RESEARCH   TRIANGLE
    Contract No. 68-02-0088
    RTI Project No. 4IU-649
N  S T I T  U T  E
        August 1972
                          FINAL  REPORT
                           FR-41U-649
                             Volume 11
    COMPREHENSIVE STUDY OF SPECIFIED AIR POLLUTION SOURCES  TO
       ASSESS THE ECONOMIC IMPACT OF AIR QUALITY STANDARDS
                   18BESTOS, DERYLLIUM,  MERCURY
                                 by

                    Richard E. Paddock, Franklin A. Ayer,

                      Alex B. Cole, David A. LeSourd
                             Prepared for

                       Division of Effects Research
                       vironmental Protection Agency
RESEARCH TRIANGLE PARK, NORTH CAROLINA  27709

-------
Research Triangle Institute
Environmental Studies Center
Research Triangle Park, North Carolina
FINAL REPORT
FR-4lU-649
Volume II
STUDY OF SPECIFIED
HAZARDOUS POLLUTION SOURCES TO ASSESS
THE ECONOMIC EFFECTS OF AIR QUALITY STANDARDS
by
Richard E: Paddock, Franklin A. Ayer,
Alex B. Cole, David A. LeSourd
Prepared for
Division of Effects Research
Environmental Protection Agency

-------
ABSTRACT
Estimates are made of the costs of controls to reduce the
emissions of asbestos fibers, beryllium, and mercury from primary
production and selected secondary sources within the Nation.
Production processes and control technology are examined. Controls
are selected to meet assumed or proposed emission standards and
costs are estimated. In addition, an extended analysis is made
where appropriate, to determine the economic impact of control costs
on each industrial source or group of industrial sources studied.
The effects on prices are also estimated. Under the assumed
implementation plan, the estimated costs are those that will be
incurred during the period of Fiscal Year 1970 through Fiscal Year
1977.
ii

-------
TABLE OF CONTENTS
ABSTRACT. . . .
LIST OF TABLES
LIST OF FIGURES
Chapter 1:
II.
III.
Chapter 2:
I.

II.
III.
IV.
VI.
VII.
A.
B.

Chapter 3:
I.
V.
..............
. . . .
.....
.....
......
.....
........
......
. . . .
8"""
Overview.
. . . .
. . . .
,.....
......
INTRODUCTION.
A.
B.
.......
. . .
......
Health Importance. . . . . .
Economic Importance. . . . .
. . . .
........
. . . .
.....
STUDY METHODS
SUMMARY

A. Asbestos
B. Beryllium
C. Mercury.
A.
B.
C.
"""'"
.......
Control Standards. r. . . . . . . . . . . . . . . . .
Cost Estimation and Economic Impact. . . . . . . . .
Report Organization. . . . . . . . . . . . . . . . .
.....
.....
........
. . . .
.....
.....
.......
.....
. . . .
......
.......
.....
. . . .
......
.....
. . . .
Asbes tos . .
. . . .
. . .
. . . .
. . . .
. . . .
INTRODUCTION
.......
. . '. . . .
. . . .
. . .
CONTROL STANDARD

CONTROL TECHNIQUES. .

ASBESTOS MINING AND MILLING
A.
B.
C.
D.
. . . .
.....
. . . .
. . . .
.......
.....
.......
Occurrence and Mining Techniques. . . . . . .
Control Techniques for Mining Operations
Asbestos Milling Processes. . . . . . . . . . . . .
Economic Impact. . . . . . . . . . . . .
ASBESTOS HANDLING
A.
B.
. . . .
. . . .
.........
Occurrence and Control
Cost of Control. . . .
..........
.....
. . . .
........
}YrnUFACTURE OF ASBESTOS-CONTAINING PRODUCTS
A.
B.
.......
Nature of the Products and Processes
Economic Analysis. . . . . . . . . . .
.....
.......
SPRAYED ASBESTOS FIREPROOFING AND INSULATION.
Nature of the Product and Process
Economic Analysis . . . . . . . .
.......
.......
Beryllium
........
..........
.....
111
Page
ii
v
vii

1-1

1-1
1-1
1-1

1-2

1-2
1-3
1-4

1-4

1-4
1-5
1-7

2-1

2-1

2-5

2-7

2-11

2-11
2-13
2-14
2-20

2-24

2-24
2-24

2-26

2-26
2-33

2-37

2-37
2-37
3-1

-------
I.
II.

III.
Chapter 4:
I.
II.
III.
IV.
TABLE OF CONTENTS (Continued)
INTRODUCTION
............
.....
......
INDUSTRY PROFILE
TECHNICAL ANALYSIS . ~ . ~
.....
.......
. . . .
..........
A.
B.
Nature of Product and. Processes .. . . . . . .
Emission Controls. . . . . . . . . . . . . . . . . .
Mercury
INTRODUCTION. . . . . . .
PRIMARY MERCURY EXTRACTION
. . . .
. . . . .
. . . . I
......
. . . .
A.
B.

C.
D.
E.
F.
Occurrence and Recovery Techniques.
Emis s ions. . . . . '". . . . . . . . . . . . . . . . .
Emission Standard. . . . . . . . . . . . . . . . . .
Emission Control Techniques and Resulting Control
Costs of Control. . . . . . . . . . . . . . . . . . .
Economic Impact. . . . . . . . . . . . . . . .
SECONDARY MERCURY RECOVERY
......
.......
A.
B.
C.
D.
E.
Mercury Recovered and Methods of Recovery
Bot tling . . . . . . . . . . . . . . . . . . . .
Costs of Control. . . . . . . . . . . . . . . .
Economic Impact. . . . . . . . . . . . . . . . . . .
Impact of Control Costs. . . . . . . . . . . .
Impact on Prices. . . . . . . . . .
F.
MERCURY-CELL CHLOR-ALKALI PRODUCTION
. . .
.....
A.
B.
C.
D.
E.
F.
G.
In trodu c tion . . . . . . . . . . . . . . . . . . . . .
Mercury Emis s ions. . . . . . . . . . . . . . . . . .
Con t ro 18 . . . . . . . . . . . . . . . . .
Preventive Controls. . . . . . . . . . . . . .
Emission Standards. . . . . . . . . . . .
Costs of Control. . . . . . . . . . . . .
Economic Impact. . . . . . . . . . . . . . . .
iv
Page
3-1
3-1
3-3
3-3
3-12
4-1

4-4

4-4
4-5
4-7
4-7
4-9
4-10

4-19

4-19
4-25
4-27
4-27
4-28
4-29

4-30

4-30
.4-30
4-33 '
4-37
4-39
4-40
4-41

-------
~
1-1
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
4-1
. . . . . 2-25
. . . . . 2-26
. . . . . 2-34
  . . . 2-36
LIST OF TABLES
Estimated Emission Levels and Control Costs for
Asbestos, Beryllium and Mercury. . . . . . . .
.....
Chemical, Mineralogical, and Physical Properties
of the Different Varieties of Asbestos Fibers. .
. . . .
Costs of Control and Resulting Emission Controls
for Asbestos Emission Sources (Nationwide) (1970 Base)
Emissions and Control Costs Associated with Asbestos
Milling (for Plants Existing in 1970;1970 Dollars)
Unit Costs for Asbestos Milling Emission Control
. . . .
Cost Estimates for Controlling Emissions from the
Asbestos Milling Industry, 1970 and 1977 . . . . .
Cost of Bags for Shipping Asbestos Fiber, 1970
U. S. Consumption of Asbestos, 1968 . . .
Asbestos Products Industry Control Costs, 1970
Plant Value of Product and Investment of the
Asbestos Products Industry. . . . . . . . .
Beryllium Production by the Primary Producers, 1970
Air Cleaner CostsD:
Sulfate Process, Ore to Be(OH)2

Fluoride Process, Ore to Be(OH)2 . .
Air Cleaner CostsD:
Air Cleaner Costs#: Air Cleaner Costs#:
Ore to Be(OH)2 .. . . . . . . . . . . .
Bertrandite
........
First Company Air Cleaner Costs#:
to Be Billets. . . . . . . . . . .
Be(OH)2
. . . .
.......
Second Company Air Cleaner Costs#:
Be(OH)2 to Be Billets. . . . . . . . . . . . .

Air Cleaner Costs#: Be Billets to Metal Forms
.....
.....
Air Cleaner Costs#:
Be(OH)2 to Alloys. . . . .
Be(OH)2 to BeO and Ceramics
. . . .
. .. . .
Air Cleaner Costs#:
Control Costs for Mercury Extraction Model Plant
. . . .
v
Page
1-6
2-2
2-10
2-18
2-19
2-19
3-2
3-15
3-17
3-19
3-21
3-23
3-25
3-27
3-28
4-9

-------
Tab 1e
4-2
4-3
4-4
4-5
4-6
4-7
4-8
4-9
4-10
4-11
LIST OF TABLES (Continued)
The Mercury Potential of the United States
at Selected Price Levels. . . . . . . . . .
Mercury Consumed in the United States, 1970 .
. . .
Average New York Price of Mercury. . . . .
. . . .
Estimated Emission Control Costs for the Secondary
Mercury Industry, 1970 ...... . . . . . . . .
. . .
Mercury Consumed in the United States in 1970 (Flasks). .
Estimate Costs for Controlling Emissions from the
Mercury Cell Ch1or-A1ka1iClndustry, 1970 and 1977
. . . .
1970 Volume and Value of Chlorine and Sodium Hydroxide
Produced by Mercury Cell and all Processes in the U.S.
U.s. Production of Chlorine:
1941-1970 .
.....
Chlorine - Price History. . .
......
.....
Sodium Hydroxide - Price History (cents per Pound)
. . .
vi
Page
4-11
4-13
4-16
4-27
4-29
4-40
4-43
4-43
4-44
4-45

-------
Figure
3-1
3-2
3-3
3-4
3-5
3-6
3-7
LIST OF FIGURES
Sulfate Process for Conversion of. Beryl are to
Plant-Grade Beryllium Hydroxide.. . . . . . .
.......
Fluoride Process for Conversion of Beryl are to
Plant-Grade Beryllium Hydroxide. . . . . . . . . . . . . .
Hypothetical Plant Process: Organophosphate Extraction
Method for Conversion of Bertrandite are to Beryllium

Hydroxide. . . . . . . . . . . . . . . . . . . . . . . . .
Conversion of Beryllium Hydroxide to Beryllium
Metal Billets. . . . . . ,. . . . . . . . . . . . . . . . .
Conversion of Beryllium Billets to Beryllium
Metal Forms. . . . . . . . . . . . . . . .
........
Conversion of Plant-Grade Beryllium Hydroxide to Alloys
Conversion of Beryllium Hydroxide to Beryllium oxide
Powder and Ceramics. . . . . . . . . . . . .
. . . .
vii
Page
3-14
3-16
3-18
3-20
3-22
3-24
3-26

-------
CHAPTER I
OVERVIEW
I.
Introduction
Under section 112 of the Clean Air Act, as amended, the
Administrator of the Environmental Protection Agency has designated
asbestos, beryllium and mercury as hazardous air pollutants.
National emission standards have been proposed for certain source
categories known to emit these pollutants. Section 3l2a of this
act requires annual estimates of the cost of efforts for controlling
air pollution.
This report provides estimates of , the costs and economic
impact of implementing air pollution control measures applied
to selected asbestos, beryllium and mercury emission sources.
A. Health Importance
Hazardous air pollutants may cause or contribute to an increase
in mortality or an increase in serious irreversible, or incapacitating
. reversible, illness. Asbestos, beryllium, and mercury are very
different in the number and type of sources, their effects on health
and the control options available.
The primary danger from asbestos is inhalation of the fibers into the
lungs. A high incidence of lung cancer, pleural or peritoneal mesothelioma,
and asbestosis are all strongly correlated with e~posure to atmospheric
asbestos. Asbestos is unique among the hazardous substances, however, in
that it does not have an acute toxicity. Beryllium is highly toxic in
all forms (except possibly beryl) causing a serious chronic lung disorder
in susceptible persons. The poisoning action of mercury is cumulative or
chronic similar to lead poisoning. Large doses of the metal or small doses
of some of its more toxic compounds, however, can produce acute poisoning.
B. Economic Importance
Asbestos, beryllium, and mercury have properties that make them
uniquely suitable for special applications in a modern industrial society.
1-1

-------
The properties making asbestos so useful are its long, extremely fine
and flexible fibers, which are thermally, electrically, and chemically
inert, high in tensile strength and have favorable frictional properties.
Asbestos is used in over three thousand products ranging from heat-resis-
tant textiles to vehicle brake linings. It is used as a filler in plastics,
for filters in many industrial applications, in insulation, and to strengthen
many products as in asbestos-cement pipe ~nd building board.
. Beryllium metal has exceptional strength and rigidity with a high
strength-to-weight ratio and important thermal and nuclear properties.
Beryllium is widely used to harden copper and aluminum much as carbon is
used to harden iron into steel.
Mercury is the only metal which is liquid at most temperatures. It
has a uniform volumetric expansion coefficient as a liquid and it is an
electrical conductor. Although it is relatively inert chemically, it will
amalgamate with nearly all other metals (iron being a useful exception).
Mercury is widely used, but the leading uses have been the electrolytic
preparation of chlorine and caustic soda, in electrical apparatus, in
paint for mildew proofing, and in industrial and control instruments.
II.
Study Methods
A. Control Standards
This study was started before control standards were propo.sed for
the hazardous pollutants. Control standards were therefore assumed
based on the best information available.
For asbestos, the control assumed for this study was to pass all
asbestos containing exhaust air from a plant or manufacturing facility
through an adequately maintained fabric filter of appropriate design.
The assumed control is somewhat more comprehensive than that subsequently
proposed under which more latitude is allowed in the choice of filter, ~d
some processes involving bound asbestos may be excluded from compliance
with dust collection device requirements.
Under the assumed beryllium standard, total emissions shall not
exceed 10 grams of beryllium in a 24 hour day, or outplant concentrations
shall not exceed 0.01 micrograms of beryllium per cubic meter of air
averaged over a 30 day period. The standard assumed for beryllium was
1-2

-------
the same as that used by the Department of Defense and the Atomic Energy
Commission for many years. This standard is the one later proposed for
beryllium.
For mercury the emission standards assumed in this study are as
follows:
Ore processing: Emissions are not to exceed 0.95 g/metric ton (0.0021
1b/ton) of ore processed and no more than 1156 g/day (2.56 1b/day) for any
one plant.
Secondary processing: Emissions are not to exceed the new industrial
hygiene TLV (Threshold Limit Value) of 2.85~5.5 mg/min. for anyone plant.
Mercury cell ch1or-alka1i plants: 'Emissions are not to exceed 0.005
1b of mercury per ton of chlorine produced and no more than 2.56 ~bs per
day from anyone plant.
The emission standard adopted for ore processing and facilities using
mercury ch1or-a1ka1i cells is that emissions to the atmosphere from sources
subject to controls not exceed 2,300 grams of mercury per 24 hour period
(5.0 pounds per 24-hour period). No standard was adopted for secondary
mercury processors. The standard adopted is significantly less stringent
than the standards assumed for this report. Costs and economic impact
significantly below those reported would appear, therefore, to be
appropriate.
B. Cost Estimation and Economic-Impact
In general, control costs were estimated by calculating the expenditure
required to increase the levels of emission control from an assumed baseline
level to the level required for compliance with the selected emission
control regulations. This approach is based on the premise that the costs
properly attributable to the implementation of the Clean Air Act, are those
costs incurred in reaching control levels not commonly being achieved at
the time the~e industries came under amendments to the Clean Air Act.
The number and type of installations currently controlling pollutants,
the level of efficiency each 1's achieving, their capacity, and other
characteristics are the bases for determining the extent and types of
control methods needed to meet the selected standards. This information
was derived from published data, trade association reports, and inter-
1-3

-------
views with industry and Environmental Protection Agency contacts.
Assumptions, including the set of control regulations, are identified in
this report. Obviously the results in terms of emissions and control costs
that are tabulated, depend on the underlying assumptions. If the assump-
tions hold, then the results as estimated will follow. In this report, 1970
was used as the base line year and coverage is for the entire United States.
Control costs are estimated in terms of the initial investment required to
I
establish control and the continuing annual expenses related to that
investment. The investment cost is the total expense of purchasing and
installing control equipment. The annual cost is the ultimate yearly charge
for capital-related costs (interest on the investment funds, property taxes
where applicable, ,insurance premiums, and depreciation charges) plus
operating (labor, utilities, and supplies) and maintenance costs. In
addition, this report presents estimates of the impact of industry costs both
on the industries themselves and on the consumers of the products produced.
C. Report Organization
This report is organized with this, the OVERVIEW chapter, followed in
order by chapters on ASBESTOS, BERYLLIUM, and MERCURY. The overview
includes an introduction to and summary of the report. The summary follows,
in general, the same format as the report. Asbestos mining and milling,
asbestos handling, manufacturing asbestos-containing products, and
sprayed asbestos-cement insulation are discussed in Chapter 2. Beryllium
extraction and processing are discussed in Chapter 3. Mercury mining
and smelting, secondary mercury, and mercury cell chlor-alkali production
are discussed in Chapter 4. Major conclusions of the report can be found
in the summary. Further and more detailed tabulations and narratives are
found in the body of the report.
III.
Summary
A.
Asbestos
Asbestos emissions would be reduced to 409 tons under the proposed
control measures from an estimated 6,382 tons in 1977 without control.
The total investment required for plants in place in 1970 would be $5.7
million. By 1977 an additional investment of $2.3 million would be
1-4

-------
required. The annualized cost would increase from approximately $1.7
million in 1970 to approximately $2.3 in 1977. As shown in Table 1-1,
the major emission reduction would be by the asbestos milling industry.
However, the highest cost would be incurred to. reduce emissions from
asbestos cement products and asbestos textile production.
Costs as indicated by this analysis do not appear to be large
enough to have a major impact on the asbestos processing industries
as a whole. Only in one case, asbestos textiles, are cost increases
per unit of production more than one percent of sales price. The
cost increase for asbestos textiles amounts to about 5 percent of sales
price which is large enough to adversely affect the industry. Adverse
effects are expected to be small, however, because of the specialized
uses for asbestos textile.
Costs among plants are quite variable, and individual plants or
companies may be adversely affected. It appears that such effects will
be minimal.
B. Beryllium
The Atomic Energy Commission and the Department of Defense recognized
the toxic nature of beryllium over 20 years ago. As the major purchasers
of beryllium and beryllium products, they issued standards for permissible
concentrations within the plants and surrounding communities of their
suppliers. EPA proposes to adopt these same standards. As a result of
these purchaser-enforced standards, emissions are adequately controlled by
the major producers of beryllium, beryllium alloys, and beryllium ceramics.
Therefore, no net costs are expected for controlling beryllium emissions,
and no economic impact analysis is necessary.
Estimates of beryllium emissions are given in Table 1-1 for metal
processing plants including primary producers, producers of beryllium
alloys, and producers of beryllium ceramics. Emissions are estimated to be
about 40 pounds per year for 1970 and are expected to increase to about 70
pounds per year in 1977 because of growth in the industry. No estimate of
uncontrolled emissions is made for 1977 since emissions are currently
Controlled.
1-5

-------
Table 1-1.
ESTIMATED EMISSION LEVELS AND CONTROL COSTS FOR ASBESTOS, BERYLLIUM AND MERCURY
Source
Asbestos
Milling
Products
Asbestos Cement
Floor Tile
Friction Material
Asbestos Paper
Asbastos Textiles
Sprayed Insulation
Sub-Totals
Bery111am
Metal Processing
Alloys
Ceramics
Sub-Totals
Mercury
Roasting
Secondary
Chlor-Alkali
Sub-Totals
1/
Emissions (Tons)
1970 1977
without with
control control
3,860
206
101
314
15
20
15
4,531
0.00562
0.00396
0.011
0.02058
50
0.03
70

120.03
5,440
290

142
441

21
28
20
6,382
II
"
If.
"
50
0.04
105
155.04
218
. . 58

28
88
2
15
o
40~
0.0124
0.0072
0.0163
0.0359
.5
0.004
6.7
7.304
Investment above that for 1970.
Control Costs ($1,000)
197~nvestmen~9771/ 197~nnua1iZ~~77~/
$
342
$ 139
$2,331
1/
"
"
"
o
.225
o
0.225
1/ Total for effects of the Clean Air Act and amendments.
1/
2,400
216
720
348
1,700
o
$5,726
]/
"
"
"
1,033
5.6

14,725

15,763.6
$
128
$
180
977
88
293
142

692
o
720
65
216
104
510
o
1,013.
91
304
146
718
o
$2,252
1/
"
"
"
586.5

11. 37
14,900

15,497.87
Current controls are adequate to meet the standards proposed, therefore,
emissions without controls are not estimated for 1977 and additional
investment and annualized costs are not required.
1-6
$1,743
1/
"
"
II
586.5
8.64

17,344
17,939.14

-------
c. Mercury
The sources of mercury emissions to the atmosphere which have been
analyzed for this project are primary mercury production plants (smelters
associated with mines), secondary mercury recovery and refining plants, and
plants using mercury chlor-alkali cells. As shown in Table 1-1, the 1970
total atmospheric emissions of mercury from these sources are estimated to
be. about 50 tons, 60 pounds, and 70 tons, respectively. By 1977 without
controls, emissions are estimated to remain the same for primary mercury
production, to increase to 80 pounds for secondary mercury recovery, and
increase to 105 tons for chlor-alkali production using mercury cells.
Controls would reduce these to 0.5 tons, 8 pounds, and 6.7 tons respectively.
The total investment required for plants in place in 1970 is estimated
to be 15.8 million. By 1977 an additional investment of $0.2 million
would be required. The annualized costs are estimated to be $17.9 million
dollars in 1970 and decrease to $15.5 in 1977. Costs as indicated by this
analysis are quite variable in their impact on the different sectors of the
industry.
Primary mercury production will be .very adversely affected by any costs
imposed to control emissions. Product prices are already depressed below
production costs and many primary producers are shutting down as a result.
Factors depressing prices are increased world production, congressional
hearings on the effects of mercury on man and the environment, cancellations
of mercury biocide registrations, reduced consumption for ch1or-alka1i
production and reduced consumption resulting from the general economic
recession. No attempt was made to analyze the economic impact of pollution
controls on mercury prices.
Estimated costs to control emissions for secondary mercury recovery
and refining plants are minimal. Little or no adverse economic impact is
expected.
1-7

-------
While the portion of the chlor-alkali industry using mercury
cells for production will have to invest almost $15 million to
control emissions, the economic impact will not be large for the
industry. Companies producing chlor-alkali are financially strong,
and enjoy a market that has been increasing almost 7 percent a
I
year. In addition, chlorine or caustic is only a minor cost in
most products where they are used and substitutes are not readily
available. Three-fourths of the companies making chlor-alkali operate
both type of cells. Only one firm uses mercury cells exclusively and
depends on chlor-alkali for the bulk of its sales.
The biggest competitive impa~t will be between the mercury
cell process and the diaphragm cell process. In the past the
mercury cell process appears to have had the competitive edge. It
now appears that emission control costs will cause the diaphragm
cell process to be economically more efficient.
Prices are expected to increase, but not by the full amount
of the cost increase. Increases are estimated to be about $0.55
to $0.65 per ton which is approximately two-thirds of the unit cost of
emission control for the industry.
1-8

-------
CHAPTER 2
ASBESTOS
I.
Introduction
Asbestos is a class of minerals which is extremely useful in a
form which requires only mechanical processing. Chemically, the entire
class is characterized as a naturally occurring fibrous silicate.
Within this class are two subclasses: (1) Serpentine, or chrysotile,
and (2) Amphiboles, including crocidolite, amosite, anthophyllite,
tremolite, and actinolite. The properties of the asbestos are
described in Table 2~1.
Commercially, chrysotile is by far the most important and most
valuable of the asbestoses, especially in "spinning" grades of fibers
(3/4" or longer). The properties which make asbestos useful are its
long, extremely fine and flexible fibers, which are thermally,
electrically, and chemically inert, of high tensile strength and of
extremely favorable frictional properties.
The asbestos air pollution problem is related to the fact that
asbestos fibers are a health hazard. This has been long recognized.
Fibers may travel significant distances because of their fine
structure and low density. Fibers are not destroyed by any known
environmental process. The primary danger from asbestos is inhalation
of the fibers into the lungs. The British Occupational Hygiene
Society has found that fibers 200~ or less in length and 3.5~ or less
in diameter are respirable- [Reference 1]. Most asbestos fiber is less
than O.5~ in diameter. Virtually all asbestos fiber is therefore
respirable [Reference 2].
It has been shown that high incidences of lung cancer, pleural
a/ d
or peritoneal mesothelioma, and asbestosis- are all strongly correlate
with exposure to atmospheric asbestos. All these diseases may be fatal
a/ Asbestosis is an asbestos-induced disease closely related to coal
miners' black lung and cotton workers' brown lung; pleural and peritoneal
mesotheliomas are tumor-like hardenings of the rib cage and abdominal
linings, respectively. In the case.of lung cancer, it has been shown
that asbestos exposure and smoking are synergistic by a factor of 8.
(Selekof et aI, 1968)
2-1

