re
i — C
•r~ *r--
•±S c
O
>> 3
S-r—
re .
O -i-
i— 3
C^-.
tO S- 4-3
CO CU CU
oja 3
t- .Q
o 3 en s-
s- c cu
-a o ••- JQ
CU to N J3"
^i -1-3
O 4-3 C 5-
re cu o o
a. 3 ••- to
«
to
0)
re co
r— CU
3-0
O T-
•r- S-
tf§
£LZ
-O
4-3 c:
o o
re -i-
CU 4-3
o; re
c
•a «r-
cu s-
CQ 0
3
"O i —
3 CU
r— Q
U.
re
3r-
o vo
ffH« ^^
P.— r«
re i
CD OJ
ID
cu r--
O 1
3 CO
CO CO
.
o re
0 T3
r— S»-
re o
O i—
•r- U_
1^
CJ -i-
CJ
c
CU 4-3
-o c
s- re
O r—
CQ Q-
«\
en
c:
•r- C CU
•o o to
O -r- 3
04-30
r— en
i— T- re
re 4-> -Q
u c
0 CU 0
— 1 > 4-3
to
cu
^-3
re
^~
3
O
•i—
4-3
C?
•o
c:
re
en
cu c
cn»r—
re o.
S- Q.
0 ••-
4-3 -E
CO CO
6-4
-------
The product storage and shipping operation creates fugitive parti-
culates, primarily during the loading of railcars and trucks with product.
The best system for control of this operation utilizes enclosure and
local exhaust ventilation ducted to baghouses during the loading operation,
No estimate of the control efficiency could be made due to the lack of
data. The limiting factor in this control scheme is the efficiency of
the collection system, i.e. the hoods and enclosure.
6-5
-------
6.4 REFERENCES
1. Hoover, J.R. Trip Report: Borden Chemical Company, Plant City,
Florida. Radian Corp. Durham, N.C. January 4, 1980.
2. Mel ing, J.L. Trip Report: Occidental Chemical Company, White
Springs, Florida. Radian Corp. Durham, N.C. December 26, 1979.
3. International Minerals and Chemical Corp., Response to Section 114
request for information, January 15, 1980.
4. Reference 2.
5. Reference 1.
6. Reference 1.
7. Ensor, D.S. Ceilcote Ionizing Wet Scrubber Evaluation. Meteorology
Research Inc. Altadena, California. November, 1979.
8. References 1,2,3.
9. References 1,3.
10. Reference 2.
11. Reference 1.
12. References 1,2,3.
13. References 2,3.
14. Reference 1.
6-6
-------
7. EMISSION DATA
7.1 AVAILABILITY OF DATA -
Relatively few emission measurements are available for animal feed
defluorination processes. There are three main sources of emission
data: 1) National Emission Data System (NEDS), 2) test data on file with
state or local agencies, and 3) information and test data obtained
directly from the animal feed defluorination industries.
Emissions and emission rates by SIC numbers for specific plants and
specific emission points can be obtained through the NEDS. Other useful
information contained in NEDS reports include control equipment, collection
efficiencies and fuel type. NEDS is not always up-to-date and the
current test results are not always available. NEDS data are basically
non-existant for the animal feed defluorination industry.
State or local agencies have information on most current test data
and permit applications. Emission test data may also be obtained directly
from the companies involved.
Available emission source test data for the animal feed defluorination
industry has been summarized in Table 7-1. No data are available from
uncontrolled facilities. The State of Florida only requires samples of
gases emitted to the atmosphere. All three of the operating plants use
similar methods for control of particulates and fluorides from the
processes.
7.2 SAMPLE COLLECTION AND ANALYSIS
The two major air pollutant emissions from animal feed defluorination
are particulates and total fluorides. There are EPA reference methods
which are applicable for these pollutants:
Method 1: Sample and Velocity Traverses for Stationary Sources.
Method 2:
Method 3:
Determination of Stack Gas Velocity and Volumetric
Flowrate.
Gas Analysis for CO^, Og, Excess Air and Dry
Molecular Weight.
7-1
-------
TABLE 7-1. EMISSION SOURCE TEST DATA1'2'3
Test
Locations
Number
of Tests
Test
Method
Feed preparation stack
Stack of 2 fluid bed
reactors and 1 kiln
Stack of 3 kilns
Stack of 2 kilns
Stack of 2 kilns and
1 dryer
Stack of-2 fluid bed
reactors
EPA Method 13B
(Wi Hard-Winter
distillation)
EPA Method 13B
(Willard-Winter
distillation)
EPA Method 13B
(Wi Hard-Winter
distillation)
EPA Method 13B
(Willard-Winter
distillation)
EPA Method 13B
(Wi Hard-Winter
distillation)
EPA Method 13B
(Wi Hard-Winter
distillation)
Note: AH of the above emission tests are measured after some type of
control device. There are no test data on uncontrolled emissions.
7-2
-------
Method 4: Determination of Moisture Content in Stack Gases.
Method 5: Determination of Participate Emissions from
Stationary Sources.
Method 13a: Determination of Total Fluoride Emissions from
Stationary Sources SPADNS .Zirconium Lake Method.
Method 13b: Determination of Total Fluoride Emissions from
Stationary Sources Specific Ion Electrode Method.
Particulate size is of importance because small particles, less
than 5 micrometers in size, are carried into the human lung. There is
no standard EPA method for determining particle size. However, the
Cascade impactor can be used for sizes between 0.4 and 10 micrometers
and recent developments such as the Coulter Counter and Thermosystems
aerosol size analyzer have been used for particles between 0.1 and
1 micrometer.
