-------
ERRATA FOR SECONDARY ZINC INDUSTRY EMISSION CONTROL PROBLEM
DEFINITION STUDY. PART 1 - TECHNICAL STUDY FINAL REPORT
Title Page - Insert the date "May, 1971" after the statement "Conducted by
the AIR POLLUTION CONTROL OFFICE, EPA, in cooperation with THE NATIONAL
ASSOCIATION OF SECONDARY MATERIAL INDUSTRIES".
Page (V-4) - At the bottom of this page, begin new paragraph by inserting
the following line:
"Findings of this study indicate that there are significant carbonaceous"
Table VII-1 (Page 1 of 2) - In Column 7, the fourth item from the top, delete
the decimal point that precedes the number "32".
Page (VIII-14), second line - Delete "thorough" and substitute "through".
-------
Final Report
on
SECONDARY ZINC INDUSTRY
EMISSION CONTROL PROBLEM DEFINITION STUDY
PART 1 - TECHNICAL STUDY
Conducted by the
AIR POLLUTION CONTROL OFFICE, EPA
in cooperation with
THE NATIONAL ASSOCIATION OF
SECONDARY MATERIAL INDUSTRIES
Prepared by
William 0. Herring
APCO
Air Pollution Control Office
Environmental Protection Agency
411 West Chapel Hill Street
Durham, North Carolina 27701
-------
Abstract
Secondary Zinc Industry
Emission Control Problem Definition Study
Part 1 - Technical Study
Table of Contents
Chapter No.
Introduction
Process Raw Materials
Products of Processes
Summary of Processes, Process Effluents and
Emission Points
Analysis of Emitting Processes and Development
of Hypothesis on Emission Generation
Emission Determinations and Correlation with
Process Variables
Process Modelling
Emission Control Systems - Past, Present
and Conceptual
Conclusions
Emission Control Concepts
I
II
III
IV
VI
VII
VIII
IX
X
Derivation of Values for Process Models
Calculation of Amounts of Compounds
Composing Particulate Emissions, for Chapter VI
Appendix
A
B
References
-------
Secondary Zinc Industry
Emission Control Problem Definition Study
Part 1 - Technical Study
Abstract
Effluents containing particulate and gaseous emissions are exhausted from
certain processes of the secondary zinc industry. Data on those emissions
with related emission-control and process data were obtained from a sampling
of plants. These data were evaluated to obtain the results of this study.
t
Process materials consumed by this industry consist principally of scrapped
items that contain metallic zinc. Small amounts of chloride fluxing compounds
are also consumed in some secondary zinc processes. • The principal processes
of the industry are sweating and distillation which are applied to recovery
metallic zinc.
Sweat processing is conducted to produce finished zinc alloys; it is also a
usual preliminary step to distillation, providing the crude zinc-alloy feed
for the latter. Distillation processes produce zinc metal and zinc oxide,
both of virtually 100% purity.
Emissions from sweat processes occur at very low rates where the processed
zinc scrap contains only small amounts of impurities. However, very significant
emissions of particulate and gaseous carbonaceous substances, and particulate
zinc oxide and zinc chloride may occur where there are substantial amounts of
impurities in the scrap. Smaller amounts of other metal oxides, metal chlorides,
and ammonium chloride may also be emitted under this condition.
-------
(Abstract - page 2)
Emissions from distillation processes contain zinc oxide as the principal
constituent. These emissions occur at significant rates, but are satisfac-
torily controlled using baghouses.
The results of this study indicate that emission factors for zinc sweat pro-
cesses range from zero to around 32 pounds of particulate per ton of scrap
material processed. The results further indicate that the emitted particulate
may be composed of zero to 34% ZnCl2; 40 to 100% ZnO; and small percentages of
carbonaceous substance (in one instance 10%). These values do not take into
account the sweat processing of scrap containing large amounts of organic
material (e.g., assemblies that contain gaskets, lubricants, etc,.), where car-
bonaceous emissions might preponderate. Such scrap is usually subjected to
preliminary sweat processing, using afterburners in some instances, that satis-
factorily incinerate the organic material and resulting carbonaceous emissions.
Sweat-process emissions are alleviated by selection of processes that appear
optimum for the type of scrap being processed and by application of established
types of gas cleaning equipment. High collection efficiencies have been obtained
in such equipment applications. However, emission control problems have not
been solved for processing all types of zinc scrap material. Limitations are
imposed on endurance and effectiveness of gas cleaning devices, used in this
way, by the following occurrences:
a. Corrosion of metallic fabricating materials and organic bag
filter materials caused by emitted chlorides.
b. Blinding of dry fabric filters by adhering carbonaceous particles
and deliquescent ZnCl_.
c. Blinding of irrigated fabric filters by adhering carbonaceous particles.
d. Abrasive wear of fibrous glass bag-filter material during cleaning.
-------
(Abstract - page 3)
e. Limited temperature resistance of organic-fabric-filter materials.
f. Tar-like deposits and ignition hazards associated with carbona-
ceous emissions.
Process model units were formulated for study of present and conceptual emission
control systems.
Several emission-control concepts were developed based on findings of this study,
and recommendations are made for further research and development.
-------
CHAPTER I
Introduction
Contents Page No.
PROBLEM AREA - PROCESSES STUDIED 1-1
SCOPE OF STUDY 1-1
APPROACH TO STUDY , 1-2
REFERENCES TO DATA SOURCES 1-2
EXTENT OF THE INDUSTRY AND POTENTIAL BENEFITS
OF EMISSION CONTROL R & D 1-3
-------
Chapter I
Introduction
PROBLEM AREA - PROCESSES STUDIED
Gaseous exhaust effluents containing deliquescent and corrosive metal chlorides
are emitted from certain processes of the secondary zinc and aluminum industries.
The study reported here was conducted to define the problems of controlling emis-
sions, so characterized, that result specifically from secondary-zinc processes.
As dealt with here, the secondary zinc industry consists of those plants that
i
process discarded and scrapped items and materials that contain metallic zinc
L
for the primary purpose of recovering that metal. The principal processes
employed are sweating and distillation. This study does not include reduction
processes applied to obtain metallic zinc from zinc oxide contained in waste
materials. In present practice, most materials of that type are processed through
primary smelting establishments. Also excluded from this study is the processing
of zinc-process wastes, which are principally of a chemical nature, to produce
chemical products - that processing usually being done in plants of the chemical
and other non-metallurgical industries. (Study of processes excluded here might
be taken up in "Reduction in Belgian Retorts" AFEM, pp. 294-6; and "Sal Skimmings"
and "Chemical Residues" Mathewson, pp. 319-21.)
SCOPE OF STUDY
The problem definition study is intended to determine a basis for research and
development to improve emission control capabilities of the industry studied.
-------
(1-2)
The results of the study may also be used as a source of data for additional
purposes brought about by legislation, including an annual report to Congress
and the setting of emission control standards required under the Clean Air
Act with Amendments of 1970.
Part 1, reported here, covers the technical phase of a study to define the
problems of controlling emissions of the secondary zinc industry. Part 2 of
this study is planned to cover the economic phase - to show the degree of
emission control that can be attained for specific process situations, the
cost of attaining that control, and situations where satisfactory control can-
not be attained because of cost.
APPROACH TO STUDY
The approach in conducting this study was to hold discussions between APCO and
NASMI representatives; review available literature; and visit a small sampling of
plants that were selected through the office of NASMI as being representative
of the range of processing and resulting emissions of the industry. Data ob-
tained through this investigation were evaluated to quantitatively define emis-
sion control problems, reveal gaps in existing emission control technology, and
develop concepts, to be considered for research and development.
REFERENCES TO DATA SOURCES
Data used in this study were obtained from plants of cooperating companies and
from published sources. The industrial plant sources are treated confidentially
and therefore not cited within the report. Published sources are cited within the
report.
-------
Cl-3)
EXTENT OF THE INDUSTRY AND POTENTIAL BENEFITS OF EMISSION CONTROL R&D
It is estimated that secondary production of zinc, aluminum, and copper alloys
accounts for 20%, 30%, and 45%, respectively, of total consumption of those
metals in the United States (NASMI Studies, p. 12; see also Minerals YB).
Although the study reported here deals specifically with zinc processing, it
is noted that emissions from the aforementioned three types of secondary metal
processing have characteristics and constituents common to all. Deliquescent,
corrosive metal chlorides are common to zinc and aluminum process emissions.
Zinc oxide makes up large percentages of particulate emissions from both secondary
zinc and copper-alloy processing. Carbonaceous emissions are common to all
three industries processing these metals. It may, therefore, be anticipated
that emission control technology, developed for secondary zinc, will be appli-
cable at least in part to emission control efforts in the other industries.
Therefore, the study reported here, while concerned specifically with the secon-
dary zinc industry, also constitutes part of a greater effort to improve emission
control capabilities of producers, which in aggregate make up a larger part of
the metal-producing industries.
-------
CHAPTER II
Process Raw Materials
Contents Page No.
ZINC SCRAP MATERIALS II-l
BUREAU OF MINES CLASSIFICATION II-l
NASMI CLASSIFICATION II-2
TECHNICAL CLASSIFICATION II-2
FLUXES II-5
FUELS II-6
-------
Chapter II
Process Raw Materials
Raw materials used in secondary zinc processing are zinc scrap materials,
fluxes, and fuels for furnaces. These are described below.
ZINC SCRAP MATERIALS
For purposes of statistical presentation, one system of classification and
nomenclature of zinc scrap materials has been developed and applied by the
Bureau of Mines (Minerals YB). For purchase specification purposes, another
system has been developed by NASMI and is applied within the industry (NASMI
NF-66). These systems do not readily lend themselves directly for the tech-
nical analyses of this study, but are listed below to help describe the ma-
terials. For the technical analyses of this study, an additional system was
formulated and is presented below. In this TECHNICAL CLASSIFICATION, zinc
scrap materials are divided into two main catagories and the sub-classifi-
cations as shown under that heading.
BUREAU OF MINES CLASSIFICATION:
New clippings Galvanizers' dross
Old zinc Diecastings
Engravers' plates Rod and die scrap
Skimmings and ashes Flue dust
Sal skimmings Chemical residues
Die-cast skimmings
-------
(II-2)
NASMI CLASSIFICATION:
Old zinc die cast scrap
New zinc die cast scrap
New plated zinc die cast scrap
Zinc die cast automotive grills
Old scrap zinc
New zinc clippings
Zinc die cast slabs or pigs
Galvanizers' slab zinc dross
Prime zinc die cast dross
TECHNICAL CLASSIFICATION:
A. Metallic Scrap. This scrap consists of metallic items, generally in the
same shapa as when manufactured or used.
1. Unplated zinc castings. Examples are reject castings, off-grade ingots,
old castings, and printers plates. Castings in this classification are free of
significant attachments made of other metals. They range from very clean cast-
ings to castings having relatively small amounts of oil and/or paint coatings,
dirt, and other impurities.
2. Plated zinc castings. These are mainly automobile grills, having chrom-
ium platings. They are considered herein as having very little, if any, contam-
ination with oils, paints, other organic materials, and dirt and only small amounts
of higher melting-point metal attachments.
3. Zinc fabricating scrap. This scrap consists of that obtained from fab-
ricating operations; examples are cuttings, punchings, chips, borings, turnings,
and routings. They are considered here as being reasonably clean, except for
coatings of oil or cutting compounds.
4. Contaminated zinc die-cast scrap. This scrap consists of assemblies con-
taining zinc die-castings; attachments made of other metals; and materials that
I
contain carbon compounds such as gaskets, electric insulation, and lubricants.
-------
(II-3)
Examples are automobile fuel pumps, carburetors, horns, and washing machine
parts. (Attachments made of metals having higher melting points than zinc may
be referred to as "unmeltable attachments" or "unmeltables.")
B. Residual Scrap Materials. These are residues and impure alloys formed in
the melting, fluxing, and application of molten-metal baths in galvanizing, die
casting, and other processes. These materials are referred to, generally, as
skimmings (or residues) and drosses. In this study, the terms skimmings and
residues refer specifically to materials that form above metal bath surfaces.
These materials are composed preponderantly of non-metallic substances including
metal oxides and residual flux, with lesser amounts of metal contained as parti-
cles (or inclusions). They are of non-metallic appearance. The term dross, in
this study, refers to materials that form within molten-metal baths, at top
surfaces and at bottoms of melting vessels. They are composed mostly of metallic
zinc and are metallic in appearance. As defined here, these terms are at variance
with some industrial usage, as is noted below.
1. Skimmings (or residues).
a. Galvanizers' skimmings (ashes). This material is formed by
oxidation of metal on galvanizing bath surfaces when no flux blanket is used.
It is skimmed from above the molten metal bath surface. Skimmings are pulveru-
lent, composed mostly of ZnO, with metallic inclusions. They are formed as
galvanized items are withdrawn from baths, creating turbulence at the bath sur-
face. Chlorine (as chlorides), derived from flux coatings on stock being galvan-
ized, may be present in these skimmings in amounts from 0 to 12%. (The processing
of "sal skimmings," formed on a galvanizing bath when a flux blanket is used, is
not included in this study. See Chapter Io)
-------
b. Die-cast skimmings. These materials are formed by accumulation
of metal oxides, ZnCl and/or NH.C1 flux, and other impurities above surfaces of
zinc-alloy baths used for die-casting and are skimmed off. Skimmings are composed
mostly of ZnO, with metallic inclusions making up 3 to 10%. Chlorine (as chlorides)
ranges from 0 to 3% (approximately equivalent to 0 to 6% anhydrous ZnCl-). The
metallic inclusions are zinc, containing copper and aluminum, derived from die-
cast alloy.
2. Dross
a. Top dross
(1) Galvanizers' top dross (also referred to as "galvanizers' top
skimmings"). This material is formed by iron-aluminum compounds floating to the
surface of galvanizing baths. These compounds result from reactions of aluminum
with iron during certain galvanizing processes where aluminum is added to the
baths to prevent a brittle layer from being formed in the coatings. They melt at
a higher temperature and are lighter than zinc; therefore, they tend to separate
and solidify at the bath surface. The product is skimmed from tops of baths and
cast into chunks. It contains uncombined zinc and metal oxides, in addition to
the aforementioned Fe-Al compounds. Top drosses contain around 90 to 95% Zn,
2 to 5% Al, are generally free of chloride flux but may contain small amounts.
(2) Die-cast dross. This material is formed at the top of
die-casting process baths and is similar to galvanizers' top dross, but with
different metal contents. Zinc content approximates 85%. Copper and aluminum
are also present. There may be some chloride-flux content.
-------
(II-5)
b. Bottom dross
(1) Galvanizers' dross* This material is formed by liquation
of an iron-zinc compound to the bottom of galvanizing baths; this settling
results from a higher specific gravity than zinc. In some galvanizing processes,
dross may rest on a layer of molten leado Removed from the bath with spoons
and cast into chunks, the resultant product contains uncombined zinc in addition
to the Fe-Zn compound. Some lead may also be present as well as chloride flux,
the latter being picked up from the bath surface during dross removal.
(References: Nonferrous, pp. 63-74; Mathewson. pp. 315-21, 469)
FLUXES
The main fluxing materials used in secondary zinc processing are ZnCl- and NH Cl.
These may be applied to the metallurgical process bath, or they may be contained
in residual scrap as obtained, as noted in the foregoing description of those
materials.
Other"smokeless fluxes" are in limited usage for processing relatively clean
scrap. Cost is considered too high and effectiveness in emission control too
limited to provide solutions to general processing and emission control problems.
Application of these fluxes does not appear to fall within the problem area of
this study and is therefore dismissed from further consideration in this report.
-------
(II-6)
FUELS
Natural gas and fuel oil are the principal fuels used in secondary zinc process-
ing. Based on the operations of plants visited during this study, it is believed
that the fuel used most is natural gas. Oil is used in smaller facilities and as
a standby fuel. Electricity is sometimes used to heat furnaces. This use of
electricity appears to be unusual and limited to processing clean scrap that does
not pose air pollution problems.
-------
Chapter III
Products of Processes
Products of secondary zinc plants are listed and defined as follows:
A. Specification zinc alloys. Standard alloys, such as die-casting types,
produced by sweat-processing and re-alloying zinc metallic scrap.
B. Zinc-content metal. Zinc-containing metal produced by sweat-processing
metallic and residual zinc scrap material in preparation f6r distillation.
C. Distilled slab zinc. Zinc-containing metal approaching 100% Zn purity
produced by distillation of metal derived from zinc scrap materials.
D. Zinc dust. Zinc-containing metal produced by distillation of metal de-
rived from zinc scrap materials. The distilled zinc vapor is allowed to condense
under conditions which form small spherical particles.
E. Zinc Oxide. ZnO approaching 100% purity produced by distillation of scrap-
derived zinc with subsequent oxidation of vapor by atmospheric combustion.
F. By-product residues containing ZnO (for reduction to metallic zinc by
primary smelters).
1. Sweat-process residues. These are residues that remain after metal has
been extracted from metallic or residual zinc-scrap material. Some chloride flux
may be retained in these residues.
-------
(III-2)
2. Residues from water-wash pre-treatment of residual zinc-scrap
material, applied to extract metal inclusions. These may be chemically treated
and/or calcined to reduce chloride content. (See PRE-TREATMENTS, Chapter IV)
G.< By-product participates containing ZnO (for agricultural soil treatment).
These are emitted particulates collected by control equipment (usually baghouses)
having a maximum chloride content of 5%. Commercial usage of collected particulates
has not been determined for collected particulates having over 5% chloride content.
H. By-product distillation residues containing Al and Cu (for use in aluminum
alloying)o These are removed from distillation furnaces.
I. Ferrous and non-ferrous unmeltable attachments to zinc base die castings.
-------
Chapter IV
Summary of Processes, Process Effluents and Emission Points
Contents
PRETREATMENTS
Page No.
IV-1
SWEAT PROCESSES
IV-2
KETTLE FURNACE
REVERBERATORY FURNACE
ROTARY FURNACE
MUFFLE FURNACE
ELECTRIC-RESISTANCE FURNACE
IV-3
IV-4
IV-5
IV-6
IV-6
DISTILLATION PROCESSES
RETORT FURNACE SYSTEM
MUFFLE FURNACE SYSTEM
List of Tables
Title
Emission Points and Effluents of Secondary Zinc-Sweat Processes
Emission Points and Effluents of Secondary Zinc-Distillation
Processes
IV-7
IV-7
IV-9
List of Figures (Process Flow Diagrams)
Title Fig. No.
Sweat Processing of Zinc-Scrap Materials in Kettle Melting Furnace 4-1
Sweat Processing of Zinc-Scrap Materials in Reverberatory 4-2
Melting Furnace
Table No.
IV-2
IV-2
(2pages)
-------
Chapter IV
Summary of Processes. Process Effluents and Emission Points
PRETREATMENTS
Treatments, preliminary to melting operations, are applied to some types
of zinc-scrap material.
Attachments that are accessible and can be removed easily are removed from
contaminated zinc die-cast scrap. Attachments and impurities that are
often not removed include gaskets, sealed-in lubricants, screws, and elec-
trical parts. Considerable organic materials and metals, other than zinc,
are therefore retained in this scrap when it is charged to melting (sweating)
furnaces.
Concentration of metallic zinc in skimmings is increased by ball-mill pul-
verizing, followed by pneumatic treatment and/or screening to remove part
of the pulverulent, non-metallic constituents.
In some instances, skimmings are crushed and then treated in the following
manner. The crushed skimmings are washed with water to separate non-metals
as a slurry and allow zinciferous metal particles to settle out; the slurry
is then treated with Na.CO to convert chlorides (mainly ZnCl ) to NaCl, form-
ing insoluble Zn(OH) . Most of the NaCl is separated from the insoluble
residues by filtration and settling; the residue is dried and calcined in a
kiln to convert Zn(OH)? to ZnO by driving off H«0 and vaporizing any remaining
ZnCl-. The calcined product is mostly ZnO and is suitable for smelting.
The kiln fume is collected in polyester fabric bag filters and recycled.
-------
av-2)
Emissions from the aforementioned pretreating processes were not studied
in depth during this investigation, since it did not appear that there were
notable difficulties in their control, there being no reports from industrial
representatives indicating any problems. These emissions are therefore not
dealt with further in this report.
SWEAT PROCESSES
In sweat processing, heat is applied to scrap materials, which may be of
either the metallic or residual types, to melt and separate metallic zinc
from metal attachments, having higher melting points, and from non-metallic
residues. Any organic materials in scrap are also burned off during sweat-
ing. Sweat processing is accomplished by charging the scrap into a melting
furnace. The charge may be worked, by agitation or stirring during melting;
and chloride flux may be present either as residual flux, in charged residual
scrap, or as flux added to the charge. Working and fluxing of the charge
are done to help effect the desired metal separation. A molten-metal bath
is formed from the metallic zinc (with dissolved alloy metals). Non-metallic
residues, along with some platings, form above the molten-metal bath surface
and are skimmed off. Unmeltable attachments settle to the bottom and are
removed. The molten metal may then be (1) cast directly into blocks for
subsequent further processing,or (2) fed directly to a distillation furnace,
or (3) it may be sampled and analyzed, and then alloyed by adding metals to
obtain specification composition, and then cast as ingots.
Types of furnaces used for sweating zinc-scrap materials are discussed in
the following order:
1. MeIting-kettle (or kettle) furnaces
2. Reverberatory furnaces
3. Rotary furnaces
-------
(IV-3)
4. Muffle furnaces
5. Electric-resistance furnaces
Of these, most usage is of the first three, which are therefore given prin-
cipal attention in this chapter,. Figures 4-1 and 4-2 are schematic flow
diagrams that show sweat processing in kettle and reverberatory furnaces.
The rotary furnace is, in effect, a mechanical modificatiin of the rever-
beratory furnace, as will become apparent in the description of that furnace
in a subsequent paragraph. Emission points of sweat furnaces and effluents
emitted from those points are shown in Table IV-1. These furnaces, their
applications, emission points, and effluents are described further in the
paragraphs below.
KETTLE FURNACE
The kettle furnace consists of a melting vessel (kettle), made of cast iron
in most cases, mounted over a combustion chamber,, Scrap materials, which
may include metallic and/or residual types, are charged into the kettle.
The metallurgical-process bath is formed as zinciferous metal is melted and
residues form above the molten-metal surface. Operating temperatures of
kettle-process baths range from 800 to 1000 F. Production is on a batch
basis, with one process heat requiring around 6 to 8 hours to process and
pour. A molten heel may be retained as finished alloy is removed from
furnaces and additional scrap (process material) charged.
Normally, products of fuel combustion are exhausted separately from emis-
sions of the metallurgical-process bath, through separate venting of the
combustion chamber. Natural gas is the generally used fuel (fuel oil being
used in a smaller number of cases).
-------
(IV-4)
As noted previously, sweat processing in a kettle furnace is shown schemat-
ically in Figure 4-1, along with emissions and emission points. Emission
points and effluents emitted from these points are detailed further in Table
IV-1.
REVERBERATORY FURNACE
The reverberatory furnace has a general box configuration with a sloped
bottom (hearth). It is used to process both metallic and residual zinc
scrap materials, which are charged into the furnace and rest on the hearth.
Burners are located in the upper part of the furnace; combustion of fuel
above the charge supplies heat to burn off organic substances, as well as
heat to melt the zinc alloys in the charge. Furnaces are designed and
burners are positioned to minimize flame impingement on the charge and to
reduce oxidation and entrainment of metal oxide particles in emissions. As
zinc alloys melt, they separate from unmeltables and flow downward over
the hearth. Bath temperatures in reverberatory furnaces are usually around
1000 F. When comparable materials are processed, these baths are usually
maintained at higher temperatures than kettle-furnace baths to increase
fluidity of molten metal and thereby improve separation from unmeltables.
Metal flows from the furnace as it melts; and, at intervals, unmeltables
are raked out and additional process material charged.
Reverberatory furnaces may be independent units or they may be integral with
distillation furnaces (Figure 205, APEM). Consideration here is limited
to the independent type of unit where molten metal from the hearth flows
through a spout into ladles or kettles. The metal may then be processed
-------
(IV-5)
further to obtain a specification alloy, or it may be fed to a distillation
furnace. Integral sweating-and-distillation furnaces are discussed further
in a subsequent section on distillation furnaces.