-------
TABLE 2-1.
CHEMICAL, MINERALOGICAL, AND PHYSICAL PROPERTIES
OF THE DIFFERENT VARIETIES OF ASBESTOS FIBERS
--------.-
--...,---------------
l.e:o,iite
V~.ril:\.}'
pr'c,pol.ty\
---------
a,,'rnic.,l
forr.\ula
Chry sot 11"
Cl'OC i<101 i La
. j,r. ~'.11c;I'1'.rl1 i to
'1'l'c"".1 ito
;.ctine.lite
4.____.-.-
:m~025i0221!20 N
to 5.%
Sllicutc of
and Fe with
some water
Na Silicate ot Fe
. ~nd ~~7 higher
iron thBn IIn-
thophyllito
~~ silicate
with iroD
percent.age
ct.ei~ici\l
(1);~,po3i-
t10n

5102' ~~

~:g0. %
FeO. "

;"2°3' "
A1203' "
li2')' %
CaO, ~~

Nil 2')' %
CaO +
N1I20. %
37 -44 49 -53 49 -53 56 -58 51 -62 
39 -44 0 - 3 1 - 7 28 -34 0 -30 
0.0- 6.0 13 -20 3' -44 3 -12 1.5- 5.0 
0.1- 5.0 17 -20       
0.2- 1.5   2 - 9 0.5- 1.5 1.0- 4.0 
12.0-15.0 2.5- 4.5 2 - 5 1.0- 6.0 0 - 5.0 
Tr. - 5.0       0 -18 
  4.0.. 8.5  ---   0 - 9 
    0.5- 2.5     
9.2-9.8     tleutro1   
Poor Good Good Very goed Good Fair
plf
::.:=i:~~~,,:o
t.o acids
IInd alka-
lies
Ionizable
nlts, ",1-
cromhos
(relative
electrical
cor.duct-
anco)
1.82
0.84
 1.34 0.58    
0 0 0   
 Iron Iron  Lirne L.ime. iron
Pri~matic. I'rhan.\tic. . Long IIntl thin (,cng IInd thin
lar..ellar to 11\:r,011l1r to col\I!I'n:!r to columnar to
U!JI:OUS fibrou s fibL'ou~ fibrous
Nonocl1nic OrthOl"hoilbic Monoclinic ~!onoclinic
Lan:011ar. LiI/!.ollar, Lonq. p1'i!Zlno1- Reticulated
CO<1rSI) to fibrous tic and ft- lol1q pl"istMtlc
lino Cibrous a!Zbo~t1- brous 1I<:J9re- crysti1ls i1nc1
:,nd ubast1- form c;jates  fibors 
tOl:m      
In cryst:l1- tn crystal- In ~~7 lir::o- In li:r.ostone
line schhts, lino !Zr.hi;ts stonc:;J:ls 111- and cl:"yst.al1 ine
etc.  IInd gneisses torat ion pL'od- schists
   uct of rnu<;;nc-  
   8illn rocks.  
   metamorphic  
   IInd i'Jnt'ous  
   rock s   
lugnetite
.:'ontent
o
-5.2
Xin~ral Iron. chro~e
i:r~uritias nickel. lime
prc3ent

Cry.tal Fibrous and
structure IIsbestiform
3.0-5.9
Xr.on
Fibrous
Crystal
!ystem
Monoclinic and Monoclinic
orthorhOIT.bic
~lineara- In veins of
l~ical s~rpentino,
structure etc.
Fib1-0IJ3 in
iron stones
Mineral
associ-
ation
In altered
poridotite
IIdJac~nt to
serpentine
IInd lir:'\cstone
. nc.!r contact
with basic
1qneous rodt.
Iron rich
sil1cious
erg illite
in quert:o:030
2-2

-------
TABLE 2-1 (Continued).
,
V"l" 1,,\"'y
Prc.pi~rt.y
--~._-'---
V"ininr CMn
1Jar~h .md
~'()''l1':<3' but
50nll,;.:'",11~st
pliable
1>J)lhq')1y11ite
...--.........._..__0...-_-
Slip or
m"~s,", {iber
:311PI m.:t5S
f.ihc:r unor i-
(.nl: "r:j "nd
111\'.or1t1cing
1'.ct inolil:e
.:...--
Gri.li'-whltc, '. Grect!i'sh
green j sh, yel- . " . .
lo-.~i~h,
1.>lut>,h
Ycl1r)'.d~h
bru...n, ~ray-
i511 "I11ite
-'-'-."-..--
IIljn:1l1 coft
to friable
-----..-.----
Vil:rcou5,
5 om(:\~'ha t
pca1'ly
Vit.r.COllS 1:0
pOClt'ly
4
.-------"---'--
5.5-6.0
5.5
6;t
5.5-6.0
High
------.--.-.-----..
'Poor
Very good
Good
Fair
Good
Fair
80,000- 100,000-
100,0001 300,0001
024,000 (r.'.a:<.) 8'/6,000 (ma:<.)
2770
0.266
Positive
510"'"
2.4 -2.6
010 pedect
_____4_-
Bii3xia1 1"o"i-
tive, extinc-
tion parallel
2100
0.201
negative
Fast
3.2 -3,3
110 porfe:ct
---
l3i,1xia1 +
IOxtinetio:1
inclined
16,000-
90,000
2550
0.193
Negative
:;tlk~'
Sl1KY
Poor
Poor
Poor
Poor
Po~r

1,000
and leas
4,000
c11d 1055
1,000-
0,000
"t',/5
2540'
;:.~oo
0.210
0.212
0.217
Negative
lIoge.tive
Fast
3.1 - 3.25
110 pel' {"ct '
Oia:<1a1 po~i-
tivc, extinc-
tion perall.,1
"---..---
NC'dium
Nedium
Negative
Medium
2.85-3.1
110 [-cdect
2.9 -3,2
110 perfect
O!."xia1 po,;.i-
tivlO, extinc-
t!e.. r31"a11el
3.0 -3.2
110 perfect
-------
fJ t".xi.:\l ncga-
tlv", c..
-------
once they become established. Asbestosis is not normally fatal if the
exposure to asbestos is eliminated before extensive lung fibrosis develops.
Asbestos is unique among the hazardous substances, however, in that it does
not have an acute ~oxicity. Asbestosis, in particular, is associated with
long-term, primariiy high-level, exposures to asbestos fibers. However,
there is no known exposure threshold below which there is zero probability
of one or another of these diseases. A recent study of lung tissue from
randomly selected autopsies of the general public indicates that asbestos
bodies (enclosed asbestos fibers) maybe found in the lungs of one-fourth
to one-half the entire U.S. population [Reference 4].
Such w~despread occurrence of asbestos fibers in human lungs very
likely'results from the varied uses for asbestos. Asbestos is used in
over three thousand industrial and commercial products ranging from special-
purpose textiles to vehicle brake linings. It is used as a filler in
plastics, for filters in many industrial applications, and an essential
ingredient in asbestos-cement pipe and sprayed asbestos-cement insulation.
Free asbestos is largely handled by air convergence during
processing because of the need to protect the easily-broken fibers
and its low density. Because asbestos easily becomes airborne, the
air conveying system must be tightly controlled to recover the
asbestos and protect the workers. However, because asbestos always
contains large numbers of small, respirable fibers, this can be an
exacting task in itself. Once the asbestos is mixed with a liquid
medium (usually a binder or filler) there is essentially no further
problem with emissions until the finishing process. If cutting,
breaking, grinding, or polishing is required during the finishing
process on an asbestos-containing material, there are usually
dusty emissions ~resent. There is no general agreement as to
whether free. asbestos is released in any particular case, and if
not, whether there is any exposure potential involved. It will be
assumed in the following discussion that these emissions must be
controlled as though they are free asbestos emissions.
The next section of this chapter discusses the proposed emission
standard, followed by sections on asbestos emission control techniques,
the asbestos milling industry, asbestos processing industries, control
costs, and economic impact of controls.
2-4

-------
II.
Control Standard
The asbestos control standard examined in this study is that all
exhaust air from a plant or manufacturing facility which contains asbestos
fibers must be passed through a fabric filter of appropriate design and
adequate maintenance. There are several reasons for choice of this
standard, among which are the following:
1) Fabric filtration is the most effective and by far the
most cost-effective method of asbestos emission control.
Fabric filtration is virtually the only high-efficiency
control technique currently used by asbestos-using industries.
An emission-type standard is not possible due to the current
lack of adequate emission monitoring techniques and equipment.
The control standard examined in this study differs from that
described in a communication from the Hazardous Pollutants Branch,
Environmental Protection Agency, July 3, 1972, which is as follows:
1) The national emission standard for asbestos:
2)
3)
a)
requires that emissions, in forced gas streams or as
local visible emissions which are not presently ventilated,
of particulate matter resulting from the milling of asbestos
ore or the manufacture of products which contain commercial
asbestos shall not exceed the amounts which would be emitted
if these gas streams or local visible emissions were treated
by a fabric filter or a wet collector which is adequately
(specified) designed, operated, and maintained;
allows certain processes (such as the printing of asbestos-
paper-based flooring products), which involve bound asbestos
and which do not release particulate matter which contains
commercial asbestos, to be excluded from compliance with
b)
c)
dust collection device requirements;
prohibits the spraying, on a building or in an open area,
of any product which contains asbestos (This is the prohi-
bition as proposed December 7, 1971. The final version of
the standard may allow the spraying of some products which
contain small amounts of asbestos and/or products which
contain bound asbestos);
2-5

-------
d)
prohibits visible emissions of particulate matter from
external conveyors for asbestos-containing materials at
asbestos mills;
prohibits visible emissions of particulate matter from
any manufacturing process (of a product containing
commercial asbest08) which is in an open area and which
is not regulated by a.specific collection device
e)
f)
requirement;
prohibits the surfacing or resurfacing of any roadway
with asbestos tailings except at asbestos mines and at
asbestos mills which are located on property contiguous
g)
to asbestos mines;
prohibits the deposition of asbestos ore or asbestos-
containing tailings into containers or vehicles intended
for "shipment of such materials on public roadways except
in containers or vehicles which, when subsequently sealed,
covered, or otherwise enclosed, prevent visible particulate
emissions derived from asbestos ore or asbestos-containing
tailings.
2-6

-------
III.
Control Techniques
As mentioned in the previous section, the only acceptable final
control technique for asbestos from manufacturing process emissions
in the proposed standard is use of a fabric filter. This will be the
only technique discussed in detail. However, other techniques which
have been used and are still being used for various reasons will be
mentioned.
Collection by dry cyclone devices is very commonly used. However,
they have low collection efficiency because of the fine size of many
asbestos fibers «1 micron diameter) coupled with their low specific
gravity (N2-3). Dry cyclone are used because the asbestos that is
collected, which may be a large fraction by weight of total process
emissions, is generally either marketable or usable in the process.
In fact, dry cyclones are essential process equipment in asbestos
milling using air aspiration. In terms of emission control, however,
dry cyclones are acceptable only as pre-cleaners to reduce fabric
filter loadings.
Wet collecting equipment, such as wet cyclones or venturi scrubbers,
are rarely used, since there are many problems associated with asbestos
slurries. Pumps, drains, etc. tend to become clogged. It is difficult
to reclaim reusable fibers from the water, a high removal efficiency
is expensive to achieve in the venturi scrubber type system. Wet cyclones
can capture asbestos fibers .for reuse, but clogging and blinding problems
generally arise in the baghouse that must follow the wet cyclone in order
to meet the standard, due to the humidity resulting from wet collection.
There are some applications for which wet collectors can provide collection
performance comparable to that of fabric filters and some applications for
whiCh fabric filters cannot be utilized.
Electrostatic precipitators have been tried as asbestos collectors,
but have proved to be inefficient collectors due to the resistivity
characteristics of most asbestoses.
Therefore, their expense is not
justified.
Fabric filters have been found to be nearly ideal for asbestos
emission control.
Some reasons are:
1)
Any asbestos fiber captured need not be further processed for reuse.
2-7

-------
4)
Once the fabric is coated with asbestos, the asbestos becomes
its own, nearly "absolute" filter.
Baghouses provide collection efficiency equal to or better
than any o,ther collection system.
\
Baghouses cost less to buy, maintain, and operate than any
system with comparable asbestos collection efficiency.
2)
3)
As a result, fabric filters are by far the most common type of
high-efficiency control system currently in use on asbestos emissions.
As an example of the efficiency of control of asbestos emissions by
baghouses, it may be noted that cleaned air from these systems using
cotton sateen bags are frequently recirculated into work areas,
especially where accurate humidity and temperature control are
necessary, or where savings in heating and cooling expenditures may
be significant. It thus appears that baghouse emission control can
3
meet or exceed the industrial hygiene TLV of 5 fib.er/cm , where "fiber"
is defined as an asbestos particle 5.0 ~m or greater in length and with a
3:1 length to diameter ratio or 1arge~ (Reference 5]. This concentration
of fibers corresponds roughly to 5 million particles per cubic foot
,
(5 mppcf), when total particles are counted rather than fibers only.
Baghouse filters have both advantages and disadvantages as compared
with other methods of collection. The prime advantages are:
1) It is a high-efficiency dry system, SO that water treatment
and water handling are not required.
2)
Larger systems, in particular may be sectionalized, so that
maintenance may be done on one section while the others
continue to operate.
collected materials may be returned directly to the process,
unless they are too fine, when they may be used in some other
application.
3)
4)
The baghouses will control other dry particulate emissions
with equal facility if required.
Some of the principal disadvantages are:
1)
2)
Large installation area per unit gas flow.
Minor bag damage can cause significant loss in cleaning efficiency.
2-8

-------
3) Cleaning efficiency is variable, being lowest just after
bags are cleaned or new bags are installed.
Bag changing is expensive, both in labor and materials.
Baghouse temperature must be kept above the dewpoint but
below bag tolerance limit, sometimes a delicate procedure.
It is clear from the wide use currently made of baghouses
in controlling asbestos emissions that the positive factors
significantly outweight the negative factors. Therefore, costs
and emission calculated for this report, as shown in Table 2-2,
are based on this method of control, except as noted.
4)
5)
2-9

-------
TABLE 2-2.
COSTS OF CONTROL AND RESULTING EMISSION CONTROLS
FOR ASBESTOS EMISSION SOURCES (NATIONWIDE)
(1970 Base) .
     PRODUCTION  EMISSIONS (Tons)    CONTROL COSTS ($1.000)  
    No. of (Tons of Fiber) 1970  1977   Investment 1/  Annualized 21
  Source Plants 1970 1977 wlo Control \1/0 Control With Control  1970 1977-  1970 1977-
 Milling  9 125.314 176.300 3.860 . 5.440  218 $ 342 $ 139 $ 128 $ 180
 Products                
 Asbestos Cement 48 412.500 580.400 206 290  58  2.400  977  720  1.013
 Floor Tile 18 201.200 283.100 101 142  28  216  88  65  91
N Friction Material 30 104.600 147.200 314 441  88  720  293  216  304
I             
..... Asbestos Paper 29 30.200 42.500 15 21  2  348  142  104  146
o     
 Asbestos Textiles 34 18.100 25.500 20 28  15  1.700  692  510  718
 Sprayed Insulation * 2.000 0 15 20  0  0  0  0  0
  Totals 168 893.914 1.255.000 4.531 6.382  409 $5.726 .$2.331 $1.743 $2.252
 1/ Invest~ent above that for 1970             
 1/ Total for effects of the Clean Air Act Amendcents           

-------
IV.
Asbestos l-lining and Milling
A.
Occurrence and Mining Techniques
Chrysotile (serpentine asbestos), which accounts for over 95% of
U.S. production, occurs in three types of formations: cross fiber,
slip fiber, and loose. Cross fiber chrysotile is that in which the
fibers span gaps (veins) in the surrounding serpentine rock-formations.
Slip fiber occupies similar gaps in the formations, but the fibers
lie parallel to the walls of the vein. Loose fibers, which occur at
only one mine site in the U.S., are "pre-milled", occurring in loose
formations near ground level mixed with various-sized aggregate. Cross
fiber chrysotile is commercially most valuable, since the highly prized
spinning length fibers 0/4" or longer) are most commonly found in such
formations. (However, the longest known fiber bundle ever mined was from a
Chinese slip fiber formation, and had a maximum fiber length of some 3 1/2 ft).
Cross fiber and slip fiber may be mined by surface (open cast or open
pit) techniques or by underground methods, while loose fiber is mined
by what is essentially an open cast process.
Open cast mining involves removal of the ore by earth-moving
equipment from shallow deposits, in one instance in the U.S. without
the need for blasting. Generally, a shallow overburden with low concen-
trations of asbestos fibers must be removed. There will be emissions
of asbestos fiber from the overburden dumps and exposed ores through
weathering, and in concentrated amounts from drilling, blasting,
overburden and ore removal, loading, and transport.
Open pit mining is similar to open cast operations except that the
workings are much deeper to follow the fiber veins. Blasting and ore
removal occur primarily on the sides of the pit along terraces which
spiral down around the sides of the pit toward the bottom. Sources of
emissions are the same as for. open cast mining except that overburden
removal will be proportionately smaller.
Underground mining of asbestos involves following the veins of ore
with shafts, galleries, and drifts, using blasting and earth moving.
However, there is no overburden removal, ore veins are not exposed to
weathering, and many dusty operations take place wlderground. Therefore,
2-11

-------
emissions from this type of asbestos mining are much lower than for
surface mining techniques. There will be significant emissions from
surface ore transfer and transportation and hand cobbing of ore.
A new method of; mining which has particular. significance in
control of emissions from open pit mining is the block caving
technique, which significantly reduces the required blasting and
eliminates the need for overburden removal. When the volume of
rock to be mined has been determined, that volume is undercut,
leaving solid support pillars to hold up the main block. As the
block caves in down the "chimney", ore is removed from underneath and
mill and mine tailings are replaced on top to maintain the downward
pressure on the block. Replacement of the tailings also reduces
tailing dump emissions and space requirements. The block caving
technique also reduces direct mining emissions to a level
comparable to "normal" underground mining operations [Reference 11].
Transporting the ore from the mine to the mill generates emissions
which are generally grouped with those of mining rather than milling.
Such emissions arise in large measure from open trucks, which are
typically 20 to 75 tons capacity, although some 200-ton units are in
use. Private mine-mill roads are frequently paved with tailings,
which liberate fibers to the environment as the trucks pass. In this
context the use of 200-ton trucks has mixed significance for emission
levels during transport. The larger truck capacity should reduce
emissions from the transported ore, with their larger volume-to-
surface ratio, and it reduces the number of trips per unit of ore.
However, the road itself takes a bigger beating from each truck pass.
The relative significance of the two effects has not been determined.
Mining of asbestos is limited in this country to the four states of.
California, Vermont, Arizona, and North Carolina. A small amount of
anthophyllite is mined at two locations in North Carolina; all the other
mines produce varying grades and types of chrysoti1e.
The four firms in Arizona produce a special low-magnetite cross
fiber chrysotile commanding a premium for electrical applications.
These firms mine asbestos at several sites using underground mining
techniques.
2-12

-------
The four mines in California use the open cast method. One of
these mines is in ore area where geological action has broken up
the deposit, so that little or no blasting is required. This mine
has some of the richest ore in the world, being in places up to
sixty percent asbestos. (The U.S. average ore content is four
percent. )
The mine in Vermont is the oldest and largest asbestos mine
in the U.S. Both slip fiber and cross fiber chrysotile is mined
there using the open pit process. The Vermont deposit is part of
the large deposits identified as the Canadian belt in southern
and southwestern Quebec.
There is concern that the six Quebec mines may be sources of
asbestos emissions that carryover into the United States,
particularly into Vermont. Further consideration of these emissions,
however, is beyond the scope of this report.
B. Gontrol Techniques for Mining Operations
The standard asbestos control technique of baghouse collection has
limited application to mining processes. Portable baghouse systems
can be used with good effect during drilling operations prior to blasting,
since this is a localized emission source. However, this is a minor
source in terms of total emissions, and some sort of wet drilling
technique might be just as effective and much less costly.
Emissions from blasting are virtually impossible to collect once
they occur. Two techniques can be used to reduce generation of emissions
to a minimum. First is use of blasting techniques and calculations of
charge which produce the minimum breakage of rock required, so that
the rock is not "blasted allover the landscape". Second, a technique
developed in France, which uses plastic capsules of water to suppress
blast dust, appears to be quite effective (twenty to eighty percent
reduction in emissions) [Reference 5]. Plastic contamination of the ore
which could not be removed by present milling methods may, however, be a
problem in applying this technique to asbestos mining.
Control of dust generated during ore loading does not appear to be
feasible and is not known to be practiced in any asbestos mining
2-13

-------
operations. Dusts generated during transport may be controlled by
(1) wetting the ore surface, (2) covering the tru~k body with canvas or
a more rigid sealing cover, (3) wetting road surfaces where they are
covered with tailings. Wetting the ore surface in the trucks bed is
I
not known to be used. In hot, dry weather it would have to be done
frequently to be effective during long trips to the mill. Canvas
covers are currently being used at several sites. Wetting road surfaces
is used in some places. Constant wetting of the surface with water is
effective but causes slippery road surfaces and requires much labor
and water. Wetting cannot be used in freezing weather. Good results
have been obtained (in increasing order of effectiveness) with road
oil, a ten to twenty-five percent water solution of liquid sulfonate,
and emulsified asphalt. Where the ore body is utilized as a mine road
chemical dust suppressants may cause unacceptable ore contamination.
It is expected that the overall control level from careful appli-
cation of the best techniques would be about eighty percent of the
potential road dust emissions.
C. Asbestos Milling Processes
1.
Introduction
Once asbestos ore has been mined, the asbestos fibers must
be separated from it. When the asbestos has been separated,
it is graded on the basis of content of various lengths of
fibers. In general, for a particular source of chrysotile, the
larger percentage of longer length fibers in the final mix bring
the higher prices. The most expensive grade of fiber is called
"crude", which is not milled, but hand-separated ("cobbed") from
the surrotmding rock into btmdles of fiber with aggregate fiber
length of 3/4" or more. Crude (long) fiber is valued for
weaving of asbestos textiles, for which shorter grades are
not suitable.
In order to maintain maximum fiber length and promote maximum
recovery, it is desirable to hold mechanical working of fibers
to a minimum. Although asbestos fibers have very high tensile
strength per unit area and per unit weight, the individual fibers
are so fine that they are rather easily broken. Compounding
the recovery problem is that asbestos fibers have the same
2-14

-------
density and chemical composition as the surrounding rock. The
solution applied in all but one of the asbestos mills in the U.S.
is to use mechanical means to free the fibers from the rock.
but accomplish the actual removal from the ore by an air aspiration
system. In order to reduce fiber losses and industrial hygiene
problems. it has become common to convey the fibers by air also.
The result is a requirement for seven to ten tons of process
"air for every ton of fiber produced. or a volume ratio of about
1600 to one. The large volumes of air required. plus the
"floatability" of asbestos fibers. leads to significant potential
emissions.
The one exception to the air aspiration milling system in the
U.S. is found in the mill which processes the loose fiber ore
found in California. At this plant a proprietary wet separation
process which presumably involves some sort of floatation technique
is used to separate the fibers from the ore tailing. As a result.
both industrial hygiene problems and air pollution potential from
the fiber extraction process are significantly reduced.
The two milling processes and their corresponding emissions
and control methods are discussed in the following sections.
Milling by Air Separation
The incoming ore is first unloaded from the arriving truck where a
cloud of dust and fiber is generally produced. The incoming
coarse ore is then typically crushed by a jaw crusher to a size
2.
that depends upon the mill. Oversize rock is separated by
rotating cylindrical trammel" screens and crushed in a secondary
crusher. usually a cone type. The ore streams in most plants
are then conveyed to a dryer (a rotary kiln in larger
installations) where moisture in the ore (up to 30% by weight)
is removed. The dried ore is then stored, with large amounts
being held to allow for variations in fiber demand and mine
production over time. At least one company departs from this
procedure by storing the ore wet, thereby smoothing out drier
operations and reducing emissions from ore storage and handling.
This company further reduces emissions and "homogenizes" its
2-15

-------
product by stripping ore for further milling from the underside
of the storage pile via a moving conveyor systems housed in a tunnel
chamber underneath the storage area.
Drie~ ore is conveyed to an additional crushing step and then
through a series of milling, shaking, and aspirating steps. The
milling, done by either hammer mills (fiberizers) or crushers,
serves to separate the fibers from the rock and from each other.
The shaking is accomplished on progressively finer screens, where
small rocks and fiber bundles pass through for further treatment,
larger rocks are retained for conveying to tailing dumps or
further crushing, and the freed fibers are removed by air flow
through powerful suction hoods.
Separated fibers are caught in dry cyclones and conveyed
to grading screens. After grading, the fibers are sent to storage
bins by grade (length). The final operations is removal of fibers
from storage bins, blending in a mixer to produce the desired
final grade, and bagging for shipment.
Residual rock, which contains a small amount of unremoved
fiber and dust, is usually transported by some means to a tailing
dump. If the block caving technique is used at the associated
mine, the tailings may be returned to the mine to assist in the
caving.
Every one of the processes, described above leads to some loss
of fibers. to the air either inside or outside the plant. Emissions
to in-plant air are usually kept to a minimum consistent with
maintaining the industrial hygiene Threshold Limit Value. This
is most effectively accomplished by operating as many pieces of
equipment as, possible in a closed condition and under a slightly
negative pressure. This is done, however, at the expense of
increasing the plant ventilation load and greater potential
atmospheric ~missions.
The major 'emissions potential at asbestos mills arises from
the cyclone collector exhaust. The series of process cyclones normally
used are of a relatively large-diameter design to limit damage to the
fibers as much as possible. The resultant overall capture efficiency
2-16

