EPA/600/7-86/016
April 1986
TECHNICAL MANUAL: HOOD SYSTEM CAPTURE
OF PROCESS FUGITIVE PARTICULATE EMISSIONS
by
Edward R. Kashdan
David W. Coy
James J. Spivey
Research Triangle Institute
RO. Box 12194
Research Triangle Park, NC 27709
and
Tony Cesta
Howard D. Goodfellow
Hatch Associates, Ltd.
21 St. Clair Avenue East
Toronto, Ontario M4T 1L9
Contract No. 68-02-3953
EPA Project Officer
Dale L. Harmon
Office of Environmental Engineering and Technology
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
AIR AND ENERGY ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC'27711
-------
TECHNICAL REPORT DATA
(Please read iHUfuctiu/is on the reverse
I. REPORT NO,
-^FEPA/600/7-86/016
2.
3. RECIPIENT'S ACCESSION NO,
TLE AND SUBTITLE
,'chnical Manual: Hood System Capture of Process
Fugitive Particulate Emissions
S. REPORT DATE
April 1986
6. PERFORMING ORGANIZATION CODE
7. AUTHOBIS)
Edward R, Kashdan, David W. Coy, James J. Spivey,
Tony Cesta*, and Howard D. Goodfellow
8. PERFORMING ORGANIZATION REPORT NO-
». PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P. O. Box 12194
Research Triangle Park, North Carolina 27700
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3953
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 9/83 - 9/85
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES £EERL project officer is Dale L. Harmon, Mail Drop 61, 919/541-
2429. (*) Cesta and Goodfellow are with Hatch Associates, Ltd,, 21 St. Clair Ave. ,
East, Toronto, M4TIL9, Ontario, Canada.
16. ABSTRACT
The manual provides to regulatory officials--charged with the responsi-
bility of reviewing hood systems for capture of process fugitive emissions—with a
reference guide on the design and evaluation of hood systems. Engineering analyses
of the most important hood types are presented. In particular, consideration is
given to design methods for local and remote capture of buoyant sources, and enclo-
sures for buoyant and intertial sources. A unique collection of case studies of actual
or representative hood systems has been included to provide insight into the evalua-
tion of existing systems or the design of a planned system.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFlERS/OPEN ENDED TERMS
C. COSATI J'lCld/GfOUp
Pollution
Processing
Leakage
Dust
Pollution Control
Stationary Sources
Fugitive Emissions
Particulate
Hood Systems
13 B
13 H
14G
11G
B. DISTRIBUTION STATEMENT
Release to Public
1«, SECURITY CLASS (THUKrporlt
Unclassified
21. NO. OF PAGES
133
20. SECURITY CLASS {This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-731
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publicatipn. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
ABSTRACT
Regulatory officials charged with the responsibility of reviewing hood
systems for capture of process fugitive emissions face a difficult task.
It is the purpose of this manual to provide these officials with a reference
guide on the design and evaluation of hood systems. Engineering analyses
of the most important hood types are presented. In particular, considera-
tion is given to design methods for local capture of buoyant sources,
remote capture of buoyant sources, and enclosures for buoyant and inertia!
sources. A unique collection of case studies of actual or representative
hood systems has been included to provide insight into the evaluation of
existing systems or design of a planned system.
This report covers a period from September 30, 1983, to November 30,
1984, and work was completed as of November 30, 1984.
iii
-------
CONTENTS
Section Page
Abstract ±±j;
Figures • . VI:L:L
Tables .............. x
Symbols. . ........ xi
Metric Equivalents .......... .... xiv
Acknowledgment . xv
1 INTRODUCTION AND SUMMARY .......... 1-1
1.1 Purpose of the Manual 1-2
1.2 Scope of the Manual . 1-2
1.3 Organization of the Manual . 1-3
2 LITERATURE REVIEW 2-1
2.1 Texts and Papers Concerning General Hood Design . . . 2-1
2.2 Papers Concerning Specific Aspects of Hood
Design 2-2
2.3 Bibliography of Industrial Ventilation 2-5
2.3.1 Industrial Ventilation: General 2-5
2.3.2 Hood Capture and Plume Theory. . 2-6
2.3.3 Natural Ventilation 2-7
2.3.4 Local Ventilation 2-7
2.3.5 Enclosures for Materials Handling
Operations ....... 2-8
3 HOOD SYSTEM CAPTURE OF PROCESS FUGITIVE EMISSIONS. .... 3-1
3.1 General Design Considerations for Hood Systems . . . 3-1
3.2 Assessment of Hooding Practices and Hood
Systems 3-7
4 DESIGN METHODS FOR LOCAL CAPTURE OF BUOYANT PLUMES .... 4-1
4.1 Design by Analytical Methods . 4-2
4.1.1 Receiving Hoods for Buoyant Sources 4-2
4.1.2 Exterior Hood (Side-draft) for Buoyant
Sources 4-5
4.1.3 Assisted Exterior Hoods for Buoyant
Sources 4-11
4.1.4 Experimental Confirmation of the
Design Equations/Performance Evaluation. . . . 4-14
4.2 Design of Hood Systems by Fluid Modeling 4-16
4.3 Design by Diagnosis/Measurement of an
Existing Hood System ....... .... 4-18
Preceding page blank
-------
CONTENTS (continued)
Section Page
5 DESIGN METHODS FOR REMOTE CAPTURE OF BUOYANT PLUMES ... 5-1
5.1 Design by Analytical Methods 5-2
5.1.1 Continuous Sources (No Obstructions,
No Cross-Drafts) ..... 5-4
5.1.2 Intermittent Sources ....... 5-9
5.1.3 Special Cases: Cross-Drafts and
Obstructions ..... ... 5-11
5.2 Design of Hood Systems by Fluid Modeling 5-13
5.3 Design by Diagnosis Measurement of an
Existing Site/Performance Evaluation. ........ 5-15
6 DESIGN METHODS FOR ENCLOSURES 6-1
6.1 Enclosures for Inertial Sources 6-1
6.1,1 Dust Generation in Inertial Sources 6-2
6.1,2 Exhausted Enclosures for Gravity
Transfer Operations . 6-4
6.1.3 Nonexhausted Enclosures . 6-6
6.1.4 Capture Performance 6-6
6.2 Enclosures for Buoyant Sources 6-7
6.2.1 Process and Layout Requirements . 6-8
6.2.2 Fume Capture 6-12
6.2.3 Mechanical Design 6-15
7 CASE STUDIES OF PROCESS FUGITIVE PARTICULATE
HOOD SYSTEMS ..... 7-1
7.1 Case I: Charging and Tapping Canopy Hood
for an Electric Arc Furnace 7-1
7.1.1 Source Description and Background 7-1
7.1,2 Design Approach for the Original
Greenfield Installation . . 7-5
7.1.3 Data Collection for System Modifications . . . 7-8
7.1.4 Design Approach for System Modification . . . 7-17
7.1.5 Design Summary 7-20
7.2 Case II: Air Curtain System for Copper
Converter Secondary Emission Capture 7-22
7.2.1 Source Description and Background 7-22
7.2.2 Design Approach 7-26
7.2.3 Performance 7-31
7.3 Case III: Basic Oxygen Furnace Secondary
Fume Capture ......... 7-36
7.3,1 Source Description and Background . 7-36
7.3.2 Design Approach ............... 7-37
7.3.3 Performance ........ 7-46
7.4 Case IV: Charging and Tapping Canopy Hood
for an Electric Arc Fu.rnace 7-51
7.4.1 Canopy Hood Design ...... 7-51
7.4.2 Hood Performance 7-53
-------
CONTENTS (continued)
Section Page
7.5 Case V: Dust Control for Clamshell Lime
Unloader Hopper . 7-54
7.5.1 Source Description and Background ...... 7-54
7,5.2 Design Approach 7-56
7.5.3 Performance 7-60
7.6 Case VI: Partial Enclosure to Control Aluminum
Rolling Mill Emissions. 7-60
7.6,1 Nature of Process Source and Hood
Selection 7-60
7.6.2 Design Procedure . 7-61
8 REFERENCES 8-1
vii
-------
FIGURES
Number ' Page
3-1 Summary of hood design process ,,.,.....,..., 3-2
3-2 Summary of general design principles .,,......., 3-5
4-1 Typical local receiving hood above vessel holding
a hot product ..... 4-3
4-2 Exterior hood (side-draft) for capture of plume
from buoyant source and analysis 4-6
4-3 Assisted exterior hood for buoyant source and
analysis 4-12
4-4 Use of design equations for predicting hood
performance and relationship to actual performance .... 4-15
4-5 Hood design data sheet local hood - receiving ...... 4-20
4-6 Hood design data sheet local hood - capture ....... 4-21
5-1 Typical shallow hopper type canopy hood (a) and pool
type canopy hood (b). Effective source-hood distance,
Z, is taken as the hood-source distance plus twice
the source diameter, D ..... 5-8
5-2 Hypothetical example of intermittent plume case.
Required hood storage volume depends on duration
of the plume surge 5-10
5-3 Average plume flow rate as a function of time
using anemometer technique at an electric
steel making furnace 5-16
5-4 Photographic scaling technique to analyze plume
velocity 5-18
5-5 Useful relationships between canopy hood performance
and rooftop opacity. In (a), actual performance is
found to lie between bounds of ideal and worst
hoods. In (b), amount of additional suction needed
to reach required opacity level can be estimated 5-20
5-6 Hood design data sheet remote hood - canopy 5-22
6-1 Mechanisms for dust generation and dispersion
during material fall in an enclosure 6-3
6-2 Schematic arrangement for BOF furnace enclosure 6-9
6-3 Enclosure for an electric arc furnace (EAF) 6-10
7-1 Original canopy hood system for control of process
fugitive emissions from an electric arc furnace ..... 7-4
7-2 Furnace tapping fume emissions ....... 7-17
7-3 Map of the plume boundaries relative to the
original hood system 7-11
viii
-------
Figures (continued)
Number Page
7-4 Observed and speculated plume flow rate during
charging ...,......,.....,., 7-12
7-5 Maximum and normal electric furnace charging and
tapping emission opacities . 7-16
7-6 Copper converter operations ............... 7-23
7-7 Converter air curtain control system . 7-25
7-8 Analysis of air curtain system ........ 7-27
7-9 Sulfur hexafluoride injection locations .... 7-33
7-10 BOF charging fume generation process and position
of local capture hood 7-38
7-11 SVS charging off-gas volume vs. heat size 7-42
7-12 SVS charging off-gas heat content vs. heat size 7-43
7-13 Charging emissions from a BOF furnace 7-47
7-14 Semi-enclosure capturing tapping, slagging, and
puffing emissions from a BOF furnace ..... 7-48
7-15 Fume hood arrangement for capture of BOF hot metal
reladling emissions ..... 7-49
7-16 Integrated secondary ventilation system for the BOF . , . 7-50
7-17 Canopy hood arrangement for capture of fugitive
emissions from the electric arc furnace 7-52
7-18 Three regions of lime drop flow patterns to be
modeled 7-55
7-19 Geometry of final configuration: baghouse flow
is drawn from back of hopper under single baffle,
which is raised off grizzly 7-59
7-20 Schematic cross section of an air-curtain hood.
Air jets prevent fumes from exfTitrating into
work areas surrounding mill 7-62
7-21 Example perimeter hood for control of aluminum
rolling mill emissions 7-63
IX
-------
TABLES
Number Page
1-1 Scope of the Technical Manual 1-4
3-1 Hooding Practices for Process Fugitive Emissions
in Various Industries 3-8
3-2 Selected Ventilation Systems for Process Fugitives
in Various Industries 3-19
4-1 Control Surfaces for Various Exterior Hood Types 4-9
5-1' Summary of Analytical Techniques for Canopy Hoods .... 5-3
5-2 Summary of Equations Governing Rise of Buoyant
Plume from a Hot Source 5-5
7-1 Overview of Case Study Selection .... 7-2
7-2 Design Summary 7-21
7-3 Summary of Hood Capture Performance 7-32
7-4 SVS System Exhaust Data 7-44
-------
SYMBOLS
Numbers in parentheses refer to sections. A few symbols have not been
included, but their meaning is given in the text.
A = Area, m2 (4-7).
Af = Hood face area, m2 (4.1.2).
A = Control surface area (4.1.2); cross-sectional area of the falling
stream, m2 (6.1).
A. = Jet nozzle area, m2 (4.1.3).
J
A = Plume cross-section at intersection of jet, m2 (4.1.3).
B = Width of metal strip being rolled, ft (7.6).
bn - Plume length scale at hood face, m (5.1.3).
C = Orifice discharge coefficient, dimensionless (4.1.1); metal coil
diameter, ft (7.6).
C = Heat capacity at constant pressure, ca1/gra-°C (4.1.1),
Cy = Hood source geometry constant, dimensionless (4.1.3).
D = Diameter of process fugitve particulate source, m (5.1.1); height
of bottom of hood above passline, ft (7.6).
d = Particle mass median diameter, m (6.1).
e = Eccentricity, m (5.1.3).
F = Buoyancy flux, m4/s3 (5.1.1).
G = Diameter of unobstructed plume at specified height above source,
ft (7,1.2).
g = Gravitational constant, m/sec2 (4-7).
H = Height of dropped material, m (6,1).
h = Thermal head of air, m (4.1.1).
K = Empirical factor, dimensionless (7,6).
-------
SYMBOLS (continued)
L = Distance from bottom of opening to location of orifice, m
(4.1.1); characteristic length (4.2); distance between rewind
reel and face of housing posts, ft (7.6).
M = Momentum flow rate, kg-m/s2 (4.1.2).
N = Number of slot widths, dimensionless (7.6).
OP = Opacity, dimensionless (4,5).
P = Source perimeter, ft (7.6).
Q = Volumetric flow rate, ms/s (4-6).
Qu = Plume volumetric flow rate at hood face, m3/s (5,1.1),
Q = Hood suction rate, m3/s (4,5).
Q. = Jet nozzle flow rate, m3/s/um"t slot length (4.1.3).
J
q = Rate of heat transfer, kcal/s (4.1.1).
q = Convectional rate of heat transfer, kcal/s (5.1.1).
q = Radiational rate of heat transfer, kcal/s (5.1.1).
R = Distance between jet and hood face, ft (7.2).
S = Model scale, dimensionless (4.2).
T = Absolute temperature of plume, K (4.1),
T. = Jet air temperature, K (4.1.3).
v
T = Air temperature in hood suction field, K (4.1.2).
&T = Temperature difference between hot body and ambient air (4,5),
t = Purge time of hood, s (5.1.2).
t, = Duration of plume surges, s (5.1.2).
*v = Plume centerline velocity, m/s (5.1.3),
CiX
V = Velocity, m/s (4-7).
-------
SYMBOLS (continued)
V. = Jet nozzle velocity, m/s (4.1.3).
-------
METRIC EQUIVALENTS
Nonmetric
°F/min
ton
1b
Btu/1b-°F
Btu/mln
cfm
ft
ft2
ft3
ft/min
in.
Times
5/9 (°F-32)
0.556
907
0,454
1.0
252
1.7
0.30
0.093
28.32
0.00508
2.54
Yields metric
°C/min
kg.
kg
cal/g-°C
cal/min
m3/hr
m
m2
L
m/s
cm
xiv
-------
ACKNOWLEDGMENT
The authors are indebted to Richard Jablin (Richard Jab!in and Associates)
and Manfred Bender (Bender Corp.) for their review of the manuscript. We
also thank Richard Perryman, John Conley, and Richard Roos (Busch Co.) for
providing the case study in Section 7.6. In addition, we appreciate the
guidance of our Project Officer, Dale Harmon.
-------
SECTION 1
INTRODUCTION AND SUMMARY
Process fugitive particulate emissions have been defined as "particu-
late matter which escapes from a defined process flow stream due to leakage,
materials charging/handling, inadequate operational control, lack of reason-
ably available control technology, transfer, or storage" (Jutze et a!.,
1977). Secondary hood systems consisting of enclosures, local hooding, or
remote hooding are the practical means of capturing process fugitive par-
ticulate emissions from many sources. Once captured, the gas stream con-
taining the particulate matter can be ducted to high-efficiency air pollu-
tion control devices. Frequently, the capture efficiency of the hood is
far less than the removal efficiency of the control device. Emissions
missed by the hood usually escape to the atmosphere.
Considering the diversity of sources classed as process fugitives, it
is not surprising that the design of secondary hood systems varies greatly;
a large range is found in size, exhaust rate, and arrangement. Regulatory
officials charged with the responsibility of reviewing hood systems for
either existing or planned sites face a difficult task. The behavior of
process fugitive particulate plumes is complex; as a result, the interac-
tion of the hood and plume is not always predictable. Moreover, most of
the traditional industrial ventilation texts do not specifically consider
process fugitive sources. The emphasis of these texts has been primarily
to provide designers with general design rules rather than with a thorough
understanding of hood design or the limitations of design methods. The
emphasis of this manual is on the design and evaluation of actual hood
systems used to control various fugitive particulate emission sources.
Engineering analyses of the most important hood types are presented which
provide a conceptual understanding of the design process: identifying
source parameters, calculation procedures, and techniques for evaluation of
hood performance. Some of the design techniques have baen introduced in
1-1
-------
technical papers by Hatch Associates and have been formalized into this
manual. Case studies of actual hood systems not only illustrate the appli-
cation of these design methods but also identify their limitations. Sev-
eral of the case studies are from the files of Hatch Associates and provide
unique insight into the diagnosis of an existing system.
1.1 PURPOSE OF THE MANUAL
The purpose of this technical manual is to provide regulatory offi-
cials with a reference guide on the design and evaluation of hood systems
to capture process fugitive particulate emissions. Much of the hood design
information is of necessity analytical, based on a mathematical or engineer-
ing approach. However, every effort has been made to explain the physical
processes in qualitative terms and to separate the formal equations.
1.2 SCOPE OF THE MANUAL
Although many names are used to type hood systems, hoods are most
conveniently classified in relation to the emission source that is con-
trolled. Three hood types may be distinguished: enclosures, exterior
hoods, and receiving hoods. Enclosures completely surround the source of
emissions. Obviously, from the standpoint of capture efficiency, enclosures
are the preferred method of control because escape of emissions is limited
to leaks through openings. However, enclosures are not always suitable,
especially in cases requiring ready access to the process source. Exterior
hoods (also referred to as perimeter and captor hoods) are so called because
they are exterior to the source. Exterior hoods function by inducing air
flow toward the suction opening. Because the "reach" of such hoods is
limited, exterior hoods are always local (i.e., close to the source).
Receiving hoods are intended to act as receptors to particulate plumes
that, by virtue of the process source, possess significant motion. Receiving
hoods may be local or remote from the source (a canopy hood is one kind of
receiving hood). An important special case is a hood system that uses air
jets to assist in the capture of particulate emissions. This design in
this manual is termed an "assisted exterior hood" (push-pull hood) because
the hood system (not the process) directs the motion of the particulate
plume.
1-2
-------
Sources of participate emissions may be classified as processes giving
rise to buoyant plumes, nonbuoyant plumes, and plumes having significant
particle inertia (a special case of nonbuoyant plumes). Sources giving
rise to buoyant plumes are hot (many are 1000° C or greater), and the initial
plume rise may reach a velocity on the order of 3 m/s, Nonbuoyant sources
are cold processes, or at least not very hot; for the nonbuoyant source,
the plume will not exhibit strong plume rise, and it 1s therefore likely to
be deflected easily by cross-drafts, even close to the source. Plumes with
significant particle inertia are generally nonbuoyant, but in addition, the
motion of the coarse particulate matter entrains additional air.
With the foregoing classification of hood types and processes, the
scope of the technical manual is summarized in Table 1-1. As shown in
Table 1-1, design of local hoods (exterior and receiving) for buoyant
sources is discussed in Section 4, design of remote hoods (receiving) for
buoyant sources in Section 5, and design of enclosures for buoyant and
inertia! sources in Section 6. Reference to the applicable case study is
also given in Table 1-1. Two situations not included in the technical
manual are exterior hoods for nonbuoyant sources and receiving hoods for
inertia! sources. Both these situations (the former typified by an open
surface tank, the latter by a grinding wheel) may be handled by industrial
ventilation guideline texts (e.g. ACGIH, 1976). In any case, neither is
generally considered a process fugitive source, and, therefore, they are
beyond the scope of this report.
1.3 ORGANIZATION OF THE MANUAL
This manual is divided into eight sections. In Section 1, the objec-
tives of this technical manual are discussed and the scope of the manual is
outlined. In Section 2, pertinent industrial ventilation literature is
summarized and a bibliography supplied. In Section 3, general design
methods are reviewed; hooding practices for many process fugitive sources
in various industries are tabulated. In Section 4, methods to design local
hoods for buoyant sources are presented, and a unique hood evaluation
questionnaire is given. In Section 5, methods to design receiving hoods
for buoyant sources are presented and another questionnaire is provided.
1-3
-------
TABLE 1-1. SCOPE OF THE TECHNICAL MANUAL
Hood type
Process fugitive source Design section Applicable case study
Exterior
Assisted
Unassisted
Assisted, unassisted
Receiving
Remote
Local
Local
Enclosures
Buoyant
Buoyant
Nonbuoyant
Buoyant
Buoyant
Nonbuoyant
(inertia!)
Buoyant
Nonbuoyant
(inertia!)
Nonbuoyant
4
4
Not discussed
5
4
Not discussed
6
6
Case II (Copper converter)
None
None
Cases I & IV (Electric arc furnaces)
Case III (Basic oxygen furnace)
None
Case V (Lime unloader)
Case VI (Aluminum rolling mill)
-------
In Section 6, design methods for enclosures for buoyant and nonbuoyant
sources are discussed. Section 7 presents analyses of six different hood
systems for capture of process fugitive participate emissions. The case
studies represent a wide range of source and hood types. Section 8 is the
references section.
1-5
-------
SECTION 2
LITERATURE REVIEW
2.1 TEXTS AND PAPERS CONCERNING GENERAL HOOD DESIGN
The following section provides a brief review of books and technical
papers dealing with design of hoods for industrial processes. The review
considers only major works. Section 2.3 provides a bibliography of signifi-
cant literature arranged by subject.
The most practical and thorough text on the subject of industrial
ventilation is P1 ant and Process Venti1ation (Hemeon, 1963). It discusses
the motion of airborne contaminants, principles of designing both local and
remote hoods, exhaust systems for carrying dusts, and dust collection. For
design of hoods, the text puts forth governing equations based on empirical
data and simplified theory. The intent of the text was to advance the
field of industrial ventilation from an essentially practical art based
only on experience to a more generalized science based on principles of
fluid flow and particle motion. The text is most valuable in providing a
conceptual basis for understanding the complex behavior of hood-source
interactions. Hemeon recognized the limitations of the design procedures,
and he never intended that the equations be applied without the benefit of
experience or judgment.
The AirPollution Engineering Manual (Danielson, 1967) discusses basic
principles of industrial ventilation extracted from Hemeon (1955). The
text attempts to provide a simple handbook. Illustrative problems demon-
strate calculation procedures. The validity of the equations from Hemeon
(1955) is not questioned, but arbitrary safety factors are recommended in
some cases.
Cheremisinoff (1976) briefly reviews and summarizes governing equations
for the design of hoods. Evidently, much is borrowed from Hemeon (1963)
and Danielson (1967). Illustrative problems demonstrate the calculation
procedures.
2-1
-------
Industrial Ventilation (American .Conference of Government Industrial
Hygienists, 1976) discusses general principles of ventilation, design of
hoods, exhaust system design, fan selection, and air cleaning devices. In
regard to the design of hoods, this manual provides rules-of-thumb for
required suction rates, positioning off-takes, control velocities, etc.
Specific hood designs for a number of processes are provided, but these are
limited to local exhaust of usually small sources. Buoyant plumes are not
discussed.
The Handbook of Ventilation for Contaminant Control (McDermott, 1976)
is intended primarily for use by industrial hygiem'sts as a practical text
accompanying the Industr_i_a 1 Venti 1 ation manual. Topics include OSHA stan-
dards, exhaust systems, hood selection, and fans. Hood design is limited
to local exhausts and enclosures for small sources.
Fundamentals of Industrial Ventilation (Baturin, 1972) is a very
different text. Translated from Russian, the text presents a phenomenological
view of industrial ventilation topics such as air jets, air curtains, and
suction openings. The treatment is theoretical with numerous references to
Russian authors. Practical applications are limited. The text is not a
design manual, and much effort would be needed to apply the theory to
actual hood design problems.
2.2 PAPERS CONCERNING SPECIFIC ASPECTS OF HOOD DESIGN
Several recent papers addressing certain aspects of hood design such
as remote capture of buoyant plumes, evaluation of hoods, enclosures for
materials handling operations, and computer-aided design are reviewed
below.
Remote capture of buoyant plumes is a common industrial ventilation
problem. From the preceding review, however, it is apparent that few
general texts deal with the problem. The procedure put forth by Hemeon
(1963) is based on empirical observations of air motion above a heated
wire. The heated wire observations provide a correlation equation to
estimate plume width as a function of height. Air entrained by the rising
plume is estimated from the convective heat loss from a hypothetical surface
having the same temperature and width as the source. This procedure does
not account for plume surges arising from intermittent fugitive particulate
2-2
-------
processes (such as charging of furnaces), nor does it account for building
cross-drafts and plume deflection around obstructions. Bender (1979)
attempted a much more rigorous approach invoking well-established plume-rise
theory (e.g., Morton et al., 1956; Morton, 1959). In Bender (1979), solutions
to the equations governing plume motion (conservation of mass, energy, and
momentum) are presented. Design considerations for canopy hoods are discussed.
In particular, spillage of plume from the hood can be avoided by providing
additional storage capacity; a baffle arrangement is suggested for that
purpose. Plume eccentricity arising from cross drafts is discussed and
requisite suction rates are recommended based on fluid modeling in a water
tank.
Evaluation of hood system performance is an important aspect of indus-
trial ventilation. To improve the performance of a working hood system, or
to judge the reasons for hood system failure, a proper diagnosis of the
hood system performance is essential. Several recent papers describe
various means for evaluating hood system performance. Hampl (1984) described
a tracer gas technique using sulfur hexafluoride. By injecting the tracer
at the process source at a known rate and measuring the quantity captured
by the hood system, a measure of hood efficiency is provided. This tracer
method was used by PEDCo (1983) in the evaluation of an air curtain system
over a copper converter. Sulfur hexafluoride was injected at four locations
above the copper converter to provide a measure of the capture efficiency
of the lateral draft hood. (See Section 7.2.) Another technique used by
Ellenbecker et al. (1983) employs a test aerosol consisting of an oil mist
injected through a diffuser. Capture efficiency is estimated as the ratio
of photometer response for the diffuser located at the process source to
the photometer response when the diffuser is placed near the hood.
More direct means of evaluating hood system performance for the actual
process source are desirable. For remote capture of buoyant plumes, estima-
tion of the plume flow rate at the hood face is critical in evaluating hood
system performance. Goodfellow and Bender (1980) describe three techniques
for estimating the plume flow rate: movie scaling, stopwatch clocking of
the plume, and anemometer measurements at the roof truss. The authors
report that, based on numerous field measurements, agreement between the
techniques may be expected, and, therefore, any of them may be used to
2-3
-------
estimate plume flow rate. A more sophisticated evaluative technique re-
viewed by Goodfellow and Bender (1980) consists of scale modeling hood
source interactions in a water tank or air system. Provided that the flow
in the scale model is turbulent and the Froude number of the model equals
the value of this dimensionless parameter for the actual hood system, scale
models permit convenient testing of hood designs or modifications to exist-
ing systems. Specific application of scale modeling of process fugitive
emissions from electric arc furnaces is provided by Bender et al. (1983).
Design of a low-level tapping hood and remote hood for charging emissions
is discussed. Fields et al. (1982) describe similar modeling of the cap-
ture of blast furnace emissions by low-level and remote hoods. For the
particular system studied, the remote hood had the most promising per-
formance.
The generation of dust during materials handling operations and attri-
tion processes is reviewed in detail by Hemeon (1963). Design procedures
for enclosing such operations are presented as well. For the case study in
Section 7.5, the generation and capture of dust dropped from a height is
pertinent and, therefore, is summarized here. According to Hemeon (1963),
when granular materials fall, each particle imparts momentum to the surround-
ing air. The macroscopic effect is an induced air stream. Exhaust systems
must take account of this induced air stream. Hemeon (1963) develops a
working equation for this induced air flow, namely, that it is proportional
to the cube root of the power generated by the falling stream of particles
(i.e., work done by the drag force over the distance fallen per unit time)
and the cube root of the stream area squared. Hemeon then presents theo-
retical equations for predicting the power generated by the falling stream
depending on the flow regime. For turbulent flow, Morrison (1971) claims
that Hemeon's equation overpredicts ventilation requirements by a factor of
three and, therefore, the constant of proportionality should be reduced
accordingly. Recently Dennis (1983) has reported on experiments with a
laboratory setup of a belt-to-chute transfer system. Based on these experi-
mental results, Dennis concludes that the induced air flow is roughly one
third of that predicted by Hemeon1s theoretical equation and further rec-
ommends a first power dependence on stream area rather than a two-third
power dependence.
