United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC
EPA 340/1-92-015d
September 1992
Revised March 1993
Stationary Source Compliance Training Series
BVEPA COURSE #345
EMISSION CAPTURE AND
GAS HANDLING SYSTEM
INSPECTION
Reference Volume 1 -
Industrial Ventilation System
Inspection Manuals
-------
EPA 340/1-92-015d
Revised March 1993
Course Module #345
Emission Capture And
Gas Handling System Inspection
Reference Volume 1 -
Industrial Ventilation System Inspection Manuals
Prepared by:
Crowder Environmental Associates, Inc.
2905 Province Place
Piano, TX 75075
and
Entrophy Environmentalist, Inc.
PO Box 12291
Research Triangle Park, NC 27709
Contract No. 68-02-4462
Work Assignment No. 174
EPA Work Assignment Manager: Kirk Foster
EPA Project Officer: Aaron Martin
US. ENVIRONMENTAL PROTECTION AGENCY
Stationary Source Compliance Division
Office of Air Quality Planning and Standards
Washington, DC 20460
September 1992
Revised March 1993
-------
CONTENTS
ITEM 1 - Technical Manual: Hood System Capture Of Process Fugitive
Particulate Emissions, EPA 600/7-86-016, April 1986
ITEM 2 - Performance Evaluation Guide For Large Flow Ventilation
Systems, EPA 340/1-84-012, July 1984
-------
ITEM 1
Technical Manual: Hood System Capture Of
Process Fugitive Particulate Emissions
April 1986
-------
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
P.O. Box 12194
Research Triangle Park, NC 27709
and
Tony Cesta
Howard D. Goodfellow
Hatch Associates, Ltd.
21 St. Clair Avenue East
Toronto, Ontario M4T IL9
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
Prepared for
Air and Energy Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
August 1985
-------
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.
n
-------
CONTENTS
Section Page
Abstract ii
Figures vi
Tables viii
Symbols ix
Metric Equivalents xii
Acknowledgment xiii
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
m
-------
CONTENTS (continued)
Section
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~H
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 Desi.gn 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 ' 7-25
7.2.3 Performance 7-31
7.3 Case III: Basic Oxygen Furnace Secondary
Fume Capture 7.35
7.3.1 Source Description and Background '. 7-35
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 Furnace 7.5^
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
-------
FIGURES
Number
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 5-3
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
-------
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 Hme drop flow patterns to be
modeled 7-55
7-19 Geometry of final configuration: baghouse flow
1s drawn from back of hopper under single baffle,
which 1s raised off grizzly 7-59
7-20 Schematic cross section of an air-curtain hood.
Air jets prevent fumes from exflltrating into
work areas surrounding mill 7-62
7-21 Example perimeter hood for control of aluminum
rolling mill emissions 7-63
vn
-------
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
vm
-------
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).
A, = Hood face area, m2 (4.1.2).
A = Control surface area (4.1.2); cross-sectional area of the falling
s 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).
b,. = 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, cal/gm-°C (4.1.1).
CM = 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, mVs3 (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, m3/s (4-6).
QH = 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/unit 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).
J
TS = Air temperature in hood suction field, K (4.1.2).
AT = Temperature difference between hot body and ambient air (4,5).
t = Purge time of hood, s (5.1.2).
td = Duration of plume surges, s (5.1.2).
Umax = P1ume centerline velocity, m/s (5.1.3).
V = Velocity, m/s (4-7).
-------
SYMBOLS (continued)
V. = Jet nozzle velocity, m/s (4.1.3).
V = Hood suction velocity, m/s (4.1).
Vu = Plume velocity, m/s (4.1).
Vcross = Cross"draft velocity, m/s (5.1.3).
W = Materials flow rate, kg/s (6.1).
X = Characteristic source dimension, m (4.1).
Y = Characteristic source-hood dimension, m (4.1).
Z = Effective height between plume virtual origin and hood face, m
(4, 5, 7.1).
a = Trajectory angle, dimensionless (4, 1, 2).
P = Deflection angle, dimensionless (4.1.3).
r£ = Pollutant rate arriving at hood face, g/s (5.3).
r., = Pollutant rate captured by hood, g/s (5.3).
e - Emissivity dimensionless, (5.1.1).
nu00(4 = Hood capture efficiency, dimensionless (5.3).
6 = Deflection angle, dimensionless (4.1.3).
p = Hot gas density, kg/m3 (4, 5).
pQ = Ambient gas density, kg/m3 (5.1).
ps = Bulk solids density, kg/m3 (6.1).
o = Stefan-Boltzmann,constant :
1.354 x lo"1^ (kcal/s-m2-K4), or
0.1714 x lo"8 (Btu/hr-ft2-°R4).
x1
-------
METRIC EQUIVALENTS
Nonmetric Times Yields metric
°F
°F/min
ton
Ib
Btu/lb-°F
Btu/mi n
cfm
ft
ft2
ft3
ft/mi n
in.
5/9 (°F-32)
0.556
907
0.454
1.0
252
1.7
0.30
0.093
28.32
0.00508
2.54
°C
°C/min
kg
kg
cal/gm-°C
cal/min
m3/hr
m
m2
L
m/s
cm
-------
ACKNOWLEDGMENT
The authors are Indebted to Richard JabUn (Richard Jablin and Associates)
and Manfred Bender (Bender Corp.) for their review of the manuscript. We
also thank Richard Ferryman, 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, Mr. Dale Harmon.
-------
SECTION 1
INTRODUCTION AND SUMMARY
Process fugitive participate 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-
ti cul ate 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 been 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 is 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 1n Section 5, and design of enclosures for buoyant and
inertial 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
*
Inertial 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 1s 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
(inertial)
Buoyant
Nonbuoyant
(inertial)
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 particulate 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
j
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 Plant and Process Ventilation (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 Air Pollution 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 hygienists as a practical text
accompanying the Industrial Ventilation 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. 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.
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 Meltshop. 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., 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.
Hanpl, 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 Materials Handling 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 1s 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 1s 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 1s a knowledge of the physical charac-
teristics of the plume necessary during both average and peak conditions,
but the process source should be examined 1n 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 1s 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 1n 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
i
Standards of
Performance
I
Source
Characterization
1
Hood Type
Selection
Design Methods
(a) Exhaust rate
(b) Hood arrangement^
Optimization
Fabrication
i
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 particulate 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. Inertial 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 inertial 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
en
General Design Principles for Hood Systems
• Design Objectives
Principle: All 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
Nonbuoyant 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. 67.
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 1n 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 inertial 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 thts 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 INDUSTRIES6
Local
Industry
Canopy
Building
Fixed Moveable Side-draft (high) Enclosure evacuation
Iron and steel
1.
2.
3.
4.
i
00
7.
8.
9.
10.
Sinter plant .
Sinter machine discharge
Sinter cooler
Blast furnace
Tap (iron)
Tap (slag)
Slag crushing
Open hearth furnace
Charge
Tap
Basic oxygen furnace
Charge
Tap
Electric arc furnace
Charge
Tap
Cold scarfing
Hot scarfing
Hot metal transfer
Pig iron (reladling)
Hot metal desulfurization
(skimming)
X
X
+
+
+
+
+
+
b P
Teeming '
Continuous casting
X
X
x = Typical control technique.
+ = In use (but not typical) control technique.
hood pr-actices are from EPA-45O/3-77-O1O unless
-
Engineering judgment.
cEPA-450/3-82-005a.
dEPA-45O/3-79-O33.
-------
TABLE 3-1 (continued)
Local
Industry
Fixed Moveable Side-draft
Canopy Building
(high) Enclosure evacuation
10
10
11.
12.
13.
14.
Coke pushing
Cold rolling6
Hot strip m1llb
Materials handling
x
x
x
x
15. Railroad car dumper
Iron foundries
1. Cupolas
Charge
Tap
2. Crucible furnace
Pouring
3. Electric arc furnace9
Charge
Tap
4. Electric Induction furnace
5. Reverberatory furnace
6. Ductile Iron 1nnocu1at1on
7. Pot furnace
8. Pouring Into molds
9. Casting shakeout
x
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
to
I
M
O
10. Cooling, cleaning castings
11. Finishing castings
12. Mold sand, binder receiving
13. Sand preparation
14. Mold mak1ngb
Steel foundries
1. Electric Induction furnace
Charge
Tap
2. Electric arc furnace
Charge
Tap
3. Open hearth furnace
Charge
Tap
4. Pouring 1n molds
5. Cooling and cleaning castings
6. Casting shakeout
J
x
x
•f
+
X
X
X
X
x = Typical control technique.
+ = In use (but not typical) control technique.
Engineering judgment.
1EPA-450/3-8l-005b and EPA-450/3-80-020a.
JACGIH (1976).
-------
TABLE 3-1 (continued)
L°Ca1
Industry
Fixed Moveable Side-draft
Canopy Building
(high) Enclosure evacuation
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.
-------
TABLE 3-1 (continued)
Local
Industry
Fixed Moveable Side-draft
Canopy Building
(high) Enclosure evacuation
Primary zinc smelting
1. Sinter machine windbox
2. Sinter machine discharge,
screens
3. Retort furnace
4. Zinc casting
5. Coke-sinter mixer
Primary aluminum smelting'
1. Anode baking
CO
2. Electrolytic reduction cell
3. Refining and casting
Secondary aluminum smelting
X
x
X
1.
2.
3.
4.
5.
6.
Sweating furnace
Reverberatory furnace
Crucible furnace
Induction furnace
Fluxing
Hot dross handling
x +
x +
x + +
x +
x = Typical control technique.
+ = In use (but not typical) control technique.
m
Engineering judgment.
EPA-450/2-78-049b.
-------
TABLE 3-1 (continued)
Local
Industry
Fixed Moveable Side-draft
Building
(high) Enclosure evacuation
Secondary zinc smelting
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 smelting0
1. Cupola
Charge
Tap
x = Typical control technique.
+ = In use (but not typical) control technique.
n
Coleman and Vandervort 1980.
'EPA-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
GO
£ Nonmetallic minerals^
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)
Canopy Building
Industry Fixed Moveable Side-draft (high) Enclosure evacuation
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 + +
OJ
M 7. Packaging + +
Asphaltlc 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
aJutze, 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-450/3-77-010 (PB272288), March.
cRevised Standards for Basic Oxygen Process Furnaces—Background Information
for Proposed Standards. 1982. EPA-450/3-82-005a, December.
dReview of Standards of Performance for Electric Arc Furnaces in Steel Industry.
1979. U.S. Environmental Protection Agency. EPA Report No. EPA-450/3-79-033
(PB80-154602) October.
9Electric Arc Furnaces in Ferrous Foundries—Background Information for Proposed
Standards. 1980. U.S. Environmental Protection Agency. EPA Report No. EPA-
450/3-80-020a (PB80-202997) May.
Environmental Assessment of Melting, Innoculation, 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-450/3-
81-005b, September.
•'American Conference of Governmental Industrial Hygienists. 1976. Industrial
Ventilation, A Manual of Recommended Practices, 17th Edition. Edwards Brothers,
Ann Arbor, Michigan.
k
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.
nColeman, 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
-------
P PAir Pollution Control Techniques for Non-Metallic Minerals Industry. 1982.
U.S. Environmental Protection Agency. EPA Report No. EPA-450/3-82-014,
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 Corporation9
.
Crucible. Inc.8
Sidbec Melt Shopb
Knoxvllle Iron Company0
CO
J-. Carpenter Steel (Reading, PA)d
VO
*
Stelco-McMaster Melt Shop6
Iscott (Trinidad)6
A
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
Dampered canopy;
partial furnace
enclosure
Moveable ladle
hood
Dampered canopy;
Internal baffles
(275 fpm)
Enclosure; air
curtain across .
roof slot
Canopy
Close hood
Canopy with
scavenger ducts
Ventilation
rate
17,600 mVmln
17,600 irVmln
17,100 mVmln
17,100 mVmln
5,900 mVmln
4,200 ma/m1n
5,100 mVmln
2,100 m'/mln
15,600 irVmln
Size Capture Control
(hood face) efficiency device
15.2 m x 13 M Reverse-air
15.2 m x 14 m 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 mx 15. 5m 95-100X Baghouse
x 10.7 m
139 m* Opacity, plume Baghouse
photography
ll.lm2 Fluid modeling Baghouse
3% maximum opacity Pulse- jet
baghouse
'Discussed In detail In Section 7.
aBrand (1981).
bHutten-Czapsk1 In EPA-600/9-81-017.
GBarkdoll and Baker (1981).
Hennlnger et al. (1984). Capture efficiency estimate
by telecon from L. Gelser to M. Bender (1984).
'Details available from Hatch Associates.
fTerry (1982).
-------
TABLE 3-2 (continued)
Industry
Republic Steel (Chicago Works)9
Republic Steel (Cleveland
Works)9
*
Stelco-Led (Nant1coke)h
Bethlehem Steel (New York)
Chiba Works (Kawasaki .
Steel Corporation) '*
oo
ro
o
Mizushima Works (Kawasaki
Steel Corporation)-1
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)
Charging
(takeoff)
Re ladling
Desulfurlzation
Des lagging
q-BOPF (180 tonne)1
Charging (escaping
furnace enclosure)
Tapping
enclosure (chain
curtains)
Reladling
ring
Ventilation Size
Design rate (hood face)
Partial furnace
enclosure;
charging hood 9,400 dmVmln
Partial furnace
enclosure; 10,100 dmVmln
charging hood 9,100 dm3 An In
Local hood 10,000 nrVmln (200° C) 13.9 m2
Movable enclosure 6,000 nrVmin (150° C)
(re ladling)
850 mVmin 9.3 m2
Local (dampered) 16,000 mVmin
Local (baffles)
Local
Booth
Part of furnace
enclosure (chain
curtains)
Part of furnace
Hoveable close-fit .
Capture
efficiency
<5% opacity
Ineffective
<5% 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
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
"Discussed in detail in Section 7.
9Ste1ner and Kertcher 1n EPA-600/9-80-012 (1980).
hBender 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
r
Tapping
Reladllng
Kashima Steel Works. . OG furnace (250 tonne)
(Sumitomo Metal s)Jl Charging (escaping
enclosure)
Oes lagging
Yawata Plant (Nippon Steel)-' BOPF (340 tonne)
Charging
Desulfurizatlon
CJ
rv> Oita Plant (Nippon Steel)^ BOPF (340 tonne)
»-J Charging
Reladllng
Des lagging
Swedish Steel'' BOPF (145 tonne)
Charging
Desulphurizatlon
Hot metal transfer
Design
Part of furnace
enclosure
Part of furnace
enclosure
Side-draft hood;
supplemental canopy
Part of furnace
enclosure (chain
curtains)
Local
Part of furnace
enclosure (chain
curtains)
Local - close
Part of furnace
enclosure
Booth (metal poured
through slot In hood)
Booth
Enclosure (doghouse)
Local (baffles)
Local side-draft
Ventilation Size
rate (hood face)
7,100 irVmln
850 mVmin per
torpedo car
60 mVmln
25 mVmin
9,200 mVmin (at 70° C)
3,300 mVmin (at 70° C) 2 m x 2 m
830 mVmin (at 70° C)
Capture
efficiency
90-95*
50-75*
95*
50-75*
50-75*
100*
100*
95-100*
75-95*
75-80*
80-100*
95-100*
Control
device
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
JRTI trip reports (1979).
kRoof monitors are ducted to baghouse for supplemental process fugitive collection.
-------
TABLE 3-2 (continued)
ro
ro
Industry
Iron and steel (continued)
Ohgishlma Plant-*
(Nippon Kokan)
Italslder (Italy)J
British Steel Corporation^
(Lackenby Works)
Titanium (Ilmenlte) Smelting
QIT, Sorel6
Lime Manufacturing
Stelco-Lede (Nantlcoke)
Process fugitive
source
BOF (250 tonne)
Charging
Scrap
Hot metal
Tapping
Deslagglng
Reladllng
BOF (350 tonne)
Charging
Hot metal transfer
Hot metal desulfurlzatton
BOF (260 tonne)
Charging
Tapping
Scavenger (supple-
mental)
Ladling
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)
Hoveable hood
Enclosure
Ventilation
rate
10,000 mVmln
(at 480° C)
3,000 nvVmln
(at 130° C)
1,800 mVmln
(at 130° C)
2,700 mVmln
4,500 m-Vmin
850 mVmln
2,100 mVmln
Size Capture
(hood face) efficiency
95-100%
50-75%
95%
60%
85-95%
2 m x 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 m
Plume flow rates
measured
Control
device
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Scrubber
Scrubber
Scrubber
Baghouse
T
RTI trip reports (1979).
eOetails available from Hatch Associates.
-------
TABLE 3-2 (continued)
rv>
CO
Industry
Secondary Lead
Test smelter
Primary Copper
Asarco-Hayden1"
*
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 mViiiln
100 nrVmln
120 na/m1n
0-lOX opacity"
2,100-3,600 n3/n1n 75-95%
Control
device
Baghouse
Baghouse
Baghouse
Preclpltators
Scrubbing towers
*D1scussed In detail In Section 7.
'coleman and Vanderyort (1980).
mEPA-450/3-83-018a.
VsMd and Edwards (1982).
°PEDCo (1983).
-------
TABLE 3-2 REFERENCES
aBrand, P. G. A. 1981. Current Trends in Electric Furnace Emission Control.
Iron and Steel Engineer. 58:59-64.
bHutten-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 and Steel
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 Steel Engineer. 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.
^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, pp. 253-271.
Bender, 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-450/3-83-018a, November.
nBeskid, 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 Asarco, Inc. Draft Report. U.S. Environmental
Protection Agency. EPA Contract Nos. 68-03-2924, Work Directive 9 and 68-02-
3546, Task Assignment No. 12.
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 Buoyant Sources
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 (m of air)
g = gravitational constant (9.98 m/s2).
4-2
-------
Local Receiving Hood
Off Take
Opening for Addition
of Product, Area A.,
Clearance
Area, A2
Maximum Thermal
Head (L)
Vessel Containing
Hot Product
Source: Goodfellow 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
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:
V =
(4-3)
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:
.T - g-^ (4-4)
where
q = rate of heat transfer from process (kcal/s)
Q = hood suction rate (m3/s)
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:
(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 oper
s o o
area for the hood openings, the updraft velocity is expressed as follows:
4-4
-------
(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 At plus A2 in Figure 4-1):
3
Qs • 2.9
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
(a)
Adapted from ACGIH, 1976.
Exhaust Hood
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
where
M = Momentum of hood suction field
M = Momentum of plume updraft
M = Resultant momentum (vector addition)
r
« = 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
where
A = control surface area (m2)
A = plume cross-sectional area (m2).
Equation (4-8) applied to the hood is written as
= (Vs)2(As)(ps)
which upon rearranging becomes
/ M
Vs \(PS)(AS) '
Substituting for (M /A ) from Equation (4-10-), the suction velocity may
be written as
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:
AS = 10 x2 + Af
4-8
-------
TABLE 4-1. CONTROL SURFACES FOR VARIOUS EXTERIOR HOOD TYPES3
b 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
S[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
Af = hood area, m2
x = distance from hood face, m (0 < 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, Qg, is given by:
s u
Applying Equation (4-8) to the plume momentum flow rate, it follows that
Mu = (vu»)(Au)(pu) ;
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
QS= Vr (VM7 UOX2 + Af)
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, A., 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 al., (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
(a)
Adapted from ACGIH, 1976.
Jet Side
Mi = jet momentum
Mu - updraft momentum
Mr = resultant momentum
0 & 0 = deflection angles
(b)
^•••^M
V
\
Exhaust Hood
Exhaust Side
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:
(P0)(V02)(AQ) = (px)(Vx2)(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:
V(A,J)(T.)
(W ^ (4'14)
where
V = average plume updraft velocity (m/s)
AU = plume cross-section at intersection with jet (m2)
A. = jet nozzle area (m2)
T. = jet air temperature (K)
T = average plume air temperature (K)
C,. = (cos 6 x tan p) + sin 6.
H
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 V(Q-)(V.)(X) (mVs/unit length of slot) (4-15)
5 J J
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 used 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
-------
Actual Hood
Performance
Linear Relationship
for Hood
8
Q100 for Hood
(Theoretical)
9 10
01234567
Hood Suction, X1000 m3/min
Adapted from Bander at at., 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 Q100- 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 Q100 (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)
V2 V2
L
•
'
PO
m N / p
V2 L4 L4 p V2 L4 L4 p
m m p Km _ p m p Kp
Lm (Pom * Pm) Lp (P0p • P
QP _ LP Pom (Pop - Pp) _ ;, Tm Tp
with q = (Q)(p)(C ) (T - T ) and -£ = =*, then
p pm p
v_ 5/
V " ^ \qm /
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 particulate
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
Particulate loading
Face of Hood
Gas composition
Gas volume
Gas temperature
Paniculate loading
Hood Off-take
Gas composition
Gas volume
Gas temperature
Paniculate loading
Part icu late Characteristics
Chemical composition
Panicle size
Particulate Emission Rate (at source)
Instantaneous
Hourly
Daily
Heat Generation Rate
Total
(Normal m/h)*
°C a
mg/Normal m°
(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
.m2
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
Paniculate loading
Hood Off-take
Gas composition
Gas volume
Gas temperature
Paniculate loading
Particulate Characteristics
Chemical composition
Particle size
Particulate Emission Rate (at source)
Instantaneous
Hourly
Daily
*
Heat Generation Rate
Total
Plume Rise Data
Velocity @ 1/2 D
@ hood centerline
Temperature
Plume cross-section
Area @ hood centerline
Nozzle Jet Data (Push-Pull only)
Nozzle air flow rate
Nozzle width
Nozzle length
(Normal m3/h)*
°C
mg/Normal m3
(Normal m3/h)
«c
mg/Normal m3
kgfc
kg/h
kg/day
kcal/s
m3fe
•c
m2
Normal m3/h
m
m
Hood
Hood Geometry Data
A m
B m
C m
D m
H m
L m
Sketch
For Push-Pull Only
F m
F m
A 0
Flanges
Hood Performance Equation*
Original Basis
Analytical
Modeling
Current Performance
Analytical
Modeling
Field Measurements
Hood Capture
Efficiency (%)
Comments
•Normal implies 20°
C, 1 atm.
tCalculation sheets attached for specific
cases.
Figure 4-6. Hood design data questionnaire-B.
4-21
-------
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
co
Source
Continuous plume
Intermittent plume
Cross-drafts
Obstructions
Hood parameters
Exhaust rate
Hood diameter
Hood storage volume
Exhaust rate
Exhaust rate
Governing equation
QH = 0.166 Z5/3 F1/3
Qs = 1.21 QH
50 percent of Z
Hood volume = t(Q^ - Qi)
P
Vcross
s H U
max
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
QH = 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) = aU
4. equal spread of buoyancy (concentration) and
velocity profiles
Dimensions
Equation
Characterizing source quantity
V
assumed
uniform
Buoyancy flux = const.
F = QA[mVs3]
Volume flow rate
Q
Center line velocity
max
m
s
5_ /18oF\
6a 5;i J
1/3
-l/3
Entrainment const.
a
0.093
Length scale
b
m
(6/5)oZ
Center line buoyancy
max
m
_
3n \lBfa
r U
,-5/3
Froude No.
= const.
Entrainment angle, 6 (approx.)
Deg.
18
Z = effective height from virtual plume origin to hood face.
F = buoyancy flux.
A = buoyancy = (g)
P 0- P
, where p = ambient density, p = plume density,
and g = gravity constant.