-------
for the series is in the range of ninety percent. The asbestos
fibers which escape are predominately the smaller, respirable and
wind transportab 1e ones. The overall emissions by the asbes tos
milling industry, which has a partial usage of fabric filters,
is 100 pounds per ton of asbestos produced, or a rate of loss
of about five percent.
The source of emissions in the mill which is more expensive
than the others to control is the ore dryer. The exhaust stream
of the dryer is typically at high humidity and about 250°F. The
associated baghouse must therefore be insulated to avoid condensation
and resultant blinding, and the bags need to be of temperature-
resistant materials, such as Or10n or Dacron.
This means,
in practice, a separate baghouse or baghouse section, and higher
investment and maintenance costs than for other emission sources
in the mill.
Milling by Wet Method
The details of operation of the one wet mill asbestos plant
in the U. S. are not known, since it is a proprietary process.
However, two important points relevant to emissions are known.
First, because of the loose nature of the ore at the source of
this mill's asbestos, little or no ore crushing is required.
Second, the product is available as either loose fiber or as
compressed pellets or balls of 1/4" - 1/2" diameter.
However, because of the nature of the ore more emissions
3.
are to be expected during transport and unloading, because so
much more of the asbestos is free fiber at this stage.
Emission
potential during milling is virtually zero since the fibers are
wet. Drying of the fiber, however, has great emission potential
since the fiber is believed to be loose at this time. If the
fibers are pelletized before drying, however, the emission
potential would be comparable to that from conventional ore
drying processes. Emission potentials from bagging and
shipping of loose fiber would be the same as for conventional
plants, but much less in the case of pelletized fiber. Emission
from tailing dumps are expected to be less, especially if the
2-17

-------
tailings are discarded wet, because of the low fraction or ore
that becomes tailings. Water pollution potential, which cannot
be dis~ussed further, is of course quite high.
4. Emissions and Control Costs for Asbestos Milling
The estimated 1970 emissions and emissions after full controls
are instituted are shown in Table 2-3 along with the associated
control costs. There are two categories shown for control costs--
"small" plants and "large" plants. This is done because the
average control costs per ton of asbestos produced were fairly
uniform within these categories, and because it was necessary
to avoid identifying individual plants. However, such a
classification does hide the variation in costs among plants.
Additional control cost per ton varies among plants from zero
to $5.96 per ton. The additional annual investment varies from
zero to $183,000.
The control costs in Table 2-3 are based on a confidential
plant examination describing the control equipment needed, and the
data in Table 2-4. Minimum costs are assumed. Units are added
rather than totally replaced, no land is acquired, and ducting is
minimal. Table 2-5 estimates to cost of controlling emissions
from asbestos milling for 1970 and 1977.
Table 2-3. EMISSIONS AND CONTROL COSTS ASSOCIATED WITH' 
  ASBESTOS MILLING (FOR PLANTS EXISTING IN 
  1970;1970 DOLLARS)   
  Total Emissions   Per Ton
Plant Size 1970 If Fully Control Costs of
Number of Plants Actual Controlled Investment Annual Capacity
Large 5 3,750 150 333,480 124,189. 0.44
Small.. 4 110 4 8,340 3,579 1.38
Totals 9 3,860 154 341,820 128,189 0.78
Average  430 17 37,980 14,243 0.78
2-18

-------
Table 2-4.
UNIT COSTS FOR ASBESTOS MILLING EMISSION CONTROL
Control Unit
Cyclone collector
Baghouse
Investment
(dollars/acfm)

$0.55
Operating and Maintenance
(dollars/acfm-year)

$0.13
Low temperature
(cotton sateen bags,
no insulation)
3.00
0.90
High temperature
(OrIon bags,
insulated)
6.00
1.30
Source:
Reference 7
Table 2-5. COST ESTIMATES FOR CONTROLLING EMISSIONS FROM THE
 ASBESTOS MILLING INDUSTRY, 1970 AND 1977 
 Estimated   
 Capacity Required Annualized 
Year (Tons) Investment Cost 
1970 163,600 $3~1,820 $128,189 
1977 116,000 367,840 137,280 
Gas flow data required to produce total costs for Table 2-3
from the unit costs in Table 2-4 were either obtained directly
from an OAP/EPA survey of plant sites or were based on capacity
data obtained in the survey combined with the seven tons of air
per ton of asbestos criterion for air flows. The seven ton value
is for milling only; air flows for drying of ore is highly dependent
on ore asbestos concentration and water content.
In any case,
actual air flow rates for ore dryers needing control were
directly available.
2-19

-------
D.
Economic Impact
1.
Industry Structure
The aspestos milling industry in the United States is made up
of nine plahts owned by nine different firms. Four of these firms
are large, vertically integrated firms that manufacture a wide
range of aspestos and other products. These firms can use their
total U.s. asbestos fiber production as raw material for further
manufacturing into finished products for consumers and industry.
The remaining five firms are much smaller and sell their production
on the open market. The largest plants are not all controlled
by the largest firms. Table 2-3 indicates the investment, the
total annualized costs, and a cost per ton of fiber produced
for the additional emission control equipment required by the
industry. From the information currently available, it has not
been possible to link individual plants with firms. This has
limited the analysis of economic impact.
2.
Production and Consumption
United States production of asbestos in 1970 was approximately
125,000 tons which was almost the same as the 1969 record high
production. Domestic asbestos production has tripled. s~nce.1956,.
a pace that has been only slightly interrupted twice, and represents
a rate of increase of about eight percent per year (Reference 1].
Domestic production normally supplies about 10 to 15 percent
of United States consumption. The remainder is provided by imports,
primarily from Canada, but with minor amounts from the Republic
of South Africa and 10 other countries. .Asbestos is traded in a
world market where demand currently exceeds production at existing
price levels (discussed below). As pressure on world supplies and
price increases, expansion of domestic production can be expected
to supply domestic buyers unable to obtain foreign supplies.
United States asbestos consumption decreased 17 percent from
1951 to 1961. From the 1961 low, consumption increased 31 percent
to 875,000 tons in 1970. This represents a rate of increase of
about three percent per year.
2-20

-------
For this study, growth in domestic production was projected
to be a five percent per year, which is about the growth in world
consumption. It has been suggested that five percent is too high
because production lias been virtually level since 1966. Domestic
asbestos consumption is tied to domestic construction activity
which has been almost level until recently. As construction has
increased asbestos consumption has increased.
3. Prices
Without yields by grades from asbestos ore, it is difficult
to determine any kind of a realistic average price per ton of
milled asbestos produced. However, the poorest grade was selling
for $65-$90 per ton in Arizona (since August 1968) and the same
grade $43.50-$90.50 per ton in Vermont (January 1970). Vermont
prices were increased 4 to 12 percent depending on grade in 1969.
Top grades sold for $386.50-$408.00 in Vermont and $1410-$1650
in Arizona. Although published Bureau of Mines data indicate
that 125,314 tons of asbestos were sold in 1970 for $10,696,000,
an average value of .$85.35 per ton, it appears that
is low. This value falls within the price range of
This may be because shipments made to plants within
are valued at cost rather than market value.
Prior to 1969, the world market for asbestos was a "buyers'
this value
the lowest grade.
the same company
market, but a signficant change to a "sellers" market has occurred
since the end of 1969. All the South African producers are so
fully committed that they are refusing to quote on substantial
tonnages of chrysotile, crocidolite, and amosite. Demand exceeds
present production capacity in the Western world and a shortage
of asbestos fiber is occurring at current prices. Accelerated
exploration, development and modernization of asbestos properties
is occurring throughout the world to meet increased demand [Reference 9].
Prices for the various grades of chrysotile asbestos are set
by the leading Canadian producers and are largely followed by
chrysotile producers in the other parts of the world. Chrysotile
prices have gradually increased over the years, keeping up with
increasing production costs and other prices. The 1970 prices
averaged about 30 percent above those in 1960.
2-21

-------
4.
Economies of Scale
Annual control costs of $5.96 per ton of fiber produced are
indicated fo~ some plants. It is believed these plants are almost
completely uncontrolled, indicating that $6 is close to the
maximum unit cost to control air pollution by asbestos plants.
Production information indicates that about 20 percent of the
asbestos fiber from ore is caught in the baghouse by plants
equipped with baghouses. Plants equipped with baghouses have
been able to sell about half of their baghouse catch. While the
price is relatively low for this grade of fiber, such sales
can add about five percent to sales revenue [Reference 7]. Therefore,
plants with sufficient baghouse catch to supply a market do realize
some economy from their size. In total the economies of scale are
not ver-r large.
5.
Price Impact
Since asbestos prices are set in the world market.and U.S.
production supplies only 10' to 20 percent of U.S. consumption,
domestic plants could be expected to have little chance of passing
increased costs along as price increases.
Exceptions would occur in case of certain fibers that are
in short supply in world markets. Producers of these fibers could
increase prices whether or not their production cost had increased.
The bases for the ability to pass on such a price increase would
be supply and demand factors rather than pollution control cost
changes. It appears that Arizona chrysotile producers could
increase prices on certain grades. Demand for Arizona filter
fiber continues to be very strong and some grades are in short
supply. This is expected to continue for the near future.
Available information indicates that the Arizona producers
probably have the small plants faced with the high control costs.
If this is true, the current strong market would certainly help
pay those control costs.
2-22

-------
6.
Impact on Competition
Control costs of $1.69 per ton of fiber or less are indicated
for some plants. These plants are believed to be already partially
controlled and therefore already absorbing much of control costs.
Uncontrolled plants are currently realizing some competitive advantage
because of their avoidance of pollution control costs. Imposition of
air pollution control regulations would eliminate such a competitive
advantage.
7.
Investment Impact
The firms in this industry should have no difficulty obtaining
the capital required to install the necessary pollution control
equipment. Six of these firms have annual sales in excess of
$1,000,000. Four of these are large diversified firms. The
three smaller firms have rather modest equipment requirements
costing less than $3,000 installed and therefore should not
have difficulty paying for control equipment from retained
earnings or from loans.
2-23

-------
v.
Asbestos Handling
.A.
Occurrence and Control
One source1of emissions which are best controlled at the mill,
although they do not occur there, are those from storing and trans-
porting bagged asbestos fiber. There are several types of bags
currently used by asbestos mills, and they vary in the amounts of
emission they permit. In addition, spillover of fibers that occurs
during bagging, plus fibers that adhere to the surface of the bags,
are also source of emissions.
. The emission from bagged asbestos are the result of tearing
or breaking of weak bags, or leakage through tears, seams, or
permeable bags. The optimum solution, which is currently being
pushed by the milling industry, may be bulk transport in sealed
railraod cars with appropriate loading and unloading techniques
to minimize losses. This technique would also reduce losses by
eliminating bagging, unbagging, and bag disposal. Until the time
this technique is generally adopted, however, the best solution
appears to be use of specially-designed bags.
B.
Cost of Control
Table 2-6 shows the costs of the various types of bags currently
used to hold asbestos fibers. Of those listed, the coated, woven
po1yo1efin-fiber bags are reported to be as good as any of the listed
bags in terms of strength and impermeability, and they are clearly
suited to the application. Furthermore, their cost is relatively
low for a woven bag. Usage of these bags throughout the industry'
would result in a significant reduction of fiber loss.
In 1970., 2,600,000 bags were required to ship the asbestos fiber
produc~d in the United States. The bags used cost an estimated
$536,432. Use of coated, woven po1yo1efin bags would have cost
$637,000, an increase in cost of $100,568 to the industry. This
increased cost amounts to slightly less than $0.04 a bag or $0.80
per ton of asbestos fiber shipped.
2-24

-------
Table 2-6. COST OF BAGS FOR SHIPPING ASBESTOS FIBER, 1970 
      Number  Total
 Type of Bag Cost per .Used,1970 Dis tribution Cost
 (100 1bs. size) 1000 Bags (Thousand) (percent) (1970)
7 oz. Jute laminated $336.00 234 9 $ 78,624
10 oz. Jute   290.00 130 5 37,700
7 oz. Jute   210 .00 780 30 163,800
5-P1y Kraft paper 200.00 156 6 31,200
4-P1y Kraft paper 170.00 520 20 88,400
3-P1y Kraft paper 130.00 416 16 54,080
Woven Po1yo1efin(coated) 245.00 208 8 50,960
Other    203.00 156 6 31,668
 Total     2,600 100% $536,432
 Source: Reference 3     
Net cost to the industry would be even lower.
Bags of coated, woven
poly olefin would reduce shipping losses from broken bags. If the average
price of the asbestos shipped was $80.00 a ton, one 100 lb. bag of
asbestos is worth $4.00. Preventing the loss of one bag of asbestos
would pay the additional cost of 100 bags. If the fiber shipped is
of higher value than $80.00 per ton (and much of it is) the net
savings could be substantial.
It is interesting to note (Table 2-6) that 14 percent of the
bags now used cost more than po1yo1efin bags, and a total of 22
percent of the bags used are of equal or greater cost.
Disposal of po1yo1efin bags would probably be no more difficult
than disposal of any other type bag. In fact, in the manufacturing
of floor tile they can be used in the product, whereas other kinds of
bags cannot.
In summary, it is believed that no net cost would accrue to the
industry from using po1yo1efin bags.
The savings might be substantial.
2-25

-------
VI. Manufacture of Asbestos-Containing Products
A.
Nature of the Products and Processes
,
Asbestos Iproducts includes an extremely broad and diverse group of
items that contain significant amounts of asbestos fiber. It is estimated
that 3.000 items fulfilling a broad range of industrial and consumer needs
are among those in this group. They include asbestos cement products,
floor tile. friction materials such as brake bands and clutch facings,
asbestos paper. and asbestos textiles. Table 2-7 gives the breakdown of the
major uses for 1968
Table 2-7.
U. S. CONSUMPTION OF .ASBESTOS. 1968
Product
Tons of Asbestos
Used
Percent
Asbestos Cement Products

Shingles. siding. flat sheets
corrugated sheets, wall board
410.000
50.2
Floor Tile
Vinyl and asphalt
200.000
24.5
Total
104.000 12.7
30.000 3.7
18.000 2.2
16.000 1.9
3.000 0.4
36,363 4.4
817,363 100.0
Friction Materials
Asbestos Paper
Asbestos Textile
Molded Thermal Insulation
Sprayed Insulation
Miscellaneous
Source:
Reference 9
2-26

-------
Although about 3000 different asbestos products are manufactured,
they can be divided into two categories for asbestos emission potential.
Either asbestos remains as essentially free fiber throughout the process
and in the final product, or the asbestos is wetted or bound into a matrix
at an early stage of processing. Production of asbestos textiles is the
major manufacturing process in the first category. Virtually all other
manufacturing processes fall into the second category.
1. Receiving, Handling, and Storage
The processes of interest which are common to the manufacture
of all asbestos-containing products are receiving, handling, and
storage of the (bagged) asbestos fiber, the removal of asbestos
from the bags, and the opening or fluffing of the asbestos fibers
prior to the other manufacturing operations. There is also a
potential emission problem from the discarded bags, which inevitably
have asbestos fibers clinging to them.
Receiving, handling, and storage of bagged fiber can best be
controlled by use of the polyolefin bags described in the section
on asbestos milling, along with adequate precautions for careful
handling, and rotation of stocks to use the oldest bags first.
It is not anticipated that such procedures will produce any
significant direct costs to the manufacturer.
Prior to their actual entry into the final
fibers must be removed from their bags, usually
product, the asbestos
fluffed or "willowed"
to loosen them from their packed condition, and conveyed to the
start of the process. The most direct method of control, unless
an air method is used for fluffing the fibers, is to totally enclose
the operations involved and operate the entire apparatus under a
slight negative pressure, with the resultant airflow cleaned by a
baghouse filter. Collected fiber from the baghouse would normally
be returned without difficulty to the production process.
If air fluffing of some sort is used, the basic control strategy
is the same. However, much higher air flows would be involved,
and the baghouse catch would be significantly greater, perhaps
requiring a dry cyclone precleaner.
2-27

-------
In either case the clean and nearly total recovery of the
. fibers, along with reduced workspace fiber loads, should be of some
advantage compared with no control; costs should be partially offset
I
by fiber recovery, increased worker efficiency, and decreased
maintenance and janitorial expenditures. Even approximate quantification
of such savings is not possible at this time, however, and no attempt
to include such offsets has been made in the control cost calculations.
Required controls for other manufacturing steps will be discussed
with the descriptions of the individual processes studied for this
report, which follow.
2.
Asbestos Cement Products
Products in this category comprise the largest total vol~me
of asbestos~ontaining materials and use the largest total
amount of asbestos of any category (410,000 tons in i968).
Specific products include wallboard, pipe, shingle, and block,
and dry mixture for sprayed insulation. Advantages of the
products over their non-asbestos counterparts are improved
tensile strength and strength-to-weight ratio, strength under
heat stress and resistance, and smoothness of finished surfaces
(critical in pipe to be used for liquid transport).
After the asbestos fiber has been prepared, it is mixed
with the cement, as about 15%-20% of the total material, either
wet or dry. If the mixing is done dry, the resulting mixture
is generally metered in a flat layer onto an open surface where
the requisite water is applied by overhead spray [Reference 3]. The
resulting layer, much thinner than the final product, i8 then wound
onto mandrels (for pipe) in a spiral mat until the requisite thickness
is built up, or is layered flat onto wallboard or shingle forms,
etc. The same winding or layering process may be used for wet-mixed
products, or the mixture may be cast.
It is not known what advantages result from either the dry
, '
or wet mixing methods, except that the dry method obviously has
higher emission potential. Once the mixture has been wetted, and
through all other processes except finishing, emission potential for
asbestos fiber is essentially zero.
2-28

-------
Finishing processes on the dried cement products vary with
product requirements and type and may include grinding, drilling,
sawing, and cutting. All of these processes are dusty, although
in varying degrees. It is not known whether any free asbestos
fiber is released by these operations, but it is assumed that it
is possible for some to be released, albeit a small amount in
proportion to the total dust generated.
Collection of emissions from finishing processes would again
be by baghouse. However, the hooding technique at these sources
would be of high-velocity, low volume type as opposed to the low-
velocity, high-volume type of system used for the fluffing and
blending stages of processing. Furthermore, it is questionable
that the collected material would be returnable to the process,
since the fibers (if free) would be very short and the cement
dust would be "spent" and useful only as aggregate. However,
since the volume of collected material is so low, it would have
very little effect on the properties of the finished product if
it were recycled to the process. Its low volume means that, in
any case, its reuse would have little if any economic value
except to eliminate disposal requirements.
Control costs were based on the estimate that 95% of total
potential emissions are currently controlled, some processes being
fully controlled and others relatively uncontrolled. This was
interpreted to mean that every plant, on the average, required a
fabric filter and hooding system on its finishing processes, which
are a small proportion of the total control requirements.
Asbestos Vinyl and Asphalt Floor Tile
Asbestos is used in vinyl and asphalt floor tiles because it
improves strength and stability without reducing flexibility and
compressibility. When used it comprises 10-30% of the total
weight of the product.
3.
Emission potential from floor tile manufacture, the second
largest user of asbestos fibers, are nil as soon as the fibers
are mixed with the hot vinyl or asphalt. The finishing processes
involve cutting tiles to size and shredding wastage and trimmings
2-29

-------
for reuse; emission of free asbestos fiber from finishing
processes should be negligible because the cutting process does
not disp~rse fibers into the air. Use of the po1yo1efin bags
previously mentioned as a possible control of handling losses would
eliminate any bag disposal problems, since these bags can be
shredded and used as part of the material input.
Control cost estimates were based upon number of plants,
average amount of asbestos used per plant, and the estimate
that 95% of potential emissions are currently controlled on the
average, some processes being completely controlled and some with
lesser or zero control.
Asbestos Friction Products (Including Gaskets)
Friction products and gaskets using asbestos contain from
30% to 80% of asbestos in some sort of (generally) organic
binder. In friction products the asbestos is used for its
strength, frictional properties, and stability at high
4.
temperatures. Gasket materials do not depend upon the frictional
advantages. The large usage of asbestos friction materials in
vehicle brakes accounts for consumption of the third largest
annual amount of asbestos of any product type.
Asbestos is used in these products in two different ways: (1)
the asbestos, as loose fiber, is mixed with the binder; or (2) the
asbestos, as either matted or woven textile, is impregnated with
the binder. The low total volume of asbestos textile production
indicates that the latter process is used only in special situations,
probably in gaskets where dimensional stability is of significance.
The further processes, such as molding, and curing, are po1lution-
free, at least with regard to asbestos. Shaping, cutting, sawing,
and other finishing processes have at least some pollution potential,
however.
Control costs were based upon number of plants, average
amount of asbestos used per plant, and the estimate that 95%
of potential emissions are currently controlled.
2-30

-------
5.
Asbestos Paper Products
Asbestos paper has essentially the same properties as normal
(cellulose-based) paper, except that it has better thermal
insulation properties and fire resistance.
Its primary use is
in insulation, although it has been reported that high-quality
bond document papers are also produced, for durability and stability.
Asbestos papers are made by the same techniques as standard
wood-pulp papers, but with the fiber being asbestos (80%-90%)
and (typically) china clay and starch or sodium silicate being
used as binders. Once the slurry is mixed in hollanders, emissions
are nil until the final slitting process. It is expected that
significant free fiber can be emitted during slitting, since
the paper matrix is not as firmly formed as for other asbestos-
containing products.
Control costs are based upon number of plants, average
amount of asbestos used per plant, and the estimate that on
the average, 95% of potential emissions are currently controlled.
The implication derived from this is that one minor emission
source, such as the slitting process, needs to be controlled.
6.
Asbestos Textile Products
Asbestos is capable of being made into the full range of
textil~ products, from nonwoven lap and felt through yarn and
cord to woven cloth and tube, and braided rope and tube.
However,
the asbestos fibers required for textiles are significantly
different from those used for other asbestos products, having
to be quite long in order to be spinnable. This fiber is sometimes
obtained in "crude" form as unopened, hand-cobbed, rock-free
fiber blocks or bundles to protect fiber length.
If the fibers are received as crude, they are opened in
edge (knife) mills into small, fiber-like bundles. These are
then milled into extremely fine fiber for flexibility. This
also makes the resultant fibers (a) more delicate and breakable,
and (b) more "floatable", leading to a greater emission potential
per weight of fiber.
2-31

-------
Once the fibers have been adequately opened and fluffed,
they may be blended with up to 20% of a cellulosic fiber such
as cotton" the specific material chosen depending upon the
applicatibn of the final product. The subsequent processes,
such as carding, lapping, roving, spinning, and weaving or braiding
(as required) are all performed on equipment essentially identical
to standard textile machinery.
The key to the. emission problem is that standard textile-type
I
equipment and processes are used. This means two things:
(1) large surface areas per unit volume of asbestos exist during
the entire processing procedure; and (2) separate pieces of
equipment that are impossible to hood for emission control, at
least economically. This latter point results from the very
large size of the machinery and the requirement for frequent access
,"
to the equipment.
Two recent developments in asbestos textile technology have
positive implications for emission control. First, it has been
found that a thin coating of a polymer on the asbestos yarn
improves the processing efficiency; in general, the coating
does not disturb the quality of the final product, and in some
cases may enhance it. It also results in about 80% reduction
in fiber emissions from processes following the coating process.
Second, control of the processing environment (temperature,
humidity, etc.), which has beneficial effects upon processing
efficiency. Close environmental control is best and most
economically handled by recirculation of the ventilation air

, ""
required for industrial hygiene. Since this can only be done
if the ventilation air is thoroughly cleaned, baghouse filters'
in the ventilation stream are used, resulting in high emission
reduction. These procedures are not yet practiced on an industry-
wide basis, however.
Costs of control are based upon total asbestos consumed,
number of asbestos textile plants, and the estimate that 85% of
potential emissions are currently controlled. 85% control was
interpreted to mean that a major source of emissions is at present
uncontrolled.
2-32

-------
B.
Economic Analysis
1.
Industry Profile
Tha asbestos products industry (SIC 3292) includes
establishments primarily engagedi.n manufacturing asbestos
textiles. asbestos building materials except asbestos paper
(industry 2661) and other products composed mostly of asbestos
except asbestos gaskets and insulation (industry 3293).
In 1967 there were 139 asbestos products establishments
(SIC 3292) employing 21,400 and shipping products valued at
$576.5 million. The number of firms had increased 12 percent,
the number of employees had increased 10 percent, and the
value of products shipped had increased 13 percent since 1963.
Value added by manufacture at $308.9 million in 1967 was 14
percent more than value added in 1963. These firms consumed 442,200
tons of crude fiber in 1967 which was 1,100 tons less than that
consumed in 1963.
Asbestos paper (SIC 2661) is manufactured by 29 plants, with
total shipments valued at $5.58 million. Normally, asbestos
paper is 80-85 percent asbestos.
This industry reported using
99,800 tons of asbestos worth $10.0 million in 1967.
The Gaskets, Packing, and Asbestos Insulations Industry
(SIC 3293) includes establishments primarily engaged in manufacturing
packing for air, steam, water, and other pipe joints, and for
engines and air compressors; insulating materials for covering
boilers and pipes; and gaskets.
Establishment primarily
manufacturing leather packing are classified in industry 3121,
rubber packing in industry 3069, and metal packing in industry

. .
In 1967, there were 300 establishments in this industry.
is only five more than in 1963 but 18 more than in 1958. The
3599.
This
value
of products shipped amounted to $350.8 million in 1967, an increase
of 26 percent since 1963 and 86 percent since 1958. Value added
by manufacture was $205.2 million in 1967, an increase of 28
percent above the value added in 1963. Employment increased 10
percent from 1963 to a total of 18,500 in 1967.
2-33