2-4
-------
Heinsohn (1982) discusses the application of computational fluid
dynamics and computer graphics to the design of hood systems for nonbuoyant
sources. To understand the significance of this approach, it is necessary
to explain the traditional method of design. As some of the analysis in
Section 4 (exterior hoods for buoyant sources) and in Section 6 (enclos-
ures for inertial sources) use concepts invoked in the traditional design
methods, a review follows. The traditional method is based on empirical
determination of the suction field in front of the hood face. The suction
field is represented as contours of equal velocity ("isovels") that decrease
in magnitude rapidly as a function of distance from the hood. The required
exhaust rate for the hood then is selected so that the induced velocity at
the furthermost point of the emission source equals a nominal control or
capture velocity. Manuals (e.g., ACGIH, 1976) provide recommended values
for capture velocities for various sources. Although this traditional
method has been used extensively for many years, it does not predict capture
efficiency nor take into account effects such as cross-drafts. Computational
fluid dynamics offers the potential for more exact solutions; computer
graphics allows designers to conveniently observe the effects of modifying
hood designs or changing process conditions. To date, applications have
been limited, and as pointed out by Heinsohn (1982), buoyant sources represent
a fundamentally more complicated problem.
2.3 BIBLIOGRAPHY OF INDUSTRIAL VENTILATION
2.3.1 Industrial Ventilation: General
Alden, J. L. Design of Industrial Exhaust Systems. The Industrial Press,
New York, New York, 1948.
American Conference of Governmental Industrial Hygienists. Industrial
Ventilation, A Manual of Recommended Practices, 17th Edition. Edwards
•Brothers, Ann Arbor, Michigan, 1976.
Baturin, V. V. Fundamentals of Industrial Ventilation. Pergamon Press,
Ltd., Oxford, Great Britain, 1972.
Caplan, K. J. Ventilation Basics. Plant/Operations Progress. 1(3):194-
201, 1982.
Cheremisinoff, P. N. and Cheremisinoff, N. P. Calculating Air Flow Require-
ments for Fume Exhaust Hoods, Total and Partial Enclosures. Plant
Engineering. 30(4):111-114, 1976.
2-5
-------
Cheremisinoff, P. N. and Cheremisinoff, N. P. Calculating Air Flow Require-
ments for Fume Exhaust Hoods, Nonenclosure Types. Plant Engineering.
30(6):143-144, 1976.
Crawford, M. Air Pollution Control Theory. McGraw-Hill, Inc., New York,
New York, 1976. Pp. 165-187.
Danielson, J. A, (ed.) Air Pollution Engineering Manual. Los Angeles
County Air Pollution Control District, Los Angeles, California, Public
Health Service Report 999-AP-30, 1967. Pp. 25-86.
Goodfellow, H, D. and Bender, M. Design Consideration for Fume Hoods for
Process Plants. Am. Ind. Hyg. Assoc. J. 41:473-484, (July) 1980.
Goodfellow, H. D. and Smith, J. W. Industrial Ventilation—A Review and
Update. Am. Ind. Hyg. Assoc. J. 43:175-184, (March) 1982.
Hemeon, W. C. L. Plant and Process Ventilation. 2nd ed. Industrial
Press, Inc., New York, New York, 1963, (1st ed., 1955).
Heinsohn, R. J. CAD for Industrial Ventilation. Mechanical Engineering.
64-69, October 1982.
McDermott, H. J. Handbook of Ventilation for Contaminant Control, Ann
Arbor Science, Ann Arbor, Michigan, 1976.
2.3.2 Hood Capture and Plume Theory
Bender, M. and Baines, W. D. Operation of an Open Canopy Fume Hood in a
Crossflow. Journal of Fluids Engineering of the American Society of
Mechanical Engineers. (June)-.242-243, 1975.
Bender, M. Fume Hoods, Open Canopy Type—Their Ability to Capture Pollutants
in Various Environments, Am. Ind. Hyg. Assoc. J. 40:118-127, (February)
1979.
Bender, M., Cesta, T., and Minnick, K. L. Fluid Dynamic Modelling of Arc
Furnace Charging and Tapping Emissions. Presented at the EPA/AISI
Symposium on Iron and Steel Pollution Abatement Technology for 1983,
Chicago, Illinois, October 18-20, 1983.
Eisenbarth, M. Secondary Dust Collection Systems in Electric and Oxygen
Converter Steel Plants. Metallurgical Plant and Technology, 5:29-39,
1979.
Fields, S. ¥., Krishnakumar, C. K., and Koh, J. B. Modeling of Hood Control
of Blast Furnace Casting Emissions. In Proceedings: Symposium on
Iron and Steel Pollution Abatement Technology for 1981, EPA-600/9-82-021
(PB83164038), December 1982.
Flux, J. H. Progress in Secondary Fume Pollution Collection in Electric
Arc Steelmaking. Steel Times—Annual Review, 691-703, 1976.
2-6
-------
Goodfellow, H. D. Solving Fume Control and Ventilation Problems for an
Electric Heltshop. Presented at the 73rd Annual Meeting of the Air
Pollution Control Association, June 22-27, 1980.
Goodfellow, H. D. Solving Air Pollution Problems in the Metallurgical
Industry. Presented at the 7th International Clean Air Conference,
Adelaide, Australia, August 1981.
Marchand, D. Possible Improvements to Dust Collection in Electric Steel
Plants by Means of Hood Extraction. Quality of the Environment and
the Iron and Steel Industry. Pergamon Press, New York, 1974.
Morton, B. R.s Taylor, G., and Turner, J. S. Turbulent Gravitational
Convection from Maintained and Instantaneous Sources. Proc. Roy. Soc.
A. 234:1-23, January 24, 1956.
Morton, B. R. Forced Plumes. J. Fluid Mech. 5:151-163, 1959.
Roach, S. A. On the Role of Turbulent Diffusion in Ventilation. Ann.
Occup. Hyg. 24(1):105-132, 1981.
•Turner, J. S. Buoyancy Effects in Fluids. Cambridge University Press,
1973.
2.3.3 Natural Ventilation
Barton, J. J. Heating and Ventilating, Principles and Practice. George
Newnes, Ltd., London, Great Britain, 1964. p. 423.
Chrenko, F. A. (ed.) Bedford's Basic Principles of Ventilation and Heating.
H. K. Lewis and Company, Ltd., 1974. p. 255.
Kreichelt, T. E., Kern, G. R., and Higgins, F. B. Natural Ventilation in
Hot Process Buildings in the Steel Industry. Iron and Steel Engineer.
53:39-46, (December) 1976.
Natalizio, A. and Twigge-Molecay, C. Ventilation of Mill Buildings—New
Directions. Iron and Steel Engineer. (July):51-56, 1980.
2.3.4 Local Ventilation
Ellenbecker, M. J., Gempel, R. F., and Burgess, W. A. Capture Efficiency
of Local Exhaust Ventilation Systems, Am. Ind. Hyg. Assoc. J. 44(10):
752-755, 1983.
Fletcher, B. Centreline Velocity Characteristics of Rectangular Unflanged
Hoods and Slots Under Suction. Ann. Occup. Hyg. 20:141-146, 1977.
Fletcher, B. Effect of Flanges on the Velocity in Front of Exhaust Ventila-
tion Hoods. Ann. Occup. Hyg. 21:265-269, 1978.'
2-7
-------
Fletcher B. and Johnson, A, E. Velocity Profiles Around Hoods and Slots
and the Effects of an Adjacent Plane. Ann. Occup. Hyg, 25(4):365-372,
1982.
Harapl, V. Evaluation of Industrial Local Exhaust Hood Efficiency by a
Tracer Gas Technique. Am, Ind. Hyg. Assoc. J. 45(7):485-490, 1984.
Heriot, N. R. and Wilkinson, J. Laminar Flow Booths for the Control Of
Dust. Filtration and Separation. :159-164, (March/April) 1979.
Socha, G. E. Local Exhaust ventilation Principles. Am, Ind. Hyg. Assoc.
J. 40:1-10, (January) 1979.
2.3.5 Enclosures for MaterialsHandling Operations
Anderson, D. M, Dust Control Design by the Air Induction Technique. Ind.
Med. Surgery, 33:68-72, 1964,
Dennis, R, and Bubenick, D, V. Fugitive Emissions Control for Solid Mate-
rials Handling Operation. J. Air Pollu. Control Assoc. 33(12):1156-
1161, 1983.
Morrison, J. N. Controlling Dust Emissions at Belt Conveyor Transfer
Points. Trans. AIME. 150:68, 1971.
Wright, R. D. Design and Calculation of Exhaust Systems for Conveyor
Belts. Screens, and Crushers. J. Mine Vent. Soc. South Africa.
19(1):1-7, 1966.
2-8
-------
SECTION 3
HOOD SYSTEM CAPTURE OF PROCESS FUGITIVE EMISSIONS
3.1 GENERAL DESIGN CONSIDERATIONS FOR HOOD SYSTEMS
The job of the hood designer is often viewed as nothing more than
devising some convenient hood arrangement and estimating the required
exhaust rate. These steps are actually only intermediate in a well-con-
sidered design process as outlined in Figure 3-1. The starting place for
hood design is defining the design objectives clearly in quantitative
terms. Once the objectives are understood and agreed upon, a thorough
characterization of the process fugitive source must take place. This step
ought not to be cursory. Not only is a knowledge of the physical charac-
teristics of the plume necessary during both average and peak conditions,
but the process source should be examined in regard to measures to reduce
emissions, planned process changes, and concurrent processes in the plant.
Selection of a suitable type of hood follows. At last, design methods come
into play to provide hood dimensions and to estimate the exhaust rate
required to meet the design objectives. Implicit in application of the
design methods is evaluation of alternative hood arrangements and required
exhaust flow rates. Technical and economic evaluations are used for optimi-
zation in the hood capture system design process. But the designer also
should be charged with the responsibility of ensuring that the hood system
after installation is reliable and accepted by all personnel. In the
following section, design objectives for hood systems are reviewed. Con-
sideration is then given to characterizing the process source. Design
methods for hood systems are then presented and discussed in general terms.
Subsequent sections provide details of the techniques used in some of these
methods. As noted below, case studies in Section 7 illustrate the applica-
tion of the design methods.
Hood systems are designed for one or more objectives. Typically, the
objective may be to reduce workplace concentrations of contaminants, or to
3-1
-------
Design Objectives
Standards of
Performance
Source
Characterization
I
Hood Type
Selection
Design Methods
(a) Exhaust rate
(b) Hood arrangement
Optimization
Fabrication
Evaluation
(a) Satisfy objectives
(b) Worker acceptance
(c) Reliability
Figure 3-1. Summary of hood design process.
3-2
-------
reduce air pollution emissions, or perhaps to recover a product. In any
case, it is essential to quantify the objectives in terms of standards,
e.g., to meet a level of workplace exposure standard, or to reach an accept-
able opacity level for participate matter escaping through the roof vents,
or to achieve a desired level of visibility in the plant. These standards
then determine the expected performance of the hood system. Hood systems,
planned or existing (greenfield or retrofit), must be evaluated with refer-
ence to the design objectives.
Attention now turns to the emission source that needs to be controlled.
Consideration should first be given to the possibility of eliminating or
modifying the contaminant generation process itself. Even when hooding is
used, changes to the process could reduce the amount of contaminant gener-
ated or simplify the hood design problems by altering the way that contam-
inant is dispersed. By the same token, the possibility of future changes
in the process conditions must be considered as well. No hood design can
accommodate increases in emission volume flow rates far in excess of the
levels it was originally intended to control. Too frequently at this
point, due consideration is not given to concurrent processes and activ-
ities in the plant. Inasmuch as every hood system is affected by air flow
patterns within the building, the opening of bay doors or drafts from
various thermal processes can degrade hood performance.
As discussed in Section 1, process fugitive particulate sources may be
broadly classed as buoyant, nonbuoyant, or inertia?. Buoyant sources are
hot processes (many are 1000° C or greater), giving rise to plumes with
initial velocities on the order of 3 m/s. Nonbuoyant sources are cold
processes, or at least not very hot; for the nonbuoyant source, the plume
will not exhibit strong plume rise and is likely to be easily deflected
even close to the source. Inertia! sources are nonbuoyant, but, in addi-
tion, consist of high concentrations of coarse particulate matter. The
motion of the particulate matter entrains additional air and determines the
behavior of the inertia! plume.
Selection among the three hood types, enclosures, exterior hoods (also
referred to as perimeter or captor hoods), and receiving hoods, is limited
by the above source category. Other general factors limiting the selection
include: planned or existing site, access to the process, amount of clear
3-3
-------
space around the emission source, and constraints on operating costs and/or
fan capacity. A summary of the general principles for designing the dif-
ferent hood types is provided in Figure 3-2, From the standpoint of cap-
ture efficiency of the hood, enclosures are always preferable to local
hoods which, in turn, are preferable to remote hoods. For large-scale
inertia! sources (e.g., materials handling), enclosures are practically the
only choice, Nonbuoyant sources are most often controlled by local hoods.
Remote hoods may be used on buoyant sources and frequently are. When plan-
ning controls for a buoyant source, however, enclosures or local hooding
should be considered. A number of specific ventilation systems, as sum-
marized in the following section, illustrate the application of hoods to
the various source categories.
Characterization of the process fugitive particulate source is neces-
sary to design the hood type that is selected. Among the important source
parameters are
1. Continuous or intermittent plume
2. Plume flow rate
3. Plume geometry
4. Source heat flux
5. Source geometry
6. Physical/chemical characteristics of the particulate matter
(especially particulate concentration)
7. Gas composition
8. Gas temperature
9. Layout of the plant.
For existing sites, these source parameters may be measured directly. For
planned sites, values of the source parameters might be estimated, or data
from similar plants may be used. These aspects are discussed more fully in
Sections 4 through 6.
Design methods use the source parameters values above in various ways
to obtain the necessary exhaust rate and dimensions of the hood. Any one
3-4
-------
to
I
Ul
General Design Principles for Hood Systems
* Design Objectives
Principle: AM hoods must be designed to satisfy certain standards of
performance.
* Selection of Hood Type
Principle: The nature of the process fugitive emissions and access to the
process determine the selection of the hood.
• Source Characterization
Principle: A thorough knowledge of the process source parameters Is essential
to the successful design of a hood system.
* Exterior Hoods
Nonbuoyanf sources: Required exhaust rate is based on contour surface and
capture velocity.
Buoyant source: Required exhaust rate may be determined from momentum
considerations.
* Receiving Hoods
Inertial source: Best hood arrangement is such that hood opening coincides
with particle trajectory.
Buoyant source: Local capture requires knowledge of heat generation rate and
gas temperature. Remote capture requires estimation of the direction and
quantity of thermally induced air flow.
• Enclosures
Inertial source: Required exhaust rate is based on air flow induced by the
motion of the materials and consideration of dust-producing mechanisms.
Buoyant source: Design is based on considerations similar to local receiving
hoods.
Adapted from Hemeon, 1963, p. 87.
Figure 3-2. Summary of general design principles.
-------
of, or combination of, five different design methods may be used. In
increasing order of sophistication, they are
1. Design by precedent
2. Design by rule-of-thumb
3. Design by analytical methods
4. Design by diagnosis of an existing site
5. Design by physical scale model.
In design by precedent, a working hood system that performs satisfac-
torily is copied. Although this method is simple, it can be powerful in
producing a design that performs satisfactorily. However, in using this
method, working designs that use excessive exhaust, and are therefore over-
designed, may be copied. Failures using this method will be because the
copied system does not match the source parameters of the system under
design. This design method is illustrated by the case study in Section 7.3.
In design by rule-of-thumb} working systems are surveyed and the
elements common to most of them are put together to form a working design
rule(s). This method is straightforward, but the design rule(s) is likely
to oversimplify matters and may result in unacceptable performance. Alter-
natively, following the design rule(s) may result in a hood system that
performs satisfactorily but uses excessive exhaust. Case studies in Sec-
tions 7.4 and 7.6 illustrate the application of this method.
Design by analytical methods uses a mathematical model to predict hood
exhaust rate and dimensions from the source parameters. Examination of
hood systems is not necessarily part of this method. Sections 4, 5, and 6
summarize analytical methods for design of local hoods for buoyant sources,
receiving hoods for buoyant sources, and enclosures for buoyant and non-
buoyant inertia! sources. The case study in Section 7,2 illustrates this
method.
Design by diagnosis of an existing system is more specialized than the
other methods. An existing system usually is not performing satisfac-
torily. Extensive observations and measurements are made in an effort to
assess the hood design. Depending on the results, certain remedies may be
applied to the hood system or an entirely new design may be necessary. A
case study illustrating this design method is provided in Section 7.1.
3-6
-------
Lastly, design by physical scale model is the most sophisticated
method and may be applied to existing or planned sites. The hood design is
scaled hydrodynamically as a physical model using water or air as the test
media. Design by physical scale model is discussed in Sections 4 and 5,
and illustrated by a case study in Section 7.5.
As mentioned above, truly successful hood designs not only meet their
expected performance standards, but remain reliable. Often times, hoods
are placed in severe environments and are subject to extreme shocks, mech-
anical and thermal. Corrosion and erosion of the hood may also be factors.
Fabrication techniques and choice of materials for the hood system there-
fore must be carefully considered. Acceptance of the hood system by opera-
tion and maintenance personnel cannot be overemphasized for ensuring ulti-
mate reliability of any hood system.
3.2 ASSESSMENT OF HOODING PRACTICES AND HOOD SYSTEMS
Regulatory officials face the difficult task of assessing hood systems
for capture of process fugitive particulate emissions. To do this task
effectively, officials should be aware of hooding practices for various
fugitive particulate sources and should have knowledge of typical ventila-
tion systems. The following section summarizes hooding practices in various
industries, and examples of ventilation systems reported in the literature.
Table 3-1 is a compilation of hooding practices for process fugitive
emission in a variety of industries. The starting place for this table was
Jutze et al., (1977), although an attempt has been made to update this work
with more recent reports. Table 3-1 provides an extensive list of process
fugitive sources in a number of industries and a survey of hooding practices
for these sources. Hooding practices have been divided into local hooding,
remote hooding (canopy), enclosures, and building evacuation. Local hooding
is further subdivided into fixed, moveable, and side-draft hoods; the
latter means that the hood draft is lateral to the source. Within the
context of the definitions in Section 1, side-draft hoods are a class of
exterior hoods and "fixed hoods" may be either exterior or receiving hoods.
Building evacuation is beyond the scope of this report but has been included
in Table 3-1 as it represents a viable option for control of some sources.
A distinction between "typical control technique" and "used, but not typical
3-7
-------
TABLE 3-1. HOODING PRACTICES FOR PROCESS FUGITIVE EMISSIONS IN VARIOUS INDUSTRIES'
Industry
Local
Canopy
Building
Fixed Moveable Side-draft (high) Enclosure evacuation
Ironand steel
1. Sinter plant .
Sinter machine discharge
Sinter cooler
2. Blast furnace
Tap (iron)
Tap (slag)
3. Slag crushing
4. Open hearth furnace
Charge
Tap
J 5. Basic oxygen furnace
a Charge
Tap
6. Electric arc furnace
Charge
Tap
7. Cold scarfing
Hot scarfing
8. Hot metal transfer
Pig iron (reladling)
Hot metal desulfurization
(skimming)
b e
9. Teeming *
10. Continuous casting
X
X
x = Typical control technique.
-*• = In use (but not typical) control technique.
All hood practices are from EPA-450/3-77-010 unless
otherwise noted.
Engineering judgment.
:EPA-450/3-82-005a.
%PA-450/3-79-033.
^
"Leaded steels only.
-------
TABLE 3-1 (continued)
Industry
Local
Fixed Moveable Side-draft (high) Enclosure
Building
evacuation
OJ
I
11. Coke pushing
12. Cold rolling5
13. Hot strip millb
14. Materials handling11
15. Railroad car dumper
Iron foundries
1. Cupolas
Charge
Tap
2. Crucible furnace
Pouring
3. Electric arc furnace^
Charge
Tap
x
x
x
x
x
4.
5.
6.
7.
8.
9.
Electric Induction furnace
Reverberatory furnace
Ductile iron innoculation
Pot furnace
Pouring into molds
Casting shakeout
x
x +
x x
x +
x + +
x
x = Typical control technique.
+ = In use (but not typical) control technique.
Engineering judgment.
Primary off-take may be adequate.
9EPA-450/3-80-020a.
hBaldw1n and Westbrook 1982.
-------
TABLE 3-1 (continued)
Local
Industry
Canopy
Building
Fixed Moveable Side-draft (high) Enclosure evacuation
o
10. Cooling, cleaning castings x
11. Finishing castings x
12. Mold sand, binder receiving
13. Sand preparation +
14. Mold makingb +
Steel foundries
1. Electric induction furnace
Charge
Tap
2, Electric arc furnace
Charge
Tap
3. Open hearth furnace
Charge +
Tap +
4. Pouring in molds x
5. Cooling and cleaning castings +
6. Casting shakeout1^ x
X
X
X
X
x = Typical control technique.
+ = In use (but not typical) control technique.
Engineering judgment.
'EPA-450/3-81-005b and EPA-450/3-8u-02Qa.
(1976).
-------
TABLE 3-1 (continued)
Industry
Fixed Moveable Side-draft
Canopy Building
(high) Enclosure evacuation
U)
I
Primary copper smelting
1. Calcine transfer
2. Calcine discharge
3. Smelting furnace
Matte tapping
Slag skimming
4. Converter
Charge, skim, pour
Primary lead smelting
1. Mixing and pelletlzlng
2. Sinter discharge and screens
3, Blast furnace
Charge
Tap
4. Lead pouring, transfer
5. Slag pouring
6. Dross kettle
7. Lead casting
8. Sinter crushing
X
X
x = Typical control technique,
+ = In use (but not typical) control technique.
1
Engineering judgment,
%PA-450/3-83-009a.
EPA-450/3-83-018a.
-------
TABLE 3-1 (continued)
Local
Industry
Fixed Moveable Side-draft
Canopy
(high)
Enclosure
Building
evacuation
Pri mary z i nc sme 1 jti ng
1. Sinter machine windbox
2. Sinter machine discharge,
screens
3. Retort furnace
4. Zinc casting
5. Coke-sinter mixer
Primary aluminum smelting"1
1. Anode baking
^2. Electrolytic reduction cell
ro
3. Refining and casting
Secondary aluminum smejting
1, Sweating furnace
2. Reverberatory furnace
3. Crucible furnace
4. Induction furnace
5. Fluxing
6, Hot dross handling
x
x
x
X
X
X
X
+
+
+
x = Typical control technique,
+ = In use (but not typical) control technique.
Engineering judgment.
nEPA-450/2-78-049b.
-------
TABLE 3-1 (continued)
Local
Industry
Fixed Moveable Side-draft
Canopy Building
(high) Enclosure evacuation
Secgndary zj nc sme11i ng
1, Reverberatory sweat furnace
2. Kettle (pot) sweat furnace
3. Rotary sweat furnace
4. Muffle sweat furnace
5. Electric resistance sweat
furnace
6. Crucible melting furnace
7, Kettle (pot) melting furnace
Secondary lead smelting
1- Blast furnace
Slag tapping
Metal tapping
Charging
Access door
2. Mold fitting
3. Pot (kettle) furnace
Charge
Tap
Secondary copper smelting
1. Cupola
Charge
Tap
x = Typical control technique.
+ = In use (but not typical) control technique.
n
Coleraan and Vandervort 1980.
5EPA-450/3-80-011.
-------
TABLE 3-1 (continued)
Local
Industry
Fixed Moveable Side-draft
Canopy
(high) Enclosure
Building
evacuation
2. Converter
Charge
Discharge (molten copper)
3. Reverberatory furnace
Charge
Tap
Ferroalloy manufacture
1. Submerged arc furnace
Tap
2. Screening
3. Crushing/grinding
Nonmetal1 i c jni nerals"
1. Crusher
2. Grinder
3. Screens
4. Conveying transfer points
5. Product loading and bagging
Portlant Cement
1. Primary crusher
2. Vibrating screen
3, Secondary crusher
x
x
X
x = Typical control technique.
+ = In use (but not typical) control technique.
PEPA-450/3-82-014.
-------
TABLE 3-1 (continued)
Industry
Local
Fixed Moveable Side-draft
Canopy
(high)
Enclosure
Building
evacuation
CO
I
4. Cement loading
5. Cement packaging
Limestone manufacture
1. Primary crushing
2. Primary screening
3. Secondary crushing
4. Secondary screening
5. Quicklime screening
6. Loading
7. Packaging
Asphalt1c concrete
1, Cold aggregate elevator
2. Dried aggregate elevator
3. Screening hot aggregate
4. Hot aggregate elevator
x = Typical control technique.
+ = In use (but not typical) control technique.
-------
TABLE 3-1 REFERENCES
Jutze, G. A.» Zoller, J. M. , Janszen, T. A., Arnick, R. S., Zimmer, C. E., and
Gerstle, R. W. 1977, Technical Guidance for Control of Industrial Process
Fugitive Partlculate Emissions, EPA-450/3-77-010 (PB272288), March.
Revised Standards for Basic Oxygen Process Furnaces—Background Information
for Proposed Standards, 1982. EPA-450/3-82-OQ5a» December.
Review of Standards of Performance for Electric Arc Furnaces In Steel Industry.
1979. U.S. Environmental Protection Agency. EPA Report No, EPA-450/3-79-Q33
(PB80-154602) October.
"Electric Arc Furnaces in Ferrous Foundries—Background Information for Proposed
Standards. 1980. U.S. Environmental Protection Agency. EPA Report No. EPA-
450/3-80-02Qa (PB80-2Q2997) May.
Environmental Assessment of Melting, Innoculatlon, and Pouring. 1982.
Research Triangle Institute. Presented at the 86th AFS Casting Congress,
Chicago, Illinois, April 19-23.
Control Techniques for Particulate Emissions from Stationary Sources—Volume
2. 1982. U.S. Environmental Protection Agency. EPA Report No. EPA-45G/3-
81-Q05b, September..
^American Conference of Governmental Industrial Hygienists. 1976. Industrial
Ventilation, A Manual of Recommended Practices, 17th Edition. Edwards Brothers,
Ann Arbor, Michigan.
1^
Inorganic Arsenic Emissions from High-Arsenic Primary Copper Smelters-
Background Information for Proposed Standards. 1983. U.S. Environmental
Protection Agency. EPA Report No. EPA-450/3-83-009a, April.
Review of New Source Performance Standards for Primary Copper Smelters,
Chapters 1 through 9. 1983. U.S. Environmental Protection Agency. EPA
Report No. EPA-450/3-83-018a, November.
Primary Aluminum: Guidelines for Control of Fluoride Emissions from Existing
Primary Aluminum Plants. 1979. U.S. Environmental Protection Agency. EPA
Report No. EPA-450/2-78-049b, December.
Coleman, R. T. , and Vandervort, R. 1980. Demonstration of Fugitive Emission
Controls at a Secondary Lead Smelter. In: Proceedings of a World Symposium
on Metal and Environmental Control at AIME. Lead-Zinc-Tin, pp. 658-692.
Source Category Survey: Secondary Copper Smelting and Refining Industry.
1980. U.S. Environmental Protection Agency. EPA Report No. EPA-450/3-80-011
(PB80-192750), May.
3-16
-------
"Air Pollution Control Techniques for Non-Metallic Minerals Industry. 1982.
U.S. Environmental Protection Agency. EPA Report No. EPA-450/3-82-Q14,
August.
3-17
-------
technique" has been made throughout Table 3-1. This distinction should be
considered more a matter of opinion than fact. Moreover, these practices
should be viewed as evolving as industries develop new control techniques.
Table 3-2 is a summary of selected ventilation systems used for process
fugitive capture in several industries. Identified in this table are the
name of the plant, process fugitive source, brief description of the hood
design, exhaust rate, dimensions, capture efficiency, and associated par-
ticulate control device. Immediately obvious in Table 3-2 is the large
amount of missing information; unfortunately, the description of the design
of hood systems is frequently sketchy. Estimates of capture efficiency
often are not provided. Capture efficiency estimates that are in Table 3-2
invariably were made by trained observers reading opacity levels of escaping
emissions, usually from the shop roof vents, but sometimes emissions inside
the shop.
Regulatory officials assessing a particular hood system installation
are cautioned against generalizing from the information in Table 3-2.
Table 3-2 is intended to provide order-of-magnitude ventilation rates and
examples of hood arrangements. Scaling from these installations to a
particular hood system under scrutiny probably will not result in meaningful
comparisons (see Section 7.3 for an example). Regulatory officials facing
a difficult assessment task are encouraged to obtain as much detailed
characterization of the plant as possible (see Section 3.1). Comparison to
other hood systems can be made successfully if detailed information is
available and the systems are similar (again, see Section 7.3 for an
example).