Adapted from Bender (1979). The assumptions are discussed 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/m3).
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
qc = (hc)(As)(AT) (5-3)
where
q = heat transfer rate due to- convection (kcal/sec)
hc = 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)(a)(T)« (5-4)
where
qr = heat transfer rate due to radiation (kcal/s)
e = emissivity (dimensionless)
AS = surface area of hot body (m2)
a = Stefan-Boltzmann constant (kcal/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
Qs = hood suction rate required for no spillage
QH = 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 if 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, Qlf 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)
Virtual Origin of Plume
\
\
\
(b)
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
"K5 times the source diameter, D.
5-8
-------
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 steelmaking 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 in 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 = td (Qp - Qg) (5-6)
where
t. = duration of plume surge (s)
Q = peak plume flow rate (m3/s)
Q = hood exhaust flow rate (m3/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, 0^.
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
-------
t 400
8
CO
* 300
o
15
ec,
200
100
I o
Q.
"Peak" Plume Flow Rate
"Normal" Plume Flow Rate
I ij I I i I I i ill
I i i i
Time, seconds
(a)
Minimum "normal"
Plume Exhaust
1000 2000 3000 4000
Hood Storage Volume, m3
(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 impractical ly 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 is simply the nominal residence time of the fume in the hood
given by the following equation:
t _ Hood volume ,. >
"
Hood exhaust rate ' >
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 from 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
Qs = QH (1+ 4.7 -) (5-8).
where
Qs = hood suction flow rate (m3/s)
QH = plume flow rate at the hood face (m3/s)
Vcross = cross~draft f°w velocity (m/s)
Umax = Plume centerline velocity, m/s at the hood face (Table 5-2).
The eccentricity (distance between the hood axis and plume axis) which
results from the cross-draft is described by the equation
V
e = 13.53 (bH) -^2Si (5.g)
U
max
where
e = eccentricity (m)
5-12
-------
t>H = plume length scale at hood face (m) (Table 5-2).
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 1s 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 1s 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
OP - Qm (s)5/3 (qp/qm)1/3 (
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 = [ 1 (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
P0 " 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-plant density
gradients can be realistically modeled.
For Intermittent plume sources as described 1n 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 1s 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
-------
IA
I
0 15 30 45 60 75 90 105 120 135 150 165 180195 210
Time, seconds
Adapted from 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. Zt, 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
Q? = o.026 W I 'it ^-^— Z2 (5-12)
where a = ^- .
Lz
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. Procedural ly, 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 an.d hood capture efficiency. Based on fluid model
5-17
-------
Time
I
I—•
en
a
Tn-1
Reference Plane
Z1
^
0
Tn
>s
\ —j- Plume
V"^- Ladle
Photograph
a
Tn
e
Tn + 1
7*y^\~-
tf
a
Tn + 1
_ — _ _
0
Fn + 2
"\
W
D
>
Tn^2
1 ^
Z2
a0 a
S = Distance traveled = Z2 - Z1 T = Time span = Tn + Tn + 1 + Tn + 1
D = Diameter
0 a
T = Frame exposure time T = Frame advance time
V = Velocity = -
Adapted from Goodfellow and Bender, 1980.
Reprinted with permission by American Industrial Hygiene Association Journal.
Figure 5-4. Photographic scaling 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:
Qs v rH
- ' '
where
Q = hood suction rate
Qu = plume flow rate at the hood face
r^ = pollutant rate captured by the hood
r£ = pollutant rate arriving at the hood face
riuood = capture efficiency of the hood
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:
(1 - QS/QH)X
OP = 1 - (1 - OPmax) " (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
-------
(a)
0.1 -
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)
"Worst" Hood
"Ideal" Hood
(b)
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)
Adapted from 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 particulate, 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 (Q,/QU) 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 cross drafts
Plume diameter
Opacity of Discharge from Building
(Normal m3/h)*
°c
mg/Normal m3
(Normal m3/h)
°c
mg/Normal rrr
>
kg/s
kg/h
kg/day
kcal/s
m3/s m/s
m3/s m/s
sec
Occurrence/min
m/s
m
%
Hood Sketch
Hood Geometry Data
A m
B m
C m
D ..,- m
F m
F m
H m
L m
W m
K m
Hood Performance Equation*
Planned Site
Analytical
Modeling
Existing System
Analytical
Modeling
Field Measurements
Hood Capture
Efficiency (%)
Comments
•Normal implies 20°
C, 1 atm.
Calculation sheets attached for specific
cases.
Figure 5-6. Hood design data questionnaire for canopy hood.
5-22
-------
SECTION 6
DESIGN METHODS FOR ENCLOSURES
The following section discusses enclosures for inertial 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 particulate 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 inertial 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
inertial sources. The following section is divided into a discussion of
dust-producing mechanisms applicable to all inertial sources, design of
enclosures for gravity transfer operations, and considerations in the use
of nonexhausted enclosures.
6.1.1 Dust Generation in Inertial Sources
Dust generation mechanisms for inertial 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
-------
-— : ,.• •
« •
Falling
Material
Leakage (
Opening \
• t
• •
• ''i*
• „
-* ^ • • *\
^^
x ^^x ;*• •
vC^.
1 \S^ • "
/ r 9 °
1 / v •
/ \ ° ° '
/ " c «
1 «0 0 '
V
^x* — Externally
Internally
/Induced
Air
— . •
/
Ai
*
Enclosure
Container
Material Splash
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):
W H2 A 2
Qi = 0.63l p / (6-1)
where
Qi = 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:
Qa = -J- (6-2)
Hs
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 rionexhausted 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
-------
HOOD
BUMPER
SECONDARY
HOT METAL CHARGMQ
FURNACE CHARGING.
(Rvtractabto)
SLAG POT
CTl
I
10
WATER COOLED HOOD
HOOD TRANSFER CAR
ADJUSTABLE SMRT
TAPPING EMISSIONS DUCT
SEAL RING
FURNACE ENCLOSURE
FLOOR
SHOP AMVORAFT
DURMQ SLAGGMG «
TAPPMQ
Source: Nicola, 1979.
Figure 6-2. Schematic arrangement for BOF furnace enclosure.
-------
ROOF SLOT DOORS
(2-SECTIONS)
rAIR JET
DIVERTER
FRONT ACCESS
DOORS
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
p = air density at Ts
C = specific heat of air at Ts
P
T = specified air temperature after dilution
T .= ambient dilution air temperature.
amb
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 1s 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 i-n 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 1s 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 Mechanical Design
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 would
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 PARTICULATE 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 1n 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
--j
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
A1r 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 electric furnace shops in the particular jurisdiction would permit
7-3
-------
ROOF EXHAUSTER
LOOK MO SOUTH
CHMOMO *BLC
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 Original 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
, - 75 ton (2,000 £) (0.11
q = 180,000 Btu/min .
Plume flow rate — Plume flow rate is calculated using an equation from
Hemeon (1963):
Q = 7.4 (Z)1'5^)173
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 + 2D, 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 (180,000)1/3
Q = 360,000 acfm.
For ladle tapping
Y = 76 (from Figure 7-2)
D = 10 (from Figure 7-2)
Z = 76 + (2 x 10) = 96 ft
Q = 7.4 (96)1'5 (180,000)1/3
Q = 391,000 afcm.
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 = Z °'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
-------
MOOULATIM DAMPEN
ISOLATION DAMKM
1< VHTUAL FLUME OHNM
LOOIMH MOUTH
VKW-'i-^
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 Collection for System Modifications
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
7-8
-------
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 plume—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-in
-------
N
f
CROC MIME AND
SOLD MIICIWS
Figure 7-3. Map of the plume boundaries relative to the original hood system.
-------
g
o
UJ
£
(E
UJ
3
O
<
O
1,600
1,400
1,200
1,000
800
600
400
200
20 SEC.
OPEN FCE., LIGHT
FUME EMISSION
PEAK THOUGHT
POSSIBLE, ESPECIALLY
WHEN CARBON ADDED
TO SCRAP
CASE 'B'
PEAK PHOTOGRA-
PHICALLY OBSERVED"?;
DURING TESTS
HEAVY FUME EMISSION
SHADED AREA REPRESENTS
MINIMUM STORAGE VOLUME
REQUIRED FOR POOL TYPE
HOOD WITH 520,000 ACFM
SUCTION
000 ACFM
WjTH POOL
\
2I2.000\ACFM
EXISTING CANOPY EXHAUST RAT
(MEASURED)
EXHAUST RATE REQUI
TYPE HOOD
Figure 7-4. Observed and speculated plume flow rate during charging.
7-12
-------
of 920,000 acfm was actually measured, but observations of much more violent
charges and experience gained from plume tests 1n 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/min
(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't 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 = -T7r4
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
P0 = air density = 0.075 lb/ft3
F = 158.000 x 32.2 x 3600 sVmin2 _ - q, n9
F 0.24 x (460 + 70) x 0.075 l'B2 x 10
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 acfm .
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
-------
0.
O
w
=J
X
UJ
a.
o
v>
O
o
100
90
80
70
60
50
40
30
20
10
0
100
90
80
70
60
50
40
20
10
0
UJ
o
ae
x
u
3.5
MIN.
UJ
ae
10
L3.5
MIN.
MAXIMUM OPACITIES
CASE A
TIME 20% OPACITY
EXCEEDED
l«t CHARGE
OF NEXT
HEAT
0
h
10 20 30
40 50 60 70 80
— ONE-2 HOUR HEAT —
90 100 110 120 130 140
«H TIME (MIN.)
NORMAL OPACITIES
CASE 'B1
u
o
(C
flC
a
o
<
u
2.6
MIN
TIME 20% OPACITY
EXCEEDED
/-!•* CHARGE
/ OF NEXT
HEAT
10 20 30
40 50 60 70 60
— ONE-2 HOUR HEAT —
90 100 110
120 130 140
•H TIME (MIN.)
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 1n 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 1n 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 (opmax)> fume volume flow
rate during period when opacity is exceeded (Q^), and fume hood suction
(Qi), are expressed in the equation below (derived from the Lambert-Beer
•law, see Section 5.3):
7-17
-------
OP • 1 - » - OPmax:
" Ql/QH} .
OP is the opacity of spilled fume when Qt (hood suction) is equal to
zero (Goodfellow and Bender, 1980). Letting X be the opacity limit and Y
the peak opacity,
= l-OP
mav
IHuX
(l-Q1/QH)xLn(l-OP)y = (l-Qi/QH)Y Ln(l-OP)x
Ln(l-OP)Y
From Figure 7-4,
(Qi)v = tne existing measured suction rate of 212,000 acfm
(QH)Y = the charging plume flow rate of
920,000 acfm-observed-case 'A1 and
1,400,000 acfm-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 'B1. Finally, the opacity limit is set at
40 percent (OPX = 0.40), and substitution into the derived equation for
case 'A' and 'B1 gives the following:
FOR CASE 'A'. OPX = 0.40, Qi = 212,000
OPy = 0.80, QH= 920,000
(Qi/QH)40% = °-755» Qi = 695,000 acfm
7-18
-------
FOR CASE 'B'. OPX = 0.40, Qa = 212,000
OPy = 0.97 QH = 1,400,000
(Qi/QH)40% = 0.876, Q! = 1,226,000 acfm
Therefore, 1f 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
I
ro
Design parameter Method of reducing secondary emissions to less than 40% capacity
Source
Charging
CASE A
Normal plume
Flow rate Q
Opacity
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%
Q!/QH = 0.876
1,400,000 acfm Q (suction)
= 1,226,000 acfm
97%
470,000 acfm Assume opacity reduced
hood from crane.
40%
Improve hood
capture technology
-Pool type hood with
45,000 ft3
491,000 acfm
(includes 25% safety
margin)
design for maximum
or upset case 1s not
not practical
to less than 20% by existing
Close
roof exhaust
during emission
NA
NA
NA
NA = Not acceptable to working conditions.
Compare to: Greenfield hood suction prediction 400,000 acfm
As installed design hood suction 216,00 acfm
As Installed measured hood suction 212,000 acfm
-------
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 is 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 Peirce-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 EMSSIONS
PRIMARY HOOD
HOOD
GATE
i
ro
oo
TUYERES
HOOD OATE
RETRACTED
BLOWING
CHARGWO
(MATTE 6 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
ro
cn
JET SIDE
MR
OJRTAM
JET
TO EXHAUST FAN
BAFFLE WU.L-*
EXHAUST SIDE
.
CONVERTER /
\ (FUME SOURCE).
x v /
LADLE
TO EXHAUST FAN
-BAFFLE WALL
Source: PEDCo, 1983.
Figure 7-7. Converter air curtein 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 if captured emissions are cleaned and/or dispersed to acceptable
levels.
7.2.2 Design Approach
The original air curtain design calculation was not available for this
assessment; therefore, the following design procedure is 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 curtarn 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
BAFFLE WALLS
BAFFLE PLATE
VIEW 'X-X'
(
fa
CONVERTER
A }
\./
('^
w
* i
2 INCH WIDE
JET SLOT-
OUTLINE
OF EXHAU
DUCT
* /' <
* \ /I3FT x
l""l
r)
JR.
\ 1 1 A
\ r:
i &/ i
&&*-
VIRTUAL PLUME
ORIGNS
Figure 7-8. Analysis of air curtain system.
7-27
-------
From fundamental heat transfer calculations, a heat release of approx-
imately 150,000 Btu/min is determined for both charging and skimming. The
resulting buoyancy flux is 1.8 x 109 ftVmin3.
Plume flow 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 •=• 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 2 = y
and V = _ 41Q ft/min
For 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 5 = 176
and V = - = 510 ft/min .
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 = OhXpiXVj) = (Q2)(o2)(V2) .
where
Q = volume flow rate
p = air density
V = velocity
I and 2 = distances from the nozzle.
For the present case (Figure 7-8) assume angle 0 = 0 (Figure 4-3), therefore,
M = M.(Sin 6)
^* J
(QU)(PU)(VU) = (QjXp.jXV.jXSin 6) .
Setting 6 = 15 degrees and assuming the worst design case of skimming where
Q = 90,000 acfm, Vy = 510 ft/min, and pu = 0.062 lb/ft2 (185° F), then
(A.)(V.)(P,-)(V,) = 11 x 106 Ib ft/min2 (since, Q. = (A.)(V.)) .
J j J J j j j
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)*5 = 8,228 ft/min
Q. = 2.166 x 8,228 =17,823 acfm .
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 QH and is estimated by
the equation (Equation 4-15):
Qu = 0.88 ((0,)(V.) (R/Slot length))15 (Slot length)
H J J
where
Qu = plume arriving at hood face
Q. = jet flow rate at origin
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)*5 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 QH is therefore approximate. The core of the jet contains
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
vj Rotate- 1n/ rotate-out
™ Blow/Idle
Overall
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/min
2.19
7.38
2.35
1.94
6.15
0.14
—
A = Tracer gas Injection 1n upper control volume.
B = Tracer gas Injection 1n 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
X
Ul
\_
CONVERTER
TOP VIEW
—SAMPLING LOCATION
JET SIDE
EXHAUST SIDE
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 participate
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-in/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 (reladling), 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, BOF 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 BOF 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
CHARGING
AISLE
HOTAIR.CO.C02
AND PARTICULATE
AISLE
CRANE GIRDER
SCRAP
Source: Shuldt et al., 1981
POSITION OF
LOCAL HOOD
FOR CHARGING
FUME CAPTURE
CHARGING LADLE
CHARGING
FLOOR
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 (Fe203, 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 CO-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 means 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.
Slagging—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 BOF 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 re!adling—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
Q Unknown
c
1
n
E
0
®
«
i
_3
£
co
C3
0
a
w
CO
O
tirfi
&\J
18
16
14
12
10
8
6
4
2
o
—
^
Kaiser
A
Oita
Kimitsu 2 • •
Led O O Fukuyama
O Italsider
Youngstown
Hilton D
A
^
-
I I I I
100 200 300 400
,. chMint »t 9i low Mt/Heat
Heat Size
Figure 7-11. SVS charging off-gas volume vs. heat size.
7-42
-------
HOT METAL CHARGE TIME
O 40 sec
A 65 sec
A 120 sec
O 240 sec
• 300 sec
D Unknown
0.5 r-
0.4
s
o
%
*^
1 0.3
o
o
I •
S
a
£ 0.2
O
co
O
0.1
Kaiser
A
Kimitsu 2 •
Led O
Oita
O Fukuyama
Hilton
Youngstown
O Italsider
Inland
a
1
1
100
Source: Shuldt et al., 1981.
200
Mt/Heat
300
400
Heat Size
Figure 7-12. SVS charging off-gas heat content vs. heat size.
7-43
-------
TABLE 7-4. SVS SYSTEM EXHAUST DATA
Plant
Stelco LEO
Fukuyama
OITA
Klmltsu #2
Inland
Stelco Hilton
Youngs town
Italsider,
Taranto
Bethlehem
Kaiser
Fontana
Start-up
date
1980
1970
1972
1971
1974
1971
1973
1969
1979
BOF
No.
2
2
2
2
2
3
2
3
2
M.T.
230
300
300
220
200
114
240
350
200
Charge
s nrVmln
40 10.000
40 10,000
300 11,300
11,200
Canopy
(7.800)
65 6,120
4,250
240 8,300
120 12,750
°C
200
200
200
200
95
315
15
90
200
Tapping
Gcal/min mVmln °C
0.316 10,000
0.316 5,000 150
0.357 8,400 80
0.354
(0.127)
0.260 Vessel hood
*(0.220)
*0.244
0.403
ReladUng
mVmln
6,000
6,000
10,100
3,700
To small
3,500
Separate
3,000
4,500
3,000
4,250
°C
150
150
150
60
95
filter
120
15
120
200
Gcal/min
0.151
0.151
0.254
0.033
0.057
0.062
*0. 167
0.062
0.134
Other
mVmln
skimming
4,000
desul. ,
9,600
desul. ,
8,350
desul.
Total
°C mVmln
16,000
16,000
deslag
0 14,500
deslag
0 12,600
11,300
16,600
(for 2
17,000
°C
135
150
87
130
120
90
vessels)
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: Shuldt et al., 1981.
Figure 7-13. Charging emissions from a BOF furnace.
-------
-~J
co
Source: Shuldt et al., 1981.
Figure 7-14. Semi-enclosure capturing tapping, slagging, and puffing emissions
from a BOF furnace.
-------
BOF-Hot Metal Reladiing Station
Offtake to
SV.S Duct
Torpedo
Car
Reladiing
Fume Hood
Hot Metal
Ladle
Source: Shuldt et al., 1981.
Figure 7-15. Fume hood arrangement for capture of BOF hot metal
relading emissions.
7-49
-------
en
o
imarv \
Primary
Hood I
.SVS Hood
SVS Hood
B.O.F.
No. 1
16000m3/min, 135° C
Emergency
Relief Vent
Fabric Filter
vvv
1 tlllt
I T Reli
'T
10000 m3/min
200° C
LL
I\|H
6000 m3/min, 150° C
B.O.F.
No. 2
n
Movable Hood
Source: Shuldt et al., 1981.
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 in 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 al., 1983). The working hood system
was deep with 60 degree sides. This feature was included in the present
7-51
-------
i
en
ro
Ladle Crane
50.7 ft.
Scavenger Duct
Ladle Crane
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 60 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 calculations 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 Source Description 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
Baghouse Flow
Operator Cab
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 et al., 1984
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 lime 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. Baghouse exhaust flow rate
2. Wind direction and velocity
3. Height of lime drop
4. Location of clamshell 1n enclosure
5. Amount of lime in clamshell
6. Amount of lime in hopper
7. Rate of clamshell opening
8. Dwell time of clam in 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, 1f not impossible, to perform. Physical modeling of the problem
and solution was therefore used as the basic design tool.
The modeling procedure is described 1n 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 in 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 is
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:
1. The dust plume results from the creation of local air flow caused
by displacment 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
V X 16'
Source: Gilbert et •!., 1984
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' Nature of 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 1n Figure 7-20 1s 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.
The manufacturer refers to it as a slotted-perimeter hood assisted by an
air curtain (Roos, 1981). In contrast to the case study 1n 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 1n 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 1n 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 1s 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
cr>
Exhaust
Isovel Pattern
Aluminum Strip
at Pass Line
Source: Roos. 1981
Figure 7-20. Schematic cross section of an air-curtain hood. Air jets prevent
fumes from exfiltrating into work areas surrounding mill.
-------
DRIVE SIDE
i
(T>
co
•OPEN POSITION
BI-FOLD 01
HOUSIN
(TV
ENTRY
HOOD
Q POST -^
P.)
\
~s*
V
'&
^ M
X
Eg
LL ^
' WINDOW
1
1
/ fL REWIND
£*• 1
r CD _
Hj
*
\
A
_ .
1
^
/
1
-8-
|
• EXIT HOOD
STRIP
TRAVEL
(3000 FPM)
OPERATOR'S SIDE
PLAN
-TELESCOPING
CLOSURE
MILL HOUSING
rMILL
FLOOR
ELEVATION
Source: Roos. 1981
Figure 7-21. Example perimeter hood for control of aluminum rolling mill emissions.
-------
The hood length is 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 ACGIH(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
-------
£ = VN
X
where
VQ = velocity at slot
Vx = velocity at any distance, x, from the slot
N = distance traveled in slot widths.
From the forgoing discussion, Vx at the floor may be taken as 100 ft/min
and N = 30, so that the slot velocity = V30 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 ftVmin. For the
entire hood perimeter of 36 ft, then, the air entrained by the jet is
estimated as 36 ft x 137 ftVmin ft = 4,950 ftVmin. 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
mill 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 and Steel 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 and Steel
Engineer. 59:11-14.
8-1
-------
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-30, 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.
8-2
-------
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.
-------
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-450/3-77-010
(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.
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. Asarco, Inc. Draft
Report. U.S. Environmental Protection Agency. EPA Contract Nos.
68-03-2924, Work Directive 9 and 68-02-3546, Task Assignment
No. 12.
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.
Shuldt, 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. 1966. 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).
-------
ITEM 2
Performance Evaluation Guide
For Large Flow Ventilation Systems
July 1984
-------
EPA-340/1-84-012
Performance Evaluation Guide
For Large Flow Ventilation Systems
by
William Kemner
Richard Gerstle
Yatendra Shah
Contract No. 68-01-6310
Work Assignment No 119
EPA Project Officer: John Busik
EPA Project Manager: Dwight Hlustick
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Stationary Source Compliance Division
Washington, DC 20460
May 1984
-------
DISCLAIMER
The information in this document has been funded wholly or in part by
the United States Environmental Protection Agency under Contract No. 68-01-
6310 to PEDCo Environmental, Inc. It has been subject to the Agency's peer
and administrative review, but it does not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
11
-------
CONTENTS
Figures v
Tables viii
Acknowledgment 1x
1. Introduction 1
2. General Ventilation and Hooding Principles 3
Design basis 5
3. Hood Design Considerations 11
Determining air flow requirements 16
4. Duct Design and Considerations 33
Transport velocities 34
Energy losses 36
Branched systems 43
5. Fan Systems 46
Fan types and operating characteristics " 46
Forced versus induced draft 55
Fan requirements for emission control system applications 57
Fan arrangements 60
Fan drives 60
Fan controls 61
Fan sizing 66
6. Ventilation System Inspection 71
Preparing for inspection 71
Safety considerations 72
Onsite company-inspector interaction 73
Inspection procedures 73
Operation and maintenance (O&M) considerations 79
111
-------
CONTENTS (continued)
Page
7. Total Furnace Enclosures 85
Electric arc furnaces 85
Basic oxygen furnaces 95
8. Special Applications 103
Coke oven sheds 103
Electric arc furnace ventilation 107
Blast furnace casthouse control 115
Control systems on basic oxygen furnaces (BOF's) 117
Building evacuation 123
Copper converters 124
References 128
Appendix A - Bibliography A-l
IV
-------
FIGURES
Number Page
1 Matrix of Hooding Applications for Participate Control 4
2 Example of Open and Closed Hooding on the Discharge End
of a Sinter Strand 6
3 Direct Shell Evacuation on an Electric Arc Furnace 8
4 Variation in Gas Flow Rate From a EOF During the Course
of a Heat 10
5 Conveyor Transfer Point Hooding Using Total Enclosure 12
6 Exterior EAF Canopy Hood 14
7 Effect of Excessive Plume Velocity 17
8 Velocity Contours (Expressed in Percentage of Opening
Velocity) and Streamlines for Circular Openings 19
9 Velocity Contours and Streamlines for Flanged Hood 19
10 Formulas for Estimating Hood Air Flows • 20
11 Dimensions Used to Design High-Canopy Hoods for Hot
Sources 21
12 Schematic Arrangement of Ladle Hood for Reladling
Emission Control 25
13 Controlled Airflow From a Heated Source 28
14 Uncontrolled Airflow From a Heated Source 28
15 General Principle of the Push-Pull (Air-Curtain) Type
System 29
16 Comparison of Face Velocity Decay for Blowing Versus
Exhausting 31
-------
FIGURES (continued)
Number Pa9e
17 Air Curtain Control System on a Capped Converter 32
18 Dew-Point of Air Containing Various S03 Concentrations 35
19 Pressure Measurements in Ducts 38
20 Simple Pressure Diagram 42
21 Taper Duct System 44
22 Centrifugal Fan Components and Layout of a Typical
Industrial Fan System 47
23 Centrifugal Fan Blade Configurations and Impeller
Arrangements 48
24 Fan Testing Procedure and Typical Characteristic Curves 50
25 Typical Characteristic Curves for a Backward-Curved-Blade
Centrifugal Fan 51
26 Typical Characteristic Curves for a Forward-Curved-Blade
Fan 53
27 Typical Characteristic Curves for a Straight-Blade Fan 54
28 Basic Principle of Induced Versus Forced Draft 56
29 Louvre Damper: (a) Parallel Blade Multilouvre; (b) Opposed
Blade Multilouvre; (c) View of Parallel Blade, Multi-
louvre Damper View Showing Linkage 63
30 GuiT->tine Damper: (a) Simplified Cross-Sectional View of
a .^illotine Damper; (b) Guillotine Isolation Damper
Using Seal Air; (c) Top-Entry Type Guillotine Damper,
Showing Operation 64
31 Butterfly Damper: (a) Simplifit Cross-Sectional View of
a Butterfly Damper; (b) Buttenly Damper Showing Hand
Oper r 65
32 Typical Furnace Enclosure 87
31 Furnace Enclosure at North Star Steel Company 91
34 Furnace Enclosure at Birdsboro Corporation 93
vi
-------
FIGURES (continued)
Number Page
35 Typical BOF Furnace Enclosure 97
36 Schematic of Basic Oxygen Secondary Emission Control
System of Kaiser Steel-Fontana 98
37 Schematic of Q-BOP Secondary (Charging) Emissions
Control System of Republic Steel, Chicago 100
38 Various Shed Configurations 104
39 Thermal Expansion of Hot Gases From a Push 105
40 Electric Arc Furnace Utilizing Partial Enclosure 108
41 Ventilation Systems for Electric Arc Furnaces 110
42 Combined Direct Shell Evacuation With Canopy Hood 111
43 Multiple Pickup Points Vented to Common Control Device 112
44 Flow Rate Required for Electric Arc Furnace Control 114
45 BOF Hood Arrangements 119
46 Canopy Hood Concept for BOF Charging Emissions 121
47 Gaw Damper (Closure Plate) Use in BOF Control 122
48 Pierce-Smith Converter 125
49 Secondary Converter Hood Configuration 126
Vll
-------
TABLES
Number Page
1 Comparison of Principal Data for 10, 30, and 100 Percent
Combustion in BOF Hooding 9
2 Range of Capture Velocities 15
3 Range of Design Velocities 37
4 Conversion Table for Duct Velocity to Velocity Pressure 39
5 Relative Advantages of Using Duct Dampers 45
6 Basic Fan Laws 58
7 Basic Damper Types 62
8 Air Density Correction Factor 68
9 Dens-i-y of Common Gases 69
10 Data on EAF Plants Designed for Total Furnace Enclosure 94
11 Flow Balancing of a Typical Furnace Enclosure With
Additional Pickup Hoods, Connected to a Common Baghouse 95
12 BOF/Q-BOP Shops Utilizing Furnace Enclosures 101
13 Example of Flow Balancing of Multiple Evacuation System
on Electric Arc Furnace 113
14 Blast Furnace Casthouse Typical Volume Requirements 116
vm
-------
ACKNOWLEDGMENT
This report was prepared for the U.S. Environmental Protection Agency by
PEDCo Environmental, Inc., Cincinnati, Ohio. Mr. Dwight Hlustick was the EPA
Project Officer. Mr. William Kemner served as the PEDCo Project Manager.
The principal authors were Mr. Kemner and Messrs. Richard Gerstle and Yatendra
Shah. Messrs. Gopal Annamraju, Gary Saunders, and Lario Yerinp prepared
specific sections of the report. The authors wish to thank Mr. Hlustick for
his overall guidance and direction on this task.
IX
-------
SECTION 1
INTRODUCTION
The purpose of this manual is to familiarize Agency inspectors with the
design principles and O&M considerations for large-scale (i.e., generally
>50,000 cfm) ventilation systems commonly found in the metallurgical indus-
try. The emphasis is on the steel industry because most large, complex
systems are found in the many individual processes used in this industry.
Applications in copper smelting and other industries are also discussed.
Inasmuch as ventilation systems are highly complex from a design standpoint
and experience plays a major role in most designs, this manual should be
considered an introductory primer rather than a detailed design manual.
Several standard publications discuss ventilation principles and fan
1-4
engineering. In general, however, these publications emphasize smaller,
more traditional applications. Furthermore, their emphasis is on ventilation
in the general sense, as opposed to air pollution control. The complicating
factors found in ventilation of large metallurgical systems are either
treated in the abstract or not at all. Literature on operation and mainte-
nance practices and inspection procedures is very limited.
Air pollution control systems in the primary metals industry, particu-
larly the steel and copper segments, rely on large capture and ventilation
systems with flow rates commonly in the range of 50,000 to 1,000,000 acfm and
greater. These systems are used primarily to control process fugitive emis-
sions from various furnaces and for building evacuation.
Because these systems are an integral feature of the compliance programs
of the industries involved, this manual was initiated to accomplish the
following:
0 To provide inspection and operation and maintenance guidance to
state and local agency personnel who evaluate the performance of
these systems.
-------
0 To provide a comprehensive treatment of the existing literature
with regard to technical and specific aspects of typical designs.
0 To provide an easy-to-read technical manual on design and operation
for the use of inspectors.
Although not treated in detail in this manual, other technologies are
available for control of process fugitive emissions in the metallurgical
industries. These technologies, which do not involve hooding and ventilation
7
systems, include roof-mounted electrostatic precipitators ' and fume sup-
pression utilizing inert gases to .suppress the oxidation of molten metal.
Both of these technologies offer promise of lower cost than conventional
ventilation approaches. They have been applied for control of charging and
tapping emissions in steel making plants and control of blast furnace casting
emissions. The former have not yet found application in the United States.
Sections 2 through 5 present technical factors of design for hooding,
ducting, and fans. Section 6 describes inspection procedures for use in
assessing the effectiveness and maintenance of ventilation systems. Section
7 deals with total furnace enclosures. In Section 8, the foregoing informa-
tion is supplemented in the context of special problems that are found in
several specific applications. The appendix contains a bibliography for
those interested in pursuing the subject matter further.
-------
SECTION 2
GENERAL VENTILATION AND HOODING PRINCIPLES
Process hooding and ventilation systems are required to capture and
transport emissions to some control device or vent. These hooding and venti-
lation systems sometimes also eliminate potential industrial hygiene problems
by reducing employees' exposure to an air contaminant and by removing heat
from the process area. Figure 1 summarizes the various types of hooding used
in processes in the metallurgical industries.
The three basic parts of a ventilation system are the hood or air in-
take, for initial capture of the emissions; the ductwork, for transport of
the gas stream to the vent or control device; and a fan, to move the gas
stream. Whereas the design of the basic hood and ventilation system is well
understood for small and medium-sized systems, the application of the same
principles to large processes often results in marginal or inadequate sys-
tems, especially when high-temperature processes are involved. This inabil-
ity to apply the same principles results primarily from the large size of the
equipment, the high heat loads, the variability of conditions in batch proc-
esses, the need for access to the process, and greater maintenance require-
ments. For larger systems, much of the design is left to the ingenuity and
experience of the designer, who must fit the hood around the process and lay
out the ductwork with minimal interferences.
Inadequate design of a ventilation system can compromise overall per-
formance. In all cases, the hood must be sized and oriented to capture the
maximum quantity of emissions without requiring excessive gas volumes (a
trade-off between performance and energy consumption). It makes little sense
to install a high-efficiency control device if a major portion of the emis-
sions are not captured initially. The hood should be as close as possible to
the point of generation without interfering with equipment movement and
process operation. It should be oriented to minimize cross-drafts and to
take advantaae of thermal drafts.
-------
Application
M
i.
£
V*.
i
c
fc
%>
i
t.
2
+>
0
Coke oven pushing
Basic oxygen furnace blowing
Baste oxygen furnace fugitives
Electric arc furnact refining
Electric arc furnace fugitives
Open hearth taoplng
Blast furnace ca<»'ig
B1*
-------
The ductwork leading from the hood (or pickup point) to the control
device must be sized to provide the needed transport velocity—generally
between 15 and 25 m/s (2700 and 4500 ft/min) —depending on particulate
loading and size distribution. Layout of ductwork should minimize energy
losses caused by bends, transitions, branches, etc., and should also minimize
air inleakage. If the source is hot, refractory lining or water-cooled hoods
and ducts may be required.
The three basic types of process hooding and venting systems are close-
fitting hoods, canopy hoods, and so-called building evacuation. (The latter
term is used loosely because in very few cases is an entire building actually
evacuated.) More than one of these three systems may be utilized on a single
process.
2.1 DESIGN BASIS
The most important of the three ventilation system components (i.e., the
hood, the duct, and the fan) is the hood. The ventilation system will not
perform well unless the hood effectively captures the emissions. The hood
design and open face area determine the amount of air that is drawn into the
system to capture the emissions. The volume of air and the process emissions
then determine the size of the ductwork, and these factors and the pressure
drop required by the control device in turn determine the size of the fan.
To minimize capital cost and fan power requirements, the designer tries to
minimize the amount of outside air drawn into the system. A face velocity of
200 to 500 ft/min is usually required through the hood's open area. Thus, to
minimize total ventilation air requirements, the hood must fit closely to the
process and have a small open area. The inability to achieve these goals is
the major problem in the applications discussed in this manual. Figure 2
illustrates the enclosure principle applied to a hood on the discharge end of
a sinter strand. In this application, nearly total enclosure is possible.
Under some conditions, the hood may be fitted directly to the process
[(e.g., direct shell evacuation (DSE) on an electric arc furnace (EAF) in
which the furnace roof serves as the hood)]; this arrangement allows only the
process gases (and minimal air infiltration) to pass through the vent system.
Even in these cases, however, high temperatures and explosive gases must be
considered. The qases must be transported at concentrations that are less
-------
INDRAFT AIR
ENCLOSING
HOOD
SINTER / ,-,
STRAND V v'
I i II
NDRAFT AIR
Figure 2. Example of open and closed hooding on the
discharge end of a sinter strand.*
Courtesy: Industrial Ventilation. 16th Ed. Published by American
Conference of Governmental Industrial Hygienists. 1980.
-------
than the lower explosive limit (LEL). In the electric arc furnace arrange-
ment just mentioned, for example, a gap is provided to allow sufficient
infiltration of air for combustion of carbon monoxide, and dilution of its
concentration (CO). This gap also permits the furnace to be tilted as illus-
trated in Figure 3.
In other arrangements, the hood is separated from the process for access
purposes or to allow outside air to mix with and cool, dilute, or combust the
process exhaust. The much larger open area of these arrangements requires
much greater total exhaust flow. Two basic designs are used in the control
of emissions from basic oxygen furnaces (BOF's), the so-called open-hood and
the more-energy-efficient closed-hood approach. Application of a ventilation
system to a BOF is more complex than most applications because the exhaust
gas is primarily carbon monoxide, which, in open-hood systems, is combusted
in the hood by the indraft air. This combustion leads to temperature and
volume variations that must be accounted for in the design of the exhaust
system. With closed-hood systems, the CO is not burned in the hood. The
CO-rich gas is cleaned and then flared or used as a fuel. Thus, for closed-
hood systems, air infiltration must be minimized to avoid explosions in the
gas-cleaning system; its major advantage is the greatly reduced size of the
gas cleaning equipment. Table 1 illustrates the difference in flow rate
associated with these BOF systems. The difference in flow rate also dictates
the type of control device used. A scrubber, for example, would be very
expensive to operate on the high flows resulting from the open-hood approach.
Another common problem in many systems is the variability of conditions.
Conditions may vary from one season to another (i.e., ambient temperature
and/or ventilation requirements), from one heat to another, or even from one
moment to the next as process conditions change. For example, Figure 4
illustrates the variation in gas flow during a heat in the BOF process. This
variation occurs over a period of 15 to 20 minutes.
-------
INFILTRATED
AIR.
GAP
EXHAUST POSITION
TILT POSITION
Figure 3. Direct shell evacuation on an electric arc furnace.
-------
TABLE 1. COMPARISON OF PRINCIPAL DATA FOR 10, 30, AND 100 PERCENT
COMBUSTION IN EOF HOODING3
Varying parameters
Total gas volume, scfm
Theoretical gas temperature
inlet hood, °F
Heat to be removed in hood,
106 Btu/h
Fan horsepower of high-energy
scrubber, kW
Type of hooding and combustion rate
Semi -open
1002
158,800
4352
889
4100
Semi -closed
30%
87,000
3992
325
2200
Closed
10%
66,700
3272
167
1640
Open
200%
318,600
b
b
8200
Reference 8.
NA = Not available.
-------
SHOP A
BLOWING TIME
Figure 4. Variation in gas flow rate f>om a BOF
during the course of a heat.
Courtesy: JAPCA, 18(2):98-101, February 1968. Article by D. H. Wheeler
entit'^d "Fume Control in L-D Plants."
10
-------
SECTION 3
HOOD DESIGN CONSIDERATIONS
The three principles of optimum hood design are:
0 Enclosure of the process or source insofar as possible.
0 Location of exterior hood in path of exhaust.
0 When exterior hood is used, minimization of interference from
cross-drafts.
The goal of good hood design is high capture efficiency. Ideally, a
process should be entirely enclosed, which would permit almost 100 percent
capture efficiency. Simple conveyor transfer hoods (Figure 5) provide an
example of total enclosure. Because frequent access to a process (to charge
materials, remove products, or perform maintenance) is usually required, most
hoods have open areas to provide this access. These open areas must be
maintained under a negative pressure by drawing air into the system, which
prevents fumes from escaping. Although this concept is simple in principle,
its application is complicated by variations in process emissions, thermal
currents from hot processes, and cross-drafts that interfere with the inflow
of air into the hood.
Hoods can be classified into three broad groups: enclosures, receiving
hoods, and exterior hoods. Enclosures usually surround the point of emis-
sion, but sometimes one face is partially or even completely open. Examples
of enclosures are paint spray booths, abrasive blasting cabinets, totally
enclosed bucket elevators, and enclosures for conveyor belt transfer points,
screens, crushers, etc. The sides of the enclosure effectively reduce cross-
drafts and also direct the plume toward the capture hood.
Receiving hoods are those in which the air contaminants are injected
into the hoods and inertia! forces carry these emissions into the hood.
These hoods are generally applied to smaller processes that impart a velocity
11
-------
ENCLOSE TO PROVIDE 150-200 fpm
INDRAFT AT ALL OPENINGS
MIN. Q = 350 efm/ft BELT WIDTH FOR BELT
SPEEDS <200 fpm
= 500 cfm/ft BELT WIDTH FOR BELT
SPEEDS >200 fpm
FOR FALLS GREATER THAN 3 FT WITH DUSTY
MATERIAL, PROVIDE ADDITIONAL EXHAUST QA
BELT WIDTH 12 in. to 36 in. QA = 700 cfm
ABOVE 36 in. QA = 1000 cfm
FLEX STRIPS
RUBBER SKIRT
Figure 5. Conveyor transfer point hooding using total enclosure,
12
-------
to the emissions, such as grinders and paint sprayers. They are not appli-
cable to the large systems discussed in this manual.
Exterior hoods must capture air contaminants that are being generated
from a point outside the hood itself. These hoods are generally used for
large systems that generate heat and require frequent access. Figure 6 is a
simplified diagram of an exterior or canopy-type hood. In this example, the
hood design is augmented with baffles and dampers to direct suction to one of
three sections, depending on the source of the emissions. This enhances
capture efficiency by decreasing the effective face area.
The total air flow into the hood system is determined by "Equation 1:
Q = A V (Eq. 1)
where Q = Total air flow, cfm ?
A = Cross-sectional area, ft
V = Air velocity perpendicular to open face area, ft/min
This simple equation is the root of inadequate system design. Because cost
is directly proportional to flow, Q, the user is continually tempted to
decrease either hood area, A, or face velocity, V. As described later, a
decrease in either of these causes rapid deterioration in hood capture per-
formance.
The desired air velocity, or capture velocity, designed into the hood
must be based on experience, but the guidelines in Table 2 may be helpful.
In many larger industrial processes, the third category—active generation
into a zone of rapid air motion--is encountered, and face velocities in the
range of 200 to 500 ft/min are required. Air motion or currents in the room
may be caused by thermal drafts from hot processes, building drafts, movement
of machinery or material, movement of the process, or rapid discharge of
gaseous emissions.
Because all of these factors cannot be accurately evaluated, a high-effi-
ciency hood must have high face velocities in which a large safety factor is
incorporated. In addition, reducing cross drafts by using partial enclosures
(both fixed and movable) will greatly enhance capture efficiency.
In the design of a hood system, it is useful to consider the concept of
a null point. This point is defined as the point where the inertia! energy
(mass times velocity) of the emission has decreased to zero or been nul-
lified. Because the mass of most emissions (gases and/or particles) is
13
-------
TAPPING
DAMPER
PERIMETER FLANGE
INTERNAL BAFFLESJ
FURNACE ROOF
IN OPEN POSITION
SIDE DRAFT HOOD
TAPPING PIT
FURNACE
Figure 6. Exterior EAF canopy hood.
14
-------
TABLE 2.
RANGE OF CAPTURE VELOCITIES
Condition of dispersion of contaminant
Examples
Capture velocity, ft/min
Released with practically no initial
velocity into quiet air.
Released at low velocity into
moderately still air.
Active generation into zone of rapid
air motion.
Released at high initial velocity into
zone of very rapid air motion.
Evaporation from tanks; degreasing,
etc.
Spray booths; intermittent container
filling.; low speed conveyor transfers;
welding; plating; pickling
Spray painting in shallow booths;
barrel filling; conveyor loading;
crushers; melting and refining
Grinding; abrasive blasting; tumbling
50-100
100-200
200-500
500-2000
NOTE: In each category above, a range of capture velocity is shown. The proper choice of values depends
on several factors:
Lower end of range
1. Room air currents minimal or favorable to capture
Upper end of range
1. Disturbing room air and thermal currents.
2. Contaminants of low toxicity or of nuisance value only. 2. Contaminants of high toxicity.
3. Intermittent, low production. 3. High production, heavy use.
4. Large hood, large air mass in motion. 4. Small hood, local control only.
Courtesy: Brandt, A. D. Industrial Health Engineering, John Wiley and Sons, New York. 1947.
Kane, J. M. Design of Exhaust Systems. Heating and Ventilating, 42, 68. November 1945.
-------
small, their momentum is soon dissipated by air resistance. Hot process
exhausts often have significant momentum due to thermal updraft. Examples
are electric arc furnace emissions and coke pushing emissions. Thermal
momentum can be misinterpreted in the sense that one might think the gases
would be easily captured because they are headed directly into the capture
hood. If the upward velocity is greater than the hood face capture velocity,
however, the gas stream will be deflected to the side as if it had struck a
barrier. This is illustrated in Figure 7.
At the null point the emissions have no momentum of their own, and if an
adequate draft or air velocity toward the hood is provided at the null point,
the contaminants will be captured. What constitutes an adequate velocity
toward the hood depends on the drafts in the area, and therefore cannot be
determined precisely.
Establishing the null point in advance for a new process is not always
possible. For existing equipment, however, direct observation will usually
establish a locus of null points. In the absence of external disturbances,
any positive velocity toward the hood at the null point will give complete
capture. In practice, however, complete capture is difficult to achieve
because of drafts and thermal currents that disturb the air flow and prevent
the formation of tl i null point. Sufficient velocity must be induced to
overcome the disturbance caused by drafts and thermal currents. Because
these drafts and thermal currents vary with the activity near the process, an
exact entrainment velocity cannot be calculated; therefore, a safety factor
must be incorporated to ensure good capture.
3.1 DETERMINING AIR FLOW REQUIREMENTS
3.1.1 Cold Processes
As shown in Figure 8, air moves from all directions toward openings
Q
und«r suction. By definition, flow contours are lines of constant velocity
in front of a hood. Similarly, streamlines are lines perpendicular to veloc-
ity contours.
The equation for air flow around free-hanging round hoods and rectangu-
lar hoods that are approximately square is :
16
-------
HOOD
DEFLECTION ' PLUME VELOCITY, 500 fpm
Figure 7. Effect of excessive plume velocity.
17
-------
» - - 2)
V " 10X2 + A
where V = Center!ine velocity at X distance from hood, ft/min
X = Distance outward along axis, ft (Equation is accurate only for
limited distance of X, where X is within 1.5D.)
Q = Air flow, cfm
A = Area of hood opening, in ft2
D = Diameter of round hoods or side of essentially square hoods
As shown in Equation 2 and Figure 8, velocity decreases rapidly with
increasing distances from the hood and varies almost inversely with the
square of the distance. The velocity decreases less rapidly with a flanged
hood, as shown in Figure 9.
Where distances of X are greater than 1.5D (as is the case in most
applications), the velocity decreases less rapidly with distance than Equa-
tion 2 indicates. Figure 10 illustrates other hood types and gives the air
volume formulae that apply.
\
In addition to canopy-type hooding systems, many other configurations of
hood systems are applied to spray booths, grinding, end open tanks. Few of
these systems have exhaust flows greater than 100,000 cfm.
3.1.2 Hot Processes—High Canopy Hoods
In hot processes, significant quantities of heat are transferred to the
surrounding air by conduction and convection, and a thermal draft is created
that causes a rising air current. The design of the hood and the ventilation
rate provided must take this thermal draft into consideration.