-------
Value of asbestos, asbestos-metallic, and asbestos-rubber
gaskets shipped totaled $62.3 million. Value of asbestos, asbestos-
metallic:, and asbestos-rubber packing shipped totaled $18.1
million.' The value of insulating materials containing asbestos
totaled $35.7 million. Manufacturers in this industry purchased
crude asbestos at a delivered cost of $10.9 million in 1967. In
1963 they purchased 31,400 tons of crude asbestos worth $5.0
million [Reference 10].
2.
Cost of Control
Table 2-8 shows the control costs for the major classes of the
secondary asbestos products manufacturers.
Total investment
required is $5,384,000.
The annualized cost is $1,615,200.
This
amounts to an average investment
annualized cost of 0.4 cents per
annualized cost per dollar value
to 5.2 cents.
of 1.4 cents or an average
dollar value of output.
The
of output ranges from 0.1 cents
Service life for control equipment was assumed to be 10 years.
Operations and maintenance cost were calculated at 10 percent of
investment. Interest~ insurance and taxes were also calculated at
10 percent of investment.
Table 2-8.
ASBESTOS PRODUCTS INDUSTRY CONTROL COSTS, 1970
   Value of  Annua~ Industry Cost
   Product Inves- 1ized Annualized Cost
  No. of (Million. ment cost To Value of Pro-
Product Group Plants Dollars) ($1,000) (1,000) duct Ratio
Asbestos Cement 48 $142,680 $2,400 720.0 .005
Floor Tile 18 37,200 216 64.8 .002
Friction Material 30 202,800 720 216.0 .001
Asbestos Paper 29 5,580 348 104.4 .002
Asbestos Textiles 34 9,900 1,700 510.0 .052
Total 159 $398,160 5,384 1,615.2 .004
.:,      
   2-34   

-------
3.
Price Impact
As indicated in Table 2-8, the price increase to consumers,
if all costs are passed on in increased prices, is significant
for only one class of products--asbestos textiles. The price
increase for that product group would be 5.2 percent. This
would cause the loss of some sales, but such losses would be
relatively small because of the specialized nature of the
applications of asbestos textiles. The exact amount of this
reduction in sales was not determinable from the data available
in this study.
4.
Investment Impact
The average plant investment required for control equipment
and the average value of plant production is given in Table 2-9.
In addition, the Investment to the Value of Product Ratio is
given. This shows the investment burden is not very large,
averaging less than one percent of sales for plants making floor,
tile and friction materials and less than two percent for plants
making asbestos cement products. For these firms the required
investment should be no problem except for firms that are already
marginal.
For plants making asbestos paper and textiles, the required
investment is 6.2 percent and 17.2 percent of product value
respectively for the average plant. These are significant. In
both groups one could expect the smaller firms to have difficulty
obtaining the required capital and closures and consolidations
may result.
Data was, however, inadequate to determine the
probable number of firms that would close.
2-35

-------
'Table 2-9. .PLANT VALUE OF PRODUCT AND INVESTMENT OF THE
,---..----..-..---- --._._"'m.'-'-' ..-.--
ASBESTOS PRODUCTS INDUSTRY
    Investmen-t
    to Value of
Product Value of Product Required Product
Group  Per Plant Investment Ratio
Asbestos Cement $2,972,500 . 50,000 0.017
Floor Tile 2,066,700 12,000 0.006
, Friction Material 6,760~000 24,000 . 0.004
Asbestos Paper 192,400 12,000 0.062
Asbestos Textiles 291,200 50,000 0.172
2-36

-------
VII.
Sprayed Asbestos Fireproofing and Insulation
A.
Nature of the Product and Process
Asbestos-cement insulation is applied to steel-frame buildings as
thermal insulation and fireproofing; it also has good acoustic insulation
properties. The sprayed mixture is approximately 30% asbestos, 55% rock
wool, and 15% cement with sufficient water for mixing and setting. The
dry mixture is delivered to site in 50-lb. bags. The spraying technique
uses either a wet slurry pumped to a nozzle or a special nozzle which
mixes water and the dry mix. Although small amounts of asbestos are
emitted from this use, it is of significant concern because of its
concentration in urban areas.
Emissions of asbestos in this process
arise from handling of the dry mixture, escape of unwetted fiber and
mixture, overspray and backsp1ash, and cleanup and disposal of wastes.
There are various measures of partial control such as premixing
with water in the bag, enclosing the sprayed area, and better cleanup
and disposal control. Of course banning of the use of asbestos leads
. '. .
to total control. There are alternative materials such as mineral wool,
ceramic fibers, calcium silicate and vermiculite alumina available for
various applications which yield comparable results at a modest increase
in cost. When it is considered that materials cost is low for the total
cost of application of insulation, and that insulation is a minor expense
in total construction, the impact is not significant. In fact, it
is cheaper to use a different material than to apply the other
techniques.
B.
Economic Analysis
The economic impact of substituting other insulating materials for
spraye~asbestos insulation appears to be negligible. The substitute
materials are not much, if any more expensive to buy or to apply. The
same process and workers apply the substitute. No retraining is required
because insulators are already applying the alternative materials for
some applications.
.
Therefore, construction costs are not changed.
2-37

-------
Sprayed asbestos insulation is not a major use of asbestos. Three
thousand tons of asbestos were used in this product in 1968. Total
consumption of asbestos was 817,363 tons that same year. . Sprayed
asbestos insulation was less than 0.4 percent of total consumption.
I
The loss of sales of 3,000 tons would not have a major impact on
either prices or the firms producing or marketing asbestos.
2-38

-------
10.
12.
ASBESTOS REFERENCES
1.
Timbrell, V. "The Inhalation of Fibrous Dusts," Annals N. Y. Academy
Science, Vol. 132, p. 421. 1965.
2.
Sullivan, Ralph J. and Yanis C. Athanassiadis. Preliminary Air
Pollution Survey of Asbestos: A Literature Review, National Air
Pollution Control Administration Publication No. APTD 69-27, Raleigh,
North Carolina. 1969.
3.
Harwood, Colin F. A Basis for National Air Emission Standards Asbestos
(Technical Review Draft), lIT Research Institute, Chicago, Illinois,
1971.
4.
"Asbestos," Environment, Vol. 11, No.2, 1969.
Selikoff, 1. J.
5.
Nicholson, William. "Proposed Standard for Workers is Questioned,"
Insulation Hygiene Progress Reports, Mount Sinai School of Medicine of
the City University of New York, Vol. 3, No.2, 1971.
6.
Grossmuck, G. "Dust Control in Open Pit Mining and Quarrying," Air
Engineering, July 1968.
7.
Environmental Protection Agency. Private communications with officials
of the Environmental Protection Agency, 1971.
8.
Clifton, R. J. "Asbestos," 1970 Minerals Yearbook, U.S. Department of
Interior, Bureau of Mines. Washington, D.C.: U.S. Government
Printing Office, 1970.
9.
Davis, W. E. and Associates. "Asbestos--Section III," National S
Inventory of Sources and Emissions: Cadmium, Nickel artd Asbestos,
Department of Health, Education and Welfare, NAPCA. Leawood, Kansas:
1970. 45 pp.
U.S. Bureau of Census, Census of Manufactures, 1967. Industry Series:
Abrasive, Asbestos, and Miscellaneous Nonmetallic Mineral Products,
MC 67(2)-32E. Washington, D.C.: U.S. Government Printing Office, 1970.
11.
Rezovsky, H.
May 1957.
"Air in Asbestos Milling," Canadian Mining Journal,
Piuze, Lionel C.
No.3. 1971.
"Asbestos," Engineering and Mining Journal, Vol. 172,
2-39

-------
Chapter 3:
Beryllium
I.
INTRODUCTION
The purpose of this chapter is to. present the engineering analysis
of air pollution controls for the industry sources producing or pro-
cessing beryllium or beryllium containing materials. An economic
discussion is not included because the engineering analysis indicates
that all potential sources of beryllium emissions are controlled within
the limits of the standards proposed at the time this study was concluded.
The primary producers of beryllium are represented under Standard
Industrial Classification (SIC) Code 3339, and include those plants that
produce beryllium metal, beryllium alloys, and beryllium ceramic inter-
mediate materials and finished products to other industries or customers.
The manufacturing industries are classified under SIC Code 2819 and 3369,
and include some 5,000 fabrication plants and machine shops that are
supplied beryllium metal, beryllium alloys, and beryllium ceramic materials
by the primary producers for further conversion to finished products.
II.
INDUSTRY PROFILE
There are two primary beryllium producers [Reference 2,3,4] and
each uses one or more of three production processes: fluoride,
sulfate, and acid-leaching, organophosphate extraction.
steps in the production process are:
The various
o
Beryl or bertrandite ores to hydroxide
Hydroxide to beryllium metal billets
Billets plus scrap to beryllium metal shapes
Hydroxide to beryllium plus scrap to alloy
Hydroxide to beryllia ceramics, powders and shapes
o
o
o
o
3-1

-------
The process steps, except from ores to beryllium hydroxide, are very
. similar in the plants of the two primary producers. These pri~ry
production pla~ts also manufacture the following general categories of
produc ts :
o Beryllium metal (billets, powders, pressed blocks, mill
products and fabricated products)
o Beryllia (powders, ceramic shapes and waves and fabricated
products) I
o Alloys (castl billets, mill products, and fabricated products)
I
I '
. Beryllium metal products are made mostly from pressed powder and are
forged, extruded, and machined. ,Beryll~a powders are pressed, extruded,
fired, and machined by ,ordinary ce~amic techniques. Finished beryllium-
copper alloy products are made from 'melts of copper and a master copper
alloy.containing 4% beryllium. Small quantities of beryllium-nickel and
beryllium-aluminum alloys are also produced. Alloy products take the form
of bar, plate, rod, wire, forgings, and billets.
Beryllium production in 1970 was about 394 tons *[Reference 2,3].
Table 3-1 presents the tonnage of beryllium found in the various
beryllium-contained products produced by the primary production plants.
These amounts of beryllium equivalent by type product are considered
as inputs to the various manufacturing plants.
:. .'
TABLE 3-1.
Berylliu~.Production by the Primary Producers, 1970~

Equivalent Be
(short tons)
Product Type
Beryllium Billets
Master Alloy
Berylli um Oxide
(Metal)
147
225
22
* Exact production figures are not published in order to avoid disclosure
of the activities of individual firms. The estimates used in this report
appeared to be the most accurate available at the time. Production has
declined in recent years, however, and it may be that these figures are
significantly high.
3-2

-------
III.
TECHNICAL ANALYSIS
A.
Nature of Product and Processes
Primary Producers(Be)
Beryllium has exceptional strength and rigidity, exceeding that
of other metals including steels. It has a high strength-weight
ratio and important thermal and nuclear properties.
Beryl (beryllium ore) is normally recovered as a co-product or
by-product from mining of other minerals. Blasted rock is hand-
cobbed and barren rock is broken off with hammers and discarded.
Beryl and other valuable minerals are recovered at this time.
Several methods for converting beryl to beryllium oxide (BeO)
have been developed. These production process methods are fluoride,
sulfate and acid-leaching, organophosphate-extraction.'
A brief description of each of the three production processes
follows as well as a description of the processing required to
1.
produce beryllium metal.
," .'
Fluoride Process
Beryl containing 10-12 percent BeO is crushed, gr?und in a wet
ball mill and filtered. It is then mixed in batches with soda ash,
sodium silicofluoride, and' sodium ferric fluoride and made into
briquats.. These briquets are dried, sintered at 750°C, cooled,
. ,
crushed, and ground in a wet pebble mill to which hot~water is
added, and the slurry is pumped to a tank for leaching. More
water is added and the mixture agitated, leached, and allowed to
settle. TIle liquid, containing soluble sodium beryllium fluoride,
is decanted to separate it from solids, which contain aluminum and
iron oxides and silica.
to precipitate beryllium
convert it to bery11ia.
Sulfate Process
Caustic soda is added to the heated solution
hydroxide, which is filtered and calcined to
Beryl containing 10-12' percent bery1lia is crushed, dried,
and melted at l600°C in an electric arc furnace. The melted beryl
is then quenched in cold water to obtain a frit, which is dried
and ground to a fine powder. Batches of powder are mixed with
3-3

-------
concentrated sulfuric acid, steamed and agitated. Water and
more steam are added to the slurry. The liquid, containing
. soluble beryllium and aluminum sulfates is filtered from the
sediment and pumped to a tank where ammonium hydroxide is added.
I
The filtrate from this operation is further treated with a

. .
chelating agent to prevent impurities from precipitating upon
subsequent addition of caustic soda. Hydrolysis follows, and
the precipitate, beryllium hydroxide, is filtered off. This
precipitate is ignited in an electric furnace to form beryllium
oxide.
Acid-Leaching, Organophosphate-Extraction
Bertrandite from the Spar Mountain, Utah, mine is the first
significant nonberyl source of beryllium (averaging 0.5 percent
Be, but containing 17 percent. moisture and creating little dust). It
is mechanically removed from the mine without blasting and stockpi'led.
The ore is hydrated bertrandite and berylliferous saponite, found
in association with the fluoride; the bertrandite is soluble in
concentrated acids. The are is first crushed in a jaw crusher that
is hooded and the dusts evacuated for scrubbing. It is next
slurried in water and pulverized in a ball mill. The sized slurry
goes to process and rejects are recycled. The slurry is next
acidulated with sulfuric acid which converts the Be to the sulfate.
The calcium in the rock is converted to gypsum and then by pH
adjustment, most unwanted solids are precipitated. The Be values
remain in the weak mother-liquor which goes to further processing.
An extensive counter current in the slurry settling system is
employed that receives various recovery and scrubbing streams.
The process used by one of the primary producer plants is
proprietary. The extraction system is basically an organic
process that yields a wet cake product of beryllium hydroxide.
The wet cake is fed into fifty-five gallon drums and shipped to
another plant. All processing is totally enclosed, including
vacuum stages, and vents are scrubbed so that stack emissions
are not visible.
As the drums are filled, they are tightly
3-4

-------
enclosed and air evacuated through a scrubber. The highest
temperature involved is 205°C. Accidental spills of powder
or slurry are hosed down into a chemical sewer system that is
flushed back to the slurry system. Standard emission control
systems are being used. Exhaust streams are treated with wet
venturi type scrubbers. Flow rates are controlled and exit
gases monitored.
Processing to Beryllium Metal
After beryllium hydroxide is isolated in satisfactory form
and purity, the processes are all rather similar in the operations
of the major producers. In general, the fluoride process produces
a technical grade beryllium oxide lower in aluminum and iron,
while beryllium oxide from the sulfate process is lower in
silicon and sodium. Both oxides are pure enough for production
of commercial beryllium alloys. Nuclear grade beryllium metal,
metal for space application, and shapes for nuclear and electronic
industries necessitate the production of high purity grades of
beryllium oxide.
The oxide or hydroxide can be dissolved in ammonium bifluoride
solution and heated to boiling, and calcium carbonate flour added
to precipitate aluminum. . The solution formed is pttrified by
precipitation and filtration, and the pure filtrate is evaporated
and crystallized to produce pure crystals of ammonium beryllium
fluoride. Next, the ammonium fluoride is driven off by heating,
leaving a residue of beryllium fluoride. .Decomposition is carried
out in a graphite-lined furnace at a high temperature (900° to 950°C).
The ammonium fluoride, which sublimes off at lower temperatures, is
collected in an air-cooled iron condenser and recycled.
The solid beryllium metal produced by this process can be
used directly as virgin metal for alloying purposes and for casting
pure beryllium. The metal is remelted for casting purposes to
remove any slag inclusions.
The remelting process is carried out
in a vacuum, without using fluxes, or at atmospheric pressure
using a flux rich in beryllium fluoride.
3-5

-------
2.
Manufacturing Industries Using Beryllium
a. General
, .
Manufacturing plants may be categorized as metal, alloy,
and ceramic. The number of plants in each category is not known.
Estimates range; in the "hundreds" of metal plants, 5,000 to 7,000
alloy plants and something less than 100 ceramic plants. It has
also been estimated that "hundreds" of foundries use beryllium
alloys [Reference 3,4] and that some use the master (4%) copper
alloy.* In determining control costs these estimates of the
number of manufacturing plants are not meaningful. Model plants,
however, can be established and control costs estimated for such
plants. In all probability the degree of control of emissions
varies from zero to full control (0.01 ~g/m3) within each plant
category. Undoubtedly emissions from melting, pouring, and casting
of beryllium alloys go uncontrolled in "hundreds" of these plants.
The operations in manufacturing plants are widely varying
in type and amount. Shops using lathes, milling machines, and similar
machinery may generate, as an example, an average of 0.05 to 0.1 grams
per minute of beryllium dust; and this dust may be drawn off with
1,000 to 4,000 cfm of air with a source concentration of about l03~g/m3.
Alloys produce less beryllium dust per machining operation but are
frequently subjected to foundry practices.
In a machining operation (e.g., lathes) a local exhaust
pick-up of dust can be applied at the tool at 1,000 to 1,500 ft/min
and the volumetric flow may be about 50 cfm depending on the inlet
area of the pick-up. A grinding operation may require a plastic
enclosure and a somewhat higher volumetric flow. A hood may be used
for grinding or spraying operation and this may require 500 cfm at a
face yelocity of 150 ft/min. A foundry ~peration may require an exhaust
of 5,000 cfm. From this it can be seen that the volumetric flow re-
quirements for a manufacturing plant are generally less than 20,000 cfm.
Small companies may work Be metal and perform turning,
cutting, drilling, milling and other operations, from which the total
exhaust throuBh manifold local pick-ups m~y be less than 1,000 cfm. .'
* EPA has estimated that approximately 90 percent of this production
is concentrated in about 15 companies.
3-6

-------
In such cases, a suitable package may be a mechanical (dry)
collector, together with a pre-filter and a HEPA filter all in
series to a stack. The collector system may have a resistance
less than 10 in w.g. It is reported that HEPA filters have an
efficiency rating of 99.9% for particle size of maximum leakage
of 0.3 microns.
From all reports it appears that the control of emissions
is thorough among government contractors but is highly variable
in private industry.
Many companies apparently believe that since they use
only a small amount of beryllium in various forms, controls are
not really necessary. Also, there appears to be a belief, possibly
erroneous, on the part of many users that "high fired" BeD is safe
and consequently controls range from zero to full.
An example of full control cited by one report [Reference 3]
is as follows:
Process ~ cyclone ~ oil mist trap ~ rough filter~ngJ.
(if required) (if required)


stackliwith + HEPA filter + bag filter
s amp ng
b.
Beryllium Metal Fabrication
Beryllium metal has unusual physical properties--high
stiffness, high heat capacity, and low density., It is produced
largely by reduction of beryllium fluoride with metallic magnesium.
Beryllium metal is generally purchased as rod, bar, or billet.
It is used in nuclear reactors as a moderator and reflector
material.
Other uses include gyroscopes, accelerometers,
internal guidance systems, parts for high-speed flight and
marine navigation, rocket propellant fuel, airplane brakes,
and heat shield for space capsules. Hore recent uses include
rotary blades and other parts for gas turbine engines, solar cell
mounting boards for satellites, mirrors, portable X-ray tubes,
3-7

-------
optical parts, X-ray diffraction and microradiography. Fabrication
operations include turning, milling, drilling, reaming, grinding,
honing, sawing, and abrasive cutting. Chemical and electrochemical
procedures are also used. Review of reports indicate that about
. .
147 short tons of; the beryllium metal were produced for these uses
in 1969 and that these figures included reclaimed scrap.
The number of machine shops that work Be metal is
not known but is reported to be in the "hundreds" [Reference 3;4].
It is also reported that these shops have emission controls from
zero to approximately 100 percent. It is estimated that in 1969,
127 tons of Be billets were produced from be(OH)? It is further
. ~

estimated that the beryllium emissions from machine shops are 8
grams/day and it is assumed that these plants operate 250 days/year
[Reference 3]. Based on these estimates and assumptions the emission
factor (EF) would be:
8
EF co 127
250
..
8
0.508
.. 15.75 grams/ton or 0.035 lbs/ton .
Be bill~ts plus scrap to metal shapes were reported as 147 tons
produced for 1969. The emissions are estimated as being 9 grams/day.
Therefore:
9
EF = 147
250
=
0.~88 = 15.31 grams/ton or 0.038 lbs/ton.
Ninety-five percent of all Be metal applications is for
the government. One-half of this Be metal is used in nuclear weapon
applications and the other half in noncommerc~l reactors. The
estimated demand for Be metal by 1972 (using 1967 as a base) is 175%.
This indicates a growth rate of about 12%. In 1970 Be metal was
selling, on the average, for about $95.00 per lb.
c.
Beryllium Alloys
To make beryllium-copper master alloy (4% Be) weighed
quantities of BeO, carbon powder, and copper are mixed in batches
and melted in an electric arc furnace. The product is impure
beryllium-copper. Additional refining or other treatment removes
the carbon and gases absorbed during reduction.
3-8

-------
.-......-~'... ~
Beryllium-copper has high electrical and thermal
conductivity coupled with strength and resistance to fatigue

. ,.

at high temperatures. Beryllium copper is used in springs,
. .
,. '- _.
bellows, diaphragms, electrical contacts, aircraft enginer parts,
. .
bushings, valves, shims, pressure gauges, plastic molds, marine
propellers, gears, bearings, precision castings, rollers, low-
sparking tools, radio and radar devices. The largest users of
beryllium-copper are in the electrical and electronic fields.
Beryllium-copper is available in wrought, cast, or forged form.
Most plants using beryllium-copper do not remelt or make basic
changes in the alloy. The processing generally involves stamping
. ." .. ..
or drawing into finished shapes.
Beryllium-nickel alloys are heat treatable and resemble
stainless steels in many respects. Uses of beryllium-nickel
alloys include surgical instruments, matrix for diamond drill
bits, parts for fuel pumps and business machines, and dies
for shaping aluminum channels, necks of bottles, stainless'
steel dinnerware, and plastics.
Beryllium confers to zinc-base alloys reduced creep,
increased tensile strength, and improved corrosion resistance.
Beryllium-zinc is identical with cold-rolled 70-30 brass in
strength properties.
Beryllium-aluminum, ticonium, beryllium platinum,
beryllium-steel, and beryllium salts are also available.
None of these alloys, however, has a commercial demand.
Reports. reviewed indicate that approximately 225 short
tons of beryllium were used in 1970 in producing beryllium alloys
and salts.
. -
The emission faotor computed below assumes that ,a plant
is emitting 13 grams/day and that the plant operates 250 days/year
in processin~ Be(OH)2 to BeO, adding scrap, and producing bery~lium
alloys [Reference 3]. Therefore,
-1l 13
EF = 225 = --- = 14.40 grams/ton or 0.032 lb/ton.
250 0.19
3-9

-------
The National Resources Council indicates that the growth
rate in the use of beryllium alloys (using 1967 as a base year)
is about 5-IS%/year.
Beryllia Ceramics

. .
. One of the best refractory materials known is beryllia.
It has high thermal conductivity, high electrical resistivity,
high melting point, and can be fabricated by normal ceramic
processes. In the fabrication process, toxic beryllium oxide dust or
fume may be inhaled. Also the machining of ceramic parts creates
dust. Except for dust. from breakage, finished ceramic articles
, .
can be stored without much danger of emissions.
d.
According to one report a method used in producing bery1lia
, '

ceramic commences with BeO and other materials which are batched
in a large floor cistern containing water. After milling, wet
screening, and adjustments, a glaze is applied to the unfired
porcelain. The glaze becomes permanently fuzed to the ceramic
article when fired. The glaze contains about 2% Be.
Bery1lia ceramics are used in high-voltage electrical
porcelains, suspension insulators, spark plugs, and microwave
windows. BeOis used as a component in special glass for high
speed tr~ns~~ssion ~f light and as ~ liner ,in, high temperature
electric furnaces and rocket combustion chambers.
The number of ceramic plants using beryllium is reported
[Reference 3,4] to be "less than a hundred".* There are five major
- .

bery11ia ceramac plants. It is reported that these major plants
use HEPA filters to control emissions.
- -."""" -.....
ceramics..
About 22 tons of beryllium were used
The emission factor was reported as
in 1969 in beryllia
being 454 grams of
[Reference 2,3].
Be or one pound per ton of beryllium processed
~. Propellants
In the manufacture of beryllium ~olid rocket fuels, ,the
beryllium used is normally received in powder form in polyethylene
containers packed in steel drums. The beryllium powder is mixed
with other materials in a dough mixer, vacuum cast, cured, and
machined into final shapes.
During the handling of the dry BeO
. '''--'''-.u ~ .--.-.--,....___._-h
* EPA estimates that approximately 5 companies account for almost
all activity in this field.
3-10

-------
powder and during the machining operation emissions occur. The
present degree of control of the emissions in propellant fabrication
plants is reported as "high" with only 1 or 2 operations that are
active. An estimated maximum of 1,000 pounds of beryllium is used
in propellants--most of this is for the High Energy Upper Stage
(HEUS) rocket motors which are tested outside the U.S.--each test
requiring about 300 to 400 pounds of beryllium. Essentially,th'e
emission controls used are the same type used in manufacturing
beryllium containing products.
The present controls of exhaust gas in static firings
are scrubbing for intermediate size grains and absolute filtration
for very small grains. This method of control is questionable.
Arguments have been put forth that the toxicity of BeD when
"high-fired" is red'uced to a safe level. This argument appears
to lack creditability as there is no known practical way to
distinguish high-fired oxide from low-fired oxide in the exhausts