Sweat processing in a reverberatory furnace is shown schematically in
Figure 4-2, along with emissions and emission points. These emission points
and effluents from these points are detailed further in Table IV-1. It
might be noted that the pouring spout of the reverberatory furnace is not
listed as an emission point; the molten-metal temperature at this point is
normally not high enough to vaporize significant amounts of zinc (APEM p. 294).
ROTARY FURNACE
The melting unit of the rotary-type furnace consists of a hollow cylinder
mounted with its lengthwise axis sloped at a small angle from horizontal.
During operation, this cylinder is mechanically rotated on that axis and
internally heated by gas or oil burners0 The principal application of the
rotary furnaces at plants visited during this study was for processing con-
taminated die-cast scrap, without application of fluxing compounds. Scrap
materials are fed into the high end of the melting cylinder. As the cylinder
rotates, zinc melts and flows out through openings in the low end, usually
into a kettle where residues are skimmed off. Unmeltables are separated from
the bath by tumbling them out of the low end of the cylinder or by manual
raking and scraping. Rotary-furnace bath temperatures are usually lower than
those of kettle or reverberatory furnaces because rotation helps (1) separate
molten zinc from unmeltables, (2) maintain molten zinc and alloy metals in
solution, and (3) use heat more efficiently by avoiding localized high temp-
erature zones, thereby allowing lower bath temperatures to be applied. The
collected zinc-containing metal may then be transferred to a distillation
furnace, or its composition may be adjusted to ?" alloy specification.
-------
(IV-6)
Observations of this study indicated no visible (or significant) emissions
at the low end of the cylinder as the molten metal or unmeltables flow from
or are being removed from the melting cylinder.
The only emission point noted (in Table IV-1) is therefore, the furnace flue
(the high end of the melting cylinder). Emissions from this point are those
contained in furnace-exhaust effluent (flue gas), which consists of the same
types of constituents as those listed for furnace flue effluents from rever-
beratory furnaces, except that emissions derived from flux would not normally
be contained in rotary-furnace effluent.
MUFFLE FURNACE
In the muffle furnace, as applied to sweating processes, combustion gases
are separated from charged zinc-scrap materials by a "muffle". (The same
principle is applied to distillation as shown in Figure 204 of APEM.) This
design permits separation of combustion products from those emissions derived
from charged zinc-scrap materials and flux. In this respect, the muffle
furnace is similar to the kettle furnace. Findings of this study indicate
little usage of the muffle furnace for sweating (although usage for distil-
lation is significant). Probably this limited usage is due to low thermal
efficiency. Because usage is not extensive, the muffle furnace as applied
to sweat processing is not subjected to detailed analysis in this study.
ELECTRIC-RESISTANCE FURNACE
Findings of this study indicate that electric-resistance furnaces are used
in a small number of plants for processing clean, scrap-derived zinc metal,
and that processing such scrap does not pose significant air-pollution prob-
lems regardless of furnace type. It appears that by avoiding fuel-combustion
products, application of electric-resistance furnaces might have an emission-
-------
(IV-7)
control advantage. However, cost of electric energy probably precludes
usage of electric-resistance furnaces for sweat processing in nearly all
situations (except for melting clean metal). This type of furnace is there-
fore not subjected to detailed analysis in this study.
DISTILLATION PROCESSES
Distillation processes are of several variations (APEM pp. 296-9; Mathewson
p. 317) which can be reasonably represented by the retort furnace system and
the muffle furnace system. Figures 203 and 204 of APEM show examples of
distillation processing by means of these systems. Table IV-2 of this report
details their emission points and the effluents emitted. Further description
is provided in the paragraphs below.
RETORT FURNACE SYSTEM
The retort furnace system consists of two units: (1) a retort furnace and
(2) a condenser (Figure 203, APEM). In the retort furnace, the distillation
retort (which may be bottle-shaped) is mounted inside the furnace closure
where fuel is burned, heating the retort and its content. The charge to the
retort may consist of molten-zinc-rich metal obtained directly from
a sweat furnace, cast zinc-content metal from a zinc-sweat process, or zinc
dross. Zinc is vaporized in the retort, and the vapor passes through a re-
factory pipe to a condenser, where it may be condensed either as zinc dust
or molten zinc (for casting into slabs), depending on the type of condenser
used. The condenser for making zinc dust consists of a bare sheet-steel
shell. Zinc vapor entering this condenser cools rapidly and therefore
condenses into small particles. The condenser for making slab zinc also
consists of a steel shell, but is refractory lined for thermal insulation.
Zinc vapor entering this condenser cools more slowly, forming liquid metal on
-------
(IV-8)
internal surfaces of the refractory lining; the liquid metal flows to
the bottom where it is tapped at intervals and cast into slabs.
Retort furnaces may also be used to produce zinc oxide by allowing zinc
vapor, from the retort, to burn in air, and collecting the resulting ZnO
product in a baghouse.
Distillation residues (mainly mixtures of aluminum and copper) are raked
out of the retort immediately after each distillation heat is completed,
while residues are still at a high temperature.
Fuel-combustion products are exhausted independently of any emissions
from the retort charge.
Emission points in the retort-furnace system are listed, with emissions
from those points, in Table IV-2 and are discussed further as follows:
A. Retort opening. During removal of distillation residues from
the retort, ZnO particles are emitted as the molten zinc that remains in
the residues continues to vaporize and oxidize. Ambient air composes the
gaseous part of effluent formed at this point.
B. Pressure-relief valve or "speise" hole of condenser. One or the
other of these devices is used to retain a positive pressure and exclude air
from condensers (APEM p. 297). Emissions from the "speise" hole consist of
nearly-pure ZnO. Emissions from condenser relief valves are mixtures of
zinc dust and ZnO. In heats where- dross is contained in the retort charge,
a small amount of chloride particulates, derived from residual flux, might
also be emitted at these points. Ambient air composes the gaseous part of
effluents formed at these points.
-------
(IV-9)
MUFFLE FURNACE SYSTEM
The muffle-furnace system consists of a melting unit (reverberatory furnace)
and a vaporizing unit (muffle furnace), combined in an integral structure,
and Includes a condenser (APEM p. 298). Materials such as contaminated zinc
die-cast scrap are charged into the melting unit. As zinc alloys melt, the
molten metal flows from that unit to the bottom of the vaporizing unit.
Fuel combustion in the vaporizing unit takes place in the upper chamber of
that unit, which is separated from the molten metal and zinc vapor (in the
lower chamber) by an arched partition (muffle). Combusion gas from the
upper chamber is exhausted to the melting unit, adding heat to that of
melting unit burners to help melt the charged scrap material. Zinc vapor
is channeled from the vaporizing unit to the condenser where it is partly
condensed to liquid metal and cast as slabs. The non-condensed vapor is
oxidized to ZnO, which is collected in a baghouse.
The melting unit is charged at one end; and unraeltable attachments and
skimmings are removed from the other. Flux is not applied to, or contained
in, the melting-unit charge.
In the melting unit incineration of organic materials in charged metallic
scrap is virtually complete so as to prevent emission of carbonaceous par-
ticulates in significant amounts.
A vibrating screen is used at the discharge end of the melting unit to
separate skimmings from unmeltables after removal from the unito This
device is a source of considerable amounts of particulate (mainly ZnO)
emissions (APEM p. 299).
Emission points and effluents from the muffle-furnace system are detailed
in Table IV-2.
-------
r'lt. i-l - Sweat ::roc:esslii6 of ilnc-3cr=p Hater-Ibis In Kettle Melting, rurnbce
r.li, tt-.u.
ri-crlcbtlnt sercji.
Jcc:ti: nln^tea c-i^-cust
Top ^-ross.
(i.lioyil1b JjCtbls ..il^.:i
t;lso LI. saceai
uf rlux appllea :
, NH^Cl.
'..tslila-l Tux In E
-------
Fit. 4-S - Sv.tct Processing of Zlnc-Scr&p I'
In Reverberatory Meltln6 Furnbce
Emissions from openings. In
fu.-r.bce. for char^lne,. fluxing,
una removal or unaclti-oles
Lnu sitlTmln^s - Contains
raettllaitjlcal process effluent.
fey Incluue fuel combustion
prouucts.
Furnuce flue JUJF -
contblns fuel comuastlon products
anc metullurfclcbl process effluent.
Tyoes of fuels used:
Fuel oil.
Types of zlne-aerbp
s procegsec:
Metallic scrap, til types.
SKlmmlritS.
Top aross.
Types ol' flux bppllea:
ZnCl , NH Cl.
fveslauc.1 flux In sltlmmir.bs
, NK4C1, fand
Types of procucts:
ilr.c-cont&lnlne, raetsl
(for Glstlllbtlon or fc
to specification).
iiy-prouuct reslcues.
-------
TABLE IV. I EMISSION POINTS AND EFFLUENTS OF SECONDARY ZINC-SWEAT PROCESSES
Process (defined by type
of furnace used)
Kettle Furnace
Reverberatory Furnace
Rotary Furnace
Emitting Process
Equipment Unit
Melt Kettle
Combustion chamber
Combustion chamber,
containing melting
hearth. (May also be
referred to as
"sweating chamber").
Rotating, cylindrical
melt unit. (May also
be referred to as
"sweating chamber" or
"combustion chamber").
Emission Point of
Process Equipment Unit
Top of melt kettle (or
surface of metallurgical
process bath, formed from
charge) .
Combustion chamber vent
Furnace flue (exhausting
combustion chamber).
Openings for charging and
fluxing; and removal of
unmel tables and skimmings.
Furnace flue (high end of
melting cylinder).
Constituents of Effluents from Emission Points
A. Emissions from process charge.
1. Products of combustion or thermal decomposition of
organic materials In charge.
2. Emissions derived from metals, fluxes, and residues
In metallurgical process bath, Including metal oxides
resulting from presence of air contacting metal.
B. Air induced or Infiltrated Into exhaust effluent stream.
A. Products of combustion of fuel (usually natural gas).
A. I. -2. Emissions from process charge (same items as
listed above).
B. Air Induced or infiltrated Into furnace, thence Into flue.
(would be in excess of air consumed In combustion).
C. Products of combustion of fuel (usually natural gas).
A. 1.-2.; B. ; and C. 'Same items as listed above for
furnace flue. These effluents are formed from (1) gases
escaping from the furnace; (2) emissions from the molten
metal and skimmings being withdrawn from furnace with
unmel tables; and (3) ambient air.
A. 1.-2.; B. ; and C. 'Same as listed above for flue of
reverberatory furnace, except emissions derived from
flux are not normally contained in rotary furnace
effluent.
-------
TABLE IV-2 (page 1 of 2) EMISSION POINTS AND EFFLUENTS OF SECONDARY ZINC-DISTILLATION PROCESSES
Process (defined by
furnace system used)
Emitting Process
Equipment Unit
Emission Point of
Process Equipment Unit
Constituents of Effluents from Emission Points
Retort Furnace System
(See Fig. 203,APEM)
Distillation Retort
Combustion chamber
Condenser
Condenser
Opening of distillation retort.
(Emissions occur during removal
of distillation residues. This
opening may be referred to as
"charging hole".
It is used for applying charge
to retort and removing residues).
A. Emissions from distillation residues.
1. ZnO (makes up most of emitted participates).
2. Oxides of other metals (mainly Al 0 ) - small amount.
B. Ambient air.
Combustion chamber vent.
A. Products of combustion of fuel (usually natural gas).
Pressure relief valve.
(Used to retain positive pressure
and exclude air from condenser).
A. Emissions from vapors distilled from retort charge,
partly oxidized by residual air In retort-and-condenser
system.
1. ZnO partlculate.
2. Metallic zinc dust (particulate).
3. Chloride particulates, derived from flux. (Very small
amounts may occur when dross is charged to retort).
B. N from residual air in retort and condenser.
"Speise" hole. (Used instead
of pressure relief valve for
same purpose. Zn vapor escaping
from condenser thru speise hole
is ignited In air, producing
ZnO particulates).
Emissions from vapors distilled from retort charge and
oxidized.
1. ZnO partlculate, nearly pure. (The charge to retort
is of molten metal which would not produce chloride
emissions - APEM p. 296).
B. N from residual air In retort and condenser.
C. Ambient air
-------
TABLE IV-2 (page 2 of 2) EMISSION POINTS AND EFFLUENTS OF SECONDARY ZINC-DISTILLATION PROCESSES
Process (defined by
furnace system used)
Muffle Furnace System
(See Fig's. 204-5, APEM)
Emitting Process
Equipment Unit
Melt Unit
(reverberator; furnace)
Vaporizing Unit
(muffle furnace)
Emission Point of
Process Equipment Unit
Flue of melt-unit combustion
chamber.
Openings in melt unit for
charging scrap material and
removing unmeltables and
skimmings.
Tap hole of vaporizing unit.
(Emissions occur during
removal of distillation
residue).
Constituents of Effluents from Emission Points
A. Emissions from melt-unit charge.
1. Products of combustion or thermal decomposition of
organic materials in charge.
2. Emissions derived from metals and residues (usually
no flux) In metallurgical-process bath, including
metal oxides resulting from presence of air.
B. Air Induced or infiltrated Into melt unit, thence Into
flue. (Would be in excess of air consumed in combustion).
C. Products of combustion of fuel (usually natural gas),
burned In melt unit and vaporizing unit.
A. 1., 2.; B. ; and C. -Same items as listed above for
furnace flue. These effluents are formed from
(1) gases escaping from the melt unit; (2) emissions
from the molten metal and skimmings being withdrawn
from melt unit with unmeltables; and (3) ambient air.
A. Emissions from distillation residues.
1. ZnO partlculate, nearly pure (APEM p. 299)
B. Ambient air.
-------
Chapter V
Analysis of Emitting Processes and Development^
of Hypothesis on Emission Generation
Contents Page No.
PROCESSES SELECTED FOR ANALYSIS V-l
DEFINITIONS OF TERMS EMPLOYED IN ANALYSIS V-2
PROCESS OPERATIONS THAT EFFECT EMISSIONS V-3
MATERIALS AND FUELS APPLIED TO EMITTING PROCESS EQUIPMENT UNITS V-4
ORGANIC CONSTITUENTS OF CHARGES V-4
METALLIC CONSTITUENTS OF CHARGES AND RESULTING COMPOSITION
OF MOLTEN METAL BATH V-5
FLUXING COMPOUNDS IN CHARGES V-7
INORGANIC IMPURITIES IN CHARGES V-7
FURNACE FUELS V-8
EMISSIONS - THEIR GENERATION AND COMPOSITION V-8
CONSTITUENTS THAT MAKE UP EFFLUENTS V-8,
GENERATION OF CARBONACEOUS EMISSIONS FROM CHARGES V-10
GENERATION OF NONCARBONACEOUS EMISSIONS FROM CHARGES V-10
FUEL COMBUSTION PRODUCTS V-17
-------
Chapter V
List of Figures
Title Fig. No.
Vapor Pressures of Principal Constituents of
Zinc Sweat Process Baths 5-1
List of Tables
Title Table No.
Data Pertaining to Melting, Vaporizing, and Condensation
of Zinc-Sweat Process Materials V-l
-------
Chapter V
Analysis of Emitting Processes and Development
of Hypothesis on Emission Generation
PROCESSES SELECTED FOR ANALYSIS
As shown in Chapter IV, chloride fluxes and organic materials are con-
tained in charges to zinc-sweat furnaces in quantities that may be sub-
stantial. By comparison, very little if any flux and no organic materials
are contained in charges to retorts or vaporizing units of distillation
furnaces. Chloride emissions are derived from chloride fluxes and
carbonaceous emissions from organic materials in furnace charges. Exhaust
effluents from zinc-sweat processes may therefore contain substantial
amounts of chlorides, as well as carbonaceous substances; very little
chlorides and no carbonaceous substances are contained in emissions from
distillation processes. Therefore, in this analysis, attention is con-
fined to the operations that make up zinc-sweat processes, the materials
entering into those processes, and the resulting emissions. It is noted,
however, that because of zinciferous constituents common to emissions
from both zinc-sweat and zinc-distillation processes, certain technological
factors are common to controlling emissions from both types of processes.
For this reason, distillation process emissions and their control will
be discussed in later chapters. In this way, emission control technology
for all secondary zinc processes will be considered together where there
is common applicability.
-------
(V-2)
DEFINITIONS OF TERMS EMPLOYED IN ANALYSIS
Effluent. The term "effluent" or "exhaust effluent" refers to the gas
stream that flows from the region of the charge or bath being processed
in a melting (sweating) furnace, and is then exhausted from the furnace.
The effluent includes any particulate or gaseous emissions derived from
the charge and any atmospheric air that enters the effluent stream. Fuel
combustion products that are mixed into that stream also become a part
of the effluent. Thus, combustion products of fuel consumed in rever-
beratory and rotary furnaces are part of the effluent; combustion products
of fuels consumed in kettle furnaces are exhausted separately and are not
part of the effluent. Fuel combustion products occuring by themselves are
not within the problem area of this study, and are therefore not included
in this definition or the following analysis.
Infiltrated air. The term "infiltrated air" refers to all atmospheric
air that enters the effluent. It includes (1) air induced into the
effluent by the furnace draft, (2) excess combustion air in reverberatory
and rotary furnaces, and (3) air that mixes with emissions escaping from
furnace openings used for charging, removing unmeltables, etc.
Metallurgical-process bath. The term "metallurgical-process bath" or "pro-
cess bath" refers to the bath formed during zinc-sweat processing, which
includes molten metal, residues, and any flux present. (Flux in the process
bath is understood to fuse with residues, the fused mixture tending to float
to the bath surface.)
Molten-metal bath. This term refers to the molten-metal portion of the metal-
lurgical-process bath.
-------
(V-3)
Emitting process-equipment unit. This term refers to the process-equipment
unit in which emissions are generated. In the kettle-sweat furnace, this
unit would be the melting kettle. In the reverberatory or rotary furnace,
it would be 'the combustion (sweating or melting) chamber. (See figures
4-1, 2 and Table IV-1.)
PROCESS OPERATIONS THAT EFFECT EMISSIONS
For the purpose of analyzing emissions, the operations that make up all zinc-
sweat processes are considered in two phases: (1) Melting the charge, and (2)
Fluxing and working the metallurgical-process bath. These operations may be
conducted sequentially, the first preceding the second, or partly or wholly
concurrently. Each phase is briefly summarized and related to furnace appli-
cations.
Melting the Charge. To recover metallic zinc, scrap materials are heated to
temperatures between 800 and 1100°F. In this temperature range, zinc is melted
and alloy metals are retained in the molten-metal bath. Heat may be applied
by conduction, as in kettle and electric-resistance furnaces. Heat may also
be applied by a combination of convection and radiation, as in reverberatory
and rotary furnaces. In the latter two furnaces, convection heating results
from hot fuel combustion gases being circulated in the charge region, whereas
radiant heating results from the furnace walls being heated by the same hot
gases then radiating heat to the charge. Also, heat may be applied to the
charge mainly by radiation, as in muffle furnaces.
Fluxing and Working the Metallurgical-Process Bath. Flux may be applied to the
charge before melting (where present in residual scrap composing part or
-------
(V-4)
all of the charge), or it may be applied during or ^fter melting. The
process bath may be worked by stirring or agitating to fuse and blend
zinc and alloy metals into a contiguous molten-metal bath and separate
the residue-flux mixture (skimmings) and unmeltables from that bath.
Both fluxing and working of process baths may be done in kettle and rever-
beratory furnaces, in single heats. Working of baths, usually without
fluxing, is done in rotary furnaces. (See Chapter IV.)
MATERIALS AND FUELS APPLIED TO EMITTING PROCESS-EQUIPMENT UNITS
In the following paragraphs, materials and fuels that are applied to sweat
furnaces and that might materially affect emissions are analyzed to deter-
mine their constituents and provide a basis for a hypothesis on how emissions
are generated and of what they are composed.
ORGANIC CONSTITUENTS OF CHARGES
Materials made of organic compounds may be present in charged metallic
scrap. Examples are gaskets, fabrics, electrical insulation, paper, and
lubricants. Although these compounds are mainly of carbon, hydrogen, and
oxygen, other elements could be present:
A. Sulfur in natural rubber and polysulfide polymers.
B. Chlorine in chloroprene and vinyl polymers.
C. Nitrogen in nitrile rubbers, polyamide plastics, and protein binders.
D. Fluorine in fluorocarbon plastics and fluorinated elastomers.
E. Silicon in silicone rubbers and lubricants.
-------
(V-5)
emissions from certain sweat processes. No special problems were reported or
indicated involving the other elements listed (A through E). For that
reason, quantities of these elements in emissions are believed to be small,
and it is therefore assumed here that organic materials contained in furnace
charges are essentially composed entirely of carbon-hydrogen-oxygen compounds.
METALLIC CONSTITUENTS OF CHARGES AND RESULTING
COMPOSITION OF MOLTEN METAL BXTHS
In all of the types of scrap materials that are charged to zinc sweat furnaces
(see Chapter II), metallic zinc is derived mainly from die-casting and gal-
vanizing alloys with these typical compositions:
A. Die-cast alloys - 94% Zn, 4% Al, 1% Cu, and 0.05% Mg; with restrictions
to maximum of 0.1% Fe, 0.007% Pb, 0.005% Cd, and 0.005% Sn.
B. Galvanizing alloys - Restricted to a minimum of 98.3% Zn, and jnaximums
of 0.08% Fe and 1.6% Pb.
Die-cast alloys may be contained in both metallic and residual scrap. Galvanizing
alloys are contained only in residual scrap.
Very small amounts of nickel may be present in residual scrap, probably derived
from alloy cast-iron melting vessels and from platings. (Mathewson, pp. 387-8,
486,315).
The assumption is made here that chromium platings, which may be contained in
charges, are essentially inert and insoluble at process bath temperatures.
It is therefore considered that they do not significantly affect emissions,
although traces of chromium might be emitted.
-------
(V-6)
It is also assumed that unmeltable attachments essentially separate from the
process bath and do not significantly affect emissions. It is noted, however,
that small quantities of metals in these attachments, such as copper, might
dissolve in the molten metal.
Galvanizing alloys contain a higher percentage of lead than die-cast alloys.
This study, however, indicated no special emission problems caused by lead.
Pertaining to the possible emission of lead vapor, the boiling points of
lead and zinc, 3160 and 1665°F respectively, indicate that any emission of
lead vapor during zinc-sweat processes would be very small. Further, as a
relatively unreactive metal, lead contained in process baths probably would
not form compounds that would be emitted in significant quantity.
Findings of this study indicated that scrap materials containing die-casting
alloys are more representative of materials subjected to sweat processing
where charges contain significant amounts of flux than those containing
galvanizing alloys. Galvanizers' skimmings are pretreated to remove residual
flux, producing clean zinciferous metals particles that can be distilled
directly or sweated with little or no flux being applied. Galvanizers
drosses that contain little or no residual flux are either distilled directly
or sweated without applying flux.
These considerations indicate that the composition shown below would be
very representative of molten-metal baths obtained from sweat processing, where
the charge contains significant amounts of flux:
-------
Metai
(V-7)
% by Weight of Metal Bath
Formed During Sweat Processing
Zn 94%
Al 4%
Cu 1%
Fe 0.1%
Mg 0.05%
Pb -s
Sn
Ni
Cd
Cr ^
> Less than 0.01% each
FLUXING COMPOUNDS IN CHARGES
Fluxing compounds are considered in this study as consisting of ZnCl» and/or
NH4C1. The double salt, zinc ammonium chloride (ZnCl2' 2NH4C1) could be
contained in residual scrap materials. This salt is assumed here as equiv-
alent to uncombined ZnCl_ and NH Cl, undergoing the same reactions and
physical changes during zinc-sweat processing.
INORGANIC IMPURITIES IN CHARGES
As applied in this report, the term "inorganic impurities" does not include
uncombined metals. It is assumed that inorganic impurities consist essentially
of oxides of the metals present, particularly ZnO. This assumption is
believed to approximate actual conditions. However, accumulations afr dlirtJ,
residues from phosphate and chromate metal-finishing treatments, fillers
and pigments of rubbers, plastics, and paints (including compounds of Si,
-------
(V-8)
Ti, Sn, Cr, Ca, Cd, and Fe) could also be present.
FURNACE FUELS
Gas or oil of 100% hydrocarbon composition are considered as being used for all
process fuel combustion. Sulfur content is assumed negligible.
EMISSIONS - THEIR GENERATION AND COMPOSITION
The following paragraphs present a theoretical analysis of the effects on
emissions of the process operations, materials, and fuels (enumerated and
described earlier in this chapter) that are applied to zinc-sweat processes.