As the heated air stream that rises from a hot surface moves upward, it
mixes turbulently with the surrounding air. The higher the air column rises,
the larger it becomes and the more it is diluted with ambient air. As illus-
trated •'n Figure 11, the rising air column expands approximately according to
the fol owing empirical formula:
Dc = 0.5 xf°'88 (Eq. 3)
where D = The diameter of the hot column of air at the level of
the hood face, ft
18-
-------
STREAMLINE
FLOW
CONTOURS
AIR FLOW
DIRECTION
(TANGENT TO
STREAMLINE)
0 50 100
% OF OPENING DIAMETER
-STREAMLINE
FLOW CONTOURS
50 100
X OF DIAMETER
Figure 8. Velocity contours (expressed in
percentage of opening velocity) and
streamlines for circular openings.
Courtesy: Silverman, L. Velocity Characteristics
of Narrow Exhaust Slots. Journal of In-
dustrial Hygiene and Toxicology, 24,
267. November 1942.
Figure 9. Velocity contours and stream-
lines for flanged hood.
Courtesy: Silverman, L. Centerline Velocity
Characteristics of Round Openings Under
Suction. Journal of Industrial Hygiene
and Toxicology, 24, 259. November 1942.
-------
rvs
o
HOOD TYPE
DESCRIPTION
SLOT
FLANGED SLOT
PLAIN OPENING
FLANGED OPENING
CANOPY
ASPECT RATIO, r
O.2 or /ess
O.2 or /ess
O.2 or greater
O.2 or greoter
and round
To suit work
AIR VOLUME
Q*37LVX
Q*2.BLVX
0* V(K)X*+A)
Q*OJ5V(K)X*+A)
Q*t.4PCV
P'* permerer of
D* height oboye work
Figure 10. Formulas for estimating hood air flows.
Courtesy: DallaValle, J. M. Exhaust Hoods. Industrial Press, New York. 1946.
Silverman, L. Velocity Characteristics of Narrow Exhaust Slots.
Journal of Industrial Hygiene and Toxicology, 24, 267. November 1942.
-------
HYPOTHETICAL
POINT SOURCE
Figure 11. Dimensions used to design high-canopy hoods for hot sources.
12
-------
xf = The distance from the hypothetical point source to the
hood face, ft (equal to y + z)
where z = (2D )1<38 and Dg = diameter of source, ft
y = Distance from top of source to hood, ft
To allow for drift in the rising column of emissions caused by cross-
drafts and air disturbances, the designer must increase the overall hood
diameter by adding 80 percent of the distance between the hood and the proc-
ess, as shown in Equation 4.
Dh = DC + O.By (Eq. 4)
where D. = Overall hood diameter, ft
Where cross drafts occur, the hood diameter may be increased still
farther end the distance between the hood and source may be decreased. When
possible, side shields in the form of steel sheets (curtain wells) or chains
suspended from the hood should be utilized to decrease cross drafts. Asbes-
tos end tarpaulin curtains have been tried, but they are rarely successful
because their light weight makes them tear easily and they do not hang
straight.
The total flow through the hood system may be estimated by the use of
Equation 5:
Vf • 7^09
-------
In addition to the volume of the hot gases rising through the hood area
defined by diameter D (Figure 3-8), room air is also drawn into the hood
through the balance of the hood area. Estimates of the desired velocity of
the air through this portion of the hood depend entirely on engineering
judgment and are based on expected cross-drafts, air disturbances, and the
toxicity of the emissions.
A velocity between 100 and 200 ft/min is recommended, and the velocity
should increase with greater air disturbances. In extreme cases in which
more violent reactions occur with sudden heat release (such as in the charg-
ing and tapping of steelmeking furnaces), even greater velocities are re-
quired:
Q - VfAc + Vr (Ah - Ac) (Eq. 6)
where Q = Total hood flow, cfm
t
Vf = Velocity of hot air, ft/mir
A = Area of hood face through which hot gases enter
C (= *Dc*/4), ft*
V = Desired velocity of air entering balance of hood
(100 to 200 ft/min)
A. = Area of total hood, ft2
The control of emissions from sources that are other than circular in
shape is best handled by hoods of the appropriate shape. Thus, a rectangular
source would require a rectangular hood to minimize the ventilation require-
ments. The equations used for circular hoods are appropriate for rectangular
hoods, but increases of 0.8 times the distance to the source (y) should be
made in both length and width.
Total hood volume determines retention time in the hood. For inter-
mittent processes of short duration, such as charging, a large hood has the
advantage of containing the exhaust gas for several minutes until the ven-
tilation system can withdraw the fumes. Partitions can be added that will
not only minimize cross-drafts, but also essentially increase hood size and
retention time. In any event, it is difficult to predict performance based
solely on theoretical design. Scale model studies can be helpful during the
23
-------
design stage, but final modifications in the field may be necessary based on
observation of system performance.
3.1.3 Hot Processes— Low Canopy Hoods
An exact distinction cannot be made between a low and high canopy hood,
but a low hood is usually defined as one in which the distance between the
hood and the source does not exceed the diameter of the source, or 3 feet,
whichever is smaller. The primary difference in the design of low hoods is
that the hood diameter and source diameter are essentially the same. A
safety factor is usually included, and for practical purposes, the hood
diameter should exceed the source diameter by at least one foot. The dimen-
sions of larger rectangular hoods should exceed the source's dimensions by at
least one foot in all directions.
For circular hoods, the total exhaust flow may be determined by the
12
following equation :
Q = 4.7(Dh)2-33(At)°'417 (Eq. 7)
where D. = Hood diameter in feet and is equal to the source
diameter plus 1 or 2 feet
For rectangular hoods, the exhaust flow may be determined by:
Q = (G.ZJW1-33 (At)°'417L (Eq. 8)
where W = Hood width, which is 1 to 2 feet larger than the source
width
L = Length, which is 1 to 2 feet longer than the source
length
Lowest flow rates are achieved with close-fitting hoods. Figure 12
illustrates the latest design for hooding a hot metal transfer operation.
Note that the open area is essentially limited to the hood slot through which
the metal is poured.
3.1.4 Building Evacuation
A building acts as a large process enclosure. By drawing air through
the building and out the roof, the building essentially serves as a hood.
24
-------
TORPEDO CAR
LADLE HOOD
ro
en
HOT METAL LADLE
Figure 12. Schematic arrangement of ladle hood for reladling emission control
Courtesy: Pennsylvania Engineering Corporation, Pittsburgh.
-------
The large size of a building compared with a more closely fitted hood re-
quires utilization of a much larger exhaust flow. Many large buildings, and
especially those that contain processes that release heat, are ventilated by
natural draft. Heat released by the processes warms the air, which rises and
is pushed out through openings (roof monitors) in the roof. Cooler ambient
air is drawn into the building through openings near ground level. Wind
action at the building openings may either increase or decrease the natural
ventilation rate, depending on the location and size of the openings and the
wind speed and direction.
A forced-ventilation system that is applied to a building must be sized
to include the air flow resulting from natural draft and also maintain suffi-
cient draft into the building to prevent any emissions from escaping through
building openings.
For buildings containing hot processes, the natural ventilation rate can
be estimated by the ASHRAE equation :
Q = 9.4(AL)°-5(Atavg)°-5 (Eq. 9)
where Q = Air flow, cfm
A = Total inlet or outlet air flow area, whichever is
smaller, ft2
L = Building height from air inlet to outlet, ft
At = Difference between average temperature in building
y and air entering building, °F
When the heat released from sources within the building can be quan-
tified, Equation 10 can be utilized:
Q = 20(0.67L)°-33(0.67H)°-33A0-67 (Eq. 10)
where Q = Air flow, cfm
L = Building height from air inlet to outlet, ft
H = Heat released within building, Btu/min
A = Total inlet or outlet area (whichever is smaller), ft2
26
-------
Building ventilation requirements may also be defined in terms of air
changes per unit of time. Again, a great deal of judgment enters into the
selection of the number of changes required, which depends on the heat gen-
erated in the building and the industrial hygiene considerations regarding
fumes and gases within the building. On the order of at least 20 air changes
per hour are required for metallurgical processes.
The distribution of ventilation is also very important. Uncontrolled
air flowing into a building as a result of negative pressure in the building
or because of poorly designed air-supply distributors not only may cause
recirculation of contaminants, but also may upset the local ventilation
systems. Therefore, the amount of air, the location of its entry into the
building, and its direction must be controlled. Figure 13 shows a controlled
air supply that results in a convective flow from a heat source (such as a
ladle of molten metal) rising to be exhausted through a roof ventilator.
Figure 14 shows an uncontrolled air supply, which results in a disrupted
plume and recirculation of the contaminant throughout the building. The
latter could cause a buildup of contaminants in the building and possible
leakage to the outside air.
Because of the huge air volumes it would require (on the order of 5 to
10 million cfm), true building evacuation is rare. Care must be taken in
closing the roof monitors or ventilators in a building and replacing this
natural ventilation with induced-draft ventilation. If the induced draft is
inadequate, both ambient dust and heat levels in the building can rise
rapidly and create health and safety hazards. This is particularly true in
hot climates.
3.1.5 Push-Pull Systems
Hood capture efficiency can sometimes be improved by the use of a push-
pull or air curtain approach. This approach involves a source of compressed
air (push) to direct the emission plume toward the exhaust hood (pull).
These applications are used to control blast furnace casthouses, copper
converters, and electric arc furnace enclosures. Figure 15 illustrates the
general principle of the push-pull system. The effective face velocity of a
27
-------
Figure 13. Controlled airflow from a heated source.
13
Figure 14. Uncontrolled airflow from a heated source.
13
28
-------
NO MAJOR OBSTRUCTIONS
IN PATH OF JET
PRESSURE SLOT
EXHAUST HOOD
PUSH PULL HOODS
Figure 15. General principle of the push-pull (air-curtain) type system.
29
-------
blowing source is sustained at much larger distances than that of a suction
source (as illustrated in Figure 16).
In one system used to control tapping emissions from a blast furnace, an
air curtain in front of the taphole directs the emissions into the capture
hood above the taphole in an arrangement very similar to that in Figure 15.
Figure 17 illustrates the air curtain application on a copper converter
furnace. The total enclosure uses an air curtain to prevent emissions from
escaping the enclosure when the doors ere opened for access. Each appli-
cation is unique in its design and must be evaluated by actual observation.
It is difficult to base suitability on design date alone.
30
-------
FAN
30 diameters
BLOWING
400 f pm
T
4000 fpm AIR
VELOCITY AT
FACE OF BOTH
400 fpm
APPROXIMATELY 10% OF FACE VELOCITY
AT 30 DIA. AWAY FROM PRESSURE JET
OPENING.
EXHAUSTING
APPROXIMATELY 10% OF FACE VELOCITY
AT ONE DIA. AUAY FROM EXHAUST
OPENING.
Figure 16. Comparison of face velocity decay for blowing versus exhausting.
Courtesy: Industrial Ventilation, 16th Ed.
31
-------
JET SIDE
EXHAUST SIDE
AIR
CURTAIN
JET
t
CONVERTER
(FUME SOURCE)
BAFFLE
WALL
TO SUCTION FAN AND
HOOD SAMPLE LOCATION
Figure 17. Air curtain control system on a copper converter.
32
-------
SECTION 4
DUCT DESIGN AND CONSIDERATIONS
The three design principles for ducting are:
0 Minimization of changes in flow direction.
0 Maintenance of smooth duct surfaces.
0 Avoidance of abrupt expansions.
Basically, a duct is a pipe or channel that conveys a gas and contained
emissions from a collection point to a more convenient point for rehandling,
cleaning, or blending. In some cases a duct also acts as a cooler.
Duct configurations range from small-diameter ducting (6 to 12 in.) to
ducting having cross-sectional areas of 600 square feet or more. Ductwork
can be a combination of circular, square, or rectangular, based on location,
space limitations, equipment or building design, and length of run to the
control point. Flows should be continuous and smooth in direction, with no
abrupt expansions or contractions. Larger duct sizes usually are fabricated
in the field in square or rectangular cross sections. Circular ducts are
limited in size by the availability of plate widths. The cost of wider
plates must be balanced against forming and welding costs. Theoretically, a
circular cross section is preferable because it minimizes friction losses and
nonstreamline flow. Duct configuration is generally dictated by economics,
however. Duct thickness in large metallurgical systems usually varies from
1/4 to 3/8 in. Strengthening ribs and expansion joints are generally re-
quired for larger sections (over about 6 feet in diameter or square). Ex-
pansion joints are required for ductwork carrying hot gases (above about
300°F).. If temperature fluctuations occur regularly, cracking of the ex-
pansion joints and the resultant leakage can be a major source of maintenance
problems.
33
-------
Ducting must be designed to deal with the following gas conditions:
temperature, abrasiveness, acidity, dust concentration, and moisture. For
temperature control, the ductwork may be lined with refractory materials or
water-cooled in its entirety. An example of this is the EOF shop, where the
hood and duct above the furnace are cooled by a variety of designs that are
discussed later.
Damage from abrasion, acidity, and moisture may be controlled by special
types of refractory or, depending on the temperature, by the use of various
coatings or special alloyed steels. Carbon steel, for example, generally is
considered suitable for gas temperatures up to 1000°F. Above this tempera-
ture, stainless steel or refractory-lined carbon steel is required. Acid
condensation (primarily from sulfates) is a problem if gas temperature falls
below the acid dew point. In copper converters, sulfur trioxide (SO.,) typi-
cally constitutes about 1 percent of the sulfur dioxide (SO^) present, or
about 0.02 to 0.1 percent of the total gas stream. This can condense and
form sulfuric acid. Figure 18 indicates the dewpoint of air containing
various SO, concentrations.
4.1 TRANSPORT VELOCITIES
The minimum design duct velocity is that required to prevent buildup in
a duct. At eH ws and other sections where the gas stream slows down, pro-
vision should be made for inspection for dust dropout and for cleaning. If
the substances in the gas tend to be sticky, or if moisture condensation is
possible, cleanout ports or flanged sections should be provided to gain
access. Buildup decrees the effective duct cross-sectional area and in-
creases transport veloc ; the latter counterbalances the former and thus
prevents further buildup. If buildup is due to stickiness or moisture,
however, it can proceed the point of total pluggage, especially in smaller
branch ducts. A further concern caused by buildup is the possibility of its
breaking loose during startups or vibration. Serious buildup can also create
an excessive mechanical load on the duct structure and supports.
Normally, the minimum transport velocity required is greater than that
required merely to prevent settling or buildup. Buildup or pluggage can
cause system upset in the main duct or in any of the branches, and this in
34
-------
OJ
tn
Figure 18. Dew-point of air containing various $03 concentrations.
Courtesy: "Combating Fuel Oil Heating Problems", Plant Engineering, January 7, 1974.
-------
turn affects the remainder of the system because of the increase in resist-
ance and decrease in flow in the blocked section. Anything that decreases
the duct size (such as damage from outside forces) affects overall collection
performance.
A change in the temperature of the gas at any point can also affect the
overall collection efficiency of the system.
The transport velocity must account for the velocity needed for gas and
particulate removal, resistance due to friction along the duct surface, and
dynamic losses due to air turbulence.
Table 3 gives a recommended range of design velocities. Metallurgical
process control systems with heavy dust loadings operate in the transport
velocity range of 3500 to 4500 fpm. When the dust consists primarily of
small particles at low concentration, lower transport velocities are pos-
sible. Higher velocities are still preferable, however, to decrease duct
diameter. Duct cost savings normally outweigh the energy penalty of the
higher velocity.
4.2 ENERGY LOSSES
Figure 19 summarizes the three pressure measurements of concern in a
duct. The sum of the energy losses that must be considered in the duct
system are:
0 Inertia - The energy required to accelerate the gas from zero to
duct velocity is equal to (V /4005)2, where V is the gas velocity
in feet per minute. This vafue is referred to as velocity pressure
or as the velocity head (h ). Table 4 can be used to convert
velocity (V or V) in feet per minute to velocity pressure (h or
VP) in inches of water.2 v
o
Straight duct ^riction - Friction loss (or pressure drop) in
straight duct runs is usually negligible relative to the pressure
drop required for elbows, branch entries, and the control device.
As shown in the following equation, the total loss in 500 feet of
10-foot diameter duct (assuming = 4000 fpm) is about 0.5 in. h20.
The equation used for clean rou- ducts is:
f_ 74 ,_ ...
f - - T-XX - (Eq. 11)
D
36
-------
TABLE 3. RANGE OF DESIGN VELOCITIES3
Nature of
contaminant
Examples
Design velocity
Vapors, gases,
smoke
Fumes
Very fine light
dust
Dry dusts and
powders
Average indus-
trial dust
Heavy dusts
Heavy or moist
All vapors, gases, and smokes
• Zinc and aluminum oxide fumes
Cotton lint, wood flour, litho
powder
Fine rubber dust, Bakelite
molding powder dust, jute
lint, cotton dust, shavings
(light), soap dust, leather
shavings
Sawdust (heavy and wet),
grinding dust, buffing
lint (dry), wool jute dust
(shaker waste), coffee beans,
shoe dust, granite dust,
silica flour, general mate-
rial handling, brick cutting,
clay dust, foundry (general),
limestone dust, packaging and
weighing, asbestos dust in
textile industries
Metal turnings, foundry tumbling
barrels and shakeout, sand
blast dust, wood blocks, hog
waste, brass turnings, cast
iron boring dust, lead dust
Lead dust with small chips,
moist cement dust, asbestos
chunks from transit pipe
cutting machines, buffing
lint (sticky), quick-lime
dust
Any desired velocity
(economic optimum
velocity usually
1000-1200 fpm)
1400-2000
2000-2500
2500-3500
3500-4000
4000-4500
4500 and up
A rule of thumb for items or contaminants not tested is to operate in the
3500-4500 fpm transport velocity range.
From Reference 14.
37
-------
TOTAL
PRESSURE
STATIC
PRESSURE
VELOCITY
PRESSURE
Figure 19. Pressure measurements in ducts.14
38
-------
TABLE 4. CONVERSION TABLE FOR DUCT VELOCITY TO VELOCITY PRESSURE15
vfl, fpm
400
500
600
700
800
900
1,000
1,100
1,200
1,300
1,400
1,500
1,600
1,700
1,800
1,900
2,000
2,100
2,200
2,300
2,400
2,500
2,600
2,700
2,800
2,900
3,000
3,100
3,200
3,300
hv. in.H20
0.010
0.016
0.022
0.031
0.040
0.051
0.062
0.075
0.090
0.105
0.122
0.140
0.160
0.180
0.202
0.225
0.249
0.275
0.301
0.329
0.359
0.389
0.421
0.454
0.489
0.524
0.561
0.599
0.638
0.678
v , fpm
a
3,400
3,500
3,600
3,700
3,800
3,900
4,000
4,100
4,200
4,300
4,400
4,500
4,600
4,700
4,800
4,900
5,000
5,100
5,200
5,300
5,400
5,500
5,600
5,700
5,800
5,900
6,000
6,100
6,200
hv, in.H20
0.720
0.764
0.808
• 0.853
0.900
0.948
0.998
1.049
1.100
1.152
1.208
1.262
1.319
1.377
1.435
1.496
1.558
1.621
1.685
1.751
1.817
1.886
1.955
2.026
2.098
2.170
2.244
2.320
2.397
39
-------
where
f = friction loss in inches of water per 100 feet
V = velocity in fpm in duct
D = inside diameter of duct in inches
Actual values can be twice as high because duct internal surfaces
are not ideally clean and smooth. Note that as duct diameter
decreases and length of ductwork increases (as in a system with
multiple miscellaneous pickup points), pressure drop can become
significant. If the system fan is not designed to provide this
pressure drop (for example, where miscellaneous pickup points have
been added letter), the duct pressure drop will result in less
suction at the hoods.
A rectangular duct can be converted to the circular equivalent in
the following manner:
1. A = the duct cross-sectional area in square feet
2. P = the perimeter in feet
3. R = -5- the hydraulic radius in feet
4. 12R = Conversion of R to inches, r
5. D = 4r = equivalent diameter in inches
Elbows - Losses for 90-degree elbows are determined as equivalent
resistance in feet of straight ;Juct. For other elbow angles, use:
60-degree elbow = 0.67 x loss for 90-degree elbow
45-degree elbow = 0.5 x loss for 90-degree elbow
30-degree elbow = 0.33 x loss for go-degree elbow
For radius of 1.5D:
Equivalent feet = 130 (jjp) (Eq. 12)
For a radius of 2D:
1.171
Equivalent feet = 89 (—-) (Eq. 13)
40
-------
For radius of 2.5D:
n '
Equivalent feet = 73 (-) (Eq. 14)
Branch Entry - Losses due to branch entry are best expressed as the
equivalent feet of straight duct of the same diameter. The equiva-
lent length is added to the actual length of straight duct and the
loss is computed from the earlier friction loss equation.
For an entry angle of 30 degrees:
n 1>214
Z = 20 (^g-) (Eq. 15)
where
Z = equivalent feet
D = diameter in inches
For an entry angle of 45 degrees:
D
Z = 32 y~) (Eq. 16)
All branches should enter the main duct at the large end of the
transition, at an angle not to exceed 45 degrees, preferably 30
degrees or less. Branches should be connected only to the top or
sides of the main duct, never to the bottom. Two branches should
never enter a main at diametrically opposite points.
c Contraction and Expansion - When the cross-sectional area of a duct
contracts, a pressure loss occurs. This loss is a function of the
abruptness of the contraction. When the cross-sectional area
expands, a portion of the decrease in velocity pressure becomes
static pressure. The increases and decreases in pressure from
expansion and contraction are calculated from equations in Refer-
ence 15:
The summation of the transport velocity requirements from all of these
losses plus the hood or entry losses plus the control device and stack dis-
charge loss determines the size and power of the fan. A pressure diagram can
be useful in characterizing and understanding a given system. Figure 20
presents a hypothetical pressure diagram for a simple system.
41
-------
DUCT. ELBOW. DAHPER, AND BRANCH ENTRY LOSSES « 1. 1
ro
/CAK
x~
UJ
of
Ul
Of
o.
f
/^
1
OPY HOOD\ /ANOPY
-O.i
2
0
-2
-4
-6
-8
-10
-12
' -0
-
^ CONT
LO
(FAB
HOOD\
• /^~">-
1O
^. DISCHARGE
PRESSURE «1.5
ROL DEVICE
55 * 10
RIC FILTER)
A
\
I
L
\
\
k-12.8
r
A
TOTAL PRESSURE REQUIRED TO SIZE FAN
Figure 20. Simple pressure diagram.
-------
4.3 BRANCHED SYSTEMS
The addition of new or additional pickup points that are in turn tied
into a main duct system has to be carefully designed to achieve the desired
flow balance throughout the system. If duct sizes are not compatible with
the static pressure throughout the system, the desired air flows will not be
achieved. Thus, it is necessary to provide a means of distributing air flow
between branches, either by balanced design or by the use of dampers. Figure
21 illustrates the widely used approach of tapered cross section for branch
entry systems. This design maintains constant duct velocity in the main
duct.
The two approaches for designing duct systems are 1) to balance duct
diameters to match desired flows in each branch, and 2) to use dampers or
blast gates to control flow. Many systems use a combination of both ap-
proaches.
Balanced design is theoretically preferable to minimize the tampering
with dampers and the dependence on operating personnel. The design calcula-
tions begin at the branch of greatest resistance and proceed as follows:
branch to main, section of main to section of main, main to control equipment
on to the fan. At each junction point, the static pressure necessary to
achieve the desired flow in one stream must match the static pressure in the
other.