3-18
-------
TABLE 3-2. SELECTED VENTILATION SYSTEMS FOR PROCESS FUGITIVES IN VARIOUS INDUSTRIES
Industry
Iron and steel
Sharon Steel Corporation
Crucible, Inc.8
Sidbec Melt Shopb
Knoxvllle Iron Company*"
oo
H-I Carpenter Steel (Reading, PA)d
«5
*
Stelco-McMaster Melt Shop8
Iscott (Trinidad)6
*
Chaparral Steel (Texas)
Process fugitive
source
Electric arc furnace
(2, 125 ton)
Charging
Tapping
Electric arc furnace
(2, 170 ton)
Charging
Tapping
Electric arc furnace
Tapping (fixed ladle)
Electric arc furnace
(2, 30 ton)
Charging
Tapping
Electric arc furnace
(20 ton steel /heat)
Charging
Tapping
Electric arc furnace
Electric arc furnace
Tapping
Electric arc furnace
Design
Canopy , two
sections
Dampened canopy;
partial furnace
enclosure
Moveable ladle
hood
Dampened canopy;
internal baffles
(275 fpn)
Enclosure; air
curtain across
roof slot
Canopy
Close hood
Canopy with
scavenger ducts
Ventilation
rate
17,600 nrVmln
17,600 «3/»in
17,100 etflMn
17,100 nrVmin
5,900 arVain
4,200 «V»1i»
5,100 ma/n1n
2,100 ms/min
15,600 ma/min
Size Capture Control
{hood face) efficiency device
15,2 nt x 13 m Reverse-air
15.2 HI x 14 • baghouse
14.6 m x 13.1 m Baghouse
14.6 m x 13.1 m
85% Baghouse
13.4 m x 7.3 m Baghouse
12.8 m x 15,5 B> 95-100% Baghouse
x 10.7 n
139 m2 Opacity; plume Baghouse
photography
11,1 m2 Fluid modeling Baghouse
3% maximum opacity Pulse- jet
baghouse
*Discussed in detail in Section 7.
aBrand (1981).
bHutten-Czapski in EPfi-600/9-81-017.
cBarkdo11 and Baker (1981).
Henninger et a!. (1984). Capture efficiency estimate
by teleeon from L. Geiser to M. Bender (1984).
eDetails available from Hatch Associates.
fTerry (1982).
-------
TftBlE 3-2 (continued)
Industry
Republic Steel (Chicago Works)*
Republic Steel (Cleveland
Works)9
*
Stelco-Led (Nanticoke)
Bethlehem Steel (New York)
Chiba Works (Kawasaki ,
Steel Corporation)
OJ
l\5
o
Mizushima Works (Kawasaki
Steel Corporation)''
Process fugitive
source
Q-BOPF
(225 ton steel /heat)
Charging
Tapping
BOPF
(250 ton steel/heat)
Charging
Tapping
BOPF (230 tonne)
Charging
Tapping
Continuous strip
galvanizing
q-BOPF (230 tonne)
Chargi ng
(takeoff)
Re ladling
Oesulfurization
Des lagging
Q-BOPF (180 tonne)1
Charging (escaping
furnace enclosure)
lapping
enclosure (chain
curtains)
Reladl tng
ring
Ventilation Size
Design rate (hood face)
Partial furnace
enclosure;
charging hood 9,400 cfarVmin
Partial furnace
enclosure; 10,100 dnrVmin
charging hood 9,100 dnrVmln
Local hood 10,000 ma/m!n (200° C) 13,9 m2
Movable enclosure 6,000 mVmfn (ISO0 C)
(reladllng)
850 «3/min 9.3 w2
Local (dampered) 18,000 nrVmin
Local (baffles)
Local
Booth
Part of furnace
enclosure (chain
curtains)
Part of furnace
Moveable close-fit
Capture
efficiency
<5% opacity
Ineffective
<5X opacity
2-9* opacity
95% effective
60-80%
95-98%
95-100%
85-95%
80-95%
50-55%
60-95%
Control
device
Venturi
scrubber
Venturi
scrubber
Baqhouse
Baghouse
Baghouse
Baghouse
Baqhouse
Baghouse
Baghouse
Baghouse
*D1scussed in detail In Section 7,
9Steiner and Kertcher in EPA-6QO/9-8Q-012 (1980).
h,
Bender et al. 1982.
j
Roof-mounted electrostatic precipltators provide
supplemental collection of process fugitives.
RTI trip reports (1979).
-------
TABLE 3-2 (continued)
Process fugitive
Industry source
Iron and steel (continued)
Mizushima Works (continued) Q-BOPF (250 tonne)
Charging
Tapping
Reladling
Kashima Steel Works, . OG furnace (250 tonne)
(Sumitomo Metals)''" Charging (escaping
enclosure)
Deslagging
Yawata Plant (Nippon Steel)'' BOPF (340 tonne)
Charging
Desulfurization
GO
i\j Oita Plant (Nippon Steel )J BOPF (340 tonne)
*""" Charging
Reladling
Deslagging
Swedish Steel J BOPF (145 tonne)
Charging
Desulphurization
Hot metal transfer
Ventilation Size
Design rate (hood face)
Part of furnace
enclosure
Part of furnace
enclosure
Side-draft hood;
supplemental canopy
Part of furnace
enclosure (chain
curtains)
Local
Part of furnace 7,100 n*/inin
enclosure (chain
curtains)
Local - close 850 ma/min per
torpedo car
Part of furnace 60 m3Mn
enclosure
Booth (metal poured
through slot in hood)
Booth 25 m3/min
Enclosure (doghouse) 9,200 nrVnin (at 70° C)
Local (baffles) 3,300 urVmin (at 70° C) 2 m x 2 m
Local side-draft 830 iB3/«1n (at 70° C)
Capture
efficiency
90-95%
SO- 75%
95%
50-75%
50-75%
100X
100X
95-100%
75-95%
75-80%
80-100%
95-100%
Control
device
Baghouse
Baghouse
flaghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baqhouse
Baghouse
j
RTI trip reports (1979).
Roof monitors are ducted to baghouse for supplemental process fugitive collection.
-------
TABLE 3-2 (continued)
Industry
Iron and steel (continued)
Ohgishima Plant''
(Nippon Kokan)
Italsider (ltaly)J'
British Steel Corporation
(Lackenby Works)
Titanium (Ilmenite) Smelting
Q1T, Sorele
Lime Manufacturing
Stelco-lede (Nanticoke)
Process fugitive
source
BOF (250 tonne)
Charging
Scrap
Hot metal
Tapping
Oes lagging
Reladl ing
BOF (350 tonne)
Charging
Hot metal transfer
Hot metal desolfurlzation
BOF (260 tonne)
Charging
Tapping
Scavenger (supple-
menta 1 )
Ladl ing
Dumping station
Design
Enclosure (chain
curtain
Enclosure (chain
curtain)
Booth
Annular hood (iron
poured through)
Semi -booth
Semi-booth (slot)
Close-fitting
local
Dampered local
local
Canopy (dampered
takeoffs)
Moveable hood
Enclosure
Ventilation
rate
10,000 wVmin
(at 480° C)
3,000 mVwin
(at 130° C)
1,800 fflVmin
(at 130° C)
2,700 ma/«in
4,500 irVniin
850 mVfliin
2,100 nVttiin
Size Capture
(hood face) efficiency
95-100%
50-75%
95%
60%
85-95%
2 m * 3 m 98%
98%
1.4 m x 6.1 m 50-75%
7.9 m x 1,8 m 80%
11.3 m x 80%
15,2 R
plume flow rates
measured
Control
device
Baqhouse
Bacjhouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Scrubber
Scrubber
Scrubber
Baghouse
JRTI trip reports (T979).
Details available from Hatch Associates.
-------
TABLE 3-2 (continued)
Industry
Secondary Lead
Test swelter
Primary Copper
Asarco-HaydenM
*
Asarco-Tacoma
Process fugitive
source
Blast furnace
Charging
Metal tapping
Slag tapping
Converter
Charging
Pouring
Converter
Charging
Pouring
Design
Local (hoist)
Local
Local
Secondary
retractable
hood
Enclosure with air
curtain
Ventilation Size Capture
rate (hood face) efficiency
340 mVmln
100 nrVmln
120 mVmin
0-10% opacity"
2,100-3,600 mVntin 75-95X
Control
device
Baghouse
Baghouse
Baghouse
Precipitators
Scrubbing towers
*Diseussed in detail in Section 7.
Coleman and Vandervort (1980).
mEPA-450/3-83-ul8a.
"Beskid and Edwards (1982),
°PEDCo (1983).
-------
TABLE 3-2 REFERENCES
Brand, P. G. A. 1981. Current Trends in Electric Furnace Emission Control.
Iron and_Steel Engineer. 58:59-64.
Hutten-Czapski, L. 1981. Efficient and Economical Dust Control for Electric
Arc Furnace. In Proceedings: Symposium on Iron and Steel Pollution Abatement
for 1980. EPA-600/9-81-017 (PB81-244-808), March.
cBaker, D. E., and Barkdoll, M. P, 1981. Retro-fitting Emission Controls on
the Electric Arc Furnace Facility at Knoxville Iron Company. Iron andSteel
Engineer. 58(8):45-50.
Henninger, J. L. , and Resh, Jr., D. P. 1984. Closing in On Arc Furnace
Emissions at Carpenter Technology. Iron and SteelEngineer. 61:26-30.
Terry, W. V. 1982. Site Visit--Chapparral Steel Corporation, Midlothian,
Texas, Electric Arc Furnaces in the Steel Industry. Letter to: Dale A.
Pahl, U.S. Environmental Protection Agency, Research Triangle Park, EPA
Contract No. 68-02-3059.
%teiner J, and Kertcher, L. F. 1980. Fugitive Particulate Emission Factors
for BOP Operations. In Proceedings: First Symposium on Iron and Steel
Pollution Abatement Technology (Chicago, IL, 10/30-11/1/79), EPA-600/9-80-012
(PB80176258), February, pp. 253-271.
hBender, M., Goodfellow, H. D., Schuldt, A. A., and Vanderzwaag, D. 1982.
BOF Secondary Fume Collection at Lake Erie. Iron and Steel Engineer. 59:
11-14.
JTrip Reports. 1979. Research Triangle Institute. Research Triangle Park,
N.C. Prepared for the U.S. Environmental Protection Agency, Hazardous Air
and Industrial Technology Branch. Contract No. 68-02-2651.
Coleman, R. T. and Vandervort, R. 1980. Demonstration of Fugitive Emission
Controls at a Secondary Lead Smelter. In: Proceedings of a World Symposium
on Metal and Environmental Control at AIME. Lead-Zinc-Tin, pp. 658-692.
Review of New Source Performance Standards for Primary Copper Smelters,
Chapters 1 through 9. 1983. U.S. Encironmental Protection Agency. EPA
Report No. EPA-450/3-83-Q18a, November.
Beskid, C. S., and Edwards, L. 0. 1982. Visible Emissions Converter Secondary
Hooding, Emission Test Report Asarco Hayden, Arizona, U.S. Environmental
Protection Agency, EMB Report 81-CUS-17, May.
PEDCo Environmental, Inc. 1983. Evaluation of an Air Curtain Hooding System
for a Primary Copper Converter, Volume 1, EPA-6QO/2-84-042a (PB 84160541).
3-24
-------
SECTION 4
DESIGN METHODS FOR LOCAL CAPTURE OF BUOYANT PLUMES
In reference to the outline of the hood design process In Section 3.1,
it is assumed that after consideration of the nature of the source and the
process operations, a local hood is a suitable choice. Attention then
turns to design methods for estimating the required exhaust rate and hood
dimensions from the parameters that characterize the source and emissions.
In the following section, three design methods are presented: design by
analytical techniques, design by fluid modeling, and design by diagnosis/
measurment of an existing site. These methods are discussed in general
terms below; the following sections then provide specific details or guid-
ance in the use of these methods.
The goal of the design methods is to arrive at a necessary exhaust
rate and the dimensions of the local hood. Although the three methods are
considered separately, they may overlap extensively. In design by analyt-
ical methods, conservation of mass, momentum, and energy equations are
applied to the source of emissions to estimate the plume flow rate arriving
at the hood face, and therefore the required exhaust rate. The values of
the source parameters used in the resulting design equations may be calcu-
lated or obtained directly as part of a field measurement program on an
existing site. In design by diagnosis of an existing site, measurements of
source parameters are obtained. Direct measurements of the plume flow
rate, and therefore the required hood exhaust rate, also may be obtained.
In such a case, it is wise to check the measured plume flow rate against
that predicted by the analytical techniques. For a planned site, field
measurements cannot be carried out, but fluid modeling techniques instead
of, or in addition to, analytical techniques may be used. If a facility
similar.to the planned facility exists, field measurements could be made in
the existing facility. In design by fluid modeling, a scale replica of the
4-1
-------
proposed hood is placed in a suitable fluid environment (e.g., water tank),
and the required hood exhaust rate is estimated by scaling up from the
performance of the model. For design of planned complex hoods, fluid
modeling is recommended. Moreover, fluid modeling may be used in conjunc-
tion with a field measurement program to diagnose causes of poor hood
performance or to test modifications to an existing hood system.
4.1 DESIGN BY ANALYTICAL METHODS
In this section, design equations for local hood capture of buoyant
plumes are presented. Hood types discussed are receiving hoods, exterior
hoods (side-draft), and assisted exterior hoods. Because of assumptions
employed in these analyses, the resulting design equations are simple and
straightforward. For the three different hood types, the following source
parameters are needed: source temperature, plume updraft velocity, and
plume area (geometry). Field measurements of an existing system Involve a
more extensive characterization (Section 4.3). As discussed at the end of
this section, exhaust rates estimated by these design equations are con-
servative.
4.1.1 Receiving Hoods for BuoyantSources
Figure 4-1 is a typical layout for a local receiving hood. The design
equations developed by applying the conservation of mass, energy, and
momentum follow. This treatment is similar to Hemeon (1963, 184-187).
First, it is seen in Figure 4-1 that the hot gas above the vessel
develops a thermal head because of the density difference between it and
the surrounding air. The required exhaust rate for capture of the hot gas
is estimated as the product of the updraft velocity of the gas due to the
thermal head and the total open area. The plume updraft velocity, V, is
estimated by the following equation:
V = C V(2g)(h) (4-1)
where
V = updraft velocity (m/s)
C = orifice discharge coefficient (dimensionless)
h = thermal head due to fluid density difference (IT, of air)
g = gravitational constant (9.98 m/s2).
4-2
-------
Local Receiving Hood
Off Take
Opening for Addition
of Product, Area A1
Clearance
Area, A2
Maximum Thermal
Head (L)
Vessel Containing
Hot Product
Source: GoodfeMow and Bender, 1980.
Reprinted with permission by American Industrial Hygiene Association Journal.
Figure 4-1. Typical local receiving hood above vessel holding a hot product.
4-3
-------
The thermal head due to fluid density difference is given by
h = ^"" (4-2)
u
where
L = distance from bottom of opening to the location of orifice (m)
AT = temperature difference between ambient air and gas inside enclo-
sure (°C)
T = absolute temperature of gas inside enclosure (K).
Substitution of Equation 4-2 into Equation 4-1 provides the following
expression for updraft velocity:
u
The temperature rise of the gas, AT, depends on the heat transfer rate
from the process (q) and the hood suction rate. Specifically, these are
related by the following equation:
where
q = rate of heat transfer from process (kcal/s)
Q = hood suction rate (nrVs)
p = gas density (kg/m3)
C = heat capacity at constant pressure (cal/gm-°C)
Assuming an air density of 1.2 kg/m3 and heat capacity of 0.24 cal/gm-°C,
Equation 4-4 reduces to the following:
A~T n
(4-5)
By substituting for AT in Equation (4-3) and using C = 0,6, which is
typical for a sharp-edged orifice, and Q = VA where A is the total open
area for the hood openings, the updraft velocity is expressed as follows:
4-4
-------
V = 2.9*M w?S • (4-6)
Since by continuity, Q = VA, the hood suction rate may be estimated
from Equation (4-6) by multiplying both sides of the equation by the total
open area, A (which equals A! plus A2 in Figure 4-1):
Equations (4-4) and (4-7) can be used to calculate the required
exhaust flow rate. The maximum heat transfer rate should be used in Equa-
tion 4-7 and can be based on actual field measurements as described in
Section 4,3 or calculated from a knowledge of the physical/chemical parameters
of the process. Using a graphical technique on log-log paper or a simple
iterative computer program, the above two equations can be solved to estab-
lish the minimum exhaust flow rates required for different hood geometry
and hood openings.
It is instructive to examine Equations (4-4) and (4-7). The terms A
and L are hood-geometry terms, whereas the terms q and T are process
variables. The latter therefore will generally be known or estimated with
less certainty. However, the cube-root dependence in Equation (4-7) implies
that errors in estimating these terms will not have a great effect on the
exhaust rate estimate.
4.1.2 Exterior Hood (Side-draft)for Buoyant Sources
Exterior hoods function by inducing air flow toward the suction opening.
The common exterior hood arrangement shown in Figure 4-2a is a side-draft
hood providing exhaust for a hot process. A receiving hood as discussed
above is clearly preferable to an exterior hood that must overcome the
thermal head (Equation (4-2)) of the plume. An exterior hood, however,
might be selected if complete access to the top of the source was necessary
(e.g., pouring metal into molds). Assisted exterior hoods, i.e., those
using air jets to direct the plume, are discussed in the next section.
4-5
-------
Source
Exhaust Hood
(a)
Source:ACGiH, 1078. (Reproduced with permission.)
a = tan'1
Mu = momentum of plume updraft
Mr = momentum of resultant
Ms = momentum of hood suction field
(b)
Figure 4-2. Exterior hood (side-draft) for capture of plume
from buoyant source and analysis.
4-6
-------
The following design method for control of buoyant sources by exterior
hoods is based on momentum considerations. This particular method, not
presented previously, is introduced by Hatch Associates. Hemeon (1963,
181-182) provides only a sketchy analysis of this hood arrangement.
The analysis of exterior hoods for buoyant sources is based on vector
addition of the momentum induced by the hood suction field and the momentum
of the plume. Momentum flow rate (momentum per unit time), first is defined
by the equation
M = (V2)(A)(p) (4-8)
where
M = momentum flow rate (m • kg/s2)
A = area (m2)
V = average velocity (m/s)
p = air density of stream (kg/m3).
Note that momentum flow rate is equivalent to the force of the jet.
In reference to Figure 4-2a, for complete capture by the exterior
hood, contaminant arising from the farthest point of the source must follow
a trajectory reaching the top of the hood at angle <* from the source. A
momentum diagram of this idea is shown in Figure 4-2b. From this diagram,
it is seen that
Ms '
where
M = Momentum of hood suction field
M = Momentum of plume updraft
M = Resultant momentum (vector addition)
« = tan"1 (Y/X)
X = source width
Y = distance between top of hood and source.
4-7
-------
The momentum flux (momentum flow rate per unit area) is assumed to follow
Equation (4-9). Specifically, with similar notation, it follows that
M M v
F = F (
5 U
where
A = control surface area (m2)
A = plume cross-sectional area (m2).
Equation (4-8) applied to the hood is written as
which upon rearranging becomes
Substituting for (M /A ) from Equation (4-10), the suction velocity may
*3 £>
be written as
(4-11)
A working design equation then is obtained from Equation (4-11) by
invoking continuity (conservation of mass) and the concept of velocity
contours (Section 2,2). Recalling that an exterior hood functions by
inducing air flow toward the suction opening, the velocity field in front
of a hood may be represented as a series of lines of equal velocity (isovels)
expressed as a function of the distance, x, taken from a direction normal
to the plane of the hood face. The velocity field has been determined
experimentally for various hood shapes as summarized in Table 4-1. To
complete this analysis, consider the simplest case, a plane unflanged hood.
The hood suction velocity is assumed to be uniform across the control
surface, A , given by the following expression:
A = 10 x2 + A.
4-8
-------
TABLE 4-1. CONTROL SURFACES FOR VARIOUS EXTERIOR HOOD TYPES3
. Aspect ratio
Hood type (width/length) Control surface
Plain opening 0.2 or greater A = 10 x2 + A~
Flanged plain opening 0.2 or greater A = 0.75 (10 x2 + A.)
Slot 0.2 or less A = 3.7 Lx
S[L = slot length]
Flanged slot 0.2 or less A = 2.8 Lx
[L = slot length]
aAdapted from ACGIH (1976, p. 4-4).
For half hoods or slots, i.e., those with a bottom edge close to the source,
control surface is one-half of the formulas.
4-9
-------
where
A = control surface, m2
A,. = hood area, mz
x = distance from hood face, m (0 5 x £ X).
The required hood suction rate follows from Equation (4-11). By the
familiar continuity equation, Q = VA, the exhaust rate required to effect
control at a distance x = X, Q , is given by:
S U
Applying Equation (4-8) to the plume momentum flow rate, it follows that
M = (V 2)(A )(p ) ;
U U lTKhu '
noting that
Pu_Ts
where T and T are the absolute temperatures of the suction and updraft
gas streams. Substituting for (M /A ), the required exhaust rate becomes
f ) (10 X2
, , , (10 X2 + A ) . (4-12)
u ^ ^
Some observations and recommendations follow in the use of Equation
(4-12). The plume velocity, V , may be estimated or measured. If measured,
then, because of large velocity gradients close to the source, the velocity
should be measured either at an elevation of one-half the source diameter
or at the hood center line elevation, whichever is greater. In designing a
hood for a planned site, the hood face area, Af, may be taken as equal to
4-10
-------
the source area as a starting value for calculations. The final hood
dimensions and shape may be limited by available space. Hood face velocity
should not exceed 30 m/s to avoid excessive noise, hood erosion, and energy
consumption. The suction temperature T can be assumed to be the ambient
temperature. Equation (4-12) may be adapted for other hood types according
to the formulas in Table 4-1.
4.1.3 Assisted Exterior Hoods for Buoyant Sources
The use of air jets in hood designs is not a new concept. The following
section is concerned with the use of air jets to direct a buoyant plume
into an exterior hood arranged laterally to the source. The topic of air
jets was examined theoretically by Baturin (1972), practically by Hemeon
(1963), and most recently, in an excellent report by Yung et a!., (1981).
Some preliminary concepts and definitions need to be addressed first.
A series of air jets, or continuous blowing slot, is called an "air curtain,"
and for this particular application, an air curtain, not a single jet,
would be necessary to direct a buoyant plume. The term air curtain frequently
provokes the misconception that the air jets create a semi-solid barrier
that the fugitive particulate matter cannot cross. This notion is false
(Hemeon, 1963). An air curtain acts by entraining surrounding air. When
used above a buoyant source, the air curtain will entrain the contaminated
air, resulting in a calculable concentration of particulate matter in the
air curtain. If the exterior hood, which is arranged to act as a receiving
hood for the air curtain, fails to provide either adequate exhaust or face
area to accommodate the flow rate and width of the curtain at the hood
face, then the contaminant will not be captured.
An idealized, assisted exterior hood arrangement is shown in Figure 4-3a.
As in the previous analysis, the design equations are derived from momentum
considerations. This analysis is restricted to "jet throw distances"
(equivalent to source width in this arrangement) less than six slot lengths,
which can be taken as a practical limit. Beyond that distance, the shape
of the jet becomes circular and is of limited use for hood capture. The
applicable momentum diagram is shown in Figure 4-3b. First, it should be
recognized that momentum is conserved at every section away from the jet so
4-11
-------
Entrainment Angle
Air Curtain
\
\
Exhaust Hood
(a)
Source: ACGIH, 1976. (Reproduced with permission.)
Jet Side
Mu
MI = jet momentum
Mu = updraft momentum
Mr = resultant momentum
0 & /? = deflection angles
X Cos 0 Tan
Mj X Sin 0
Exhaust Side
(b)
Figure 4-3, Assisted exterior hood for buoyant source and analysis.
4-12
-------
that the total rate of air flow in the stream in relation to the primary
flow from the nozzle is found by applying Equation (4-8) to two cross-sections:
(P0XV>(A0) = (PX)(V/)(AX) (4-13)
where
(p )(V 2)(A ) = momentum flow rate at the nozzle
(p )(V 2)(A ) = momentum flow rate at distance x from the nozzle.
Considering the geometry shown in Figure 4-3b, the required jet nozzle
velocity may be found from the following equation:
where
V = average plume updraft velocity (m/s)
A = plume cross-section at intersection with jet (m2)
A. = jet nozzle area (m2)
W
T. = jet air temperature (K)
J
T = average plume air temperature (K)
Cy = (cos 8 x tan p) + sin 8.
The entrained air volume at the exhaust hood is calculated by using the
governing equation for a continuous slot or, equivalently, line jet (Bender,
1979):
Q = 0.88 ^(Q.)(V-)(X) (mVs/unit length of slot) (4-15)
v vl
where
Q = hood suction rate (m3/s)
Q. = jet nozzle flow rate (m3/s/unit length of slot)
*J
V. = jet nozzle velocity (m/s)
%j
X = entrainment distance (m).
4-13
-------
The entrainment distance is usually taken as the distance between the
nozzle and exterior hood. The entrainment angle of the jet, which defines
the boundaries of the jet, has been found experimentally to have an approxi-
mate value of 24 degrees (Bender, 1979). From Figure 4-3a, it is seen that
the minimum hood height is the entrainment distance multiplied by the
tangent of the entrainment angle.
Application of the above design equations is presented in the case
study in Section 7.2. Recommended practices in the use of these equations
are as follows. Because of the vector addition of forces, if the jet
nozzle is directed horizontally, the resultant force always will be above
the nozzle elevation. Consequently, it is recommended that the nozzle be
pointed downward at an angle of 15 to 25 degrees from the horizontal. Air
jet velocities at the nozzle should not exceed 30 m/s to avoid excessive
noise or energy consumption. Finally, the interaction of the air curtain
and the exhaust hood can be complex, especially if the velocities of the
suction field of the hood are of the same magnitude as the jet velocity
values in the vicinity of the hood. Assisted exterior hood designs'there-
fore often require considerable adjustments to the nozzle angles and slot
widths to achieve acceptable performance.
4.1.4 Experimental Confirmation of the Design Equations/Performance
Evaluation
The preceding analytical techniques always should be ysed with judgment
and, if possible, experience. The design equations must not be considered
as providing totally accurate predictions. The reason for this caution is
partly because the theory is simplified to one-dimensional flow. But even
if more sophisticated mathematical modeling was performed, serious limita-
tions still would exist because all analyses of this type are predicated on
an idealized model of the actual hood system. The idealized mathematical
model provides only a limited or incomplete description of the actual
hood-source interaction. However, some experimental confirmation of the
validity of the design equations is afforded by fluid modeling (see also
Section 4.2).
The design equations for the required hood exhaust rate give the
required hood exhaust flow rate to achieve 100 percent capture efficiency.
4-14
-------
100
.
o
LU
£
CO
O
•o
o
o
X
20-
Actual Hood
Performance
Linear Relationship
for Hood
1
8
for Hood
(Theoretical)
9 10
2345 67
Hood Suction, X1000 m3/min
Adapted from Bender et al., 1983.
Figure 4-4. Use of design equations for predicting hood performance
and relationship to actual performance.
4-15
-------
On a plot of hood efficiency against hood exhaust rate, as shown in Figure
4-4, this operating point is depicted as QKJO- On a flow rate basis,
estimates of the hood efficiency at lower exhaust rates may be made by
connecting with a straight line the point QKJQ (100 percent) and the origin
(assuming a linear relationship between hood efficiency and exhaust rate).
Figure 4-4 also shows the actual hood performance as determined by fluid
modeling studies. Actual hood performance is generally found to be concave
downward so that the assumed linear relationship provides a conservative
estimate of hood capture efficiency. Therefore, for a given operating
exhaust rate of a hood system, the linear relationship in Figure 4-4 may be
safely used to predict improvements in hood performance by increases in
exhaust rate.
4.2 DESIGN OF HOOD SYSTEMS BY FLUID MODELING
The general theory behind the use of scaled models to represent the
flow behavior of a full-scale prototype (in this case, hood system) is
clearly beyond the scope of this report. Therefore, only an outline follows
of the approach used in fluid modeling of hood systems.
Hood systems are typically modeled in a water tank using salt solution
to represent the buoyant motion of the plume (e.g., Goodfellow and Bender,
1980). By establishing dynamic similarity between the test model and the
prototype (hood), data measured in the model flow may be related quantita-
tively to the prototype flow. Two conditions are necessary to establish
dynamic similarity:
1. Exact geometric similarity, which requires that the linear
dimensions of the model are in the same proportion as the
corresponding dimensions of the prototype.
2. Kinematic similarity, which requires that the flow regimes
be the same for model and prototype.
Kinematic similarity is achieved by matching governing dimensionless groups
which describe the flow regime. For modeling hood systems, the governing
dimensionless groups are the Reynold's number and the Froude number. But
because almost all industrial operations involve very turbulent flow, for
which there is little Reynold's number dependence, the Reynold's number
criterion can be achieved simply by ensuring that the flow in the model is
4-16
-------
turbulent. For processes involving hot gases (i.e. buoyancy driving forces),
the Froude number similarity criterion yields the required prototype exhaust
rate as follows.
Froude (Model) = Froude (Prototype)
LV2 \ / v2
(4-16)
.--, - H-5- -
m ^ / p
V2 L4 L4 p V2 L4 L4 p
mm p *m _ p m p Kp
L (p - p ) L (p - p
m oiti m p op p
o. (pQP " PP} = ss Tm (TP * TQP)
m Pop ^om Pm p m om
P T
with q = (QKpXC ) (T - T ) and -£ = =2, then
M pm p
Q 3 /q
.!£_= csf J£
Q7 \q
xm \Hm
where
o = ambient conditions
L = representative dimension
C = specific heat at constant pressure
S = the model scale (= 10 for 1:10 scale model)
Q = representative volume flow rate
T = representative hot gas temperature
p = gas density
q = heat transfer rate.
4-17
-------
The required prototype flow rate at the hood off-take (subscript 1) follows
1/3
Important observations can be made concerning the use of Equation 4-17.
First, the estimated exhaust rate for the prototype varies directly with
the model flow rate. Second, the prototype exhaust rate has a strong
dependence (5/3 power) on the model scale. Both these parameters, however,
may be measured with accuracy. Third, the prototype exhaust rate does not
have a strong dependence (1/3 power) on the heat flow rates which are the
most difficult to determine. In general, fluid modeling of hood systems
offers the potential to take account of factors difficult to handle by
analytical techniques (e.g., building cross-drafts) and further, to do
convenient evaluation of hood design modifications.
4.3 DESIGN BY DIAGNOSIS/MEASUREMENT OF AN EXISTING HOOD SYSTEM
Frequently, hood systems used to capture process fugitive participate
emissions are judged to be performing unsatisfactorily. Sometimes, new
stricter standards are being enforced—standards that may far exceed the
original design objectives of the system. At other times, the original
design basis of the hood system was faulty or too limited, and remedial
measures were never taken. Also possible are changes in process conditions
since the original design was conceived, or perhaps the initial charac-
terization of the source was in error. For any of these reasons, a field
measurement program of the performance of the hood system may be carried
out. Because such measurement programs are very site-specific, only general
guidance is provided here. Unique questionnaires have been included to
summarize information obtained from a field measurement program. The
questionnaire may also be used for a planned facility with measurements
being obtained in a plant similar to that being planned.
A field measurement program should begin with characterization of the
source which, of course, is also a crucial step in the design of a system
for a planned site (Section 3.1). Source sampling should include measure-
ments of gas composition, volume, temperature, and particulate loading.
4-18
-------
The same measurements should be made at the hood face and exhaust off-take.
The plume flow rate at known distances above the source may be estimated by
photographic techniques as described in Section 5.1. The heat generation
rate may be obtained by measuring temperatures at elevations above the
source. Data sheets summarizing the important measurements to be obtained
in the field program testing of local receiving hoods and assisted exterior
(push-pull) hoods for buoyant sources are shown in Figures 4-5 and 4-6.
Worthy of consideration at this point is an alternative approach. In
some situations, an entirely new hood system may be necessary for the
source. Installing a temporary hood above the source permits direct evalua-
tion of a new design. Connecting a fan and duct to the hood, for example,
may establish the required exhaust rate to meet the new design objectives.
In this approach, care must be taken to ensure that the maximum plume flow
rate has been observed, or at least, accounted for.
4-19
-------
Hood Design Data Sheet
Local Hood: Receiving
Description of Point of Emission,
Duration and Frequency of Emission
and Contaminant Description
Emission Source
Gas composition
Gas volume
Gas temperature
Paniculate loading
Face of Hood
Gas composition
Gas volume
Gas temperature
Particulate loading
Hood Off-take
Gas composition
Gas volume
Gas temperature
Particulate loading
Particuiate Characteristics
Chemical composition
Particle size
Particulate Emission Rate (at source)
Instantaneous
Hourly
Daily
Heat Generation Rate
Total
(Normal m/h)*
°C
mg/Normal m3
(Normal m/h)
°C
mg/Normal m3
(Normal m/h)
°C
mg/Normal m3
'kg/s
kg/h
kg/day
kcal/s
Hood Sketch
Hood Geometry Data
Face area
Hood height
Off-take area
Openings area
.m"
.m
m
Hood Performance Equation*
Original Basis
Hood Capture
Efficiency (%)
Analytical
Modeling
Current Performance
Analytical
Modeling
Field Measurements
Comments
'Normal implies 20° C,1 atm.
^Calculation sheets attached for specific
cases.
Figure 4-5. Hood design data questionnaire-A.
4-20
-------
Hood Design Data Sheet
Local Hood: Exterior
Description of Point of Emission,
Duration and Frequency of Emission
and Contaminant Description
Emission Source
Gas composition
Gas volume
Gas temperature
Part icu late loading
Hood Off-take
Gas composition
Gas volume
Gas temperature
Particulate loading
Participate Characteristics
Chemical composition
Particle size
Part icu late Emission Rate (at source)
Instantaneous
Hourly
Daily
Heat Generation Rate
Total
Plume Rise Data
Velocity @ Vfc D
@ hood centerline
Temperature
Plume cross-section
Area @ hood centerline
Nozzle Jet Data (Push- Pull only)
Nozzle air flow rate
jLt J '>J*fc«
iMozzie wiuin
Nozzle length
(Normal m3/h)'
°C
mg/Normal rrr
(Normal m3/n)
°C
mg/Normal m3
kg/s
kg/h
kg/day
kcal/s
m3/s
OQ
m2
Normal m3/h
m
m
Hood Sketch
Hood Geometry Data
A m
B m
C,
m
D m
H m
L m
For Push-Pull Only
E,
m
F m
a o
Flanges
Hood Performance Equation'*'
Original Basis
Analytical
Modeling
Current Performance
Analytical
Modeling
Field Measurements
Hood Capture
Efficiency (%)
Comments
* Normal implies 20° C,
1 atm.
^Calculation sheets attached for specific
cases.
Figure 4-6, Hood design data questionnaire-B.
4-2!
-------
SECTION 5
DESIGN METHODS FOR REMOTE CAPTURE OF BUOYANT PLUMES
In reference to the design process outlined in Section 3, it is assumed
that after consideration of the process fugitive source, a remote hood has
been selected for control of the buoyant source. Remote hoods are always
termed "canopy hoods," or sometimes qualified as "high canopy hoods" to
distinguish them from "low canopy hoods." In this manual, canopy hoods and
remote hoods are identical; a low canopy hood is simply a local receiving
hood (Section 4,1).
Canopy hoods are intended to act as receiving hoods to plumes having
buoyant motion arising from the associated hot process source. As discussed
in Section 3.1, remote capture of such plumes is the least desirable means
of control. Nevertheless, canopy hoods present little interference with
process operations, which undoubtedly accounts for their wide application
(Section 3,3). As the performance of canopy hoods is often unsatisfactory,
it is useful to first list common performance failures before discussing
design procedures. Typical failures of canopy hoods include:
1. Spillage
2. Plume deflection by cross-drafts
3, Plume spreading.
A discussion of each failure follows so that hood designers and reviewers
can consider them in reference to the particular case at hand. Spillage
occurs when the plume flow rate to the hood exceeds the hood suction rate,
i.e., fume simply spills out of the hood. For intermittent process fugitive
plumes, such as charging of furnaces, copious amounts of fume are produced
in a short duration often resulting in spillage from the hood. As the
buoyant plume rises from the source, dilution with clean air (entrainment)
decreases the plume velocity thereby allowing the plume to be deflected by
building cross-drafts. These drafts do not have to be excessive to cause
5-1
-------
the plume to be partly or totally deflected away from the hood. The last
cause of failure of canopy hoods is plume spreading around obstructions.
Because the path between the process fugitive source and the remote hood
often is obstructed by cranes, walkways, etc., the rising plume, in diverging
around such obstacles, spreads out with the result that the hood face area
may not accomodate the ultimate plume width. Although it is virtually
impossible mathematically and reliably to predict plume spreading, field
observations (of an existing site) or fluid modeling may be used to take
account of this problem.
Design methods for canopy hoods include the following: analytical
techniques, fluid modeling, and field measurement of an existing site. The
goal of these design methods is to obtain exhaust rates and hood dimensions
necessary to satisfy design objectives. Obviously, the first two methods
may be applied to planned or existing sites, whereas a field measurement
program may be carried out for an existing site only. Although treated
separately in the following sections, the design methods may not be that
distinct in actual practice. For example, analytical methods and fluid
modeling rely on the use of source parameters that could be measured as
part of a field program. Alternatively, a field measurement program may be
carried out to diagnose failures of a particular hood system. Possible
modifications to the system could then be readily evaluated by fluid model-
ing techniques, or the field data measurements could be compared to estimates
obtained from analytical techniques.
5.1 DESIGN BY ANALYTICAL METHODS
Analytical methods for design of canopy hoods for buoyant sources use
source characteristics (heat release rate and source-hood geometry) to
estimate the required exhaust rate and hood dimensions. The discussion
logically divides into uses involving continuous plumes, intermittent
plumes, and special cases of obstructions and cross-drafts. Each case is
discussed in detail in the following section. Table 5-1 summarizes the
governing equations with references to the text.
5-2
-------
TABLE 5-1, SUMMARY OF ANALYTICAL.TECHNIQUES FOR CANOPY HOODS
Source Hood parameters
Continuous plume Exhaust rate
Hood diameter
Intermittent plume Hood storage volume
Cross-drafts Exhaust rate
en Obstructions Exhaust rate
i
OJ
Governing equation
Qu = 0.166 Z5/3 F1/3
n
Qs=1.21QH
50 percent of Z
Hood volume = t(Q " Qi)
n - n nil? cross.
S n U
Perform fluid modeling or
diagnosis of existing
site
Reference
Eq. (5-1)
Eq. (5-5)
Section 5.
Eq. (5-6)
Eq. (5-8)
Section 5.
1.1
1.3
-------
5.1.1 Continuous Sources (No Obstructions, No Cross-Drafts)
The following analysis assumes that the canopy hood is located at a
distance greater than two source diameters above the source and that the
difference between ambient temperature and plume temperature is less than
100° C. If the roof to ground (source) temperature gradient is less than
17° C, the analysis may overestimate plume flow rate at the hood face by
3 percent. If the temperature gradient exceeds 17° C, the overestimation
increases, but it is conservative not to correct for the gradient. Larger
temperature (density) gradients, however, may cause the plume not to reach
the hood face, so fluid modeling is recommended in that case (see Section 5.2
for a discussion of this problem). In this analysis plume motion is also
assumed to be dominated by buoyant convection with no momentum flux from
the source (i.e., jets). To be termed continuous, a source must produce a
buoyant plume lasting at least 30 seconds.
The effective height, Z, between the canopy hood and a virtual plume
origin is taken as the distance between the hood and source plus one and
one-half times the source diameter (Bender, 1979). The plume flow rate at
the canopy hood face is calculated from an equation for a point plume
(Table 5-2) as follows (see Figure 5-1).
QH = 0.166 (Z)5/3 (F) (5-1)
where
Qj_l = plume flow rate at the hood face (m3/s)
Z = effective height from the virtual plume origin to the hood
face (m)
F = buoyancy flux (m4/s3).
The buoyancy flux is calculated using the following equation:
F =
5-4
-------
TABLE 5-2. SUMMARY OF EQUATIONS GOVERNING RISE OF BUOYANT
PLUME FROM A HOT SOURCE
Assumptions:
1. gaussian velocity profiles
2. small density difference
3. entrainment velocity: V(b) = all
4. equal spread of buoyancy (concentration) and
velocity profiles
Dimensions
Equation
Characterizing source quantity
V
assumed
uniform
Buoyancy flux = const.
F = QA[m4/s3]
Volume flow rate
Q
18Fa1/3 ,5/3
Center line velocity
U
max
m
s
5_ /18oF\l/3 7-l/3
6a ^ 5?T I
Entrainment const.
a.
0.093
Length scale
b
(6/5)otZ
Center line buoyancy m
\ax **
Froude No.
Fr = U /VA b
max ^ max
Entrainment angle, 8 (approx. ) Deg.
5 /5nWd/n
3n \18Fa / ^a j
V5/a = const.
18
z-5/3
2 = effective height from virtual plume origin to hood face.
F = buoyancy flux.
po " p
A = buoyancy = (g) , where p = ambient density, p = plume density,
^o and g = gravity constant.
Adapted from Bender (1979). The assumptions are discussad more fully in
Turner (1973).
5-5
-------
where
q = heat release rate (kcal/s)
g = gravitational constant (m/s2)
C = specific heat of air (cal/gm - °C)
T = absolute temperature (°K)
p = air density (kg/i3).
Emissions are usually in the form of an opaque fume which absorbs a signifi-
cant portion of radiant heat. Therefore, the heat transfer rate, q, should
consider both convective and radiative heat loss (in contrast to Hemeon,
1963, who recommended considering only convective heat loss). Governing
equations then for estimating these heat loss components are
= (hc)(As)(AT) (5-3)
where
q = heat transfer rate due to convection (kcal/sec)
h = natural convection heat loss coefficient (kcal/m2_°C)
A = surface area of heat source (m2)
AT = temperature difference between hot body and room air,
and for radiative heat loss,
qr = (e)(As)(ff)(T)4 (5-4)
where
q = heat transfer rate due to radiation (kcal/s)
e = emissivity (dimension!ess)
A = surface area of hot body (m2)
0 = Stefan-BoItzmann constant (kca1/s*m2-K4)
T = absolute temperature (K),
The preceding heat loss equations are familiar to most engineers, and,
typically, handbooks are used to obtain values for the coefficients (in
5-6
-------
consistent units). Heat transfer rates may also be determined directly
from some sources by measuring the temperature drop as, for example, in a
ladle of molten metal. Heat transfer rates may be badly underestimated if
exothermic chemical reactions occur in the source (e.g., ladle additions).
5.1.1.1 Required Hood Exhaust Rate--
For 100 percent capture of the buoyant plume and no spillage, the
required exhaust rate is obtained from Equation (5-5):
Qs = 1.21 QH (5-5)
where
Q = hood suction rate required for no spillage
Qy = plume flow rate estimated from Eq. (5-1).
The factor 1.21 is not an arbitrary safety factor, but was determined from
fluid model studies of the capture efficiency of canopy hoods (Bender,
1979). Spillage takes place in canopy hoods 1f the hood exhaust rate
exactly matches the plume flow rate because a mixture of plume and ambient
air circulates within the hood volume and spills from dead spaces of the
hood that do not receive the plume. The hood exhaust rate, Q4, divided by
the hood face area should provide a minimum face velocity of 1.5 m/s, or the
plume may overturn and spill from the hood.
5.1.1.2 Hood Dimensions--
As shown in Figure 5-la, the canopy hood face area must be sufficient
to accommodate the plume width at the height of the hood. Using the entrain-
ment angle of 18 degrees in Table 5-2, the plume boundaries may be estimated
by trigonometry and the hood sized accordingly. Alternatively, the hood
diameter may be chosen simply as one-half the value of the effective height,
which results in a somewhat more conservative value of hood diameter.
Storage capacity of the hood, and therefore the shape, is not important for
continuous sources. A typical hopper type canopy hood is shown in Figure
5-la.
5-7
-------
a)
\
\
(b)
I
\/
Virtual Origin of Plume
Source: Bender, 1979
Reprinted with permission by American Industrial Hygiene Association Journal.
Figure 5-1. Typical shallow hopper type canopy hood (a) and pool type canopy hood (b).
Effective source-hood distance, Z, is taken as the hood-source distance plus
1.5 times the source diameter, D.
5-i
-------
5.1.2 Intermittent Sources
Frequently the source of the buoyant plume does not produce a steady
plume, but rather, a huge volume surge lasting only a few seconds. Charging
electric steel making furnaces with scrap is a typical example of this type
of intermittent process fugitive source. The design of hoods for intermit-
tent sources is quite different than for continuous sources. From the
preceding discussion, a hood design based on exhausting a rate in excess of
the plume flow rate (Eq. (5-5)) would be totally impractical and excessive
for intermittent sources. A practical alternative is to use a canopy hood
with a sufficient reservoir ("pool-type hood") to temporarily store the
intermittent surge of fume (Figure 5-lb). The following section outlines
the techniques used to estimate the hood storage requirements for a pool-
type hood. Practical experience 1n the use of these hoods is also provided.
For intermittent sources, it is necessary to establish the maximum or
peak plume flow rate conditions that can be expected during the course of
process operations. Figure 5-2a shows a hypothetical case with the peak
plume flow rate represented as a step function above normal conditions.
The canopy hood volume required to store this surge can be expressed by the
following equation:
Hood Volume = trf (Q - Qg) (5-6)
where
t. = duration of plume surge (s)
Q = peak plume flow rate (nrVs)
Q = hood exhaust flow rate (ms/s).
Using example values of Q = 400 m3/s and t = 5 and 10 s, Figure 5-2b shows
storage volumes as a function of hood exhaust rate, Qj.
Various combinations of hood exhaust rate and hood storage volume can
be selected above the minimum exhaust line. The cost and layout restrictions
for providing a large storage canopy hood must be compared to the cost of
the hood exhaust system. The final selection is made to minimize the
overall cost.
5-9
-------
E_ 400
I
•§ 300
o
X
» 200
•**
<0
cc
§ 100
EL
2 0
Q-
"Peak" Plume Flow Rate
'Normal" Plume Flow Rate
Time, seconds
(a)
Minimum "normal"
Plume Exhaust
1000 2000 3000 4000
Hood Storage Volume, m^
(b)
Figure 5-2. Hypothetical example of intermittent plume case. Required hood storage
volume depends on duration of the plume surge.
5-10
-------
Note how the exhaust and storage requirements drastically increase if
the plume surge duration doubles from 5 to 10 s. If the surge lasts 30 s,
for example, the hood volume would have to be Impractically large to be
able to operate at a hood exhaust flow rate below the surge flow rate. The
source would then be considered "continuous" for practical design purposes.
Pool-type hoods, even when sized properly, can suffer from certain
performance failures. Turbulence of the stored fume can result in the fume
overturning and falling back out of the hood. A baffle arrangement as
described by Bender (1979) can be installed within the hood to prevent this
spillage. Another consideration is the frequency of the fume surges.
There must be sufficient time between surges to purge fume from the hood.
The purge time 1s simply the nominal residence time of the fume in the hood
given by the following equation:
. _ Hood volume ,* 7-,
~ Hood exhaust rate " l }
When the plume surge enters the hood, air from inside the hood is displaced.
If the displaced air still contains fume from the previous surge, this fume
will spill as the new surge enters the hood.
A very deep hood is sometimes used (as, for example, a hood formed by
the building roof trusses). In such cases, it has been found that the peak
fume surges can be stored without overturning, and, consequently, a baffle
arrangement is not necessary. Hood face velocities as low as 0.5 m/s are
adequate with this type of deep hood.
The case study in Section 7.1 illustrates the design of a pool-type
hood to improve the performance of a system originally designed as a hopper
type. The case study in Section 7.3 illustrates the use of a very deep
pool-type hood.
5.1.3 Special Cases: Cross-Drafts and Obstructions
The preceding analyses of continuous and intermittent sources are
predicated on the assumption that building cross-drafts and obstructions
between the canopy hood and process source are not present. In practice,
both cross-drafts and obstructions can significantly interfere with the
5-11
-------
operation of canopy hoods. The following section discusses measures to
take account of or reduce these effects.
Canopy hoods act as receiving hoods. The hood suction velocity field
induced by the canopy hood extends only a short distance. Therefore, even
light gusts within the building may deflect the plume away from the hood.
In general, the best solution for plume deflection by cross-drafts is to
shield the area with solid walls or curtains. Obviously, such shielding
must be placed so as to minimize interference with process operations.
In some cases, building cross-drafts may have a prevailing direction
and intensity, or a draft may be purposely generated by mechanical ventila-
tion. The possibility then arises of locating the hood eccentric to the
plume center!ine. The following equations are adopted front model experiments
performed by Bender (1979) for predicting hood requirements in a cross-draft.
The hood exhaust required to give the best theoretical collection efficiency
is described by the equation
V.
'max
where
Q = hood suction flow rate (m3/s)
Q,. = plume flow rate at the hood face (m3/s)
V = cross-draft flow velocity (m/s)
U = plume centerline velocity, m/s at the hood face (Table 5-2).
fllClX
The eccentricity (distance between the hood axis and plume axis) which
results from the cross-draft is described by the equation
Qs = QH (1 + 4.7-) (5-8)
max
where
e = eccentricity (m)
5-12
e = 13.53 (bH) - (5-9)
-------
b,, = plume length scale at hood face (m) (Table 5-2).
H
This equation holds with adequate accuracy for ratios of source diameter to
hood distance from the source of less than 1/5, for plume deflection angles
of less than 45 degrees, and for a hood face diameter equal to or less than
Use of these equations shows that even a light cross-draft will displace
the plume significantly. It is important to know building air flow ventila-
tion patterns. They may be predicted by considering the location and
velocity of all air inlet openings,
If the plume strikes an obstruction (e.g. an overhead crane} on its
ascent to the canopy hood, the plume will spread and entrain more air than
predicted by Equation (5-5). Depending on the size of the obstruction, the
plume could be deflected beyond the hood shape selected for the nonobstructed
case. It is difficult to predict analytically the degree of deflection.
Therefore, field observations or scale modeling should be used for setting
the hood shape when obstructions are expected to deflect the plume. If the
hood face area is increased to accommodate the deflected plume, the minimum
hood face velocity of 1.5 m/s should still be applied to prevent spillage.
5.2 DESIGN OF HOOD SYSTEMS BY FLUID MODELING
The use of fluid dynamic models to establish the sizing and performance
of canopy hoods is well established. Details of the modeling systems and
design/test procedures are presented in references such as Bender (1979),
Goodfellow and Bender (1980), and Fields et al. (1982).
The modeling procedure to be followed is as described in Section 4.2.
The resulting design equation for establishing required exhaust rates is
based on matching the Froude number of the model to that of the prototype
(canopy hood). The required exhaust rate for the hood is given by
Qp = Qm (S)5/3 (qp/q (5-10)
where
Q = canopy hood volume flow rate for the prototype
5-13
-------
Q = canopy hood volume flow rate for* the model
S = model scale (e.g., 10 for a 1:10 scale model)
q = heat flow rate for the prototype
q = heat flow rate for the model.
m
This equation can be rewritten in terms of temperature or buoyancy
flux instead of flow rates as follows:
where
T = representative hot gas temperature; the subscript o
denotes ambient conditions.
F = buoyancy flux = — (g)(Q)
Q = plume source flow
p = ambient density
p = source density
g = gravity constant.
If the modeling test medium is water with saline solution as the
PO " P
buoyant plume source, the buoyancy term (g) is selected to provide
po
an appropriate time scale.
In Section 5.1.1 it was mentioned that in plants where high roof-to-
ground temperature gradients (air density gradients) exist, plumes may not
reach the canopy hood face before being dispersed. This problem may be
observed in so-called closed process plants. In many of these plants,
ventilation to the atmosphere is avoided because of agreements with environ-
mental regulatory agencies. Air changes in these facilities are primarily
determined by the amount of process and fugitive emission control system
exhausts. This tends to leave the process building greatly underventilated.
Bender (1984) has demonstrated, in fluid dynamic model tests using salt
5-14
-------
water to scale the plant heat release, that the effects of in-pi ant density
gradients can be realistically modeled.
For intermittent plume sources as described in Section 5.1.2, water
models have been used successfully to simulate the process. The mean hood
capture efficiency can be determined accurately using a new technique
described in a paper by Bender et al. (1983),
5.3 DESIGN BY DIAGNOSIS/MEASUREMENT OF AN EXISTING SITE/PERFORMANCE
EVALUATION
As already mentioned, canopy hoods frequently perform poorly. The
capture efficiency of these hoods may be degraded by many factors, includ-
ing deflection from building cross-drafts, spreading around obstructions,
or spillage of captured fume. While observing the performance of an exist-
ing hood system is undeniably valuable, quantitative measurements are
necessary to prescribe remedies to what is often a confounding set of
problems. Since canopy hoods act as receiving hoods that rely on the
motion of the buoyant plume for collection, specific techniques for measur-
ing plume velocities are described in the following section. In addition,
a useful technique for relating hood capture efficiency to roof monitor
opacity is presented. Lastly, a design questionnaire summarizes the impor-
tant source characteristics and performance measurements that should be
part of a field measurement program or an intensive review of an existing
system. Use of the design questionnaire for a planned new facility is also
appropriate when measurements may be made in an existing facility that is
similar to the planned facility.
Goodfellow and Bender (1980) describe three field measurement techniques
for determining plume velocities. These techniques are: propeller anemometer,
stopwatch, and photographic scaling. A grid of propeller anemometers can
be arranged at the roof truss level. Usually six to eight anemometers
provide an adequate sampling. The plume velocity distribution is determined
as well as the average velocity using this technique. However, accumulation
of dust in the propeller bearings shortens the useful lifetimes of the
anemometers. As an example, Figure 5-3 is a plot of average plume flow
rates measured at roof truss level as a function of time for a typical
tapping operation on an electric steelmaking furnace.
5-15
-------
r\
X
0 15 30 45 60 75 90 105 120 13S 150 165 180195 210
Time, seconds
Source: Goodfellow and Bender, 1980; Reprinted with permission by American Industrial Hygiene Association Journal,
Figure ,5-3. Average plume flow rate as a function of time using
anemometer technique at an electric steelmaking furnace.
5-16
-------
The stopwatch technique for determining emission volume flow rate is
based on measuring the elapsed time for fume to rise between two known
levels (e.g. Zl5 Z2) with a stopwatch. For this test procedure to be
valid, the test must be carried out in a region where the rising fume
clearly exhibits buoyancy-dominated plume behavior. The calculation proce-
dure depends on of the location of the plume virtual origin and the heat
release for the process (see Figure 5-1).
At elevation Z2 above the plume virtual origin, the plume volumetric
flow rate is given by
=0.026 ( ^2 " MJ \l ±^_S - Jz2~ (5-12)
zt
where a = =—
'2
The emission flow rate from an electric-arc tapping process has been esti-
mated at any level above the steel ladle using the stopwatch technique in
conjunction with the plume velocity (Goodfellow and Bender, 1980).
Photographic scaling is perhaps the best of the three techniques.
Provided that the plume is properly illuminated, the average plume flow
rate and plume behavior may be determined. Procedurally, the plume should
be illuminated at an oblique angle to the camera; also, an object suitable
for scaling should be included in the scene. Although a standard movie
camera (18 frames/s) with 8 mm or 16 mm color film may be used, superior
results are obtained with a motor-driven 35-mm camera. The velocity of the
plume can be estimated by scaling from the speed of the film. The plume
diameter as a function of distance above the source is obtained by scaling
against the reference object. Figure 5-4 illustrates the photographic
scaling technique.
Failures in canopy hood performance are often realized as emissions
that escape through the roof monitors of a shop. Indeed, emission standards
for many process sources are expressed in terms of the opacity levels of
these emissions. It is therefore desirable to relate roof monitor opacity
to hood suction rate and hood capture efficiency. Based on fluid model
5-17
-------
Time
a
, Tn - 1
Reference Plane
Z1
OD
xx
0
Tn
X^
\ —/Plume
V-4-LadIo
Photograph
a
Tn
0
Tn-M
/~5»«\ '
w
a
Tn + 1
1
n
0
Tn + 2
™
vJ
D
>
!
a
Tn + 2 .
1 *-
Z2
a 8 a
S = Distance traveled = Z2 - Z1 T = Time span = Tn + Tn 4-1 + Tn + 1 V = Velocity =
D = Diameter
@ a
T = Frame exposure time T = Frame advance time
Source: Good fellow and Bender, 19BO; Reprinted with permission by American Industrial Hygiene Association Journal.
Figure 5-4. Photographic seating technique to analyze plume velocity.
-------
studies of the performance of canopy hood systems, the following generalized
expression may be used to summarize canopy hood performance as a function
of the ratio of plume flow rate at the hood face to hood suction rate, or
the ratio of captured pollutant to total pollutant arriving at the hood
face:
X
where
Q = hood suction rate
QLI = plume flow rate at the hood face
n
ru = pollutant rate captured by the hood
n
rp = pollutant rate arriving at the hood face
Hu -4 = capture efficiency of the hood
nOOQ
and X depends on the- hood type as follows —
Ideal hood: X = 2 (spills fume of low concentration from plume fringe)
Actual hood: 1 < X <_ 2 (intermediate between ideal and worst)
Worst hood: X = 1 (spills fume of average concentration).
This relationship is illustrated in Figure 5-5a. It is seen there that, in
general, actual canopy hoods performance lies between limits represented as
ideal and worst of hoods. This notion may then be extended to relate hood
performance to roof monitor opacity by the following relationship:
S
OP = 1 - (1 - OPmax) S H (5-14)
where
OP = observed or desired opacity level
OP = the maximum opacity observed for zero hood suction for
max an existing installation.
5-19
-------
1.0
0.9
0.8
-. °'7
""a
o °*6
X
J 0.5
0.4
0.3
0.2
0.1
(a)
11 i \ i \ i i**-^*^
0 0.1 0.2 0.3 0,4 0.5 0.6 0.7 0.8 0.9 1.0
Ratio of Hood Suction to Plume Flow Rate
if\ /rt.. \
"Worst" Hood
— "Ideal" Hood
(b)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Ratio of Hood Suction to Plume Flow Rate
(QS/QH }
Source: Goodfellow and Bender, 1980; Reprinted with permission by American industrial Hygiene Association Journal.
Figure 5-5. Useful relationships between canopy hood performance and rooftop opacity.
In (a), actual performance is found to lie between bounds of ideal and worst
hoods. In (b), amount of additional suction needed to reach required opacity
level can be estimated.
5-20
-------
Equation 5-14 can be derived from.the Lambert-Beer law with the fraction of
light transmission, i.e., (1-OP), a function of the light path length, the
concentration of participate, and certain other physical and optical proper-
ties of the particulate.
Figure 5-5b then is a plot of this relationship for the two limits of
X = 1 and X = 2. In the use of Figure 5-5b, the maximum opacity (zero hood
suction rate) must be measured. Then for a particular hood system, the
amount of additional hood suction (0 /Q,,} required to reduce the opacity to
S n
a certain level (OP) may be found from the figure. The case study in
Section 7.1 provides a detailed illustration of the use of this method.
Lastly, Figure 5-6 summarizes the important source and plume charac-
teristics which should be examined in the analysis of canopy hoods.
5-21
-------
Hood Design Data Sheet
Remote Hood: Canopy
Description of Point of Emission,
Duration and Frequency of Emission
and Contaminant Description
Emission Source
Gas composition
Gas volume
Gas temperature
Particulate loading
Hood Off-take
Gas composition
Gas volume
Gas temperature
Particulate loading
Particulate Characteristics
Chemical composition
Particle size
Particulate Emission Rate (at source
Instantaneous
Hourly
Daily
Heat Generation Rate
Total
Plume Rise Data
Normal plume volume and velocity
Peak plume volume and velocity
Duration of peak
Frequency of peak
Direction of cross drafts
Velocity of crass drafts
Plume diameter
Opacity of Discharge from Building
(Normal m3/h}*
°C
mg/Normal m3
{Normal m3/h)
°C
mg/Normal m3
\
kg/s
kg/h
kg/day
kcal/s
m3/s mis
rn3/s m/s
sec
Occurrenee/min
m/s
m
%
Hood Sketch
Hood Geometry Data
A m
B m
C m
D m
E m
P m
H m
I m
W m
K m
Hood Performance Equation1'
Planned Site
Analytical
Modeling
Existing System
Analytical
Modeling
Field Measurements
Hood Capture
Efficiency <%}
Comments
'Normal implies 20°
C, 1 atm.
t Calculation sheets attached for specific
cases.
Figure 5-6. Hood design data questionnaire lor canopy hood.
5-22
-------
SECTION 6
DESIGN METHODS FOR ENCLOSURES
The following section discusses enclosures for inertia! process fugitive
sources and for buoyant sources. In reference to the discussion of general
design considerations in Section 3.1, enclosures represent the preferred
method for control of process fugitive emissions because escape of emissions
is restricted to gaps or openings in the enclosures. Therefore, considera-
tion always should be given to the use of enclosures in planning a ventila-
tion system, although they may not always be practical where ready access
to the process source is necessary.
Design of enclosures for inertial sources is completely different than
design for buoyant sources. The dust produced by inertial sources arises
from the motion of the particulate matter itself, rather than from the
thermal head of the air in the case of buoyant sources. In Section 6.1,
dust-producing mechanisms of inertial sources are discussed. Emphasis is
placed on a common and significant application of enclosures to gravity
transfer operations of bulk materials. In Section 6.2, design considerations
for enclosing buoyant sources are presented. Since buoyant sources are to
be controlled, the discussion closely parallels Section 4.1. However, the
use of enclosures for buoyant sources entails its own set of difficulties.
Therefore, design procedures for enclosures are outlined and practical
experiences are summarized.
6.1 ENCLOSURES FOR INERTIAL SOURCES
Enclosures are practically the only hood suitable for large scale
inertial sources such as bulk materials handling operations. Unlike buoyant
sources, dust generated by these operations does not travel in predictable
paths, and the range of travel is usually limited. These considerations
preclude the use of remote hooding. Local hooding, i.e., receiving hoods,
6-1
-------
are sometimes used for inertia! sources that have a single direction of
travel, such as particles projected from a grinding wheel. But for large
scale inertia! sources, dust generation takes place in all directions.
Exterior hoods are generally found to be unable to alter the motion of
coarse participate matter that is projected away from the hood.
As with any hood system, design methods are used to obtain required
exhaust rates and hood dimensions. For enclosing inertia! sources, various
mechanisms of dust production that arise from the motion of the particulate
matter determine the required exhaust rate. The enclosure dimensions are
also affected by these mechanisms, although process requirements are factors
as well. Positioning the exhaust off-take (connection between hood and
branch duct) is of special importance in the design of enclosures for
inertia! sources. The following section is divided into a discussion of
dust-producing mechanisms applicable to all inertia! sources, design of
enclosures for gravity transfer operations, and considerations in the use
of nonexhausted enclosures.
6.1,1 DustGeneration inInertia!Sources
Dust generation mechanisms for inertia! sources have been reviewed by
Hemeon (1963). All these mechanisms of dust production arise from the
motion of the particulate matter and therefore are dependent on the size
distribution of the materials, adhesiveness of the material, moisture
content, friability of the material, and other factors. The main mechanisms
of dust generation are air induction, material splash, air displacement,
and air entrainment, as shown in Figure 6-1.
Air induction is probably the most important consideration in the
design of enclosures for inertial sources. During the motion of coarse
particulate matter, each particle imparts momentum to the surrounding air
stream. The macroscopic effect is an induced air stream. On reaching an
enclosure, the air streams outward through openings (e.g., access doors,
chutes, gaps, etc.), carrying dust with it. This phenomenon of air induc-
tion is familiar, as it is observed around a shower bath. For the quant-
ities of materials handled in industrial applications, the volume of air
induced is substantial.
6-2
-------
Externally Induced Air
Leakage
Opening
Enclosure
Falling
Material
Internally
Induced
Air
Material Splash
Container
Figure 6-1. Mechanisms for dust generation and dispersion during
material fall in an enclosure.
6-3
-------
Material splash refers to the violent escape of air and dust when
falling materials suddenly impact a hard surface. Obviously, the effect is
important for gravity transfers of material, although no quantitative
measure of the effect has been established. As pointed out by Hemeon
(1963), escape of dust by the action of material splash is local to the
compacting pile and therefore may be distinguished from escape by air
induction which occurs throughout all openings regardless of location.
Air displacement refers simply to the air displaced by the material as
the material is discharged into a container. The velocity and direction of
expelled air depends on the geometry of the container and amount of open
area. Generally, the volume of displaced air will be small compared to the
volume of induced air, but the quantity is easily calculated.
Air entrainment of dust occurs when any secondary air movements cause
further dispersion. The source of such currents may be random air currents
or external winds. Entrainment of dust can be an important consideration,
especially when the cause of secondary air motion is the primary dust-
producing machine (e.g., a pneumatic chisel).
Of the dust-producing mechanisms above, air induction and air entrain-
ment are important for determining the exhaust rate for enclosures. Air
displacement and material splash are important for determining the size and
shape of the enclosure.
6-1-2 Exhausted Enclosures for Gravity Transfer Operations
A common and important application of exhausted enclosures is to bulk
materials transfer points such as at chutes, bins, and dumping sites.
Design equations for estimating the required exhaust rate follow. Con-
sideration is also given to sizing the enclosure and positioning the off-
take.
The exhaust rate for an enclosure for controlling emissions from a
falling materials operation should equal the sum of the following quant-
ities;
1. Flow rate of air induced by the falling material. (This quantity
is typically much larger than air displacement; however, air
displacement may also be separately estimated.)
6-4
-------
2. Flow rate of air entering the enclosure by entrainment.
3. Flow rate sufficient to provide a working indraft velocity of air
through all openings (i.e., control velocity).
As pointed out in Section 2.2, Hemeon (1963) developed equations for estimat-
ing the volumetric flow rate of induced air based on the power generated by
the stream of falling particles, that is, the work done by the drag force
over the distance fallen per unit time. The recommended equation for
estimating the induced air flow rate is the following (Morrison, 1971 and
Dennis, 1983):
1/3
=0.631| , / 1 (6-1)
where
Q! = flow rate of Induced air (m3/s)
W = material flow rate (kg/s)
H = drop height (m)
A = cross-sectional area of falling stream (m2)
p = bulk solids density (kg/m3)
d = particle mass median diameter (m).
The flow rate of displaced air is given simply by the materials flow rate
divided by the bulk density:
Q2 = ~ (6-2)
Lastly, for a recommended control velocity of 0.5-1.0 m/s through the total
area of the openings, A, the flow rate is given by:
Q3 = A x v (6-3)
where
V = control velocity (0.5 m/s for well-protected sources; 1.0 m/s for
vigorous motion operations).
Sizing the enclosure is more important than might first appear. If
the enclosure walls are close to the compacting pile of material, material
6-5
-------
splash effects will cause losses through openings in these walls. Therefore,
the use of a larger enclosure allows the velocity of these air streams to
decrease before reaching the walls. Since no quantitative estimates may be
made as to the magnitude of the material splash effects, field observations
of an existing system and experience are the only guides. Air entrainment
becomes a factor when the enclosure has large areas or complete sides that
must remain open. Winds or local air currents then can enter and exit the
enclosure, thereby removing dust. The flow rate of ingress air can be
calculated in a straightforward manner from the wind velocity, open area,
and entry loss coefficient of the opening. However, the ingress air is
usually found to be quite large so that it may not be practical to attempt
to counteract it by enclosure exhaust alone. Positioning the exhaust
off-take close to the active zone of dust generation may capture the most
concentrated portion of airborne dust before recirculation and mixing with
entrained air can occur, thereby reducing needed exhaust to that for air
induction and control velocity only.
Selection of the off-take position is important from the standpoint of
the amount of material removed. Locating the off-take in the proximity of
the material stream or at points of splash will result in greater removal
of materials. This positioning may be desirable as a means to control
splash effects provided that the off-take velocity is kept low.
6.1.3 Nonexhausted Enclosures
Nonexhausted enclosures may be used to contain dust arising from
inertial sources and to protect against entrainment by winds. All the
difficulties attendant in the use of exhausted enclosures apply equally
well to nonexhausted enclosures. Since nonexhausted enclosures do not
maintain an inward air flow through openings, tight sealing is the only
means for restricting escape of dust. No design procedures for nonexhausted
enclosures can be given, but provisions should be made for removal of
settled dust and for access to any equipment inside the enclosure.
6,1.4 Capture Performance
Capture efficiency on an existing enclosure installation can be esti-
mated by measuring the portions of captured and spilled dust. The measure-
6-6
-------
ment program can be quite involved depending on enclosure size, intermittance
of operation, dust settlement in the enclosure, and the extent of air
entrainment. The measurement program would have to be custom designed to
best suit the operation,
6.2 ENCLOSURES FOR BUOYANT SOURCES
Enclosures are used in many industries to capture emissions from
buoyant sources (Examples are provided in Section 3.2 and Table 3-2). The
following discussion concerns large enclosures used on metallurgical process
vessels. Many of the design equations and procedures developed for buoyant
source hoods apply to large enclosure design. Process vessels successfully
using enclosures include electric arc furnaces, top- and bottom-blown
oxygen steel conversion furnaces, and nonferrous industry converters.
Use of enclosures for capture of fugitive emissions offers the follow-
ing advantages:
1. Total capture of emissions is possible and is as effective
as building evacuation. The enclosure, unaffected by in-plant
drafts, offers total containment.
2. As a side benefit with electric arc furnaces, the enclosure
offers a great potential for noise control.
3. Working conditions outside the enclosure are drastically im-
proved. The bulk of heat, fume, and dust from the furnace
are contained within the enclosure.
4. On small and low production furnaces, the enclosure can be
used as both primary and secondary control, thereby reducing
the need for other hardware.
The main disadvantage of using an enclosure is the potential for
interference with the normal operation and maintenance of a furnace. A
major design effort is required to overcome this disadvantage. All aspects
of the furnace operation must be considered. Lines of sight for furnace and
crane operators, access for crane-held ladles and buckets, furnace movements,
and maintenance access must be accommodated by the enclosure design. This
is more easily achieved in a new installation. Enclosure design on existing
sites becomes very difficult and may require a compromise between furnace
operation and fume capture performance.
6-7
-------
Figures 6-2 and 6-3 illustrate examples of enclosures on electric arc
furnaces and oxygen steel conversion vessels. Many different enclosure
shapes for similar furnaces demonstrate that enclosure design is very site
specific. Several patents exist for enclosures. Use of a patented enclosure
does not automatically imply success. Each installation is different and
requires that proper design procedures be followed.
Considerations in the design of enclosures for buoyant sources divide
into three areas: process and layout requirements, fume capture, and
mechanical design. The following discussion emphasizes electric arc furnace
enclosures.
6.2.1 Process and Layout Requirements
When planning an enclosure for a metallurgical furnace the following
questions should be asked:
1. Are primary and/or secondary emissions to be controlled by
the enclosure?
2. What is the extent of furnace and related equipment movements?
3. How will the enclosure affect the furnace operation process
control?
4. Where must enclosure openings be located?
6.2.1.1 Fume Control System—
On oxygen steel conversion furnaces, primary fume control is usually
achieved by a separate close capture hood positioned over the vessel mouth.
The enclosure then is used for secondary fume control on charging, turndown,
tapping, and slagging emissions.
On electric arc furnaces, an enclosure can be used for both primary
and secondary fume control; however, for large high-production furnaces, it
is more economical to provide separate direct evacuation control for primary
melting emissions. Use of separate gas cooling equipment to handle the
heat content of primary emission off-gas on high-production furnaces is
often less expensive than directly quenching with large amounts of dilution
air from an enclosure. The amount of air dilution is dictated by fabric
filter temperature limitations. Generally, if the enclosure exhaust rate
6-8
-------
BUMPER
SECONDARY HOOD
HOT METAL CHARGING LADLE
FURNACE CHARGING DOORS
OtetrKtitri*)
SLAG POT
WATER COOLED HOOD
MOOD TRANSFER CAR
ADJUSTABLE SKIRT
TAPPING EMISSIONS DUCT
SEAL RING
FURNACE ENCLOSURE
OPERATING FLOOR
LADLE
SHOP AW INDRAFT
(DURWG SLAGGING *
TAPPING
Source: Nicola, 1979.
Figure 62. Schematic arrangement for BOF furnace enclosure.
-------
r ROOF SLOT DOORS
(2-SECTIONS)
Note the use of an air curtain across the roof slot.
Figure 6-3. Enclosure for an electric arc furnace (EAF).
6-10
-------
for secondary fume capture is similar to or greater than that required for
primary control, the enclosure system is designed to handle both emissions.
Where primary control is afforded by the enclosure and fume leaves the
furnace via electrode openings, the extra wear and tear on electrode holding
equipment must be taken into account. This problem is particularly evident
on Ultra High Power (UHP) furnaces where the holding equipment would be
constantly exposed to high-temperature flame. As a possible solution, the
furnace could be equipped with a roof-mounted water-cooled stub stack which
naturally draws fume from the furnace and into the enclosure, thus diverting
fume and resulting damage from electrode equipment.
When primary fume capture is performed by the enclosure, furnace
off-gas combustion efficiency is lower than that for furnace direct evacua-
tion control. The off-gas (rich in carbon monoxide) rises from furnace
roof openings, partially burns, and cools with enclosure air. Significant
levels of CO have resulted in enclosures and exhaust ducting from this type
of combustion. These levels are not explosive but present a potential
hazard to personnel working in the enclosure or in downstream fume cleaning
equipment.
Therefore, as a final consideration, environmental regulations that
limit the amount of CO discharge from a meltshop may force primary emissions
to be handled by a high-combustion-efficiency fume control system.
6.2.1,2 Furnace and Related Equipment Movements—
Various furnace movements must be accommodated by the enclosure.
Furnace tilting for tapping and slagging, electrode vertical lift, and
direction of furnace roof swing must be accounted for in the design of the
enclosure shape and the location of the exhaust off-take.
Movement of related equipment must also be considered. The size and
position of doors and openings in the enclosure are determined by the
following furnace operations and associated rigging: tapping ladle and
slag pot positioning, charge bucket positioning, and routine removal of
furnace roof and water-cooled panels. Also, emergency measures must be
anticipated throughout the design. For example, a full ladle trapped in
the enclosure, because of a door jam, must be removed before the metal
solidifies.
6-11
-------
6.2.1.3 Furnace Operation and Process Control —
The following items affecting furnace operation and process control
should be addressed as the enclosure shape is considered:
1. Line of sight for crane operators and furnace attendants
2. Furnace control points and attendants location
3. Method of charging additives
4. Furnace ancillary equipment location
5. Equipment maintenance access requirements.
Furnace control points and ancillary equipment location may be positioned
in or out of the enclosure. If the bulk of furnace ancillary equipment is
located in the enclosure, layouts must allow for proper servicing. If
attendants must work in the enclosure during furnace operation, emission
capture design must provide a relatively fume-free work environment.
6.2.1.4 Enclosure Openings--
In general, enclosure opening requirements should be minimized during
the layout stage.
Bucket charging of an electric arc furnace requires a roof slot for
crane access. Sliding doors can be used to cover these openings. After the
bucket has entered the enclosure, the side doors are closed; however, roof
slot doors remain open. An air curtain blowing across the roof slot can be
used to prevent charging emissions from escaping through the roof slot.
Ample clearance is required to fit doors and air curtain equipment on the
enclosure roof. A roof slot is also required during tapping if the ladle
is held by the crane.
6.2.2 Fume Capture
Fume capture is accomplished by a combination of the following enclosure
features:
1. Containment and storage of the emission
2. Air extraction from the enclosure
3. Air curtain and exhaust off-take.
6-12
-------
If acceptable working conditions must be maintained in the enclosure during
the furnace operation, attention must be given to internal air flow patterns,
i.e., minimization of fume recirculation in the enclosure.
6.2.2.1 Containment and Storage of the Emission—
The main function of the physical enclosure is to contain secondary
furnace emissions from tapping, slagging, charging, and, perhaps, primary
emissions from melting. These emissions are thermally entrained against
the enclosure roof. If the enclosure is not built tightly, these emissions
can overcome the indraft effect of the extraction system. Gaps around roof
slot doors can also present a severe leakage problem. When the roof doors
are open for crane rope access, an air curtain can be effectively used to
contain emissions.
The enclosure is also capable of storing fume surges during bucket
charging. With proper design, the top of the enclosure will fill with fume
while the lower working level remains clear. The key to producing this
effect is to reduce fume recirculation in the enclosure by proper placement
of the air curtain with respect to the exhaust off-take.
Tapping, slagging, and melting are prolonged operations, and, therefore,
the enclosure should not be used for fume storage during "these periods.
The enclosure exhaust capacity must be greater than the emission plume flow
rate to avoid fume buildup in the enclosure during these operations.
6.2.2.2 Required Exhaust Rate--
To determine the air exhaust rate from the enclosures the following
steps are recommended:
1st Step—Primary Emission Heat Content
The heat content of furnace emissions and the temperature limitation
on the fume collector are considered for this step. The off-gas heat
content is calculated for furnace reactions during melting and refining
periods. (This lengthy calculation procedure is not covered in this manual.)
Assuming a fabric filter collector is used with polyester cloth, a 250° F
temperature limit is imposed for continuous operation.
The fume volumetric flow rate after dilution is then determined from
the following equation (equivalent to Equation 4-4):
6-13
-------
Q = q/((p)(Cp)(Ts -
where
Q = actual volume flow rate after dilution
q = heat transfer rate from furnace off-gas
0 = air density at Ts
C = specific heat of air at Ts
T = specified air temperature after dilution
T . = ambient dilution air temperature.
For a high production furnace, the fume volume flow rate, after air
dilution to 250° F, will be considerably higher than for secondary fume
control by the enclosure, and a separate primary fume capture system would
be used.
For the remaining steps, a small low-production furnace is under
consideration, with both primary and secondary emissions being captured by
the enclosure.
2nd Step—Secondary Emission Plume Flow Rate
The fume flow rate for charging and tapping is then predicted by
methods covered in Section 5 and in Sections 7.1 and 7.2. The enclosure
height is taken as the limit of plume rise. The plume rise from the open
furnace before charging should also be calculated. This event is a prolonged
emission.
3rd Step—Enclosure Exhaust Rate
The volume flow rate from prolonged emissions during roof swung open,
melting, and tapping sets the minimum exhaust rate required to ensure a
relatively fume-free enclosure environment. The fume volume flow rate
after dilution (from 1st Step) is compared to the highest of the calculated
plume flow rates from the prolonged emissions. The greater of these two
rates determines the enclosure exhaust rate.
Although the charging plume flow rate can be higher than tapping, it
does not set the enclosure exhaust rate. Instead, the enclosure is used to
store this approximately 30-second surge.
6.2.2.3 Air Curtain and Exhaust Off-take—
Air curtain design and exhaust off-take location are very important
considerations.
6-14
-------
The air curtain is applied on roof openings which are typically 2.5 to
3.0 m wide, and used for crane rope access. The opening may extend
over the length of the enclosure and should therefore be served by two sets
of independently working doors—one for tapping and one for charging. This
feature minimizes the open area when one of the two events occurs, as shown
in Figure 6-3.
The optimum position for the exhaust off-take is directly opposite the
air curtain discharge. Rising fume with the highest concentration is
directed straight into the off-take without excessive recirculation in the
enclosure.
The main purpose of the air curtain is to contain the vertical updrafts
from charging and tapping emissions. Because of the upward momentum of
these emissions} the air curtain slot discharge should therefore be pointed
downward (15 to 25 degrees from the horizontal) to achieve an approximate
horizontal resultant flow.
The air curtain design procedure is outlined in Section 5 and illus-
trated in the case study in Section 7.2. The plume data for furnace charging
is used in this step. Note that the plume volume flow impinging on the
width of the slot should be used rather than the whole plume flow.
During melting, the air curtain should efficiently direct fume towards
the exhaust off-take without allowing recirculation within the enclosure.
The air curtain design should therefore also account for this fume trajectory
when a lower updraft velocity from melting is experienced.
The air curtain supply air can be taken from either inside or outside
of the enclosure; however, there is a net flow advantage to taking this air
from the inside.
Elevated work area temperatures in the enclosure at operating floor
level may be a problem. Limited louver openings or wall fans can be used
for man cooling if operators must normally spend prolonged periods in the
enclosure.
6-2.3 MechanicalDesign
The success of an enclosure installation depends heavily on acceptance
by operations and maintenance personnel. Mechanical and structural relia-
6-15
-------
bility must therefore be designed into the enclosure. The following are a
few design details to be considered:
1. After opening locations and proper clearances have been
established, the enclosure frame support system should be
considered. Major support beams placed at the edge of
openings will provide extra strength against the rubbing of
crane cables. The overall construction should be light,
which allows fast easy repair in the event of collision with
crane held objects. Collision with a robust enclosure woyld
still result in damage and probably be more difficult to
repair.
2. Enclosure doors should be designed with generous clearances
and be easily operated by simple mechanisms. Wheels, guide
rollers, and pneumatic cylinders can be used as part of door
mechanisms.
3. To minimize leaks, roof doors that are susceptible to fume
updrafts should overlap the inside of the enclosure shell.
All roof construction must be tightly sealed.
4. Access for easy maintenance must be provided. Removable
roof panels for access to furnace subassemblies are desir-
able. Water cooled equipment, electrode and roof movement
mechanisms, etc., all require overhead access for'proper
maintenance. Small jib cranes may have to be located in the
enclosure.
5. Material selection for the enclosure shell should consider
environment corrosiveness. Aluminized sheeting is preferred
over zinc coated materials in a steel production environment.
6. The damaging sound levels produced by an electric arc furnace
can be contained within a furnace enclosure if a proper
acoustical design is carried out. Any design should be
made, or at least checked, by a acoustical engineer. The
following points should be considered:
The material should be sufficiently heavy. In most
cases structural requirements already ensure this.
The cladding should be sufficiently stiffened or damped
to preclude resonances at the furnace frequency and its
first few harmonics.
The inside of the enclosure should be lined with sound
absorbing material (eg. fiberglass) selected for the
frequencies involved and suitably protected from damage.
Holes, openings, and air leaks should be minimized,
treated, or at least located away from people where
'. possible.
6-16
-------
Operating practices should minimize the amount of time
operators have to spend inside the enclosure or near an
opening while the furnace is operating.
6-17
-------
SECTION 7
CASE STUDIES OF PROCESS FUGITIVE PARTICIPATE HOOD SYSTEMS
The following section discusses a unique collection of hood system
designs. Each design is treated as a case study. The studies represent a
varied range of industries, hood types, and design methods. The intent of
this section is to provide insights into the design and/or analysis of
either actual installations or representative examples.
An overview of the case study selection is given in Table 7-1. Case
studies I and II illustrate analytical techniques described in previous
sections. Case studies III and IV Illustrate design by precedent, i.e.,
using a working system as a model for the case at hand. Case V illustrates
the use of physical scale modeling in the design of an enclosure. Case VI
illustrates the use of design by rule-of-thumb, although the rule has been
tested and modified by the designers. The intent is that the reader gain
an appreciation of the difficulties in design of hood systems that no
simple, textbook-type problems can provide.
7.1 CASE I: CHARGING AND TAPPING CANOPY HOOD FOR AN ELECTRIC ARC FURNACE
7.1.1 Source Description and Background
7.1.1.1 General —
Case I is a canopy hood installation on an electric arc furnace melt-
shop. The shop operates one 18-ft diameter, 80-ton furnace powered by a
35 MW electrical supply. Since startup in 1975, the feed to the furnace
has been 100 percent scrap charge. The fugitive particulate emission
source is furnace tapping and charging. These buoyant emissions are cap-
tured by a canopy hood. The canopy hood and furnace direct evacuation
share a common fume collection system.
The major objective of this case study is to demonstrate an analytical
technique for calculating the amount of additional hood suction required to
7-1
-------
TABLE 7-1. OVERVIEW OF CASE STUDY SELECTION
Case Hood type
Process
fugitive source
Method
Highlights
i
ro
I. Canopy hood
II. Assisted exterior hood
III. Local receiving hood
IV. Canopy hood
V.
Enclosure
VI. Assisted exterior hood
Electric arc furnace
Charging
Tapping
Copper converter
Charging
Skimming
Basic oxygen furnace
Charging
Electric arc furnace
Charging
Tapping
Clamshell unloader
Lime dust
Aluminum rolling mill
Lubricant aerosol
Diagnosis of an
existing site
Performance
evaluation
Design by
precedent
Design by
precedent
Physical scale
modeling
Design by
rule-of-thumb
Mapping plume behavior
Plume storage
Eliminating cross-drafts
Air curtain theory
Tracer evaluation
Opacity measurements
Survey of installations
Combustion effects
Hood storage volume
Scavenger ducts
Opacity measurements
Effects of variables
Positioning off-take
A1r curtain uses
Field verification
-------
reduce the opacity of emissions from the shop roof to a specified level.
This technique is applicable when air pollution regulations are based on
opacity levels from the shop roof.
The method requires field measurement of opacity, hood suction, and
plume flow rate data at an existing installation. The data presented here
were collected for an electric arc furnace shop during a detailed study of
charging and tapping roof emissions.
This case study is well documented and includes a discusssion of the
design approach for the original installation, details on the as-installed
system, observed and measured hood performance, and the design approach for
hood modifications for meeting a predetermined opacity level. A final
design summary allows comparison of the various canopy hood performance
parameters which are developed through the course of this example.
7.1.1.2 Canopy Hood System—
The canopy hood is built into the roof truss space and divided into
three sections, as shown on Figure 7-1. Power operated dampers in the hood
are remotely controlled to function as follows:
1. Furnace Meltdown—The charge and tap side dampers are open, while
the top section modulates and supplies quench air for cooling
direct evacuation gases from the furnace.
2. Furnace Charging—The top and charge side dampers are open.
3. Furnace Tapping—The top and tap side dampers are open.
7.1.1.3 Regulatory Standards-
Regulations affecting the control of fugitive emissions from electric
arc furnace operations fall under both ambient air and workplace agencies
(e.g., EPA and OSHA).
During design and installation of the original fume control system in
1975, there were no applicable ambient air regulations regarding opacity of
charging and tapping emissions. The degree of control required was based
on allowable process weight emissions from the collection system stack and
suspended particulate (ground level concentrations) regulations. As a
result of proposed 1979 environmental law revisions, the opacity regulation
for e^ctric furnace shops in the particular jurisdiction would permit
7-3
-------
LOOKHS SOUTH
CHAIMMOMStE
Note; Also shown is the proposed pool-type hood addition.
Figure 7-1. Original canopy hood system for control of process fugitive
emissions from an electric arc furnace.