A hypothesis is developed, based on this analysis, on how emissions are generated
and of what they are composed. The analysis incorporates certain assumptions,
which are- believed to represent actual occurrences and conditions. In a
later chaper the hypothesis developed here is applied in interpreting
data obtained from emission tests and thus is subjected to verification.
CONSTITUENTS THAT MAKE UP EFFLUENT
Generally, any carbonaceous emissions in effluents occur during melting
operations when organic materials in the charge are burned off. Noncar-
bonaceous emissions occur during fluxing and working operations when vapor-
ization, oxidation, and entrainment involving constituents of the process
bath and surrounding gases take place. Emission of ZnO particulates
(noncarbonaceous emissions) may also take place during melting of the
charge due to vaporization and oxidation of elemental zinc taking place in
kettle and reverberatory furnaces (particularly in the latter).during
that phase of the process.
-------
(V-9)
Considering sweat processing in kettle, reverberatory, and rotary furnaces,
organic materials might be contained in charges to all of these. On the
other hand, flux might be contained in charges to kettle and reverberatory
but not normally in charges to rotary furnaces. It then follows that the
possible combinations of constituents making up effluents from these furnaces
may be listed as follows:
Furnace Type
Kettle *
Reverberatory
Constituents of which Effluents may be Formed
Carbonaceous emissions from the charge, during melting.
Noncarbonaceous emissions from (1) the charge
during melting and (2) the metallurgical process
bath. Could include chlorides.
Infiltrated air.
Carbonaceous emissions from the charge, during melting.
Noncarbonaceous emissions from (1) the charge, during
melting and (2) the metallurgical process bath.
Could include chlorides.
Infiltrated air.
Fuel combustion products.
Carbonaceous emissions from the charge, during melting.
Noncarbonaceous emissions from (1) the charge, during
melting and (2) the metallurgical process bath, without
flux. No Chlorides contained in emissions.
Infiltrated air.
Fuel combustion products.
Detailed consideration is given below to the generation and composition of
carbonaceous and noncarbonaceous emissions, and fuel combustion products,
which-along with infiltrated air makeup effluents as shown, in-the above
tabulation. .
Rotary
Fuel combustion products from kettle furnaces are exhausted separately from the
effluent that flows from the process bath, and are therefore not listed here.
(See Effluent definition, page V-2)
-------
(V-10)
All of the considerations below apply tp sweat processing in kettle and yever-
beratory furnaces where chloride flux is included in charged process materials.
The considerations that pertain to noncarbonaceous emissions do not apply
to rotary furnaces, since chlorides, derived from flux, are not normally emitted
from those furnaces. Also, findings of this study indicate that because of
lower bath temperatures and absence of flux in rotary furnaces, noncarbonaceous
emissions from that type of furnace are not significant. Rotary furnace
emissions are therefore considered here as being entirely carbonaceous under
usual process conditions.
GENERATION OF CARBONACEOUS EMISSIONS FROM CHARGE
As material charged to furnaces is heated, metal melts and any organic compounds
present undergo decomposition, oxidation, vaporization, and/or mechanical
entrainment into the exhaust effluent gas stream. Carbonaceous emissions
thus formed are cooled by infiltrated air and radiation of heat from ducts
carrying the effluent. Where emission control equipment is used, cooling may
also occur in that equipment, such as through the cooling action of a scrubbing
fluid. Finally emissions may be cooled when effluent is exhausted to the at-
mosphere. During cooling, carbonaceous vapors condense, and viscosities of
liquids increase. Resulting carbonaceous emissions may therefore be composed
of combinations of carbonaceous fly ash, solid carbon particles, liquid
droplets of carbon compounds having consistencies ranging from oily to tar-
like, and gaseous carbon compounds. It follows, from considerations of
organic constituents of charges, that these emissions would be made up essentially
of inert ash and compounds of carbon, hydrogen, and oxygen.
GENERATION OF NONCARBONACEOUS EMISSIONS FROM CHARGE
As a process bath is fluxed and worked, noncarbonaceous emissions may result
-------
(V-ll)
from chemical reactions and physical occurences involving the following
substances:
A. Fluxes applied to bath or contained in charged residual scrap.
B. Metals in molten-metal bath.
C. Residues that accumulate (fused or mixed with any flux present)
at the top of the process bath above the molten metal, these residues
being introduced as inorganic impurities in the charge or formed by
oxidation of charged metal.
D. Oxygen in infiltrated air. (Atmospheric nitrogen is not considered
as a possible reactant because it is normally unreactive at zinc-sweat
process temperatures.)
Fluxes consist of ZnCl. and/or NH Cl. Metals contained in the molten-metal
bath are considered in the subsequent analysis as being limited to zinc and
aluminum. This simplification is justified by data, shown previously,
indicating that other metals would not be present in sufficient quantities to
significantly affect the nature or amount of emissions. For similar reasons,
metal oxides in residues are considered as consisting of ZnO and Al_0_. It
is immediately apparent that A1C1_ might be produced within the process bath
by reactions of ZnCl_ with aluminum (APEM. p. 287). It follows from these
considerations that the metals and compounds in process baths that may enter
into chemical reactions and physical occurences forming noncarbonaceous
emissions consist essentially of Zn, Al, NH.C1, ZnCl_, A1C1,, ZnO, and A100,.
4 23 ^ J
Physical properties of these materials that pertain to emissions are tabulated
in Table V-l and shown graphically in Figure 5-1.
Vapor pressures plotted in Figure 5-1 show that at the temperature of molten
metal, NH.C1 would vaporize at a high rate and that ZnCl- and metallic zinc
-------
(V-12)
would also vaporize significantly. Data in Table V-l show that ZnO and Al?0,
are refractory at the temperatures of the process bath. Therefore, any ZnO
or Al_0, in residues would not vaporize but might enter the effluent by
mechanical entrainment due to the velocity of the effluent stream. Agitation
and skimming of the process bath would promote vaporization of metal and flux as
well as mechanical entrainment of residue particles into the effluent. Also
in this temperature range, NH Cl decomposes to NH and HC1 gases, which may
*T »5
recombine when cooled, by infiltrated air, forming small particles of NH Cl fume
(APEM, p. 294).
Owing to the presence of the molten metal, vapors and gases formed as noted
above, the following reactions could take place in the immediate region of the
molten metal bath surface:
(1) NH4C1 (flux) + NH3 (gas) + HC1 (gas)
(2) 2 HC1 + Zn (liquid or vapor) •+
ZnCl- (particles - liquid at process bath temp.) + H_
(3) 2 Zn (liquid or vapor) + 0, (atmospheric) •+ 2 ZnO (particles, solid)
(4) 2 H2 + Q (atmospheric) •+ 2H2Q (vapor)
As emitted particles cool (approaching 212°F) because of infiltrated air and
radiation, further reactions could occur:
(5) ZnCl_ (particles, liquid) -+ ZnCl2 (particles, solid)
(6) NH3 (gas) + HC1 (gas) -»- NH4C1 (particles, solid)
(7) ZnCl (particles) + 2 HOH (from Reaction 4, atmosphere and/or fuel com-
bustion products) ~t Zn(OH) + 2 HC1. (Reaction takes place in particles
because of deliquescence of ZnCl .)
(8) HC1 (contained in particles from Reaction 7) + NH, (from Reaction 1) -*•
NH Cl.(Reaction takes place in particles formed in Reaction 7. Resulting
particles contain NH4C1, ZnCl 2> HOH, Zn(OH)2> and HC1.)
(9) Zn (vapor) -»• Zn (particles, solid)
-------
(V-13)
Pertaining to aluminum, further reactions could occur in the immediate
region of bath surface:
(10) 2 Al (dissolved in bath) + 6 HC1 (gas) + 2 A1C13 + 3H2
(11) 4 Al (dissolved in bath) + 3 0 (atmospheric) -»• 2 Al-0 (particles, solid)
(12) 2 Al + 3 ZnCl2 -> 2 A1C13 (gas) + 3 Zn
In view of these reactions, it appears that the small amounts of aluminum
likely to be contained in molten metal baths would not significantly influence
the essential characteristics or quantities of emissions. Therefore, as a
simplification of this analysis, aluminum is not considered further except to
note that small quantities of aluminum compounds might be present in emissions.
Data presented in Chapter VI will show that cooling of emissions within the effluent
stream, resulting from dilution with infiltrated air and radiation from ducts
carrying effluent, reduces temperatures of effluents below the melting
points of metallic zinc and ZnCl- and the sublimation (decomposition) point
of NH4C1.
The generation of noncarbonaceous emissions from the process bath, as they
exist in the effluent stream at the point of either being exhausted from
the stack or entering a gas cleaning device, may then be summarized as follows:
A. ZnCl^ particulate (solid) result from the following occurrences:
L
1. ZnCl« vapor is formed as ZnCl- flux vaporizes and as HC1 (derived from
NH Cl) reacts with elemental zinc vapor.
2. ZnCl- liquid particles are formed as ZnCl_ vapor condenses, during
cooling from the bath temperature range (800 to 1,100°F) to temperatures
just above the melting point of ZnCl2 (689°F).
3. ZnCl solid particles are formed as the ZnCl_ liquid particles solidify
-------
(V-14)
on cooling to temperatures below the ZnCl melting point (689°F).
B. ZnO particulate (solid) result from vaporization of metallic zinc followed
by atmospheric oxidation of ainc vapor.
C. NH Cl particulate (solid) result from the following occurrences:
1. NH_ and HC1 gases are formed as NH Cl decomposes at temperatures above
662°F.
2. NH.C1 solid particles are formed directly as NH_ and HC1 gases
recombine on cooling to 662°F.
D. Particulate containing ZnCl2, Zn(OH),^ HC1, NH Cl, and HOH are formed
as a result of the deliquescent adsorption of water by ZnCl- particles, as
emissions cool to temperatures around 212°F and below. NH gas (formed in
Reaction 1) that does not recombine,with gaseous HC1 is probably adsorbed mostly
into these particulates, forming NH Cl as in Reaction 8.
E. Metallic zinc particulate could be emitted through condensation of zinc
vapor if a sufficiently reducing atmosphere in the furnace is obtained.
F. Particulate containing combinations of ZnCl^, ZnO, and NH.C1 are emitted
as a result of mechanical entrainment of the residue-flux mixture into the effluent.
ZnCl- in these particulates would be subject to the reactions with HOH and NH_
noted previously.
G. NH gas that does not recombine with gaseous HC1 and is not adsorbed into
partially-hyrolyzed particles, as described in £ above, remains in the effluent
gas stream as a gaseous emission.
-------
(V-15)
For further consideration of the possible emission constituents several additional
assumptions are believed justifiable:
1. Particles of ZnCl» and ZnO are formed generally separate from each other
(not as particles containing mixtures of large percentage of both compounds),
as ZnCl vapor condenses to droplets that solidify on further cooling and as
zinc vapor oxidizes directly to solid ZnO particles. This assumption is based
on considerations of (1) the cyclic nature of the processes where flux or
flux-containing residual scrap is manually applied at intervals, ZnCl« vapor-
ization occuring in highest concentration immediately after flux application,
and (2) the difference in vapor pressures of ZnCl2 and elemental zinc
tending to increase the ratio of ZnCl_/ZnO emitted, at lower temperatures,
the converse occuring at higher temperatures. Since ZnO particles would instan-
taneously result from zinc vapor oxidation those particles could act as nuclei
on which ZnGl- vapor could condense. Some mixing might take place in this way.
2. Metallic zinc particles in the effluent are of negligible content'because
an oxidizing atmosphere in the effluent would normally be maintained by infiltrated
air.
3. Gaseous NH, in the effluent is of negligible content because most NH_
o o
would either recombine with gaseous HC1 or be adsorbed into ZnCl- particles as
these particles adsorb moisture and hydrolyze to HC1 and Zn(OH)_.
The foregoing considerations therefore lead to the hypothesis that non-
carbonaceous emissions resulting from melting, fluxing and working zinc-
sweat process baths (at points in effluent streams where equilibrium
mixtures are approached) consist mostly of combinations of particulate formed
-------
(V-16)
as shown in Items A, B, C, D, and F above (E and G omitted). These emissions
may then be summarized as follows:
(1) ZnCl particulate (solid)
(2) ZnO particulate (solid)
(3) NH4C1 particulate (solid)
(4) Particulate containing ZnCl2, Zn(OH) , HC1, NH.C1, and HOH (resulting
from vaporization of molten zinc and flux, followed by oxidation, condensation,
other chemical reactions and physical occurrences).
(5) Particulate containing combinations of ZnCl_, ZnO, and NE.C1 (resultin
from entrainment from residue-flux mixture). Could include products of
hydrolysis and other reactions of ZnCl2
Pertaining to particle size, it follows from the assumption listed as number
one (above) that size of ZnCl? particles would depend on the time particles are
maintained in the liquid state (above the melting point temperature of ZnCl_),
in which particle growth and agglomeration would most readily occur. High
effluent flow rates and large volumes of infiltrated air impose rapid cooling
and would therefore produce small particles. ZnO particles formed from oxidation
of zinc vapor would be small because their formation in the solid phase
would be instantaneous at the temperatures involved, with infinitesimal time
available for particle growth.
-------
(V-17)
FUEL COMBUSTION PRODUCTS
Mixing of air with fuel at burners is assumed satisfactory to effect
complete combustion. It then follows that fuel combustion products
consist essentially of CO^, H_0, and the residual N? that would remain after
atmospheric 0- is consumed in combustion. The content of nitrogen oxides
and the possible presence of sulfur oxides and carbon monoxide in combustion
products are not considered here. Those emissions do not appear significant
to the problem area of this study. In any future work where it is desired
to consider their affect on secondary zinc industrial emissions, results of
studies specifically covering those gases should provide sufficient data.
(Duprey, pp. 6-7)
-------
TABLE V-J - DATA PERTAINING TO MELTING; VAPORIZING, AND CONDENSATION OF
ZINC-SWEAT PROCESS MATERIALS
Process Material
Constituent
Zinc
NH4C1
ZnClg
ZnO
Aluminum
A1C13 (or A12C16)
A12°3
Melting. Point
^C
419**
dec. 350
365
>1800
660
2,000
op.
786
dec. 662
689
>3272
1,220
3,630
BolllnK Point
OG
907
732
2,056
subl. 178
2,210
Op
1,665
1,350
3,734
subl. 352
4,010
Vapor
Pressure at
Pouring
Temperature
mm. Hg.. *
<1 to 15.2
> 760
1 to 74
0
<1
>760
0
* Pouring temperature of zinc sweat-process baths (APEM, p 293) : 800 to 1,100°P
** All data In this table Is from Perry.
-------
12 5265
IO X IO TO THE INCH
KEUFFEL ft ESSER CO.
"ADI IN II 9.A.
-------
Chapter VI
Emission Determinations and Correlation vith Process Variables
Contents Page No.
SUMMARY VI-1
EMISSION FACTORS - PARTICULATE VI-2
EMISSION COMPOSITION - PARTICULATE ?I-5
COMBINED EMISSIONS FROM ONE REVERBERATORY AND SEVERAL
KETTLE FURNACES VI-5
COMBINED EMISSIONS FROM SEVERAL KETTLE FURNACES VI-8
PARTICLE SIZE AND CHARACTERISTICS VI-11
EMISSIONS FROM CHARGES OF MIXED RESIDUAL ZINC-SCRAP
MATERIAL - KETTLE-SWEAT PROCESSED VI-11
EMISSIONS FROM CHARGES OF MIXED METALLIC ZINC-SCRAP
MATERIAL - KETTLE-SWEAT PROCESSED VI-11
ZINC OXIDE PARTICULATE VI-12
ZINC CHLORIDE PARTICULATE VI-14
EFFLUENT TEMPERATURE. FLOW RATES. AND GAS COMPOSITION VI-15
EFFLUENT TEMPERATURE VI-15
EFFLUENT FLOW RATES VI-16
EFFLUENT GAS COMPOSITION VI-17
-------
Chapter VI
LIST OF TABLES
Title Table No.
Summary of Results of Process and Emission Data Evaluation VI-I
Determination of Emission Factors of Secondary Zinc-Sweat Processes VI-2
Visual Comparison of Emissions Resulting from Various Process
Materials - Secondary Zinc-Sweat Processes VI-3
Determination of Exhaust Effluent Flow Volumes as Functions of
Process Weight; Stack Gas Analysis VI-4
LIST OF CALCULATION SHEETS
(included in Appendix B)
Title Sheet No.
Calculation of Amounts of Compounds Composing Particulate Emissions 1-3
-------
Chapter VI
Emission Determinations and Correlation with Process Variables
SUMMARY
Processes of sweating zinc scrap materials are dealt with here. Data on processes
and emissions were obtained and evaluated. Results are summarized in Table VI-1.
The data were from phases of processing that included fluxing and did not include
burning off large amounts of organic materials.
The data indicated that under the applied process conditions, mixtures of ZnCl?
(hydrated) and ZnO make up most of the particulate emissions with carbonaceous
particulates sometimes present in small amounts. Data also indicated that the
gaseous part of effluents from kettle furnaces approaches atmospheric composition
and that gases in reverberitoryfurnace effluents are mixtures of air with combus-
tion products «f fuel (usually natural gas). (Fuel combustion products from
kettle furnaces are exhausted separately from metallurgical process effluents and
are not considered here.)
It should be noted that during initial heating of charged metallic scrap, most of
any organic materials in the charge would burn off, possibly resulting in carbon-
aceous particles making up most of the particulate emissions as well as emissions
of gaseous compounds of carbon and other elements. As indicated above, no data
were obtained on the highly carbonaceous emissions arising from this process
phase, and results reported in Table VI-1 do not reflect such data. However,
-------
(VI-2)
carbonaceous emissions in smaller amounts may occur after initial heating and burning-
off because of many possible variations of process procedures and mixtures of scrap
material charged. Such emissions are reflected in the results reported here. Chemical
analysis of particulate emitted from one heat of residual scrap material indicated
roughly 10% carbon compounds, the balance being mostly ZnCl. and ZnO. This carbon-
aceous content probably resulted from a quantity of metallic scrap containing organic
substances in the charge.
The results reported here are believed reasonably representative of the range
of zinc scrap material normally processed as stated. However, the data show emissions
averaged over certain test periods. Also, these processes are cyclic; there are
variations in operator judgement and in manually performed process operations. For
these reasons, peak loadings might exceed those reported herein. Conversely, there
are inactive periods during heats when emissions are insignificant. Further, the
data sampling was small; there are data gaps; and certain conjectures were applied
in arriving at results. These are detailed in the chapter text. Results should be
useful in theoretical studies of emission control concepts, but more comprehensive
data gathering and/or testing are needed to confirm results and fill data gaps.
EMISSION FACTORS - PARTICULATE
Emission factors were calculated from particulate emission rates of zinc-sweat processes
as shown in Table VI-2. The emissions were from processing several different mixtures
of residual and metallic scrap materials in kettle and reverberatory furnaces.
Visual observations comparing emissions from kettle-sweat processing of several
scrap material mixtures are shown in Table VI-3. (See REFERENCES TO DATA SOURCES.
Chapter I.)
-------
(VI-3)
These emission factors and emission observations Indicate correlations that
may be summarized as higher emissions, and consequently higher emission factors,
resulting from process variables noted as follows:
Ao Metal bath temperature and amount of flux. Higher temperatures increase
the vaporizing rate of both ZnCl? and Zn metal. Larger amounts of chloride flux
also increase the amount of ZnCl. vaporized. These occurrences Increase the amount
of partlculate emissions of ZnCl. and ZnO.
B. Residual scrap in charge. Dross derived from the tops of zinc processing
baths requires relatively high temperatures and large amounts of flux to melt and
separate zinc from Fe-Al compounds and metal oxides. Also, die-cast skimmings
contain considerable residual flux. Thus, the necessary high processing temperatures
and the large amounts of flux entailed in residual scrap processing cause high
emission rates for the same reasons noted in Item A above.
C. Amounts of impurities in charges of metallic scrap. Impurities include
organic materials, non-metallic residues (mainly aetal oxides), and platings.
Organic Materials add directly to emissions where combustion is incomplete.
Inorganic impurities require flux for removal (larger amounts of flux being needed
where amounts of impurities are larger) with resulting Increased chloride emissions.
Emission factors shown in Table VI-2 are higher for reverberatory furnaces than for
kettle furnaces, if both furnaces process scrap materials that appear comparable.
However, conclusions on the relative advantages of the two furnace types in reducing
emissions cannot be drawn from these limited data. A full comparison would be
considerably complex, and only a few comments will be made here. Where scrap charges
contain carbonaceous substances, it appears that reverberatories could be advantage-
ously operated to effect more complete combustion more readily than kettle processing.
Probably the main cause of ZnO emissions from reverberatories is turbulence of hot
-------
(VI-4)
gases contacting and vaporizing zinc during initial melting when there is little
or no flux-residue cover over metal. However, in the same reverberatories, by
applying flux after initial melting, with fuel combustion stopped or reduced,
flux in the charge would not be subjected to as high turbulence of hot gas, and it
appears that emission of chlorides could thereby be kept at minimum levels. Thus,
whereas emission factors might be higher for reverberatories, chlorides (which are
the most destructive constituent of emissions) are probably of lower concentration
in reverberatory emissions as compared with emissions from kettle furnaces processing
comparable charges. Reverberatory furnace design, as it affects heating of the
charge by radiation vs. convection and hot gas velocity in charge region, probably
has a considerable effect on emissions. As used in production, the two furnace
types are each uniquely suited for certain types of work depending upon the type of
charged scrap material.
In summary, the emission factors shown below (obtained from Table VI-2) appear
representative of the types of processes and process materials listed;
Type of Furnace Process Raw Material Emission Factor
Ib./ton Process Material
Kettle-Sweat Metallic Scrap;
General Mixture 10.8
Kettle-Sweat Residual Scrap: Mixed 24.5
zinc die-cast dross &
die-cast skimmings
Reverberatory-Sweat Metallic Scrap;
General Mixture 13.
Reverberatory-Sweat Residual Scrap; Mixed
die-cast dross & skimmings 32
-------
(VI-5)
EMISSION COMPOSITION - PARTICULATE
Farticulate emission samples were collected from two separate zinc-sweat process
runs and chemically analysed. Results are shown In the paragraphs that follow.
COMBINED EMISSIONS FROM ONE REVERBERATORY AND SEVERAL KETTLE FURNACES
Charges consisted of die-cast dross and/or die-cast skimmings. Some zinc metallic
scrap, having carbonaceous content, might also have been included. Fluxing was
by means of residual chloride flux and a small amount of NH.C1. Two samples of
emitted particulates were obtained and analysed. Results of analyses were reported
as follows.
All metal analyses were performed by atomic absorption spectroscopy. The samples
were not completely soluble in any of several solvents tested including acids,
aqua regia, organic solvents, and sodium carbonate fusion. Hot hydrochloric acid
was found to dissolve 90% of the particulate. The insolubles appeared to be a
carbonaceous material. Two working samples were taken from each container,
dissolved in the acid, filtered and diluted to accommodate the linear range or
the instrument. Results were as follows:
Sample #1 Sample 12
NH* 0.43% 0.36%
Cl" 8.93Z 8.32%
Zn 47.50% 44.50%
Al 1.43% 0.54%
Cu 0.04% 0.05%
Fe 0.40% 0.21%
Pb 0.14% 0.16%
(Listing continued on next page.)
-------
(VI-6)
Cd
Mn
Cr
Sample til
0.02%
0.03%
0.01%
Sample #2
0.03%
0.01%
0.004%
Totals
58.97%
54.18%
To determine the moistare content of the samples, specimens were weighed before
and after oven drying, and results were obtained as shown below:
Drying Time & Temperature
4 hrs. at 120°C (248°F)
4 hrs. at 140°C (284°F)
Overnight (around 12 hrs)
at 160°C (320°F)
% Volatile
Sample 1
3.0
3.4
5.2
Sample 2
7.1
7.5
10.2
During collection, the particulate contacted stainless steel parts of the collector,
which underwent some corrosion. Infiltrated air entering exhaust ducts from the
furnaces waar:of sufficient volume to reduce effluent temperature to 108°F.
(Radiation from exhaust ducts probably accounts for a small part of the heat loss.)
The collected particulates had been allowed to stand uncovered at least overnight
before samples were extracted and sealed in containers to await analyses. The
samples had the appearance of light gray, low-density powder.
Because of atmospheric moisture in the large amount of infiltrated air the
following effects probably occurred. The emitted ZnCl. was fully hydrated
to ZnCl '4H-0 (Mathewson p. 661) while particulates were still in the effluent
gas stream. Further moisture adsorption by deliquescence of the hydrated
ZnCl,, occurred during overnight atmospheric exposure of the collected parti-
culate. Any other metal chlorides present, particularly A1C1_ and
-------
(VI-7)
MgCl9, underwent similar moisture adsorption. The effect of moisture adsorption
was probably accentuated by the extreme porosity of the collected mass caused
by the presence of extremely small acicular ZnO particles (APEM pp. 271-73).
Besides those metals included in the analyses listed above, other metals that
could be present in small quantities (derived from alloys processed, melting
vessels, corrosion of stainless steel, and impurities) include Mg, Sn, Ni, Si,
Ca, and Na.
To obtain an approximate, complete analysis shoving compounds that make up particu-
late emissions, the following assumptions that are based on above considerations and
believed to approximate actuality are applied:
A. All M* is combined as NH.C1.
4 4
B. The balance of Cl , not combined as NH.C1, is combined as ZnCl.. (Actually,
Cl probably combines preferentially with Al and Mg. However, quantities of these
metals likely to be contained In process baths are considered small enough to
permit this simplification.)
C. ZnCl- is present as the hydrate ZnCl.-4H 0. (In the tabulations that follow,
amounts of this compound are shown as equivalent anhydrous ZnCl_, with water of hydration
itemized separately.)
D. 10% of the particulate samples are carbonaceous material (Indicated in test
report as noted above).
£. All other compounds composing particulates are metal oxides, aluminum being
present as Al.O., iron as Fe_0_, and lead as PbO.
F. Metals other than Zn, Al, Fe, and Pb in collected particulates total 1X> and
their oxides total 2Z of sample weight.
G. Moisture adsorbed by deliquescence (in excess of water of hydration) equals
the pe«cent volatile determined by drying specimens overnight at 320°F, shown above.
-------
(VI-8)
Compositions of samples are then calculated as shown in Calculation Sheets 1 and 2
and are tabulated as follows:
Component
ZnCl-
ZnO
NH Cl
JV
Pb3 3
HO (in ZnCl -4H 0)
Oxides of Mgf Snf Ni, Si, Ca, Na
Carbonaceous Material
Moisture (deliquescent)
Sample 1 (X)
15.3
50.0
1.4
2.7
.6
.2
8.1
2.0
10.0
5.2
Sample 2 (%)
14.5
46.9
1.1
1.0
.3
.2
7.7
2.0
10.0
10.2
Total 95.5% 93.9%
COMBINED EMISSIONS FROM SEVERAL KETTLE FURNACES
Emissions were from charges of mixed die-cast dross and skimmings. Fluxing was by
means of residual chloride flux. Effluent particulates were sampled in the header,
which connects furnace exhaust ducts to the stack, and analysed. Sampling was by means
of extraction thimbles. Results of analysis were as follows:
17.7% chlorine as chlorides (Cl~)
48.3% free and combined zinc (total Zn)
No additional data were obtained on composition of particulates. The test report
noted that emissions were sampled at times when melt kettles were being skimmed to
obtain peak loadings. Effluent was exhausted by means of natural draft and
consisted principally of air and particulate emissions. Analytical methods were
not detailed in the test report, but it is presumed that standard techniques were
applied with samples sealed in containers immediately after collection to await
chemical analysis.
-------
(VI-9)
To obtain an approximation of amounts of principle compounds in sampled particulates,
it is assumed that (A) all Cl is combined as ZnCl ', (B) all ZnCl- is present as
the hydrate ZnCl.'AH-O; and (C) all Zn other than that in ZnCl is present as ZnO.
Calulations are then made as shown in Calculation Sheet 3. Results obtained are
as follows:
Compound
ZnCl2
ZnO
H20 (in ZnCl2-4H20)
Amount (Z)
33.8
40.0
18.0
Total 91.8Z
The balance (8.2Z) is probably made up of about 42 other metal oxides in quantities
approximating those shown for the previously discussed samples and about 42 water
of deliquescence and/or carbonaceous particles.
Comparing analyses of emissions from reverberatory and kettle furnace combinations
(discussed first above) with those of kettle furnace emissions tends to confirm
that percentages of ZnCl. in reverberatory emissions are likely to be lower
(and ZnO percentages higher) than in kettle furnace emissions (as suggested by
considerations noted in discussion of EMISSION FACTORS above).
The emissions studied here are from processes where amounts of chloride flux,
expressed as percentages of the total charge, approach the maximum for the range
of zinc scrap materials that are normally sweat-processed. It appears therefore
that percentages of chlorides in the resulting emissions would also approach the
maximum. Conversely, percentages of ZnO would approach the minimum. The opposite trend
of emission composition would be anticipated for processing clean, unplated castings
-------
(VI-10)
where little or no chloride flux is contained in the charge. It then follows (consid-
ering data shown above) that the range of composition of particulate emissions
from zinc-sweat processes may be summarized by the following approximations:
Type of Scrap Material
Making Up Charge
Constituents of Particulate Emissions
ZnCl,
ZnO
Clean, uaplated Zn castings
Residual Zn scrap
Approaches 0%
34%
Approaches 100%
402
That is, ZnCl- would range from 0 to around 34% and ZnO from around 40 to 100%
of partlculates emitted from all zinc-sweat processes, the actual amounts depending
upon type of scrap processed and amount of flux applied.
It should be noted, however, that zinc-sweat processing is highly cyclic. The
ranges above are representative of emissions averaged over periods of time that
include fluxing. Therefore, peak loadings may exceed those ranges.
It should also be noted that the above ranges, being applicable to phases of
processing that include fluxing, are based on the supposition that any carbonaceous
participates present would be of relatively small amount. Undoubtedly, where
charges contain large amounts of organic materials, preponderances of carbonaceous
particulate emissions result. It appears that such highly concentrated carbonaceous
emissions would take place mainly during initial heating of charged material and
that flux would be added after most organic materials are burned off.
-------
(VI-11)
PARTICLE SIZE AND CHARACTERISTICS
EMISSIONS FROM CHARGES OF MIXED RESIDUAL ZINC-SCRAP MATERIAL - KETTLE-SWEAT
PROCESSED,
A particulate sample was obtained from the effluent during a process run using a
standard impinger containing alcohol. Fluxing was by means of residual chloride
flux in charged die-cast skimmings. Particles were measured by means not noted
in the test report but presumed to be an optical microscope. Sizes were reported
by percent distribution as
602 - 0 to 10 microns
17% - 11 to 20 microns
23% - Larger than 20 microns
During this test, stack gas was analysed. Results were reported as
CO - 0.01Z by volume
CO. - none found
02 - 19.7% by volume
This analysis indicates little or no carbonaceous substance in the charge during the
test and probably no carbonaceous particles in the sampled particulates. No
chemical analysis of particulates was obtained. However, analysis of particulates
from similar processing (covered under EMISSION COMPOSITION) indicated those
considered here would be mostly a mixture of ZnCl. and ZnO with analysis about
the same (around 17.7% Cl~ and 48.3% Zn; or 33.8% ZnCl , 40Z ZnO, and 18% water
of hydration).
EMISSIONS FROM CHARGES OF MIXED METALLIC ZINC-SCRAP MATERIAL - KETTLE-SWEAT
PROCESSED
Particulate samples were obtained from the effluent during a process run using a
-------
(VI-12)
stainless steel nozzie-thimble holder connected through a stainless steel probe to
a Greenberg-Smlth impinger (500ml). Fluxing was by means of chloride flux added
to the charge. Most particulates were found in thimbles, very little in impingers.
Particles collected in thimbles were measured using a conventional
microscope (oil immersion). Three determinations were made of geometric mean
values (50% less than size). Results were reported as follows:
1.1 microns
1.3
1.6
Many agglomerates were observed in the microscope field that were not included in
the particle size determination.
The report stated that qualitative and semi-quantitative tests of sampled particles
indicated significant amounts of Zn, Mg, and Na, with traces of Al, Cu, and Fe. It
did not mention any indication of carbon content in emissions. It was noted that the
furnace operator was instructed to flux the furnace at least once during the
sampling period, which indicates that most of any carbonaceous material in the
charge was probably burned off before particles were sampled.
As in the previously discussed sample, the particulates considered here are
probably mostly a mixture of ZnCl- and ZnO but with a lower percentage of Cl
and a higher percentage of Zn because of the difference in the type of charge,
relatively less flux usually being contained in metallic scrap charges than in
residual scrap charges (See Tables VI-2 and VI-3).
ZINC OXIDE PARTICULATE
To study the ZnO-particulate constituent of zinc-sweat process emissions, it is
useful to consider emissions from secondary processing of copper-zinc (brass)
-------
(VI-13)
alloys that involve "blowing" of alloy baths with compressed air to remove impurities,
such as iron, by oxidation. Other alloy metals that may be present are lead and
tin (APEM p. 271). Of the alloy metals composing the process bath, zinc has the
lowest boiling point at 1663°F as compared to 3160, 4120, and 4703°F for lead, tin,
and copper, respectively. Fluxes for brass processing are generally inert materials
used in small quantities to cover the bath, prevent oxidation at the bath surface,
and mix withiMnetal oxides that are formed. Resulting emissions contain as high as
96Z ZnO, which results from vaporizing followed by oxidation of Zn. Thus ZnO partic-
ulate is generated during brass blowing through the same mechanism as
in zinc-sweat processing - vaporizing followed by atmospheric oxidation. Other
particulate emitted from brass blowing would Include oxides of lead and tin and
flux particles, mechanically entrained in effluent gas, that would not have
deliquescent or fusing properties similar to ZnCl.. Under these conditions,
characteristics of ZnO particles in effluents are less obscured and interfered
with by other effluent constituents than is the case in zinc-sweat emissions.
Therefore ZnO particulate as emitted from brass blowing may be studied under
more ideal conditions.
Particle sizes of ZnO fumes are reported as 0.03 to 0.3 micron in APEM (pp. 271-3).
The same reference shows electron photomicrographs of fume from a zinc smelter
and a yellow brass furnace. These figures show star shapes of ZnO particles with
thin rod-like formations radiating from the nuclei. A loose agglomeration of these
particles is shown in Figure 185 (APEM).
Particle size measurements of emissions sampled during the blowing operation of a
secondary brass-process heat showed a geometric mean size of 1.1 microns, measured
by optical microscopy. Chemical analysis of particles sampled showed 462 Zn, 8% Pb,
and 52 Sn.
-------
(VI-14)
The above data indicate that as initially formed, ZnO particles are of sub-micron
size and that these particles form mechanically attached agglomerates (not bonded
by liquification of fusion). These agglomerates appear as particles of approximately
one micron in normal optical microscopic observation and measurement.
ZINC CHLORIDE PARTICULATE
Data on effluent streams having particulates with primarily ZnCl_ content were not
obtained during this study. An estimate of ZnCl_ particle size is arrived at below
by applying data presented in previous paragraphs involving partlculate ZnCl» and
ZnO emitted in mixtures.
Data cited previously showed that particulates emitted from a kettle-processed,
chloride-fluxed, zinc-scrap charge (which would have consisted mainly of ZnCl_ and
ZnO particles) were of around one micron, which was about the same size indicated
for ZnO particles emitted from brass processing and zinc smelting. It follows
from this comparison of sizes between mixed-ZnCl_-and-ZnO particles and ZnO particles
(the latter being in approximate isolation) - all being around one micron - that
size of emitted ZnCl? particles must also approximate one micron. This applies to
emissions from kettle-sweat processing. Average participate ZnCl emitted from
reverberatory sweat processing might be somewhat larger because combustion products
in the effluent could cause slower cooling during part of the processing, with
ZnCl™ particles being maintained longer in the liquid phase, allowing particle
growth. However, this effect would tend to be balanced by higher exhaust flow
rates required to remove fuel combustion products from reverberatories.
-------
(VI-15)
EFFLUENT TEMPERATURE. FLOW RATES, AND GAS COMPOSITION
In the kettle sweat processes studied in this investigation, fuel combustion productd
were exhausted separately from process bath emissions. Melting kettles (or kettle-
sweat furnaces) were enclosed in hoods with doors that were opened during a consider-
able part of process time to permit charging, fluxing, sampling, skimming, and
removal of unmeltables and through which air could infiltrate into the effluent.
Reverberatory furnaces had openings for charging, fluxing, skimming, and removal
of unmeltables. Hoods at such openings were used during process operations that
Involved emissions. The effluent was composed of flue gas containing bath emissions,
fuel combustion products, and infiltrated air as well as emissions and air
collected at hoods.
In the following paragraphs, temperatures, flow rates, and gas composition of
process exhaust effluents, resulting from the occurrences mentioned above, are
shown and discussed.
EFFLUENT TEMPERATURE
Temperatures of all effluent streams from zinc-sweat processes that were obtained
during this study are shown with related data as follows:
Furnace
Type
Kettle
Kettle
Kettle
Kettle
Kettle
Draft
Type
Induced
Induced
Induced
Natural
Natural
Process
Scrap-Mat '1.
Type
Residual
Residual
Residual
Metallic
Metallic
Effluent
Flow Rate
SCFM
1,850
395
2,180 ACFM
3,960 *
4,540 *
Effluent
Temp.
°F
176^
118
147
195
105J
;> Avge
148
-------
(VI-16)
Furnace
Type
Reverberatory
(APEM p. 307)
Reverb. -Kettle
Combination
Draft
Type
Induced
Induced
Process
Scrap-Mat "1.
Type
Metallic
Residual
Effluent
Flow Rate
SCFM
7,680
13,400
Effluent
Temp.
°F
190
108
* Concurrent processing in same plant
All of the above temperature measurements were made in stacks or in headers leading
directly to stacks. There had been no treatment of effluent before temperature
measurement, so the only cooling of effluent had been by air infiltrated at the
furnaces and by radiation from exhaust ducts.
EFFLUENT FLOW RATES
Exhaust effluent flow rates of zinc-sweat process runs, involving kettle and
reverberatory furnaces, are tabulated in Table VI-4. From these flow rates,
effluent flow factors (defined here as cubic feet of exhaust effluent flow per ton
of scrap material processed, i.e. process weight) were calculated and are shown
in the table. In all of the tabulated processes, effluents were exhausted by
induced drafts. It is assumed that flow rates induced were those required to
efficiently exhaust emissions that result from the process conditions shown in the
table.
Effluent flow factors provide a means of relating process weight rates with
flow rates required to exhaust resulting effluents. Those shown below, obtained
from Table VI-A, appear representative of the types of processes and materials listed.
However, it should be noted that they were derived from a small sampling of data and
are therefore subject to confirmation through more comprehensive data gathering
and analysis.
-------
(VI-17)
Process Effluent Flow Factor
Furnace Scrap-Mat11. - Cu. Ft. Exh. Effluent
Type Type Per Ton Process Mat'l.
Kettle Residual 222,000"! A
J> Avge.
Kettle Metallic 185,000j 200'00°
Reverberatory Residual . 840,000
Reverberatory Metallic 443,000
These values indicate that higher effluent flow rates, and consequently higher
effluent flow factors, result fromr
-------
(VI-18)
radiation, it seems that the consistently low effluent temperatures would result
mainly from air infiltration.
Indications that free NH_ might be present in effluents were not noted in any test
reports obtained during this investigation. Also, there was no mention of the
presence of NH in discussions with company representatives, and indications of
emitted NH. were not noticed during visits at plants. It appears most likely
that any free NH , generated from reactions involving NH.C1 flux in zinc-sweat
processing, is adsorbed with moisture into ZnCl. and other metal chloride fumes.
Such adsorption appears to oddur in emissions from galvanizing processes where
"fumes escaping contain an excess of ammonia" (APEM p. 402) and where fume analysis
shows small percentages of NH, (APEM p. 405).
Reverberatory Furnaces
Chemical analyses of the gaseous constituents of exhaust effluent from reverberatory
furnaces were not available during this investigation. Analyses of such gas
mixtures were estimated for one reverberatory furnace by considering the known
combustion rate of the natural gas fuel, applying the following assumptions, and
calculating results shown below and in Columns 8-11, Table VI-4.
Assumptions:
A. Air entering the furnace during burner operation is 20% in excess of the
theoretical natural gas combustion requirement.
B. Additional infiltrated air entering the effluent during burner-off time
equals 10% of total effluent.
C. All moisture in the air entering the effluent is adsorbed by particulate metal
chlorides. Air entering the effluent is therefore considered dry.
-------
(VI-19)
D. The volume of emitted participates (including adsorbed moisture) is small
enough to be neglected. Total effluent flow rate is therefore considered equal
to effluent gas flow rate.
B. The charge is free of carbon compounds or other combustible or volatile,
non-metal compounds that would add to gaseous emissions.
Then, during any processing period (designated here as a run), the following
equations are applicable:
A. Effluent gas, during burner-on time • natural fas combustion products
including 20Z excess air. (Let this combustlon-product-and-alr mixture be
designated "C.P.")
B. Effluent gas, during burner-off time - dry air.
C. Effluent gas, for total run « C.P. that occurs during fraction of run
time burners are operated + additional infiltrated air entering effluent during
burner-off time. (Let fraction of run time burners are operated be designated
"F".)
D. Total effluent gas flow rate » (flow rate of C.F., during burner-on time)
times (F) + (.10) times (total effluent gas flow rate).
("Total effluent gas flow rate" expresses an average value for the total run
including burner-on time and burner-off time.)
Where:
Total effluent gas flow rate - total effluent flow rate - 7000 cu. ft/min.
(Col. 4, Table VI-4)
Flow rate of C.P. during burner-on time - 13.5x1,600 - 21,600 cu. ft/min.
(Col. 8, Table VI-4; and APEM p. 882)
-------
(VI-20)
Then, substituting in equation I):
7000 « 21,600xF - .10x7000
F - 7000 - .10x7000 = .29 or 29%
21,600
That is, 29% of the total effluent gas has the composition of C.P.
Determination may then be made of effluent gas composition during burner-on time
and of percentages of C.P. (constituent) gases in the total effluent, as follows:
C.P. Gases Vol. of C.P. Gases Z C.P. Gases in % C.P. Gases
(See APEM from 1 cu. ft. Effluent During in Total
p. 882) Natural Gas Burner-on time Effluent
CO, 1.13 Cu. Ft. 1.13 - 8.4Z 8.4x.29 - 2.4%
13.52
HO 2.08 2.08 - 15.4% 15.4x.29 - 4.5%
13.52
N 9.87 9.87 • 72.9Z 72.9x.29 - 21.2%
13.52
0 0.44 0.44 - 3.3% 3.3x.29- .9%
13.52
Total 13.52 Cu. Ft. 100%
The remaining 71% of the total effluent has the composition of dry air.
Effluent gas during burner-off time has the composition of dry air, which is
listed below. Also listed are percentages of dry air (constituent) gases in
total effluents, calculated as shown:
% Dry- Air Gases
Dry-Air in Effluent During % Dry-Air Gases
Gases Burner-off Time in Total Effluent
N2 78% 78x.71 - 55.4%
02 21% 21x.71 - 14.9%
Other 1% lx.71 - .7%
Total 100%
-------
(VI-21)
Combining the above, an estimated analysis of total effluent gas (stack gas)
from a reverberatory-sweat process run is obtained as follows:
CO. * "2° Z N Z 0 Z Other
21.2
55.4
.9
14.9
.7
2.4 4.5
2.4 4.5 76.6 15.8 .7
Sum of above equals 100Z
It should be noted that in arriving at the above approximation, combustion charac-
teristics of natural gas shown in Table D7, p. 882, of the APEM are applied. The
values in Table D7 do not include nitrogen oxides and the probably-insignificant
amounts of CO and sulfur oxides contained in natural gas combustion products (Duprey,
p. 6). This simplification is believed justified in view of the problem area of
this study. A more rigorous treatment would be needed in evaluating total secondary
zinc industrial emissions.
-------
Table VI-1 - Summary of Results of Process and Emission Data Evaluation
Emission Properties
and Related Process
Variables
Emission Rates - Perticulati
-Ib/hr.
Emission Factors - Parti-
culate, Ib/ton Process Mat'l
% Chlorine (Cl~) in charge,
derived from flux.
Composition of 1 %ZnCl^
Emitted Parti- \J,ZnO '
culates
Particle Size of/ZnCl,,
lates
Effluent (Exhaust Gas)
Temperature - F
Effluent Flov Rates
- Approximate SCFM
Effluent Flow Factors
- cu. ft. Exhaust Effluent ]
Ton of Process Material.
Effluent Gas Composition:
Burners on )
\
Burners Off /
I
Total Run <
(Average) [
Type of Process Furnace and Zinc-Scrap Material Charged
Melt Kettle (Sweat Process)
Metallic, Clean
Approaches 0
Approaches 0
Approaches 0
Approaches 0
Approaches 100%
Metallic, General | Residual
it to kit for various furnace
combinations
10.8 24.5
to 0.15% to 3.0%
0 to 34% 34%
100 to 40% 40%
Reverberatory (Sweat Process)
Metallic, Clean
Approaches 0
Approaches 0
Approaches 0
Approaches 0-
Approaches 100%
1C
Metallic, General
13.5 for 1
furnace
13
No data
Residual
20 for 1
furnace
32
to 3.0%
15% or less
50% or more
No data - Probably around 1 micron and less
• 0.03 to 0.3-micron particles; 1-micron agglo
105 to 195 F for various furnace combinations.
2,000 to 16,000 for various furnace combinations
No data
er
Around 200,000
(For charged zinc scrap materials that contain no or
Approximates Atmospheric Air
Around 78% NZ
Around 21% 02
Air (Same as above)
Ait (Sane as above)
No data
No data
No data
190 °F for
1 furnace
7,680 for
1 furnace
No data
7,000 for
1 furnace
mic substances.)
No data
No data - would approximate Air
No data
8.4% CO
15.4% HO
72.9% N^
3.3% 0
78.0% N'
21.0% 0,
2.4% CO,
4.5% HO
76.0% N,
15.8% 0^
**"Residual" zinc-scrap materials consist of drosses and skimmings.
* "General" metallic-zinc scrap materials are meant to include mixtures of all types of metallic scrap
(without excessive amounts of impurities), the resulting emission characteristics approximating the
average for metallic scrap.
-------
Table VI-2 - Determination of Emission Factors of Secondary Zinc-Sweat Processes
123456 789 10 11 12
Type of Furnace
Used During Test
Melting Kettle (6)
Melting Kettle (1)
Melting Kettle (2)
Reverberatory (1)
Melting Kettle (7)
Reverberatory (1)
5 ft. 9 in. Wx 6 ft.
(Reference APEM p. 307)
Particulate emission
factor for "Zinc sweating
furnace" shown in Table 24,
Compilation of Air
Pollution Emission Factors
(PUS No. 999-AP-42) 1968.
Type of
Scrap
Material
Processed
Die- cast
dross and
skimmings
mixtures
same
same
Die- cast
dross
Unplated
and plated
castings
and
fabricatin
scrap.
Zinc
castings
Type of
Flux
Applied
Residual
flux in
skimmings .
same
same
NH^Cl
(Residual
flux might
also be
present)
ZnCl2
Unknown
(Probably
ZnCl2 or
NH4C1)
Process Weight Rate
Ib scrap/hi
per furnace
500
(k ton)
800
(.4 ton)
500
C-t ton)
1000
(*5 ton)
1,290
(.65 ton)
2,080
(1.04 ton)
Ib flux/hr
per furnace
(Approxi-
mations)
10*
16*
10*
L-lb NH^Cl/hi
1- unknown
residual
flux
0.66
Unknown
(Probably
small
amount)
Particulate Emission
Rate
Ib/hr per
all
furnaces
63.7
47.6
4.06
7.23
18.3
3.78
16**
,'tr' ' r-r
13.5
Ib/hr per
furnace
Column 5 7
Ho. of
furnaces)
10.6
7.94
4.06
7.23
9.2
1.89
16
13.5
Emission
Factor
Ib parti-
culate per
ton process
material
(Col. 6 ^
Col. 3)
42
32
10
18
37
Z,
,,24.5
Ave .
32
- x -
(Assv.-n'id.
See rfte
-•- '-'st
•:<.-i '. .•• •:
13
Sampling
Method
Extraction
thimbles
Std.
Impinger
con tainin
alcohol
Greenberg-
Smith
impinger
ASTM
PT. C21
Unknown •
Unknown Unknown • • 10.8
Tne type jf process furnace tnc; scrap is.terli.1 prucesseu ure not snown In tne reference.
The reference ealuslcn factor (10. &) Inolcates B lew to medium emission level, by
compt-rlEon with the other emission fiictors shown auovu. Tbole VI-3 Inalcutes that such
emission levels are chartctfcribt Ic of JtetUe-&*cct pr-ocesslne of Oener6l mixtures of
j^ttlilc scrap (nut cjiUclniiit, lar^e o-aov;nts _.f 1 apurl ties) . It Is tnerefore oellevec:
tnat Use fuucur (lO.e) can ue rtai l3llct.i ly af^ileo -o Uie au~ve mel t lnt-K6ttie
prucesr. ln& uf -''Ui'.plt'-eu MIC pletec ^aiitlnt3 ".";'• rocrli;ut;nb scrap."
Bath
Temperature
°F
900
900
1000
900
Unknown
Notes
* Fluxing is by means of resi-
dual flux. These figures were
calculated assuming ^ charge is
of die-cas c skimmings , containing
4% ZnCl2> so chat total charge
would contain 2% by weight of
ZnCl?. (Reference: Mathewson
P- 315)
** Effluent of this reverberatory
was combined with that of several
kettle furnaces, processing mixed
die- cast dross and skimmings ,
and the combined effluent stream
sampled. Results shown here were
calculated by applying emission
factor 24.5 to determine rate of
particulate emission from kettles,
then subtracting that rate from
the total particulate emission
rate (determined by sampling. )
-------
Table VI-3 - Visual Comparison of Emissions Resulting From Various Process Materials - Secondary Zinc-Sweat Processes
Type of Furnace
Melting Kettle (1)
Melting Kettle (1)
Melting Kettle (1)
Melting Kettle (1)
Type of Scrap Material
Processed
Clean zinc scrap -
unplated castings,
off-grade ingots.
"New" and/or "old
zinc". Mainly
unplated zinc castings.
Relatively clean, with.
small amount of
organic material.
Mixtures of elated
Zinc castings and
contaminated zinc
die-cast scrap.
Zinc die-cast dross
Type of
?lux Appliec
ZnCl2
ZnCl2
ZnCl2
ZnCl2
Process Weight Rate
Lb scrap/hr
>er furnace
2000
2100
1400
1700
Ib flux/hr
per furnace
(Estimations
.17
.2 to 1
1.7 to 2.5
5
(0.3% of
charge wt.)
(0.15% C1-)
Observations
of Emissions
Practically no
emissions
Low level of emissions.
Includes small
amount of fly ash.
Medium level of
emissions
Highest level of
emissions of the
materials shown here
Bath
Temperature
(Approxi-
mate)
°F
850
850
850
1000
-------
Table VI - it (page
of 2) - Determination of Exhaust Effluent Flow Volumes as Functions
of Process Weight***; and Stack Gas Analysis
Type of Furnace
Used During Test
Melting Kettle (2)
Melting Kettle (8)
Reverberatory (1)
Reverberatory (1)
5 ft. 9 in. W x 6 ft.
4 in. L x 4 ft. H
(Reference APEM p. 307)
1
Type of
Scrap
Material
Processed
Die-cast
dross and
skimmings
mixtures.
Unplated
and plated
castings &
fabricating
scrap.
Die-cast
dross
Zinc
cas tings
2
Process
Wt. Rate
Ib scrap/hr
per furnace
500
(k ton)
.65 ton
1.000
(h ton)
2,080
(1.04 ton)
3
4
Exhaust Effluent*
Flow Rate
cu . f t , /min .
from all
furnaces
2,220 @
176°F
1,850 @
70°F
16,000
@ 70°F(std
temperature
assumed)
7,000 **
@ 70°F
7,680
@ 70°F
cu. f t . /min.
per furnace
(Col. 3 - No
5
Exhaust Effluent
Flow (cu. ft.) per
Ton of Process Mat'l.
(Col. 4 - Col. 2)
.
of furnaces)
925
2,000
7,000
7,680
(Effluent Flow Factor)
222,000 -|
> 200,000
AVRC.
185,000 J
840,000
443,000
6
Exhaust
Effluent
Temp.
Op
176
unknown
108
[temp, of
combined
stream from
reverb. &
kettles)
190
* See definition of effluent (exhaust effluent) in Chapter V.
A* Effluent from this reverberatory was combined with that of several kettle furnaces,
processing mixed die-cast dross and skimmings; and flow rate of the combined (total)
effluent stream was measured by pitot traverse. Results shown here were calculated
by applying the effluent-flow factor 200,000 to determine effluent flow rate
from the kettles, then subtracting that rate from the total measured rate to obtain
the rate shown for the reverberatory.
*** This function is defined as effluent flow factor as illustrated in Column 5.
-------
Table VI - 4 (page 2 of 2)
Type of Furnace
Used During Test
Melting Kettle (2)
Melting Kettle (8)
Reverberatory (1)
Reverberatory (1)
5 ft. 9 in. W x 6 ft.
4 in. L x 4 ft. H
(Reference APEM p. 307)
7
Stack Gas
Analysis
(Chemical
analysis o
gas sample
Volume
CO .01 •
CO - none
02 - 19.6
8
Natural Gas
Combustion
Rate During
Burner
Operation
cu. ft. /min.
1,600
9
Produc ts -of -Combust ion
Flow Rate During
Burner Operation
(20% excess air
assumed)
cu.f t ./rain.
(13.5 x Col. 8)
21,600
10
Fraction of Pun Time
"urners arc Operated
(107, Infiltrated
air assumed)
(Col. 4 - .10 x Col. 4)
'Col 9
0.29
(29%)
11
Stack Gas Analysis
(Calculated, based on
assumptions of
Columns 9 and 10)
% by Volume
(Avge. for total run)
C02 - 2.4
H20 -4.5
N2 -76.6
0 -15.8
Other - .7
12
Stack Gas Analysis
(Calculated - See
text Chapter 6)
% by Vol.
(During burner-on time)
C02 - 8.4
H20 - 15.4
N2 - 72.9
02- 3.3
13
Stack Gas Analysis
(Calculated - See
text Chapter 6)
Z by Vol.
(During burner-off time]
N2 - 78
0 - 21
Other - 1
-------
Chapter VII
Process Modelling
Contents Page No.
DEFINITION OF PROCESS MODEL UNITS VII-1
COMPOSITE EMISSIONS VII-1
PROCESS CONDITIONS NOT INCLUDED IN MODELS VII-2
MELTING CLEAN ZINC CASTINGS VII-2
ROTARY SWEAT PROCESSING OF CONTAMINATED DIE-CAST SCRAP VII-2
OCCURRENCE OF A REDUCING ATMOSPHERE IN REVERBERATORY FURNACES VII-2
MAKING UP PLANT PROCESS MODELS FROM PROCESS MODEL UNITS VII-3
PROPOSED USE OF MODELS VII~4
Table
Title Table No.
ZINC SWEAT PROCESS MODEL UNITS VII-1
(pages 1,2)
-------
Chapter VII
Process Modelling
DEFINITION OF PROCESS MODEL UNITS
Process model units, defined by characteristics of secondary zinc production
processes and resulting emissions, are formulated here. The purpose of these
units is to provide a basis for studies of emission control concepts.
Five model zinc-sweat process units were formulated and are shown in
Table VII-1. Each of these units represents an application of one furnace
and is defined, basically, by the type of furnace, types of materials
composing the charge, and process weight rate. (For brevity in the following
discussion, the model units will be referred to simply as models.)
The models are defined in further detail by values, shown in Table VII-1,
representing further process and emission data. These values, as filled in
for each model, were derived through consideration of data from several
actual process runs (covered in previous chapters) where process conditions
appeared reasonably comparable, and through theoretical considerations. These
values are believed to reasonably approximate data that would be obtained
from processes conducted under the stated conditions. Key considerations in
deriving the tabulated values for the models are fehown in Appendix A.
COMPOSITE EMISSIONS
It should be noted that during a process heat, carbonaceous emission may
partly or entirely precede non-carbonaceous emissions (See Chapter V).
Therefore, the values shown in Table VII-1 pertaining to emissions should
be regarded as applying to composite emissions taking place over periods
-------
(VII - 2)
of time involving melting, fluxing, and working process baths when both
types of emissions night occur.
PROCESS CONDITIONS NOT INCLUDED IN MODELS
Several process conditions that were.not considered in formulating the
models are listed below with reasons for emission.
MELTING CLEAN ZINC CASTINGS
Clean zinc scrap may be melted in all types of zinc sweating furnaces
without generating significant emissions. Emission control equipment is
not required under this condition. Therefore processing clean scrap is
not included in process modelling.
ROTARY SWEAT PROCESSING OF CONTAMINATED DIE CAST SCRAP
Findings of this study indicated that charges to rotary-sweat furnaces usually
do not contain chloride flux and that emissions from those furnaces are
usually satisfactorily controlled by afterburners. Therefore, there was
no apparent need for a rotary furnace process model to study chloride
emissions. Also, no data were available during this study on rotary furnace
emissions. Findings were based on experience of plant operators.
OCCURRENCE OF A REDUCING ATMOSPHERE IN REVERBERATORY FURNACES
If the atmosphere in the sweating chamber of a reverberatory furnace were
not sufficiently oxidizing because of insufficient infiltrated air, com-
bustible effluent mixtures including carbonaceous particulates, gases such
-------
(VII - 3)
as CO, and metallic zinc particles could result. This might occur if
the furnace were closed during melting of the charge; however, no indications
of this occurrence were found during this study. Also, because of process
operations conducted on furnace charges, there would nearly always be sufficient
infiltrated air to prevent a reducing atmosphere. (The low effluent
temperature shown for reverberatory furnaces on page VI-16 is an indication
of such high dilution of reverberatory furnace effluents with infiltrated
air.:) Since the process condition cited here seems unlikely, it was not
considered in process modelling. This condition is noted because it poses
the possibility of a fire hazard, which, though remote, should not be over-
looked.
MAKING UP PLANT PROCESS MODELS FROM PROCESS MODEL UNITS
To make up hypothetical plant process models for the purpose of studying
r
emissions and emission control concepts, the model units defined in Table
VII-1 may be combined. For example, the following combinations might be
made for such study:
12 Model A units
12 Model B units
8 Model C units
4 Model D units
4 Model E units
6 Model A units and 2 Model D units
-------
(yii-4)
Furnace capacity, indicated by process weight rates of the models, range
from 0.5 to 0.65-ton per hour for the kettle type and 0.5 to 1.0-ton per
hour for the reverberatory type. Kettle furnaces having capacities as
high as 1.0-ton per hour are also in use as shown in Table VI-3.
PROPOSED USE OF MODELS
The models developed here are believed to approximate representative industrial,
secondary zinc, sweating processes that emit deliquescent, corrosive chlorides.
The approximations are believed close enough to bring out the general make-up
and physical characteristics of those emissions., Therefore, the models should
be suitable for preliminary studies of emission control concepts. For advance
work involving prototype and demonstration devices, data directly applicable
to specific existing plants should be obtained.
-------
Table VII - 1 (page 1 of 2) - Zinc Sweat Process Model Units
Model
Unit
Ident.
Letter
A
B
C
D
E
Furnace Type
Kettle Furnace
Kettle Furnace
Kettle Furnace
Reverberatory
Furnace
Reverberatory
Furnace
1
Emitting
Process-
Equipment
Unit
Melting
kettle
Melting
kettle
Melting
kettle
Sweating
chamber
Sweating
chamber
2
Type of
Scrap
Material
Processed
Residual
scrap , mixed
(no organic
material)
Residual
and metallic
scrap ,
including
organic
ma t ' 1 . i mixed
Metallic
scrap , mixed
-containing
organic
ma t ' 1 .
Residual
and metallic
scrap , incl .
organic
mat '1. , mixed
Metallic
scrap , mixed
-containing
organic
ma t ' 1 .
3
Type of
Flux
Applied
Residual
ZnCl.
Residual
ZnCl.
Residual
ZnCl2
ZnCl.,
4 5
Process Weieht Rate
ton scrap/hr
per furnace
1/4
1/4
0.65
1/2
1
Ib flux/hr
per furnace
10
8
0.66
16
1
6
Emission
Top of
mel ting
kettle
Top of
melting
kettle
Top of
molting
kettle
Flue and
hooded
furnace
openings ,
comb ined .
Flue and
hooded
furnace
openings ,
combined .
7
Emission
Ib. parti-
culate per
ton Process
Mat '1.
24.5
24.5
10.8
-32
13
8
Particulate
Emission
Rate
Ih/hr per
furnace
(Col. 4 X
Col. 7)
6.1
6.1
7.0
16
13
9 10 11 12
Particulate Comoosition - %
ZnCl,
34
28
5
15
4
ZnO
40
40
77
65
89
H n
— 2—
ZnCl;
4H.O'
" I.
18
15
3
8
2
H n
-2-
Deli-
que-
sccnce
4
3
1
6
1
Other
Metal
Chlorides
Oxides
4
4
4
6
4
Particulates
0
10
10
0
0
-------
Table VII-1 (page 2 of 2) - Zinc Sweat Process Model Units
Model
Unit
Ident .
Letter
A
B
C
D
E
Particle S
Particle Size
ZnCln and ZnO
formed from vapor
1 micron - may include
agglomerates up (.0
10 microns. Makes
up 60% of particulate
emissions .
Same except particles
in this size range
make up 55% of
particulates .
Same as B.
Same as A.
Same as A.
ize and
Distribution
ZnCl, and ZnO
Larger than 10
microns. Makes
up 40% of
particulates .
Same except
particles in this
size range make up
35% of particulates.
Same as B.
Same as A.
Same as A.
3 4
Carbonaceous
Particulates
None
Particles range
frora micron-size
droplets to visibl
fly ash. (10%
of particulate
emissions . )
Same as B.
Hone
None
5
Exhaust-
Effluent
Gas
Composition
78% N
21% 0,
(dry air)
78% N
21% 0,
.1% CO
.1% OT
.1% H^O
Same as B.
73-78% N
3-21?: 0,
0- 8% CO
0-15% H20
73-78% N
3-21% 0
0-8% CO,
0-15% H 0
6
Effluent
Flow Factor
Cu.ft.Effl
per ton
Process
200,000
Same
Same
800,000
400,000
7
Effluent
Flow Rate
SCFM
(Col. 4, p. 1
X Col. 6,
830
Same
2,200
6,700
6,700
8
Process
Bath
TemE-
F
900
Same
900
1000
1000
9
Exhaust
Effluent
Temp^
°F
150
150
150
190
190
[
-------
Chapter VIII
Emission Control Systems - Past, Present and Conceptual
Contents
Page No.
EXHAUST SYSTEMS
PAST ATTEMPTS TO CONTROL SWEAT-PROCESS EMISSIONS
SCRUBBER, JET IMPACT (PLAIN WATER)
SCRUBBER, VENTURI (PLAIN WATER)
SCRUBBER, PACKED TOWER ( PLAIN WATER)
BAGHOUSES
EMISSION CONTROL SYSTEMS IN CURRENT USE
SCRUBBER, IMPINGEMENT PLATE (PLAIN WATER)
SCRUBBER, VENTURI (PLAIN WATER)
BAGHOUSES
ELECTROSTATIC PRECIPITATORS
AFTERBURNERS
VIII-1
VIII-1
VIII-2
VIII-2
VIII-3
VIII-3
VIII-4
VIII-4
VIII-7
VIII-8
VIII-10
VIII-12
CONCEPTUAL EMISSION CONTROL SYSTEMS
VIII-12
-------
Chapter VIII
Emission Control Systems - Past, Present, and Conceptual
EXHAUST SYSTEMS
Exhaust effluents from kettle sweat-process furnaces are collected by means
of hoods that cover and enclose melting kettles. Openings in these hoods
are provided to allow process operations such as charging, fluxing, and
skimming. When they are not in use these openings are covered to minimize
air infiltrationi• Usually a plant incorporates several such furnaces that
are operated at the same time. Typically, ducts connected to the tops of
furnace hoods carry effluent to a header through which the combined effluents
are carried to emission control (gas cleaning) equipment. Dampers in the
duct system allow furnaces not being used to be closed off and also allow
effluent to be carried directly to the stack when clean scrap is being
processed or during an emergency shutdown of the gas cleaning equipment.
Drafts are induced by exhaust fans of the emission control equipment and/
or by the natural draft characteristics of the stakk.
Effluents from the reverberatory sweat-process furnace are collected by meand
of the furnace flue and hoods covering other openings in the furnace. These
effluents, which may be combined with those of other reverberatory or kettle
furnaces, are carried to gas dealing equipment or a stack through duct sys-
tems similar to that described above.
Effluents from rotary sweat-process furnaces are exhausted through the furnace
flues of each furnace directly to an afterburner, thence to a stack.
PAST ATTEMPTS TO CONTROL SWEAT-PROCESS EMISSIONS
Gas-cleaning devices used in past attempt* to control emissions from zinc-sweat
-------
(VIII - 2)
processes are described below, with reasons for discontinuing their use.
Except as otherwise noted, all of this equipment had been applied to pro-
cesses approximating combinations of Models A and B (Chapter VII), residual
scrap being contained in furnace charges, resulting in particulate emissions
of high chloride content.
SCRUBBER, JET IMPACT (PLAIN WATER)
This (pilot model) scrubber had a rated capacity of 500 CFM. It operated with
a pressure drop of 27.5 in.w.g. at gas flow rates around 430 SCFM. Collection
efficiencies were satisfactory, exceeding 98 percent. Consideration of the
following attributes of the equipment, hov/ever, led to discontinuance of its use:
A. Unit could not withstand corrosiveness of fumes. (Fabricating material
of the unit is unknown but is presumed to be mild steel.)
B. High electric power requirement.
C. Bigh maintenance requirement.
D. High water consumption.
E. Problem of polluted waste water disposal.
F. No performance guarantee.
SCRUBBER, VF.NTURI (PLAIN WATER)
This scrubber had a rated capacity of 2200 CFM. It operated at pressure drops
between 46 and 58 in. w.g. at gas flow rates of approximately 2000 SCFM,
obtained by means of a fan with a 50-HP motor. Collection efficiencies
between 92 and 99% were attained. This system was discontinued because cost of
the electric power was considered prohibitive for the limited gas-handling capacity.
-------
(VIII - 3)
SCRUBBER, PACKED TOWER (PLAII WATER)
The maximum collection efficiency of this equipment was between 6.0 and 65%;
at this efficiency the local code could not be met.
BAGHOUSES
Bag filters made of modacrylic, acrylic, cotton, and silk were applied in efforts
to control emissions. Use of all of these was discontinued because Vabrid
endurance was limited to a maxiumum of ft month. The bags failed as fabrics corroded
(became pasty and disintegrated). Although process and emission data were not
obtained for these emission control attempts, it is believed that the processes
approximated Models A and B, kettle furnaces being used with resulting emissions
of high chloride and low carbonaceous content.
Orion (acrylic) fabric bag filters have been used to collect paniculate emissions
from sweat processing of metallic scrap materials in reverberatory furnaces as
shown in Table 87 of APEM. Data were not provided on flux applied, organic
material in the charge, and any resulting chloride and carbonaceous content in
the emissions. 'Hie considerable amount of particulate emissions (shown by data
in Table 87) indicated that substantial quantities of impurities were being
removed from metal being sweated and that flux would have been applied to help
effect that removal. Also, organic material was probably contained tn the
charge because some coating or content of organic material is characteristic
of scrap items other than clean castings. Since bags were not Minded by the
deliquescent effects of chloride or by carbonaceous particulate it seems
likely that chloride content was low and that incineration of organic material
was nearly complete. It thus appears that process and emission conditions
of this baghouse application approximated Model E.
-------
(VIII - 4)
Endurance of the Orion bags in the above application was not reported.
However, the test results indicate that a baghouse system can be used
satisfactorily to collect emissions from zinc-sweat processing in rever-
beratory furnaces where particulate chloride concentration is low enough
and effluent temperature is high enough to prevent harmful deliquescent ef-
fects.
Orion bag filters have been used for collecting emissions from secondary
brass processes, where particulates were mostly ZnO. Apparently the bags
were satisfactory for this usage. (AE, Nov. '67)
During this study, no direct information was obtained on modacrylic or acrylic
bag filters in current usage with secondary zinc or brass processes. It
appears that polyester fabric is preferred in current practice.
Glass cloth bag filters have been used to control emissions from reverberatory
sweat processing of zinc scrap, as reported in APEM, p. 299. Their use was
discontinued because glass fibers failed to withstand abrasion of threads
during bag cleaning.
filters cannot be used where carbonaceous emissions are at a level that
would pose a fire hazard and where droplets of liquid carbon compounds would
blind the bags.
EMISSION CONTROL SYSTEMS IN CURRENT USE
SCRUBBER, IMPINGEMENT PLATE (PLAIN WATER)
A scrubber of this type is being used to clean exhaust effluents from a plant
system that consists of several units approximating Model C. Materials pro-
-------
(VIII - 5)
cessed consist of metallic scrap mixtures including unplated and plated castings
and fabricating scrap. Content of organic materials and other impurities is mod-
erate orcrelatively low. Amounts of applied ZnC^ flux and resulting chloride
emissions are of an intermediate or relatively low level, as Model C indicates.
This three-stage scrubber is similar to that described in Control Techniques,
para. 4141214. Nozzles, plates and baffles are stainless steel. The scrubber
housing is coated internally with polyvinylchloride. Pipes that circulate
scrubbing fluid are of mild steel. Gas is exhausted from the scrubber by a
fan requiring a 45-1/2 HP motor. A 2-HP motor is used for pumping scrubbing
water.
The scrubber is operated at all times that «etal is being melted, fluxed, and
worked. Therefore, carbonaceous emissions that occur during initial heating
of charged material are collected, as well as ZnClj and ZnO emissions occuring
during subsequent fluxing and working of the bath.
The scrubber operates with a pressure drop of 11 in. w.g. at a pas flow rate
of 16,000 CFM.
Collection efficiency of this equipment has not been determined by particulate
sampling tests. Observations of the plume, indicate that 90 to 95% of part-
iculate is collected 90% of the time. Observations of the plume under process
conditions known to be conducive1 to emission of ZnCl. and ZnO, individually,
indicate that the scrubber collects ZnCK very efficiently but that it is not
very efficient in collecting ZnO. The 10% of the time when collection efficiency
is apparently less than 90% is probably accounted for by excessive ZnO emis-
sions occuring at high molten metal bath temperature.
The stainless steel and PVK-coated parts of the scrubber have satisfactorily
withstood corrosion by the scrubbing fluid, although infrequent failures of
the nozzles apparently have been caused by stresses involving welds; and there
-------
(VIIT - 6)
has been minor corrosion of the mild steel pipes. These results tend to
confirm that emissions from the processes involved are of low chloride
content (probably around 5%, as assumed for Model C, or less).
The low resistance of stainless steels and of iron to corrosion by ZnCl
solutions (Perry 4th ed., pp. 23-29, 30) indicates that those materials
would not withstand that corrosion except at low ZnCl_ concentrations.
Because of low corrosion resistance, therefore, the scrubber could be
used satisfactorily to control emissions only from the processing of
relatively clean scrap, that does not entail considerable flux application.
Use of this scrubber is further restricted by emission of ZnO in excess
of the capacity of the scrubber for satisfactory removal, which occurs at
high molten mefcal temperatures. Thus, emissions from processing residdal
scrap cannot be satisfactorily controlled because high temperatures are re-
quired to sweat-process that type of scrap. Also, in sweating some metallic
scrap, particularly with considerable unmeltable attachments, temperature
distribution within the bath is uneven, with high temperature regions emitting
ZnO excessively. Using instruments to obtain meaningful bath temperatures
under this condition is generally precluded, so that maintaining satisfactory
temperatures and mixing the bath depends on visual and manual techniques of the
operator. This dependency further restricts the range of metallic scrap that
can be processed since manual techniques are of limited accuracy in maintaining
bath temperatures at levels where ZnO emissions are controllable.
Passage of emitted carbonaceous particles through the scrubber has been evidenced
by deposits of small quantities of tar-like substances (probably mixtures of
hydrocarbons and ZnO) on the scrubber exhaust fan and fan closure. These deposits
resulted in imbalance of a previously used fan, which caused destructive vibration.
This problem was overcome by replacing that fan with a "standard duty wheel".
-------
(VIII-7)
The tar-like deposits have not occured during several .months since this
replacement. These carbonaceous particulate emissions, controlled at low
levels by selection of relatively clean scrap, probably do not pose an air
pollution problem. However, the limited capacity of the scrubber to control
carbonaceous particulate emissions, shown by the tar-like deposits, restricts
the content of organic material that can be tolerated in scrap material pro-
cessed.
In summary, this scrubber is limited to handling emissions from the pro-
cessing of selected types of scrap materials, excluding (1) metallic scrap
that contains more than small amounts of organic impurities, (2) all
residual scrap, (3) metallic scrap having large amounts of unmeltable
attachments. These limitations were determined through experience with
kettle-furnace processing. Although no information was obtained on use
of the scrubber with reverberatory furnaces, this scrubber would be
relatively ineffective in controlling the high rates of ZnO emission that
are expected from processing most scrap materials in reverberatory furnaces.
A further limitation is that there is no known use for the scrubber waste
slurry.
SCRUBBER, VENTURI (PLAIN WATER)
This study indicates that venturi scrubbers are not being used in the sec-
ondary zinc industry. However, effluents containing particles that
consist mostly of ZnO, emitted from a secondary brass plant, are being con-
trolled by a venturi scrubber (MC, 6/70). These emissions are from rever-
beratory furnaces wherein "blowing" operations are conducted. (See pages
VI-12, 13).
Part of this venturi system is made of stainless and part of mild steel.
The system can operate with a total pressure drop up to 54 in. w.g. at
gas flow rates up to 23,000 ACFM.
-------
(VIII - 8)
Operation of the system has been satisfactory. Emission tests have shown
collection efficiencies between 92 and 99% at effluent gas temperatures
around 500°F and at a resulting saturated gas temperature of 90°F. There
has been no corrosion or abrasion of either the stainless or mild steel
parts of the system. The ZnO apparently provides a protective coating effect
over the bare metal of the scrubber system. (The emissions are probably of
close to neutral pH.)
The system incorporates a filtering device that discharges collected parti-
culates as wet filter cake.
BAGHOUSES
Dacron (polyester) mu1tifilament bag filters are being used to collect emis-
sions from a combination of several reverberatory and distillation retort
furnaces. (Emission points are as indicated in Tables IV-1 and IV-2). Materials
processed include contaminated die-cast scrap and galvanizers' drosses, which
may contain very small amounts of chloride flux. No additional flux is
applied. The collected particulate material contains about 95% ZnO and
may include small percentages of chlorine as chlorides.
The installation (duct size, length, etc.) and process operations of this
system are such that the effluent temperature at the point of baghouse
collection is around 250°F. By maintaining this temperature and the low
percentages of chlorides in particulates being collected, blinding of the
bags and caking of collected material due to deliquescence are prevented.
Carbonaceous emissions are prevented by virtually complete incineration
of organic material in the reverberatory furnaces. The bags are satisfactorily
cleaned by means of a pneumetic air shaker.
-------
(VIII-9)
The company using this system reports that at temperatures as high as 212 F,
ZnCl2 particles become visibly wet and might blind bags by adsorption of
atmospheric moisture. Hence, the higher effluent temperature of 250°F is
maintained.
This baghouse installation incorporates about 1700 square feet of cloth
surface, with an air-to-cloth ratio of 2 to 1, maximum. Flow through the
system is about 28000 cfm, and pressure drop across the bags is 2 in. w.g.
A 50-HP motor drives the exhaust fan. The system collects about 350 pounds
of particulate per day.
In another installation, a baghouse using polyester fabric bags collects
emissions from reverberatory furnaces where contaminated die cast scrap is
processed without applying flux. Molten metal from this furnace is sub-
sequently distilled. Again, organic material in charged scrap is satisfactorily
incinerated in the reverberatory furnace, and baghouse collection is satis-
factory.
Experience in baghouse applications indicates that polyester fabric is the
most acceptable bag material in present use for collecting emissions from
zinc-distillation processes and limited zinc-sweat processes.
Limitations of polyester-fabric filters have not been established, but it
seems evident that at maximum atmospheric moisture content usage would be
limited to certain ranges of chloride concentration in the emitted particulate,
those ranges being defined with effluent temperature and air-to-cloth ratios
as the principal parameters. The experience reported above indicated that ant
maximum ZnCl- oSncentratAan^betwien 2 .and 15% at an ef f luentetempefcatiffrei of .i250 F
and a maximum air-tb-cl«th rationof 2'to il foils"Wifchincan:aCCeptab.&«.range.
-------
(VIII-10)
Model E probably represents marginal process and emission conditions, in
which a baghouse might or might not perform satisfactorily.
Polyester-fabric bag filters could not be used where charges containing
considerable amounts of residual scrap are being sweat processed in either
reverberatory or kettle furnaces because of chlorides that would be emit-
ted. Use of the bag filters would be further precluded where metallic scrap
charges containing significant organic material are processed in kettle
furnaces because of carbonaceous, as well as chloride, particulate likely
to be emitted. Thus all models in Table VII-1 except Model E seem precluded
from emission control by means of baghouses.
ELECTROSTATIC PRECIPITATORS
An electrostatic precipitator is used to clean exhaust effluents from
processes consisting of combinations of units approximating Model A, B, and
D. Charges to furnaces consist mainly of residual scrap mixtures having con-
siderable residual flux content. Small amounts of metallic scrap containing
organic materials may also be included. Therefore, chloride content of
emitted particulate is high, and small percentages of carbonaceous parti-
cluate may also be emitted.
This precipitator has the general configuration of the flat-surface type
shown in Control Techniques Figure 4-45. Electric conductors in the pre-
cipitator that are exposed to emissions are made of stainless steel. An
electric heater prevents accumulation of moisture in the precipitator.
Electric power consumption by the precipitator is low, most being consumed
by 15-HP fan motor used to exhaust the precipitator, very little being
-------
Cvin -
consumed at the electrodes.
During operation, the pressure drop is 0.3 in w.g. across the total exhaust
and emission control system at an effluent flow rate of 15,000 ACFM and
120°F effluent temperature.
Results of an emission test showed a collection efficiency exceeding 98%,
reducing emissions to less than .01 Gr./SCF, with :alxHi>t -57 pounds of'
particulate per hour being collected at a total process weight rate of
about 2.4 tons per hour.
Some corrosion of the stainless steel electric conductors has resulted from
contact with the collected chloride particulates. The extent of this
corrosion is significant, but not so great as to prevent use of the system.
Taere is some possibility that carbonaceous particles, if present in emissions
being collected, might be ignited by electrical discharges within the
precipitatorr
The presence of ZnCl_ probably imparts electrical conductivity to the
particulates that enables their efficient collection in electrostatic
precipitators. Because of high resistivity, there is probably a limit on
concentration of ZnO in particulat* at which precipitator collection
efficiencies are satisfactory. fDonoso p. 8).
Following are the principal limitations and disadvantages of the electrostatic
precipitator for the application described:
A. Corrosion of stainless steel parts.
B. Possible ignition hazards where charges contain organic material
and where furnace conditions are not conducive to complete incineration
of those materials. (See Chapter VII)
-------
(VIII - 12)
C. Lack of a known use for the mixtures of ZnCl- and ZnO with small
amounts of other substances collected in the precipitator.
D. Restriction of efficient collection to a certain range of particulate
composition and effluent temperature. High ZnO concentration and high temperature
could reduce electrical conductivity and cause inefficient collection; high
ZnCl~ concentration (with low ZnO content) could cause excessive corrosion.
AFTERBURNERS
Exhaust effluents from rotary furnaces are passed through afterburners to
burn carbonaceous particulate and gaseous emissions. Additional gas clean-
ing is usually not done because concentration of noncarbonaceous emissions
in these effluents is low.
One problem was reported on afterburner usage. At times during sweat pro-
cessing with afterburners in operation, significant carbonaceous emissions
do not occur. Because no control devices or methods are available to detect
this condition and responsively stop afterburner operation, fuel is wasted
duriiig such periods.
CONCEPTUAL EMISSION CONTROL SYSTEMS
Information is given here concerning the only system investigated during this
study that,has been tested as a new approach to controlling emissions
from zinc-sweat processes. In the following chapters on Conclusions and
Recommendations, emission control concepts will be developed further, based
on the total findings presented in this report.
The conceptual system that had been tested was a laboratory-scale modification
of the system for removing solid and mist particles shown in PEP 3/67. In
this modified system, part of the effluent from a zinc-sweat process approximating
-------
(VIII - 13)
Model C was filtered through a water-irrigated, polyester-felt diaphragm
(approximately 12-inches diameter and 1/2-inch thick). During the tests,
process conditions were varied to yiald emissions having the composition
characteristics shown below:
Resulting Characteristics of
Applied Process Conditions Particulate Emission Composition
A. High ZnCl_ flux content in charge and High Chloride (ZnCl_) content.
normal molten metal bath temperature.
B. High molten metal bath temperature and High ZnO content.
no ZnCl- flux in charge.
C. High organic material (oil) content in High carbonaceous (oil droplet) content,
charge.
Findings of this test were limited to these visual observations:
A. The irrigated filter was very effective in removing ZnCl- Particulate
emissions generated by process condition A_, as shown by clean air downstream
of the filter.
B. The irrigated filter was only partly effective in removing ZnO
particulate emissions generated by process condition B, as shown by visible
smoke downstream of the filter and white clouding of the scrubbing water, both
occurring at the same time.
C. The irrigated filter was blinded very quickly, increasing greatly
the pressure drop across the filter, by oil droplets emitted during process
condition C. After the test, notable white ZnO could be seen on the downstream
side of the filter, while greater intensity of the black burned-oil color
appeared on the upstream side.
Although no. quantitative data were available from this test, the data in
the reference article are significant. In a pilot unit, thin, water-irrigated,
felt filter pads were tested. The effluents being cleaned were exhaust gases
from pigment calciners that contained sulfuric acid mist and TiO_
pigmentary particles at about 167°F temperature. Filter pads made of
-------
(VIII - 14)
polyester, polypropylene, and polyacrylonitrile fibers were tried. The
housing in which the filter was mounted and thorough which the effluent
passed was made of polyester glass resin.
In several tests in which polyester and 'polypropylene felt filters were
used (individually), collection efficiencies were reported between 94 and
100% for TiCL removal; and between 83 and 94% for acid mist removal. During
these tests, effluent gas flow rates ranged from 75 to 170 ft/min.[(cu ft/
min)/sq ft] and pressure drops were from 6.5 to 9.6 in. w.g. The reference
article comments that a minimum flow rate of 8 gallons of spray liquor per
1,000 cubic feet of gas was required for high mist removal. Loading of acid
mist in gas being cleaned was as high as 1.74 grains/per standard cubic foot.
The relative suitability of the felt materials tried for this application
was^not discussed in the reference article; endurance of a polyester mat-
erial was especially noted, however. This material lasted until its support
screen failed after a 59-day test period. It was also noted that polypropylene
and polyester are very much more resistant to wetting than polyacrylonitrile,
with polypropylene slightly more resistant than polyester. (As noted above,
high collection efficiencies were attained with both polypropylene and
polyester felt filters.) A PVC-coated glass mesh screen provided satisfactory
support for the filter pad.
The test results shown above, considered together, indicate that ZnCl_ particulates
can be removed from zinc sweat process effluents at very high efficiencies and
low energy consumption (low pressure drop) by means of this conceptual system.
The principal limitations of the irrigated fibrous felt system for cleaning
zinc sweat process effluents are blinding of the filters by carbonaceous
particulates in effluents being cleaned, and low efficiency of ZnO particulate
removal.
-------
Chapter IX
Conclusions
Contents Page No.
PROCESSES OF SECONDARY ZINC PRODUCTION IX - 1
EFFECTS OF PROCESSES ON EMISSIONS IX - 1
SELECTION OF PROCESSES TO MINIMIZE EMISSIONS IX - 2
EMISSIONS FROM SECONDARY ZINC SWEAT PROCESSES IX - 4
EMISSION CHARACTERIZATION IX - 4
CORRELATION OF CHARGED MATERIALS WITH IX - 5
EMISSION COMPOSITION
EVALUATION OF EMISSION DETERMINATIONS AND
PROCESS MODELS IX - 6
EMISSION CONTROL SYSTEMS IX - 7
CONTROL OF HIGH-CARBON-CONTENT EMISSIONS BY
CURRENT TECHNOLOGY IX - 7
CONTROL OF LOW-CARBON-CONTENT EMISSIONS BY
CURRENT TECHNOLOGY IX - 7
PROBLEMS OF UTILIZING AND DISPOSING OF COLLECTED
PARTICULATE MATERIAL IX - 9
AVAILABILITY OF DATA FOR PHASE II WORK IX - 9
-------
Chapter IX
Conclusions
PROCESSES OF SECONDARY ZINC PRODUCTION
EFFECTS OF PROCESSES ON EMISSIONS
The effects of processes, as defined by the type of furnace used, on resulting
emissions are characterized and compared as follows:
Emission Characteristics of Reverberatory Furnaces
1. Carbonaceous emissions are controlled satisfactorily by incineration of
organic material within the furnace, where furnace design and operation are satis-
factory.
2. Noncarbonaceous particulate emission rates are higher than those of kettle
or rotary furnaces processing comparable charges.
3. Effluent flow rates are higher than those of kettle furnaces processing
comparable charges.
Emission.Characteristics of Rotary Furnaces
1. Carbonaceous emissions are not controlled satisfactorily by incineration
within rotary furnaces of current design but are controlled satisfactorily by
application of afterburners.
2. Noncarbonaceous emissions are not significant, where flux is not applied
and where furnace design and operation are satisfactory.
3. No data were obtained on effluent flow rates of rotary furnaces.
-------
(IX - 2)
Emission Characteristics of Kettle Furnaces
1. Carbonaceous emissions are not controlled satisfactorily by incineration
within the furnace but are usually held at levels low enough that afterburning
is not required by charging scrap material mixtures that do not contain large
amounts of organic materials.
2. Emissions of noncarbonaceous particulate are significant when the
furnace charge contains significant amounts of flux and inorganic impurities,
but are lower than those of reverberatory furnaces processing comparable charges.
3. Effluent flow rates are lower than those of reverberatory furnaces
processing comparable charges.
Emission Characteristics of Distillation Furnaces
1. Carbonaceous emissions do not occur.
2. Emissions of noncarbonaceous particulate are significant but are limited
to nearly pure ZnO (no significant chloride content).
3. No data were obtained on effluent flow rates from distillation furnaces.
SELECTION OF PROCESSES TO MINIMIZE EMISSIONS
Emission characteristics of the several types of secondary zinc processing fur-
naces can be applied by selecting specific combinations of furnaces to minimize
emissions and effluent flow. This approach to emission control is effective
in certain applications, as shown by two notable examples:
1. Scrap containing considerable organic material is processed without
flux in reverberatory furnaces, and the resulting sweated crude zinc alloy
is distilled to obtain pure zinc. In this way, carbonaceous and chloride
emissions are prevented and noncarbonaceous emissions (free of chlorides) that
occun can be collected in conventional baghouses.
-------
(IX - 3)
2. Scrap containing considerable organic material is processed without
flux in rotary furnaces with afterburners. The resulting crude zinc alloy
is then processed further with application of relatively small amounts of
flux in kettle furnaces. In this way, carbonaceous emissions are controlled
and noncarbonaceous emissions arei minimized, and thus the requirement for
gas-handling capacity of any equipment used to clean those effluents is re-
duced.
Minimizing emissions through furnace selection is well established and appears
to have been optimized in the secondary zinc industry. Emissions cannot
be controlled completely by this approach, however, as the examples
illustrate. A further example of the limits of this approach to emission
control is that high-chloride-content particulates are emitted from processes
of sweating residual scrap where considerable flux is present, regardless of
furnace selection. (Pretreatments to remove chlorides from residual scrap
are of limited economic practicality;) .• Besides considerations of emission
control, process selection depends on materials to be processed, products to
be made, and process economics. Conservation of alloy metals (mainly aluminum
and copper) may be a further consideration in process selection. These
metals are utilized automatically where zinc alloy is sweated from scrap,
then realloyed to meet a zinc alloy specification. The alloy metals might
be partly or wholly lost where scrap-derived crude zinc alloy is distilled.
It is concluded that, with optimum process selection to minimize emissions,
problems remain in control of emissions from processes represented by the
process models developed in Chapter VII.
-------
(IX - 4)
EMISSIONS FROM SECONDARY ZINC SWEAT PROCESSES
EMISSION CHARACTERIZATION
Emissions from zinc sweat processes may be classified in two general catagories:
1. Emitted mixtures mostly of carbonaceous particulate and gaseous sub-
stances. (Noncarbonaceous particulates might be present, but only in minor
amounts.) These mixtures occur most characteristically in emissions from
rotary furnaces.
2. Emitted mixtures mostly of noncarbonaceous particulate substances,
principally ZnCl,, and ZnO. (Carbonaceous particulate. and gaseous substances
might be present in amounts which, though relatively small, could be significant
in emission control.) These mixtures occur most characteristically in
emissions from kettle and reverberatory furnaces.
Conclusions in the paragraphs that follow pertain only to the preponderantly
noncarbonaceous emissions of the second catagory.
Values approximating data that characterize these sweat-process emissions were
developed in Chapter VII and are shown for process model units in Table VII-1.
Following are the approximate ranges of quantitative properties characterizing
these emissions:
1. Emission factors range from negligible for processing clean scrap in
kettle furnaces to around 38 pounds of particulates emitted per ton of process
material for processing residual scrap in reverberatory furnaces.
2. In typical sweat processing of zinc scrap materials, the percentages of
equivalent anhydrous ZnCl- in particulate emissions range from none to around
34%, the percentages of ZnO from around 40% to 90% or more, and the percentages
of carbonaceous particles from none to around 10%.
-------
(IX - 5)
3. Emitted particles of ZnCl- and ZnO are mostly around 1-micron
diameter or lesfei)1 these small particles result from" valorization of
-
flux and metallic zinc followed by (1) condensation of ZnCl- vapor to
liquid droplets that solidify as cooling takes place and (2) atmospheric
oxidation of zinc vapor to solid ZnO particles. Some of the emitted particles
of these compounds may be somewhat larger than 1 micron because of agglomeration.
Also, larger particles of these and other oxiaB*1* and chlorides are emitted
through entrainment of pulverulent -material in the residue-flux mixture that
forms atop the process bath.
CORRELATION OF CHARGED MATERIALS WITH EMISSION COMPOSITION
Since chloride emissions result from chloride flux in charges to zinc
sweat processes, and ZnO emissions result in large part from metallic
zinc charged to those processes, the following relationships would hold:
A. The particulate ratio ZnCl?/ZnO increases as the charge ratio ZnCl,,/ Zn
B. The particulate ratio Cl /Zn increases as the charge ratio ZnCl_/Zn
where
n . _ ._„..,__ amount of ZnCl,, in emitted particulates
Particulate ratio ZnCK/ZnO = - - _ . 2 - r- — -3 - *— -. - - -
- 2t-s: — amount of ZnO in emitted particulates
_ . . .. . . .-,-/, ++ amount of Cl in emitted particulates
Particulate ratio Cl /Zn = - - — ,. .. ++ . — -..^ s - r- - i — r— •
- ' - amount of ZnTT in emitted particulates
_, _ ,,, /^ amount of ZnCl» in charge
Charge ratio ZnCl./Zn = - - — ? - r— ?T' — * — • — n -
- a - - - 2 - amount of metallic Zn in charge
These relationships would apply where other process conditions are constant.
Although these relationships were indicated by results of this study, data
were not sufficient for developing them as equations or graphical presentations.
-------
(IX - 6)
EVALUATION OF EMISSION DETERMINATIONS AND PROCESS MODELS
The conclusions characterizing emissions and relating them to process models
were derived from a small sampling of data; where data were not available,
theoretical considerations were applied. Deficiencies in data could be
enumerated from information on formulating process models in Chapter VII.
Two such deficiencies are noted here:
1. No chemical analyses were obtained on emissions from metallic scrap
processing. All available chemical analyses applied to emissions from residual
scrap processing.
2. All data on amounts of applied flax were •Approximations; 'flux allied
to processes was not weighed, and residual scrap was not chemically analyzed
to determine flux content before charging to sweat furnaces. The approximations
were based on familiarity of plant operators with usual processing techniques
and on published analyses of residual scrap.
The following conclusions pertain to the process models that were derived
from the process and emission determinations of this study and shown in
Chapter VII:
1. The process models indicate the general makeup of emissions from repre-
sentative zinc-sweat processes. Values shown for the models approximate
characteristics of those emissions closely enough that they may be used in initial
screening and evaluating studies of emission control concepts leading to
development and testing of pilot equipment.
2. The models establish a format for any further data gathering needed
for research and development.
3. Because of the limited data sampling, it would be benificial to obtain
additional data to confirm and enhance the accuracy of the models.
-------
(IX - 7)
4. Complete data on emissions and related processes would be
needed for design of prototype or demonstration equipment. These data
should be obtained at the plants where this equipment would be tested.
EMISSION CONTROL SYSTEMS
CONTROL OF HIGH-CARBON-CONTENT EMISSIONS BY CURRENT TECHNOLOGY
The preponderantly carbonaceous emissions, characteristic of rotary-
furnace sweat processing of contaminated zinc die-cast scrap, are con-
trolled with afterburners. Only one limitation of this emission control
system was indicated during this study. That is, fuel combustion
efficiency of the afterburners is limited by lack of devedes. and/or
methods for stopping their operation when carbonaceous emission rates
reach insignificant levels.
CONTROL OF LOW-CARBON-CONTENT EMISSIONS BY CURRENT TECHNOLOGY
Equipment used in past and present attempts to control emissions of
relatively low carbon content (represented by Models A through E of
Chapter VII) is of limited effectiveness. Principal limitations are
shown as follows:
Corrosion of Metals by High-Chloride-Content Emissions
Emissions having high chloride concentration as represented by Models A, B,
and D, cause considerable corrosion to mild steel and stainless steel parts of
emission control equipment. This limitation has been experienced with plain-
water scrubbers and electrostatic precipitators. Where chloride concentration
of emissions is low enough these corrosive effects are tolerable.
-------
(IX - 8)
Blinding of Fabric Filters by High-Chloride-Content Emissions
Emissions having high chloride concentration, as represented Models A,B, and D,
cause blinding of fabric filters due to deliquescence of ZnCl . This limitation
has been experienced with bag filters. Another effect of deliquescence of
ZnCEiiiis corrosion of some organic-fiber, bag filter materials by collected
particulates. Bag filter blinding (and conversely the range of conditions
in which bag filters are useful for collecting emissions) is mainly a
function of ZnCl_ concentration in particulate emissions and effluent temperature.
Use of higher effluent temperatures reduces deliquescence, but this practice
is limited by temperature resistance of the bag material. (Carbonaceous
particulates, as emitted from Model B, might also contribute to bag blinding.)
Wear of Fibrous-Glass Bag Filters
The usual fibrous-glass fabric bag filters, when used to collect low-
chloride-content emissions (represented by zinc-distillation processes and
by model E), do not withstand, abrasive wear of woven threads that occurs
during bag cleaning.
Carbonaceous Deposits in Emission Control Equipment
Carbonaceous emissions, as represented by Models B and C, may cause oily or
tarlike deposits in gas cleaning equipment. This problem has been experienced
at exhaust fans of plain-water scrubbers, but has been reduced to a manageble
level by fan selection. Carbonaceous deposits could also occur in baghouses
and electrostatic precipitators.
Ignition Hazard
Carbonaceous emissions, as represented by Models B and C, could be ignited in
emission control equipment, where carbonaceous deposits accumulate and
at intermittent periods when carbonaceous emissions occur at higher rates than
-------
(IX - 9)
indicated for the models . Bag filters and electrostatic precipitators appear
susceptible to this hazard.
Resistivity of Particulates Outside the Collecting Range of
Electrostatic Precipitators
The range of particulate emission compositions and the corresponding range of
process conditions wherein particulates can be collected at satisfactory
efficiencies in electrostatic precipitators are not known. Effectiveness
of electrostatic precipitators depends largely on resistivity of particles
being collected, a property that varies with composition of particulate
emissions. Therefore for some particulate compositions collection in
precipitators probably would not be satisfactory.
Blinding of Irrigated Felt Fabric Filters by Carbonaceous Particulates
Carbonaceous emissions, represented by Models B and C, cause blinding of
felt filters irrigated with plain water.
PROBLEMS OF UTILIZING AND DISPOSING OF COLLECTED PARTICULATE MATERIAL
Commercial usage of collected particulates having less than around' 95% ZnO
has not been determined. Where large amounts of chlorides are present in
the collected'particulates, disposal of this material presents problems
because the chloride constituents are corrosive.
AVAILABELITY,OF DATA FOR .PHASE II WORRK
The data obtained during this study apply most substantially to sweat
processing of residual scrap and mixtures ^<>f :TesiJdual and "metallic scrap
with chloride flux being contained in charges and -with.resulting emissions
containing considerable chloride particulates. These data are therefore the
most available for subsequent (Phase II) work.
-------
CHAPTER X
Emission Control Concepts
Contents Page No.
CONCEPT OF PROCESS MODIFICATION TO REDUCE NON-CARBONACEOUS X-l
EMISSIONS. BASED ON FINDINGS OF THIS STUDY
GASEOUS FLUXING AND WORKING OF SEALED PROCESS BATHS X-l
CONCEPTS OF EMISSION CONTROL SYSTEMS. BASED ON
FINDINGS OF THIS STUDY X-3
CHLORIDE CONDENSATION SYSTEMS X-A
CHLORIDE SOLUTION SYSTEMS X-6
VENTURI SCRUBBERS X-ll
BAG FILTERS X-13
ELECTROSTATIC PRECIPITATORS X-16
AFTERBURNERS X-16
GRANULAR BED DEVICES X-17
CONCEPTS OF UTILIZING COLLECTED PARTICIPATE MATERIAL X-17
CONVERSION OF COLLECTED PARTICULATE TO ZINC OXIDE AND
ZINC CHLORIDE AS SEPARATE PRODUCTS X-17
REMOVAL OF CHLORIDE FROM COLLECTED PARTICULATE BY
MEANS OF-AMMONIA AND STEAM X-18
ZINC OXIDE ENRICHMENT OF COLLECTED PARTICULATE X-19
SMELTING OF COLLECTED PARTICULATE USING HYDROGEN X-20
CORRELATION OF CHARGED MATERIALS WITH EMISSION COMPOSITION X-20
RECOMMENDATIONS FOR FURTHER RESEARCH AND DEVELOPMENT X-21
-------
Chapter X
Emission Control1Concepts
CONCEPT OF PROCESS MODIFICATION TO REDUCE NON-CARBONACEOUS EMISSIONS.
BASED ON FINDINGS OF THIS STUDY
GASEOUS FLUXING AND WORKING OF SEALED PROCESS BATHS
To minimize the problems of controlling the non-carbonaceous emissions that result
from fluxing and working zinc-sweat process baths after any organic material has
been burned off, it would be desirable to minimize both particulate emission rates
and effluent flow rates. In an ideal system for accomplishing this, the bath would
be contained in a sealed chamber and heated by conduction or radiation. Such a
system would minimize (1) oxidation of molten metal due to contact with moving air,
(2) entrainment of particles from the flux-residue mixture that accumulates atop
the process bath, and (3) volume of effluent gas (infiltrated air being excluded
and no fuel combustion products being added to the effluent). In current practice.
the approach to this ideal is limited mainly by the need to conduct operations on
the process bath such as applying flux, stirring, and skimming.
The ideal process defined above might be approached by modifying current processes
as follows: After the charge is melted and most of any organic material in the
charge is burned off, the bath would be enclosed to seal out air. The only opening
in the bath closure would be a vent for exhausting effluent. Gaseous chlorine would
be lanced into the process bath causing formation of ZnCl., as C12 reacts with
molten zinc. (A1C1_ gas and MgCl~ solid would also be formed in smaller amounts.)
-------
(X-2)
Thus ZnCl« flux would be formed deep beneath the bath surface.
In this conceptual process, chloride emissions would be reduced since the
procedure would eliminate the vaporization loss that occurs immediately when
ZnCl,, is applied to the process bath surface, as in current processing. Air
infiltration and resulting particulate entrainment would be reduced, since the
closure would not be opened for application of flux and for manual mixing of
flux into the bath. The bath would be agitated and mixed by gas bubbling through.
To increase this effect, if needed, inert gas such as N.. might be lanced in
addition to the Cl_. To increase this effect further, determination might be made
of the positions of the lances with respect to the vessel containing the bath,
and the flow rates of the lanced gases that would optimize stirring effects.
Initial tests might be made of lances installed projecting into the side of a
chlorinating vessel (which might have a general vertical, cylindrical shape with
hemispherical bottom) so as to apply the gases in a direction between radial and
tangential, at controlled flow rates.
Fluxing and working the bath by this chlorinating method might be done in the
original melting furnace (modified to suitably enclose the bath after burning off
any organic material in the charge) or it might be advantageous to transfer the
bath from the melting furnace to a separate vessel that could be closed during
chlorination. This transfer of molten metal could be made with ladles or by
means of a launder or piping system using gravity or mechanical pumping.
Since the reaction of Cl_ with the molten metals would be exothermic, it probably
could be conducted in a relatively simple vessel, with no provision for applying
-------
(X-3)
additional heat except perhaps to the lances, where heat might be needed to
prevent solidification of metal at times when gases are not being applied.
It appears possible that most, if not all, of the operations of fluxing and
working process baths might be done by means of this process modification
(although provision would have to be made for skimming baths), and that by
this means particulate emissions and effluent flow could be reduced considerably.
This process modification appears to be potentially applicable to processing all
metallic and residual scrap. Flux would generally be formed by lancing with
chlorine. However, where residual flux content of a scrap charge is sufficient
for fluxing needs, N2 alone might be lanced to accomplish the needed agitation
and mixing of the enclosed process bath.
CONCEPTS OF EMISSION CONTROL SYSTEMS. BASED ON FINDINGS OF THIS STUDY
High-carbon-content emissions from zinc sweat processes can be controlled
satisfactorily with afterburners. Therefore, the principal objective of research
and development pertaining to cleaning the exhaust effluent gases from secondary
zinc processes should be to develop devices capable of -cleaning effluents having
low carbon content but with significant chlorides; such emissions may be defined
as those from any combination of Process Model Units A through E (Chapter VII).
The most urgent need for such devices is for application to emissions from processing
charges that contain residual scrap with or without metallic scrap.
The above considerations lead to the following criteria for gas cleaning devices
to be developed:
-------
(X-4)
1. The devices should be capable of removing the following particulate
constituents from effluents:
a. Carbonaceous substances (liquid and solid) present as relatively
small fractions of total particulate (around 10% average, with varia-
tions from 0 to 20%).
b. ZnCl2(0 to 40% of total particulate).
c. ZnO (0 to 100% of total particulate).
2. These constituents should be removed in a minimum number of stages
(gas cleaning equipment units).
To meet these criteria several concepts of gas cleaning devices are
presented below.
CHLORIDE CONDENSATION SYSTEMS
The occurrence of a liquid state during formation of ZnCl- particulate (see
Page V-12) suggests that settling chambers, condensers, or reflux columns might
be developed for removing ZnCl? from effluent gas and separating that chloride
from ZnO.
In a system incorporating a settling chamber for that purpose, effluent would be
allowed to cool slowly to 689°F (the melting point of ZnCl2) to increase the
tendency to particle growth through condensation of ZnCl- vapor on surfaces of the
initially formed particles and through agglomeration of those particles. Agglo-
meration would be promoted by Brownian motion (at elevated temperatures) increasing
the collision and fusion of the liquid particles. The application of sonic agglo-
meration might be considered to further promote this effect. This system might
-------
(X-5)
be designed to allow particles to settle while still liquid or to allow them
to cool below 689 F, after attaining suitable size, and settle as solid parti-
cles. Such a system might include two chambers: the first to maintain an
elevated temperature to allow particle growth and agglomeration; the second to
allow cooling and settling, then removal, perhaps through a hopper.
A condenser or reflux column used for this purpose might contain baffles and/or
packings to provide surfaces on which ZnCl- droplets could impinge, collect, and
flow. Heated ducts might also be used in such units to maintain the needed
temperature range, while also providing surfaces for droplet impingement. If a
condenser were used, liquid ZnClp would be allowed to flow out for recovery. If
a reflux column were used, liquid ZnCl_ would flow back to the process bath for
recycling.
Waste heat from furnaces, afterburners, or other sources in secondary zinc plants
might be used to maintain the needed temperature in the conceptual settling,
condensing, and reflux devices described above. For example, effluent from a
reverberatory furnace might be channelled through a chamber constructed on top of
that furnace. Heat transferred from the top of the furnace into the chamber might
maintain the elevated effluent temperature needed to bring about the desired ZnCl-
particle growth. The effluent might then be channelled to an adjacent chamber for
cooling and settling.
Particles of ZnO would not have the tendency of ZnCl,, to grow and agglomerate by
vapor condensation and liquid fusion. It is, therefore, expected that 2hO particles
would pass through the settling chamber, condenser, or reflux column, and might be
collected in a baghouse, electrostatic precipitator, or scrubber. The separation,
-------
(X-6)
of course, would not be complete. There would be some mixing of ZnO with the ZnCl?
removed in the first (settling, condensing, or reflux) stage and of ZnCl_ with the
ZnO removed in the second gas-cleaning stage.
Most of the carbonaceous particles would probably be removed with the ZnCl« that
condenses or settles out, since the solid (fly-ash) type of carbonaceous particles
would generally be relatively large and liquid carbonaceous particles would tend
to agglomerate forming larger particles. With the reflux column, the collected
carbonaceous material would be returned to the furnace (with ZnCl_), where it
might be ultimately consumed in combustion.
Ceramic materials appear most promising for construction of these systems, but
metals including mild steel might be tried especially for parts like hoppers,
where dry particles would be handled.
CHLORIDE SOLUTION SYSTEMS
Zinc chloride is soluble in water, and ZnCl- particles are wetted readily by
liquid and vaporous moisture. In contrast, ZnO is insoluble in water and resists
wetting. This comparison of properties suggests that ZnCl_ particulate might be
efficiently removed from effluents by means of aqueous scrubbers that operate at
low energy loss (low effluent gas pressure drop); and that ZnCl» might thereby be
separated from ZnO particles in the effluent, the latter requiring a high energy
loss for removal in aqueous scrubbers. (See Page VIII-7). A scrubbing system
for this purpose might include one or more chambers and/or towers containing water
sprays, water films, irrigated packings, and/or baffles alone or in combination.
-------
(X-7)
For example, a spray chamber or irrigated packed tower might prove sufficient, or
such units might be used in combination. (See Control Techniques 4.4.) As ZnCl '
particles emitted from a zinc sweat process passed through the scrubber, they would
absorb moisture, increasing their weight and tendency to agglomerate. The result-
ing increased inertia would increase their interception by water drops, water films,
packings, and/or baffles in the scrubber. The small ZnO particles would retain
their low inertia and tend to pass through the scrubber.
The potential of the conceptual chloride solution system suggested above is further
indicated by results of tests of water-irrigated diaphragm filters. (See Page VIII-13.'
These tests showed that ZnCl- particulate emissions can be efficiently removed
from effluents, at .low energy loss, by interception at water-irrigated fibers;
while ZnO, in large part, passes through.
The indicated potential of this concept is also supported by results of tests
in which HC1 gas was efficiently collected at low energy loss by several types of
packed aqueous scrubbers (KSB).
Particles of ZnO that pass through the conceptual aqueous scrubbing system,
probably with some ZnCl_ retained in the effluent, might be collected in a
baghouse, electrostatic precipitator, high-energy venturi scrubber, or other
suitable equipment. If a dry collector such as a baghouse or electrostatic
precipitator were used, it might be necessary to first dry the effluent and/or
raise its temperature above the dew point. A dry bed or similar device might be
used to eliminate mist. Waste heat from furnaces or afterburners in the plant
might be used to raise the effluent temperature.above the dew point. For example,
-------
(X-8)
the effluent (after passing through the scrubber) might be passed through a
chamber or other heat exchange device positioned above a reverberatory furnace,
thus applying heat from the furnace to the effluent. A heat exchanger might
also be devised that would utilize heat from the exhaust gas of a rotary furnace
afterburner for that purpose.
In devising a scrubbing system to eollect ZnCl_, the various wet collectors
and ancillary equipment shown and discussed in Control Techniques, 4.4, should
be considered. Plain water, NH.OH, and non-foaming detergent solutions might
be considered as scrubbing media for ZnCl2 removal.
Scrubbing with plain water would seem to generally preclude use of irrigated
filters approximating the felt diaphragm type described on Pages VIII-12,13
because of blinding by carbonaceous particles. However, other low-energy-loss
scrubbers shown in the above-mentioned reference might be considered.
Scrubbing solutions of NH.OH, maintained neutral or slightly alkaline, would
probably overcome the acidic corrosion caused by hydrolysis of ZnCl- and NH.C1.
Collection of carbonaceous particles might be increased because of reduced sur-
face tension resulting from NH.OH in the scrubbing solution. However, alkalinity
might cause precipitation of enough Zn (OH)_ to clog some types of scrubbers.
Therefore, scrubbing with NH.OH solution might restrict selection of scrubbing
devices to those that resist clogging, such as spray types.
Scrubbing solutions of non-foaming detergents including NaOH and sodium phos-
phates might be tried in an irrigated-filter scrubber, such as the felt diaphragm
-------
(X-9)
type described on Pages VIII-12,13, the detergents being used to remove
carbonaceous particles (mainly oil droplets) and prevent blinding of the filter.
The pH of such solutions would have to be controlled to prevent Zn(OH)?
precipitation. Attempts might be made to accomplish this by determining a
satisfactory buffered-solution composition.
Zinc chloride and NH,C1 collected in plain water or NH.OH scrubbing solution,
as suggested above, might be recovered. Settling and filtration might be
tried to remove insoluble (mainly carbonaceous) substances. Removal of insoluble
substances in this way would be mainly applicable to plain-water scrubbing systems.
Resulting solutions might be evaporated, perhaps using plant waste heat, to
recover the chlorides.
The efficiency of the first (scrubbing) stage of this conceptual system in
removing carbonaceous emissions as well as chlorides (mainly ZnCl_), would be
important to determining feasibility of this system because of the possible
blinding of filtering devices and the ignition hazard that would be posed by
carbonaceous substances in the subsequent ZnO removal stage.
As mentioned earlier, the corrosiveness of the chlorides handled by the scrubber
might be overcome by scrubbing with NH.OH solution. Pertaining to plain-water
scrubbing (as suggested above) materials that have been found satisfactory in
scrubbers handling corrosive chlorides include mild steel coated with certain
rubbers and plastics, fiberglass-reinforced polyester plastics, and acid-resistant
brick (CEP 3/67, Donoso, KSB). Stainless steel has been used successfully for
handling solutions of low ZnCl. concentration, but has corroded significantly in
-------
(X-10)
direct contact with collected particulate having high chloride concentration
and with solutions of high ZnCl- concentration (see Pages VIII-5,6,11). Both
ceramic and plastic packings have been used in scrubbers collecting corrosive
chlorides. Mild steel duct work has been satisfactory for carrying emissions
from furnaces to stacks and gas cleaning equipment, showing that solid ZnCl-
particles do not present a corrosion problem in the absence of moisture.
Consideration might be given to including an afterburner as the first stage of
this conceptual emission control system, to be operated when carbonaceous emissions
are high and by-passed when not needed. Waste heat from this afterburner might
be used for any of the purposes mentioned herein. Conceivably, such an after-
burner might be operated as a source of heat for those purposes, even when not
needed to control carbonaceous emissions.
In evaluating this concept, the effects of several variables on collection
efficiency of scrubbers used to remove ZnCl_ should be determined. These deter-
minations could lead to optimizing those variables for maximum collection effi-
ciency. These variables seem particularly pertinent:
1. Temperature of effluent entering the scrubber.
2. Concentration of ZnCl- in effluent entering the scrubber (which would
tend to increase as air infiltration rate decreases). Might be expressed as grains
ZnCl™ per standard cubic foot.
3. Phase or phases of ZnCl- in the effluent entering the scrubber (solid,
liquid, vapor, or combination of these). This variable would be determined by 1
and 2.
4. Retention time of the effluent at an elevated temperature. (Length of
retention time at temperatures between the melting and boiling points of ZnCl2 would
probably determine particle growth, other variables being constant).
-------
(X-ll)
Plant waste heat might be utilized to obtain effluent conditions (exemplified
by variables 1, 3, and 4) determined as optimum for maximum efficiency of ZnCl_
removal.
VENTURI SCRUBBERS
Overcoming the corrosion problems experienced in attempts to control chloride-
containing emissions with venturi scrubbers (noted in Chapter VIII) might be
accomplished by two approaches to research and development.
Use of alkaline scrubbing solutions to neutralize the acidic corrosive
effects of the chlorides being collected.
Alkaline compounds that might be tried in such scrubbing solutions include NH.OH,
Ca(OH)», NaOH and sodium phosphates. Also, investigators might try aqueous mixtures
of insoluble alkaline-reacting substances, such as granulated limestone and the
collected alkaline waste dust from portland cement processes (AEMPC, p.17). The
success of this concept would depend on whether a solution composition and pH
could be determined (the maintained) that would prevent corrosion but not cause
precipitation of Zn(OH)9 to an extent that would clog the gas-cleaning equipment.
Use of corrosion-resistant material.
Venturi scrubbing systems in which the venturi scrubbing unit and the cyclone
separating unit (see Control Techniques, Fig. 4-18) are made of corrosion-
resistant materials might be tried. Several conceptual material applications
could be considered for this purpose.
-------
(X-12)
1. Molded or laminated, fibrous glass reinforced, thermosetting plastics.
Of the methods of fabricating the venturi and cyclone units with fibrous-glass-
reinforced polyester, epoxy, or other thermosetting plastics, spray forming would
probably be the least costly. Molding fiber-filled resin, mat laminating, and
filament winding are further possibilities. Inserts made of corrosion-resistant
metals or alloys, such as titanium, niobium, or other "reactive metals" might be
used in the venturi throat and other areas of high gas velocity, if needed to
resist abrasive wear and help dissipate heat (from friction and the sensible heat
of the effluent). Aluminum and its alloys might also be tried as insert material
(see Perry, p. 23-28). A further variation of this basic concept would be the
use of thin shells or platings of corrosion-resistant metal over interior surfaces
of the venturi and cyclone units. Also, metallic fibers might be incorporated in
the plastic to conduct and dissipate heat.
2. Molded thermoplastics. Thermoplastics of some types might be molded
to produce the units. A key consideration in selecting thermoplastics for this
use would be the effluent temperature of the process to which the units would be
applied. Findings of this study indicated that the usual temperatures of kettle-
sweat process effluents are low enough to permit use of a wide range of thermo-
plastics, including some low-cost polyolefins; the higher temperatures of rever-
beratory-sweat process effluents would narrow that range (applicable to those
processes), and could require usage of more expensive thermoplastics. Where
conditions permit the use of polyolefins, the units could probably be used as
"disposable" items, with frequent replacement, because of the low material cost.
Rotational molding appears to be a promising method of fabricating the units.
Supporting parts made of a low-cost metal might be used with the units to provide
-------
(X-13)
rigidity and help dissipate heat. For example, the venturi and/or cyclone separator
of a venturi system might be made of a molded plastic, and these parts might be
enclosed in shells made of mild steel or aluminum, which could be opened to allow
removal and replacement of the plastic parts when necessary. Carbon black incor-
porated in formulations might reinforce and otherwise enhance properties of
plastics (particularly polyolefins) used in venturi scrubbing systems. Fibers of
glass and/or metal might also be incorporated in the plastic for reinforcement
and/or heat dissipation. Corrosion-resistant metallic inserts could be used in
the venturi throat and other points of high gas velocity if needed to resist
abrasive wear and to dissipate heat. Use of such inserts might be appropriate if
a relatively expensive plastic were used.
3. Plastic-coated metal. The units might be made of low-cost metals such
as mild steel or aluminum alloy, with coatings of rigid or elastomeric plastic
that would withstand corrosion and abrasion. Coatings of polyurethane or vinyl
plastisol might be tried for this purpose. The metal in such structures would
provide rigidity and help dissipate heat.
4. Corrosion-resistant metals. The unit might be made entirely of
corrosion-resistant metal. Metals that might be considered for this application
include aluminum (see Perry p. 23-28), titanium, other reactive metals, and alloys
of those metals. Mild steel coated with lead poses a further possibility.
BAG FILTERS
To overcome the problems incurred in bag filter application (noted in Chapters
VIII-IX), two related determinations might be made. These are listed below, with
concepts of bag filter development and application.
-------
(X-14)
Determination of bag filter material
A determination would be made of bag filter materials having properties that show
the most promise of resisting (1) the wetting, adhering, and corrosive effects of
ZnCl_ with adsorbed moisture from the effluent gas (mainly atmospheric moisture of
infiltrated air), (2) frictional wear that occurs in cleaning fibrous-glass bag
filters, and (3) elevated temperatures that might be maintained in effluent to
minimize moisture adsorption.
Of currently availabel fibrous materials that might be considered for this
determination, polyesters should be included. Also, fluorinated polymers should
be investigated both as potential fiber materials and as coatings that might be
applied to other fibers such as fibrous glass. Initial investigations of fluori-
nated material might include determining the availability or possible development
of fibrous glass treated to obtain a fluorinated silicone coating on the glass
fibers. Fluorinated-silicone-coated glass fibers might have the desired properties
of resisting wetting, adhesion, corrosion, frictional wear (between fibers of woven
threads), and high effluent temperatures.
Determination of limitations of ZnCl,, particulate concentration for
satisfactory bag filter collection
After determining the most promising bag filter materials, a determination should
be made of the concentrations of ZnCl~ in emitted particulate that limit use of
bag filters (made of those materials) within the temperature range at which the
effluent might be maintained. Higher temperatures would reduce the tendency of
-------
(X-15)
ZnCl- to adsorb moisture. Therefore, one would expect correlation between
effluent temperature and the maximum ZnCl. concentration in particulate that
could be collected in a bag filter, higher ZnCl_ concentration being tolerable at
higher effluent temperatures. Effluent temperature is limited by the temperature
resistance of the bag filter material. This determination might best be presented
as a graph showing maximum ZnCl_ concentration, as percent of emitted particulate,
versus temperature of effluent entering the bag filter, bag filter materials being
shown as parameters, and effluent moisture content (which would be determined
mainly by ambient relative humidity) held constant at the maximum expected level.
Other variables, including air-to-cloth ratio, and particulate emission rate,
could also affect this correlation; thus several graphs might be required to
represent a realistic range of process conditions. In this way, the useful range
and the practicability of bag filters for controlling zinc sweat process emissions
could be established.
Bag filters appear more promising for use with reverberatory-sweat furnaces than
with kettle-sweat furnaces because of the following characteristics:
1. Combustion efficiency of reverberatory-sweat furnaces is higher, with
consequent elimination of carbonaceous emissions where furnace design and operation
are satisfactory. (Carbonaceous particulate could blind bags and pose a fire
hazard in a baghouse.)
2. Effluent temperature of reverberatory-sweat furnaces is higher because
of combustion products entering the effluent, which would reduce moisture adsorption
by emitted ZnCl_. This advantage might be offset in some instances by water vapor
introduced by fuel combustion.
-------
(X-16)
3. Dilution of emitted ZnCl_ with ZnO particles is greater in reverberatory-
furnaces, which generate more ZnO particles because of turbulence of air and
combustion gas, and higher molten metal temperature.
ELECTROSTATIC PRECIPITATORS
To more fully establish the applicability of electrostatic precipitators and to
optimize their usage for controlling zinc sweat process emissions, determinations
might be made of the following:
1. Optimum materials, devices, and/or methods for overcoming corrosion
of precipitator parts caused by chlorides collected in the precipitator (noted in
Chapters VIII-IX).
2. The nature of any ignition hazards resulting from carbonaceous or
other flammable emissions and limitations thereby imposed on use of electrostatic
precipitators.
3. The ranges of particulate composition and other variables in which
emissions control by means of electrostatic precipitators is satisfactory. This
determination might be presented as graphs of collection efficiency versus ZnCl_
concentration, with other variables such as effluent temperature, flow rate, and
other chlorides present, held constant or shown as parameters.