Damper adjustments permit the desired flow to be varied through each
portion of the system after initial design. This permits adjustments to be
made by trial and error to accommodate variation in actual conditions. The
design calculations begin at the branch of greatest resistance, and pressure
drops are calculated through the branches and duct sections to the fan. At
each point where two gas streams meet, the combined flow is then used, and
when it reaches the main duct, this combined flow is added to the main duct
flow with no attempt to balance the static pressure in the joining gas
streams. Branches are sized for the desired minimum duct or transport veloc-
ity at the desired flow rate.
Advantages and shortcomings of these two methods are further outlined in
Table 5.
43
-------
SIZE FOR BALANCE
AND TRANSPORT
VELOCITY
BRANCH
DUCTS
TO FAN
Figure 21. Taper duct system.
Courtesy: Industrial Ventilation.
-------
TABLE 5. RELATIVE ADVANTAGES OF USING DUCT DAMPERS
14
Balance without dampers
Balance with dampers
Air volumes cannot be easily
changed by the operator.
Small degree of flexibility for
future equipment changes or addi-
tions; the ductwork is "tailor-
made" for the job.
Choice of exhaust volumes for a
new unknown operation may be in-
correct; in such cases some duct-
work revision is necessary.
No unusual erosion or accumulation
problems.
Ductwork will not plug if veloci-
ties are chosen wisely.
Total air volumes are slightly
greater than design air volumes
because of the additional air
handled to achieve balance.
Poor choice of "branch of greatest
resistance" will show up in design
calculations.
Layout of system must be in com-
plete detail, with all obstructions
cleared and length of runs accu-
rately determined. Installations
must follow layout exactly.
Air volumes may be changed rela-
tively easily.
Greater degree of flexibility for
future changes or additions.
Correction of improperly esti-
mated exhaust volumes is easy
within certain ranges.
Partially closed dampers or blast
gates may cause erosion and
thereby change the degree of
restriction or cause accumula-
tions of material.
Ductwork may plug if persons have
tampered with the blast gate
adjustment.
Balance may be achieved with
design air volume.
Poor choice of "branch of
greatest resistance" may remain
undiscovered. In such case the
branch or branches of greater
resistance will be "starved."
Leeway is allowed for moderate
variation in duct location to
miss obstructions or interfer-
ences not known at time of
layout.
45
-------
SECTION 5
FAN SYSTEMS
Centrifugal fans are normally used in conjunction with large emission
control systems because of the large flow rates and high pressure drops in
these systems. Although some axial fans can deliver large volumes of air at
high resistance, they are best suited for clean-air applications. The pres-
ence of dust causes rapid erosion of axial fen components because of the high
tip speed of the fan and the high air velocity through the fan housing.
Centrifugal fans can be designed for differing gas characteristics encountered
in various emission control applications. This section discusses the major
aspects of centrifugal fans with regard to their applicability for large
ventilation systems.
5.1 FAN TYPES AND OPERATING CHARACTERISTICS
A centrifugal fan is used to transfer energy to gases by centrifugal
action. Figure 22 shows the components of the centrifugal fan end layout of a
typical industrial fan system. It consists of a wheel or rotor that is ro-
tated by an electric motor in a scroll-shaped housing. The gases enter the
housing axielly, make a right-angle turn, and are forced through the bledes of
the rotor end into the housing by centrifugal rorce. The centrifugal force
imparts velocity pressure to the air, and the Diverging shape of the scroll
conver a portion of the velocity head into static head.
Certrifugal fans are classified according to the following blade configu-
rations:
0 Backward-curved blade
0 Forward-curved blade
Straight blade
Figure 23 shows a few basic centrifugal fan blade configurations and impeller
arrangements.
46
-------
COUPLING INBOARD END
DRIVE UNITVEAVRING
IMPELLER
OUTBOARD END
STEEL PEDESTAL
BEARING
STEEL PEDESTAL
MAIN SHAFT
GRADE LINE
CONCRETE PEDESTAL
Figure 22a?
Courtesy: Hydrocarbon Processing, June 1975. Article by J. W. Martz
and R. R. Pfahler entitled "How to Troubleshoot Large Indus-
trial Fans."
SCROLL SIDE
BLACKPLATE
BLADES
OUTLET
OUTLET
AREA
\^
CUTOFF
INLET
SCROLL
SUPPORTS
INLET COLLAR
BEARING SUPPORT
Figure 22b.
Courtesy: "With permission of the American Society of Heating, Refrig-
erating & Air Conditioning Engineers, Inc., Atlanta, GA.
Figure 22. Centrifugal fan components and layout of a typical industrial
fan system.
47
-------
RADIAL BLADE FAN
RADIAL TIP FAN
SINGLE INLET
AIRFOIL FAN
DOUBLE INLET
AIRFOIL FAN
Courtesy:
Figure 23a.
Hydrocarbon Processing, June 1975. Article by J. W.
Martz and R. R. Rfahler entitled "How to Troubleshoot
Large Industrial Fans."
BACKWARD-
INCLINED
STRAIGHT
AIRFOIL
FORWARD-
CURVED
Figure 23b.
Courtesy:
Excerpted by special permission from CHEMICAL ENGINEER-
ING (date of issue) Copyright (c) (year), by McGraw-
Hill, Inc., New York, N.Y. 10020.
Figure 23. Centrifugal fan blade configurations and impeller arrangements.
48
-------
The size, shape, and number of blades affect the operating characteris-
tics of the fan. Fan performance is characterized by the volume of gas flow,
pressure, fan speed, power requirement, and operating efficiency. The rela-
tionship of these parameters is measured according to the testing methods
sponsored by the National Association of Fan Manufacturers or the American
Society of Mechanical Engineers. The fan is tested from shutoff conditions to
free-delivery conditions. At shutoff, the duct is completely blanked off; at
free delivery, the outlet resistance is zero. Between these two conditions,
various flow restrictions are placed at the end of the duct to simulate vari-
ous operating conditions. The operating parameters are measured at each test
point and plotted against volume on the abscissa. Figure 24 illustrates the
fan testing procedure and shows the typical fan characteristic curves.
Each fan type has a different performance characteristic. The fen per-
formance curves are used in the selection of a fan type. Generally, the
characteristics of geometrically similar fans are identical. The fan manu-
facturers can predict the performance of a large fan from the tests on a
smaller but geometrically similar fan.
5.1.1 Backward-Curved-Blade Fan
The blades in a backward-curved-blade centrifugal fan are inclined in a
direction opposite to the direction of rotation. The blades (usually 14 to
24) are supported by a solid steel backplate and shroud ring. The scroll-type
housing permits efficient conversion of velocity head into static head.
The characteristics of a backward-curved-blade fan are shown in Figure
25. The static pressure of this fan rises sharply from free delivery to about
50 percent volume point. Beyond this point and up to the no-delivery point
the pressure remains approximately constant. Maximum efficiency occurs at
maximum horsepower input. The horsepower requirement is self-limiting; it
rises to a maximum as the capacity increases and then decreases with addi-
tional capacity. This self-limiting horsepower characteristic of the back-
ward-curved-blade centrifugal fan prevents overloading of the motor when the
fan load exceeds its design capacity. The operating efficiency of the back-
ward-curved-blade fan is high, and this fan develops higher pressure than the
forward-curved-blade fan.
49
-------
}*—BLANKED OFF
RESTRICTED
—LESS RESTRICTED
FLOW STRAIGHTENERS
WIDE OPEN
VOLUME FLOW RATE, Q /
FREE DELIVERY
Figure 24a.
Courtesy: ASHRAE Handbook. Equipment Volume.
SP VP
MEASURING
STATION
i i i i I I i i
0 10 20 30 40 50 60 70 80 90 100
PERCENT OF WIDE OPEN VOLUME
Figure 24b.
Source: U.S. Environmental Protection Agency. Standards Support Documents:
An Investigation of the Best Systems of Emission Reduction for Elec-
tric Arc Furnaces in the Steel Industry. (Draft) June 1974.
Figure 24. Fan testing procedure and typical characteristic curves.
50
-------
VOLUME
Figure 25. Typical characteristic curves for a
backward-curved-blade centrifugal fan.
51
-------
Because the backward-curved blades are conducive to buildup of material,
they are not recommended for dirty streams. These fans are generally used in
ventilating applications where large volumes of clean air are to be handled on
a continuous basis. When used for emission control applications, the back-
ward-curved-blade fan must be installed on the clean-air side of the control
system.
5.1.2 Forward-Curved-Blade Fan
The forward-curved-blade fan generally has 20 to 64 blades. The blades
are shallow, and both the heel and tip are curved toward the direction of
rotation. The rotor of the forward-curved-blade fan is known as a "squirrel-
cage" rotor. A solid steel backplate holds one end of the blade, and a shroud
ring supports the other end. The scroll design is similar to that of the
backward-curved-blade fan.
As shown in Figure 26, the static pressure of this fan rises from a free
delivery to a point at approximate maximum efficiency, drops to about the 25
percent volume point, end then rises back up to the no-delivery point. Horse-
power requirement increases with volume. Because horsepower increases rapidly
with capacity, there is a danger of overloading the motor if system resistance
is not accurately estimated. Forward-curved-blade fans are designed to handle
large volumes of air at low pressures. The fan speeds ere relatively low, end
the pressures developed by forward-curved-blade fans are generally insuffi-
cient for emission control system applications. These fans ere used exten-
sively in heating, ventilating, and air conditioning applications.
5.1.3 Straight-Blade Fan
Straight-blade fans are the simplest of all centrifugal fens. The fan
usually has 5 to 12 blades, which are generally attached to the rotor by a
solid steel backplate or a spider built up from the hub. The rotors are
relatively large in diameter.
Figure 27 shows the performance characteristics of the straight-blade
fan. The static pressure of this fan rises sharply from free delivery to a
maximum point near ro delivery, where it falls off. Mechanica, efficiency
rises rapidly from no delivery to a maximum near maximum pressure, and then
drops slowly as the fan capacity approaches free delivery.
52
-------
VOLUME
Figure 26. Typical characteristics curves for a forward-curved-blade fan.
53
-------
VOLUME
Figure 27. Typical characteristics curves for a straight-blade fan.
54
-------
The straight-blade fan can be used in exhaust systems handling gas
streams that are contaminated with dusts and fumes. Various blade designs and
scroll designs have been developed for specific dust-handling applications.
5.1.4 Backward-Inclined Blade Fan
The two types of backward-inclined blades in the centrifugal fan class
are air foil blades and flat blades.
5.1.5 Rim-Type Wheel Fans
With abrasive material, or materiel that tends to stick, a rim-type wheel
provides more structural integrity; some of these have back 'plates on them.
Literally dozens of rim-type fans are on the market. Each fan typically has
at least six blades. The redial tip tends to develop more static pressure so
it can develop the same gas flow as the straight-blade radial at higher pres-
sure drop.
5.2 FORCED VERSUS INDUCED DRAFT
The terms "forced draft" and "induced draft" come from boiler technology,
where they refer to either forcing air through the boiler (with a blower) or
pulling air through the boiler with a fan located on the exhaust side. Figure
28 illustrates the meaning of these terms in the context of air pollution
control ventilation systems. Essentially, the control device takes the place
of the boiler. Two considerations are important in the use of forced-draft
systems:
1. Whether the fan is exposed to cleaned gas or dirty gas
2. Whether the control device is under pressure or suction
The suitability of the centrifugal fan for dirty applications has already
been discussed. The axial fan would clearly only be suitable in an induced-
draft system. The forced-draft system is generally preferred in large ap-
plications, particularly those controlled by a fabric filter, because control
costs are lower, even when the higher fan cost and maintenance are taken into
account. Many large (i.e., 300,000 cfm and greater) systems treat relatively
clean oas because the process gas has been diluted with indraft air. Thus,
55
-------
DUCT
SOURCE
C>
TO
STACK OR
MONITOR
CONTROL
EQUIPMENT(UNDER PRESSURE)
FORCED-DRAFT
FAN (Dirty)
SOURCE
TO
STACK
CONTROL
EQUIPMENT(UNDER SUCTION)
INDUCED- DRAFT
FAN (Clean)
Figure 28. Basic principle of induced versus forced draft.
56
-------
even the "dirty" gas inlet to the fan is relatively low in dust concentra-
tions, i.e., 1.0 gr/scf or less. Also, this type of system does not usually
require a stack. Systems that use scrubbers or electrostatic precipitators,
on the other hand, usually have induced-draft fans. In scrubber systems with
variable pressure drops, the induced-draft fan is especially needed to control
pressure drop.
Forced-draft fans that are used on the "dirty" gas side are usually
larger and rotate at a slower speed than fans on the clean side to cut down on
the abrasion of the fan blades.
5.3 FAN REQUIREMENTS FOR EMISSION CONTROL SYSTEM APPLICATIONS
Basically, a fan can develop static pressure without delivering much gas
volume, or it can deliver very little static pressure at high gas volume.
Unfortunately, it cannot do both at the same time. Any given fan cannot
perform beyond the limitation of its operating curve.
Increasing the rotation speed and the gas volume through the system
doesn't shift the system resistance curve, but it permits the fan to overcome
more resistance and move more gas through the system by shifting the fan
operating curve. A fan is basically a volumetric energy machine. A certain
gas volume is carried between each blade. Therefore, if the number of times
the blades pass the outlet is increased, more gas is moved and at an increase
in pressure. Any increase in fan rotation speed increases fan horsepower
requirements. Two ways to control capacity are 1) to put in dampers to waste
static pressure across the damper to reduce gas flow, and 2) to reduce fan
rotation speed. If dampers are added, the fan is still expending energy to
develop pressure drop; therefore, reducing fan rotation speed is more effi-
cient because it develops a lower horsepower and uses less energy than a
damper. Table 6 summarizes the basic fan laws.
In the design of fans for emission control system applications, con-
siderations must be given to the nature of the gases being handled. Emission
control system fans may be subjected to one or more of the following operating
conditions:
57
-------
TABLE 6. BASIC FAN LAWS
Variable
When speed changes
When density changes
Volume
Varies directly with speed ratio
RPM
Does not change
Pressure
Varies with square of speed ratio
RPM 2
Horsepower
p = P
P2 Hl
Varies with cube of speed ratio
3
HP2 = HP1
Varies directly with density ratio
P = P (—\
' o 'i \T> 1
c ID,
Varies directly with density ratio
58
-------
0 High temperatures
0 Corrosive gases
0 Dust-laden gases
0 Presence of abrasive particles
0 Presence of explosive materials in the gases
Special construction materials can provide protection against high tem-
peratures and corrosive properties. Bronze alloys are used for handling
sulfuric acid fumes and other sulfates, halogen acids, various organic gases,
and mercury compounds. Stainless steel is the most commonly used corrosion-
resistant metal for impellers and fan housings. Protective coatings such as
bisonite, cadmium plating, hot galvanizing, and rubber covering provide re-
sistance to corrosive gases. Depending on the particular application, soft,
medium, or firm rubber can be bonded to the metal. A good bond will yield an
adhesive strength of 700 pounds per square inch. Rubber-covered fans have
proved exceptionally durable.
When a fan must handle explosive gases, the construction material should
be such that it does not produce a spark if accidentally struck by other
metal--e.g., bronze and aluminum alloys. In most applications, it is prefer-
able to combust explosive gases prior to entering the ductwork. In the case
of a BOF controlled by the closed-hood system, however, the gas is not com-
busted, so the inherent heating value rising from the contained carbon mon-
oxide can be retained.
Fans handling dust-laden gases must be protected from wear due to abra-
sion. When wear is expected, some manufacturers equip their fans with wear
strips and weld beads. The strips consist of a thick, cross-hatched hardened
floor plate welded to the blade at the centerplate or backplate, and the plate
is built up at the edges with weld beads. The inertia of the dust particles
carries them toward the backplate or centerplate, where the wear plate with-
.stands most of the abrasion. The weld beads, which are angled along the edge
of the wear plate, break up the particle flow and prevent impingement and
scouring. For severe abrasion conditions, fans can be equipped with full-
blade liners or heavy-duty rotors with thick wear plates bolted or welded to
the full face of the blades. The fan construction is such that worn wear
plates can be replaced on site.
59
-------
5.4 FAN ARRANGEMENTS
Whenever possible, the fan should be installed on the clean side of the
emission control system, where it will be subjected to less severe operating
conditions. Plants with multiple emission control systems generally have
separate fans for each control system. If a fen is shared by more than one
emission control system, fan operation can be difficult to control. At such
installations, the fan may operate at less than optimum conditions because it
may be subjected to fluctuating loads.
For applications where fan load is expected to vary widely and the peak
fan load is relatively large, multiple fans may provide better control over
the system operation. A parallel fan system will increase reliability and
flexibility. The performance of two fans in parallel can be predicted by
combining the ordinates and abscissas of the pressure-volume curves of both
fans. At low loads, the parallel system can also be operated as a single fan
system. Individual fan loads can be adjusted to optimize the performance of
the system at a given load.
5.5 FAN DRIVES
Large centrifugal fans are generally driven by three-phase, alternat-
ing-current motors. The two principal types of motors used for driving fans
are 1) the squirrel-cage induction motor, and 2) the wound-rotor induction
motor. Although both motor types are self-starting, special starting controls
are needed in many applications to limit starting time tc acceptable values.
Both types operate at rated load with very little slip.
The squirrel-cage motor takes its name from the rotor construction.
Among the various standard designs, the one designated as Design B is usually
used on fans. This motor is suitable for continuous operation at rated load,
and its starting current is relatively low.
Wound-rotor motors, also known as slip-ring motors, can provide adjust-
able-speed drive if desired. Speed reductions below 50 percent are not recom-
mended, however. Tc^ speed and efficiency are about the seme as fo- the
squirrel-cage motor.
60
-------
5.6 FAN CONTROLS
The throughput of fans may be changed by varying the speed or by changing
the pressure condition at the inlet and/or outlet. Variable inlet vanes
provide control on the inlet side of the fan, and dampers can be used on
either the inlet or outlet side. Each of the fan control methods entails a
loss in efficiency over most of the operating range. Because most large-ca-
pacity fans are ordinarily driven by constant-speed motors, it is usually not
possible to control the fan by varying the speed. Variable inlet vanes can
provide gradual load adjustment; however, these must be purchased with the
fan. Dampers offer flexibility of location and control. A fan system with
dampers can be designed to meet a wide range of load fluctuations. Table 7
lists the basic damper types, and the following subsections provide brief
descriptions of the major types.
5.6.1 Louvre
Louvre or multiblade dampers may be of the opposed-blade or the paral-
lel-blade design (Figures 29a end 29b). An overall view of a parallel-blade,
multilouvre damper is shown in Figure 29c. Parallel-blade louvre dampers can
be closed tight, but they offer little modulation ability. Conversely, op-
posed-blade louvre dampers offer excellent modulation ability, but they cannot
be closed as tightly as the parallel-blade louvre dampers can. Louvre dampers
may be used to regulate and isolate gas flow. For isolation, two dampers can
be used together and sealed by pressurizing the chamber formed by the ductwork
between the dampers with a seal-air fan. A single damper may be used for gas
flow regulation.
5.6.2 Guillotine Damper
A guillotine damper may be a top-entry or bottom-entry design, with or
without seal air (Figure 30). Guillotine dampers for system isolation may be
equipped with a seal-air blower, single- or double-bladed, to pressurize the
sealing space and thus ensure against gas leakage past the damper.
5.6.3 Butterfly Damper
Butterfly dampers are often used for secondary duct runs (Figure 31).
They are mounted by a center shaft that crosses the duct, and the damper plate
61
-------
Generic
TABLE 7. BASIC DAMPER TYPES
Specific
Common designs
Louvre
Guillotine
Butterfly
Blanking plate
Parallel-blade multilouvre
Opposed-blade multilouvre
Top-entry guillotine
Top-entry guillotine/seal-air
Bottom-entry guillotine
Bottom-entry gui11otine/seal-air
Single louvre
Double louvre
Double louvre/seal-air
Single louvre
Double louvre
Double louvre/seal-air
62
-------
\
\
(a)
(b)
(c)
Courtesy: Frisch Division, DAYCO Company, Chicago, Illinois.
Figure 29. Louvre damper: (a) parallel-blade multilouvre; (b) opposed-
blade multilouvre; (c) view of parallel-blade, multilouvre damper showing
linkage.
63
-------
(a)
DAMPER BLADE
STAINLESS STEEL
SEALS
COVER PLATE
PRESSURIZED CHAMBER
GAS
0 C>
r-t
PRESSURIZED
CHAMBER
FAN INLET
T|
(b)
Courtesy: Frisch Division, DAYCO Company, Chicago, Illinois.
F jre 30. Guillotine damper: (a) simplified cross-sectional view of a
c,jillotine damper; (b) guillotine isolation damper using seal air;
(O top-entry type guillotine damper, showing operation.
64
-------
BUTTERFLY
(a)
(b)
Courtesy: Frisch Division, DAYCO Company, Chicago, Illinois.
Figure 31. Butterfly damper: (a) simplified cross-sectional view of a
butterfly damper; (b) butterfly damper showing hand operator.
65
-------
rotates about the shaft from a plane parallel to the gas flow (open) to a
plane perpendicular to the gas flow (closed). Butterfly dampers are used most
often to regulate gas flow, but they have been used for opening and closing
off ducts and fan systems.
5.6.4 Blanking-Plate Damper
The most basic damper is the simple blank-off plate. Blanking plates are
essential when a process must be isolated for the protection of the mainte-
nance crew. Should it become necessary for persons to enter any section of
the ductwork, the blanking plate ensures isolation of that section. When used
in conjunction with positive-ventilation air purge, the blanking plates ensure
the safety of personnel.
Blanking plates are similar to guillotine dampers in that they cut across
the duct opening; however, the track design for a blanking plate is intended
only to guide the plate as it is put in place and bolted down.
5.7 FAN SIZING
The air (base) horsepower requirement of a fan can be calculated by
Equation 17:
Air hp = Qh/6356 (Eq. 17)
where Q = inlet volume, ft3/min
h = total static pressure rise, in. H20
Fan brake horsepower can be obtained by dividing the air horsepower by
the mechanical efficiency of the fan. Mechanical efficiency for most centri-
fugal fan operating points will be 50 to 65 percent. Fan size can also be
determined from the fan manufacturer's catalog. Manufacturers' catalogs
generally include multirating tables that give the operating parameter ranges
for different fan models. Many larger fans, however, are custom-built and
thus not found in multirating tables. If the inspector needs to do a detailed
analysis on a custom-built far he/she must obtain the fan ratings ?*d per-
f Ttionce curves from either the manufacturer or the plant.
66
-------
The multirating tables are generally based on standard conditions. When
air is not at standard conditions, corrections must be made to volume, pres-
sure, and horsepower corrections to select a fan at an "equivalent" volume and
pressure. For a manufacturer's rating table to be used, the fan requirements
must be converted to the density used in the ratings. The density correction
is generally made by using the ratio of air density at standard conditions to
the actual density of the air at the fan inlet. Table 8 provides air density
correction factors at various temperatures and altitudes. Table 9 lists the
densities of various common gases. It is usually sufficient to assume the
density of air because it is the predominant gas, but in some cases (e.g.,
saturated gas from a scrubber), a density correction for composition may be
necessary. A further density correction is theoretically required because the
air at the fan inlet is under suction, which lowers the density. For example,
the correction factor from -60 in. HLO to sea level is about 1.18. Before a
fan is chosen from the multirating tables, the following adjustments should be
made:
1. Determine density factor (d = 0.075 for air at 70°F and 29.92 in.
Hg)
2. If the gas flow is indicated at standard conditions, convert it to
actual fan conditions:
Q -(•) x (, x Vscfm
where Q = actual volume of air entering the fan
A = barometric pressure corresponding to fan
site altitude, psia
T = inlet temperature, °F
Vscfm = volume of air at standard conditions
(70°F end 14.7 psia)
3. Multiply static pressure by the density factor, d
4. Using corrected static pressure and actual gas flow, Q, select fan
from multirating tables
5. Divide the fan bhp selected in Step 4 by the density factor.
67
-------
TABLE
AIR DENSITY COPRECTION FACTOR, d
Altitude, ft.