7-3
-------
7.3 REFERENCES
1. Borden Chemical Company, Plant City, Florida. Stack Compliance
Test Results. October 2, 1978 through October 26, 1979.
2. International Minerals and Chemical Corporation, Mulberry, Florida.
Emission Test Results. September, 1979.
3. Occidental Chemical Company, White Springs, Florida. Point Source
Test Results. February 16, 1977 through August 10, 1978.
7-4
-------
8. STATE AND LOCAL EMISSION REGULATIONS
All three of the operating animal feed defluorination plants are
located in Florida. The State of Florida Department of Environmental
Regulation (DER) has promulgated regulations which require control of
particulate and fluoride emissions from the thermal defluorination of
phosphate rock for use as an animal feed supplement.
The emission limits specified by these regulations are the same
for new facilities and existing facilities. The regulations limit
fluoride emissions to 0.37 pounds of fluoride per ton of phosphate
produced expressed as tons of PO^B* Based on the characteristics of
the fluoride rock processed in Florida, typically 3.5 percent fluoride
and approximately 35 percent P2°5» these regulations require that the
effluent gas streams be controlled by systems having control efficiencies
of 99.8 percent or greater for fluorides.
Particulate emission limits are on the basis of a process weight
regulation. For a process weight rate up to 30 tons per hour, the
o fi?
allowable emissions are expressed by E = 3.59 P where E equals the
allowable emission rate in pounds per hour and P equals the process
weight in tons per hour. For weight rates greater than 30 tons per hour,
the expression is E = 17.31 p°*16. Florida also has a visible emissions
regulation limiting stack densities to less than 20 percent opacity.
However, this visibility regulation does not apply to emissions regulated
by the process weight limitations. The regulations also require that
reasonable precautions be taken to prevent fugitive particulate emissions
2
from any source.
Using the process weight limit equation for particulates and the
allowable of 0.37 pounds per ton P20g fluoride limit, typical emission
rates for the fluidized-bed reactor process and the rotary kiln process
are as follows.:
8-1
-------
Process
FTuidized-Bed
Reactor
Rotary Kiln
Ty pi cal
Process Rate (ton/hr)
8.0
3.5
Participate Fluoride*
Emissions (Ib/hr) Emissions (Ib/hr)
13.0
7.8
1.04
0.45
*based on feed raw material
content of 35 percent
The State of Florida has a very active enforcement branch. Plants
are required to submit compliance stack tests twice a year, with state
people present during the tests. The state has a fairly extensive file
on each of the defluori nation plants, with detailed construction and
operating permits and all compliance testing data.
The DER plans to review the regulations for thermal defluori nation
this year to determine if changes in the particulate regulations should
3
be made. They feel no need to change the fluoride emission regulations.
Local emission limits, where they exist, are adopted from the state
regulations. These agencies, usually on a county level, act as a
representative of the State DER.
The Florida State regulations came about in response to complaints
about damage to vegetation, especially citrus trees, and to animals from
ingestion of fluorine-contaminated vegetation. Over the last 25 to 30 years,
the regulations have required improvement of the control systems to
achieve almost complete removal of fluorides. According to state sources,
there have. been very few complaints in recent history, and none in the
last two to four years.
8-2
-------
8.1 REFERENCES
1. Rules of the Florida Department of Environmental Regulation,
Chapter 17-2s Air Pollution.
2. Reference 1.
3. Reference 1. •
8-3
-------
TECHNICAL REPORT DATA .
(Please read Instructions on the reverse before completing)
REPORT NO.
FPA-450/3-8n-m 5
TITLE AND SUBTITLE
Source Category Survey:
Animal Feed Defluorination Industry
6. PERFORMING ORGANIZATION CODE
. RECIPIENT'S ACCESSIOt
. REPORT DATE
May 1980
AUTHOR(S)
.1 R,
8. PERFORMING ORGA
d J. 1.
Mel inn
; AND Ad
PERFORMING ORGANIZATION NAME
Radian Corporation
P.O. Box 8837
Durham, NC 27707
DRESS
1O. PROGRAM ELEMENT
11. CONTRACT/GRANT NO.
68-02-3058
2. SPONSORING AGENCY NAME AND ADDRESS
DM for Air Quality Planning and Standards
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency
n~ .o._u TM.;-,«„•! ~ D^v.1, M P 97711
13. TYPE OF'REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
5. SUPPLEMENTARY NOTES
Park
04-
EPA 200/04
6. ABSTRACT
This study investigated the need for new source performance standards (NSPS)
for processes that defluorinate phosphate rock to produce an animal feed supplement.
This defluorination is accomplished by heating phosphate rock in a kiln or fluidized
bed to about 1370 C to drive off fluoride as HF.
It is recommended that NSPS not be developed-." , The only three domestic plants,
all in Florida, have a very small growth potential. Any expansion within the next
five years would probably be a single kiln or fluid bed furnace added to existing
facilities at one of these plants. The total combined fluoride emissions are only
27 tons/year 'for the three plants. The best demonstrated control technology
(a fabric filter or a wet scrubber) is already used at each plant, and the
Florida State regulations will continue to make this type of control a practical
necessity.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Phosphate rock defluorination
Fluorides
Animal feed supplement
Air pollution control
Stationary sources
13B
18. DISTRIBUTION STATEMENT
unlimited
19. SECURITY CLASS (ThisReport)
unclassified
42
20. SECURITY CLASS (Thispage')
unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |