EPA-450/3-74-010
May 1973
         NATIONAL EMISSIONS
                    INVENTORY
                    OF  SOURCES
                AND EMISSIONS
                               OF
                    MAGNESIUM
      U.S. ENVIRONMENTAL PROTECTION AGENCY
         Offire of Air and Waste Management
       Office of Air Quality Planning and Standardg
      Research Triangle Park, North Carolina 27711

-------
                                      EPA-450/3-74-010
NATIONAL EMISSIONS INVENTORY
                      OF
      SOURCES AND EMISSIONS
                      OF
                MAGNESIUM
                       by

                  GCA Corporation
                GCA Technology Division
              Bedford, Massachusetts 01730
                Contract No. 68-02-9601
            EPA Project Officer:  David Anderson
                    Prepared for

           ENVIRONMENTAL PROTECTION AGENCY
             Office of Air and Water Programs
          Office of Air Quality Planning and Standard*
            Research Triangle Park, N. C.  27711

                     May 1973

-------
This report is issued by the Environmental Protection Agency to report technical
data of interest to a limited number of readers. Copies are available free of
charge to Federal employees, current contractors and grantees, and nonprofit
organizations - as supplies permit - from the Air Pollution Technical Information
Center, Environmental Protection Agency, Research Triangle Park, North
Carolina 27711, or from the National Technical Information Service, 5285 Port
Royal Road, Springfield, Virginia  22151.
This report was furnished to the Environmental Protection Agency by GCA Corp-
oration, Bedford, Massachusetts, in fulfillment of Contract No.  68-02-9601.  The
contents of this report are reproduced herein as received from GCA Corporation.
The opinions, findings, and conclusions expressed are those of the author
and not necessarily those of the Environmental Protection Agency. Mention of
company or product names is not to be considered as an endorsement by the
Environmental Protection Agency.
                    Publication No. EPA-450/3-74-010
                                    11

-------
                             TABLE OF CONTENTS




SECTION                   	TITLE	PAGE


I


II








III



IV
V
VI


VII
ABSTRACT
ACKNOWLEDGEMENT
INTRODUCTION
A. PURPOSE AND SCOPE
B. CONCLUSIONS
OVERALL U.S. MATERIAL FLOW CHART FOR MAGNESIUM
A. MINING OF MAGNESITE, BRUCITE AND OLIVINE
B. MINING OF DOLOMITE
C. RECOVERY OF MAGNESIUM FROM SEAWAIER AND
WELL BRINES
D. IMPORTS AND EXPORTS
E. MAGNESIUM OXIDE
F. MAGNESIUM METAL
G. REFRACTORY MAGNESIA
H. MAGNESIUM CHEMICALS
SOURCES AND ESTIMATES OF MAGNESIUM - CONTAINING
EMISSIONS
A. DATA PRESENTATION AND ACCURACY
B. DEVELOPMENT OF EMISSIONS ESTIMATES - 1970
C. SUMMARY OF PRINCIPAL EMISSIONS
REGIONAL DISTRIBUTION OF PRINCIPAL SOURCES AND
EMISSIONS
NATURE OF EMISSIONS
UPDATING OF EMISSIONS ESTIMATES
A. VERIFICATION OF CURRENT ESTIMATES
B. PERIODIC REVIEW OF ESTIMATES
REFERENCES
V
vi
1
1
2
4
4
4
4
6
6
6
7
7
9
9
13
20
21
25
29
29
29
31
                                111

-------
                       LIST OP TABLES AND FIGURES
TABLE NO.
   4a

   4b
              SOURCES AND ESTIMATES OF MAGNES lUM-
              EMISSIONS
SUMMARY OF PRINCIPAL SOURCES AMD EMISSIONS OF
MAGNESIUM

REGIONAL DISTRIBUTION OF PRJNCIBAL SOURCES AUD
EMISSIONS

DESCRIPTION OF MAGNESIUM MI81RALS AH) PRODUCTS

PROPERTIES OF MAGNESIUM METAL
PAGE


 10


 20


 22

 26

 26
FIGURE NO,
              MAGNESIUM MATERIAL FLOW - 1970
                                    IV

-------
                                ABSTRACT

     A national inventory of the sources and emissions of the element
negnesium was conducted,  , The study included the preparation of an over-
all material flow chart depicting the quantities of magnesium moving
from sources of mining and importation through all processing and  repro-
cessing steps to ultimate use and final deposition.   All major sources
of magnesium-containing emissions were identified and their magnesium
emissions into the atmosphere estimated.  A regional breakdown of
these sources and their emissions was also provided.   The physical and
chemical nature of the magnesium-containing emissions was delineated to
the extent that information was available, and a methodology was recom-
mended for updating the results of the study every two years.

-------
                             ACKNOWLEDGMENT

     The continued cooperation and dedication of Mr.  Carl Spangler of
EPA, who served as Program Monitor until his death,, is deeply appreciated,
     GCA would like to extend thanks to Mr.  David Anderson and Mr. James
Southerland of EPA for their cooperation in the preparation of this
study.
     In addition, special thanks are also due to Mr.  E.  Chin, Conmodity
Specialist, Bureau of Mines; Ms. Marie Harris, Dept.  of Commerce; and
The National Lime Association, Washington, D.C., who provided significant
technical inputs to this program.
                                   VI

-------
I.   INTRODUCTION

     A.   PURPOSE AND SCOPE

          The Monitoring and Data Analysis Division,  Office of Air
Quality Planning and Standards of the U.S. Environmental Protection

Agency (EPA) has contracted with GCA Technology Division to conduct a

national inventory of the sources and emissions of the element magnesium.
The purpose of the study was to define as accurately  as possible,  based

on existing and available published and unpublished information,  the

levels, nature, and sources of magnesium-containing emissions  for defined

geographic regions throughout the United States.

          The scope of this program is outlined below.

               .  Develop an overall material flow chart
                 depicting the quantities of magnesium
                 moving from sources of mining and
                 importation, through all processing  and
                 reprocessing steps to ultimate use and
                 final disposition as far as the movements
                 can be traced.

               .  Identify all major potential magnesium-
                 containing emission sources and estimate
                 the total quantity of magnesium emitted
                 to the atmosphere from each source.
                 Emission factors and level, and types
                 of air pollution control, are also
                 provided for each of these sources to
                 the extent that available information
                 permits.

               .  Define those sources which contribute
                 at least 80% of the total emission of
                 magnesium.

               .  Provide a regiona1 breakdown of these
                 major sources and their emissions.

               .  Present the nature of the magnesium-
                 containing emissions for each of these
                 major sources,  including a delineation
                 of their physical and chemical form
                 and particle size distribution,  to the
                 extent that information is available.

-------
               . Provide recommendations as to a methodology
                 for updating the results of this study
                 every two years.
     B.   CONCLUSIONS
          1.   Mater ial
               Based on all available data,  905,000 tons of magnesium
was consumed in the U.S. in 1970, in various chemical and pure forms.
The sources of magnesium included sea water  and other brines, the mining
and importation of magnesite, and the mining of dolomite.   By far, the
largest portion of the magnesium was consumed in the form of fire brick
and other refractory materials, especially within the iron and steel
industry.  Very little magnesium was consumed as metal, and of this,
only about 10 percent was recycled as scrap.
          2.    Principal Emis s ion Sources
               Due to a large emission of flyash from the combustion of
coal, and a small yet significant concentration of magnesium in the
flyash, the largest source of magnesium emitted to the atmosphere was
found to be the combustion of coal.   Almost  57% of all U.S. emissions
of magnesium was from this source.
               Within the magnesium and magnesium compounds industries,
the largest source of emissions was estimated to be the production of
refractory materials, accounting for 10 percent of the total U.S.
emissions.  Taken together, mining, processing, and calcining of
magnesite and dolomite were estimated to be  responsible for about 17%
of all U.S. emissions of magnesium.
          3.    Regiona 1 Emis s ions
               The region of the U.S. in which most of the estimated
magnesium is  emitted is Region 5* (Ohio-Minnesota), with about 33 per-
cent of the U.S. total.  The region in which the total emission per
square mile is greatest is Region 3 (Pennsylvania -Virginia).  Although
the emissions from these areas are strongly  weighted by the estimated
emission from coal combustion, similar conclusions would be made in the
absence of coal combustion emissions.
*See page 21 for a list of regions.

-------
          4.   Mature of Emissions
               Magnesium is believed to be emitted predominantly in the
carbonate and oxide chemical forms, usually not alone but included in
most particles along with ash, slag, and other mineral material.  A wide
size range of particles containing magnesium is generated.   However,
after particulate control equipment, and/or gravitational settling, the
remaining particles containing magnesium are typically on the order of
1 micron diameter.  "With regard to the magnesium content of these
particles, they probably tend to be both stable and relatively inert.
          5.   Degree of Control
               The overall level of control of magnesium emission is
estimated to have been about 87 percent in 1970.   All of the large
sources of emissions were controlled at levels ranging from 82 to 99
percent.  Of these large sources, coal combustion, the largest,  was also
one of the most poorly controlled at an estimated 82 percent.   The
production of refractory materials was estimated  to utilize a control
level of only 80 percent.

-------
II.  OVERALL U.S. MATERIAL FLOW CHART FOR MAGNESJEIM
     Figure 1* presents a flow diagram depicting the total quantities
of magnesium products moving from sources of mining and importation
through the processing and reprocessing steps to ultimate use and final
deposition.  Each of these sources is discussed below.  Process descrip-
tions are given  in Section II.
     A.   MINING OF MAGNESITE., BRUCTTE, AND OLIVINE
          Magnesite and brucite are presently mined at only the basic
open-pit mine in Gabbs, Nevada.  Total ore production in 1970 was
631;000 tons.     Based on an approximate magnesium content of 35 percent,
the magnesium content of the produced ore is estimated at 221,000 tons.
Operations at or near the mine, after removal of a large overburden,
include blasting, crushing, screening, washing, and drying.
          divine, another magnesium ore, is presently mined in
Washington and North Carolina.  No data could be found to estimate its
production.  Its predominant use is as a molding sand in foundries and
it is not in demand specifically for its magnesium content.  No other
magnesium ores are presently mined, except dolomite.
     B.   MINING OF DOLOMITE
          Deposits of dolomite are quarried throughout the United States,
using processes similar to the above.  It is found extensively with
limestone and shares many of the same properties.  Total reported pro-
duction of dead burned dolomite for use as a magnesium compound totaled
               (2)
1,373,000 tons.     Based on a magnesium content of 13 percent, magnesium
from dolomite mining totaled 173,000 tons.
     C.   RECOVERY OF MAGNESIUM FROM SEAWATER AND WELL BRINES
          All magnesium metal in produced by electrolysis of magnesium
chloride recovered from seawater.  In addition magnesia (MgO) and magnesium
*
 Data in Figure L and in this section are left unrounded, for purposes
 of information control.  On the average, the typical statistic is
 accurate to within 10 percent, in the opinion of the authors.

-------
Magnesium  Material  Flow-1970
(thousand  tons contained  Mg)


-------
chloride recovered from seawater are used to produce other magnesium
compounds.  The total reported recovery of magnesium compounds (MgO
                                                     (3)
equivalent) from seawater and brines is 625,618 tons.  '  Converting to
magnesium content, magnesium compounds from seawater and well brines
totaled 378,000 tons of magnesium.  In addition, 112,000™' tons of
magnesium metal was produced from seawater.  Thus, total production of
magnesium and magnesium compounds from seawater and well brines was
490,000 tons.
     D.   IMPORTS AMD EXPORTS
          Imports of magnesite in 1970 totaled 128,193 tons.K '  Con-
verting this into total contained magnesium, assuming a 29 percent
magnesium content^ * gives an import of 37,000 tons.
          Total contained magnesium in imported magnesium compounds
reached 8,000 tons, predominantly in the form of epsom salts (magnesium
sulfate).
          U.S. exports of magnesium and magnesium-containing materials
totaled 98,937 tons^ ' ' or 29,000 tons of contained magnesium.
     E.   MGHESIUM OXIDE
          Magnesia (magnesium oxide) consumed other than in refractories
and in the production of other magnesium compounds is reported at 198, 00(
tons (119,000 tons of contained Mg).     It is produced predominantly
from seawater and well brines by calcining.  A primary use is as a
stabilizing or vulcanizing agent in rubber.  Chemical manufacturers use
it in the production of roofiag cement.  The distribution of uses appear'
ing in the flow chart Is based on a percentage breakdown provided by
                         (2)
the U.S. Bureau of Mines. v/
     F.   MAGNESIUM
          Magnesium metal is presently produced solely from seawater,
although it has been recently, and may possibly in the future be
produced from dolomite in a thermic process.  By far the major consump-
tive use has been in the production of alloys.  Its uses are divided
between structural and nonstructural.  Major  notistructural uses

-------
include use as a reducing agent and for cathodic protection of other metals,
particularly iron and steel, in underground pipes, water tanks, and water
heaters.  The predominant structural consumers are the aluminum industry,
which uses magnesium in the production of alloys; and Volkswagenwerk A.G.,
which uses magnesium extensively in its engines.  The distribution in
the flow chart is based on data from the U.S. Bureau of Mines.W
          Secondary magnesium is obtained from new scrap and old scrap.
New scrap consists of borings, skimmings, slags, drosses and defective
articles produced in mills and fabricating plants.  Old scrap is metal
recovered from old aircraft and other obsolete equipment.  Secondary
magnesium is used in alloying, rather than in pure material.  Secondary
metal production totaled 12,042 tons in 1970.' '
     G.   REFRACTORY MAGNESIA
          Refractory grade magnesia is produced by high temperature
calcining from magnesite, seawater, and from dolomite.  It is the
major consumption of magnesium compounds, equal to 483,000 tons of
                    (2)
contained magnesium.     The refractory material is used primarily in
steel and nonferrous furnaces for the parts exposed to molten metal and
slag.  Raw dolomite is also used as a flux and for patching open hearth
furnaces.  Magnesia is also used in conjunction with chromium to produce
chrome magnesite or magnesite-chrome brick.
     H.   MAGNESIUM CHEMICALS
          1.   Magnesium Chloride
               Magnesium chloride is the primary product from all brines
                                                (21
and seawater.   According to the Bureau of Mines,v ' 513,029 tons con-
taining 131,000 tons of magnesium were shipped and utilized for non-
metal producing uses.   It should be noted, however, that almost three
times that quantity was utilized as a feed material in the manufacture
of magnesium metal, refractory materials, and primary magnesium chemicals.
          2.   Magnesium Hydroxide
               Magnesium hydroxide is a common intermediate product in
the production of several magnesium compounds and in metal production.

-------
Its major uses are in sugar refining and pulp and paper,  with some use
as a pharmaceutical.   These three applications consumed 36,000 tons of
magnesium.
          3.   Magnesium Carbonate
               Magnesium carbonate is used as a thermal insulator for
boilers and pipes, and in table salt to prevent caking, as veil as in
Pharmaceuticals and cosmetics.   The total reported to be shipped or
used in 1970 by the U.S. Bureau of Mines was 6,799 tons or 2,000 tons
                       (2)
of contained magnesium.
          4.   Magnesium Sulfate
               Magnesium sulfate or epsom salt is a major pharmaceutical.
It is also used in explosives and matches, in the pulp and paper indus-
try, and in dyes.  A major portion (34,939 tons) was imported in 1970.
                                       ( 2)
This contained 4,000 tons of magnesium.  '  Magnesium sulfate is also
produced from magnesium hydroxide by treatment with tiulfuric acid.  A
total consumption of 6,000 tons of magnesium sulfate was estimated for
the U.S. in 1970.

-------
III. SOURCES AHD ESTIMATES OF Mft.GNESIUM-CONIAINING EMISSIONS
     A.   DATA MES1OT4TION AND ACCU1ACX
          Table 1 presents a summary of the data from which emissions
were estimated for all major potential sources.  Each of the columns
comprising this table will be discussed below.
          1.   Emission^Factors
               Except where indicated, this gives the pounds of total
particulates emitted per ton of production.  Such considerations as:
                    . variations in process conditions among
                      individual plants comprising a source
                      category
                    . inaccuracies in existing data
                    . a limited quantity of existing data
may however, result in an average emission factor for a source category
varying by more than an order o£ magnitude from the value presented.  In
recognizing the need to indicate the level of accuracy of these emission
factors, a reliability code is presented along with each emission factor
value appearing in the table.   This reliability code system is described
below and is based on the system utilized in EPA Document No. AP-42,
"Compilation of Air Pollutant Emission Factors";^ '
                    A; Excellent
                       This value is based on field measurements
                       of a large number of sources.
                    B; Above average
                       This value is baaed on a limited number
                       of field measurements.
                    C: Average
                       This value is based on limited data and/
                       or published emission factors where the
                       accuracy is not stated.
                    D: Below average
                       This emission factor is based on
                       engineering estimates made by know-
                       ledgeable personnel

-------
                     TABLE 1
SOURCES AND ESTIMATES OF MAGNESIUM-CONTAINING EMISSIONS
Source
tanSS, &. Olg^PKOjCgSSIHG
Mining, crushing, and drying
Of dolomite
Mining, crushing, drying,
and briquettiog of raagneslte
OXIDE FK0DUCTIOH
Dolomite - vertical kiln*
Dolomite - rotary kiln*
Magnesite - rotary kiln*
Hydroxide - rotary kilns
METAiLUggt
Magnesium production
Alloying and refining
REFRACTORIES
Grinding and mixing
Electric casting
IHAIITE&tESHT SOURCES
Iron and Steel
Sinter Procea*
Blast Furnace
Open Btarth
Baale OxvgMt
Electric Arc
Coal
Oil
Asbestos
Cement
Uncontrolled Particulate Efciasion Factor
(Ib/ton)

465
475
7
180
180
180

0
4

150
75

20 Ib/ton
sinter
130 Ib/ton
pig iron
17 Ib/ton
steel
40 ib/tas
•teal
10 Ib/ton
steel
H/A
H/A
N/A
(kg/kgxlO3)

233.0
238.0
3.5
90.0
90.0
90.0

0
2

75.0
37.5

10.0
65.0
8.5
20.0
5.0
H/A
H/A
H/A
H/A
Reliability
Code

(C)
(W
(B)
(C)
(C)

(D)
(C)

(C)
(C)

(B)
W
(B)
(B)
(B)
Total*
Production
level
(tona/yr)

2,200,000
631,000
137,300
1,236,000
125,000
625,618

136,500
124,000

483,000
44,000

51,000.000
88,800,000
65,800,000
48,000,000
16,800,000
33,800,000°
287,000C
6,579C
7,790,000°
% Mg in
Emissions

13 (A)
40 (B)
17 (C)
17 (C)
50 (C)
60 (B)

60 (D)

See Note a.
See Note a.

O.fr-6.0 (C)
0.1 •* 3.6 (C)
0,2 - 0.7 (C>
0.4-0.7 (C)
0.2-9.2 (C)
0.70 (A)
0.15 (C)
19.0 (A)
0.07 - 1.15(C)
Mg Emissions
Before
Control*
(tone/yr)

66,500
59,950
82
18,911
5,620
33,700

0
149

36,200
1,650

(3,060 - 30.600)
15.3006
(5,770 - 207,900)
69,300b
(1,130 - 3,5)70)
1,700*
(3,840 - 6,700)
4,800b
(168 - 7,700)
4.200b
236,000
430
1,250
(5,400 - 89.000)
31.1006
Estimated
Level of
Control

95.0
90.3
39.0
81.0
80.0
95.0

0
90.0

80.0
85.0

90.0
99.0
40.0
99.0
78.0
82.0
0
0
88.0
Mg Eodssionj
After
Control*
(tona/yr)

3,320
5,790
50
3,593
1,130
1,690

0
15

7,250
250

(306 •* 3,060)
l,530b
(58 - 2,079)
693b
(670 - 2,340)
l,000b
(38 •* 67) b
(37- •» 1,700)
42.600
430
1,250
(650 - 10,700)
3.73011
586,842 87-°* 75'2«
MOTES: a. Badifion factor multiplier equal to tons of Mg processed or handled annually.
b. Intermediate value (see text).
c. Particular generated, before control. N/A - Not applicable.

-------
           2«   Level  of Production.Activity
               This column  depicts the quantity of material produced  (un-
 less  otherwise stated) annually.  When multiplied by the emission factor,
 an  estimate of the total particulate emissions for that source in pounds
 per year  is obtained.

               The values in  this column are based on the material flow
 calculations presented in  Section II,  Consequently, they have the same
 accuracy  as those material flow values which are estimated at + 10%.

           3.   Percentage of Metal in Emissions
               The method of analyzing or assaying a dust sample for  the
 amount of metal it contains determines to a large extent the reliability of
 the data.  For example, analytical chemistry techniques for dust containing
 substantial fractions of metal can be accurate to within a small percentage.
 On  the other hand, optical spectroscopy methods for determining concentra-
 tions or  the order of parts per million can be inaccurate by a factor of 2,
 Because of this variability, the reliability codes discussed above for the
 emission  factors are also  utilized to estimate the relative accuracy of
 the percentage values listed in Column 3.

          ^*  Level of Magnesium Emissions Before Control
              The values in this column are derived by multiplying the
 values in Columns 1 through 3.  The result is converted to tons/year of
 emissions before control.

          5.  EstimatedLevel of Emission Control
              The overall effectiveness of control for a source category
 is based on two factors:
               . the portion of the processes that are under control
               . the typical degree of control
For example, if 60 percent of vertical roasters have some type of particu-
late emission control and these include both scrubbers and precipitators
                                    11

-------
such that the apparent weighted average efficiency of control is 85 per-
cent, the overall control effectiveness is estimated to be 60 x 85 = 51
percent.

              The accuracy of control efficiency data varies with the
degree of control.  For a set scrubber operating at 80 percent efficiency
(i,e,, pass 20 percent material) the actual emission may safely be assumed
to be between 15 and 25 percent because of the relative ease of making de-
terminations at this level.  Thus, the emissions after control may be as-
sumed to be accurate within + 5/20 or 25 percent.  On the other hand, for
a baghouse reported as being 99 percent efficient, or passing only 1
percent of the material, the actual emission may vary from 0.5 to perhaps
2 percent because it is frequently difficult to make low-level measure-
ments with accuracy. In such cases, the resulting emission data could, be
in error by a factor of 2,

              Unless otherwise specified, it is assumed that the reported
overall level of particulate control applies equally to all magnesium-
containing particles, independent of size, resistance, and other impor-
tant collection parameters.  This assumption results in a correct esti-
mate of magnesium emissions after control when the particulate is chem-
ically homogenious; i.e., the magnesium is contained in the same concen-
tration in all particles.  If however, magnesium is concentrated in cer-
tain particles and in addition, the efficiency of the control equipment
is not uniform for all particles, then the utilization of an average con-
trol level is less valid for calculating magnesium emissions after control.
Data on the preferential control of magnesium-containing particles is
seldom available, but is included in this report, when possible.

              The accuracy of estimating the level of control for a spe-
cific source category is dependent on the quality of available data.  The
investigators feel that, in general, the level of control data will con-
tribute an accuracy to the resulting emission estimates within + 25 per-
cent.
                                    12

-------
           6.   Level  of Metal Emissions After Control
               The values  in this column are derived by multiplying the
values  in  Column 4 by the value  (100 minus estimated level of control).

      B.   DEVELOPMENT OF  EMISSIONS ESTIMATES - 1970

           1.   Mining of Dolomite
               Dolomite and limestone are mined and treated much in the
same manner,  in that dolomite is the double carbonate of calcium and
magnesium  (CaCO-'MgCO,,).  Emissions are generated from open-pit mining;
crushing and  screening; and washing and drying operations, preparatory
to burning.

               Mining is estimated to generate emissions on the order of
1 Ib/ton and  is uncontrolled.  Crushing and screening are estimated to
generate 24 Ibs/ton, and  since most emissions are uncontrolledj the over-
all level  of  control effectiveness is estimated at 20 percent.    Drying
the washed, crushed  rock  generates an estimated emission of 440 Ibs/ton
                                                    (Q\
which is well  controlled  at an estimated 99 percent.     These three
emission sources combine  to 465 Ibs/ton, before control, and 24.6 Ibs/
ton after  about 95 percent control.  The typical magnesium content in
dolomite is 13 percent.     An estimated 2.2 x 10  tons of rock was pro-
cessed in  1970, of which  only about two thirds was converted to magnesium
oxide.  Further processing of dolomite will be discussed under magnesium
oxide production (Section 4).

           2.  Mining of Magnesite
              At present, only one magnesite and brucite mine is operating
in the U.S.  In preparing magnesite, the steps are similar to dolomite;
i.e.,  crushing and drying.  An overburden and waste tonnage equal to
twice  the magnesite produced, triples the emission up to the point of
drying,  but simultaneously reduces the magnesium content of the dust by
the same proportion.   At  this mine, the mining, primary crushing, and
screening operations are understood to be controlled, resulting in 3332Q
                                   13

-------
tons of magnesium emitted.  The drying operation is estimated to be con-
trolled at 96 percent efficiency, resulting in an emission of 2,220 tons
of magnesium.  After drying, binding ingredients are added to the magne-
site concentrate to make  it the consistency of heavy dough.  It is then
pressed into briquettes,  and dried in rotary kilns at moderate tempera-
tures.  It is next crushed, screened, and stored.  The drying and crushing-
screening operations are  estimated to generate 100 Ibs/ton and 25 Ibs/
ton of emissions, respectively, at 99 and 96 percent control for a net
emission of 631 tons of dust or 250 tons of magnesium.  Thus, the mine
and plant together are estimated to produce a total of 5,790 tons of
magnesium emission, after an average control level of 90.3 percent.

          3.   Recovery.. From Salt Brines
              Sea, well,  or lake waters containing magnesium chloride are
treated with ground, roasted oyster shell or with a magnesium compound to
form a precipitate of magnesium hydroxide.   This precipitate is subse-
quently calcined to MgO (Section 4) or treated and electrolytically separ-
ated to magnesium metal (Section 5).   Although substantial quantities of
precipitate are produced, no emissions are expected except in the case of
plants that dry the precipitate for sale or prior to calcining.  Emissions
from drying in a rotary drier would be expected on the order of 100 Ibs/ton,
and would be fairly well  controlled (90 to 95 percent, probably).  However,
no data on the drying of  precipitate was obtained.  It is assumed that any
emissions from this process are included with calcining emissions.

          4-   Magnesium Oxide Production
              A variety of oxides are produced for various uses by
varying the feed material and the temperature of calcining.  Magnesite
or MgOH calcined at temperatures below 900 C produces caustic-calcined
magnesia, a relatively reactive material used in cements, Pharmaceuticals,
and the rubber industry.  Dolomite is normally calcined at about 1100 C.
However,  dead-burned dolomite is produced in rotary kilns at about 1700 C,
and dead-burned magnesite at up to 1460 C.   Dead-burned magnesias thus
produced are used largely in refractory materials.

                                   14

-------
              The  quantity of emissions from  these processes probably
depends more  on  the preparation of  the feed material and the type of cal-
cining equipment than  on  the temperature of operation.  Material prepara-
tion and  equipment used are similar to the lime industrys for which data
is more readily  available.  Emissions factors are estimated at 180 Ib/ton
for rotary kilns  and  7 Ib/ton for  vertical kilns, with  levels of control
at 81 and 39  percent,  respectively.   ' *      It is estimated that 10 per-
cent of the dolomite is processed in vertical kilns, and that all hydroxide
precipitate and  magnesite are processed in rotary kilns.  The magnesium
content of the particulate is taken to be midway between those of the
feed and  the  final product.  The levels of emission control for hydroxide
and magnesite calcining are estimated at 80 and 95 percent, respectively,
for 1970,  Recently, the  installation of new  control equipment has, no
                                                    /Q \
doubt, raised these levels of control substantially.     The total emis-
sion estimate for  production of magnesium oxide is 6,463 tons of magnesium
into the  atmosphere.

          5,  MagnesiumMetalProduction
              In 1970, the only process for producing primary magnesium
was the electrolytic reduction of magnesium chloride to magnesium and
chlorine.  The precipitated hydroxide (Section 3) is concentrated by
filtration, then redissolved in hydrochloric  acid to form a much more
concentrated  solution  of magnesium  chloride than the original salt brine.
This solution is then  fed to electrolytic cells for reduction.

              In addition to the production of commercial magnesium by
this process by  a  single plant in Texas, between one and three plants in
the titanium industry  also use the  process to recycle magnesium for their
own use.  The production of chlorine gas from all these processes requires
that the  emission be well controlled.  Even before control, only minute
quantities of magnesium would be emitted with droplets in the gas stream.
Therefore, it is assumed that emissions from  the production of magnesium
metal were negligible  in 1970.
                                   15

-------
          6•  Refining and Alloying
              Since molten magnesium reacts vigorously with oxygen and
nitrogen, precautions are taken to avoid contact with air.  The two
methods for doing this are:  (1)  use of fluxes  floated on top of the molten
metal in the open pots or crucibles and (2) melting in a closed pot con-
taining sulfur dioxide.  No magnesium emissions are expected from the
molten metal in alloying.  However, magnesium chloride is a basic con-
stituent of the fluxes used with the open pot crucible.  The chloride
forms magnesium oxide when in contact with water, and slight emission of
fluxing magnesium may result.  The emission factor of 4 Ib/ton is for a
pot furnace.     The assumption is made that all primary plus secondary
magnesium metal (124,000 tons) was refined, alloyed, or smelted by the
open pot process.  An estimated level of control of 90 percent is based
on controls commonly used in alloying and secondary metal refining.  The
resulting emission estimate is a negligible 15 tons.

          7.  Refractory Material Production
              Fire brick and both wet and dry patching compounds are made
in large quantity from magnesia derived from both magjnesite and dolomite.
After being calcined at 900 to 1700 C, the magnesia is finely ground,
classified, and mixed with other brick ingredients such as chromite or
clay.  To mix well, drying may be necessary.  These processes appear to
be major sources of emission, although data are scant.  Subsequently,
the mixture is cast into brick and dried in stationary kilns at low
temperatures; or if fused-cast, melted in an electric arc furnace at up
to 2500 C and then poured and cast.  During the drying in kilns and
through the miscellaneous process of handling, the bricks and compounds
appear to generate negligible emissions.  The electric casting operation
is estimated to be a substantial source of emission, however.

              Based on emissions estimated for clay refractory materials,
and for similar grinding and melting operations, emission factors of 150
Ib/ton for all mechanical operations, and an additional 75 Ib/ton for
melting and casting, are assumed.  Of the 483,000 tons of magnesium used

                                   16

-------
in refractories, it is estimated that about 10 percent is used in patch-
ing materials,     and that only about 10 percent of the brick is cast,
or about 44,000 tons.  The level of emission control should be higher than
in the case of clay refractories (64 percent) because the materials are
more valuable.  A level of control of 80 percent for the mechanical pro-
cesses is assumed, and 85 percent for the electric melting and casting
processes.  The resulting total emission estimate is 7,500 tons of mag-
nesium emitted to the atmosphere; the largest estimated emission from the
magnesium industry,

          8.  Magnesium Chemicals
              Emissions from the production and use of magnesium-based
chemicals are estimated to be negligible.

              a.  Magnesium Chloride

                  In addition to usage in the production of oxide and
metal, MgCl is also used in oxychloride cements for flooring and
plaster, as a fireproofing agent for wood, as a dust binder, and in
fluxes for alloying.   A small segment of the wood pulpina; indus-
try uses magnesium-based chemicals for separating wood fibers.  None of
these processes appear to generate emissions of any magnitude.

              b.  Magnesium Hydroxide
                  This is used  in the pharmaceutical, milk o£ magne-
sia, in sugar refining, and by the paper industry.   None of these pro-
cesses appear to generate emissions of any magnitude.

              c.  Magnesium Carbonate
                  This is used in the production of insulation for steam
pipes; in paints, plastics, and paper; and as a soil conditioner.  The
total quantity of carbonate consumed is very small.
                                  17

-------
              d.  Magnesium Sulphate
                  This Is used as spsom  salts, dyes, drying agents  for
organic solvents, tanning agents , and  fireproof ing ,  The  total con.siitnp-
tioa is small.
              a.  Iron: and Steel Production
                  The iron and steel industries use raw dolomite  and
limestone (containing small amo'mcs of magnesium) as fluxes in  the  pro-
duction of iron and s<;ael.  Magnesium metal is also used  as a scavenger
    deaxidixar .  Ffrctharmoita, the furnaces and containers  used  for  most
molten iron and steel ere lined v;ith refractory material,  much  of which
contrains magnesia ,  Open hearth furnaces require about 10 Ib of  msgnesite
brick per ton of .steel produced, for example.      It is,  of course,  not
expected  tha<: all these sources of magnesium will  result in emissions
from the fieas-ces.  Since magnssiuro has a much lower vaporization tempera-
ture than most of the Q'cher elements present, at least some magnesium
emission is expected.

                  Table 1 gives the quantities of iron arid steel  produced
by the major processes , together with generally accepted  emission and con-
trol factors.   Data concerning the amount of magnesium contained  in the
emission was found to be as follows:
Sintering
Blast
Open
Basic
fun
iscc
hearth
oxygen
Electric arc



a.
b.
c.
0.
0,
0.
0.
0.
6-6a
1-3
2-0
4-0
2-9
Reference
Reference
Refer en
ce
a
,7s
,7a
a
7
12
14
R
0.
2-3. 6b 0.
0.23° O.le 0,
0.
2-9d 9.2,4S8C 1.2e 0.
d. Reference 13
e. Reference 15

6-
1-
gj
6
3.
2-0.
4-0.
2-



9.



s
6
7
7
2



Est.
3.
1,
0.
0.
5.



Ave
0%
2%
^&/
5%
01



                                   18

-------
These  Figures  indicate a  total estimated  average of 4,195 tons of
magnesium after control,
             b.  Coal Combustion
                 Coal consumption in 1970 amounted to 517,000,000
tons.      The  particulate generated has been estimated at 33,800,000 tons
of which  82% was controlled, leaving 6,100,000 tons of emission to the
atmosphere.     Spectrochemical analysis  of 373 samples of coal from
across the U.S. indicated an average content of 0.70 percent maenesium
in the ash.      Thus, it is estimated that 236,000 tons of magnesium
before control or 42,600  tons of magnesium after 82 percent control, was
emitted into the atmosphere.  This is almost six times greater than the
next largest source of magnesium identified in this study (refractory
production).

              c.  Oil Combustion
                  It is estimated that 287,000 tons of particulate are
generated  by the combustion of residual oils in the U.S. and that this
particulate is not under any significant degree of control.     The con-
tent of magnesium in this particulate has been reported as 0.27 percent
(two samples, Ref. 7); 0.053 percent (three samples, optical emission
spectroscopy, Ref. 17), and 0.20 percent  (one sample, several analytical
techniques, Ref. 18).  Using an average 0.15 percent, results in an esti-
mated emission of 430 tons.

              d.  Asbestos
                  Emissions of asbestos are estimated to have been 6,579
tons in 1970, after control.     Chrysolite, the predominant fiber of
asbestos,  is a silicate of magnesium containing approximately 19 percent
magnesium.  This results in an estimated emission of 1,250 tons of mag-
nesium in  1970.

              e.  Cement Production

                  It is estimated that 934,000 tons of particulate after con-
trol were generated from the various processes of cement production in 19?0.

                                   19

-------
Based on limestone, which is distinguished from dolomite in that it con-
tains a low percentage of magnesium, one expects that cement emissions will
contain significant, but highly variable amounts of magnesium,  Partieulate
concentrations of magnesium are reported to be:  0,07 percent (three
samples of kiln and clinker cooler dust, Ref. 19); C.4 percent (three
samples from air separators, Ref. 19); and 1.15 percent (kiln dust, Ref.
7).  Based on these percentages, the range of magnesium emissions is 650
to 10,700 tons/year after 88 percent controls with an intermediate value
of 3,730 tons/year,

      C.  SUMMARY OF PRINCIPAL EMISSIONS
          Table 2 summarizes the major sources and estimated emissions
of magnesium, as developed in Table 1 and accompanying discussion.  The
sources are grouped Into two categories,  those directly originating within
the magnesium industry, and those having no relationsiiip to the magnesium
industry, called inadvertent sources.  The latter category represents
about 62 percent of the total estimated emissions.  These principal esti-
mates are examined further in section IV of this report.
                                 TABLE 2
         SUMMARY OF PRINCIPAL SOURCES AND EMISSIONS OF MAGNESIUM

Inadvertent Sources
Coal combustion
Iron and steel production
Magnesium Industry Sources
Refractories production
Magnesite mining and processing
Dolomite mining and processing
Dolomite calcining, rotary kilns

U.S. Tons/
Year of Mg
42,600
4,195
7,500
5,790
3,323
3,593

% of U.S.
56.5
5.6
10,0
7.7
4.4
4.8
89.0
- - i
                                   20

-------
 IV.  REGIONAL  DISTRIBUTION OF PRINCIPAL SOURCES AND EMISSIONS
     For  the purpose  of  showing geographical distribution, the  United
 States was  divided  into  ten  regions identical  to the Regional Branches
 of EPA.

          Region                           States
            I          Conn., Me., Mass., N.H., R.I., Vt.
            II          N.J.,  N.Y., P.R., V.I.
          Ill          Del.,  Md., Pa., Va., W.Va., D.C.
            IV          Ala.,  Fla., Ga., Ky., Miss., N.C., S.C., Tenn.
            V          111.,  Ind., Mich., Minn., Ohio, Wis.
            VI          Ark.,  La., N.M., Okla.,  Tex.
          VII          Iowa,  Kans., Mo., Nebr.
          VIII          Colo., Mont., N. Dak., S. Dak., Utah, Wyo.
            IX          Ariz., Calif.,  Nev., Hawaii and the South Pacific
            X          Alaska, Idaho,  Oreg., Wash.

     The  principal  emission  sources listed in  Table 2 are distributed among
 these ten regions,  as  shown  in Table  3.  Also, the number of plants pro-
 ducing the  emissions are shown in the table, when such information is
 available.

     When the  estimated number of plants was greater than 100, and the
 distribution of plants was not known, the regional breakdown was made on
 a different basis,  such as population, geographical area, or shipments
 reported, as most appropriate for that category.  Whether the distribution
 was by plant size,  number of plants,  or another statistic, the distribu-
 tion is believed to be accurate to within 10 percent in most cases.

     The  accuracy of the distributions by region varies with the category.
 The number  of  plants per category varied from  one to several thousand in
 this study.   When the number of plants was less than 100, an attempt  was
made to identify each plant and plant location, and include it in one of
 the ten regions.  When production or capacity  figures for these plants
were available, total production or capacity for each region was computed,
                                   21

-------
                        TABLE  3





REGIONAL DISTRIBUTION  OF PRINCIPAL SOURCES  AND EMISSIONS
Principal
Sources
Inadvertent Sources
Coal combustion
Iron and Steel
Production
Magnesium Industry
Sources
Refractories
production
Magnesite mining
Dolomite mining
Dolomite calcining

TOTALS :

EPA REGION
1
298
0.7
232
23
357
3
0
0
237
2
257
2
1381
1.8
2
242S
5.7
334
33
833
7
0
0
0
0
0
0
3595
4.8
3
9,287
21.8
1121
111
2025
17
0
0
475
4
513
A
13,421
17.8
4
8,989
21.1
395
39
357
3
0
0
119
1
128
1
9,988
13.3
5
17,422
41.3
1477
146
2975
25
0
0
1539
13
1668
13
25,081
33.3
6
597
1.4
201
20
0
0
0
0
0
0
0
0
798
l.l
7
174
4.1
50
5
120
1
0
0
119
1
128
1
2,164
2.9
8
1406
3.3
20
2
357
3
0
0
237
2
257
2
,220
2.9
9
298
0.7
294
29
476
4
5790
1
594
5
642
C
-)
8,094
10.8
10
128
0.3
70
7
0
0
0
0
0
0
0
0
198
0.3
TOTAL (units)

42,600 (tpy)
100 (% of U.S.)
4,195 (tpy)
415 (U.S.
plants)
7,500 (tpy)
63 (No. Plants)
5,790 (tpy)
1 (No.
Plants)
3,320 (tpy)
28 (No.
(Plants)
3,593 (toy)
28 (No.
(Plants)
66,997
«9-0 (% of U.S.)
Ref.

4
20
21
4
22
22



-------
and the U.S. emission estimate for that category was distributed by region
accordingly.  When production or capacity figures were not available, the
emission was distributed by the number of plants in each region.  If the
number of plants was very small or there was reason to believe that cer-
tain plants were larger or produced more emission, distributions were
weighted accordingly.

     Specifically5 the largest emission estimate; i.e., from the combus-
tion of coal, was distributed percentagewise, according to coal shipments
in 1970, by state of destination.  Emissions from iron and steel produc-
tion were distributed by total number of iron and steel producing plants,
by state.  This is an approximation since emissions from certain processes,
notably sintering and open hearth steel production, are larger than from
other processes.  Information is not available as to numbers and sizes
of sintering equipment by state in the U.S., nor for open hearth production.
It is implicitly assumed that the distribution of these largest emission
processes will be similar to that of the overall iron and steel industry.

     Emissions from production of magnesium-containing refractories are
distributed by number of plants reporting products of that type, by state,
as no information as to the relative sizes of the plants was available.
Only one plant was engaged in the mining of magnesite in 1970.  Nearly
all dolomite mining and calcining were done in the same locations, with
the possible exception of a small quantity of additional calcining per-
formed by manufacturers of refractory materials to meet their own speci-
fications.  Ihus, emissions from mining and calcining were distributed
in the same way.  A map of locations of plants operating rotary kilns for
dolomite calcining was obtained from the National Litne Association.  The
list was supplemented by a list of five additional plants producing dead-
burned dolomite, also obtained from the same source.  Again, distribu-
tion by number of plants does not reflect the relative sizes of the
plants, nor the relative cleanliness of their operations.
     As a result of these distributions, Region 5 was estimated
to release the largest amount of magnesium to the atmosphere,
                                    23

-------
about 33 percent of the U.S. total.   Reeions  3  and  4 were  next,
with another 31 percent combined.  These results are largely due to the
estimates regarding emissions from coal combustion, which contributed
roughly two-thirds of the emissions from each of these regions.

     Considering the geographical area of the ten regions, the most con-
centrated emission of magnesium, averaged over the area, was in Region 3,
with 0.11 tons of magnesium per square mile per year, closely followed
by Region 5 with 0.075 tons per square mile per year.
                                   24

-------
V.         OF EMISSIONS
    The physical characteristics of the particles emitted to the atmo-
sphere containing magnesium, and the chemical form of magnesium, are
the results of the exact process variables, feed materials, and the nature
of magnesium itself.  Tables 4a and 4b describe some of the important min-
erals containing magnesium, and some of the properties of the metal.  The
boiling, or vaporization temperature of the various minerals and of mag-
nesium is probably the most important property from the standpoint of
emissions.  Processes that exceed that temperature may be expected to
generate large fractions of the magnesium compound present, while pro-
cesses operating at lower temperatures should be relatively non-emitting
of that magnesium compound as a vapor.

    Magnesium is among the most readily oxidized elements, which is one
of the reasons it is used as a scavenger in steel making.  It can be ex-
pected to form chemical bonds with oxygen and other elements and element
groups rather than remain in metallic form.  This is highly probable
when hot finely divided particulate is exposed to oxygen.  Subsequently
the oxide may absorb water or carbon dioxide.

    The combustion of coal generates the largest emission of magnesium
to the atmosphere.  The form of magnesium in the coal, before combustion,
is not established, but is probably a carbonate.  Combustion at tempera-
tures on the order of 1500 G is expected to partially reduce the carbon-
ate to the oxide.  Since, however, the oxide has a vaporization tempera-
ture on the order of 3600 C, the oxide will probably not be emitted as a
vapor, but as an inclusion with the flyash.

    Analysis of several sizes of flyash particles for magnesium con-
tent showed a uniform 0.6 percent magnesium contained in all parti-
cles greater than 1.7 micrometers, and 0.8 percent in particles smal-
                         (25)
ler than 1.7 micrometers.      Flyash particles themselves vary in
size from submicron to greater than 100 micrometers, but average
(mass median) diameters are typically from 3 to 10 micrometers.
                                   25

-------
                   TABLE 4a<8<23»24>
DESCRIPTION OF MAGNESIUM MINERALS AND PRODUCTS
Mineral
Brucite
Dolomite, raw
Dolomite, calcined
Dolomite, dead-
burned
Magnesite, raw
Magnesite, dead-
burned
Magnesia
Magnesia, caustic
calcined
Magnesium peroxide
Olivine
Periclase
rl
Formula
MgO' 1^0
CaC03MgC03
	
	
MgC03
. 	
MgO
MgC03 or MgOH
Mg02
(Mg,Fe)2Si04
MgO
Process
Temp.
Natural
Natural
1100°C
1700°C
Natural
Same as magnesia
Over 1460°C
950°c or less
	
Natural
1700°C
(Also natural)
Hardness
2.5
3.5-4.5
	
—~—
3.5-4.5

	
	
„__
6.5-7.0
Specific
Gravity
(g/cc)
2.4
2.8-3.0
,__
	
2.95-3.2

	
	
	
3.3-3.4
3.58
Natural materials include varying amounts of other ingredients.
Temperature required fot process completion.
Standard minerology scale of hardness.
                   TABLE 4b (Ref.  23,  Table 3-169)
         PROPERTIES OF MAGNESIUM METAL
  Melting point:          650°C
  Boiling point:          1107°C
                         (MgO boiling point:  3600 C
  Density;               1.74 g/cc
  Atomic weight:          24.3 a.w.u.
  Heat of vaporization:   32.5 kg-cal/g atom	
                      26

-------
Most of the equipment used to control flyash emissions tends to remove
the larger particles and allow the smaller ones to escape.  Particles
on the order of 1 micrometer and smaller may be expected to travel in-
definitely (i.e., tens of miles) before being removed from the air by
natural processes (settling, agglomeration, washout, etc.).  No physi-
cal or chemical changes of the magnesium contained in the particles are
expected to take place during this period, except for possible absorp-
tion of water or carbon dioxide.

    The next largest estimated source of emission, the production of re-
fractory materials, is due mostly to mechanical processes including
grinding, crushing, screening and classification mixing, etc.  Grind-
ing may be expected to generate a substantial quantity of particles on
the order of 10-micrometer diameter, as well as some on the order of
1-micrometer diameter.  The smaller ones will tend to escape, while the
larger, if not immediately controlled, will tend to settle in the vi-
cinity of the grinders.  The particles that escape will be non-spherical,
and will contain magnesium predominantly as an oxide.  Screening and
mixing operations, being less intensive in terms of the stresses applied
to the particles, will tend to emit 10- to 100~mierometer particles and
agglomerates, while a relatively small fraction of the material in
particles 1 micrometer and smaller will escape the vicinity of the plant.

    Emissions from mining and calcining are expected to include both the
carbonate and oxide forms of magnesium.   The emissions from a concentrate
drier ahead of the kiln are reported to be 95 percent less than 5 mi-
crometers,  85 percent less than 3 micrometers, and 10 percent less than
             / Q \
1 micrometer.     The literature reports analysis of particulate emitted
                                            (*? £c\
from 20 dolomite kilns in Russia as follows:
                30- to 70-micrometers diameter before control
                40 to 50 percent potassium oxide plus magnesium
                10^ to 10  megohm-cm electrical resistance
Both calciner or kiln and drier particles are expected to be non-spherical
and fairly inert chemically.
                                    27

-------
    Dusts in the vicinity of iron works have been analyzed for particle
size distributions and MgO content.  The average particle sizes ranged
from 1 to 4 micrometers; the MgO content ranged from 3,3 to 6,8 percent.
The variation of MgO concentration with particle size was not reported,
however.
    In conclusion, the sizes and chemical forms of magnesium containing
particles are such that most of the particulate emitted without control
measures, and a modest fraction of the particulate emitted despite con-
trol measures, will probably not travel far from the emission source.
However, a large portion of the U.S. emissions of magnesium will exist
in micrometer and smaller particles and will travel long distances from
the plant.  This magnesium is expected to be chemically inert except
for some possible absorption.
    The magnesium content of city air is reported in one instance as
                                                  3
 0.42 (total particulate concentration = 44.7 pg/jn ) to 7,21 micro-
grams per cubic meter (total particulate not measured) in particles
averaging (mass median diameter) 4.5 to 7.2 micrometers.  Between 17
and 23 percent of the mass was contained in particles less than 1
                    (27)
micrometer diameter.
                                  28

-------
VI.  UPDATING OF EMISSIONS ESTIMATES

     The following recommendations are made for periodically updating the

estimates made in this study;


     A.  VERIFICATION OF CURRENT ESTIMATES

         1.  Verify that the estimated concentrations of magnesium
             contained in emissions from the production of iron and
             steel are representative.  Although the concentrations
             are small, the tonnages of emissions from these pro-
             cesses are potentially large enough to affect the con-
             clusions of the study, if the concentrations are in
             fact higher than estimated.  If the concentrations
             prove to be higher, the emissions estimates should be
             distributed by region, by types of iron and steel
             processes, rather than number of plants of all types.

         2.  Verify that the emission factors for the production
             of refractory materials, and for the calcining of
             dolomite and magnesite, are representative.  The es-
             timates used in this study are based on a limited
             quantity of data which also showed considerable vari-
             ation from data source-to-source.
     B.  PERIODIC REVIEW OF ESTIMATES

         1.  The Bureau of Mines estimates for material flow, in-
             dustry practices, and trends provide the best esti-
             mates of the size of the industry.

         2.  EPA activities are currently generating the best
             emissions data and should be reviewed using:

             a.  Overall industry studies (e.g., Reference 25),
             b.  The Source Test Program in which specific indi-
                 vidual plant emissions are measured.  This in-
                 formation provides emission factors for speci-
                 fic examples of typical industrial operations;
                 and also provides some analyses of the particu-
                 late usually including trace metal content and
                 particle size.

             c.  NEDS (National Emissions Data System) is steadily
                 being enlarged and improved.  This system can
                 provide emission factors for specific plants and
                 plant operations, the type of particulate control
                 equipment in use, and the actual, or estimated,
                 control efficiency.   The system may eventually be
                 expanded to include description of the emissions.

                                    29

-------
3,  The magnesium Industry should be consulted for Its
    opinion and suggestions on the most recently pub-
    lished estimates.  This may be best accomplished by
    interviewing the Magnesium Commodity Specialist,
    Division of Non-ferrous Metals, Bureau of Mines, in
    Washington, B.C.; or by interviewing one or more of
    the principal companies in the industry.

4.  The literature should be reviewed, using (a) indus-
    trial views as published from time to time in
    Chemical Engineering, for example and (b) environ-
    mental views as summarized in Pol1ut ion Abstracts,
    for example.

5.  Individual companies or plants may be approached for
    opinions, data, or cooperative tests of their own
    operations.  This is a difficult approach to the
    problem of obtaining fresh information due to the
    natural reluctance of the plants to discuss environ-
    mental problems.  However, data thus obtained have
    a relatively high degree of reliability.

6.  State agencies in which specific plants are located
    may be able to provide useful information, and
    should be contacted.
                          30

-------
VII.  REFERENCES

      1.  1971 Engineering and Mining Journal International Directoryoj
          Mining and Mineral ProcessInformation, Mining Information
          Services, New York (1971).

      2.  Preprint from Mineral Yearbook /'Magnesium Compounds," U.S.
          Bureau of Mines, Washington, D.C. (1971).

      3-  1971 Statistical Abstracts. U.S. Dept. of Commerce, Washington,
          D.C. (1971).

      4.  Minerals Yearbook. 1971, U.S. Bureau of Mines, Washington, D.C.
          (1971).

      5.  Hazel B. Cornstock,"Magnesium and Magnesium Compounds, A Material
          Survey," U.S. Bureau of Mines Information Circular 1C 8201,
          Washington, D.C. (1963).

      6.  "Compilation of Air Pollutant Emission Factors," U.S. Environ-
          mental Protection Agency, AP-42, Research Triangle Park, N.C.
          (1972).

      7.  A.E. Vandergrift, et al., "Particulate Pollutants Systems Study,"
          Handbook of	Emlssion Properties, Vol.  Ill, Midwest Research
          Institute,  Kansas City, Mo.  (1971).

      8.  Personal Communications with members of the Magnesium industry.

      9.  Oliver Bowles,  "Limestone and Dolomite," U.S.  Bureau of Mines
          Information Circular IC7738, Washington, D.C.  (1956).

     10.  NEDS (National  Emissions Data System)  data was used to supple-
          ment data more  readily available in the literature.

     11.  D.J. Kusler and R.G.  Clarke, "Impact of Changing Technology on
          Refractories Consumption," U.S.  Bureau of Mines Information
          Circular No.  8494 (1970).

     12.  V.  Masek, "On the Composition of Dusts on Work Locations and in
          the Near Vicinity of Iron Works," Staub, 31(2), 27 (Feb. 1971).

     13.  W.W, Campbell and R.W.  Fullerton, "Development of an Electric
          Furnace Dust Control  System," JAPCA,  12(12),  574 (Dec.  1962).

     14.  Southern Research Institute,  Birmingham,  Ala.,"A Manual of
          Electrostatic Precipitator Technology," Report to NAPCA (EPA),
          Contract No.  CPA-22-69-73 (August 1970).

     15.   Public  Health Service,  U.S.D.H.E.W., Air Pollution Engineering
          Manual,  99-AP-40 (1967).


                                   31

-------
16.  R.F. Abernathy, et al. , "Major Ash Constituents in U.S. Coals,"
     Bureau of Mines, U.S. Dept» of the Interior, Report of Investi-
     gation No. 7240 (1969).

17.  A. Levy, et al., "A Field Investigation of Emissions From Fuel
     Oil Combustion for Space Heating," Report by Battelle Institute,
     Columbus,Ohio, to American Petroleum Institute, API Proj.  No.
     SS05 (1 Nov. 1971).

18.  D.J. Lehmden, R.H. Jungers, and R.E. Lee, "The Determination of
     Trace Elements in Coal, Flyash, Fuel Oil, and Gasoline," Part 1.
     Presented at the American Chemical Society Meeting, Dallas,
     Texas (April 1973).

19.  Source Test Reports of emissions from specific industrial  opera-
     tions, obtained in part from EPA.

20.  American Iron and Steel Institute, "Iron and Steel Producing
     and Finishing Works of the U.S.," a table in the Directory of
     Iron and Steel Works of the U.S. and Canada (1970).

21.  Product Directory of the Refractories Industry in the U.S..
     The Refractories Institute, Pittsburgh, Perm.  (1968).

22.  Commercial Lime Plants in U.S. and Canada (Map), National  Lime
     Association, Washington, D.C. (1970).

23.  Perry's Chemical Engineers Handbook, 4th Edition, McGraw Hill
     Inc., New York (1963),

24.  "Mineral Facts and Problems 1970," U.S. Bureau of Mines Bulle-
     tin 650, Washington, D.C.

25.  R.E. Lee, "Trace Metals in Fly Ash as a Function of Particle
     Size," a single page unpublished table, obtained via R.E.  Lee,
     Research Triangle Park, N.C.  (EPA).

26.  L.C. Chally, "Physics - Chemical Properties of Dust in Waste
     Gases During Firing of Refractory Raw Materials in Rotary  Kilns,"
     Stal, 30(10), 931, Oct.  1970 (Russian), English abstract,
     APTIC No. 28311.

27.  R.E. Lee, et al., "Particle Size Distribution of Metal Compo-
     nents in Urban Air," Env.  Sci. and Tech., 2(4}, 288 (April 1968).
                              32

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1 REPORT NO, 2.
EPA-450/3-74-010
4. TITLE AND SUBTITLE
National Emissions Inventory of Sources and
Emissions of Magnesium
7. AUTHOH(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA Corporation
6CA Technology Division
Bedford, Massachusetts 01730
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
Research Triangle Park, N. C, 27711
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
May 1973
6, PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
2AE132
11. CONTRACT/GRANT NO.
68-02-9601
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A national inventory of the sources and emissions of the element magnesium
was conducted. All major sources of magnesium-containing emissions were iden-
tified and their magnesium emissions into the atmosphere estimated. Also, a
method for updating the results of the study every two years was recommended.
17. KEY WORDS AND DOCUMENT ANALYSIS
•». DESCRIPTORS b.lDENTIFIERS/
Magnesium
Air Pollution
Emission
Inventories
Sources
13, LISTmBUTION STATEMENT 19. SECURITY C
Uncl
Release Unlimited 20. SECURITY c
Uncl
OPEN ENDED TERMS C. COSATI Field/Group
)
i
L.ASS (This Report/ 21. NO. OF PAGES
assified 32
LASS (This page) 22 . P R I CE
assified
EPA Form 2220-1 (9-73)
                                                          33

-------