Barometer, in.
Air temp. , °F
Hg
Wg
-40
0
40
70
100
150
200
250
300
350
400
450
500
550
600
700
800
900
1000
-1000
31.02
422.2
1.31
1.19
1.10
1.04
.98
.90
.83
.77
.72
.68
.64
.60
.57
.54
.52
.47
.44
.40
.37
Sea
level
29.92
407.5
1.26
1.15
1.06
1.00
0.95
0.87
0.80
0.75
0.70
0.65
0.62
0.58
0.55
0.53
0.50
0.46
0.42
0.39
0.36
1000
28.86
392.8
1.22
1.11
1.02
0.96
0.92
0.84
0.77
0.72
0.67
0.62
0.60
0.56
0.53
0.51
0.48
0.44
0.40
0.37
0.35
2000
27.82
378.6
1.17
1.07
0.99
0.93
0.88
0.81
0.74
0.70
0.65
0.60
0.57
0.54
0.51
0.49
0.46
0.43
0.39
0.36
0.33
3000
26.82
365.0
1.13
1;03
0.95
0.89
0.85
0.78
0.71
0.67
0.62
0.58
0.55
0.52
0.49
0.47
0.45
0.41
0.37
0.35
0.32
4000
25.84
351.7
1.09
0.99
0.92
0.86
0.81
0.75
0.69
0.64
0.60
0.56
0.53
0.50
0.47
0.45
0.43
0.39
0.36
0.33
0.31
5000
24.90
338.9
1.05
0.95
0.88
0.83
0.78
0.72
0.66
0.62
0.58
0.54
0.51
0.48
0.45
0.44
0.41
0.38
0.35
0.32
0.30
6000
23.98
326.4
1.01
0.91
0.85
0.80
0.75
0.69
0.64
0.60
0.56
0.52
0.49
0.46
0.44
0.42
0.40
0.37
0.33
0.31
0.29
7000
23.09
314.3
0.97
0.89
0.82
0.77
0.73
0.67
0.62
0.58
0.54
0.51
0.48
0.45
.0.43
0.41
0.39
0.35
0.32
0.30
0.28
8000
22.22
302.1
0.93
0.85
0.79
0.74
0.70
0.65
0.60
0.56
0.52
0.49
0.46
0.43
0.41
0.39
0.37
0.34
0.31
0.29
0.27
9000
21.39
291.1
0.90
0.82
0.76
0.71
0.68
0.62
0.57
0.58
0.50
0.57
0.44
0.42
0.39
0.38
0.35
0.33
0.30
0.28
0.26
10,000
20.58
280.1
0.87
0.79
0.73
0.69
0.65
0.60
0.55
0.51
0.48
0.45
0.42
0.40
0.38
0.36
0.34
0.32
0.29
0.27
0.25
en
oo
Standard air density, sea level, 70°F = 0.075 lb/ft3.
-------
TABLE 9. DENSITY OF COMMON GASES
Gas
Hydrogen
Oxygen
CO
co2
N2
Benzene
Ammonia
so2
Water vapor
Air
lb/ft3
0.0052
0.0828
0.072
0.1146
0.0728
0.2017-
0.0446
0.1697
0.0466
0.075
69
-------
The fan brake horespower obtained from the above calculations is divided
by the motor efficiency to obtain motor horsepower. Motor efficiency for
three-phase motors such as those used on large systems is typically 85 to 90
percent.
Usually, fan horsepower is designed for so-called cold-start conditions,
in which case the density correction factor is not used. This enables the
motor end fan to pull the full quantity of "cold" (i.e., ambient) dry air upon
startup without overload. After startup, when the air reaches operating
conditions of temperature and humidity, it is less dense and the horsepower
load is reduced.
70
-------
SECTION 6
VENTILATION SYSTEM INSPECTION
Careful preparation and planning are vital to a successful inspection
and evaluation of ventilation systems. An .inspection will be meaningful only
if the inspector knows what information he/she wants to collect and is famil-
iar with the equipment at the site. Time invested in a file review will
reduce the inspector's field time and that of the source representative.
Also, if the inspector can obtain all the required data during the inspec-
tion, later time-consuming efforts to secure missing data can be avoided.
Furthermore, if the inspector has performed his/her homework, the plant
personnel are more likely to view the inspector as a professional and to
provide the information and cooperation the Agency needs. This section
presents guidelines to assist the inspector in conducting a successful in-
spection.
6.1 PREPARING FOR INSPECTION
When inspecting a ventilation system, the inspector must record the data
on site for later use in evaluating compliance practices. The following
items will help to ensure that the inspection is complete and that the perti-
nent information is obtained while the inspector is on site:
Plot Plan
The plot plan should show entrances, major buildings, and the process
area to scale and include other appropriate details that provide orien-
tation.
Equipment Drawings
Photographs or sketches of the equipment configuration are useful for
reference or comparison. These should show major process and control
equipment for easy reference at a later date.
71
-------
Process Flowsheet and Equipment Checklist
These should provide the inspector with a clear idea of the operating
procedures, factors affecting emissions, a listing of necessary data to
collect for determining compliance, and data collection methodology.
Before beginning the inspection, the inspector should review the work-
sheets and process flows with the plant's representative at the plant to
assure that the information obtained during the file review is accurate and
up to date. This also informs the plant representative of the inspection
procedures so that he/she can assist the inspector in collecting information.
6.2 SAFETY CONSIDERATIONS
To aovid injury while conducting a thorough inspection, the inspector
must:
0 Wear the requisite safety equipment
0 Be aware of the safety hazards
0 Respect the company's safety procedures
0 Never become overconfident
The last point is particularly important. A relatively new inspector or
engineer may begin to feel like an "oldtimer" after the first few trips to
the plant, and this can be dangerous. For safety reasons, the inspector
should make it a practice to stay with the plant escort.
6.2.1 Safety Equipment
For proper fit, the inspector should have his/her own safety equipment.
The following equipment is recommended:
o
o
o
o
o
o
Hard hat
Safety glasses with side shields or full-cover goggles
Steel-toed safety shoes
Fire-resistant pants and jacket
OSHA-approved respirator (fit-tested)
Heavy-duty gloves
Although incidents may seem ulikely, when the inspector's attention is
often focused on observing emissions and/or operating procedures, he/she can
easily become unmindful of potential hazards. Part of the preparation of the
observation procedures should be to review the location of the observation
point, any required movements, the expected activities in the immediate
72
-------
vicinity, and the availability of an escape route. It is generally advisable
to stay as far away from moving equipment as possible.
Unfortunately, a thorough inspection entails some risk. Respect for the
hazards, familiarity with operations, and constant concern for safety will
minimize the chance of an unpleasant or fatal accident.
6.3 ONSITE COMPANY-INSPECTOR INTERACTION
The success of an inspection depends greatly on the interaction between
the inspector and plant representative. Upon arrival at the plant, the
inspector should be prepared to discuss the following:
0 Authority for the inspection
0 Agency organization
0 Scope, timing, and organization of the inspection (preferred in-
spection agenda)
0 Treatment of confidential data
It is also important to inquire about the operational status of the
equipment to be inspected and the kinds and frequencies of any malfunctions.
If equipment is not operating at or near normal conditions, the inspector
should note the reasons and when the plant expects it to be operating nor-
mally (for followup inspection scheduling).
Before collecting any data, the inspector should observe process opera-
tions for a while to become familiar with process variations, wind patterns,
plume characteristics, etc.
6.4 INSPECTION PROCEDURES
The inspection and testing of large ventilation systems present some
problems not normally associated with smaller ventilation systems. Most of
the problems are associated with access to the ductwork and the physical size
of the system. In many instances, access limits the measurements that can be
taken.
The first criterion in the evaluation of the performance of large sys-
tems is to determine whether capture is proper at the source. Poor lighting
sometimes limits visual observation of capture capabilities and makes it
difficult to evaluate the effectiveness of the system. The second criterion
is to determine the integrity of the ductwork to the control equipment.
73
-------
Again, access conditions may limit this visual inspection to what can be
observed from the ground or from a nearby roof. Nevertheless, excessive
inleakage is occurring and the need for maintenance should be determinable.
6.4.1 Duct System Inspection
The inspector should have the following before he/she begins the inspec-
tion:
0 Knowledge of the operations in the process.
0 Knowledge of the physical/chemical nature of the process charge
materials.
0 A layout of the operations showing the controlled and uncontrolled
areas.
0 A line sketch showing the elevation(s) and layout of the ductwork;
the locations of the collector, fan, control equipment; and the
flow patterns.
0 Layout(s) and sketch(es) of the duct size (length and diameter)
showing the main duct, branch or interconnecting ducts, and their
respective flow (actual cubic feet per minute); temperature (°F) of
actual gas flow; slope ratio for transition points of branch duct-
ing to the main or interconnecting ducts and their respective
angles; type and thickness of duct lining; and the location, type
and size of blast gates and dampers and how they are controlled.
0 Copies of plant tests for velocity, pressure drop, and temperature,
and the production or process rate during the test.
0 The production or process rate during the inspection.
During the inspection, the inspector should note the condition of the
duct (e.g., erosion, corrosion, rusted-through openings), flanges, expansion
joints, fits of swing-away joints, etc., all of which affect the workability
of a ducting system.
He/she should check to see that any emergency air inlet dampers, such as
those located in the duct on electric arc furnace systems, are closed. These
are designed to open when temperatures are high, and they should close auto-
matically and stay closed during normal operation.
For inspection purposes, the ventilation system is defined as the duct-
work leading from the emission points to the control devices. It is recom-
mended that static pressure taps be made throughout the length of this
74
-------
ductwork to provide data on air inleakage and ductwork plugging, two of the
major problems with ventilation systems. Both will change the static pres-
sure and temperature profiles of the ductwork system. If a problem is indi-
cated, the inspector should carefully examine the ventilation system to
pinpoint the problem.
Duct blockage, which is characterized by a large increase in static
pressure between the blockage and fan, reduces hood face velocity and results
in failure of the hooding to capture fugitive emissions. Blockage typically
occurs in ductwork bends and cooling loops and may be attributable to insuf-
ficient duct velocity (improper duct sizing), sedimentation of dust par-
ticles, or excessive cooling of the gas stream (which tends to change the
particulate matter into sticky particles that deposit in the duct).
Air inleakage can be a major contributor to excessive cooling of the gas
stream. Although air inleakage may not occur at a single point, it is char-
acterized by lower static pressures and lower temperatures downstream of the
inleakage points. Excessive inleakage may result in fugitive emissions due
to decreased collection efficiency at the emission point and, as noted above,
increased potential for duct blockage.
The inspector should check the face velocity and positioning of all
fugitive hooding. Improper positioning of hooding or hood damage can reduce
capture efficiency. Also, a negative pressure of at least 24 to 49 Pa (0.1
to 0.2 in. H?0) should be maintained at fugitive and process emission points
to accommodate any surges in gas volume and emissions.
6.4.2 Fan Inspection
Lower duct static pressures can also indicate an undersized or underper-
forming fan. The fan system components should be inspected for wear or
corrosion and excessive dust buildup or grease accumulation removed. Fan
couplings should be inspected for loose bolts or misalignments. Bearings
should be clean and lubricated. Shaft seals should be inspected for leakage.
Although a certain amount of leakage is tolerable, excessive leakage may
indicate a need for seal replacement.
A major factor to be checked during an inspection is fan vibration.
Vibration is a function of fan speed. Normal vibration amplitudes at dif-
ferent fan speeds are as follows:
75
-------
Normal vibration
Speed, rpm amplitude, in.
400 0.003
800 0.002
1200 0.0013
1800 0.0008
3600 0.0005
Although vibration amplitudes up to 2.5 times the normal value are accept-
able, corrective measures are required when this limit is exceeded.
Excessive vibration can result from several things, including material
buildup on a blade or bearing wear. Excessive vi.bration requires immediate
attention, as it may quickly lead to catastrophic failure of the fan. An
out-of-balance fan wheel rotating at 300 to 700 rpm can fracture the shaft
and break through the housing. This can damage other equipment in the area
and endanger the welfare of people in the immediate vicinity. The inspector
should vacate the area immediately and promptly report any severe vibration
to the plant.
Performance tests may be necessary to ensure the proper fan operation
after major maintenance,, including rebalancing the fan wheel if any repairs
were made to rotating components. Using a pitot tube and manometer, an
inspector can measure velocity pressures at various points in the duct and
calculate the volume flow rate from these readings after correcting them for
density at the operating temperature and pressure. Fan horsepower can be
calculated from readings of voltage and current supplied to the motor, but it
also must be corrected for actual density. The horsepower and static pres-
sure should be plotted on the fan's original characteristic curves for com-
parison.
Each fan has a set of isolation sleeves, the M'ze of which depends on
the fan size. Cracks or tears in the inlet sleeves permit in-draft air to be
pulled in by the fan, which reduces suction from the hood. In high-tempera-
ture operations, inlet sleeves are made of an asbestos compound. In lower-
temperature operations, they are made of neoprene type rubber. Rubber
sleeves should never be painted; any type of paint will attack rubber.
Fans are designed to rotate in a given direction. If electrical connec-
tions are reversed, the fan will still move air, but not efficiently. Fan
rotation direction should be confirmed against design direction.
76
-------
Fan curves cannot be used directly on a scrubber that has water droplet
carryover, as the water droplets represent a mass that the fan has to accel-
erate. A wet fan indicates that a mist eliminator isn't working properly.
This will require the fan to have higher horsepower, and the fan curve will
inaccurately project that more gas is going through the system than there
actually is. Uncondensed moisture in the gas stream also must be corrected
because it changes the density. At some sources, a Fyrite test kit can be
used to check the change in oxygen content throughout the system. Severe
inleakage will cause a sudden jump in oxygen content. This technique cannot
be used for an ambient system.
6.4.3 Gas Flow Check
Gas flow is a useful parameter in evaluating system performance because
it provides an indication of the capture velocity at the hood(s). In gen-
eral, as the quantity of the gas moved through the fan increases, the re-
quired horsepower also increases. This increase is reflected by an increase
in motor current, which is often measured in a control room. Although motor
current sometimes can be measured by a portable clamp-on ammeter, inexperi-
enced personnel should not attempt this kind of measurement, particularly at
high operating voltages. The current flow is a useful indicator for all
types of systems, regardless of the flow control method (i.e., damper type or
speed control). Static pressure across a fan operating at a fixed speed and
without inlet spin-vane dampers also will indicate gas flow; gas flow de-
creases as static pressure or resistance increases.
The use of fan curves is one method of determining gas flow. This
requires the measurement of fan rotation speed, gas temperature, and motor
horsepower or static pressure across the fan. This method cannot be used if
inlet spin vane dampers are used.
The use of a pitot traverse is a more traditional approach to deter-
mining gas velocity. In this method (outlined in EPA Reference Methods 1 and
2) a pitot tube is used to measure velocity profile across the duct and the
known cross-sectional area of the duct for calculating gas volume through the
system. Again, the limiting factor may be access to the ductwork. Long,
difficult-to-handle pitot tubes may be required to obtain the measurements in
large ducts.
77
-------
Once gas volume is estimated, the average hood capture velocity may be
estimated by comparing the measured gas volume with the cross-sectional area
of the hood. This indicates whether the average hood face velocity is ade-
quate to provide the desired capture efficiency. When multiple-hoods are
interconnected, the flow in each branch may have to be measured to establish
the proper flow balance from each hood. The difficulty with this method is
that it provides only an average hood face velocity, which may be substan-
tially different from localized velocities in large ventilation hoods.
Pitot tubes generally will not accurately indicate hood face velocity at
the relatively low velocity encountered at most ventilation hoods (250 to 750
fpm). Therefore, other types of instrumentation must be used such as hot-
wire and low-pressure gauges. Hand-held instruments such as vane and propel-
ler-type anemometers cannot be used to measure localized face velocity.
Hot-wire anemometers may be attached to long probes for measurement of face
velocity at various points across the hood face, but the results would have
to be corrected for the additional wire length (resistance), which would
affect the readings. Again, access would have to be provided for these
tests, which could be impractical or unsafe during source operation. A test
of this nature, however, would establish minimum and maximum hood face veloc-
ities, flow distribution, and average hood face velocity, whereas measurement
of the gas volume/hood opening area provides only an average velocity.
6.4.4 Visual Observations
If no measurements are made, inspection is limited to visual observation
of the system performance, including hood and duct integrity, the effects of
cross-drafts, and operational procedure1- that may affect the capture effi-
ciency of the ventilation sy:',em. Hoods should remain intact and as close to
the source as possible, not only to capture emissions, but also to reduce gas
volume requirements. Ducts should be checked for damage, which can cause
pressure losses, wear, and inleakage. Cross-drafts should be minimal. If
the system is not designed to operate with cross-drafts, it may not perform
satisfactorily. Wind direction and open doors and windows should be noted.
78
-------
6.5 OPERATION AND MAINTENANCE (O&M) CONSIDERATIONS
Because the fan systems used in large ventilation systems represent
substantial capital outlay and are costly to operate, it is in the best
interest of the plants to maintain this equipment in a fashion that lowers
the incidence of failures that lead to excessive downtime and extends the
useful life of the equipment. The fan system should be considered as includ-
ing the fan, the motor and drive systems, the inlet and outlet duct systems,
and any flow-control dampers used to control the quantity of gas being moved
through the ventilation system.
The successful operation and maintenance of the fan system on large
industrial ventilation systems does not depend on any one item, but on the
proper design and operation of several components. Although some items are
more critical than others, all must operate as a unit for the ventilation
system to deliver the desired gas volume most efficiently.
For economic reasons, most single fan installations are limited to gas
volumes of approximately 1.0 to 1.3 million acfm at nominal static pressure
drops. If greater gas volumes are needed than can be economically delivered
by a single fan, two or more fans arranged in parallel may be used. This
arrangement allows the use of several smaller "off-the-shelf" fans, which are
less costly than custom engineered fans; however, when multiple fans are
used, care must be taken to see that the ductwork and fans match, particu-
larly when fans of different sizes are used. This approach permits the use
of smaller individual fan drive motors and also allows the ventilation system
to remain partly operational in the event of the failure of one of the fans
or fan drive motor systems.
Although many components in the fan system may be subject to failure,
two areas are of the greatest concern: 1) improper balance and the resulting
excessive vibration of the fan, and 2) failure of the drive motor system.
Either can result in expensive repairs or replacements. Proper design and a
preventive maintenance program can reduce the incidence of such failure.
Some drive systems are quite large and require motor sizes in excess of
500 hp (up to as high as 8000). The initial purchase is generally propor-
tional to the motor size. On the other hand, the larger systems tend to be
more efficient, and energy savings over the life of the unit may offset the
79
-------
initial cost. Proper matching of motor size with horsepower requirements
also will usually increase motor efficiency.
Most drive motors are at their highest efficiency and lowest power
factor losses when operated at 80 to 90 percent of their maximum rated load.
If motors are always operated at 100 percent of rated load, they may require
more maintenance and may not have sufficient reserve for occasional periods
when more horsepower is needed. The "service factor" rating of the motor,
which usually ranges between 1.0 and 1.15, is an indicator of the sturdiness
of the motor and its ability to run at higher than rated load for an extended
period of time without experiencing damage.
Because many of these fans are used in systems that control fugitive
emissions varying in gas temperature (and gas volume to be captured), the
maximum and minimum gas temperatures must be considered in sizing the fan
motor. If the fan is sized only for gas conditions at the maximum operating
temperatures, problems with motor overload may occur when gas temperature de-
creases. This results in a denser gas, and more energy is required to move
it. Eventually, such continual overloading of the motor may lead to burnout
of the motor windings and motor failure if efforts are not made to minimize
the problem.
Motor startup is also related to motor overload. The larger the motor
size is, the higher the operating voltage required to keep the current flow
at a reasonable level. Operating voltage will typically be 440 to 460 for
motors having 350 hp or less (but in some cases up to 500 hp). For larger
motors, operating voltages generally increase to 2130 or 4260. When a fan
motor is started "across the line," however, it can draw 6 to 7 times its
normal operating current (regardless of voltage) during acceleration of the
fan wheel to its normal rotation speed. Although this current surge is of
short duration and diminishes as the fan approaches its running speed, it
could damage the motor on startup with a cold gas stream and cause circuit
breakers to trip, both at the fan and in other areas of the plant that might
be affected if the surge of current into the fan circuit were to reduce
operating voltages to unacceptably low voltages and cai:e undervoltage trips.
Two methods that can be used to reduce the possibility of circuit
breaker trips or damage to the motor upon startup are reduced-voltage starts
and closed-damper starts. Reduced-voltage starts allow the fan wheel to
80
-------
accelerate to a portion of its rotation speed, and then full voltage is
applied to accelerate the fan wheel and gas to full speed. Although the
current will still surge upon startup, the levels of surge should be more
acceptable, particularly if other circuits are involved. Closed-damper
starts allow the motor to accelerate the fan wheel without simultaneously
moving the gas; much of the current surge results from accelerated gas flow
through the system. Opening the dampers gradually to allow gas to flow after
the fan wheel is rotating permits the flow of current to be controlled.
A third method is to control the rotation speed. This method is similar
to the reduced-voltage start. The primary difference is that this method
uses a variable-speed motor or a transmission coupling, which allows variable
fan rotation speeds. A gradual increase in rotation speed controls the
current surge.
Excessively high currents in the motor create heat in the windings,
which can destroy the winding insulation and result in the loss of windings
because of short-circuiting. Heat buildup in the motor can be a major reason
for motor failure. Other reasons for heat buildup in the motor include
improper or restricted ventilation due to the location of the motor or exces-
sive dust buildup. Motors must be kept clean to maintain the flow of cooling
air through them. Additional cooling considerations may be necessary if fans
are located where ambient temperatures are high. Normally, solid-state
controls for variable-speed motors also must be protected from high tempera-
tures.
Transmission of the motor energy to the fan shaft is usually accom-
plished by direct drive, by V-belts, or through a variable-speed trans-
mission. The belts must be tensioned properly and kept free of grease or oil
to prevent slippage and belt damage. The fluid levels of fluid-drive trans-
missions for variable-speed operation must be maintained (and possibly
cooled) for reliable operation.
Most fan shafts are supported by bearings at the fan housing and at the
drive connections. Worn bearings can cause excessive fan vibration and
increase energy costs. Bearings for smaller fans can be installed with
grease seals and a grease fitting for routine lubrication. The bearing
lubrication should be checked at least daily to ensure that it is adequate.
81
-------
On fans operating with high-temperature gas streams, it is particularly
important to keep the bearings cool to prevent breakdown of the lubricant.
Heat can be transmitted through the shaft to the bearing assembly.
Although most fans are designed with some inleakage of air around the
shaft to the fan housing, this may not be enough for adequate cooling. In
this case, heat fins may be installed on the fan shaft where it exits from
the fan. These fins rotate with the shaft and provide extra cooling surface
by conducting heat away from the bearings. In situations where this approach
is inadequate, a continuous lubrication system that utilizes circulating oil
can be applied. This system, which consists of a pump, circulation lines, a
filter, and a cooling system (usually water), simultaneously provides bear-
ings with continuous lubrication and cooling. More complex than simple
grease lubrication, this system has the disadvantage of requiring continuous
operation.
On large fan systems, the use of vibration monitoring equipment can
prevent continued operation of a fan that is unbalanced or has worn bearings.
If a severely vibrating fan is allowed to continue operating, a "fan explo-
sion" could occur when the fan is no longer able to withstand the stresses
and simply comes apart during operation. This can be extremely dangerous to
personnel working in the area and also costly to repair.
All fan systems are equipped with inspection hatches and the fans should
be inspected through these hatches at least annually to evaluate the severity
of fan wheel wear. Generally, evidence of fan blade wear appears on the
edges of the fan wheel, on the side opposite the inlet. The chance of fan
wear and blade buildup is greater on fans that are i stalled in'the gas
stream prior to the air pollution control equipment than on fans placed down-
stream of the control equipment, and they should be checked more frequently
for wear and vibration. When blades are replaced, the fan usually must be
rebalanced.
Particulate buildup on fan blades can lead to improper fan balance.
Because the buildup on fan wheels tends to be rele'-vely uniform, it does not
aff:-ct the fan wheel balanced until it flakes off, at which time a substan-
tial change in fan bale e can occur. This, in turn, leads to fan vibration
and bearing wear, and Ca.i eventually cause fan wheel failure. This buildup
problem can be severe when the dust is sticky or oily or if the fan follows a
82
-------
wet scrubber, in which case wet gas conditions enhance the buildup problems.
(Water can be applied to a fan that is used after the scrubber to prevent
particulate buildup.) The higher the rotation speed of fans, the more sus-
ceptible they are to the severe inbalance problems of particulate buildup.
Fans with linings should be inspected as often or more often than other
fans to ascertain that the lining is still intact. Care should be taken not
to damage the lining during inspection. Because fans are most often lined
for corrosion resistance, these linings must be kept in good condition to
prevent excessive fan wheel or housing corrosion. The key to the successful
application of lining is good surface preparation, which seems to be as much
an art as a science.
Another area of concern is the delivery of proper gas flow through the
ventilation system. This flow control is usually accomplished through the
use of dampers or by controlling the rotation of the fan speed. The latter
method is the most energy-efficient. The quantity of gas delivered by a fan
is proportional to the change in rotation speed, but the horsepower required
changes by a cubic relationship. Thus, very small changes in rotation speed
can result in significant changes in horsepower. The initial cost of the
speed control equipment will offset the energy savings somewhat.
The next most efficient method of controlling gas flow is the use of
inlet spin vane dampers. As the vanes close, gas is spun in the direction of
the fan wheel rotation. The quantity of gas delivered by the fan is a func-
tion of the difference in inlet and outlet gas velocity. Because the gas is
spinning in the direction of rotation, the difference in velocity is smaller
and the quantity of gas delivered is smaller with no waste in the static
pressure developed by the fan. These dampers should be checked periodically
to guard against improper functioning and excessive wear of the vanes.
The least efficient but simplest method of controlling gas flow involves
the use of blade-type dampers. These dampers place additional resistance
(variable) into the system to control the gas flow. Even at reduced flow
rates, the fan still delivers a high static pressure and thus requires more
energy to move the reduced gas volume. This method is also the simplest to
maintain unless excessive wear occurs.
Air inleakage is one reason why systems may not perform as designed.
Air can flow into the system wherever a leak occurs, e.g., through a hole in
83
-------
the ductwork, an expansion joint, or open branches that should be closed.
The fan will pull gas through the point of least resistance, and this can
leave the ventilation system with inadequate draft. Any damage should be
repaired before it leads to inadequate control of emissions, corrosion (lead-
ing to increased air inleakage), and higher energy costs for transporting
unwanted dilution air though the system.
Ductwork design also affects how much gas the system will deliver. The
shortest, straightest duct is the most desirable for the design, operation,
and maintainance of the system. On large systems, rectangular ducts are
often easier to install than round ones because it is easier to weld large
flat sections of steel; however, particulate matter has a tendency to build
up in the corners of rectangular ducts and close off usable area. For this
reason rectangular ducts may require periodic cleanout. On the other hand,
some fans (e.g., double-inlet centrifugal fans) are more adaptable to rectan-
gular ducts.
Improper design of the inlet design of the ductwork can cause a swirling
gas flow to the fan, which creates an effect similar to that obtained from
inlet spin vane dampers. This can cause reduced gas flow and inadequate
capture by the ventilation system. This is particularly true if smooth
transitions are not provided to the fan, in which case flow-straightening
vanes or redesign of the ductwork may be required.
84
-------
SECTION 7
TOTAL FURNACE ENCLOSURES
This section discusses total furnace enclosures in the metallurgical
industries. It is assumed that the reader is generally familiar with the
metallurgical processes that use these enclosures. The basic ventilation
principles outlined in previous sections are valid for these special appli-
cations.
7.1 ELECTRIC ARC FURNACES
The recent trend in electric arc furnace (EAF) emission control is to
totally enclose the furnace operations. This system allows the collection
of both primary and secondary emissions from EAF operations. Several
smaller furnaces (<100 tons), have installed total furnace enclosures. The
essential features of these enclosures are sliding doors that create access
for the crane and scrap charging bucket and an air curtain to block the
escape of fumes from the roof when the roof doors are parted for crane cable
access during charging. The advantages of total furnace enclosure are:
0 Effective fume capture.
0 Low volumes of air handled as emissions collected at the source.
0 Capture of both primary and secondary emissions.
0 Access for furnace maintenance.
0 Lower noise levels outside the enclosure
0 Lower capital and operating costs.
0 Better working environment and lower roof temperature in the EAF
shop.
0 Minimal effect of cross winds because the entire operation is
enclosed.
85
-------
0 Lower maintenance of cranes and other equipment because of reduced
dust-fall within the shop.
Figure 32 illustrates a typical furnace enclosure.
7.1.1 Description
A typical enclosure should have adequate space around the furnace to
provide a suitable working area and reduce the effect of heat on the enclo-
sure structure and furnace components. The enclosure structure is generally
lined with hi-rib aluminized sheeting and joints sealed with closure strips.
Bi-parting vertical doors in front of the furnace are controlled by air
cylinders to allow the overhead crane to enter the enclosure. A rectangular
roof slot provides clearance for crane cables when the crane is operated
within the enclosure. This slot is closed by roof doors that are pneumati-
cally operated. An air curtain under the roof slot area contains and guides
the emissions to the pickup hood located opposite the air curtain nozzles.
The air curtain fan is usually located on the roof of the furnace transformer
room. Removable roof panels above the roof doors allow access for furnace
maintenance. The furnace operations, air curtain, and pickup hood can be
observed safely from the control room. The control room usually has a full
glass window that forms a part of the enclosure wall. Doors to the enclosure
allow access to the furnace for oxygen lancing and for taking metal and slag
samples.
To charge the furnace, the crane operator positions the charging bucket
in front of the enclosure and aligns the crane cable in front of the roof
slot. The furnace operator swings the furnace roof aside and opens the
front vertical bi-parting doors and the roof doors of the enclosure. The
crane operator then brings the charging bucket into the enclosure. At
ground level, a furnace operator guides the crane operator in positioning
the charge bucket directly over the furnace. The crane operator dumps the
scrap by opening the bottom of the bucket and then reverses the crane out of
the enclosure. The control room operator closes the vertical and roof doors
and swings the furnace roof in position to seal the furnace. The melting
operation commences after the furnace roof is in position. During the
ertire period when the bi-parting doors and roof doors are open, the air
curtain is in operation along with the enclosure exhaust. During melting,
86
-------
CO
LOCAL TAP HOOD
AIR CURTAIN
PICKUP HOOD
BI-PARTING DOORS
TO BAGHOUSE
CRANE RAIL LEVEL
FURNACE LEVEL
GROUND LEVEL
Figure 32. Typical furnace enclosure.
(Not to scale)
-------
oxygen lancing, deslagging, and sampling, the doors are closed and the
enclosure is exhausted to a gas cleaning device, generally a baghouse.
A stack in the fourth hole position of the furnace roof leads the
emissions generated during melting and refining closer to the exhaust hood.
Depending on the design of the enclosure, the furnace tapping occurs either
within or outside the enclosure. In most cases a transfer car brings the
teeming ladle into the tapping position underneath the furnace. If tapping
takes place outside the enclosure, a local hood is provided to capture
tapping emissions. In the design where tapping takes place within the
enclosure, the enclosure exhaust evacuates the tapping emissions.
7.1.2 Design Considerations
Charging, melting, refining, and tapping emissions occur during the
operation of the EAF. The maximum emissions (85 - 90%) occur during melt-
1 O
down or refining. Direct shell evacuation or side draft hoods adequately
control these primary emissions. Several techniques including canopy hoods
and building evacuation control secondary emissions. Although canopy hoods
have few operating restrictions, they have the disadvantages of high volume
requirements, high capital, and high operating costs. Crosswinds within the
shop affect capture efficiency. Local hoods have provided limited success
in collecting tapping fumes. A compromise solution to the high volumes
required for canopy hood systems is a modified canopy approach where fixed
curtain walls form a partial enclosure around the furnace to act as a
chimney to direct charging emissions to the canopy hood, while local hoods
are utilized to control tapping emissions.
The ultimate control of furnace emissions involves furnace enclosure
technology. On small electric furnaces, complete emissions control can be
achieved by collecting emissions at the source by adopting furnace enclosure
technology. Emissions generated during all phases of the EAF operations can
be withdrawn from the enclosure using relatively low flow volumes. If
emission volumes are high as a result of the use of UHP furnaces, oxyfuel
burners, or larger furnaces, however, additional primary controls such as a
DSE (fourth hole) system may be required. The approximate volume of air
required for a jrnace enclosure in a small furnace will be equivalent to
the volume required by a side-draft hood system; a larger furnace will
88
-------
18
require an amount equivalent to the volume required by a DSE system. The
actual air flow required will depend on many factors such as furnace size,
oxygen lancing rate, exhaust-duct position, scrap cleanliness, furnace
power, types of steel produced, operating practices, and enclosure volume.
7.1.3 Retrofit Installations
Because each retrofit installation is unique, enclosures must be
tailored to meet individual requirements, furnace practices, and shop
layout. Examples of two retrofit EAF enclosures in the United States are
found at the North Star Steel Company and Birdsboro Corporation.
North Star Steel Company—
This company has two 60-ton furnaces that are controlled by DSE and a
roof canopy system. The capture efficiency of the canopy system was low
during charging and tapping because of the flat design of the roof canopy
and insufficient extraction volume. Each furnace was retrofitted with an
enclosure. The furnace enclosure is a steel structure c-lad with ribbed
aluminized sheets and sealed. Mechanical bi-parting side doors and roof
doors allow entry of the crane for charging. The bi-parting vertical doors
are closed during charging and an air curtain seals the roof. The entire
available flow is used to extract the charging emissions through a high-
level evacuation duct. Emissions from melting, oxygen lancing, and slagging
operations are controlled by a DSE system. During tapping, mechanical
bi-parting doors on the tapside of the enclosure are opened to permit entry
of the tap ladle. These doors have a top horizontal roof with a slot for
cables holding the tap ladle. Once the ladle is positioned, the bi-parting
doors close to contain the fumes. During tapping, fume is evacuated by a
high-level duct. On the charge side, the mechanical roof door, the roof
section complete with air curtain, and the charge section evacuation duct
can all be mechanically wheeled aside for maintenance operations such as
replacement of furnace parts and removal of the shell and roof. Salient
19
data on this plant are as follows:
Furnaces: Two: 15-ft diameter; 60-ton capacity each
Transformer: 30,000 to 33,000 kVA
89
-------
Meltdown rate: 55 tons/hour
02 consumption: 300 ft3/ton
Approximate enclosure
size: 40 x 36 x 31 ft high
Extraction volume: 170,000 acfm per enclosure
Air curtain volume: 6000 acfm
Figure 33 shows the enclosure at the North Star Steel Company.
Birdsboro Corporation--
This new facility initially designed with a side-draft hood for primary
emission control had to be redesigned to control both primary and secondary
emissions. The furnace enclosure design allowed an acceptable solution with
an exhaust volume of 75,000 acfm for a 40-ton furnace. Tapping and charging
are carried out from the same side. A mechanical hood 'car moves away from
the enclosure to provide an opening for the scrap bucket and tap ladle.
Three cylinder-operated movable roof doors seal the crane cable entry slot.
These doors are linked to dampers in the air-curtain duct system. The air
curtains seal the opening when the doors are opened for charging or tapping.
Overhead removable panels at the enclosure roof level aid maintenance
operations. In the initial startup, the enclosure was not effective during
lancing. Measures adopted to overcome this problem were better sealing of
the furnace enclosure, upgrading the air curtain fan, adjusting air nozzles
to deflect the lancing emissions into the extraction duct, and installing a
19
deflector plate. Salient data on this shop are as follows:
Furnaces: One: 13-ft, 6 in. diameter; 40 tons capacity
Transformer: 10,000 kVA
Maximum meltdown rate: 18.6 tons/hour
Oxygen lancing: 250 scfm (design)
Enclosure size
(approximate): 36 x 30 x 28 ft high
Extraction volume: 75,000 acfm (design); 81,600 acfm
operating
90
-------
Figure 33. Furnace enclosure at North Star Steel Company.
91
-------
Temperature: Normal 150°F; Maximum 275°F
Air curtain volume: 9,900 acfm (design) increased to
11,700 acfm.
Figure 34 shows the furnace enclosure at Birdsboro Corporation.
7.1.4 New Installations
Some of the new EAF installations required to meet NSPS have adopted
furnace enclosures for smaller furnaces (<100 tons). Table 10 is a list of
the essential features of new EAF shops designed for furnace enclosures.
Plant 3 has additional pickup points including a local tapping hood, an
additional hood to evacuate tundish lancing emissions, and a scavenging hood
located above the roof slot of the enclosure, all connected to a common
baghouse. Balancing the draft for such multiple pickup points is difficult,
but adjustments are made with operating experience. A typical flow balance
during various phases of the EAF operations is shown in Table 11. Dampers
on each hood control the flow.
7.1.5 Inspection
To achieve good evacuation, the enclosure should be well sealed. This
will depend on the integrity of the enclosure. The inspector should observe
and note the general condition of the enclosure walls in regard to any loose
sealing strips, misalignment of the bi-parting doors (which creates large
gaps), holes in the panels, damaged or warped panels, open access doors, and
any other openings that reduce the capture efficiency of the enclosure.
During scrap charging and at other periods when the air curtain is operated,
the inspector should observe the air curtain as it diverts the thermally-
driven plume into the pickup hood. Improperly placed air nozzles in the air
curtain may not completely seal the roof door area and thus allow emissions
to escape. The air curtain operation can be safely observed from the
control room. The performance of the enclosure can be assessed by simultane-
ous observations by two inspectors, one inside the control room and the
other outside the enclosure, during all phases of a heat cycle. The inspector
in the control room should observe and note the fit and closure of the roof
doors, i'nd th- buildup of emissions within the enclosure and how effectively
they are evacuated at various stages of the furnace operation. The inspector
92
-------
•EXHAUST DUCT
REMOVABLE PANELS
OVERHEAD MAINTENANCE ACCESS
ROOF DOORS
NORTH
CHARGING AND
TAPPING HOOD CAR
LADLE ADDITIVE CHUTE
ROOF AND HOOD CAR
CONTROL PANEL
FURNACE CONTROL PANEL
EAST ROLL-UP DOOR
SOUTH ROLL-UP DOOR
FURNACE CONTROL PANEL
Figure 34. Furnace enclosure at Birdsboro Corporation.
93
-------
TABLE 10. DATA ON EAF PLANTS DESIGNED FOR TOTAL FURNACE ENCLOSURE20'21
Plant 1
Plant 2
Plant 3
Number of furnaces
Furnace size
Transformer
Type of steel made
Meltdown rate
Approximate enclosure
size
Extraction volume
Air curtains at the
roof slot
Other features
1
12-ft,-6 in diameter;
28-32 tons/heat
16,800 kVA
Specialty
s20 tons/hour
42 x 51 x 35 ft
150,000 acfm
Yes
Vertical bi-parting
doors and roof slot with
doors for crane cable
access.
16-ft diameter;
60 tons/heat
33,600 kVA
Carbon steel
=72 tons/hour
34 x 53 x 44 ft (curved
at the top, dome shaped)
175 to 200,000 acfm at
150°F
Yes
Vertical bi-parting doors
and roof slot with doors
for crane cable access.
1
15-ft diameter;
55 tons/heat
17,000 kVA
Medium-carbon steels
s!6 tons/hour
Not available
150,000 acfm at 130°F
Yes
Vertical bi-parting doors
and roof slot with doors
for crane cable access.
A separate tapping hood
and a scavenging hood above
the crane slot, evacuated
to the same baghouse.
'Each furnace has a separate enclosure.
-------
TABLE 11. FLOW BALANCING OF A TYPICAL FURNACE ENCLOSURE WITH
ADDITIONAL PICKUP HOODS, CONNECTED TO A COMMON BAGHOUSE
(150,000 acfm)
Operating Mode
Furnace charging
Melting and refining
Tapping
Lancing at Lancing
Station
Roof
scavenging
hood
73,000 acfm
73,000 acfm
73,000 acfm
43,000 acfm
Furnace
enclosure
pickup hood
77,000 acfm
77,000 acfm
22,000 acfm
77,000 acfm
Local
tapping
hood
Closed
Closed
55,000 acfm
Closed
Lancing
hood
Closed
Closed
Closed
30,000 acfm
NOTE: Lancing hood cannot be operated during charging or when air curtain is
operated, and during tapping. Roof scavenging hood can evacuate at a
higher rate than 73,000 acfm, if needed.
outside the enclosure should observe and note the shop layout, any leakage
of emissions from the enclosure, crossdrafts inside the shop, emissions
fromancillary operations, additional pickup points, and the overall flow
balance under different operating modes. Although crosswinds generally do
not affect the performance of an enclosure, emissions that build up near the
shop roof above the enclosure will escape the building. If a separate
scavenger hood is provided above the enclosure, its capture efficiency
should be observed.
7.2 BASIC OXYGEN FURNACES
Primary emissions from basic oxygen furnaces are captured in specially
designed hoods. The hood is erected above the mouth of the vessel to
control the primary emissions when the vessel is in the vertical position.
The primary hood is not as effective, however, when the vessel is tilted for
various operations such as charging, sampling, tapping, and deslagging.
Additional hoods or enclosures are required to control these secondary
emissions. The trend is toward total enclosures.
95
-------
7.2.1 Description
The vessel is enclosed on all sides, with a door or bi-parting doors in
the front to facilitate scrap and hot metal charging. Since these
operations occur at and above the vessel, natural convection will permit a
plume of hot dusty gases to escape during charging. To reduce this
possibility, charging should occur as close to the vessel as possible and
under the hood. A separate charging hood is provided within the enclosure
for capturing the charging emissions. The enclosure can extend partially or
completely at the tap side. Tapping is carried out at and below the level
of the vessel, and the hot dusty gases have a tendency to escape in the
natural draft induced by the process heat. In the newer designs, a permanent
tap hood is installed at the back of the enclosure. Figure 35 shows a
typical arrangement of a BOF vessel enclosure. The furnace enclosure
extends below the charging floor, and the only openings are for the ladle
car. These openings can be effectively reduced by the addition of vertical
shields on the end of the ladle car;
In the bottom blown BOF (Q-BOP), oxygen is blown through tuyeres.
During charging and turndown, gas (either oxygen or nitrogen) must be blown
through the tuyeres to prevent liquid metal, slag, or solids from entering
and clogging the tuyeres. This generates heavy emissions and makes capture
of the secondary emissions more difficult; hence, invariably all the Q-BOP
furnaces are completely enclosed.
7.2.2 Gas Cleaning Systems
The gas cleaning system employed to control secondary emissions may be
an extension of the primary control system. One hood designed to collect
charging emissions and another for collecting tapping emissions could be
ducted to the same primary gas cleaning system. The gas flow would be
adjusted for the different demands of the heat cycle. Another alternative
is to duct the charging and tapping hoods in the furnace enclosure to a
secondary c- ^.rol unit, generally a bag filter. Figure 36 indicates a
schematic OT the secondary control system at Kaiser Steel in Fontana,
Cal iform'a.
96
-------
BUMPER
SECONDARY HOOO
HOT METAL CHAAGMO LADLE
FURNACE CHARGMO DOORS
SLAOPOT
WATER COOLED HOOD
HOOD TRANSFER CAR
ADJUSTABLE SKWT
TAPPMG EMISSIONS DUCT
SEAL RING
FURNACE ENCLOSURE
OPERATING FLOOR
TEEMMG LADLE
SHOP AM
•CRAFT DURMC
SLAOGMG ft
TAPPING
•TD-
Courtesy:
'Figure 35. Typical BOF furnace enclosure.
Pennsylvania Engineering Corporation.
-------
I
HOT METAL
TRANSFER
1
V
SLAG
SKIMMING
1
\f
HOT METAL
TRANSFER
2
\ /
SLAG
SKIMMING
2
\
FURNACE
CHARGING HOOD
1
FURNACE
CHARGING HOOD
2
V
TAPPING HOOD
\
TAPPING HOOD
BAGHOUSE
1,020,000 m3/hr
MAXIMUM TEMPERATURE 230°C
EXHAUST FANS
Figure 36. Schematic of basic oxygen secondary emission control system
of Kaiser Steel-Fontana.
98
-------
A common baghouse with a I,020,000-m3/h (600,000 acfm) design flow
serves not only the charging and tapping hoods within the furnace enclosure
but other areas such as the hot metal transfer and slag skimming stations.
In the closed-hood system that intermittently handles a flammable gas
(CO), use of the primary system to control the secondary sources could
result in an explosion.
At the Q-BOP plant of Republic Steel in Chicago, which has a suppressed
combustion or closed-hood system, charging emissions are safely collected by
diverting the emissions to the primary system of the nonoperating vessel.
The shop has two vessels, and one furnace is operated at a time. Figure 37
shows the schematic arrangement of the gas cleaning operation. When both
gas cleaning system fans are operated, a flow rate of 634,000 m3/h (373,000
pp
acfm) is available at the charging hood during hot metal charging. Fumes
captured in the charging hood bypass the quencher and pass directly to the
venturi scrubber. The design pressure drop of the venturi during furnace
22
charging is 218 cm (56 inches) water column.
Table 12 lists the BOF/Q-BOP shops that utilize the furnace enclosure
for emission control.
7.2.3 Inspection
The integrity of the furnace enclosure is important in obtaining good
evacuation within the enclosure. The inspector should observe and note the
condition of the bi-parting doors to detect any misalignments or warpage
that leads to large gaps between the doors. The inspector should also check
for gaps between the water-cooled sections of the duct work and inspect the
condition of the pickup hoods within the enclosure and all the associated
ductwork leading to the cleaning unit.