-------
emissions of not more than 20 percent opacity except for 40 percent opacity
for not more than 4 min/hr/furnace,
7-1-2 Design Approach for the Orlgi nal Installation
7.1.2.1 Calculation Procedure —
The volume of fumes rising into the roof hood during charging and
tapping of the furnace were calculated based on
1. Height of hood above the furnace and ladle
2. Furnace and ladle diameter
3. Rate of heat release from the furnace and ladle.
A simple calculation procedure (below) showed 360,000 acfm of air would
rise into the canopy hood at the meltshop roof level. The design calculation
procedure used for this application follows.
Heat release — Assuming a rate of temperature drop of ladle and furnace
as 10° F/min, the rate of heat release is
„ - 7B ton 2,000 * 0.12
q = 180,000 Btu/min .
Plume flow rate — Plume flow rate is calculated using an equation from
Heraeon (1963):
Q = 7.4 (Z)1'5(q)1/3
where
Q = fume volume reaching the canopy hood (acfm)
q = heat release (Btu/min)
Z = height of canopy hood above the virtual plume source
Z = Y + 20, where Y is the distance from the top of the source to the
hood face, and D is the source diameter in feet.
7-5
-------
Hood Face
2B
Furnace
Virtual Plume Origin
For furnace charging
Y = 55 ft
D = 18 ft
Z = 55 + (2 x 18) = 91 ft
Q = 7.4 (91)1'5 (180,QOO)1/3
Q = 360,000 acfm.
F°r ladle tappi ng
Y = 76 (from Figure 7-2)
D = 10 (from Figure 7-2)
I = 76 + (2 x 10) = 96 ft
Q = 7.4 (96)1'5 (180,000)1/3
Q = 391,000 afern.
Hood design—The hood shape and cross-sectional area can be determined
by considering the following:
1. Plume diameter at the hood face
2, Plume deflection by building cross-drafts
3. Hood face velocity.
The diameter of an unobstructed plume at a specified height above the
source can be determined using the following equation from Hemeon (1963):
G = I °-88/2
For furnace charging, this theoretical diameter is about 27 ft; however,
the plume is greatly obstructed by the scrap bucket and crane. It would be
difficult to predict analytically the plume spread around these obstructions.
Physical modeling, or observations in similar plants, could help determine
the expected plume bifurcation.
The design basis of the original hood shape was determined by the
fume-collection system supplier. The final hood dimensions were determined
by experience in other meltshops with similar obstructions in the path of
the plume rise.
7-6
-------
MODULHWG DAMPED
ISOLATION DAMPERS
!(•—- vmruAt PLUME onoiN
LOOKIHB NOHTH
Note: Also shown is the proposed retractable chain curtain system.
Figure 7-2, Furnace tapping fume emissions.
-------
7.1.2.2 Final as Installed Design--
Calculations indicated that furnace direct evacuation control during
melting required 180,000 acfm. Of the total fume emission from an electric
arc furnace, about 93 percent occurs during meltdown and only 7 percent
during charging and tapping. A significant expense was required to capture
only 7 percent of the total emission.
To meet 1975 environmental regulations pertaining to mass discharge,
the capture of melting fumes was sufficient. It was assumed that with most
of the total emissions captured, any remaining visual emissions would
probably be acceptable as well.
As a result, the system was designed with a total capacity of 216,000
acfm available to the canopy hood during charging and tapping, and 180,000
acfm available to the direct evacuation during meltdown. The full size
canopy hood was installed as shown on Figure 7-1.
7.1.3 Data CoHect 1 on f or Sys temMod i f i cat i o n s
After the system had been in operation for 5 years, increased concern
regarding emission opacity made it necessary to undertake a detailed study
of roof emissions. The performance of the "as installed" canopy hood
system was evaluated to verify design parameters and to define new require-
ments for upgrading the meltshop fume collection system.
The canopy system flow was measured at 212,000 acfm during charging
and tapping (design was 216,000 acfm). As a revised operating practice to
ensure enough air for proper combustion of furnace direct evacuation gases,
the isolation dampers in the three-section canopy were left open. This
revision in operating practice reduced the evacuation rate for the charge-
side portion of the hood during charging and the tap-side portion of the
hood during tapping, thus reducing the hood's effectiveness in capturing
charging and tapping emissions.
Observations were made to establish the plume size and behaviour
during charging and tapping. Charging plume velocities near the roof truss
were measured using a plume photographic technique, while an analytical
approach was used to evaluate the tapping plume flow rate. The opacity of
spilled fume discharging through roof exhaust fans was measured using an
7-*
-------
opacity monitor. All of these steps were used to define requirements for
complying with opacity regulations,
7.1.3.1 Field Observations—
Charging plume—When charging the furnace, the crane operator places
the scrap bucket above the open furnace. The fume already rising from the
furnace flows around the still closed bucket, impinges on the charging
crane> and spills from the canopy hood. The plume spreads beyond the hood
face area causing some of the rising air to miss the hood. Furnace fume
emissions increase noticeably as the crane operator slowly opens the scrap
bucket. When the bulk of the scrap drops, a large cloud emerges from the
annul us between the furnace and the bucket, and some fume emerges from the
furnace door. The crane operator moves the bucket away from the furnace as
fume starts rising through the bucket. Fume is dragged away with the
moving crane and bucket.
The plume velocity varies greatly. The plume rises slowly when the
furnace is relatively cold and the scrap contains a minimum amount of
combustibles. As the fume rises slowly, it is subject to dispersion by
building air cross currents. The plume is usually dense and dark brown.
The plume rises most quickly when the furnace is hot, particularly
when there is a hot metal pool in the furnace and when the bucket contains
combustible materials. A fast rising plume with a ball of fire engulfs
much of the charging crane. Particulate entrainment in such a plume is
significant.
Both extreme plume cases were observed in this plant. In either
situation, the capture efficiency of the canopy hoods is low, with the
spilled fume leaving the building through roof fans. The typically observed
charging plume contour is shown in Figure 7-1. It is apparent that the
fume hood size is not adequate for the actual generation rate, taking into
account the crane obstruction.
The easterly perimeter of the charging hood appears to be in an optimum
location relative to plume trajectory, although much of the fume misses
this edge because of deflection by the crane trolley. With a deeper hood,
acting as a storage reservoir, and greater suction, more of this fume could
7-9
-------
be captured. In the north and south direction, a significant amount of
fume misses the hood partly due to crane obstruction and hood size.
Tapping piume--The distance from the top of the ladle to the roof
trusses is 76 ft. An undisturbed plume would just rise in between the
crane bridge. However, cross-drafts and the crane trolley cause fume to
spill out from the sides of the crane, as shown in Figure 7-2.
During tapping, a southerly building cross-draft causes fume to spill
on the north side of the hood. With a northerly cross-draft, fume escapes
on the south of the hood. An increase in hood face area to accommodate
fume being spilled by these cross-drafts was not recommended. An extended
hood would be taking in clean air on the upwind side of the deflected
plume, resulting in lower overall fume collection for the hood.
Due to deflection by the crane trolley, large volumes of fume miss the
hood on the west side. Should a tapping canopy hood modification be required,
a hood on this side could be considered.
7.1.3.2 Plume Flow Rates and Hood Evaluation—
Observations confirm that proper fume hood design has to take into
consideration any obstructions the fume might encounter on its way to the
fume hood. Theoretical calculations based on simple plume flow rate equa-
tions do not predict plume growth around obstructions. For a greenfield
site, fluid dynamic scale modeling can be used for such predictions. In an
existing plant, visual observations of the problems using the plume photo-
graphic technique can be used to measure the plume characteristics.
Figure 7-3 shows the degree to which the cranes and other structures
block the canopy hood face. The outline of the plume edges are shown as
they cross the hood face after passing the obstructions, A large percentage
of the area is made up of solid walkways attached to the crane bridge (as a
small improvement to the fume capture, these walkways could be replaced by
grating). Figure 7-3 helps determine the plume cross-sectional area used
in determining flow rate from velocity measurements, and establishes the
proper location for a modified hood.
Furnace charging—Photographic scaling of charging plumes was used to
generate the fume flow-rate diagram in Figure 7-4. A peak charge flow rate
7-10
-------
34-t
1
N
tXISTMS HOOO
OUTLINE
Figure 7-3. Map of the plums boundaries relative to the original hood system.
-------
2 1,600
u.
u
g 1,400
0
x 1,200
Ul
^J 20
OPEN FCE., LIGHT
FUME EMISSION
PEAK THOUGHT
POSSIBLE, ESPECIAL
HEAVY FUME
CASE 'A1
IFV^.
WHEN CARBON ADDED .' \
TO SCRAP / \
/ \ r
SEC. i
EMISSION
,
-SHADED AREA REPfi
MINIMUM STORAGE
1,000
u. 800
UJ
3 600
Q.
o
I 40°
cr .
5 200
PEAK PHOTOGRA-
PHICALLY OBSERVED
DURING TESTS
CASE 'B1
REQUIRED FOR POOL TYPE
HOOD WITH 520,000 ACFM
SUCTION
X
52D.OOO ACFM
EXHAUST RATE REQUIRi
TYPE HOOD
WJTH POOL
\
__ 2i2,000\ACFM
EXISTING CANOPY EXHAUSf RAT
(MEASURED)
1
O
iy
u
u.
2468
UJ
HO
10 12 14 16 18 20 22 24 26 28 30 TIME
I I I (SEQ
UJ
CO
o
UJUJ<
zxz
-------
of 920,000 acfm was actually measured, but observations of much more violent
charges and experience gained from plume tests in other steel plants suggest
a peak flow rate of about 1,400,000 acfm for a few seconds.
The following shows the measurement approach:
1. From the fume interference diagram, Figure 7-3, the plume cross-
sectional area at the canopy hood face level is estimated to be
1740 ft2.
2. Measured plume velocity as shown on Figure 7-1 is 530 ft/m1n
(from plume photography).
3. Peak plume flow rate is therefore 1740 x 530 = 922,000 acfm.
It is not practical, nor necessary, to design a fume system to have a
suction flow rate equal to the peak charging fume flow rate. Well designed
fume hoods compensate for peak fume generation rates by temporarily storing
the fume (Section 5.1.2). This technique allows the fume control fan to be
considerably smaller. Excessive hood face area is as undesirable as insuf-
ficient hood storage volume. Hood area resulting in face velocities of
less than 300 ft/min tend to spill fume. The technique for determining the
optimum hood storage volume which minimizes the hood suction requirement
will be demonstrated for the charging hood.
From field observations and Figure 7-3, the approximate hood face
cross-sectional area has been established as 1740 ft2. With 300 ft/min as
the minimum face velocity, a 520,000 acfm hood suction requirement is
calculated by multiplying the hood face area times the nominal face velocity.
(More recent experience has shown that a face velocity of 100 ft/min can be
tolerated if the hood is deep enough.) The hood storage volume is determined
by referring to the charging plume flow-rate diagram in Figure 7-4. The
area above the 520,000 acfm horizontal line and under the plume flow-rate
curve, for case 'B', represents the minimum volume required for storing the
plume surge.
For the present example, the area under the curve (obtained by inte-
gration) represents 30,000 ft3. A pool type hood incorporated into the
existing roof structure will provide a total hood storage of 45,000 ft3.
The shape of a proposed pool type hood is shown in Figure 7-1.
7-13
-------
Furnace Tapping—Analytical considerations involving the
ladle heat release, plume theory, and the meltshop geometry were used to
predict the tapping fume flow rate. This prediction was confirmed by
observations of the plume as shown in Figures 7-2 and 7-3.
The following shows the analytical approach:
1. Heat release for an 80-ton tapping ladle is estimated at 158,000
Btu/min from fundamental heat transfer calculations.
2. Both the radiation and convective portion of ladle heat release
are assumed to heat the plume. A significant portion of radiant
heat is absorbed by the opaque iron oxide fume.
3. Buoyancy flux (Equation (5-2)) is calculated from:
F - (fl)(q)
- (Cp)(T0)(P0)
where
q = heat transfer rate (Btu/min)
g = gravity constant = 32.2 ft/s2
C = specific heat of air = 0.24 Btu/lb °F
TQ = absolute air temperature = 530° R
PQ = air density = 0.075 lb/ft3
F = 158.000 x 32.2 x 3600 sVrain2 _ - „ 1Q9 ft4/m1n3
h 0;24 x (460 + 70) x 0.075 1'9Z 10 ft /m1n '
The plume volume at the existing hood face is calculated from an
equation for a point plume (Equation (5-1)):
Q = 0.166(Z5/3)(F1/3)
where Z = height from virtual plume origin to the hood face. Therefore,
Z = 91 ft (from Figure 7-2, 76 ft + 15 ft)
Q = 0.166 x 915/3 x (1.92 x 109)1/3
Q = 380,000 acfm (similar to Section 7.1.2.1 result).
7-14
-------
Ladle additives which produce exothermic reactions can significantly
increase the plume flow rate. As an upper limit, some additions may double
the buoyancy flux, and the plume flow would increase as follows:
Q = 380,000 x (2)1/3 - 479,000 acfra .
From the interference diagram in Figure 7-3, it is evident that modifi-
cations to the canopy hood to help capture tapping emissions are necessary.
Extensions to the hood face to cover fume deflection from the crane trolley
and cross-drafts would result in an excessive face area. A higher hood
suction rate combined with baffles would be requird to maintain a reason-
able face velocity.
An alternative to major hood modifications is a high level curtain
enclosure to contain the tapping fumes. This concept is shown in Figure 7-2.
The capture of tapping fume could be improved if the face of the existing
canopy hood could be lowered by use of the curtain. The volume of the
tapping plume would be reduced from 380,000 to 175,000 acfm with a four-
sided enclosure hung 16 ft below the tapping crane.
7.1.3.3 Opacity Measurement—
In order to develop a design basis for fume control system modifica-
tions, an opacity monitoring program was performed on the meltshop roof
exhaust fan emissions. The instrument used was a Lear Siegler RM 41P
opacity monitor with a recorder. The location of the roof exhaust fan with
respect to the canopy hood is shown in Figures 7-1 and 7-2.
The opacity measurement results are summarized in Figure 7-5 for
maximum and normal emissions. Conclusions are drawn as follows:
1. The first charge rarely exceeds 20 percent opacity.
2. The second and third charge frequently exceed the 40 percent
opacity limit.
3. Tapping rarely exceeds the 40 percent opacity limit.
4. The second and third charge combined opacity can exceed the
4-min/h allowable limit.
5. The third charge and tapping opacity can exceed the 4-min/h
allowable limit.
7-15
-------
100 •
iO -
80-
70
>
t—
i 60
a.
o
H 50
l_
to
< 40-
X
X
" 30-
o
£ 20
10
/%
- — ^
y
2
K
X
. TO-
wZ
< 60
0.
o
. 50 .
t— JW
1 40-
•«•
X
UJ
u. 30-
o
o
s*
s
s
S
S
S
UJ
<
X
u
T3
^
3.5
MIN.
0*
1 !
x
X1
X
X
X
X
J-
X"
X
X
X
x
X
X
X
X
x^
X
X
X
X
X
X
X
X
X
X
X
X
MAXIMUM OPACITIES
CASE A
/-TIME
20% OPACITY
/) EXCEEDED
/
/
/
/
/
/
/ f
/ '
I
C9
£
a.
!S
/ 4.9,
/ MIN.
^3.8
MIN.
i i i i
10 20 30 4O 50 60 70 80 90 100
nNC-Tuniio MFAT
p^
S
S
/
s1
/
/
/
s
i^
jS
^
/•
/
x^
^~
• |$t CHARGE
/ OF NEXT
^ HEAT
X^
"/
»
^
x
x
^
x
0
U
^
,2.6
MIN.
i i i
/-TIME
20% OPACITY
/I EXCEEDED
/
/
w
1
c?
•e
2.6 «
MIN.
*"
i
/
/
/
/
/
/
/ .
/ f
/ 3.
I
|
en
z
a.
S
**
8
/ MIN.
2.6
""""lyll}^
i t ii
10 20 30 40 50 60 70 80 90 100
.i ONF - ? Mm 1R UP AT —
^_
X
<"•
X"
x
x
^
X
•*—
•f
_/~or NEXT
>j HEAT
s
s-
^ I 1
IIO I2O 130 140
T1UP fMIM \
Figure 7-5. Maximum and normal electric furnace charging and tapping
emission opacities,
7-16
-------
6. With the first charge being ignored, tapping and the follow-
ing second charge are spaced about 45 min apart (less than
1 h). Their combined opacity, when greater than 20 percent,
can exceed the 4-min limit.
7, Charging alone rarely exceeds 20 percent opacity for more
than 4 min.
8, When charging emissions are In the 20 percent to 40 percent
opacity range, they exceed the 20 percent limit usually for
less than 2 min. Two subsequent charges therefore usually
do not violate the regulation even though their opacities
might be in the 20 percent to 40 percent opacity range.
9. Tapping alone can exceed 20 percent opacity for 4 min.
10. A roof exhaust opacity of 100 percent is expected to occur
without canopy fume hood suction.
7.1.4 Design Approach For System Modification
The design conclusions arrived at from performance analysis are the
following;
1. With charges spaced about 25 min apart, the roof exhaust opacity
has to be reduced to less than 40 percent in order to satisfy air
pollution regulations.
2. With a charging exhaust capacity adequate enough to reduce the
opacity to less than 40 percent, the tapping opacity will most
certainly be less than 20 percent with the proposed hood modifi-
cations shown in Figure 7-2.
An analysis of possible ways to reduce electric furnace secondary
emissions to less than 40 percent opacity was carried out. Three methods
(and their combinations) of improving secondary emission control are des-
cribed as follows.
7.1.4.1 Increase Canopy Exhaust Capacity—
The increase in canopy hood exhaust capacity required to reduce charg-
ing emissions to less than 40 percent opacity is determined by referring to
Figures 7-4 and 7-5 and by using the following calculation procedure:
Opacity (OP), as a function of peak opacity (OP ), fume volume flow
max
rate during period when opacity is exceeded (QH), and fume hood suction
(Qi), are expressed in the equation below (derived from the Lambert-Beer
law, see Section 5.3):
7-17
-------
OP = I - (1 - OPmav)(1 " Ql/QH3 .
iilclX
OP is the opacity of spilled fume when Qt (hood suction) is equal to
nicix
zero (Goodfellow and Bender, 1980). Letting X be the opacity limit and Y
the peak opacity,
(l-OP)(1"Ql/QH) = 1-OP
(l-Qi/QH)xLn(l-OP)y = (1-Q1/QH)Y Ln(l-OP)x
Ln(l-OP)
From Figure 7-4,
(Qi)y = the existing measured suction rate of 212,000 acfm
(Qu)y = the charging plume flow rate of
920,000 acfnrobserved-case 'A' and
1,400,000 aefm-maximum-case 'B1
Note: (QH)X = (QH)y .
From Figure 7-5, OPy = the measured normal opacity 80-percent-case 'A1 and
maximum opacity 97-percent-case 'B'. Finally, the opacity limit is set at
40 percent (OPy = 0.40), and substitution into the derived equation for
case 'A' and 'B1 gives the following:
FOR CASE 'A', OPX = 0.40, Qj = 212,000
OPy = 0,80, QH= 920,000
(Qi/QH)40% = 0-755, Qi = 695,000 acfm
7-18
-------
FOR CASE 'B1, OPX = 0.40, Qt = 212,000
OPy = 0.97 QH = 1,400,000
(Qi/QH)40% = 0.876, Qt = 1,226,000 acfm
Therefore, if plume volumes are only 920,000 acfm and exhaust opacities are
correspondingly low, only 75.5 percent of 920,000 acfm is needed to satisfy
the 40 percent opacity regulation (695,000 acfm). If plume volumes are
1,400,000 acfm only 87.6 percent of 1,400,000 acfm is needed to satisfy the
40 percent opacity regulation (1,226,000 acfm). The same factors (75.5 and
87.6 percent) apply if suction requirements are reduced because of fume
storage allowances.
7.1.4.2 Improve Hood Capture Technology--
The analysis shows that without canopy hood modifications, 695,000
acfm canopy hood suction (three times more than the present exhaust rate)
is needed to capture normal electric furnace charging emissions. This
would ensure that emissions normally have less than 40 percent opacity
within the 4 min/h time limit. It is uncommon to design furnace charging
emission control systems to capture maximum emissions unless fume system
sharing between several furnaces can be achieved, which is not the case
here.
The investigation discussed in Section 7.1.4.1 shows that a modified
canopy hood is needed. The present hood has a storage capacity of less
than 15,000 ft3. A pool-type hood with a hood face area similar to the
present hood has a storage volume of about 45,000 ft3. Such a hood could
achieve high capture efficiency of fume with 520,000 acfm. For 40 percent
allowable charging emission opacity, this volume flow rate could be reduced
to 75.5 percent or about 393,000 acfm.
It is important to note that the large number of assumptions (espe-
cially those regarding peak opacity, spillage characteristics of new vs old
hood, and fume volume flow rates) suggest a safety factor in system sizing,
A factor of 25 percent above the 393,000 acfm lower limit is recommended,
i.e., design the system for 491,000 acfm.
7-19
-------
7.1.4.3 Close Roof Exhaust During Charging--
Closing roof exhaust fans in order to meet environmental regulations
would be attractive for obvious cost reasons. The possibility of closing
the roof exhausters for the 15-min interval following charging and tapping
has severe repercussions. The roof ventilators must be positively closed
and not merely turned off in order to prevent the escape of fume due to
natural draft. Plant ventilation would therefore be curtailed for 30 min
out of every 2-h heat. This would be detrimental to building air and
working conditions. Furthermore, suction demand by the primary fume system
following charging would make the building air problem more severe.
7,1.5 Design Summary
Table 7-2 summarizes the various calculated and measured canopy hood
performance parameters. The problem has been examined using both measured
data and theoretical calculations for opacity predictions.
The study conclusions were as follows:
1. A pool type hood over the furnace charging operation with a
suction of 491,000 acfm is required to satisfy a roof discharge
opacity limit of 40 percent.
2. In order to limit tapping roof emissions to below 20 percent
opacity, a hood extension hung from the crane is required to
improve capture. The hood suction required for charging when
applied during the tapping operation would then certainly produce
an adequate reduction in opacity.
The case study suggests a 491,000 acfm pool canopy hood exhaust capac-
ity is required to satisfy a 40 percent opacity limit. Before undertaking
a new installation, the analytical result should be further refined by
testing the assumptions using a fluid dynamic scale modeling technique.
A survey of canopy hood installations on a similar size electric arc
furnace would show hood suction volumes in excess of 500,000 acfm as typ-
ical (Stiener, 1975). This gives further confidence to the proposed
solution.
It is worth noting that the oversimplified original greenfield calcu-
lation technique was not able to predict the hood suction required for
marginal capture (40 percent opacity) using the hopper type hood (400,000
vs. 695,000 acfm).
7-20
-------
TABLE 7-2. DESIGN SUMMARY
ro
Design parameter Method of reducing secondary emissions to less than 40% capacity
Source
Charging
CASE A
Normal plume
Flow rate Q
Opac i ty
CASE B
Maximum plume
Flow rate
Opacity
Tapping max
Plume flow
Tapping max
Plume opacity
Increase canopy
Characteristic hood exhaust
-Existing hopper hood
15,000 ft3 volume
Qi/QH = 0.755
920,000 acfm Q (suction)
= 695,000 acfm
80%
Qi/QH = 0.876
1,400,000 acfm Q (suction)
= 1,226,000 acfm
97%
470,000 acfm Assume opacity reduced
hood from crane.
40%
Close
Improve hood roof exhaust
capture technology during emission
-Pool type hood with NA
45,000 ft3
491,000 acfm NA
(includes 25% safety
margin)
design for maximum NA
or upset case 1s not
not practical
to less than 20% by existing
NA = Not acceptable to working conditions.
Compare to: Greenfield hood suction prediction 400,000 acfm
As installed design hood suction 216,00 acfra
As installed measured hood suction 212,000 acfra
-------
7.2 CASE II: AIR CURTAIN SYSTEM FOR COPPER CONVERTER SECONDARY EMISSION
CAPTURE
7.2.1 Source Description and Background
7.2.1.1 General-
Case II is an air curtain system installed on a primary copper converter
for capture of low level fugitive emissions. The installation is at ASARCO's
Tacoma Smelter and 1s the first domestic full-scale prototype air curtain
hood on this type of application.
The air curtain capture efficiency was evaluated during an extensive
testing program by PEDCo Environmental, Inc. (PEDCo, 1983). The results of
this program have been used to describe the hood performance in Section
7.2.3.
The original air curtain design calculation was not available for
assessment. Section 7.2.2 presents a design approach for an air curtain
based on application of an analytical technique to the existing site.
7.2.1.2 Converter Operation--
Copper converting is the process of transforming copper matte produced
by a smelting furnace into blister copper. A Pelrce-Smith copper converter
is used and consists of a horizontal refractory-lined steel cylinder (13 ft
diam x 30 ft long) with an opening in the center (called the converter
mouth). The converter vessel is rotated into various positions during its
operation. Figure 7-6 shows the converter position for charging, blowing
and skimming.
7.2.1.3 Converter Emissions—
During converter blowing, oxygen-enriched air is passed through tuyeres
into the shell interior. Emissions generated during blowing are captured
by a primary hood and routed to a sulphur dioxide recovery plant. Fugitive
emissions (not captured by the stationary primary hood) are generated
during converter charging, skimming, and pouring. During a typical 12 h
converter cycle, secondary emission occurrences can total 30 min with an
average duration of 4 min each.
Charging of copper matte and cold scrap is done by an overhead crane
and ladle (a box may be used for scrap). Emissions during charging of cold
scrap are the most severe.
7-22
-------
— FUGITIVE EMISSIONS
I
l\3
OJ
HOOO GATE
RETRACTED
PRIMARY HOOO
HOOO
GATE
TUYERES
BLOWING
CHARGING
(MATTE 8 SCRAP)
SKIMMING a POURING
Figure 7-6. Copper converter operations.
-------
When the converter is rotated out for skimming and charging, the
primary hood gate is raised, and injection of air continues until the
molten bath level is below the tuyeres. Similarly, before the converter is
rotated back to the blowing position, the air is turned on. It is during
these converter movements that significant amounts of off-gases, blown out
by the air, are released thus contributing to the overall fugitive emissions.
The primary hood suction is switched off when air to the tuyeres is
shut off. Therefore, no partial capture of fugitive emissions occurs
through the primary hood.
7.2.1.4 Air Curtain Hooding System—
Four basic methods of converter secondary emission capture exist.
These methods include enclosures, push-pull systems (air curtain), movable
hoods, and fixed hoods. From considerations of effectiveness, reliability,
cost, maintenance, and operating interference, the push-pull systems have
been shown to be superior.
The air curtain as applied to the converter uses the principle of
blowing an air jet across the open space above the fugitive source. Contact
between the rising fumes and the air jet causes the fumes to be directed to
a suction plenum located directly opposite. Figure 7-7 illustrates the
copper converter air curtain system.
7.2.1.5 Regulatory Standards—
Regulations affecting the control of fugitive emissions are under the
jurisdiction of both occupational health and environmental protection
agencies.
Under OSHA regulations, the intent is to provide an in-plant working
environment which is relatively free of contaminants. Without secondary
emission control (with building ventilation as the only means of diluting
contaminant emissions) personnel are regularly exposed to contaminant
levels above the OSHA standard. The contaminants of concern are sulfur
dioxide, copper dust, lead, and inorganic arsenic.
Under EPA regulations, the present intent is to maintain ambient air
quality standards for sulfur dioxide. Sulfur dioxide ground level concen-
trations in violation of ambient standards can result from fumigation by
7-24
-------
FROM AIR JET FAN
-•J
I
INJ
Ol
JET SIDE
EXHAUST SIDE
TO EXHAUST FAN
BAFFLE WALL
TO EXHAUST RAN
BAFFLE WLL
Source: PEDCo, 1983.
Figure 7-7. Converter air curtain control system.
-------
converter fugitive emissions. Opacity regulations, which promote zero
visible emissions from process buildings during all operating conditions,
would also be violated by uncaptured fugitive particulate emissions.
The successful operation of a capture hood for converter secondary
emissions can certainly satisfy indoor regulations, and also ambient re-
gulations 1f captured emissions are cleaned and/or dispersed to acceptable
levels.
7.2.2 DesignApproach
The original air curtain design calculation was not available for this
assessment; therefore, the following design procedure 1s based on applica-
tion of analytical technique to the existing site. This same approach has
been successfully used to design the air curtain hood component of various
electric arc furnace enclosures.
The first step requires determination of the emission plume flow rate
and velocity. The air curtain is then located by careful consideration of
fume source characteristics and converter operating requirements. The
final air curtain configuration is determined by applying the theory of jet
behavior (Section 4.1.3).
Figure 7-8 shows the fugitive emission plumes, originating from charg-
ing and skimming activities, with respect to the as-tested air curtain.
The dimensions and location of the air curtain have been pieced together
from sketchy information but are more than appropriate for verifying the
air curtain design. It is assumed that the jet blows horizontally, although
this may not be the optimum design.
7.2.2.1 Plume Flow Rate and Velocity—
The volume and velocity of fume rising to the air curtain level during
charging and skimming are predicted by using the same procedure as in
Case I, Section 7.1.3.2.2.
Heat release—Heat is released from the following locations:
1. converter mouth
2. hot metal stream
3. surface and sides of the ladle.
7-26
-------
JET SUPPLY | j
BAFFLE PLATE
EXHAUS^
BAFFLE WALLS
CONVERTER
2 INCH WIDE
JET SLOT
OUTLINE
OF EXHAUST -*j
DUCT
SKIMMING
(POURING)
VIRTUAL PLUME
ORIGINS
Figure 7-8, Analysis of air curtain system.
7-27
-------
From fundamental heat transfer calculations, a heat release of approx-
imately 150,000 Btu/m1n is determined for both charging and skimming. The
resulting buoyancy flux is 1.8 * 109 ftVmin3,
P1ume f1ow rate—The vertical distance between the virtual plume
origin (Section 5.1.1) and the air curtain elevation from Figure 7-8 is
21 ft for charging and 38.5 ft for skimming. The plume flow rates are then
calculated to be
charging: Q = 32,000 acfm at T - 480° F
skimming: Q = 90,000 acfm at T = 185° F .
T is calculated by a simple heat balance on the plume volume.
Plume velocity--The mean velocity of the rising air column at the
intersection with the jet elevation is found by
V = Q -r A (plume cross-sectional area at the jet elevation).
For charging, the plume will spread around the ladle as it rises, and
the cross-section area is based on a diameter of 10 ft (an ideal plume with
an entrainment angle of 18 degrees cannot be assumed):
A = 10 x 10 x J = 78 ft2
and V = 32^°° = 410 ft/mi n .
f°r skimming, the plume will spread under the influence of the hot
converter shell, and a cross-section area based on a 15-ft diameter is
assumed:
A = 15 x 15 x 2 = 176
and V = 90^000 = 51Q ft/m-n
The above analytical approach could be supplemented with data collected
from plume photography in the case of an existing site.
7-28
-------
7.2.2.2 Air Curtain Design—
The general principles of optimized air curtain design as applied to
controlling buoyant emissions from a typical metallurgical process are
based on summation of plume momentum (Section 4.1.3). The momentum exerted
by a rising buoyant plume, when added to the momentum of the intercepting
jet, produces a resultant flow direction which must be considered when
locating and sizing the exhaust plenum.
If the nozzle jet is directed horizontally, then the resultant will
always be above the nozzle elevation. Conversely, if the nozzle is pointed
downward at an angle of 15 to 25 degrees, the resultant can be directed
below or at the nozzle elevation. The latter arrangement requires less jet
flow rate and is often the most practical for layout considerations. This
principle was illustrated in Figure 4-3b,
It has been established by theory and experiment that momentum of the
total jet stream is the same at all sections at whatever distance from the
nozzle:
Momentum Flux = (QiXPiXVi) = (Q2)(p2)(V2) .
where
Q = volume flow rate
p = air density
V = velocity
1 and 2 = distances from the nozzle.
For the present case (Figure 7-8) assume angle p - 0 (Figure 4-3), therefore,
Mu = MjCSin 8)
(QU)(PU)(VU) = (Q
Setting 8 = 15 degrees and assuming the worst design case of skimming where
Qu = 90,000 acfm, Vu = 510 ft/min, and py = 0.062 lb/ft3 (185° F), then
(A-XVjXpjXVj) = 11 x 1Q6 ib ft/min2 (since, Q. = (AjXV..)) -
Assuming the jet slot width and length from Figure 7-8, and density, then
Slot width = 2/12 ft
Slot length = 13 ft
7-29
-------
p. = jet air density at ambient temperature of 70° F is 0,075 1b/ft3
A, =2.166 ft2
V. = (67.7 x lO6)1* = 8,228 ft/mln
Q. = 2.166 x 8,228 =17,823 acfm .
M
This compares to 18,000 acfm for the nozzle velocity on the as-tested
prototype air curtain.
Next, the entrained air volume and jet velocity at the receiving hood
are calculated by using the governing equation for a line jet. Volume flow
rate at distance R from the slot is represented by Q., and is estimated by
the equation (Equation 4-15):
Qu = 0.88 ((Q,)(V.) (R/Slot length))55 (Slot length)
n 33
where
Q,, = plume arriving at hood face
Q = jet flow rate at origin
>J
V| = jet velocity at origin
R = distance from slot.
The distance R is established by considering the influence of the
exhaust plenum capture zone and the baffle plate. Entrainment is judged to
occur between the jet and the edge of the baffle plate on the exhaust side.
Beyond that point, entrainment is blocked by the upper baffle plate, and
the plume updraft is captured by the influence of the exhaust off-take
velocity field.
From Figure 7-8, R = 12 ft, therefore
QH = 0.88 x ((17,800/13) x 8,200 x 12)"* x 13 = 133,000 acfm .
In order to capture all of the entrained air, the minimum exhaust
volume would have to be 133,000 acfm. The hood as tested exhausts 126,000
acfm.
Use of the above equation requires the assumption of a small density
difference between the jet air and the air being entrained. In this case
the average updraft temperature, estimated to be 165° F as compared to jet
air assumed to be at an ambient temperature of 70° F, yields densities of
0.062 and 0.075 lb/ft3, respectively, or a difference of about 20 percent.
7-30
-------
The estimate of Qu is therefore approximate. The core of the jet contains
n
most of the intercepted fugitive gas, while the top fringe contains clean
air; therefore, a partial exhaust of 80 percent only may be necessary for
effective capture of fugitives.
By applying different nozzle angles and adjusting slot width, the
overall design can be optimized with respect to minimizing jet and exhaust
capacity. Experience has shown that to avoid excessive noise and energy
consumption by the air jet, the jet slot velocity should not exceed 6,000 ft/
min.
7.2.3 Performance
An estimate of the air curtain capture efficiency and fugitive emission
factors for the overall converter cycle and specific operational modes was
performed by PEDCo Environmental, Inc. under U.S. EPA Contract Nos. 68-03-
2924 and 68-02-3546.
Three separate converter cycles were evaluated during the extensive
test program. Hood capture efficiency was evaluated by three methods:
tracer gas study, visual observations of opacity, and measurement of opac-
ity. Fugitive emission factors were developed from measurement on emis-
sions captured by the hood for the following; sulfur dioxide, particulate,
selected trace elements, and particle size distribution. Table 7-3 sum-
marizes the various hood capture efficiencies and the S02 fugitive emission
factor for the overall cycle and specific modes which are pertinent for
assessing the hood capture performance.
The main conclusions reached by the test program with respect to hood
capture performance are
1. A 90 percent or better fugitive emission capture was claimed
achievable for the overall converter cycle and specific
operating modes.
2. Converter and crane operations are significant variables in
the generation and capture of fugitive emissions.
3. The fugitive emission generation rate is significantly
greater during cold additions and rotating-in/rotating-out
operating modes.
7-31
-------
TABLE 7-3. SUMMARY OF HOOD CAPTURE PERFORMANCE
Hood capture
efficiency (%)
by tracer
gas study
Operating mode
Matte charge
Cold additions
Slag skimming
Copper pour
Rotate- 1n/ rotate-out
Blow/idle
Overal 1
A
94
99
95
89
__
96
94.6
B
62
62
84
81
—
44
66.6
Hood capture effective-
ness (%) by visual
observation of opacity
Observer 1
94
95
78
92
77
96
88.6
Observer 2
91
85
82
85
76
90
84.8
Measured
opacity (%)
above hood
14
21
18
9
--
—
15.5
Sulfur dioxide emission
Average
Ibs event
9.5
32.0
11.0
7.4
23
3.5
—
Average
Ibs/rnin
2.19
7.38
2.35
1.94
6.15
0.14
—
A = Tracer gas injection in upper control volume.
B = Tracer gas injection in lower control volume.
-------
7.2.3.1 Tracer Gas Study--
Tracer gas methodology, using sulfur hexafluoride (SF6) as the tracer
gas, was shown to be a feasible means of estimating the air curtain capture
efficiency. This technique was used to establish the air curtain volume
and to determine the effect of converter operation as a variable on the air
curtain efficiency. During tracer gas tests, the exhaust hood flow rate
was set at 126,230 acfm during converter rotate-out activities and 75,500
acfm during blowing and idling.
Tracer recovery tests of the air curtain hood system were performed by
injecting tracer gas in the area immediately above the converter in lower
portions of the control volume, as shown in Figure 7-9, A summary of test
results is listed on Table 7-3. The results have been interpreted as
follows:
1. For the upper volume tests (tracer injection point shown in
Figure 7-9), the converter operating mode had no adverse
effect on the tracer recovery efficiency. An overall recov-
ery of 94,6 percent was determined for this test case. The
high recovery efficiency indicates that the air curtain is
very effective in capturing fugitive emissions that pass
directly under the air curtain. However, tracer gas injected
into the upper control volume does not account for spillage
outside the control area. Therefore, upper control volume
tests do not provide a direct measurement of hood capture
efficiency for operations which spill fume outside this
upper injection zone.
2. For lower control volume tests, the converter operating mode
had a definite effect on the tracer recovery efficiency.
This effect was mostly caused by the location of the tracer
injection nozzle with respect to the emission source.
During charging, the injection probe was located below the
source, while during skimming and pouring the probe was
located above the source. The location of the thermally
driven plume source, which contains the fugitive emission,
with respect to the tracer injection point plays a signif-
icant role in affecting tracer recovery efficiency. During
the blow/idle mode the hood exhaust rate was reduced to the
lower setting and therefore explains the low tracer value of
44 percent.
7.2.3.2 Visual Observations of Opacity--
Two trained independent observers characterized hood performance by
estimating overall hood capture effectiveness, approximate opacity, dura-
tion, and significance of any visible emissions observed.
7-33
-------
UJ
PRIMARY HOOD
en
3
x
u
•CONVERTER
TOP VIEW
—SAMPLING LOCATION
JET SIDE
EXHAUST SIDE
AIR
CURTAIN
JET
BAFFLE WALL
TO EXHAUST FAN
UPPER CONTROL
VOLUME TESTS
BAFFLE WALL
LOWER CONTROL
VOLUME TESTS
ELEVATION
Source: PEDCo, 1983.
Figure 7-9. Sulfur hexafluoride injection locations.
7-34
-------
A record of hood capture effectiveness for the various modes is listed
in Table 7-3. As expected, quantitative values of effectiveness for some
modes of operation differ between the two observers. Overall, the values
are in agreement.
Visual emission observations revealed how converter and crane opera-
tions introduced significant variability in hood capture efficiency.
During skimming, ladle position and rate of slag discharge affected the
hood capture performance. Rapid slag discharge into a ladle placed on the
ground resulted in considerable fume spillage into the converter aisle. If
the ladle was held by the crane and slag discharged slowly, capture perform-
ance increased considerably. Tracer recovery tests performed in the upper
control volume could not distinguish between the two modes of slag discharge.
For this reason, visual observations were used in conjunction with the
tracer to quantify capture effectiveness.
Fume spillage during the rotate-in and rotate-out operation was also
significant but could not be detected with the tracer method.
7.2.3.3 Measured Opacity-
The opacity of emissions escaping the air curtain were monitored at a
point above the hood (but inside the building) and recorded. Although it
is difficult to correlate hood efficiency/effectiveness with the opacities
recorded by the transmissometer, a judgment can be made when considering
the visibility of any spilled fume discharging through the converter roof
ventilators.
Table 7-3 shows a peak opacity value of 21 percent above the air
curtain hood during cold additions. Considering that dilution with con-
verter building air occurs while the emission rises to roof level, the roof
discharge opacity would be expected to be much less than 20 percent.
Therefore, with respect to discharge opacity to the environment, the hood
effectiveness is judged to be adequate,
7.2.3.4 Discussion—
The fugitive emission rate varies greatly for the various converter
operation modes. This is illustrated by the S02 emission values listed in
Table 7-3. The emission rates for cold additions and rotate-in/rotate-out
7-35
-------
modes are approximately three times greater than for matte charging, slag
skimming, and copper pouring. This relationship also applies to particulate
fugitive emissions if a constant ratio of particulate to S02 in the gas
stream is assumed.
In general, the rate of fugitive emission from the converter is propor-
tional to the heat released to the plume carrying the emission. In terms
of hood capture performance, the worst cases (most difficult to capture)
are therefore cold additions and rotate-in/rotate-out operations. (The
plume momentum arriving at the air curtain increases with heat release
which in turn increases the air curtain requirement to overcome this force.)
The measured and observed capture performance for cold additions is
excellent, whereas for rotate-in/rotate-out, capture is significantly less.
For cold additions, the fume source is directly under the air curtain,
whereas the converter mouth during rotate-1n/rotate-out is remote from the
effect of the air curtain, as is the skimming operation. Although the
overall performance of the air curtain was judged to be adequate, areas of
Improvement could be considered for the rotate-in/rotate-out and converter
skimming operating modes.
7.3 CASE III: BASIC OXYGEN FURNACE SECONDARY FUME CAPTURE
7.3.1 Source Description and Background
7.3.1.1 General--
Case III Is secondary fume control system on two 250-ton (230 metric
ton) basic oxygen furnaces (BOF). The major reference for this case is a
published paper by Schuldt et al. (1981).
BOF secondary emissions are generated during transfer of blast furnace
molten iron between vessels (reladHng), charging of molten iron and scrap
into the refining vessel, and slagging and tapping of steel. Oxygen blowing
can also cause secondary emissions due to splashing slag at the vessel
mouth caused by the boil within the vessel. These emissions are captured
by local hooding with the secondary ventilation system (SVS). Process
gases generated during steel making are handled in a separate particulate
removal facility.
7-36
-------
Of particular interest in this example is the design approach used in
sizing the capture system. Similar plants in Western Europe and Japan have
successfully captured secondary emissions by using local hooding only, and
local hooding plus partial building evacuation (Coy and Jablin, 1979), With
an appreciation of how key design parameters affected the system size, a
survey of existing installations and capture technology was used as the
basic design tool. The success of this described approach has been proven
in practice. The system of local hooding for Case III performed better
than expected.
7.3.1.2 Regulatory Standards--
The plant is situated in a new industrial area, A zero visible emis-
sion standard was part of a stringent environmental design requirement for
this area. High priority was also given to the workplace environment.
Therefore, in order to comply with both outdoor and indoor requirements,
fume source capture efficiencies approaching 100 percent had to be achieved.
As a result, BQF secondary emissions control received high priority as part
of the environmental control strategy for a greenfield facility.
7.3.2 Design Approach
7.3.2.1 Nature of BQF Secondary Emissions—
The major sources of secondary BOF shop emissions are
1. Charging (molten iron/scrap)
2. Tapping
3. Slagging
4. Puffing
5. Molten Iron reladling.
Charging—Fume is generated during the charging of molten iron into a
furnace that already contains scrap. Figure 7-10 illustrates the fume
generation sources for the BOF vessel operation. The following mechanism
may produce fume during BOF charging:
1. Entrained air which enters the vessel with the molten iron and
oxidizes the charge
7-37
-------
FURNACE
CHARGES
HOTA1R,CQ,C02
AND PARTICULATE
AISLE
CRANE QIRDER
POSITION OF
LOCAL HOOD
FOR CHARGING
FUME CAPTURE
SCRAP
CHARGING LAOLE
CHARGING
FLOOR
Source: Schuldt et al., 1981. (Reproduced with permission.)
Figure 7-10, BOF charging fume generation process and position of local
capture hood.
7-38
-------
2. Iron oxide scale on scrap reacting with molten iron
3. Combustion of oil or other materials mixed with the scrap.
Important variables which affect the off-gas evolution rate are
1. Molten iron charging rate
2. Scrap composition (Fe20g, oil, moisture, bulk density)
3. Molten iron composition (carbon, silicon)
4. Molten iron/scrap ratio
5. Slag retained in vessel
6. Amount of slag retained with molten iron
7. BOF vessel temperature.
Calculations can be carried out to estimate gas volumes, gas composi-
tions, and temperatures at the vessel mouth. Depending on the assumptions,
a wide range of flow rates can be estimated. Although the calculation
procedures indicate sensitivities of off-gas flows to changes in specific
parameters, at the time of design, it was clear that calculation techniques
had not reached a level of sophistication where one could consider establish-
ing system volumes with absolute confidence.
It is difficult to establish charging hood volumes because of the
following;
1. It is a combustion process and hence one must account for turbu-
lence, residence time, degree of mixing, temperature, and percent
combustibles in the gas.
2. Charging occurs over a short period of time, and gasflows and
temperatures fluctuate rapidly. Hence, transients, not steady-
state conditions, are important. This makes analysis more complex.
The designer of a secondary fume system must clearly recognize that
the basic system design parameters must adequately account for the com-
bustion process in terms of temperature, flow, and oxygen levels in the
gases. Three important design considerations are residence time, extent
and type of refractory lining for the ducting and hooding near the vessel
mouth, allowance for thermal expansion of hood and ducting, and safety
aspects to eliminate explosion concerns.
7-39
-------
Provided that the off-gas contains .excess air, the combustion charac-
teristics are then dependent on time, temperature, and turbulence. The
hood and off-take configuration will enhance mixing or turbulence. Suffi-
cient mixing to support combustion is usually achieved with normal hood
geometry. Temperature of the off-gas at the charging hood is normally high
enough to support combustion.
For metallurgical processes such as those generating carbon monoxide
and when off-gases are exhausted through ducts, a conservative design resi-
dence time for complete combustion is 0.2 to 0.3 s. Typical residence
times calculated for the refractory lined combustion section of SVS systems
have been found to be 0.75 to 1.0 s. This healthy safety factor is required
because of rapid surges which occur in the fume generation rate during
charging. The safety factor ensures that during these surges the refractory
section is long enough to protect downstream steel ducting from high tempera-
tures.
An important design criterion for the charging process is to ensure
that there is always an excess amount of combustion air. A single hood
off-take has the advantage of helping to promote combustion. Mixing of the
combustion air with combustibles occurs in the same duct. In comparison, a
system with two off-takes may result in one off-take carrying a CD-rich gas
while the other contains primarily air. At the point where these flows
combine, ignition has been known to occur with explosive force.
Another prime factor in fume generation is the rate of pouring molten
iron into the vessel (the faster the pour, the higher the fume generation
rate). It is common to specify the maximum allowable pouring rate in order
to identify the system limits. From an operating point of view, this
usually raeans a compromise.
Tapping—During tapping operations, fume evolution is normally fairly
steady; however, if ladle additions such as ferrosilicon or ferromanganese
are made, the fume generation may be higher by a factor of two.
S1aggi'ng--Also during slagging operations, fume generation can vary
widely. Factors such as steel grade, slag volume, and use of additives
strongly influence fume release. Also, slagging fumes tend to be relatively
7-40
-------
cold. They have minimal buoyancy. This often makes them difficult to
capture in an over-head canopy.
Puffing—Another source of EOF secondary emissions is puffing. Puffing
results from short-lived pressure pulses during oxygen blowing. With an
adequately designed primary fume system, these puffs produce a small amount
of fume. The secondary ventilation system plays an important role in
capturing these puffing emissions, and control may be readily incorporated
into the system design.
Molten iron reladling—Finally, molten iron reladling from a torpedo
car is another source of secondary emissions requiring careful attention in
a BOF shop. Experience has shown that the amount of exhaust volume required
to control these emissions with local exhaust ventilation is primarily a
function of the degree of enclosure of the transfer point. With a tight-
fitting hood, exhaust volumes can be kept to a minimum. The rate of molten
iron transfer is a factor as well but is of less importance.
7.3.2.2 Review of Secondary BOF Fume Control Technology—
Table 7-4 summarizes system design data available from the literature
on recent secondary ventilation systems and compares it to the actual Case
III installation. Up until 1978, Japanese steel plants had the largest
secondary fume system in operation. More recently, one installation in the
United States, which started up in 1978 with a rated capacity of 600,000 acfm,
is marginally larger. A full description of the Italsider system, operating
in Italy, is contained in Coy and Jablin (1979).
In order to have a common denominator for fume system size compari-
sons, it is convenient to consider a basic shop parameter such as heat
size. Figure 7-11 shows a plot of charging hood volume versus heat size.
The Fukuyama system was the basis for the Case III design.
The other Important parameter is the total heat content of the secon-
dary ventilation gases after combustion. It dictates the amount of cooling
required to lower off-gas temperatures to an acceptable level for gas
cleaning by a fabric filter (baghouse). Figure 7-12 is a plot of charging
off-gas heat content versus heat size for the data in Table 7-4.
7-41
-------
HOT METAL CHARGE TIME
O 40 sec
A 65 sec
A 120 sec
O 240 sec
• 300 sec
O Unknown
e
I
n
E
1
B*
3
"3
M
«
3=
O
w
"§
CO
U
* _ . , „. _
-------
HOT METAL CHARGE TIME
O 40 sec
± 35 sec
A 120 sec
<> 240 sec
• 300 sec
Q Unknown
0.5 r-
0.3
1
o
u
**
as
01
x
31 0.2
O
S)
"s
6
0.1
Kaiser
A
Kimitsu 2 •
Led O
Oita
O Fukuyama
Hilton
Youngstown
O Italsider
inland
D
100
200
300
400
Mt/Heat
Source: Schuldt et al., 1981. Heat Size
(Reproduced with permission.)
Figure 7-12. SVS charging off-gas heat content vs. heat size.
7-43
-------
TABLE 7-4. SVS SVSTEH EXHAUST DATA
Start-up
Plant date
Stelco LED
Fukuyama
OITA
Kiraitsu #2
Inland
Stelco Hilton
Youngs town
Italsider,
Taranto
Bethlehem
Kaiser
Fon tana
I960
1970
1972
1971
1974
1971
1973
1969
1979
BOF
No.
2
2
2
2
2
3
2
3
2
M.T. s
230 40
300 40
300 300
220
200
114 65
240
350 240
200 120
Charqe
fli3/ni1n
10,000
10,000
11,300
11,200
Canopy
(7,800)
6,120
4,250
8,300
12,750
°C
200
200
200
200
95
315
15
90
200
Tapping
Gcal/mln nra/«Hn °C
0.
0.
0.
316 10,000
316 5,000 150
357 8,400 80
0.354
(0.
0.
*(0.
*0.
0.
127)
260 Vessel hood
220)
244
403
Re ladling
nrVmin
6,000
6,000
10,100
3,700
To snail
3,500
Separate
3,000
4,500
3,000
4,250
Of.
150
150
150
60
95
filter
120
15
120
200
Gcal/min
0.151
0.151
0.254
0.033
O.OS7
0.062
*0. 167
0.062
0.134
Other Total
m3/min °C ma/m1n °C
16,000 135
skimming 16,000 150
4,000
desul. , deslag
9,600 0 14,500 87
desul. , deslag
8,350 0 12,600 130
11,300 120
desul. 16,600 90
(for 2 vessels)
17,000 200
*Assumed values
-------
7,3.2,3 Selection of Hood Capture System-
It Is important to recognize that the performance of the charging hood
(capture effeciency for a given hood suction) is influenced by scrap quality
(cleanliness and bulk density), hot metal pouring rate, and geometry. This
makes it difficult to guarantee the performance of the total system if hood
suction is adopted from an installation and applied without considering the
other influencing factors.
Although vessel size is being used as a common factor for comparing
hood capture systems, it is the amount of hot metal and scrap charged and
their chemistry which are the important variables. By using vessel size,
it is assumed that the metallurgical practice is similar for most of the
BOF operations surveyed, (e.g., the full vessel weight capacity is used and
charged with 30 percent scrap and 70 percent molten iron.) It is also
assumed that the molten iron is added in one charge. Note that Figures 7-11
and 7-12 were prepared to establish a design benchmark to help make an
engineering decision. The graphs were not intended to directly correlate
hood suction and heat release to vessel heat size.
The design of the secondary ventilation system was a compromise of a
number of objectives set by operators, designers, and suppliers of equipment.
These objectives include
1. Desire to use all types of scrap
2. Maximum possible charging rate
3. Avoidance of explosions
4. High capture efficiency
5. Cost-effectiveness
6. Tight performance guarantees.
The two main steps leading to the selection of the hood capture system
for BOF charging by using other systems' design data are as follows.
Step 1 - Compare Magnitude of Emission Source—The two main factors
affecting the magnitude of the emission source (velocity, flow rate, and
temperature), are vessel size and hot metal charge time. A logical compari-
son for Case III operation is the Fukuyama plant. The vessel size is
7-45
-------
similar, while the desired hot metal charge time is identical (Figure 7-11).
To ensure similar capture performance, the hood geometry with respect to
vessel mouth must be constructed similarly.
Step 2 - Compare Off-Gas Heat Content—The off-gas temperature is
important in specifying the gas cleaning equipment. If a fabric filter
(baghouse) is used, with polyester bags, for example, the gas must be kept
below 275 °F (135 °C) at the filter.
The off-gas heat content for hot metal charging must be estimated to
predict the off-gas temperature at a specific hood suction flow rate. The
main factors affecting heat release are again vessel size and hot metal
charge time. Figure 7-12, constructed from information in Table 7-4,
displays a range of heat release values for the BOF hood installations.
The Case III heat release was similar to the Fukuyama operation, based on
identical charge time requirements and similar vessel size.
7.3.2.4 Capture Hooding—
The BOF charging fume emission is captured by a refractory lined local
hood positioned over the ladle as shown in Figures 7-13 and 7-14. Tapping,
slagging, and puffing emissions are captured by a semi-enclosure formed
around the furnace by heat shield partitions. The partitions extend down
to slag and tap ladles, which help direct fume up into the semi-enclosure.
Above the charge floor, the enclosure is open on the tap and charge sides.
Suction for these operations is provided through the main charging hood
off-take at the rate of 350,000 acfm.
The molten iron reladling operation is partially enclosed by a three-
sided fume hood as shown in Figure 7-15. The hood sits over the ladle and
accepts molten iron from a torpedo car on the open side. The top of the
hood is closed and serves as the off-take. A 212,000-acfm suction volume
is applied to this hood. The integrated secondary ventilation system is
shown in Figure 7-16.
7.3.3 Performance
The charging hood performs better than expected as shown in Figure 7-13.
When charging 176 tons of molten iron in 40 s, nearly all of the fume is
captured. For practical purposes, all fume is effectively captured when
charging at a faster rate of about 30 s. It should be noted that the
7-46
-------
Source: Schuidt et al., 1981. (Reproduced with permission.)
Figure 7-13, Charging emissions from a BOF furnace.
-------
00
Source: Schuidt et al., 1981. (Reproduced with permission.)
Figure 7-14. Semi-enclosure capturing tapping, slagging, and puffing emissions
from a BQF furnace.
-------
BO.F-H0t Metal .Rejadltng. Station
Offtake to
Duct
Torpedo
Car
Hot Metal
le
Source: Schuldt et al., 1B81. (Reproduced with permission.)
Figure 7-15. Fume hood arrangement for capture of BOF hot metal
relading emissions.
7-49
-------
tn
O
Primary
Hood I
SVS Hood,
16000 m3/min, 135° C
I Emergency
3 f „ Relief Vent
Fabric Filter
V
V
V
-LK
10000 m3/min
200° C
Primary
Hood
\
, 6000 m3/min, 150° C
B.O.F.
No. 2
Movable Hood
Source: Schuldt et al., 1981. (Reproduced with permission.)
Figure 7-16. Integrated secondary ventilation system for the BOF.
-------
molten iron transfer rate of 5.8 tons/s with complete capture of emissions
is probably the best 1n the industry.
The Integrated secondary ventilation system (Figure 7-16) is well-suited
for the steelmaking shop. The system is capable of handling process varia-
tions, and it is remarkably efficient in capturing secondary emissions.
Visually, it is estimated that nearly all of the reladling emissions are
captured while the vessel hood is more than 95 percent effective.
Furthermore, because fume capture was treated as a combustion process
as well, problems with combustibles have so far not materialized. Measure-
ments have shown an abundance of excess air, and there is evidence that the
design promotes rapid combustion and dilution of exhaust gases. Combustibles
are low throughout the system, and, as a result, potentially explosive
conditions have not been encountered.
7.4 CASE IV: CHARGING AND TAPPING CANOPY HOOD FOR AN ELECTRIC ARC FURNACE
This case study examines another canopy hood system for capture of
charging and tapping fumes from an electric arc furnace. The original
design basis is provided, and the included results of recent performance
tests suggest excellent capture efficiency,
7.4.1 Canopy Hood Design
The meltshop under consideration contains two ultra-high-power elec-
tric arc furnaces with capacities of 115 and 150 tons. The 150-ton furnace
was added to the existing 115-ton furnace to increase shop capacity. It
was commissioned in December 1981.
Direct evacuation is used to control emissions from the furnaces
during melting and refining. The canopy hood system shown in Figure 7-17
is used to capture process fugitive emissions during charging and tapping
of the 150-ton furnace. Emissions from the furnaces are ducted separately
to a mixing chamber and then to baghouses. With the installation of the
newer furnace, baghouse capacity was increased by incorporating a negative-
pressure pulse-jet baghouse into the air pollution control system.
The canopy hood system geometry was based on the designer's observa-
tions of one working system (Walli et a!., 1983). The working hood system
was deep with 60 degree sides. This feature was included in the present
7-51
-------
in
ro
Scavenger Duct
Ladle Crane
Ladla Crane
50.7 ft.
r "* -i -,50 ton
1 f Furnace
Partition
Figure 7-17, Canopy hood arrangement for capture of fugitive emissions
from the electric arc furnace.
-------
design as shown in Figure 7-17, From the discussion in Section 5.1.2, it
might be anticipated that the 60 degree sides would produce a hood with
storage capacity greater than a shallower hopper-type hood, thereby reducing
plume spillage. The width of the hood was determined by projecting a line
15 degrees from the vertical, from the furnace roof ring and ladle lip to
the desired height of the canopy hood (Walli et al,, 1983). Selected hood
face dimensions were 72 x so ft. Design exhaust rate was determined by
multiplying a nominal face velocity of 150 ft/min by the hood face area
resulting in a value of 650,000 acfm.
Other features of this system include solid baffles and a scavenger
duct system shown in Figure 7-17. The scavenger duct system was installed
at the request of the State regulatory agency who reviewed the design. The
solid baffles are sheet metal partitions suspended from the meltshop roof
to the level of the crane on purlins. The purpose of the baffles is to
create a secondary collection zone around the hood and furnace. The scav-
enger ducts located on either side of the canopy hood contain 20 Hp fans.
Any emissions that escape the canopy hood are caught in the secondary
collection zone and returned to the canopy hood by the fans. From the
discussion in Section 5.1.3, it might be expected that the baffles also
reduce the effects of building cross-drafts.
7.4.2 Hood Performance
Recent tests of this canopy hood system indicated that the design
performs quite well: over two days of testing, the highest 15-s interval
opacity observed at the roof vent was 15 percent, and the highest 6-min
average opacity was 3.5 percent (Terry, 1982). Operating exhaust rates
were 550,000 acfm through the canopy and 50,000 acfm through the scavenger
ducts.
It is tempting to perform simple calculation's to estimate the required
exhaust rate for this system as in the case study in Section 7.1. For
example, assuming a rate of 10° F/min for the temperature drop of ladle and
furnace and for effective height, Z = 99 ft. , an estimated exhaust rate of
Q = 520,000 acfm results. Although this calculation might suggest a cor-
rect order-of-magnitude estimate, it is not really appropriate. This is
partly because the temperature drop is assumed and not measured; but more
7-53
-------
importantly, the calculation is inappropriate because the effects of ob-
structions (cranes), intermittent plumes (charging), and site-specific
features are not taken into account. Detailed examination of these fac-
tors, as shown in Section 7.1, is quite involved.
7.5 CASE V: DUST CONTROL FOR CLAMSHELL LIME UNLOADER HOPPER
7.5.1 SourceDescription and Background
Case V design review Involves dust control on lime transfer by a 15-
ton capacity clamshell into an enclosed hopper. This case is an example of
fugitive particulate control on a nonbuoyant source. The source is typical
for bulk materials handling at receiving terminals throughout Industry.
Large amounts of loose material is handled in the open, thus making control
of dust generation and dispersion a constant challenge.
The major reference for this case is Gilbert et al. (1984). The paper
describes a modeling technique used to improve capture of lime dust from
the clamshell unloading operation. To design an accurate physical model,
it was necessary to identify important variables that were affecting the
fugitive emission problem. The paper contains a detailed account of the
variables affecting performance, which makes it an excellent reference for
demonstrating the design aspects for this type of nonbuoyant source. The
paper also has a qualitative description of performance before and after
modifications to the hood system.
7.5,1.1 Lime Unloading Operation—
The lime unloading operation consists of using a clamshell to unload a
barge. The lime is carried by the clamshell onto an enclosed unloader
hopper and dropped. From this transfer point, the lime is carried by
conveyors to storage silos.
Figure 7-18 illustrates the lime dumping hood. A three sided en-
closure contains the discharge area over the hopper. The top is fitted
with a slot for the clamshell trolley. In the original design, the exhaust
duct to the dust collection baghouse is located at the enclosure midpoint.
7.5.1.2 Description of Fugitive Emissions—
The following sections, which describe the fugitive emissions, are
taken directly from the referenced paper;
7-54
-------
Trolley and Clamshell
Wind
X
Operator Cab
vw
Baghouse Flow
Grizzly Bar System
Legend
A Air
B Baghouse
C Clamshell
D Drag
F Field
G At grizzly
L Lime or Sand
M Model
W Wind
V Velocity
Source: Gilbert«ai., 1984 (Reproduced with permission.)
Figure 7-18. Three regions of lime drop flow patterns to be modeled.
7-55
-------
During the lime unloading operation when the clamshell is
dumped into the hopper inside the enclosure, fugitive emissions
of Hme dust can sometimes be seen escaping over the front lip of
the hopper, escaping at the middle and upper elevation out the
front of the enclosure, escaping through the open trolley slot at
the top of the enclosure, and/or pulled out in the wake of the
clamshell. There are many variables that effect the flow patterns
inside the hopper and the enclosure to cause these fugitive
emissions.
There are several important characteristics of the flow
patterns and dust generation that are obvious from watching the
field unit in operation. Almost all of the entrained lime dust
comes up out of the hopper from below the grizzly starting about
1 to 2 sec after the lime starts to fall through the grizzly.
The amount of dust, the plume velocity, and the region where it
comes up out of the grizzly depend on where the load was dropped,
how large a load was dropped, and the elevation of the clamshell
above the grizzly. The plume travels upward in the enclosure and
sometimes directly out the front of the enclosure. As the plume
rises in the enclosure, it is caught by the wind swirl patterns
and carried higher in the enclosure where it can escape through
the front or out of the trolley slot at the top of the enclosure.
As the plume rises it may move in front of the clamshell, into
the clamshell, in back of the clamshell, or to the sides of the
clamshell depending on where the drop was made. Because the
clamshell is brought out of the enclosure as soon as it is empty,
it will generally push or carry out lime dust as it exits from
the enclosure. From field observations, it was also obvious that
a full clamshell load drop produced more dust in the enclosure
than a partially full clamshell. For a severe dust generation
drop, it would take 30 to 40 seconds for the enclosure exhaust
flow to clear the enclosure of airborne dust.
7.5.2 Design Approach
Cost-effective control of dust problems arising from bulk materials
handling requires an initial examination of the overall handling. Factors
influencing dust generation and dispersion must be understood in order to
achieve a proper design.
A number of steps can be taken to minimize dust generation and disper-
sion. For the clamshell case, an active containment design was pursued for
minimizing dispersion. Active containment relies upon an inflow of air
into some type of enclosure (Section 6.1).
A list of important variables affecting dust control during clamshell
unloading was established in the referenced paper as follows:
7-56
-------
1. Saghouse exhaust flow rate
2. Wind direction and velocity
3. Height of lime drop
4, Location of clamshell In enclosure
5. Amount of lime in clamshell
6, Amount of lime in hopper
7. Rate of clamshell opening
8. Dwell time of clam 1n enclosure
9. Location of enclosure ventilation openings
10, Degree of material dampness
11. Enclosure open area control velocity.
7,5.2.1 Original Design-
Original design calculations for this example were not available.
Control velocities on enclosures are generally recommended at 100 to 200
ft/min by dust control design manuals. For the original design, a 60,000 acfm
exhaust flow induced an inward velocity of 96 ft/min through the enclosure
entrance and trolley slots. This was not sufficient to overcome plume
trajectories aimed outward or to overcome the effect of moderate wind
levels.
7.5.2.2 Modified Design—
A design based on the enclosure open area control velocity does not
consider all the other variables listed as affecting dust control. Calcula-
tion procedures to predict many of the other variables would be very com-
plicated, if not impossible, to perform. Physical modeling of the problem
and solution was therefore used as the basic design tool.
The modeling procedure is described in Gilbert et al. (1984). A
one-sixth scale model of the unloader hopper was selected so that flow
patterns in the enclosure could be evaluated. Smoke was used to simulate
the behaviour of the lime dust in the enclosure. Since the lime dust was
relatively fine (mass median diameter less than 13 urn), submicron smoke was
a conservative representation. The lime drop from the clamshell was simu-
lated by releasing coarse sand, thus modeling the flow patterns caused by
the volume displacement and the air entrainment. The effect of local wind
7-57
-------
direction and magnitude on the enclosure was simulated by common window
fans, A total of 26 tests were run and documented photographically by two
synchronized cameras.
Conclusions concerning the causes of the fugitive emissions were
developed from extensive model testing. The emissions escaped from the
enclosure by direct plume trajectory and by wind dispersion. Lime dropped
into the back of the grizzly (steel grate of rectangular openings) created
a plume towards the front of the enclosure. A drop near the front-produced
a plume to the rear. The plume was caused by the rapid displacement of air
and dust from the hopper. Winds were found to create a vortex inside the
enclosure that drew dust high up 1n the enclosure and out the front.
Conclusions concerning the elimination of fugitive dust escape were
also developed from model testing. The baghouse capacity of 60,000 acfm 1s
sufficient to capture most of the emission by implementing the following
remedies.
1. Capture of dust is improved by repositioning the exhaust duct at
a lower elevation closer to the grizzly. The original location
of the exhaust duct at a high elevation tended to draw dust up
toward the clamshell and its wake.
2. By dropping lime in front of the hopper the dust plume is directed
to the back where a baffled off-take effectively captures the
lime dust.
3. A downward flowing exhaust through the grizzly and into the
hopper directly counteracts the plume velocity.
4. Slow opening of the loaded clamshell at low elevations minimizes
emissions.
The final recommended configuration for improving dust capture is
shown in Figure 7-19. The design change was rather simple and the model
test showed a significant reduction in visible fugitive emission.
7.5.2.3 Discussion—
This design review example has illustrated the following points:
I. The dust plume results from the creation of local air flow caused
by dlsplacment of air and dust from the hopper by the lime dumping.
7-58
-------
18*
Sloped Roof
26'
Solid Baffle
Raised 6" Off
Grizzly
8'3-11/16"
Opening Size
r x ir
Source: Gilbert etal., 1984 (Reproduced with permission.)
Figure 7-19. Geometry of final configuration: baghouse flow is drawn from back of
hopper under single baffle, which is raised off grizzly.
7-59
-------
2. Winds had a significant effect on fugitive emission releases.
Emissions increased with increasing velocity and depended on the
direction of the wind.
3. Capture system performance on a nonbuoyant source is influenced
by enclosure (hood) design and location of the exhaust point.
In the present example, by understanding the factors influencing dust
generation and dispersion, a useful rule-of-thumb may be inferred that the
control velocity should be applied through the grizzly by exhausting from
the hopper.
7.5.3 Performance
The modifications shown in Figure 7-19 were installed in the field
unit. Reports from field unit operators and observers indicated that the
significant improvement shown by the model tests is realized in the field.
The fluid modeling technique has thus been proven as a useful design tool.
7.6 CASE VI: PARTIAL ENCLOSURE TO CONTROL ALUMINUM ROLLING MILL EMISSIONS
The following case study examines the use of a hood assisted by an air
curtain to control emissions from an aluminum rolling mill. Although the
example does not represent an actual installation, dimensions and conditions
are typical of a single-stand cold rolling mill. The authors are indebted
to Busch Co. for providing this case study (Perryman, 1984).
7.6.1 Natureof Process Source and Hood Selection
Aluminum rolling mills are used to reduce the thickness of aluminum
sheet. Both hot and cold rolling mills require that a fluid be applied to
the strip to serve as both a lubricant and a coolant. In cold rolling
mills, a mineral oil coolant similar to kerosene is used. In hot rolling
mills, the coolant is usually a very dilute oil and water emulsion. In
both mills, the rotary movement of the rolls and linear movement of the
strip generate fine liquid particles (mechanical atomization). Also,
rolling the metal generates sufficient heat by friction to vaporize a
fraction of the coolant. Coolant particles are objectionable because of
worker exposure to hydrocarbons, reduced in-plant visibility, and potential
fire hazards. Because of the differences in coolants, cold mills usually
have some form of hooding; hot mills often are uncontrolled.
7-60
-------
The hood design depicted in Figure 7-20 is used for both hot and cold
rolling mills. This hood design is difficult to classify within the scheme
used in this manual but is probably best defined as a partial enclosure.
Thi manufacturer refers to it as a slotted-perimeter hood assisted by an
air curtain (Roos, 1981). In contrast to the case study in Section 7.2,
the air curtain shown in Figure 7-20 does not direct the emissions into the
hood, but rather serves to contain the emissions and deflect, unwanted air
currents. It should be borne in mind that this design evolved from modifi-
cations to simpler exterior hoods, which often were not very effective.
7.6.2 Design Procedure
The following example calculation indicates the design procedure for
an assisted slotted-perimeter hood for a single-stand aluminum cold rolling
mill. The required exhaust rate and hood dimensions are calculated by a
rule-of-thumb method (ACGIH, 1976) modified for this application; the air
entrained by the air curtain is estimated by a procedure in Hemeon (1963).
The conceptual layout of the hood design 1s shown in Figure 7-21. For
the exit hood, the following source dimensions'are needed:
1. Width of metal strip being rolled (B) =3.0 ft.
2. Height of bottom of hood above passline (D) = 4.0 ft.
3. Distance between rewind reel and face of housing posts (L) = 12.0
ft.
4. Metal coil diameter (C) =6.0 ft.
5. Width of mill inside housing posts = 5.0 ft.
6. Width of mill outside housing posts =6.5 ft.
7. Height of passline above mill floor level = 3.5 ft.
From these source dimensions, the hood dimensions are calculated as follows
(ACGIH, 1976). The hood width is taken as 80 percent of the hood height
above the passline plus the source (strip) width:
Hood Width = 0.8 D + B
= 0.8 (4.0) + 3.0
=6.2 ft.
7-61
-------
Air Curtain
I
CTJ
rv>
Exhaust
Isovel Pattern
Aluminum Strip
at Pass Line
Source: Roos, 1981 (Reproduced with permission.)
Figure 7-20. Schematic cross section of an air-curtain hood. Air jets prevent
fumes from exfiltrattng into work areas surrounding mill.
-------
DRIVE SIDE
-OPEN POSITION
" REWIND
r
WINDOW
f — '
1
, CD _ _
--!•
L
1 I
1 A JU
, M._Lj
1
/
I
-B-
• EXIT HOOD
STRIP
TRAVEL
C3OOQ FPMJ
CLOSURE
OPgRATOR'S SIDE
PLAN
MILL HOUSING
MILL
FLOOR
ELEVATION
Source: Roos, 1981 (Reproduced with permission.)
Figure 7-21. Example perimeter hood for control of aluminum rolling mill emissions.
-------
The hood length 1s taken as the source length plus 40 percent of the hood
height above the passline:
Hood Length = 0.4 D + L + C/2
= 0.4 (4.0) + 12 + ^
= 16.6 ft.
Therefore, overall hood dimensions are 6.2 ft by 16.6 ft.
The required exhaust rate, Q, is estimated by the following equation
modified from ACG!H(1976):
Q = 1.4 KPDV
where
K = empirical factor (dimensionless)
P = source perimeter (ft)
D = height of hood above passline (ft)
V = control velocity (ft/min).
The source perimeter is found to be 36 ft from the source dimensions above
(i.e., 2(L + C/2 + B)). Similarly, the height of the hood above the pass-
line is 4.0 ft. Assuming air currents are moderate, a control velocity of
250 ft/min may be used. The empirical factor, K, varies between 0.26 and
1.88 and depends on the passline height, cross-drafts, and effects of the
air curtain. For this case, K = 0.52. Hence, the required exhaust rate is
estimated as
Q = 1.4(0.52) (36)(4)(260)
= 27,256 ftVmin .
The air curtain supply rate is selected so that the velocity of the
jet at the floor is a nominal value of 100 ft/min. (Higher velocities at
the floor result in the jet "bouncing," thereby reducing collection.) A
slot width of 3 in. is typically used so that the distance the jet travels
is 90 in. or 30 slot widths. The air entrained by the jet in Its travel is
estimated by the following equation from Hemeon (1963, p. 203) for two-
sided expansion:
7-64
-------
where
V = velocity at slot
V = velocity at any distance, x, from the slot
N = distance traveled in slot widths.
From the forgoing discussion, V at the floor may be taken as 100 ft/rain
and N = 30, so that the slot velocity = ^30 x 100 = 550 ft/min, A 3-in.
slot has an area of 0.25 ft2 per foot, so that the discharge rate of the
slot per linear foot is 0.25 ft2 x 550 ft/min = 137.5 ft3/min. For the
entire hood perimeter of 36 ft, then, the air entrained by the jet is
estimated as 36 ft x 137 ft3/min ft = 4,950 fta/min. It is seen that the
hood exhaust rate is sufficient to accommodate the air entrained by the air
curtain.
Despite the application of this hood design to many mills, final
installation generally is not the straightforward application of theory
that the above example suggests. Factors such as obstructions beneath the
hood (e.g., mechanical, structural, or electrical elements) and site-specific
rail! characteristics (e.g., speed of mill, type of coolant, and type of
material rolled) require that the system operating conditions be "fine-tuned
" in the field. Air curtain nozzles, for example, are made to be very
adjustable. In this regard, it is recognized that Hemeon's air entrainment
ratio estimates are high, as recently confirmed by Yung et al. (1981).
Nevertheless, these estimates are considered usefully conservative in
providing an upper limit.
7-65
-------
SECTION 8
REFERENCES
Alden, J. L. 1948. Design of Industrial Exhaust Systems. The Indus-
trial Press, New York, New York.
American Conference of Governmental Industrial Hygienists. Committee
on Industrial Ventilation, Box 16153, Lansing, Michigan 48901.
1976. Industrial Ventilation, A Manual of Recommended Practices,
14th Edition.
Anderson, D. M. 1964. Dust Control Design by the Air Induction
Technique. Ind. Med. Surgery. 33:68-72.
Baker, D. E., and Barkdoll, M. P. 1981. Retro-fitting Emission
Controls on the Electric Arc Furnace Facility at Knoxville Iron
Company. Iron andSteel Engineer. 58(8):45-50.
Barton, J. J. 1964. Heating and Ventilating, Principles and Practice.
George Newnes, Ltd., London, Great Britain, p. 423.
Baturin, V. V. 1972, Fundamentals of Industrial Ventilation. Pergamon
Press, Ltd,, Oxford, Great Britain.
Bender, 1984. Bender Corporation Letter to D. W. Coy, Research
Triangle Institute. December 16, 1984.
Bender, M, 1979. Fume Hoods, Open Canopy Type—Their Ability to
Capture Pollutants in Various Environments. Am. Ind. Hyg. Assoc. J.
40:118-127, February.
Bender, M. , and Baines, W. D. 1975. Operation of an Open Canopy Fume
Hood in a Crossflow. Journal of Fluids Engineering of the American
Society of Mechanical Engineers, June, 242-243.
Bender, M., Cesta, T., and Minnick, K. L. 1983, Fluid Dynamic Modelling
of Arc Furnace Charging and Tapping Emissions. Presented at the
EPA:AISI Symposium on Iron and Steel Pollution Abatement Technology,
Chicago, Illinois, October 18-10.
Bender, M., Goodfellow, H. D., Schuldt, A. A., and Vanderzwaag, D.
1982. BOF Secondary Fume Collection at Lake Erie. Iron arid Steel
Engineer. 59:11-14.
3-1
-------
Beskid, C. S. , and Edwards, L. 0. 1982. Visible Emissions Converter
Secondary Hooding, Emission Test Report Asarco Hayden, Arizona,
U.S. Environmental Protection Agency, EMB Report 81-CUS-17, May
1982.
Brand, P. G. A. 1981. Current Trends in Electric Furnace Emission
Control. Iron and Steel Engineer. 58:59-64.
Caplan, K. J. 1982. Ventilation Basics. Plant/Operations Progress.
1(3):194-201.
Cheremisinoff, P. N., and Cheremisinoff, N. P. 1976. Calculating Air
Flow Requirements for Fume Exhaust Hoods, Total and Partial
Enclosures. Plant Engineering. 30(4)-.111-114.
Cheremisinoff, P. N., and Cheremisinoff, N. P. 1976. Calculating Air
Flow Requirements for Fume Exhaust Hoods, Nonenclosure Types.
Plant Engineering. 30(6):143-144.
Chrenko, F. A. (ed.). 1974. Bedford's Basic Principles of Ventilation
and Heating. H. K. Lewis and Company, Ltd., p. 255.
Coleman, R. T., and Vandervort, R. 1980. Demonstration of Fugitive
Emission Controls at a Secondary Lead Smelter. In: Proceedings
of a World Symposium on Metal and Environmental Control at AIME.
Lead-Zinc-Tin, pp. 658-692.
Coy, D. W., Carpenter, B. H., Spivey, J. J., and Jablin, R. 1985.
Engineering Evaluation to Examine Air Control Technology Used in
Foreign Practice of Steelmaking, EPA-600/2-85-071 (PB85216596/AS),
June.
Crawford, M. 1976. Air Pollution Control Theory. McGraw-Hill, Inc.,
New York, New York, pp. 165-187.
Danielson, J. A. (ed.) 1967. Air Pollution Engineering Manual. Los
Angeles County Air Pollution Control District, Los Angeles,
California, Public Health Service Report 999-AP-3Q, pp. 25-86.
Dennis, R., and Bubenick, D. V. 1983. Fugitive Emissions Control for
Solid Materials Handling Operation. J.Air Pollu. Control Assoc.
33(12):1156-1161.
Ellenbecker, M. J., Gempel, R. F., and Burgess, W. A. 1983. Capture
Efficiency of Local Exhaust Ventilation Systems. Am. Ind. Hyg.
Assoc. J. 44(10): 752-755.
Fields, S. F., Krishnakumar, C. K., and Koh, J. B. Modeling of Hood
Control of Blast Furnace Casting Emissions. In Proceedings:
Symposium on Iron and Steel Pollution Abatement Technology for
1981, EPA-600/9-82-021 (PB83164038), December 1982.
3-2
-------
Fletcher B. , and Johnson, A, E, 1982. Velocity Profiles Around Hoods
and Slots and the Effects of an Adjacent Plane. Ann. Occup.Hyg.
25(4):365-372.
Fletcher, B. 1977. Centerline Velocity Characteristics of Rectangular
Unflanged Hoods and Slots Under Suction, Ann. Occup. Hyg.
20:141-146.
Fletcher, B. 1978. Effect of Flanges on the Velocity in Front of
Exhaust Ventilation Hoods. Ann. Occup. Hyg. 21:265-269.
Gilbert, G. B., Hunter, T. E., and Ross, D. 1984. An Experimental
Model Evaluation to Optimize the Ventilation System for a Clamshell
Lime Unloader Hopper. Presented at the 77th Annual Meeting of
the Air Pollution Control Association, June 24-29.
Goodfellow, H. D. 1980. Solving Fume Control and Ventilation Problems
for an Electric Meltshop. Presented at the 73rd Annual Meeting
of the Air Pollution Control Association, June 22-27.
Goodfellow, H. D. 1981. Solving Air Pollution Problems in the Metal-
lurgical Industry. Presented at the 7th International Clean Air
Conference, Adelaide, Australia, August.
Goodfellow, H. D., and Bender, M. 1980. Design Consideration for
Fume Hoods for Process Plants. Am. Ind. Hyg. Assoc. J. 41:473-
484, July.
Goodfellow, H. D., and Smith, J. W. 1982. Industrial Ventilation—A
Review and Update. Am. Ind. Hyg. Assoc. ^J. 43:175-184, March.
Hampl, V. 1984. Evaluation of Industrial Local Exhaust Hood Efficiency
by a Tracer Gas Technique. Am. Ind. Hyg. Assoc. J. 45(7):485-490.
Heinsohn, R. J. 1982. CAD for Industrial Ventilation. Mechanical
Engineering. 64-69, October.
Hemeon, W. C. L. 1955. Plant and Process Ventilation. 2nd ed.
Industrial Press, Inc., New York, New York, 1963.
Henninger, J. L., and Resh, Jr., D. P. 1984. Closing in on Arc
Furnace Emissions at Carpenter Technology. Iron and Steel Engineer.
61:26-30.
Heriot, N. R., and Wilkinson, J. 1979. Laminar Flow Booths for the
Control of Dust. Filtration and Separation, 159-164, March/April.
Hutten-Czapski, L. 1981. Efficient and Economical Dust Control for
Electric Arc Furnace. In Proceedings: Symposium on Iron and
Steel Pollution Abatement for 1980, EPA-600/9-81-017 (PB81-244-808),
March.
8-3
-------
Jutze, G. A., Zoller, J. M., Janszen, T. A., Amick, R. S., Zimmer, C.
E., and Gerstle, R. W. 1977. Technical Guidance for Control of
Industrial Process Fugitive Particulate Emissions, EPA-45Q/3-77-01Q
(PB272288), March.
Kreichelt, T. E., Kern, G. R., and Higgins, F. B. 1976. Natural
Ventilation in Hot Process Buildings in the Steel Industry, Iron
and Steel Engineer. 53:39-46, December, t
McDermott, H. J. 1976. Handbook of Ventilation for Contaminant
Control. Ann Arbor Science, Ann Arbor, Michigan.
Morrison, J. N. 1971. Controlling Dust Emissions at Belt Conveyor
Transfer Points. Trans. AIME. 150:68.
Morton, B. R. 1959. Forced Plumes. J. Fluid Mech. 5:151-163.
Morton, B. R., Taylor, G., and Turner, J. S. 1956. Turbulent Gravita-
tional Convection from Maintained and Instantaneous Sources.
Proc. Roy. Soc. A. 234:1-23, January 24.
Natalizio, A. and Twigge-Molecay, C. 1980. Ventilation of Mill
Buildings—New Directions. Iron and Steel Engineer, July:51-56.
Nicola, A. G. 1979. Best Available Control Technology (BACT) for
Fugitive Emissions Control in the Steel Industry. In: Third
Symposium on Fugitive Emissions Measurement and Control (October
1978, San Francisco, CA), EPA-600/7-79-182, August, pp. 281-312.
PEDCo Environmental, Inc. 1983. Evaluation of an Air Curtain Hooding
System for a Primary Copper Converter, Volume 1, EPA-600/2-84-Q42a
(PB 84160541).
Perryman, R. W. 1984. Busch Company. Letter to E. R. Kashdan,
Research Triangle Institute, November 9.
Roach, S. A. 1981. On the Role of Turbulent Diffusion in Ventilation.
Ann. Occup. Hyg.. 24(1):105-132.
Roos, R. A. 1981. Control of Emissions Generated by Hot and Cold
Rolling Operations in the Aluminum Industry. Presented at the
36th Annual Meeting of the American Society of Lubricant Engineers,
Pittsburgh, Pennsylvania, May 11-14.
Schuldt, A. A., et al., 1981. BOF Secondary Fume Collection at Stelco's
Lake Erie Works, Nanticoke. Presented at the 20th Annual Conference
of Metallurgists, Ontario, Canada, August.
Socha, G. E. 1979. Local Exhaust Ventilation Principles. Am. Ind. Hyg.
Assoc. J. 40:1-10, January.
8-4
-------
Steiner J., and Kertcher, L. F. 1980. Fugitive Particulate Emission
Factors for BOP Operations. In Proceedings: First Symposium on
Iron and Steel Pollution Abatement Technology (Chicago, IL,
10/30-11/1/79), EPA-600/9-80-012 (PB80176258), February 1980,
pp. 253-271.
Steiner, B. 1975. State-of-the-Art for Electric Arc Furnace Secondary
Emission Control. Presented to the Committee on Environmental
Affairs,' International Iron and Steel Institute, Brussels, Belgium,
June 17.
Terry, W. V. 1982. Site Visit—Chapparral Steel Corporation, Midlothian,
Texas, Electric Arc Furnaces in the Steel Industry. Letter to
Dale A. Pahl, U.S. Environmental Protection Agency, Research
Triangle Park, EPA Contract No. 68-02-3059.
Trip Reports. 1979. Research Triangle Institute. Research Triangle
Park, N.C. Prepared for the U.S. Environmental Protection Agency,
Hazardous Air and Industrial Technology Branch. Contract No.
68-02-2651.
Walli, R. A., and Rostik, L. F. 1983. A Market Mill Approach to
Environmental Control—Chapparal's Experience. Presented at the
1983 Spring Conference of the Association of Iron and Steel
Engineers, Dallas, Texas, April 11-13,
Wright, R. D. 19S6. Design and Calculation of Exhaust Systems for
Conveyor Belts. Screens, and Crushers. J. Mine Vent. Soc. South
Africa. 19(1):1-7.
Yung, S-C., Curran, J., and Calvert, S. 1981. Spray Charging and
Trapping Scrubber for Fugitive Particle Emission Control. U.S.
Environmental Protection Agency, EPA Report No. EPA-600/7-81-125
(PB82115304).
8-5
------- |