".
of experimental motors. Presently theDOD policy is to conform
with DHEW Public Health Service policy which is as follows: In
the event that the national defense requires planned, limited
emissions of beryllium combustion products to the atmosphere prior
to the availability of containment systems, the emission sites
should be remote from places of human habitation and should conform
with the following criteria: For intermittment exposure of off-site
human population to any compounds of beryllium, the maximum exposure
of 75 mg/min Be/m3 of air may be tolerated with the limits of
10 to 60 minutes, accumulated during any 2 consecutive weeks.
Reports reviewed indicate that there were approximately
ten facilities throughout the U.S. at which static firings of
beryllium-enriched motors have been conducted. Most of these
facilities have no beryllium motor static firing activity at the
present time. One company at Bacchus, Utah, currently engages
in beryllium-rich rocket grain fabrication.
3-11

-------
f.
Disposal of Solid Waste Containing Beryllium
Solid wastes contaminated with small quantities of
beryllium, but which may constitute dangerous sources of emissions, .
include clothes, rags, filters, filter aids, tar paper, mops,
kraft paper, ¥ipers of all varieties, brushes, plastic bags, etc.
A 1arg~ amo~nt of the solid waste is handled under
contract by "professional" handlers of.wastes. In most cases,
~ _. --.~.. ,'-' .,.- ..- _. - - .. &.... ~.....' "
the wast.eis.packaged in polyethylene bags or drums and labeled
,for disposal. No fir~ control as to Where the wastes are to be
disposed really exists. Other waste disposal practices that exist
-"~' --. ..---- .-- ..-..
are:
o
o
o
o
o
o
o
o.
o
Return solid waste to basic ,supplier of Be
Bury on company site
Bury in city or county dump
Bury in the desert or in landfills
Store in abandoned clay mines
Encase in concrete ,and bury when irradiated
Bury at approv~d government sites
Open-air burning at company site and then covered
Burning in company incinerators.
It is estimated that about 0.1 percent of the yearly produced
beryllium is returned to the environment as solid waste. No
estimate of emissions to the atmosphere is given. Initial
analysis of this problem indicates that emission controls vary
from none to adequate.
B.
Emission Controls
Beryllium is highly toxic in all forms, with the possible exception
of beryl, and is a serious hazard in production and use. It was not
until rather recently, since World War II, that it was recognized that
. small quantities of beryllium dust and fumes can cause a serious
chronic lung disorder in susceptible persons. In 1950, the ABC
issued standards as to permissible concentrations within its supplier
plants and in the surrounding communities. A monthly average
concentration of not more than 0.01 }Jg/m3 of bery11ium'in the 8Iltbient
air was ." established as the standard.
3-12

-------
In the discussion preceding this section, basic processes were
identified and discussed for primary beryllium production. NO~-1, the
air cleaning equipment presently being used and the associated control
costs will be covered. Block diagrams of each of the steps in the
basic processes, exhaust gas volumetric flow rates and emission
controlling equipment are shown in Figures 3-1 through 3-7. Gost
estimates for the items of air cleaning equipment as applied to
each process are given in Tables 3-2 through 3-9.
as of March 1971.
These costs are
3-13

-------
. /~
SULFATE PROCESS FOR CONVERSION OF BEltYL ORE TO PLANT-GRADE BERYLLIUM HYDROXIDE.
Proce~s Steps
Beryl ore
I'
"
112S04'
... silica
NH40H
"'alum
NaOH-
chel~ting
agents
<'II
...
cU
...
...
...
,,...
....
fORy lNG,
L::~CKAG ING .
Emission Control, Equiplncnt
r>
B
SOOOctm
oj
Hill
2 SOOc tm
[>-2 500c fm
[>
DC
SOOOcfril
FT.F
[>-4S0cfm
F'rF
[>--;MOO HST EVS'
r>-,'" cfm ,2ea .~e~ '

. ~460 PTS . , PTS
. 2ea. ," I .
dm ~~ 6400cfm
. . ., /- .
.!I.
o
cU
...
en
...
o
o
....
I
o
o
('t'\
8S00cfm
(
. (
,
'FTF - fabric tube fit ter

~VS - ejector venturi scrubber
I .
,DC - dry cyclone
"

,HST - hydraulic sCTubbing tower
, I
1 PTS - packed tower scrubber

VS - venturi scrub~er
Plant-Grade Beryllium Hydroxide )
-.-
SQurce:
Reference 1
I
Figure 3-1
3-14

-------
TABLE 3-2.
AIR CLEANER COSTS#:
SULFATE PROCESS, ORE TO Be(OH)2
      ,   
Class I 'fypc* Air **  Maintenance Powur *** Instt\11cd  Annual
  Flow Rate  Cost/Yr Cost/Yr Cost  Operating
         Cost
 .        
FTF  Shaker 5,000 cfm $ 470 421) $ 16,130 $ 890
EVS-2 ea  High Energy 1,250 ea  190 2,970 16,800  3,130
DC  High       
  Efficiency 2,500  70 210 3,460  280
F'TF  Pulse Jet 450  SO 55 2,880  105
FTF  Shaker 5,000  470 420 16,130  890
PTS-2 ea  r-1cd. HiDh    .   
  Energy 1,200 ea  165 950 11 , 520  1,1J.5
EVS-2 ea  HiDh       
  Energy 1,200 ea  165 2,860 16,800  3,000
HS'£-2 ea  Med. lI1.9h       
  Energy 1,200 ea  165 640 6,910  805
PTS  Med. High       
  Energy 6,400  450 2,530 11,520  2,980
 ,        
     $2,195 $11,055 $102,150 $13,195
. **
Among wet collectors, PTS and HST are here considered medium high energy
types--PTS because of its 0.5-1.0 in. w.g. pressure drop per ft height, and
HST because of pumping power requirements.
Actual flow rates. Rated flow rates may be as much as 30% higher. Rated
flow rates, where known, are used for installed cost estimate.
*** Makeup water is included in power cost.

FTF c'fabric tube filter
BVS c ejector-venturi scrubber
DC c dry cyclone
PTS c packed tower scrubber
HST c hydraulic scrubbing tower
*
#Costs apply to March 1971.
Source:
Reference 1
3-15

-------
FLOORIDE PROCESS FOR CONVERSION OF a'ERYL ORE TO PLANT-GRADE BERYLLIu.t HYDROXIDE
-Process steps
,Emission Control Equipment
"
.. 'q
12,600 ct'm
-' 1000 clm...
to atmosphere
FTF '
Na2siF6
Na2c03
oil.water
FTJi'
5000 clm
~
o
('Ij
+'
tJ)
..
o
o
'H
I
o
o
N
B
water.
2000 clm.....
. to atmosphere

~270~EJ-TF .'
cfm . 3900 clm--'
.-. . .... . to atmosphere
[> '. . - ,
FTF
.
ultra
collector
.
steam LEACHII\G
(NH4)2S20a
18-R
6000 clm--
to atmosphere
.... metal
salts

NaOH
Air from other processes
PRECIPITATING
FTF - fnbric tube filter-
IX: . .dry cyc lone.
VS . venturi scrubber
HST . hydraulic scrubbing tower
steam
Fn.TFJUNG
DRYING,
PACKI\G ING
Pl~nt Grade n~ry)lium Hydroxide
)
Source: . Reference I
Figure 3-2
3..16

-------
TABLE 3-3. AIR CLEANER COSTS#:
FLUORIDE PROCESS, ORE TO Be(OH)2
Class Type *  Air ** r>1aintcnance Power ***. Installed Annual
  Flow Rate Cost/Yr Cost/Yr Cost Ope-rating
 .      Cost
FTF Reverse Jet 12,600 cfm $1,200 $2,100 $25,400 $3,300
DC High      
 Efficiency 750 25 6S 2 , :310 90
DC: FTF CombiTlC!d;      
 cOllvcy.ing 1,000 130 1,240 6,910 1,370
FTF Reverse Jet 5,000 475 940 10,700 1,415
VS High Energy 2,000 155 2,500 13,860 2.655
DC lIigh      
 Efficiency 2,700 80 220 3,455 300
FTF Pulse Jet 3,900 355 470 10, 700 825
HS'f Mcd. High      
 Energy 6,000 215 1,720 23,040 1,935
    2,635 9,255 96,375 11,890
 .      
***.11-  I    
1/3 F'IF ShcJ-?i11?t
process, 1/6 from a research facJ.1J_ty, and 1/3 from a Bc(OH)2 punfJ.catJ.on
process. ..
**
«.**
****
FTF = fabric tube filter
DC = dry cyclone
V5 = venturi scrubber
HST = hydraulic scrubbing tower
DG:FTF = a unitized dry cyclone, fabric bag
also for pneumatic transfer of dust
draft 1055.
filter (manual shaker t>~c), used
collection at about 60 In. w.g.
II
Costs apply to March 1971.
Source;
Reference 1
3-17

-------
. .'
HYPorHntIcAL PLAN'r. PR~ESS: ORGANOPHOSPHATE EXTRACTI~ METHOD.
FOR CCNVERSIrn OF BERTRANDITE ORE TO BERYLLIUM HYDlWXlDE
'. .
. . Process Steps
I.
. Emission Control Equipment
    DC FTF 
H2S04   I> E} 
flocculant      
 SETTLING AND  BVS  
 DECANT ING   
NariS      
 . JRON REDUC ING .[> EVS  
!     EVS 
    1 ea per ~
EIIPA saNENT    stage u
    lIS
EXTRACT ING*     ~
    VI
(multistage)    EVS 
~rnffinatejl;HPA   1 ea per 
   stage 
--
FIt,TER ING
(Na2Be02
solution)
[>
FTF
ult ra
. collector
dry air diiution
Beryllium hydroxide cake
(to 99% purity) '.
*EHPA solvent is O.2SN di-2-ethylhexyl phosphori.c acid
with 2 wt vol percent isodecyl alcohol in kerosene
.
DC
EVS -
FTF-
dry cyclone .
ejector venturi scrubber
fabri~ tube filter
Source::,. Reference 1
Figure 3-3
3-18

-------
TABLE 3-4. AIR CLEANER COSTS II: BERTRANDITE ORE TO Be(OH)
. 2
.           
Class * Type Air Flow** Maintenance Power*** Installed Annual
  Rate, cfm Cost/Yr Cost/Yr Cost OpcTi\ting
          Cost 
DC-4 ea. High 600 ea. '$ 70 $ 200 $ 7,375 $ 270
 Efficiency          
FTF-2 ca. Shaker 1,200 ea.  235  240  11 , 520  475
EVS-16ea. High 600 ea.  685  11,410  83,200 12,095
 Energy          
  ..         
FTF Shaker 3,000   285  255  11 , 520  540
FTF Shaker 30 ,000   3,320  4,400  46,080 7,720
    $4,595 $16,505 $159,695 $21,100
*
EVS is 8" si2c, oparating at 100 psig water pressure and providing 4 in.
w.g. pressure drop.
Actual flow ratcs. Rated flow rates may be as much as 30% higher.
Rated flow rates, where known, are used for installed cost estimate.
Makeup water is included in power cost.
'..
**
***
DC = dry cyclone
EVS = ejector-venturi scrubber
FTF = fabric tube filters, of which one
precoated with asbestos floats.
.
(at 30,000 cfm) is an ultra collector,
#
Costs apply to f.larch 1971.
Source:
Reference 1
3-19

-------
COOVERSIrn OF BERYLLroM HYDROXIDE TO BERYLLIUM P.tETAL BILLETS
: .~
(Showing alternative state-of-the-art em~ssioncontrol equipment)
First Company
Emission' Control Equipment

Plant-grade.
Be(OH)2
NH4F, HF
DISSOLVING

CaC03 , PbO
'Process Steps
1000 ~
HST

-------
TABLE 3-5. FIRST COMPANY AIR CLEANER COSTS#: Be(OH) TO Be BILLETS
2
Class '!YP(!* Air ** r-1.:'1:in tl'n.U1ce I'owC!t .It.* Inst.d,lcu  AnmnJ.
  Flow Rate Cost/Yr Cas t/Yl'  Cost OpHrnt.iI1U
 .         Cost
HST Med. High         
 Ener~JY 1,000 cfm $ 70 $ 230 $ 3,120 $ 300
HST r-Ied. High         
 Energy. 1,325  95  310  3,810  405
EVS-2 ca High         
 Energy 1,250 ea.  190  2,970  16,800  3,160
VS High         
 Energy 1,600  120  1,900  11,520  2,020
EVS-6ea High         
 Energy.. 270 ea.  120  1,930  18,000  2,050
VS-2ca High         
 Energy 4,500 ca.  640 10,630  34 , 560 11,270
EVS-5ea High         
 Energy 1,.500 ea.  590  7,430  42,000  8,020
FTF Bag         
 Collapsing 9,500  900  790  20, 7 SO 1,690
DC High         
 Efficiency 600  25  55  1,850  80
   2,750 26,245  152,410 28,995
****          
1/6 FTF Shaker 1/6 x 6.5,000 1,015  1,240  13,450 2,100
.
Among wet co.Ucctors, rTS and HST are here considered medium high energy
types--PTS because of its 0.5-1.0 in. w.g. pressure drop per it height,
and !-1ST uccause of pumping pONcr rcquirements.
Actual flow rates. Hated floN rates may be as much as 30% higher.
Rated flow rates, where known, are used for installed cost estimate.
Makeup water is included in pow(:r cost.
This fabric tube filter is ternwd an "ultra collector." It is a shaker-type
compartmented filter, having orlo~ bags precoa~ed by asbestos floats. The
u1 tra collector. is a secondary clc-:mer of "dry" exhaust gases. It handles
about 65,000 cfm, distributed 1/3 from the fluoride proccss, 1/6 from the
Bc(OH)2-to-bi11et-process, 1/6 from a research .facility, and 1/3 .from a
Bci(OH).., purification process.
...
HST'= hydraulic scrnhbing to\~cr; EVS = ejector venturi scruuber
VS = venturi scrubber; FTP = fabric t.ube filter
nc = dry cyclone
*
**
***
****
II CostsaLJply to H.:\rch 1971.
Source:
Reference 1
3-21

-------
." .'
. -
>'.
'..1-
CONVERSION OF BERYJ~LruM BILLETS TO BERYLLnJM METAL FORMS
Process Step's
BmissionControl Equipment
. .
spent salt recovery
berylli~m billets

PICKLING,
WASHING
scrap chips
~lCOO
, ctm
J/.
o
10
ofj
en
ofj
,It-f
I
o
\0
M
43,000. c,fm
FTF
2ea
to atmosphere
.>



~ Il~;a~.200 elm

, DC - dry cyclone
.
Finished Beryllium Forms
HST - hydraulic scrubbing
tower
FTF - fabric tube filter
~ource:
Reference 1
Figure 3-5
3-22

-------
TABLE 3-6. SECOND COMPANY AIR CLEANER COSTS#: Be(OH)2 TO Be BILLETS
ClaEoS Typc* Air** Maintenance Pow~r*** Installed Annual
   Flo\' Ra.te C.ost/'lr Cost/'lr Cost Operating
 .     Cost
PT5 Med. High 7,000 cfm $ 500 $ 2,750 $ 18~470 $ 3,250
 Energy   
FTF Pulse ,Jet  1,500 140 200  .340
PTS-9ca toted. High 160 ea 215 620 14,530 835
 Energy  ;   
05 Lm\' Energy 10,000 710 1,580 8,650 2,290
EV5-2ea High Energ 1~200 ea 180 2,860 16,800 3,040
PT5-2ea Med. High 7,000 ea 995 5,545 34,560 6,540
 Energy     
FBS-2~a toted. High 1,200 ea 165 475 8,::K)0 640
 Energy     
P1'5 f.1cd. High 21,000 1,495 8,315 27,700 9,810
 Energy     
FTF Shaker 17,000 1,610 1,~10 25,370 3,020
.i    6,010 23, '/55 154,380 29,765
*
Among wet collectors, PTS and HST a.rc here considered m~diul11 high ener.gy
types--PTS because of its 0.5-1.0 in. w.g. pressure drop I~r ft height,
and HST becD-lIse of pumping power requil:croents.
Actual flm'l rates. Rated' flow rates may t>e a.s much as 30% higher.
Rated flow rates, where known, are used for installed cost estimate.
Makeup water is included in power cost.
**
***
PTS
FTfo'
OS
EVS
FBS
=
packed tower scrubber
fabric tube filter
orifice scrubber
ejector-venturi scrubber
floilting bed scrubber
=
=
=
=
# .
Costs apply to f.1ilrch 1971.
Source:
Reference 1
3-23

-------
CONVERSICX'l OF PLANT-GRADE BERYLLIUM HYDROXIDE TO ALLOYS
(Showing state-of-the-art omission control equipment)
Process Steps,'
Bmission Control Equipment
P1nnt-grado
. 8e(01l)2
. a

O()O c:f'm
DROSS
STOR ING
-'l
o
CIS
....
U1
....
o
o
'H
, I
11'1
t-.
roof
dross
, .
22,000 cfm
HE/\T TREATING,
SHAPING AND
FINISHING
1> 6000 c fm
[> 2200 cfm
[>.--J12,000 cfm
~4200 cfm
~
..... to atmosphere
....
.
4% Be master alloy
t
4% Be master alloy
copper chips
.
~
o
CIS
....
U1
....
o
o
~
I
o
11'1
-'-2% Be alloy
. .
.f>-7S00~' . 12,000 elm

PSC FTF
. 3en ., .
. I
.F'ff - :fabric tube filter
'PSC particle settling chamber
Finished 2~ Be stockf~rms
Source:
Reference 1
Figure 3-6
3-24

-------
II
TABLE 3-7. AIR CLEANER COSTS: Be BILLETS TO Be METAL FORMS
Class Type *   Air,lH(.  1 Maintenance I Powcr*** Installed Annual . '
 "   }'"lo\'l i~a te . Cost/'ir Cost/'ir Cost Operating
.      , I    Cost  
        ,      
 .             
 ..      ,.       
DC High'   1,000 cfm $ 35  $ 90 $,2,300 $ 125 
 Efficiency    , .   . ..'   
       ,    .,   
HST Med. High 6,000   215   1,715 23,040 1,930 
 Energy    .,       
           t ; '"'   
DC-10ea HiUh   600 ea,  320.   900 ~3,180 3.,220'
 Efficiency           
     ,  6,270, .    
FTF-2ea Reverse Jet 21,000 ea 4,070  50 , 700 10,3110 
 , ,     4,640  8,975 109,220 1'3,615 
*
Among wet collectors, PTS and liST arc here considered medium high enertlY
. types--PTS because of its 0.5-1.0 in. w.g. pressure drop per ft height, ,
. and liST beCa1l5C of pumping powel' requirements. .
ActUrtl flow ratf!S. Ri\tC:'d flow :,ates may be as l1luch as .30% higher.
Rated flow rates, where known, arc used for installed cost estimate.
. ,Makeup water is included in pow~r cost.
**
***
DC
liST
FTf.'
=
dry cyclone
hydraulic scrubbing tower
:fabric tube filter
=
:::
#
Cost~ apply to March 1971.
.
Source:
Reference 1
3-25

-------
COOVJms t<:N OF mmYLLWM IIYDROXIDRiTO BERYLLIUM OXYDE PCMDSR AND CERAMICS "
Process ~teps,
Emission Control Bquipment
. '
high purity
Be(OH)2
El
li
1100 C£"l


-20,000cfm
, ",..".",......" ,
~~~~::P 1700 C£" ... ....
81800~ .. .... ... . ... . 2BOO ctm
gr'300 cfm--- to atmosphere
. '
'PRESSING . ' , '.
20,000cfm , '



[>-, 7150 Me , ' '
6ea' "',
. . .'

~' ,',', '"
.
Finished ceramic forms
FTF .' fabric tube filter
PTS -packed tower scrubber
MC mist collector
Source:
Reference 1
Figure 3-7
3-26
.
,!II!
U
1\1
...
CII
...
\j.f
,
o
o
('11
.
.!II!
U ' .
1\1 '
...
U!
....
....
I
o
Ion
-+
.!II!
u
1\1
...
U!
....
\j.f
I
Ion.
t--

-------
TABLE 3-8. AIR CLEANER COSTS#: Be(OH)2 TO ALLOYS
Clnss Type Air* M...intcnance Power** I Installod Annual
  Flow R~te Cost/Yr Cost/Yr Cost Operating
 .       Cost
FTJ.' Shaker 1,500 cfm $ 140 $ 130 $ 5,760 $ 270
FTF-2ea Shaker' 2,500 ea  475  420 16,130  895
 . ,        
FTF  400 ea  35  35 2,880  70
DC High 5,000  155  420 ' 5,070  575
 Efficiency        
DC High 400  12  35 1,380  47
 Efficiency        
FTF-2ea Reverse Jet 11,000 ea 2,090 3,190 41,470  5,280
PSC-3ea Low 2,500 ea  95  35 920  130
 Efficiency        
       ..  
FTF Reven;c Jet 12,000 1,135 1,740 21,890 2,.875
   4,137 6,005 95,500 10,142
*
Actual flow rates. Rated flow rates may be as much as 30% higher.
Rated flow rates, where known, arc use~ for installed cost estimate.
Makeup water is included in power cost.
**
.
FTF =
DC =
PSC =
fabric tube filter
dry cyclone
particle settling chamber
#Costs apply to March 1971.
Source:
Reference 1
3-27

-------
 TABLE 3-9. AIR CLEANER COSTS#: Be(OH)2 TO BeO AND CERAMICS  
Class Typo- Ai:r*- .  Maintenanco. Power-.. Installed Annual
  Flow Rate /' Cost/Yr .'. Cost/Vr Cost Operating
          Cost
FTF Shaker 1,000 clm  '$ 105  $ 155 $ 6,340 $ 260
PrS Med. High 3,000   215:'  1,190 11 ,520  1,410
 Bnergy        
PIS , Med. High 5,000 .   355   1,980 16,140  2,335
 Energy          
PTS Med. High 12,000 . .   855  4,750 25,370 . 5,605
 Energy          
FIF-2ea . Reverse Jet 300 ea   60   . 45 6,920  105
FTF Shaker 1,100   105'  155 6,340  260
FTF Reverse Jet 1,800   180   265 7,490  445
FTF Pulse J0t 300   35   45 2,300.  80
MC-6ca. Mi~t 7, 150   215   980 4,610 1,195
 Collector          
. .
2,125
9,565
87,030
11 , 695
-
Among wet collectors, PTS 'and HST nre here considered lI1edium high energy
types--PTS because of its 0.5-1.~ in. w.g..pressure drop . per It height,
and HST because of pumping power requirements.
Actual flow rates. Rated fl~~ rates may be as much as 30% higher.
Rated flow rates, where known, arc used lor installed cost estimate.
Makeup water is included in power cost. .
..
***
FTF
PTS
MC
.
= fabric ~ube filter
= packed tower scrubber
I: mist collectors, for operations such ~s '.'we.t..grinding,"
625 cfm, one for 1,050 ,cfm, .and one . fo;: .;3.600 clm.

"'Costs apply to March 1971.
four sized for
II
Source:
Reference 1
\'
3-28

-------
10.
11.
12.
BERYLLIUM REFERENCES
1.
Environmental Protection Agency, Bureau of Stationary Source
Pollution Control, Division of Compliance; Control Techniques 
for Beryllium Emission, First Draft Copy, Durham, North Carolina,
July 1971.
2.
Lev!n, Harry. .A.BasisforNationalAirEmission Standards on
Beryllium. Camarillo, California: Litton Environmental Systems,
April 7, 1971.
3.
Levin, Harry. Interim Submittal on Major Sources of Beryllium
Emissions. Camarillo, California: Litton Environmental Systems,
February 28, 1971.
4.
u.s. Department of the Interior, Bureau of Mines. Mineral Facts
and Problems, 1970, Bulletin 630. Washington, D. C.: U.S. Government
Printing Office.
5.
"National Inventory of Sources and Emissions, Arsenic, Beryllium,

Managanese, Mercury, and Vanadium," Beryllium, Leawood, Kansas:
W. E. Davis & Associates, April 1971.
6.
Celenza, G. L. "Designing Air Pollution Control Systems," Chemical'
Engineering Progress, November 1970, pp. 31-40.
7.
U.S. Department of Health, Education, and Welfare, Public Health
Service. Air Pollution Engineering Manual. (PHS Pub. No. 999-AP-40).
Cincinnati, Ohio: 1967.
"
8.
Mancuso, Thomas F. "Relation of Duration of Employment and Prior
Respiratory Cancer Among Beryllium Workers." Environmental Research
Volume 3, Number 3. New York and London: Academic Press, pp. 251-275.
9.
and Francis L. Bunyard. A Systematic Procedure
Cost of Controlling Particulate Emissions from
U.S. Department of Health, Education, and
Edmiston, Norman G.
for Determining the
Industrial Sources.
Welfare, June 1969.
u.s. Department of Commerce. 'Trends in Usage of Beryllium and 
Beryllium Oxide (AD 382 579). Washington, D. C.: National Research
Council, February 1968.
u.s. Department of Interior. Materials Survey - Beryllium.
D. C.: Bureau of Mines, September 1953.
Washington,
Beryllium Technology, Metallurgical Society Conferences, Volumes 1
and 2. New York: Gordon and Breach, Science Publishers, Inc., 1966.
3-29

-------
26.
27.
28.
29.
13.
Kjellgren, ,Bengt R. F. '~Beryllium" 'Rate 'Metals 'Hartdbook, ,Second
Edition. Hampel, Clifford A." (ed.) ,Reinho1dPublish~ng, Corp.
London: . Chapman & Hall, ,Ltd., 1961;pp. 32-57.
, .. ... .. .." ,. "., .

14. White,:,D,.. W.t'Jr.;and J. E..Burke.(ed.);"Hea1th:Hazards,":The
, 'Metal : Beryllium. : ClevelandJOhio: . : The American Society. for1fetals,
1955, pp. 620-640. . .
,. ., ,. .. .. . .. ,. . " .
15.
F1y,M. A.and,E.. L."Semenova. ' 'Hartdbook 'of 'the 'RareE1emertts; 'Trace
"Elemertts'and'Light'E1ements, Volume 1. MacDonald & Co., Ltd.,
pp. 167.
16.
McMillan, P. W.; Glass-Ceramics, Vol. 1. Nelson Research Laboratories.
Stafford, England: English Electric Co., Ltd., 1964.
17.
Siemens-Konzern, Rimbach, and Michel.
Corp., 1929.
Beryllium.
Reinhold Publishing
18.
"Engineering Outline 142, Solid Propellants," Engineering, Volume 206,
UDC 662.3. September 13, 1968, pp. 405-408.
19.
"Coor's Volume Tops Million Pounds per Month," Ceramics Industry,
Volume 91. October 1968, pp. 39-41.
20.
"Forecast '70," Ceramics Industry, Volume 92.
June 1969, pp. 43-49.
21.
"ARC Hard at Work on Development of Beryllium Powder," Missiles 
and Rockets. May 27, 1963, p. 27.
22.
"Dust Collection Pays Its Way," American Machinist, Volume 114(2).
January 26,1970, pp. 114-118.
23.
Schilling, S. A., "Beryllium," Engineering and Mining Journal, Volume
171. March 1970, pp. 116-117.
24.
"Design and Application of Be Structures," Beri11ium Technology,
Metallurgical Society Conference, Volume 2. pp. 972-975.
25.
"Beryllium Grows Up, Thanks to New Designs," Product Engineering,
Volume 40. December 1, 1969, pp. 67-68.
"Beryllium and Its Alloys," Engineering, Volume 210, No. 5447.
October 2, 1970, pp. 360-363.
Weaver, Henry C.
"Beryllium," Minerals Yearbook, 1968, pp. 207-209.
"Modified Process Gives Superior Beryllium," SAE'Journa1, Volume 76,
No.9. September 1968, pp. 66-67.
"Making the Most of Beryllium," Battelle Technical Review, Volume 17,
No.9. September-October 1968, pp. 15-19.
3-30

-------
30.
"Atmospheric Monitoring of Toxic Levels of Missile Propellants,"
American 'IndtistrialHygene 'AssoCiation'Jotirnal. February 1964,
pp. 77-80.
31.
McSouth, M. E. and J.,P.Terry., "Air, Pollution Control at Cape
Kennedy," American Indtistrial'Hygene'Association'J6tirnal ,Volume 26,'No. 2.
April 1965, pp. 172-176.
32.
"Environmental Engineering in Handling Toxic Materials," Air Engirteering,
Volume 9, No. 10. October 1967, pp. 30, 31, and 33.
33.
Hasenc1ever, Dieter. "What May Be Demanded of High Efficiency Filters?"
Staub, Volume 26, No. 10. October 1966 [English Edition], pp. 22-26.
34.
Lund, Herbert F. "Industrial Air Pollution Control Equipment Survey:
Costs and Procedures," Journal of the Air Pollution Control Associa-
tion, Volume 19, No.5. May 1969, pp. 315-321.
35.
"Beryl-Salient Statistics," Chemical Economics Handbook.
California: Standford Research Institute, July 1969.
Menlo Park,
36.
"Beryllium: Promise and Problems," Iron Age, Volume 206, No.6.
August 6, 1970, pp. 61-63.
37.
"Takeover Candidates," Finance.
June 1969, p. 56.
38.
u.S. Department of Health, Education, and Welfare, National Air
Pollution Control Administration. Preliminary Air Pollution Survey
of Beryllium and Its Compounds, Washington, D. C.: u.S. Government
Printing Office, October 1969.
39.
Spangler, C. V., Trip Report to Burch Beryllium Corporation at Delta,
Utah, 29 April 1970.
40.
"Beryllium - Hazardous Air Pollutant," Environmental Science and
Te'chno1ogy, Vol. 5, No.7, July 1971. pp. 584 & 585.
41.
Brodovicz, Ben A., "Air Quality Criteria for Pennsylvania," Journal
of the Air Pollution Control Association, Vol. 18, No.1, January
1968. pp. 21-23.
42.
Rehm, Alexander, Jr., "New York State's Classification - Ambient Air
Quality Objectives System", Air Pollution Control Association Journal,
Vol. 15, No. 11, pp. 519-522.

Magan, John A. and John R. Goldsmith, M.D., "Standards for Air Quality
in California," Air Pollution Control Association Journal, Vol. 10,
No.6, pp. 453-455, 4467.
43.
3-31

-------
CHAPTER 4
MERCURY.
1.
Introduction
Mercury is a silver-white metal, the only metal which is a liquid at
most temperatures, normally encountered in the environment. Other prop-
erties which make it useful in various applications are its ability to
amalgamate (dissolve) with nearly all other metals (iron being a useful
exception), uniform volumetric expansion coefficient (as a liquid), high
electrical conductivity, and (for some applications) toxicity.
Unfortunately, the toxicity of mercury and its compounds make it a
hazardous substance to use, especially where its toxicity is not the
reason for its use. Mercury is a relatively inert chemical. Its poison-
ing action is usually similar to that of other heavy metals (antimony, lead,
etc.) in that it is a cumulative or chronic poison. Large doses of the metal
or small doses of some of its more toxic compounds, however, can produce
acute poisoning.
Persons seem to vary in their tolerance to mercury as a poison. One
historical report claims that a person swallowed one pound (about two table-
spoons) of mercury as a laxative; the treatment worked, after about three
days, and there were minimal reported side effects. Other medical reports
of persons swallowing, inhaling, or injecting metallic mercury "accidentally",
in connection with work, or to commit suicide, indicate a wide range of
toxic levels. Specific effects include skin sores, lung damage (when in-
haled), kidney and brain damage, and psychological symptoms ("hatter's
disease"). The wide variation in intake levels required to produce clinical
mercury intoxication, particularly in chronic exposure cases, appears to
be ~ependent upon the body's tolerance level and the slow but variable rate
of excretion of the mercury rather than the specific exposure rate.
Elemental mercury has been identified by EPA as a hazardous atmos-
pheric emission for several interrelated reasons.
First, of course, is
4-1

-------
its known toxicity. Second, mercury is highly volitile leading to
industrial hygiene problems. The most widely-used method for maintaining
low ambient air concentrations (50 g/m3 TLV or less) inside plants is to
ventilate work spaces; with large amounts of air, thereby moving the
mercury from inside to outside. There are frequently large process
losses to the atmosphere in conjunction with the ventilation losses, or
even where there are no ventilation losses. Industrial hygiene threshold
. .
limit values (TLV) are bas~d upon working-day exposures; workers who live
near sources of atmospheric mercury. are therefore exposed in excess of
the exposure assumed by the TLV. Third, it is believed by some research-
ers that, no matter how mercury is freed to the environment, it eventually
ends up in the waters, where it can become concentrated in the aquatic
food chain in particularly toxic forms (alkyl mercury compounds), endanger-
ing both higher animals and humans. Fourth, with the general population
at risk, however small the risk, consideration must be given to the
possible, and as yet unknown, synergistic effects of mercury and other
environmental pollutants, particularly the other heavy metals.' Finally,
mercury is a widely-used industrial material.
The sources of mercury emissions to the atmosphere which have been
analyzed for this project are: 1) primary mercury extraction plants;
2) secondary mercury recovery and refining plants; and 3) chlor-alkali
production plants using mercury cells. These sources have been studied
intensively by EPA, and proposed emission standards have been published.
. ,

It should be noted that the estimated total atmospheric emissions of mercury
from these sources in 1970 were about 50 tons, 60 pounds, and 70 tons,
respectively. This is small compared with approximately 500 tons from coal
combustion (assuming average mercury content of coal is 1 ppm) and about
150 tons from incineration and other waste disposal not associated with
" -

mercury processing. However, emissions per source are much more localized
from the 18 largest mercury smelters and 32 mercury-cell chlor-a1ka1i
plants than the 400 fossil-fueled power plants and the thousands of incin-
erators and other waste disposal sites.
4-2

-------
Additionally, technology for adequate control of direct mercury
emissions to the atmosphere from. the three types of sources studied is
now commercially available or soon will be. It is therefore rea-
sonab1e to begin economic analysis of cost impacts generated by control
of these mercury emissions. The object of this portion of the final
report is to present the basis used for estimating control costs to each
of the three industries, the results of imposing those controls, and
analyses of the expect~d economic impacts.
!
The remainder of this chapter discusses, for each of the three in-
dustries in turn, the operating techniques and parameters of the industry,
the resulting uncontrolled emissions, applicable emission control
techniques, emission standards, current control practices and levels of
control (1970), costs of controls necessary to meet the standard, resulting
emissions, and the various economic impacts of emission controls.
4-3

-------
II.. Pr!~r}7Mer~ury Extraction
Occurrence and Recovery Techniques
Me~cury is,foun~ in many minerals throughout the earth's crust
including severaLthat contain commercial quantities of other metals.
The primary commercial source of metallic mercury throughout the
world, however, is red cinnabar, which contains mercury as the sulfide.
Ores used for commercial p~oduction in the U. S. typically average
4-5 lb of mercury per ton; a few'~xtr~~tionplants have oc~asio~a~l}7_-
,p'r0d.~:ed f.r~~C?Ee, as low as t.w~ pounds per ton.
In the United States mercu~y is pr~duced exclusively by pyro-
metallurgical processes, either by indirect or direct fired methods.
, Indirect firing (retorting) generally requires ores of 20 pounds or
more of mercury per ton to be economically attractive. Little if
any such ore is available on a regular basis in the U. S. Virtually
all primary, or "prime virgin" mercury produced in the U. S. comes

.
from ores that average 2-8 lbs of mercury per ton, produced by direct~
fired processes.
The dominant 'direct-fired process used in the domestic primary
mercury industry involves-use of a rotating kiln inclined 5° to 15° above
horizontal. Firing is done with the hot gases flowing up the kiln
and ore flowing downward. Thirty to fifty percent excess air is ,
admitted to ensure that the sulfur is oxidized, but still ensuring
that temperatures remain well above the boiling point of mercury
(357°C) until the gases reach the condenser. A pull-through hot
fan is used to prevent escape of mercury vapor by. providing a slight
negative pressure to the kiln at all points. Spent or burnt ore
leaves the kiln at about the sublimation point of cinnabar, 583.5°C.
(1_~_~2~3.°F.)'; The burnt ore is removed to a dump pile after natural, con-
vact1ve__cooling.
A.
After passing through the kiln, the hot gases, containing mercury
vapor, entrained dust and soot, and corrosive gases such as sulfur
oxides, pass through a cyclonic collector to remove most of the
particulates. The captured dust may be treated for recovery of
4-4

-------
adsorbed mercury by returning it to the kiln or to a separate retort.
From the cyclonic collector the gases pass into the condenser for
mercury recovery.
The condenser typically consists of two banks of vertical inverted
U-tubes. Cast iron condensers predominate, although, a ceramic first stage
is used in some plants. Mercury is condensed as the gas is cooled; cooling
of the condenser is usually accomplished by ambient air, but (especially in
hot weather) may be aided by spraying the condenser tubes with water. The
condensed mercury collects at the bottom of the condenser tubes, where it
settles into water in a container called a "launderer." The gases leave the
condenser proper at 32°-49°C normally, saturated with mercury vapor.
Generally, the gases then pass into a redwood settling chamber, which may
contain baffles to assist in removing the entrained mercury droplets as the
gases cool further by contact with the walls and by expansion.
Direct
water sprays are sometimes used to cool the gas.
a stack to the air.
At this stage, the mercury is usually contaminated with soot and
other particles that escape the cyclone collector. It is, therefore,
transported to a hoetable where it is mechanically mixed with lime
to clean it. The cleaned mercury coalesces, collects at the low point
of the table, and is bottled in iron flasks (76 lbs of mercury
per flask) for shipment. This prime virgin mercury is pure enough
Gases then pass out
for most uses.
Hydrometallurgical processing of mercury ores has been investigated
and shows some promise, in that recovery cost for the two methods are
comparable. The industry has shown little interest, however, primarily
because of the technical complexity of the process as compared with
roasting, and the high initial investment. If regulation of atmospheric
emission of mercury from roasting imposes major control costs, the
virtually emission-free hydrometallurgical process may become more
attractive.
B.
Emissions
Atmospheric emissions of mercury during the roasting process
may occur at several points, including the burnt ore bin and bin
4-5

-------
discharge, tailing dump, hoetable,and condenser exhaust. These
emissions are essentially mercury vapor at concentrations below
saturation vapor pressure except for the condenser exhaust, which
is saturated with vapor and may contain entrained mercury droplets
I
and mercury-containing particulates as well.
Emissions from the hot burnt ore can be virtually eliminated,
and operating costs reduced, by using the hot rock to preheat
the combustion air. Furthermore, yield of mercury might be
increased slightly. Emissions from the tailing dump can be
reduced by burying the tailings (returning it to worked-out
portions ,of the mine); this would also reduce water pollution
problems that typically occur in connection with tailing dumps.
The hoetable can be controlled by enclosing it and using the
requisite ventilation air as combustion air, a~ain increasing yield
slightly. Siimi1ar techniques may be used at otha-r '4passive~'
emission points.
The most significant source of merct.1ry emis.ions 'by, fa,r, howeftr,
is the condenser exhaust. Hot gases cominS directly from the kiln
are generally not aatt.1rated with mercury vapor; however, since tbe
method of recovery involves cooling the gas bel0W tba boiling point
of mercury and its resultant condensation, the stack aaseswi1l be
saturated with vapor, at least at the minimum temperature reacbed,
and may a180 contain entrained droplets of mercury and mercury-
containing ash and dust. The emissions calculated are basad on lOS&QS
of mercury as the,8aturated vapor only, and therefore are the minimwm
emissions expected.
Data made available to RTI by EPA indicatathat, with 50% e~cass
stoichiometric air, 470 m3 of air are exhausted per ton of ore at
the typica~ condenser exhaust temperature of 49.C (120.F). The
saturation concentration of , mercury at 49.C (120.F) is .118 81m3
(7.45 X 10-6 1b/ft3). Stack emissions of mercury vapor per ton
of ore processed are. therefore approximately 56 g (.12 lb). The best
control currently used at any mercury plant involves furth~r cooling

,
of the gases to 13°C (55°1) by u&e of direct water spray iata the
. .

gas in settling chanbers. With th. contdned redaction in 8M w11lll'Jli
4-6

-------
to 417 m3/ton and the lowering of vapor pressure to .0073 g/m3,
the resulting mercury vapor emissions would be 3 g/ton. With the
assumption that 2.25 kg (5 1bs) mercury per ton of ore is the
average content of U. S. ore, and noting that U. S. prime virgin
mercury production in the U. S. in 1970 was 27,281 (76 1b) flasks,
approximately 410,000 tons of ore were smelted in 1970. Nationwide
1,240 to 23,100 kg (2,750 to 51,200 1b) of mercury would then be
emitted by smelters from exhaust stacks alone. The actual figure is
quite likely closer to the larger value, since (1) emissions rise
rapidly with increasing temperature, and (2) no estimate of entrained
droplets and mercury containing particulates has been included.
Independent estimates based on other data supplied to RTI by
EPA indicate that 2 to 3 percent of the mercury originally in the
ore is lost in stack gases, or 18,000 to 24,000 kg (40,000 to
60,000 1b) in 1970.
C.
Emission Standard
The mercury emission standard selected for this study of
primary mercury recovery is .95 g/ton (.0021 1b) per ore roasting
(0.0021 1b/ton) facility and in no case more than 1156 g/day (2.56
1b/day).
The standard proposed subsequent to the completion of this study
allows 2,300 grams (5.0 pounds) per 24 hour period. This should reduce
control costs very significantly from the control costs estimated in
this study.
D.
Emission Control Techniques and Resulting Control
Control techniques which may be applied to the exhaust gas
stream include: (1) further cooling followed by demisting;
(2) alternative (1) followed by treated activated charcoal (TAC)
adsorption; (3) conversion to hydrometa11urgica1 recovery;
(4) chemical scrubbing using sulfuric acid.
Alternative (1) involves an extension of current recovery
techniques used by the industry, and is therefore least likely
to cause technical difficulties or industry objections.
Alternative (2), if required for additional control,
4-7

-------
likewise-presents few technical problems,but has apparently
not been tried with S02-contaminated gases.
Alternative (3) is a known technology involving floatation
separation and electro~ytic reduction; it is technically much more
complex and has been shunned thus far by the U. S. industry. It
should, however, drastically reduce emissions.
Alternative (4) has been reported in use at only one plant,
in Finland.' The process involves passing the cooled gas through a
demister-type tower with a countercurrent flow of 90 percent
sulfuric acid over the demister elements; this is followed by
scrubbing of the gas with a 30 percent sulfuric acid solution.
Passing the gas through the demister causes the mercury (and
selenium) present in the gas to be converted to insoluble sulfides;
the scrubbing process removes entrained acid mist and sulfur oxides.
The mercury- and selenium-bearing acid is treated in a thickener, with
clarified and strengthened acid returning to the proce~s. The mercury
selenium sludge is retorted to recover mercury and the residue. is sent
to a selenium recovery plant.
Cost have been developed for alternatives (1), . (2), and (4).
Alternative (1) will not meet the emission standard of .0021 pounds
per ton of ore with a maximum of 2.56 pounds per day. Alternative (2)
is capable of 'meeting the standard with a comfortable margin, with
as little as .0004 pounds of mercury emitted per ton of ore, and
allows a plant of up to 7,000 tons .of ore per day to operate within
the 2.56 pounds per day maximum. (A 7,000 tons per day plant would
produce about 166,000 flasks per year.) Alternative (3) involves
replacement of an entire plant's equipment. While the investment
c~sts are c~mparable to those of new pyrometallurgical plant equipment,
required operator retraining and technical complexity make this
alternative extremely unattractive to the industry. The same comments
apply to alternative (4), plus the comment that its economics appear
to depend upon selenium recovery, Which requires additional capital
expense and which may not be feasible if U. S. mercury ores do not
contain selenium or some other recoverable metal.
4-8

-------
E.
Costs of Control
Assuming selection of alternative (2) as described above, the
costs shown in Table 4-1 are estimated to apply to a "large" plant
(average of the 13 largest producers in 1970) of 72 tons ore/day
processed, giving production of 4.75 flasks/day or 1,710 flasks/yr.
It is reasonable to assume that the mercury collected in the
water-spray-demist operation is readily recoverable at almost 100%
efficiency with essentially no: additional cost. At the 1970 average
price of $404 per flask, the additional 40 flasks recovered for sale
are worth approximately $16,000. Thus, this portion of the control
is more than able to pay its own way.
TABLE 4-1.
CONTROL COSTS FOR MERCURY EXTRACTION MODEL PLANT
 Investment Annual Operating 
Control Step Costs($l,OOO) & Maintenance Charge ($1,000)
Cooling and $35.0 $10 . 0 . 
Demis ting 1./   
TAC Adsorption~/ 9.9 15.5 
Totals $44.9 $25.5 
1/ Cooling and Demisting: The gases are cooled to 4°C (4~OF)
by a direct spray of chilled well water, followed by demistl.?g
to removed condensed mercury droplets and mercury-laden water
spray. Operating and maintenance costs based on $6 per flask
produced.
2/ TAC Adsorber: Annualized costs are based on one carbon
change per year, with no recovery value for the carbon.
4-9

-------
F.
. E conomi c Impac t
1.
Industry Structure
Seventy-nine mines reported mercury production in 1970 compared
to 109 in 1969. 'fhis isa reduction of about 28 percent. Domestic
production was 27,303 flasks having a value of $11.1 million.
This is an average of about 346 f.lasks per mine having an approximate
value of $141,000.
Twelve mines produced 1,000 flasks or more, 5 produced between
500 and 999 flasks, and 10 produced 100 to 499 flasks,with 52
producing the remainder. California was the major producing
state, contributing 68 percent of domestic mercury production,
followed by Nevada, Idaho, and Oregon. Table 4-2 shows the
location of U.S. mercury reserves.
Many of the mines that became inactive were small producers
who found it uneconomical to continue production at the prices
that prevailed during 1970. Production decreased only about
.

8 percent indicating a trend to fewer but larger producers.
Some exploration and development work continued during 1970,
however. Ore quality declined slightly with recovery averaging
about 4.8 pounds per ton treated.
The mercury mining industry is operating well below the
levels possible at higher prices. Since mid-1970 demand has
. declined leading to lowered prices and lowered output. While
1970 production was only 8 percent below 1969 production, 1971
production is expected to be about 33 percent below 1969 production.
The dependence of production upon price is illustrated by Table 4-2
which estimates the U.S. mercury potential at various price levels
(based on 1961 prices, costs, and technology).
Recovery of mercury as a by-product from the smelting of
other ores is becoming significant although it has not been
in the past. In 1969 280 flasks of mercury were recovered from
.zinc are (mostly of New York origin) by a Pennsylvania smelter.
In 1970 similar mercury by-product recovery from copper and
silver ore was started in Ireland. There an annual recovery of
about 1500 flasks is expected, although the mercury content of
the ore is reported to be so low that it has not been assayed.
4-10

-------
Table 4-2. THE MERCURY POTENTIAL OF THE UNITED STATES AT
SELECTED PRICE LEVELS 1/
   Mercury price per flask (76 pound flasks)
State $100 $200 $300 $500 $1,000 $1,500
Alaska  14,500 33,500 69,500 124,500 142,500
Arizona   2,500 8,000 12,000 13,500
Arkansas    4,500 6,500 7,500
California 39,000 112,500 258,000 566,500 853,000 941,000
Idaho   20,500 37,500 65,000 72 ,000
Nevada 7,000 13,000 34,500 71,500 115,500 157,500
Oregon   12,500 34,500 55,000 71 ,000
Texas   17,500 35,000 54,000 59,.000
Utah     500 500
Washington    500 1,000. 1,000
Total2/ 46,000 140,000 379,000 827,000 1,287,000 1,465,500
Based on 1961 costs and technology.
1/
~/
Totals are cumulative.
Source:
Reference 5
4-11

-------
2.
Market Structure
Mercury, sold in 76-pound flasks, is traded on a New York
commodity market, , However, about 90 percent of sales is. priv~t::e.:::.._--
sales between producer and user with the price established from
that reported by the commodity market. U. S. production usually
makes up about 40 percent of U. S. consumption with imports
accounting for the remainder of domestic consumption. A duty
of $12.92 per flask was imposed on imports until January 1, 1971
when the rate was reduced to $11.40 per flask, in accordance
with provisions of the General Agreement on Tariffs and Trade.
This amounts to approximately five percent of market value.
Table 4-3 indicates the major uses for mercury. In this
table any mercury required for initial inventory and start-up
or expansion of electrolytic preparation of chlorine and caustic
soda is included in "other". Consequently, Table 4-3 substantially
understates the importance of chlorine and caustic soda. This
has been the largest single end use for mercury when purchases
for both current consumption and initial inventories are considered.
Use in electrical apparatus is second in importance. Almost all of
this is used for 'mercury batteries, by one manufacturer. The
third largest use is for mildew proofing and antifouling paint.
These three end-uses comprise over two-thirds of total 1970
domestic consumption. Because of its importance and the pollution
problems involved, electrolyt~c production of chlorine and
caustic soda will be discussed separately in a later section.
The mercury market is not at all concentrated on the buyer's
side. There are a number of companies using mercury in or for
production of a number of widely varying products. The largest
purchaser consumed less than 5 percent of world production. For
most users, the cost of mercury is a very minor part of total
production cost.
In mid-1970, the Federal Government charged eight chlor-alkali
producing companies with dumping significant amounts of mercury
into waterways. Other significant users of mercury, such as
4-12

-------
Table 4-3.
MERCURY CONSUMED IN THE UNITED STATES, 1970
Use
Consumption
(76-pound Flask) Percent
l/?/
Agricu1ture- =-
1,811 3.0
219 0.4
2,238 3.6
2 , 2 86 3.7
15,952 25.9
15,011 24.4
1,806 2.9
4,832 7.9
Amalgamation
Catalysts
Dental Preparations
Electrical Apparatus
Electrolytic Preparation of Chlorine
and Caustic Soda
General Laboratory Use
Industrial and Control Instruments
Paint: 2/
Antifouling
198
0.3
Mildew Proofing
10,149
16.5
?/
Paper and Pulp Manufacture=-
226
0.4
Pharmaceuticals
690
1.1
Otherl/
5,858
9.5
Total Known Uses
61,276
Total Uses Unknown
227
0.4
/
Grand Tota1-
61,503
100.0
1./
2/
Includes fungicides and bactericides for industrial purposes.
1971 and 1972 decisions by EPA will essentially eliminate these
markets.
1/
Includes mercury used for installation and expansion of chlorine
and caustic soda plants.
4-13

-------
agriculture and the paper industry, have since been charged with
contributing to pollution of the environment. Subsequently, the
total use of mercury has dropped about one-third. No new chlor-
alkali plants using ~ercury cells have been completed since early
1970 and at least thtee have closed or announced plans to close.
Through improved plant operating procedures those remaining have
reduced mercury consumed as processing losses by about 27 percent.
Consumption of mercury per ton of chlorine produced was 0.43
pounds in 1970, compared with a level between 0.5 and 0.6 pounds
prevailing during the previous 8 years.
In addition, several industries have curtailed or eliminated
the use of mercury in their processes. The paper industry has
accelerated its trend away from mercury and no longer uses mercury
compounds for in-process slime control. Methyl mercury compounds
are no longer licensed by the U. S. Department of Agriculture to
be used in treating seeds for agriculture. One major mining
company has discontinued use of the mercury amalgamation process
for gold recovery.
The mercury industry operates in a world market where the
U. S. consumes about 22 percent of world production, with 60
percent from imports. Major world producers, in order of importance,
are the Soviet Union, Spain, Italy, Mexico, United States, and
Canada. Canada became a major mercury producer in 1968 with the
opening of a highly mechanized mine producing ore similar in quality
to U. S. ore. In 1970 Canada contributed about 9 percent of world
production. Three-quarters of this was exported to the U. S., replacing
Spain in 1970 as the major mercury exporter to the U. S. Spain is
second and Italy is third in supplying mercury to the U. S. Both of
these 'countries have been producing mercury from extremely high grade
ore (compared to that of the rest of the world) for several hundred
years. Mexico is another source of U. S. imports and a major contributor
to world production. However, the importance of Mexican production
is difficult to estimate because production statistics from Mexico
4-14

-------
are undependable.
Trade sources estimate as much as half of
Mexican production moves through illegal channels and is
unreported to avoid Mexican income, production, and export
taxes.
3.
Prices and Trends
Total U. S. consumption of mercury in 1970 dropped 22 percent
from 77,988 flasks in 1969 to 61,503 flasks, and consumption
in 1971 is running at an annual rate of about 55,000 flasks.
As consumption has fallen, so have prices. Average price per
flask was $505.04 in 1969, $407.77 in 1970, and closed at $225
at the end of trading in 1971 although prices had averaged
$285 or higher for the first three quarters of the year
(Table 4-4). Since 1965 the overall price trend has been
down, especially since the mercury pollution publicity of
mid-1970.
Factors believed to have contributed to price decline
are increased world production, U. S. Senate hearings on the
effects of mercury
of registration of
resulting from the
on man and the environment, cancellation
48 mercury biocides, and reduced consumption
general economic recession.
Impact of Control Costs on the Industry
For domestic firms, break-even production cost averages
between $300 and $400 per flask, varying with the quality of
ore being worked and the difficulties encountered. Ore bodies
that can be worked with open pit methods are less expensive
to operate. Deeper mines that require underground mining methods
may encounter water and ventilation problems. For most mines
exploration and development costs are in addition to the break-
4.
even production costs reported.
Operations can continue in
areas explored and developed until they are exhausted. Firms
must then either make additional investments in development
and eventually exploration or close.
Current prices are so far below costs that domestic producers
have abandoned operation until only a few of the largest remain.
4-15

-------
Table 4-4. AVERAGE NEW YORK PRICE OF MERCURY
  Price
Year  ($/F1ask)
1960  $ 210.76
1961  197.71
1962  191.21
1963 .  189.45
1964 .  314.79
1965  570 . 75
1966  441. 72
1967  489.36
1968  535.56
1969  505.04
1970  407.77
1971 1st Quarter 341. 34
 2nd Quarter 285.89
 3rd Quarter 288.30
 4th Quarter 253.46
 Year 295.00
Source:
References 3 and 4
4-16

-------
One is operating on ore already broken out (and therefore ore
with a very low marginal cost) and plans to close when that is
exhausted. Another producer has reduced output to one-quarter,
having already closed two mines. This company reported a loss
last year and does not expect to remain open much longer. A
third company continues to operate at a loss hoping prices will
increase since water incursion may make it unlikely that the
mine can ever be re-opened if it is shut down.
Future prospects for the U. S. industry are dim. Competition
,

from Spanish, Italian, and Canadian producers that have either
higher quality ore or highly mechanized operations is intense.
Some foreign producers operating very high grade ore have curtailed
operations and withheld production from the market because of low
prices. GSA stockpiles declared surplus are large. Industry
stocks are currently low, but the mercury released for market
each time a mercury cell ch10r-a1ka1i plant closes is large. A
revival of the domestic mercury industry within five years is
doubtful.
Any costs associated with required air pollution control
equipment will serve to hasten the demise of the few firms
continuing to operate. For those that have closed, air pollution
control costs will raise costs should they try to open again and
thereby help prevent such attempts. Further, the larger firms
will be favored because they come the closest to having, or
being able to raise, the necessary financial resources. Air
pollution control costs will also favor certain process changes.
Currently, mercury is extracted from unconcentrated ore close by
each mine. A requirement for more elaborate extraction plants
will favor floatation concentration of ore and shipment of the
concentrate to a central extraction plant owned cooperatively
or by a large firm.
4-17

-------
5.
Price Impact
Any analysis of the economic impact of air pollution controls
imposed on an industry that is shutting down because prices no
longer cover operaFing costs is largely academic. Producers'.~.H
absorb the costs from nonexistent profits nor can they pass the
cost on in the form of price increases. Some firms may continue
operating without controls until enforcement causes them to
close.
4-18

-------
III.
Secondary Mercury Recovery
Mercury Recovered and Methods of Recovery
Mercury is recovered from many products and pieces of equipment
in which it has been used,when their useful life is over. In addition,
it is often highly purified for special uses, as in scientific
instruments. These operations are all undertaken by the industry
known as secondary mercury recovery.
A.
The major amount of mercury recovered currently is received in
liquid form, containing primarily dissolved metallic impurities.
Prime virgin (primary) mercury is occasionally purified by the same
techniques as "used" mercury. Solids from which mercury is recovered
include dental amalgams, chemical sludges, scrap dry cells and batteries,
and junked equipment in which mercury was used, such as mercury stills
and mercury switches~ For accounting purposes, releases of mercury
from government stockpiles are reported in the literature as part of
secondary mercury production.
Recovery of mercury from contaminated liquid is really a purification
process from which the contaminants are occasionally recovered as
by-products. The three techniques by which the purification is
accomplished are: (1) distillation; (2) solution purification;
and (3) oxygenation ("oxification" in the trade). Mercury is recovered
from solids by methods similar to those used in primary recovery,
generally indirect-fired retorting. These four processes, each
of which can lead to significant emissions, are described separately
below, along with the bottling process where the recovered mercury
is packaged for shipping.
Other emissions can occur in the secondary recovery plant from
miscellaneous handling processes, suCh as filling of apparatus.
This may be especially true if hot operating equipment, such as a
stil1pot or retort, is not adequately cooled before cleaning and
refilling. Also, emissions are expected from accidental spills,if
these are not properly cleaned up or covered.
4-19

-------
1.
Distillation
Dirty mercury is distilled by heat under a vacuum of approxi-
mately .01 torr, the vacuum being drawn on the condenser side
of the still. Cooling of the distilled vapors is typically
I
done in a water-jacketed condenser. Where very-high purity
mercury is required, a double or even triple distillation of
the mercury may be performed. In installations where multiple
distillation is done frequently or routinely, the equipment may
be an enclosed series of stills which require no external transfer of
liquid from one distillation step to the next. Alternatively,
the mercury may undergo a mixture of the various purification
processes, depending upon the particular contaminants.
Major emissions of mercury from the distillation process
itself come from the vacuum pump exhaust, which is generally
vented to the plant's workroom air. The advantages of distillation
under a vacuum presumably are (1) a lowered boiling temperature
of mercury resulting in lower heating requirements and/or
more rapid distillation, and (2) removal by the vacuum system
of dissolved vapors and gases. Typical ventilation flow rates
for a 3000 flask/year plant are 2000 to 4000 CPM. Emissions
from the workroom are then expected to be 2.85-5.5 mg/mtn at
the new TLV or 5.7-11 mg/min at the current TLV. At the maximum
exhaust rate and the current TLV this results in emissions of
.0348 1b/24-hr day or 12.5 Ib/yr if the plant operates
continuously. Emissions from the distillation process itself
must be much below this, for the total emiss~ons to inp1ant air
also include those from uncollected spills, transfers, bottling, etc.
The total estimated loss amounts to a maximum of .0035% of the
total plant throughput of 3000 flasks and may be considered a
negligible loss, at least economically.
The prime difficulty with the above calculation is the
assumption that tbe industrial TLV is being observed. There are
no data to indicate that it is not being met, but neither are
there data verifying that it is. However, the overall loss
rates estimated above seem small, in view of the potentials
4-20

-------
for emission which exist from leaks, spillages, and other accidents.
If spills are not properly cleaned up or covered, evaporation from
the large surface areas exposed will lead to emission problems.
Nonetheless, it appears that losses from the production
I '
process itself will be very small, as indicated above. It is
anticipated that no control for vapor losses from production
processes would be required unless new data giving a radically
altered picture are obtained. However, if controls are required,
the following techniques would seem applicable:
(1) enclosed working compartments for the transfer of
both dirty and clean mercury and removal of sti11pot
residues (which are up to 95% mercury). This will
reduce spills and evaporative transfer discharges
to an absolute minimum;
control of the vacuum pump discharge with an activated
(2)
carbon adsorption filter.
,

Installation of these controls would reduce overall plant emissions to a
truly negligible level after the work area is adequately cleaned of
existing residues from spills. It is anticipated that ventilation
requirements would be reduced to those required for normal worker
comfort unless a major ac~ident occurred. Costs of control systems
are estimated to be very low. Activated carbon can adsorb 10%
its own weight of mercury before becoming spent. Supposing that
half the total emissions of 12.5 lbper year come from the vacuum
pump exhaust, or 6.25 lb" a total of 62.5 lb of charcoal/yr would
be required if all the mercury is'adsorbed. In this small
quantity, the required treated, activated carbon may cost as
much as $125 per year. Assuming small, "cartridges" of carbon
requiring 6 changes per year, one hour's total labor for between
$5 and $10 is reasonable. Installed cost, mostly labor, would
be $50-$100 and necessitate no downtime. Annualized costs would
then be less than $150 per vacuum pump.
4-21

-------
2. ' Solution Purification
Solution purification involves removal of (primarily metallic)
impurities by dissolving the impurities in dilute nitric acid.
The processing: takes place in a stainless steel leach tank.
. I .
The dirty mercury is ~ptied into the tank, covered with a layer
of dilute acid, and air is bubbled (sparged) through the mercury
and acid to promote mercury-acid contact. The gas flow rate is
generally 1-5 cfm. After a sparging time of several hours the
cleaned mercury is separated from the contaminated acid by
decantation; the mercury is then water-washed to remove all
traces ,of acid, and then bottled.
Although there are several points within the process stream
which may emit mercury to the air, only the sparging air stream
carries mercury directly into the outside air. The other sources
result in emissions to the workroom, which must be kept at or
below the industrial TLV. Maximum emissions carried by the
sparging air may be calculated assuming that (1) the air becomes
saturated with mercury during the bubb1~ng process, and (2) no
mercury droplets are entrained in the exhaust air. With these
assumptions, and assuming an air temperature of 86°F(containing
1.8 x 10-6 1b mercury vapor/ft3) and flow rate of 5 cfm, the
maximum mercury emissions would be .013 1b per 24 hr day. Lower
air temperatures, partial "scrubbing" of the air by the nitric
acid layer and/or lower spargi,ng rates would all lead to lower
atmospheric emission rates.
Losses from the other emission points lead to contamination
of the workroom air, which is exchanged with atmospheric air to
produce the required new TLV of 50 ~g/m3. Ventilation flow rates
of 2000 ft3/min are reported as typical of solution purification
plants. The losses from ventilation then would be 0.01 1b/24-hr.
day, leading to a maximum plant-wide loss rate of 0.023 1b per
day.
Emission can occur by evaporation from the spent leaching
acid and waste wash water. These are not considered in this
4-22

-------
analysis since either the acid will be stored, covered for reuse
or discarded with the waste water.
Emissions from the sparging process can be reduced by passing
the sparging air through a treated activated carbon adsorber.
In order to prevent difficulties with the adsorber which may be
caused by nitric acid vapors, it may be necessary to water wash
("bubble") the sparging air before adsorption. If the air is
cooled during washing, this might result in a reduction in the
mercury content as well, but it would not be a significant
reduction. Costs of control would be low, even as a percentage
of current plant investment; uncontrolled emissions are so small,
however, that even this expense is probably not necessary.
[Emissions from the process of bottling the cleaned mercury
are discussed elsewhere.]
Oxygenation
The oxygenation cleaning process is accomplished in a manner.
similar to that of the solution purification process. The
differences are: (1) oxygenation depends upon oxidation of
impurities to filterable solids,' so that no acid leachant is
used; (2) there is no washing process, but the cleaned mercury
is filtered; (3) the sparging rate is much lower than that of

. .
solution purification, typically 0.2 cfm.
The sparging air from the oxygenation process is typically
scrubbed with water and cleaned with activated charcoal before
venting. As a result emissions would be negligible, down from
the maximum possible at 86°F of .00052 lb/day. Emissions from other
sources would be similar to those from solution purification,
which were calculated as 0.01 lb/day.
Control of the emissions from the oxygenation process
itself would not seem to be necessary, in view of their low quantity.
However, the current use of water scrubbing on the sparging
gases indicates that there may be a problem with entrained
mercury droplets, which would be recoverable using the water
scrubbing technique. There are no further data on this point,
3.
however.
4-23

-------
.Additional mercury emissions associated with oxygenation,
but,not directly connected with the process, arise from recovery
of mercury from the filter cake (usually by retorting), and
bottling. Both of ~hese processes are discussed in later
sections.
4.
Retorting
The technique of retorting for mercury recovery is used
when the mercury is contained in a solid or semisolid material,
such as dental amalgams, mercury electrical cells, and sti11pot
residues.
The retorting operation is similar to both distillation
and primary smelting. If the scrap contains a high percentage
of mercury, the material is generally treated in an externally
heated pot; no process air is required, since the mercury does
not have to be chemically liberated, and the non-mercury
residual may also be a valuable by-product. Where low concen-
trations of mercury are involved, the mercury-bearing material
is placed in an oven-type cabinet on trays. In either process,
as the material is heated, the mercury volatilizes and passes
into a water-cooled condenser. In some cases, the gases may
be further treated by additional chilled water cooling and/or
water sprays. The residual gases and vapor are then exhausted
to the atmosphere.
Data on exit gas concentrations of mercury, recovery rates,
etc., are not available. It is expected, on the basis of process
type, that recovery rates of mercury from the hot vapor would
be comparable to those of the primary smelting industry,
parti~u1ar1y from closed, externally-heated vessels. Emissions
from oven-type retorts, which use direct-contact heating,
are much harder to estimate because of the highly variable
mercury content and heat capacities of the charged materials,
which will result in highly variable vapor concentrations in
the gas stream. Also, contamination of the gases by other
volatile materials, or by fine particulate which may carry
4-24

-------
mercury away, is impossible to specify. Gas flow rates per unit
of mercury charged or produced are not calculable, but are
expected to be lower than for comparable primary smelting
production.
Need for controls, therefore, is almost impossible to
specify, since production rates, ratios of closed to oven
type systems, etc., are not known. Applicable control
systems beyond those already in use might be refrigerated
cooling and demisting of the gas stream or use of treated
activated carbon or other adsorbant.
However, there is
no indication of their applicability in the literature.
Calculations of control costs, therefore, have not been made.
B.
Bottling
,
The mercury produced by the various recovery and
purification processes is stored in reservoirs until enough
has been accumulated to be bottled, usually in the standard
76-lb iron flask. Emissions of mercury vapor to the atmosphere
can occur from the exposure of the mercury to the atmosphere
plus splashing during the filling process.
About 96% of the secondary mercury processed in the U.S.
is bottled by a metering apparatus which is open to the workroom
atmosphere. In some installations the bottling apparatus
is in the same room as other secondary processing equipment,
while in other plants it occupies a separate room. Losses from
natural evaporation, etc., are too small to measure directly.
With a maximum room ventilation rate found of 4000 cfm at the
3
5011g/m , however, the losses would be a maximum
required TLV of
of 0.02 lb/day.
The remaining 4% of U.S. production of secondary mercury is
bottled using the same apparatus, but enclosed in a box of about
0.5 m3 volume. The box contains a door, a window, and glove ports.
Operation of the box involves opening the door to remove full flasks
4-25

-------
and insert empty ones, -sealing the door, and filling the flasks.
Emissions from the box, assuming that the air inside is saturated
with mercury at warm room temperature (77°F) and that the air
I
is completely exchanged when the. door is opened, are, .00002 lbper_._.-
flask transfer cycle. Enough flasks can be contained in a box
of 0.5 m3 size that 2 transfer cycles/day would be adequate
for most plants. Emissions would then be 4xlO-5 lb/day.
Emission controls on bottling would not seem to be necessary
inmost cases, unless the bottl~ng operated continuously for a
long period of time. If the closed box filiing arrangement
is used, that would certainly be adequate control. The cost of this
box is estimated at roughly $300-$500. There would be some
economic advantages to using the box, since it allows better
control and recovery of spills, reduced labor cost for contr?l
of spills, and would allow a somewhat more relaxed employee
attitude because of reduced emissions concern. Whether these
factors would allow recovery of the cost of the box over its
useful life may depend upon the sLtuation in which it is used,
but it very well could. In any case, it would seem to be a
minor expense.-
4-26

-------
C.
Costs of Control
Control costs for the secondary mercury industry
Some plants already have, as good operating practice,
equipment. Others do not have it and. do not need any
equipment to meet control standards. Table 4-5 shows
are minimal.
some of the suggested
additional
the estimated
control costs for the industry based on the assumption that every
firm in the industry must install control equipment. Actual
expenditures would be much less.
Table 4-5.
Estimated Emission Control Costs for
the Secondary Mercury Industry, 1970
Investment
Cost
Annual Operation
& Maintenance Charge
Annualized
Costs
Activated Carbon
Absorption Filter

Bottling Box
$
2,400
12,800
$ 4,800
800
$ 5,280
3,360
Totals
$ 15,200
$ 5,600
$ 8,640
D.
Economic Impact
1.
Industry Structure
There were 32 active firms reported in 1970, each with a
single plant.
Firms, as would be expected, are located close
to sources of junk mercury or mercury containing scrap. They
are located primarily in. the New York-New Jersey, Chicago,
Los Angeles, and San Francisco Bay industrial areas.
Much of the mercury processing by the industry is on a
consignment or fee-for-service basis. The mercury or mercury
containing scrap is sent to the mercury processor to be cleaned,
and the processed mercury is then returned without a change in
ownership. The processor charges for what he has done.
2. Market Structure
Mercury recovered by the secondary industry is (except for
redistilled mercury) no different from that produced by the
primary industry. The product of each industry is directly
4-27

-------
substitutable for the product of the other industry, and therefore
the product of the secondary industry is sold in the same markets
and at the same prices as primary mercury.
Redistilled m~rcury is somewhat different in that the cost
of the added processing must be covered by the price. However,
redistilled mercury can be produced from mercury recovered by
,the secondary industry or manufactured by a primary producer.
Therefore, the higher price for redistilled mercury in only enough
to cover the added cost plus a reasonable profit on the
service. There are enough firms in the industry to prevent
the realization of monopoly profits. In addition, a firm
requiring redistilled mercury can buy mercury on the open
market and have it redistilled if the price differential for
redistilled mercury appears to have become unreasonable.
Table 4-6 shows the 1970 consumption of primary, secondary,
and redistilled mercury. Electrical apparatus is the principal
use for both redistilled and secondary mercury. Other major
uses for redistilled mercury in order of importance are for
industrial and control instruments, dental preparations, and
general laboratory use. These same industries are important
to secondary mercury, but industrial and control instruments
Eank- f'!~1:'_~h_--i.nstead of second.
E.
Impact of Control Costs
Control costs for the secondary mercury industry, both as a whole
and for individual firms, are so low that the impact will hardly be
noticed. Based on 1970 secondary mercury production of 8,051 flasks,
and assuming every firm requires control equipment, investment
costs would" at most amount to about $2.00 per flask. Annualized
costs would be at most about $1.00 per flask of 1970 production
(1970 price averages $407.77). Costs are so minimal that firms
would have no difficulty making the necessary investment and recovering
the increased costs from lower operating cos'ts, by slightly higher
charges for processing mercury for others, or by slightly lower
4-28

-------
Table 4-6.
MERCURY CONSUMED IN THE UNITED STATES IN 1970
(FLASKS)
1/2/
Agriculture --
Amalgamation
Catalysts
Dental preparation
Electrical apparatus
Electrolytic preparation of
chlorine and caustic soda
General laboratory use
Indust2}a1 and control instruments
Paint :-
Antifouling
Mildew proofing 2/
Paper and pulp manufacture-
Pharmaceuticals
Other
Total known uses
Total uses unknown
Grand total
Primary

1,811
206
1,916
166
11,432
14,749
689
2,124
193
10,149
223
280
5.668

49,606
15
49,621
Redistilled
Secondary
Total

1,811
219
2,238
2 ,286
15,952
15,011
1,806
4,832
198
10 ,149
226
690
5.858

61,276
227
61.,503
1/
1:../
Includes fungicides and bactericides for industrial purposes.

1971 and 1972 decisions by EPA will essentially eliminate these
markets.
Source:
Reference 3
3
225
1,372
3,469
10
97
748
1,051
495
2,353
262
622
355
5
362
12

8,296
69
3
48
178

3,374
143
8,365
3,517
payments for mercury and mercury containing scrap. It appears that
much of the cost for those firms that must install control equipment
would be recovered from lower operating costs.
Impact on Prices
Air pollution control costs for the secondary mercury industry
would have no impact on the price of mercury. These prices are
determined competitively in a world market and the domestic secondary
mercury industry is too small to have much impact. Then, too, the
change in cost is small. At most, a slightly increased fee schedule
for processing mercury for others and slightly lower payments for
scrap mercury might result. Any such changes would be so small as
to go virtually unnoticed.
F.
4-29

-------
IV. . MERCURY-CELL CHLOR-ALKALI PRODUCTION
A.
Introduction
I
Virtually all chlorine used in the U. S. is produced by one of
the available electrolytic methods, primarily electrolysis of brines
and of fused chlorides. By far the largest production is from
brines of alkali-metal chlorides, with the production of chlorine
gas at the anode and the alkali metal (sodium or potassium) at the
cathode. In the diaphragm cell process the alkali metal reacts
directly with the water to form the corresponding hydroxide in
solution, and hydrogen gas. The diaphragm, characteristic of the
process, serves to separate the brine and caustic solutions. In the
mercury or amalgam cell process a layer of mercury over the normal
carbon serves as the cathode, the mercury-brine interface acting
analogously to the diaphragm. The alkali metal amalgamates with
the mercury and the amalgam is tapped from the cell. The amalgam is
then passed through a second cell known as a denuder, in which the
alkali is stripped from the mercury by pure water, forming hydroxide
and hydrogen gas. The cleaned mercury is then returned to the cell.
The mercury cell process is the one which is considered in this report.
The chlorine produced by the two processes is essentially identical
and interchangeable; but the caustic is not in some applications.
Caustic from the diaphragm cell process is generally of relatively
low concentration (15%), and contains chloride or hypochlorite ions,
even if in low concentrations, making it unsuitable for use in
synthetic fiber production. Mercury cell caustic, on the other
hand, although free of chloride and hypochlorite contamination,
and of 50%~73% concentration, contains mercury. It is therefore
. unsuitable for photographic chemical applications and some other
uses.
B.
Mercury Emissions
There are several points from which mercury can escape to the
atmosphere in a mercury-cell chlor-alkali operation. The actual
amount of mercury emission from each type of source may in fact vary
4-30

-------
from plant to plant, since several types of cells are in use, and the