AFTERBURNERS
Consideration might be given to developing a device and/or method for conserving
afterburner fuel by stopping fuel combustion in afterburners when carbonaceous
emission rates reach insignificant levels.
-------
(X-17)
GRANULAR BED DEVICES
Granular bed devices may offer possibilities of controlling emissions containing
ZnCl_ because the moving granules of inert mineral would resist the corrosion and
blinding that occurs with other filtering devices. (See GBP)
CONCEPTS OF UTILIZING COLLECTED PARTICULATE MATERIAL
In present emission control practice, emissions from zinc sweat processes that
contain considerable chlorides are collected in two forms (1) aqueous mixtures
from scrubbers containing ZnCl., ZnO, and sometimes small amounts of carbonaceous
substances, and (2) mixtures of solid particles from electrostatic preclpltators
containing ZnCl. (with some adsorbed moisture), ZnO, and sometimes small amounts
of carbonaceous substances. Developing uses for these materials would alleviate
the problem of their disposal and allow more favorable economics of emission
control systems. Four concepts of systems for utilizing these materials are
presented and discussed.
CONVERSION OF COLLECTED PARTICULATE TO ZINC OXIDE AND ZINC CHLORIDE
AS SEPARATE PRODUCTS
Collected particulate might be treated with steam in attempts to effect the
reaction:
(1) ZnCl2 + 2 HOH -»• Zn(OH)2+ + 2HC1+
Since Zn(OH)7 decomposes at 257°F,- this reaction 'appears possible within the
steam temperature range 212 to 257°F. Zinc hydroxide obtained from reaction
(1) might be heated-tzfo produce'ZnO at 257°F:
-------
CX-18)
(2) Zn(OH)2 ^ ZnO + HOHt
As an alternative approach, collected particulate might be treated with steam at
temperatures of 257 F and higher to produce ZnO directly, through the reaction.:
(3) ZnCl2 + HOH •* ZnCH + 2HClt
The HC1 vapor formed in reaction (1) or (3) might be dissolved in water and the
resulting acid reacted with additional, untreated, collected particulate to effect
the reaction:
(4) ZnO + 2HC1 ->• ZnCl2 + HOH
Settling, filtration, and/or evaporation might be applied to products to the
above reactions to remove water.
Waste heat from secondary zinc processing furnaces and/or afterburners might be
used as needed for this conceptual process, e.g. for the above-mentioned evaporation.
Thus, reactions (1), (2), and (3), by removing the chloride constituent from
collected particulate mixtures, afford possible means of producing a mixture
having high enough ZnO concentration and low enough chloride content to have
commercial use. Reaction (4), by converting ZnO in collected particulate to ZnCl ,
might provide a highly concentrated ZnCl. mixture that could be recycled as flux.
REMOVAL OF CHLORIDE FROM COLLECTED PARTICULATE BY MEANS OF AMMONIA AND STEAM
Collected particulate might be treated with NH_ and steam in attempts to effect
the reaction:
-------
(X-19)
(5) ZnCl2 + 2NH + HOH -»• 2NH.CH- + ZnO*
The NH_-steam mixture would have to be applied at a temperature higher than
212 F so that NH.C1 could be separated as a vapor from ZnO. (See Figure 5-1
and Perry p. 3-44)
The objective of this concept is to produce a mixture having high enough ZnO
concentration and low enough chloride content to be of commercial use. The
NH.C1 produced might be recycled as flux.
ZINC OXIDE ENRICHMENT OF COLLECTED PARTICULATE
Metallic zinc might be added to the material collected in scrubbers or precipitators
to react the zinc with ZnCl- and water as follows:
(6) ZnCl., + 2HOH -»• ZnfaH)^ + 2HC1 (§olution)
(7) 2HC1 + Zn -+• H t + ZnCl-
One possible source of metallic zinc for this use would be the unmeltable attachments
discharged from zinc sweat processes having residual crude zinc-alloy coatings.
These reactions indicate that as elemental zinc is added to a given quantity of
collected particulates, the amount of ZnCl_ in the mixture would remain constant
but the amount of zinc present in the -mixture as Zn (OH),, would increase.
The resulting mixture of collected particulate and products of the reactions
would be heated to evaporate water and decompose Zn(OH)0 to ZnO, at 257 F (Perry.
£
p. 3-22). Plant waste heat might be used for this evaporation. Hydrogen from
-------
(X-20)
Reaction (7) above might be utilized to produce heat or for other purposes
(including a conceptual smelting system shown below). If the ZnO content of
the collected particulate were enriched, the resulting material might be
commercially acceptable.
SMELTING OF COLLECTED PARTICULATE USING HYDROGEN
In this concept, the material collected in electrostatic precipitators would
be treated with H at high temperature to effect the reactions:
(8) H2 + ZnCl ^ 2HCH- + Zn+
(9) H2 + ZnO £ H20t + Zn+
Free energy relationships should be investigated to determine whether these
reactions appear promising.
CORRELATION OF CHARGED MATERIALS WITH EMISSION COMPOSITION
Relationships of charged materials with emission composition were shown in
Chapter IX (page IX-5). Equations or graphs developed from those relationships
would be useful for estimating the composition of emissions as percentages of
ZnCl? and ZnO, or chlorine as Cl and total zinc, using data on amounts of ZnCl
flux and zinc charged to a sweat-process heat. Emission compositions estimated
in this way might be used for further definition and refinement of process models
and for selecting emission control systems, on theoretical bases, without the
necessity of emission tests.
Developing these equations or graphs would depend on obtaining additional process
and emission data. As additional data are obtained, perhaps incidental to sub-
-------
(X-21)
sequent development work, equations and graphs correlating charged materials
with emission composition might be developed.
RECOMMENDATIONS FOR FURTHER RESEARCH AND DEVELOPMENT
The concepts developed here are based on indicated effects that are immediately
apparent from characteristics of secondary zinc process emissions, brought out
in this study. These concepts represent only a brief, initial treatment of those
emission characteristics and indicated effects. Mainly, they reveal principles
that can be utilized in developing emission control systems. Many variations
and combinations of the concepts, and principals brought out by them might be
used in the ultimate engineering of systems that will alleviate problems of
controlling secondary zinc industrial emissions.
A more rigorous treatment of potential chemical reactions and physical occur-
rences, involving the emissions, than that shown here should be performed for
the purpose of developing additional and/or alternative concepts that appear
feasible.
Engineers and other specialists of NASMI, APCO, other concerned individuals
and groups should be invited to recommend additional and/or alternative concepts.
The concepts, obtained as stated above, should be subjected to a screening study
that would consist of theoretical and laboratory investigations, including
consideration of technical and economic factors (the latter from Part 2 of
this problem definition study; see Chapter I). Based on findings of that screening
study, the most promising concepts should be selected and recommendations made for
development and in-plant tests of prototype and/or demonstration emission control
equipment.
-------
(A-l)
Appendix A
Derivation of Values for Process Models
Contents Paee No-
APPLIED FLUX A-l
EMISSION FACTORS A-2
COMPOSITION OF CHARGES AND RATES OF APPLYING PROCESS MATERIALS A-2
PARTICULATE EMISSION RATES A-3
COMPOSITION OF PARTICULATE EMISSIONS A-4
PARTICLE SIZE, AND PARTICLE SIZE DISTRIBUTION A-5
EFFLUENT GAS COMPOSITION A-6
EXHAUST EFFLUENT FLOW RATES AND TEMPERATURE A-6
APPLIED FLUX
Processing with ZnClj flux was assumed for all models. This condition seems
reasonably representative as indicated by general findings of this study
(see Table Vl-2 and VI-3) and by the snail amount of NH^"*" indicated in
analyses shown on page VI-5. Further, the fluxing action of ZnCl and NH^Cl
are essentially the same, as shown in Chapter IV.
-------
(A-2)
EMISSIONS FACTORS
Emission factors that were developed and presented in Chapter VI (page VI-4)
were tabulated in column 7 of Table VII-1, page.l. Process conditions corres-
ponding to those emission factors, as indicated by findings of Chapter VI,
were tabulated in the preceding columns.
COMPOSITION OF CHARGES AND RATES OF APPLYING PROCESS MATERIALS
A residual scrap mixture, containing residual flux and no organic material, was
assumed as making up the charge to Model A (Table VII-1, page 1, column 2).
Metallic scrap mixtures containing organic materials were assumed as making up
part of the charges to Models B, C, D, and E (Table VII-1, page 1, column 2).
For Model A, the rates of charging scrap material and of applying flux were
assumed as 1/4 ton/hour and 10 Ib./hour respectively (Table VII-1, page
1, columns 4-5). These assumptions agree with the data shown for the first
test tabulated .in Table VI-2 and correspond with the emission factor 24.5.
Model B is similar to Model A except that metallic scrap containing
organic material was assumed as being included in the residual scrap
charge. To arrive at a rate of applying flux to Model B, the rate for Model
A was reduced from 10 to 8 to account for dilution of the flux content of
the residual scrap charge by metallic scrap. The emission factor 24.5, as-
sumed for Model A, was also assumed applicable to Model B because car-
bonaceous emissions resulting from organic material in the charge would
tend to compensate (in quantity) for the reduction of chloride emissions due
to the reduced amount of flux in the charge.
-------
(A-3)
To establish the values representing process data for Model C (Table VII-1,
page 1, columns 1-5), the process data shown for the fifth test tabulated
in Table VI-2 were assumed. The published emission factor 10.8 (the last
item tabulated in Table VI-2) was chosen for this model in preference to
the emission factor 10.3, determined for the fifth test. This choice
was made to be consistent with the literature and because the two emission
factors are not significantly different, considering the general limited
precision of emission tests and the approximation of process data noted in
Table VI-2, column 4.
The mixture of the charge to Model D was assumed to be the same as that of
Model B. The flux application rate for Model D (16 Ib./hour was derived
by increasing the rate shown for Model B (8 Ib./honr) in proportion to the
higher rate of charging scrap material to Model D.
For Model E, the rate of charging scrap material was assumed as 1 ton/hour.
This value approximates the rate shown for the sixth item in Table VI-2.
Process conditions of Model E (shown in Table VII-1) correspond to the
emission factor 13, of the same sixth item. The rate of applying flux was
assumed as 1 Ib./hour to agree with the rate shown for processing similar
scrap material in the second processing condition tabulated in Table VI-3.
PARTICULATE EMISSION RATES
Particulate emission rates for all models were calculated from the tabulated
process weight rates and emission factors, as shown in Table VII-1, page 1,
column 8.
-------
(A-4)
COMPOSITION OF PARTICULATE EMISSIONS
A chemical analysis of particles emitted under process conditions similar
to those of Model A was determined in Chapter VI and shown on page VI-9.
This analysis was therefore tabulated for Model A in Table VII-1, page 1,
columns 9-12.
The organic materials in the charges to Models B and C (kettle furnaces)
were assumed to cause 10% carbonaceous content of particulate emissions
(Table VII-1, column 12). This percentage was suggested by results of the
particulate analysis shown on page VI-5.
The organic materials in the charges to Models D and E (reverberatory fur-
naces) were assumed to be completely incinerated. No emission data were
obtained during this study to support this assumption. It is based on the
experience of plant operators and on consideration of conditions in reberber-
atory furnaces that appear conducive to high combustion efficiency.
The amounts of ZnCl- and ZnO particulate emissions shown for Model B are
based on those shown for Model A, but the percentage of ZnCl in particulates
is reduced to account for the reduced content of ZnCl flux in the charge and
for the added content of carbonaceous particles to emissions. The amounts
of water of hydration and of deliquescence in emissions from Model B were
estimated to accord with the ZnCl content.
Since the scrap material charged to Model C and E is comprised of only me-
tallic scrap and the amounts of flux applied are therefore small compared
with the flux content of residual scrap, the amounts of emitted ZnCl. would
also be relatively small. Particulate emissions of 5% and 4% ZnCl were
-------
TA-5)
assumed representative of this condition for Models C and E respectively.
These percentages correspond to an assumption that about half of the applied
ZnCl flux is emitted as particulates. The balance of the particulate
emission constituents tabulated in Table VII-1 were calculated for these
models to accord with the ZnCl content, as shown in columns 9-12.
The composition shown for particulate emissions of Model D (columns 9-12)
is an approximation derived from the analysis of particulates emitted from
a combination of one reverberatory and several kettle furnaces shown on page
VI-8. The approximation was made by omitting the content of carbonaceous
material from the analysis (of page VI-8) and recalculating percentages of
ZnCl , ZnO etc. The percentages were weighted to increase the ratio of ZnO
to ZnCl« because conditions in reverberatory furnaces indicate a tendency
to higher ZnO emission rates, as compared with ZnCl . (See page VI-3.)
PARTICLE SIZE AND PARTICLE SIZE DISTRIBUTION
In tabulating particle size in Table VII-1, page 2, it was assumed that
of the emitted ZnCl and ZnO particles, those formed from vapor condensation
and oxidatiori approximate 1-micron size, and that agglomerates of those
particles reach sizes up to 10 micron. The balance of emitted particles
would then consist mostly of entrained ZnCl_ and ZnO particles, which would
be larger, and of carbonaceous particles of various sizes. Distribution of
the sizes of these particles was assumed as shown. These approximations
appear reasonable in view of findings of Chapter VI. Not noted in this
tabulation are the other chlorides and oxides, such as Aid , that would
be present in small amounts.
-------
(A-6)
EFFLUENT GAS COMPOSITION
The effluent gas of Model A is considered as having the composition of dry
air in accordance with findings of Chapter VI.
The effluent gas of Models B and C is also considered as having the compo-
sition of dry air except that 0.1% CO , 0.1% CO, and 0.1% HO content are
assumed, to account for combustion of part of the organic material in the
charge.
The range of effluent gas composition shown for Models D and E is the same
as that shown for reverberatory sweat process emissions in Table VI-1. It
is assumed that combustion of organic material in the charge is complete, and
that the amount of that material is not large enough to significantly affect
effluent gas composition.
EXHAUST EFFLUENT FLOW RATES AND TEMPERATURES.
Effluent flow factors for all models shown in Table VII-1, page 2, column 6
were obtained from Table VI-1. Effluent flow rates shown in column 7 were
calculated from those flow factors, as shown. Process bath temperatures shown
in Column 8 were obtained from Table VI-2. Exhaust effluent temperatures
shown in column 9 were obtained from pages VI-15 and 16.
-------
Appendix B
Calculation of Amounts of Compounds Composing
Particulate Emissions for Chapter VI
Calculation Sheet 1
Calculation Sheet 2
Calculation Sheet 3
-------
Appendix 3
Calculation Sheet 1 - Calculation of Amounts of Compounds Composing Particulate Emissions for Chapter VI
Basis: 100 grams (each) of Samples 1 and 2
Atomic and Molecular Weights
Calculation of Amounts,,o.f_ Compounds in. Basis Sample
Sam::.; 2
tm.
Atomic Wts.: N = 14; H = 1; Cl = 35.5
Molecular wt. of NH.C1: 14 + 4 + 35.5 = 53.5
4
Combined atomic wt. oE NH, : 14 + It = IE
4
0.47 g NH^ present in basis sample
NH,+ _ IS _ .47g
NH^Cl ~ 53.5 ~ wt. of IIH^CI in lOOg sample
wt. of NH Cl in lOOg sample • .4?R X 53.5
4 18
0.36fi present' in basis sample
18 -36g
53.5 wt. of NH.C1 in lOOg sample
wt. of NH Cl in lOOg sample = .36 X 53.5
4 18
Cl-
Atomic wt.: Cl = 35.5
8.93g Cl present in basis sample
8.32g Cl present in basis sample
Amount of Cl in NH.Cl in basis sample
= 1.40 - .47 = .93g
Amount of Cl- in 1IH.C1 in basis sample
= 1.07 - .36 = .71g
Zn
Atomic wt.: Zn = 65.4
Molecular wt of ZnCl = 65.4 4- 2 X 35.5
= 65.4 + 70 = 135.4
47.5g Zn present in basis sample
Amount of Cl in ZnCl2 = 8.93g - -93g = 8.0g
2 Cl _ 2 X 35.5 _ Amount of Cl in ZnCl in basis sample
2" 13
= 8.0E
44.5g Zn present in basis sample
Amount of Cl in ZnCl2 = 8.32 - .71 = 7.61g
2C1 2 X 35.5 Amount of Cl - - -
ZnCl_ 135.4 Amount of ZnCl- in basis sample
ZnCl2 135.4 Amount of Zn
= 7.61
Amount of ZnCl, -
Amount of ZnCl in basis sample = 8.0g X 135.4
2 X 35.5
= 15.3 g ZnCl.,
Amount of ZnCl - -
Amount of ZnCl in basis sample = 7.61 X 135.*
2 X 35.5
14. 5g ZnCl,.
Amount of Zn in ZnCl = 14.5 - 7.6 = 6.9g
Amount of Zn in ZnO = 44.5 - 6.9 = 37.6g
7,n _ 65.4 _ Amount of Zn in ZnO in basis sample
ZnO 81.4 Amount of ZnO in basis sample
= 37.6B
Atomic wt.: 0 = 16
Molecular wt. of ZnO = 65.4 + 16 = 81.4
Amount of 7.n in ZnClj = 15.3 - 8 = 7.3g
Amount of Zn in ZnO = 47.5 - 7.3 = 40.2g
Zn 65.4 _ Amount of Zn in ZnO in basis sample
ZnO 81.4 Amount of ZnO in basis sample
Amount of ZnO in basis sample
Amount of ZnO in basis sample = 40.2g X 81.4
65.4
= 50.Og ZnO
Amount of ZnO -
Amount of ZnO in basis sample = 37.6gX 81.4
65.4
-= 46.9g
-------
Appendix B
Calculation Sheet 2 - Calculation of Amounts of Compounds ^exposing Participate Emissions for Chapter VI
Basis: 100 grans (each) of Samples 1 and 2
Atomic and Molecular Weights
Calculation of Amounts of Compounds in Basis Sample
Al
Atomic wt.: Al = 27.0
Molecular wt. of Al^ = 2 X 27.0 + 3 X 16.0
= 5^.0 + 48.0 = 102.0
1.43 R Al present in basis sample
2 Al _ 2 X 27.0 = 1.43g
102
Amount of Al 0 in basis sample
0.54 g Al present in basis sample
2A1 2 X 27 0.54g
' —
Amount of Al 0 in basis sample = 1.43gX 102
2 X 27
102 Amount of Al 0 in basis sample
*- .* £. J
Amount of Al 0 in basis sample = .S4g X 102
2 X 27
• 1.02 g AJ.O.
Atomic wt.: Fe = 55.8
Molecular wt. of Fe^ = 2 X 5i.8 + 3 X 16.0
= 111.6 + 48.0 = 159.6
0.40gFe present in basis sanple
2 Fe _ 2 X 55.8 .40g
0.21g Fe present in basis sample
2 Fe 2 X 55.8 _ .21g
Fe 0 159.6 Amount or Fe 0 in basis sample
Amount of Fe-0 in basis sample = .40g X 159.6
2 X 55.8
= 0.57 g Fe203
Fe 03 159.6 Amount of Fe 0 in basis sample
Amount of Fe 0 in basis sample - .21g X 159.6
2 X 55.8
= O-JJOgFe-O,
Pb
Atomic wt.: Pb = 207.2
Molecular wt. of PbO = 207.2 +• 16
= 223
0.14g Pb present in basis sample
Pb _ 207 _ .14g
0.16 g Pb present in basis sample
PbO 223 Amount of PbO in basis sample
Amount of PbO in basis sample = .14g x,-7
= 0.15e PbO
Pb = 207 _ .16g
PbO 223 ~ Amount of PbO in basis sample
Amount of PbO in basis sample = .16g X 223
- 0.17g PbO
H..O - Water of hydration inZnCl.-4H.O
—
Molecular wt. of HO = 2 + 16 = i8
1 mol
18g
1 mol ZnCl2 = 135. '.g ZnCl2
mols ZnCl- in basis sample
= 15.3 g ZuCl., X 1 mol ZnCl,, _ 15.3 mol ZnCl,
135.4 g ZnCl 135.4 i
For every mol of ZnCl in basis sample, there would be 4 mo
H20 (water of hydration)
= 15.3 mol g-nCl X > mol I! 0 X 18g HO
135.4 J iuul Znci-2 mol II^0
= 8.1 g H;0
mols ZnCl in basis sample
14.5
X 1 mol ZnCl
135.4 s-grrC
14.5 mol ZnCl
135.4 *
Is of -.20.
HO (water of hydration)
= 14.5_ X 4 X 18
135.4
=7.7 g H;0
-------
Appendix B
Calculation Sheet 3 - Calculation of Amounts of Compounds Composing. Particulate
Emissions-for...Chapter VI - Basis lOOg of sample
Jj
17. 7g Cl present in basis s
48. 5g Zn present in basis sa
2 Cl 2 X 35.5 Amount of
ZnCI™ 135.4 Amount of
= 17. 7R
Amount of ZnCl_ in basis s
Amount of ZnCl0 in basis earn
= 33. 8g ZnCI..
Amount of Zn in ZnCI- =33.8
Amount of Zn in ZnO = 48.3 -
Amount of ZnO in basis sampl
H.O (water of hydration) = 3
. L 1
i pie
pie
Cl i
ZnCl0
£
:.
n pie
]e =
- 17.
16.1
= 3i
.8 )
f .4
n ZnCI
in ba
17. 7g
7 = 16
= 32.1
.ig x
4X1
i
-f
£.
1
11
' C
'f
in b
s sa
135.
X :
g (1
( ir
.4
.4
V
asis s<
nple
4
5.5
n basi
«
basis
40. Og
= 18g
m
7
s
Z
li
Jle
.7 X
71
samp
ampl
nO
2o
135.4
le)
e)
-------
Secondary Zinc and Aluminum Industries,
Emission Control Problem Definition Study
Notation
Used in Text
References
(page 1 of 2)
Publication
APEM
Mathewson
NASMI Studies
Minerals YB
NASMI NF-66
Nonferrous
Control Techniques
CEP. 3/67
AE. Nov. 196T
Perrj
Donoso
MC/6/70
Zinc in 1969
Air Pollution Engineering Manual (Public Health Service
Publication No. 999-AP-liO) 1967
Zinc. The Science and Technology of the Metal, its Alloys
and Compounds, ed. C. H. Mathewson, American Chemical
Society Monograph No. ll*2 (Reinhold) 1959
Studies of Dislocation Factors: No. II, The Secondary
Material Industries and Environmental Problems, ed^
S. Wakesberg (NASMI) 1968
Minerals Yearbook (Bureau of Mines)
Standard Classification for Non-Ferrous Scrap Metals,
Circular NF-66 (NASMI) 1966
The Nonferrous Scrap Metal Industry (NASMI) 1967
Control Techniques for Particulate Air Pollutants (NAPCA
Publication No. AP-51) 1969
Chemical Engineering Progress, March '67, "Removing Solid
and Mist Particles"by Morash, Krouse, and Vosseller,
pp 70-7lt.
Air Engineering, Nov. '67, "Zinc Oxide Emissions at Baltimore
Smelter Controlled by Dust Collection System", pp 18-20
Perry's Chemical Engineers' Handbook, ed. Perry, Chilton and
Kirkpatrick, Fourth Edition (McGraw-Hill) 1963.
NASMI's First Air Pollution Control Workshop. 1967, "Utiliza-
tion of Scrubbers and Precipitators" by J. Donoso.
Modern Casting, June '70, "Gas Cleaning for the Nonferrous
Foundry Industry" by E. W. Stenby, pp 52-1*.
Minerals Industry Surveys (U.S. Bureau of Mines) "Zinc in
1969".
-------
References (Continued)
(page 2 of 2)
-cion
ed in Text Publication
KSB "Performance of Commercially Available Equipment in
Scrubbing HCI gas" by Kempner, Seiler and Bowman, JAPCA
March 1970, pp 139-1^3.
GBP Evaluation of Granular Bed Devices, Final Report, Contract
No. PH 86-67-51, Phase III, AVATD-010T-69-RR, June 1969,
Avco Division, Lowell, Mass. (NAPCA).
AEMPC Atmospheric Emissions from the Manufacture of Portland
Cement by Kreichelt, Kemnitz and Cuffe (PHS) Ho. 999-AP-17.
1967.
------- |