condition of the cells may vary with age and plant maintenance policy.
However, the emissions described below are believed to be representative.
Total amounts of emissions, where given, refer to a plant producing
100 tons of chlorine per day unless otherwise noted.
Purified and nearly saturated brine solution and mercury returned
from the denuder are fed concurrently from the inlet end box into the
cell. The end box serves to keep a reserve of mercury and brine ready
for smooth operation and to keep the mercury covered with brine. The
spent brine and amalgam are collected in the outlet end box. The brine
is tapped off from the end box and usually dechlorinated and is at
least partially recycled.
The chlorine gas, which cont,ains water vapor and perhaps some
mercury vapor, is dried by scrubbing with concentrated sulfuric acid.
Inert gases are removed. The spent sulfuric acid contains essentially
all of the mercury from the chlorine gas, so that there is virtually
no loss with the chlorine product. The chlorine gas is then compressed
and occasionally liquified.
Ventilation air supplied to the outlet end box contains mercury
and mercury compounds, when it leaves the end box, in amounts which
depend largely upon the amount of ventilatiQn air used. The amount
of air used, in turn, depends upon cell design, plant age, and plant
operating procedures. Current trends are to reduce end box ventilation
flow as much as possible in new cell designs and by modifications to
existing cells. Currently, it is believed that the maximum total
outlet end box ventilation flow for a 100 ton/day chlorine plant is
1500 cfm. If the gas is saturated with mercury vapor at 60°F and contains
no particulate mercury, this leads to a mercury loss from the plant of
1.2 lb per 24-hr day in the absence of any controls.
The alkali amalgam is bled from the outlet end box into the
denuder or decomposer, where it is reacted with purified water in
a short-circuited "cell" to produce the alkali caustic at a high
purity and about fifty percent, by weight, concentration at the
final outlet. The by-product hydrogen gas results from this reaction,
4-31

-------
and is typically saturated with mercury at 210°F. The minimum treat-
ment known to be applied to the hydrogen stream is cooling to about
110°F followed by demisting to remove the condensed mercury droplets.
~e approximate hydrogen ;flow rate at this temperature is 750 cfm from
a 100 ton per day chlorine plant. At this flow rate, if the hydrogen
stream is then flared or discharged, the mercury emissions are
approximately 4.1 lb per 24-hour day.
The alkali caustic solution from the denuder is filtered to
remove suspended mercury droplets and other impurities. The resulting
filter cake is retorted in a manner very similar to that used in
secondary mercury processing. The recovered mercury is returned
to the cells for reuse. Mercury recovered from the spent chlorine-
drying sulfuric acid may also be treated in this way. Emissions
from the retorting operation are estimated to average .002 lb per
100 tons of chlorine gas produced.
Further processing'of the caustic solution may be undertaken
to make the concentration 73% by weight, or the material may be
heated to produce anhydrous caustic alkali. No significant mercury
emissions are indicated to result from this process.
Probably the largest amount of mercury emissions occurs from
ventilation of the cell room. Mercury is "routinely" spilled in
small amounts, lost as vapor in leaks, may by-pass pump seals in
mercury handling systems, and escapes as vapor when cells are
broken down for maintenance. In addition, significant losses
may occur, on an infrequent basis, during catastrophic cell failure.
Aside from cell failure, however, the direct losses of mercury vapor
are probably small, although they will vary with cell design and
equipment condition and age. Because of the high cost of mercury,
these losses are kept at a reasonable minimum. Losses from spilled
liquid, however, unless it is immobilized or cleaned up, can mount
up at a rate depending upon the temperature of the cell room (usually
quite warm because of pOWer usage) and the exposed surface area of
the mercury. Losses to the atmosphere measured from these sorts of
4-32

-------
data are difficult to estimate.
However, ventilation rates have been
obtained for cell rooms ranging from 100,000 to 1,000,000 cfm for
each 100 tons per day of chlorine produced. On the assumption that
3
the required TLV of 50 ~g/m represents the actual airborne mercury,
the range of mercury losses in the cell room air at these flow
rates is 0.45 1b to 4.5 lb per day.
Data currently available
indicate that cell room mercury emissions probably increase with
plant size, but not in direct proportion to plant capacity. Although
for smaller plants it would appear that mercury emissions in cell
room air could readily exceed those from all other sources combined,
it would appear to be less likely ~hat this would occur in larger
plants, where the other losses scale up directly with the production
rate.
C.
Controls
Mercury emission control techniques for the hydrogen gas stream,
end-box ventilation air stream, and cell room ventilation air
stream fall into three general categories: cooling and demisting
of air streams saturated or nearly saturated with vapor, chemical
scrubbing, and adsorption. The following sections discuss each
technique and its relative effectiveness.
1.
Cooling and Demisting
""-.
\
If the partial pressure of mercury in the gas stream reaches
the saturation point ("dewpoint") at a reasonably high temperature,
mercury can be removed from the gas by cooling it below the
saturation temperature to as low as desired or feasible and
removing the condensed mercury droplets or supersaturated vapor
with demisting equipment. Highly efficient demisting removes
up to ninety-nine percent of the suspended droplets and super-
saturated vapor, leading to high reductions in mercury vapor
content because of the strong temperature dependence of the
vapor pressure curve. For example, cooling the hydrogen gas
stream from 210°F to 110°F reduces the vapor pressure of the
mercury, and therefore the total mercury vapor per unit volume
4-33

-------
of gas, by 97.5%, by weight. 100% demisting is required, of
course, in order to remove all 97.5% of the mercury.
An extension of this technique involves compression of the
gas'before cooling and demisting. The demisting must be
accomplished before: the gas is allowed to expand to avoid
revaporizing the co~lesced mercury. As an example of this
starting at 86°F, compression of the hydrogen stream to
3.5 atmospheres absolute followed by cooling to 68°F
with demisting can reduce the mercury vapor by 55% by weight
(Battelle Memorial Institute, 1971). It should be noted that.
when the hydrogen gas is allowed to expand to one atmosphere
pressure the mercury concentration would be below saturation.
Cooling the gas to 37°F rather than 68°F would result in a
90% reduction in contained vapor.
Although full-scale application of these techniques to
end-box ventilation air has not been reported as yet, it seems
reasonable that they would be attempted in the near future.
Chemical Scrubbing

A depleted brine scrubbing system has

to treat the hydrogen stream of a domestic
3
Exit mercury contractions of 0.1 mg/m are

facility.
2.
been used since 1961
chlor-alkali facility.
reported for this
The use of "chlorinated brine" to reduce mercury content of
the gas stream below its saturated vapor pressure has been
examined. The system used a solution of sodium hypochlorite,
sodium chloride, and sodium'hydroxide to produce soluble mercury
salts which are then absorbed into the solution. Pilot plant
3
operations indicated a reduction of mercury vapor to 1 mg/m
in the hydrogen gas stream, corresponding to a reduction of
ovet 95% below cooling-demisting step results [Ref. 2].
Construction of a full-scale system was started. Subsequently,
it was converted to a depleted brine system.
A hypochlorite scrubbing system is currently being used
by the British Petroleum Company to treat both the end-box and
hydrogen streams at one of their facilities.
Chemical scrubbing
appears to have promise for controlling mercury emissions at
chlor-alkali plants.
4-34

-------
3.
Adsorption
The technique of chemical-physical adsorption of mercury is
mechanically the simplest type of collection system. The
contaminated gas stream is passed through a bed of the adsorbing'
medium and the mercury is preferentially trapped by either a
,
chemical or physical adhesion process to the surface of the
medium [Reference 2]. Depending upon the nature of the medium
and the'details of the adsorption process, the adsorbent may or--
may not be regenerable. In some cases'it may also be possible to
recover the mercury for reuse.
Adsorption of mercury on an approptiate medium occurs at
virtually any concentration of the vapor down to almost trace
levels. With adequate bed thickness and sufficiently low face
velocity, and with the bed itself kept reasonably below saturation,
it should be possible to reduce mercury concentrations in any gas
to very low levels. Although some applications, notably for
use on cell room air, still face full-scale trials, adsorption
emission control techniques promise the ultimate in control of
low-concentration mercury emissions.
Two types of media are currently being studied for their
applicability to mercury emission control. The one upon which
the most work has been done is activated carbon, either "plain"
or "treated". The other medium is a molecular sieve, believed
to be a zeolite-type material similar to those used in water
deionization.
Activated charcoal, as noted above, may be used either with
or without "treatment". The treatment appears to consist of
impregnation of the charcoal with iodine and potassium iodide.
Two manufacturers of charcoal are currently doing extensive
research and development on this application of charcoal
adsorption; one of them claims that its particular charcoal
is regenerable. There is no indication as to the disposition
made of the removed mercury. If non-reusable charcoal is used,
there may be a significant disposal problem of spent charcoal,
since the charcoal appears to become saturated with mercury after
capture of about 10% mercury by weight. This could become a very
high operating expenditure, since the charcoal becomes saturated
4-35

-------
so readily. The charcoal adsorption
most effectively in "final" clean-up
low concentration ~f mercury vapor.
Mercury emission control with a molecular sieve is a
relatively new proprietary process for which many details
are not available. However, it appears to operate in much the
same manner as the charcoal adsorption system. Currently the
manufacturer claims. an installation in a 135 ton per day
plant on the hydrogen gas stream. Two units are operated
simultaneously with one acting as the collector while the
other'is being regenerated. Gas to be stripped of mercury
vapor is first cooled and demisted to as low a concentration
as practical. The gas is then passed through the collecting
unit and either utilized or vented. Regeneration in the second
unit reportedly involves bleeding off part of the cooled,
demisted stream, heating it to about 650°F and passing it
through the bed being regenerated to strip out the mercury.
The mercury-laden regeneration gas is then returned to the
coo1ing-demisting step for recovery of the mercury.
The exact emission reduction achieved by the molecular
sieve process is not known; however, the manufacturer claims that
the hydrogen gas leaving the process contains 0.50 mg/m3 ppb
.025 ppm) of mercury (private communications with EPA). Losses
at this rate in the hydrogen from the plant where the process
is currently being operated amount to approximately 4 1b/year.
Presumably this is near the levels achievable by the activated
charcoal adsorbents.
The primary advantages that the molecular sieve adsorbent
has over activated charcoal are the demonstrated recoverability
of the mercury, and the. reusability and relatively long life of
the medium (the molecular sieve manufacturer is reportedly willing
to guarantee the bed for a minimum two-year lifetime). This may
be offset by the difficulty of handling the heated hydrogen
technique would be used
of gases which have a
4-36

-------
regeneration stream and the added costs of heating and then
cooling the regeneration hydrogen.
Technically, it would seem feasible to control both end-box
ventilation air and cell room ventilation air with either one
of the adsorption media. Even cell room ventilation air,
3
with maximum permitted levels of 50 ~g/m of mercury, is reducible
by at least a factor of five, based on the available data for
pilot scale charcoal adsorption equipment operating on the
hydrogen gas stream. However, it is not known at present
whether trace gases found in end-box ventilation air and cell
room ventilation air (predominantly ambient air) either
positively or negatively affect the mercury absorption character-
istics of either the charcoal or the "molecular sieve."
D.
Preventive Controls
The extremely large volumes of cell room ventilation air that
are currently used as the primary means of achieving the industrial
hygiene TLV, plus the known sources of emissions into the cell room,
suggest that "arrest-at-the-source" emission control procedures may
be both more effective and more economical for cell room air. Sources
of mercury emissions into the cell room air include the following:
1. Routine (process-required) emissions
a. End-box sampling procedures.
b. Removal of mercury butter from end boxes.
c. Cell maintenance and rebuilding operations.
2.
d. Other mercury-connected maintenance
e. Cell and mercury pump leaks.
Accidentia1 (catastrophic) emissions)
a. Accidental mercury spills.
work.
c.
Cell failure.
Other unusual circumstances.
b.
4-37

-------
It would appear that the emissions from category 1. are at least
conceptua1.ly amenable to reduction by cell modification, improved
maintenance procedures, and improved auxiliary equipment. It is not
known whether any applica~le technology has been utilized to the
utmost with a view to reducing mercury emissions. To the extent
that it is cheaper merely to provide adequate ventilation than it
is to reduce emission losses, technological improvements are
probably at a minimum. In.addition, once adequate ventilating
capacity is installed it is difficult to justify expensive
improvements that would result in merely underutilizing the
installed capacity.
Non-routine losses can be reduced by
handling requirements and (2) instituting
in instances of cell failure.
It is believed all of these steps are now being followed in
varying degrees in maintaining the mercury level in the cell-room
air below the new TLV of 50 ~g/m3. Some of the available means
(1) reducing mercury
improved shutdown procedures
for implementing these practices are applicable to both new and
existing facilities. A partial listing of the means which may be
applied, for example, to the minimization of incidental spills
includes:
o
Devising and putting into practice improved routines for
all normal cell-room operations that contribute to the
incidental release of me.rcury.
o
Providing special equipment required, for example, for
the wet removal of cell butter.
Improving maintenance procedures.
. Detecting and minimizing hot-hydrogen leaks.
Devising procedures to deal with brine and caustic
leaks from the cells and auxiliary systems.
Taking steps to minimize leaks from the mercury pumps by
improving pump maintenance procedu~es, or isolating the pumps
from the cell-room atmosphere by enclosing them and collecting
mercury that has leaked.
o
o
o
o
4-38

-------
Some of the means for improving housekeeping procedures that have been
suggested or applied include:
o Lining floors, sewerage, and channels with PVC or epoxy
coatings.
o Removing mercury spills from floors and channels with
o
vacuum sweepers.
Rapid hosing down of floors and drains after leaks.
Using mercury scavengers to inactivate spills that occur
in hard-to-reach locations.
For new construction, the following steps have been recommended by
Bouveng and Ullman in a paper presented in Stockholm, Sweden, in 1969:
o The use of improved cells designed to minimize leakage of
the cell contents.
o
o
Installation of numerous small mercury traps
locations in the cell-room drain systems.
The use of V-shaped rather than rectangular
channels.
at strategic
o
floor
o
The use of most of the means of controlling mercury emissions
listed above.
It is suggested that individual operators need some latitude in
choosing from among the control options suggested above to suit
individual situation and needs [Reference 2].
E.
Emission Standards
Standards for mercury emissions from mercury cell chlor-alkali
plants used for this report are as follows:
o No more than .005 lb of mercury per ton of chlorine produced.
o No more than 2.56 lbs per day from anyone plant.
A revised emission standard is currently under consideration
by EPA. It limits mercury emissions to a maximum of 2300 grams (5 Ibs)
per day from anyone plant (1300 gm/day for the cell room and 1000 gm/day
for the combined process streams). Consideration of this standard is
not included in this report. It would seem that significantly reduced
control, and therefore costs and economic impact, than those shown
below, would be required of the industry if the newer standard is adopted.
4-39

-------
F.
Costs of Control
The investment required to control emissions to the atmosphere by
mercury-cell plants of average size (approximately 240 tons per day)
is estimated to be $14,725:,000 (Table 4-7). Annualized costs are
I
estimated to be $17,344,200. This amounts to an investment of
$1,950 and an annualized cost of $2,300 per ton per day of chlorine
plant capacity.
Control cost will increase plant investment from five to ten
percent and will increase production costs about five percent for
mercury-cell plants. Cost changes of this magnitude will have a
significant impact on the economics of mercury cell production versus
. .

diaphragm cell production. The two processes appear to be very
similar in production costs. However, mercury cell production was
apparently favored in recent years since the majority of new
construction (through 1970) has been mercury cell plants. Addition
of control cost to mercury cell plants will cause diaphragm cell plants
to be more economical since the latter require no controls. This will
cause new construction to be shifted to diaphragm cell plants.
Modern mercury cell plants are efficient enough so that they can
be adequately controlled and remain competitive in the industry.
Only those mercury cell plants that are already obsolete and marginal
will be abandoned as a result of the imposition of air pollution
emission control costs.
Costs of control to consumers will decrease by
investment of $12,650,000 and an annualized cost of
even though total industry production will increase
tons of chlorine.
1977 to an
$14,900,000,
to 13,375,000
1970 .
1977
Table_4-7. ESTIMATED COSTS FOR CONTROLLING EMISSIONS FROM
THE MERCURY CELL CHLOR-ALKALI INDUSTRY. 1970 AND 1977
Estimated Chlorine
Production
Mercury Cell

3 ,180 ,400 tons,

2,663,800 tons
Industry
8,895,200 tons
13,375,000 tons
Required
Investment
$14,725,000
$12,650,000
Annualized
Cost

$17,344,000

$14,900,000
Year
4-40

-------
These estimates are based on the assumption that industry production
will increase at the rate of six percent per year but that mercury cell
production will decrease at the rate of 2.5 percent per year through
1977. The assumption of decreased mercury cell production is based
on a shift in economics that favors diaphragm cells, causing industry
to construct diaphragm cell plants to replace mercury cell plants as
they become obsolete.
G.
Economic Impact
'1.
Industry Structure
For the chlor-alkali industry as a whole, there were 35
firms manufacturing and selling chlorine and caustic in the open
market. These firms operated 85 plants. Thirty-one of these
are mercury cell plants owned by 16 different firms. Seven firms
operated only a single mercury cell plant. With the exception of
one firm, the single plant firms are among the major industrial
firms in the country. Most chlor-alkali manufacturers are large
multiproduct chemical companies or are vertically integrated
industries that use chlorine, caustic, or both, in their
manufacturing process. Most of the chlorine plants in the
country have a captive market for at least a part of their
production.
The chlor-alkali industry is moderately concentrated
and displays remarkably stable price schedules. However,
effective long run collusion does not appear likely even
though the nine largest chlorine-caustic producers were charged
with conspiracy and price fixing. Almost every producer uses
some or most of his production in his own operation and sells
the excess. The barriers to entry into production are not great.
The minimum efficient plant is not large compared to total
production and the technology is readily available. Finally,
the products are highly standardized and little production
differentiation is possible. In the long run, effective
competition can be expected because new sellers can enter the
4-41

-------
market or consumers can integrate backwards into chlor-alkali
production (and perhaps sell their excess production) restoring
competition, if monopoly profits develop.
Production and;Consumption
The weighted average capacity of mercury cell chlorine plants
in the United States is 244 tons per day. In 1970 approximately
8.9 million tons of chlorine valued at $667.1 million and 10.1
million tons of caustic valued at $664.9 million were produced
by all processes in the U. S. (Table 4-8). Mercury cell production
was 3.2 million tons of chlorine and 3.5 million tons of caustic
valued .at $238.5 million and $230.9 million, respectively.
In 1970, production of chlorine was down 5.5 percent from
the 1969 record high production of 9.4 million tons (Table 4-9).
This, however, is still above 1968 production. Growth in
production has been interrupted only three times since World
War II and has increased at a compound rate of 6.7 percent
since 1960.
2.
Chlorine and caustic consumption are expected to continue
growing for the next several years. The estimated rate of
growth for the next. five years approximates six percent. No
major technological changes are anticipated in the next 10 years
that will adversely affect the total demand for these products.
One new technological development promises increased electrical
efficiencies to both mercury and diaphragm cells. It is a
rhodium-plated titanium anode. .
3.
Prices and Trends
Current prices are 3.75 cents per pound or $75.00 per ton
for ,chlorine (Table 4-10) and 3.30' cents per pound or $66.00
per ton for caustic (Table 4~11). Chlorine prices are up from
last year; caustic remains the same. This continues a rather
remarkable record. There has not been a decrease in published
prices for either product since 1939, although discounting
arrangements may have at times reduced actual prices. Prices
4-42

-------
Table 4-8.
1970 VOLUME AND VALUE OF CliLORINE AND SODIUM HYDROXIDE
PRODUCED BY MERCURY CELL AND ALL PROCESSES IN THE U.S.
Tons of Product
Value
Mercury Cell Process

Sodium
Hydroxide
All Processes

Sodium
Hydroxide
Chlorine
3t498t200
$230t881t200
Table 4-9.
Chlorine
8t895t200
$667,140tOOO
10t073t700
$664,864t200
3t180t400

$238t530tOOO
U.S. PRODUCTION OF CHLORINE: 1941-1970
Year
Thousands of Short Tons
1941
1942
1943
. 1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
800.8
989.8
1t214.4
1t262.4
1,19.2~1
1t16.5 :1
1t443~2
1 't 640: 0
1,767 :0
2 t 084...2
2t51'7.9
2t608.7
2 , 797: 3 .
2 ,903~}
3,421.1
3,797. 7
3,947.7
3,604.5
4;347.1
4~636;9
4,600.8
5t142.9
5,464.1 .
5,945.2
6,517
7,204
7 ,680
8,428
9,422
8,895
Source:
Bureau of the Census, Current
Industrial Reports; Stanford
Research Institute, Chemical
Economics Handbook, 732.40308
4-43

-------
Table 4-10. CHLORINE - PRICE HISTORY
1930
1931.
1932;
1933
1934

1935
1936
1937
1938.
1939

1940
1941
1942
1943
1944

1945
1946
1947
1948
1949

1950
1951
1952
1953
1954

1955
1956
1957
1958
1959

1960
1961
1962
1963
1964

1965
'1966
1967
1968
1969

1970
Price Bases Are:
1930-1932
1933
1934-1970
Liquid
Cents
Per Pound
2.40
1.75
1. 75
1.75
1.85

2.00
. 2.15
2.15
2.15
1.75

1.75
1.75
1.75
1.75
1.75

1.75
1.75
2.00
2.25
2.40

2.55
2.70
2.70
2.70
2.93

2.93
3.05
3.15
3.15
3.15

3.25
3.25
3.25 ..
3.25
3.25

3.25
3.25
3.35
3.45
3.65

3.75
Dollars
Per Ton
48.0
. 35.0
35.0
35.0
37.0

40.0
43.0
43.0
43.0
35.0

35.0
35.0
35.0
35.0
35.0

35.0
35.0
40.0
45.0
48.0

51.0
54.0
54.0
54.0
58.6

58.6
61.0
63.0
63.0
63.0

65.0
65.0
65.0
65.0
65.0

65.0
65.0
67.0
69.0
73.0

75.0
Tanks. Works
Tanka. Works. Freight Equalized
Tanks. Single Units. Works.
Freight Equalized
Prices are given in the source as dollars per
100 pounds. and the price is calculated
from this. .
. Source;
.Q!l. ~ ~ Drug Reporter
4-44

-------
Table 4-11.
SODIUM HYDROXIDE - PRICE HISTORY
(CENTS PER POUND)
Liquid
Solid
(76% Na20)
1930
1931
1932
1933
1934

1935
1936
1937
1938
1939

1940
1941
1942
1943
1944

1945
,1946
1947
1948
1949

1950
1951
1952
1953
1954

1955
1956
1957
1958
1959

1960
1961
1962
1963
'1964

1965
1966
1967
1968
1969

1970

Liquid:
2.25
2.25
2.25
1.98
1.95

1.95
1.95
1.95
1.95
1.95

1.95
1.95
2.10
2.40
2.40

2.40
2.55
2.55
2.55
2.55

2.70
2.80
2.90
2.90
2.90

2.90
2.90
2.90
2.90
2. 90 ~

2.90
2.90
3.00
3.00
3.30

3.30
2.95
.2.55
2.55
2~90
2.60

2.60
2.60
2.60
2.30
2.30

2.30
2.30
2.30
2.30
2.30

2.30
2.30
2.50
3.05
3.05

3.20
3.35
3.35
3.70
3.70

3.85
4.10
4.30
4.80
4.80

4.80
4.80
4.80
4.80
4.80

4.80
4.80
5.15
5.15
5.35

5.80
Price Bases Are:
1934-1946
1947-1970
49-49% NaOH. Sellers'
Tanks. Works
50% NaOH. Sellers' Tanks.
Works. Dry Basis
Solid (76% Na20): Price Bases Are:
1930-1933
1934-1966
1967-1970
Drums, Car10ts
Drums, Carlots, Works
700-pound Drums, Carlots,
Works
"
//
//
,
Source: Qil, ~ ~ Drug Reporter
/'
/
/
/
4-45
,

/1

-------
have increased at a compound rate of 1.4 percent for chlorine
and 1.3 percent for caustic since 1960.
The lower purity' caustic from diaphragm cells is normally
sold at a discount with the amount of discount depending on the
I .
quantity sold, destination, and the region of the country.
Along the Gulf Coast the market is softer than in the Northeast
and the discount may reach 10 percent. In the Northeast, the
discount is generally five to seven percent (Charles River
Associates, 1969).
Price Impact
Prices for chlorine and caustic can be expected to go up.
The industry as a whole is competitive and profit margins are
so narrow that the cost increase cannot be entirely absorbed.
The portion of the market served by mercury cell plants,
about 35 percent, is so large that diaphragm cell plants cannot
serve the entire market without shortages that would put upward
pressure on prices. This would allow at least some mercury
cell plants to survive but not without an overall increase
in prices.
The size of the price increase is a question not easily
answered. An increase of $0.92 per ton across the entire
4.
1970 production of chlorine and caustic would cover the
increased production costs and would be the upper limit to
price increases. However, about 65 percent of the industry
is not affected by the cost increase. If a price increase
in the range of $0.90-$0.95 was attempted, price cutting
by diaphragm cell plants to capture a bigger share of the
mark~t could be expected so that an increase of this size
could not be maintained. Where and when prices would
eventually stabilize is unknown. An increase of $0.55
to $0.65 might be a reasonable estimate.
4-46

-------
5.
Impact on Competition
Control costs of the amount reported in Table 4-7 should
have little adverse affect on the companies producing chlor-alkali.
Most have captive markets where chlorine or caustic are minor
costs in the product produced for at least part of their
production. Three-fourths of the companies have both mercury
cell and diaphragm cell plants. While eight firms use only the
mercury cell process, only one depends on chlor-alkali production
for the bulk of its sales. Only one multiplant firm has only
mercury cell plants.
The biggest impact on competition will be on that between
mercury cells and diaphragm cells. It appears that the
diaphragm cell will now be more economical but the mercury
cell will not be so adversely affected that it will be
entirely abandoned.
6.
Investment Impact
Firms in the chlor-alkali industry should have no
difficulty obtaining the necessary capital to install air
pollution control equipment. Sales from the average
size chlor-alkali plant are in excess of $12 million. The
construction costs of such a plant are high enough that
only financially strong companies are in the industry. For
most firms in the industry, chlor-alkali sales make up only
a minor part of total sales.
4-47

-------
Selected References
1.
Charles River Associates, Incorporated, 1969, Economic Analysis of
Mercury, Prepared for Property Management and Disposal Service,
General ServiceaAdmlnistration, Washington, D. C., July 1969.
I
2.
Battelle Memorial Institute, 1971, Basis for National Emission
Standards for Mercury, Unpublished first draft of the final
report for EPA Contract EH 50-71-33, July 1971.
3.
Bureauo£ Mines, Minerals Yearkbook. 1970, U.S. Depaitment of the
Interior, Washington, D. .C., Volume I, 1971.
4.
Bureau of Mines, Mineral Industry Surveys, U. S. Department of the
Interior, Washington, D. C., Quarterly 1970, 1971, and 1972.
5.
Bureau of Mines, Mercury Potential of the United States,U. S. Department
of the Interior, Bureau of Mines Circular 8252, Washington, D. C.,
1965.
6.
Bureau of Census, Current Industrial Reports, U. S. Department of
Commerce, Washington, D. C., 1967.
4-48

-------