During scrap charging, the vessel tilt angle should be estimated along
with the effective capture efficiency of charging emissions in the secondary
charging hood. During hot metal charging the tilt angle and the pouring
rate of hot metal play an important part in the effective capture of the
secondary emissions. The primary hood capture and enclosure evacuation
should be observed during slopping. In top-blown furnaces, lanse hole
emissions or fume rollout (puffing) from the primary hood indicate in-
sufficient draft at the primary hood. This will increase the load on the
secondary controls within the enclosure.
99
-------
o
o
NO. 1
Q-BOP
FURNACE CLOSURE
/^\-
(
\
••••
1
;
X
L^^
X
r*
-t
^ i
V
">
f
j\
2
1
L
I
r:
vi/ "
SECONDARY
HOOD NO. 1
SHUTOFF NO. 1
OPEN ,-BELL VALVE NO. 1
|H|
CLOSED" 1
-». if^ -*. c^—
-------
TABLE 12. BOF/Q-BOP SHOPS UTILIZING FURNACE ENCLOSURES9
Company and location
Type of enclosure
Type of gas cleaning
Startup date
Republic Steel Corp.
Chicago, Illinois
Bethlehem Steel Corp.
Burns Harbor, Ind.
CF&I Steel Corp.
Pueblo, Colorado
Kaiser Steel Corp.
Fontana, California
Granite City Steel Oiv.
Granite City, Illinois
Rouge Steel Co.
Dearborn, Michigan
Sharon Steel Corp.
Farrell, Pennsylvania
4-sided enclosure with
mechanized biparting
charging doors
4-sided enclosure with
mechanized biparting
charging doors
3-sided enclosure with-
out charging doors
4-sided enclosure with
single mechanized
charging door
4-sided enclosure with
single mechanized
charging door
3-sided enclosure with-
out charging doors
4-sided enclosure with
single mechanized
charging door
High-energy scrubber (420,000
acfm at 143°F)
High-energy scrubber (350,000
acfm at 145°F)
Fabric filter (540,000 acfm at
275°F)
Fabric filter (500,000 acfm at
600°F)
Electrostatic precipitator
(900,000 acfm at 550°F)
Electrostatic precipitator
(1,050,000 acfm at 500°F)
High-energy scrubber (320,000
acfm at 168°F)
1976-77
1977
1978
1978
1980
1981
1982
Source: Pennsylvania Engineering Corporation.
-------
The inspector should observe and note the shop layout, gas cleaning
equipment, and multiple-pickup points connected to the same gas cleaning
source and become familiar with the flow-balancing scheme under different
operating modes. This is very important as each plant is unique and equip-
ment operation varies from plant to plant.
102
-------
SECTION 8
SPECIAL APPLICATIONS
The basic principles of ventilation outlined in previous sections are
valid for all applications. This section discusses some additional con-
siderations of interest in common control systems in the metallurgical indus-
tries. It is assumed that the reader is already generally familiar with the
processes involved.
8.1 COKE OVEN SHEDS
8.1.1 Description
The coke oven shed for control of coke oven pushing emissions is a
special kind of hooding. Many variations in configuration are possible
(Figure 38) but all are designed to meet three basic objectives:
0 To contain excessive emissions during the brief (45-second) period
of a single push; cooling and exhausting of these emissions take
place over a longer period.
0 To arrest the upward travel of the plume and to roll the plume
within the shed; heavier particles fall out to the ground.
0 To contain all of the emissions from the coke side of the battery,
including door emissions, coke spillage emissions, and hot car
emissions.
Shed control systems require ducting, fans, and a control device, usu-
ally a fabric filter. This discussion is limited to the ventilation princi-
ples and O&M considerations associated with the shed itself.
Figure 39 presents a general illustration of the thermal expansion of
hot gases from a push that the shed must accommodate. In the original shed
designs, a duct ran along the length of the shed, and the exhaust suction was
distributed along the entire length. These sheds required high flow rates
(300,000 acfm or more) to achieve reasonable face velocity. Newer designs
103
-------
Figure 38. Various shed ;onfigurations.
104
-------
1MIN
2MIN
3MIN
o
tn
•ASE AMBIENT TEMP.
4MIM
TIME
Figure 39. Thermal expansion of hot gases from a
-------
entail a partition approach in which dampers are used to concentrate the
suction in the segment of the shed where the push is occurring. In either
design, an inherent design principle is the holding of it within the large
confines of the shed until it can be fully exhausted. The design relies on
the theory of fully exhausting a given push before the next push occurs.
Because containment of the plume is an important factor, several dif-
ferent shed configurations have been offered to maximize containment. A
common feature of all designs is an enlarged upper portion of the shed, which
serves as a plenum (see Figure 38). The containment volume must be designed
to match the size of the push and pushing rate expected at the battery.
A potential weak point in shed control is the influence of wind cur-
rents, especially at the ends of the shed. Even in the absence of wind
disruption, emissions from "dirty pushes," i.e., heavy plumes, have a tend-
ency to "roll out" from under the shed near the ends. The only two solutions
to this problem are
1) to close the ends insofar as possible to interrupt wind currents, and 2)
to extend the overall shed structure beyond the ends of the battery. Sheds
can be designed to extend beyond the battery toward the quench tower to
capture emissions during travel of the hot car.
8.1.2 Inspection
In the inspection of a coke-side shed, the inspector should consider the
following:
1. The integrity of the shed structure and any missing roof panels or
holes in the roof.
2. Leakage between the battery top and the shed interface.
3. Integrity and cleanliness of the baffle plates and heat shields.
4. Effect of cross-drafts and heavy-partic
-------
8.2 ELECTRIC ARC FURNACE VENTILATION
8.2.1 Description
Canopy hoods are widely used to control secondary emissions from elec-
tric arc furnace (EAF) refining. The advantages of these hoods are that they
do not interface with the furnace; they provide ventilation during charging,
tapping, and slagging off; they do not affect furnace metallurgy; and their
maintenance costs are low. The disadvantages are that they require high air
volume; they are subject to cross-flow air currents that interface with fume
control; and the plume can interfere with crane operator's line of vision
during charging and tapping.
The need for high air volumes can be overcome by combining the canopy
with a partial enclosure, as illustrated in Figure 40. The three-sided
enclosure reduces the effect of cross-flow interference and directs the plume
toward the hood. As shown earlier in Figure 6 and explained in the discus-
sion regarding coke oven sheds, another approach is to use a baffled hood
that concentrates suction towards the heaviest area of emissions.
The ultimate expansion of this approach is total furnace enclosure which
was previously discussed in Section 7-
A complicating factor in EAF ventilation is the control system required
to combine fourth-hole evacuation and a canopy hood on the same fan system.
A balanced draft is difficult to achieve, and it must be established by trial
and error. For example, the canopy hood is often dampered off when the
furnace cover is in place so that suction will be concentrated on the fourth-
hole takeoff. In this configuration, the control system directs total system
suction to the canopy hood when the cover is off for charging and when the
furnace is tilted for tapping. In some cases, however, the charging emission
plume is still rising toward the hood while the cover is being quickly re-
placed. Thus, when the cover has been replaced and suction is taken off the
canopy hood, the rising plume is not captured. In such an arrangement, it
would be desirable either to replace the cover more slowly or to build an
electrical delay into the automatic damper control system.
107
-------
iPtftATIMG FLOOR
SLAGGING HOOD'
SLAG POT
Figure 40. Electric arc furnace utilizing partial enclosure.
Source: U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle
Park, North Carolina 27711
108
-------
Figure 41 shows the basic types of ventilation systems used on electric
arc furnaces. Figure 42 illustrates a popular approach in which the fourth-
hole direct evacuation is combined with a canopy hood. In some cases, a
third stage of capture is included by closing off the roof monitors in the
area of the furnace and connecting a take-off duct to the roof. This third
stage captures smoke that has escaped the canopy.
In the combined direct evacuation canopy system, baghouse temperature is
controlled by mixing the cool gas from the canopy with the hot gas from the
furnace. A cool-air emergency inlet damper is usually located prior to the
baghouse. If this damper remains open, it will cause a ventilation "short-
circuit" and rob the hood of suction.
As shown in Figure 43, two or more furnaces are typically combined on a
system, which can produce complex flow balancing requirements. Table 13
presents an example of the ideal flow balance on a two-furnace system.
Actually, flows through various duct sections must be monitored to confirm
their desirability. All elements of the system must be kept in good condi-
tion to maintain the ideal balance. Misalignment of the gap in the DSE duct
(see Figure 43) will result in the indraft of more combustion air at this
point and lower suction elsewhere. Figure 44 summarizes total flow rate
requirements for various EAF control approaches and indicates the rapid
increase in flow requirements with the move from direct evacuation to build-
ing evacuation. These data are based on averages of various systems and
should be used only as a general indicator of reasonable flow requirements.
Compromises were made in most plants where hoods had to be retrofitted.
The furnace and bay configuration, clearance for canopy hoods above the crane
gantry level, and ductwork layout became important factors because they could
mitigate the efficiency of the canopy hood. Failure to consider activities
that increase the fume generation rate, such as oxygen blowing and oxyfuel
burners, during the initial design of the hoods can cause the system to be
overloaded and capture efficiency to be reduced.
8.2.2 Inspection
The inspector should consider the following when inspecting EAF's:
109
-------
FOURTH HOLE
SIDE DRAFT
COMBINATION HOOD
CANOPY HOOD
Figure 41. Ventilation systems for electric arc furnaces.
-------
CLEANED GAS LOUVERS
CANOPY HOOD
DIRECT SHELL EVACUATION
ROOF TAP (WATER COOLED)
FAN
COMBUSTION AIR GAP
BAGHOUSE
\/\/\7
COLLECTED
DUST
Figure 42. Combined direct shell evacuation with canopy hood.
in
-------
POLLUTION CONTROL SYSTEM
BAGHOUSE
CANOPY
HOOD
SLAG POT
HOOD
Figure 43. Multiple pickup points vented to common control device.
112
-------
TABLE 13. EXAMPLE OF FLOW BALANCING OF MULTIPLE EVACUATION
SYSTEM ON ELECTRIC ARC FURNACE
Item
Furnace 1 charging
Furnace 1 slagging hood leakage
Furnace 1 tapping hood leakage
Furnace 1 direct shell leakage
Furnace 2 oxygen lancing
Furance 2 slagging (simultaneous with
oxygen blowing)
Furnace 2 canopy leakage
Furnace 2 tapping hood leakage
Total system capacity
Air flow, cfm
800,000
12,000
18,000
25,000
265,000
118,000
44,000
18,000
1,000,000
113
-------
1200
100 200 300 400 500
H = SUM OF HEAT SIZES IN SHOP, tons
600
Figure 44. Flow rate required for electric arc furnace control.
-------
1. The integrity of the canopy hoods, the side-draft hood, and all the
duct work leading to the cleaning unit. In the case of an enclosed
or sealed roof, any large holes or missing panels will reduce the
capture efficiency.
2. Additional thermal currents produced by ancillary operations such
as overflow of slag and/or metal on dumping ladles should be noted.
These operations increase the load on the evacuation system.
3. Furnaces and bay layout have an important bearing on the capture
efficiency of the hoods.
4. Multiple-furnace operations, additional pickup points, and the
overall flow balance under different modes of operation should be
observed.
5. Crosswinds inside the shop can severely disrupt the capture effi-
ciency of the hoods.
8.3 BLAST FURNACE CASTHOUSE CONTROL
8.3.1 Description
Although recently developed technology does not use ventilation for
control of casting emissions, numerous ventilation-based systems are in-
stalled or planned on blast furnace casthouses. Sources of fugitive emis-
sions in the casthouse are the tap hole, trough, skimmer, runners, hot metal
spouts, and slag spouts. These various sources require a multiple hooding
system or total building evacuation. Local control is favored because total
flow requirements are less; however, the design of many older casthouses
precludes their use. Table 14 lists the typical flow volume requirements for
a large blast furnace casthouse system.
These systems require flow balancing, proper design, and the other
factors discussed earlier. Also, the frequent removal and replacement of
close-fitting hoods or runner covers invite both damage and misplacement.
Because of the proximity of the hooding to the molten metal, frequent main-
tenance of the refractory linings and runner covers is required. Large blast
furnaces with multiple tapholes and casthouses require a carefully designed
flow balancing and damper system for switching from one casthouse to the
other. Such systems have been designed for full evacuation of one taphole at
a time. If damper settings are such that two tapholes are evacuated simulta-
neously, capture efficiency will be curtailed.
115
-------
TABLE 14. BLAST FURNACE CASTHOUSE TYPICAL VOLUME REQUIREMENTS
Location
Taphole
Skimmer
Iron runner
Hot metal spouts
Slag spouts
Total
Volume
requirements ,
cfm
70,000
35,000
70,000
75,000
50,000
300,000
Volume
requirements ,
m3/min
1978
989
1978
2119
1413
8477
116
-------
8.3.2 inspection
When inspecting blast furnace casthouse systems, the inspector should
consider the following:
1. The performance of the taphole hood during various stages of the
tap (i.e., immediately after taphole opening, during heavy metal
flow, and during slag and metal flow). The capture efficiency of
the hoods (covers) at the trough, skimmer, runners, and slag and
metal spouts also should be observed during various stages of the
cast and under different conditions, such as high and low metal
temperature and high and low slag basicity.
2. When there are multiple casthouses, the flow balance between the
casthouses should be observed. The flow is sometimes equally split
between the casthouses, which significantly reduces the capture
efficiency.
3. The furnace layout, casthouse layout, and ductwork location have an
important bearing on the efficiency of the evacuation system.
4. Cross-currents inside the casthouse mitigate the hood capture. It
is common practice to remove and replace the siding in the cast-
house, depending on the season. A system designed for a closed
structure is not likely to perform adequately if the structure is
opened.
8.4 CONTROL SYSTEMS ON BASIC OXYGEN FURNACES (BOF's)
8.4.1 Description
Primary capture systems on BOF's use a specially designed hood over the
mouth of the furnace. This usually operates effectively when the furnace is
in an upright position; however, when the furnace is tilted for charging,
tapping, or testing, emissions generally escape the hood (for the reasons
discussed earlier). Capture of these secondary emissions requires additional
hooding or enclosures. The total enclosure of BOF was previously discussed
in Section 7.
Water-cooled primary hoods are sometimes used to generate steam because
the gas temperatures in the hood are initially over 3000°F. These hoods are
exposed to the most severe operating conditions of all the process applica-
tions discussed—fluctuating high temperatures and the possibility of physi-
cal damage from cranes or ejections of material from the furnace. The hood
may be elevated several feet above the furnace to allow indraft of combustion
air (open), or they may be closely mated with the furnace to conserve the
117
-------
carbon monoxide inherent in the exhaust gas (closed). Figure 45 illustrates
these BOF hood arrangements.
Many different hood constructions have been tried to cope with the
severe operating conditions. The membrane or "tube-bar" type is usually
favored today. Table 15 presents the performance characteristics of various
hood constructions. Table 16 lists the various hood types that are used for
full combustion and limited combustion systems in the United States, Canada,
25
and Mexico.
Control of secondary emissions is achieved by local hoods or enclosures.
Local hoods are generally ineffective because they are too far from the
source, are too small to encompass the emissions, and because the evacuation
rate is usually insufficient. Figure 46 illustrates a simple external hood-
ing arrangement for charging emissions.
Figure 47 illustrates the Gaw damper approach, which enables the primary
hood to function as a secondary hood for charging emissions. The movable
damper increases hood suction by restricting the hood face area. Maintenance
of the damper is high because of the harsh environment. As in the case of
the local hood discussed previously, effective use of the Gaw damper requires
a slow pouring rate of charged metal to decrease fume generation and a mini-
mum tilt angle of the vessel to generally direct the fumes toward the hood.
8.4.2 Inspection
When inspecting BOF ventilation systems, the inspector should consider
the following:
1. Primary hood capture efficiency during operations such as slopping.
2. The tilt angle and hot metal pour rate durin- hot metal charging,
which greatly affects the capture efficiency.
3. The performance and capture of secondary controls.
4. Emissions from the lance hole or fume rollov (puffing). from the
primary hood, which indicate insufficient draft on the hood.
5. In closed hood systems, the snug fit of the hood skirt at the
furnace. (Indraft of air above the design level will overload the
fan and cause the variable-throat scrubber to open, which reduces
pressure drop.)
118
-------
/ RELIEF DOOR
QUENCHER
^- OAS COOLER
LANCE
WATER-COOLED FLANGES
SANO SEAL
MOVEABLC SKIRT
REMOVABLE HOOO
H- FIXED NOOO
ZZZT
VESSEL
CLOSED
/ RELIEF DOOR
QUENCHER
f— OAS COOLER
WATER-COOLED FLANGES
SANO SEAL
REMOVABLE HOOO
p- FIXED HOOO
N08KOTHERC
VESSEL
OPEN
Figure 45. BOF hood arrangements.
Courtesy: BOF Steelmaklng, Volume III. Published by American Institute of Minina
and Petroleum Engineers. y*
-------
TABLE 15. PERFORMANCE CHARACTERISTICS OF DIFFERENT EOF HOOD CONSTRUCTIONS
25
Initial cost
Ability to take
high tempera-
ture
Ability to take
temperature
change
Resistance to
slag buildup
Resistance to
scaling
Maintenance Cost
Refractory-
lined
Lowest
Poor
Poor
Poor
--
Very high
Water-
cooled
plate
panels
Low
Fair
Fair
Good
Poor
High
Formed
panels
Moderate
Good
Good
Good
Fair
Fair
Double
pass
High
Very good
Very good
Very good
Good
Low
Water wall
boiler
High
Very good
Very good
Fair
Good
Fair
Membrane
High
Very good
Very good
Good
Good
Fair
TABLE 16. BOF HOOD CONSTRUCTION DESIGNS IN USE
IN THE UNITED STATES, CANADA, AND MEXICO25
Construction
Tube
Panel/jacket
Membrane
Not known
Total
Full combustion
,jnber
10
41
30
4
85
Percent
12
48
35
5
100
Limited combustion
Number
3
4
13
0
20
Percent
15
20
65
0
100
Total
Number
13
45
43
4
105
Percent
12
43
41
4
100
120
-------
.FURNACE
CHARGING AISLE:
*-CRANE GIRDER
CANOPV HOOD
CHARGING
LADLE
Figure 46. Canopy hood concept for BOF charging emissions.
121
-------
CLOSURE PLATE CONCEPT
_FURNACE
CHARGING ABLE
CRANE
GIRDER
RETRACTED,
POSrTION
CHARGING
LADLE
Figure 47. Gaw damper (closure plate) use in BOF control
122
-------
8.5 BUILDING EVACUATION
The use of total building evacuation to control air pollution contamina-
tion or to improve the work area environment is normally a difficult and
costly approach.
The problem areas are:
1. Achieving the necessary number of air changes per hour (20 or more,
depending on the contaminants).
2. The ability to reach the specific work areas to supply these air
changes. (Some areas have dead air pockets, and either ducting or
forced ventilation has to be provided.)
3. The volume of air required. (A building 800 ft long, 80 ft wide,
and 40 ft high has a volume of 2,560,000 ft3. With 20 air changes
per hour, :850,000 scfm of air is required. If contaminants, open
doorways, heat emissions, and air currents caused by thermal proc-
esses or vehicle movement are added, the actual requirement for
effective ventilation of such a building would be 3 to 4 x 10
scfm.)
4. The existing building structure not being designed for total en-
closure. (The weight of siding, ducting, and wind loads may re-
quire strengthening of the building columns and roof trusses and
the addition of more purlins, struts, and bracing.)
5. The location of process equipment and the flow patterns of mate-
rials not being conducive to the proper collection of emissions.
(Lighting may also be a problem.)
6. Need for redundancy of the fans. (Clean-out mechanisms should be
built into the ducting that are readily accessible and repairable.
Ducting is long and huge.)
A newly designed building in which process equipment is located specifi-
cally for total building evacuation can greatly reduce these problems.
A variation of building evacuation that does not involve local hooding
26
and ventilation systems is the roof-mounted electrostatic precipitator (REP).
This system has not yet been applied in the United States, but it has been
?6
successfully used in Japan. The technique involves roof modification to
help channel the natural plume rise of BOF fugitive emissions into an REP,
which collects the process fugitive emissions. The system has no fan, and
the plates are generally cleaned by use of a water spray. The REP specifi-
cation provided by Sumitomo Heavy Industries for controlling two BOF's, each
with 300-ton capacity, offers a design efficiency of 91.5 percent.
123
-------
8.6 COPPER CONVERTERS
8.6.1 Description
Capture of emissions from Peirce-Smith type copper converters presents
some unique problems. The converter is essentially a horizontal cylinder
with a circular opening on top. This cylinder is rotated around its longi-
tudinal axis so the opening can face toward the ladle. This rotation makes
it impossible to achieve an absolutely tight seal between the hood and the
converter. Also, when the converter is rotated for charging and tapping, the
opening is no longer under the hood, and fumes tend to escape into the con-
verter building. The heat and collisions with the ladle suspended from the
crane tend to cause the hood to warp, and eventually it does not fit well.
Figure 48 shows a converter and the primary hood.
At copper plants, the emissions from all converters (usually three to
five units) are ducted together to a common particulate control device and
sometimes to a sulfuric acid plant. Dampers in the individual breeching
control the draft on any single converter. These dampers require routine
maintenance to ensure proper functioning.
Primary converter hoods have sliding gates on the front side as shown in
figure 48. These gates are really movable hood extensions that serve to
cover the converter opening and improve capture efficiency.
Some smelters have installed secondary hoods (Figure 49) or air curtain
systems to ca;ture fumes when rotation ca es the converter opening not to be
unde- the main hood. By providing a strong blast of air to blow fumes into a
suction hood, the air curtain provides an open area for locating the charging
and tapping ladle and the crane hooks and cables. Secr ary hoods suspended
above the converter opening and ladle have the same w sses found in other
processes. For example, because of their distance aoove the converter and
ladle, they are affected by thermal and cross-drafts and are not effective
unless face velocities of the hood are high.
124
-------
(END VIEW)
ro
in
(SIDE VIEW)
dj
(1) SHELL; (2) HOOD; (3) AIR BAFFLE; (4) CONVERTER MOUTH; (5) ROLLERS;
(6) TURNING RINGS; (7) AIR - SUPPLY DUCT;
Reprinted from the Kirk-Othmer Encyclopedia of Chemical Technology, 3rd edition,
Fig. 9, Page 836. Copyright© John Wiley & Sons, Inc. 1979.
Figure 48. Peirce-Smith converter.
-------
TO SECONDARY
HOODING
MAIN DUCT
Figure 49. Secondary converter hood configuration.
126
-------
8.6.2 Inspection
Converter hoods are inspected by making a visual assessment of their
physical condition and the emissions that escape during the various converter
cycles. The hood and duct system should not contain any openings that allow
air to be drawn in and reduce the suction at the converter. The hood lip
should come within about 6 inches of the converter opening during blowing.
The hood gate or slide should be extended when the converter is in the up-
right position to ensure that the converter opening is covered. During
converter rollout, the draft should be reduced so that a high-suction air
flow can be maintained on the converters in the blowing mode.
127
-------
REFERENCES
1. American Conference of Governmental Industrial Hygienists. Industrial
Ventilation. 16th Ed. 1980.
2. Hemeon, W. C. L. Plant and Process Ventilation. The Industrial Press,
New York. 1963.
3. Department of Health, Education and Welfare. Air Pollution Engineering
Manual. J. A. Danielson (Ed.) Public Health Service Publication 999-
AP-40, 1967.
4. Buffalo Forge Company. Fan Engineering. 8th Ed. Buffalo, New York.
1983.
5. Jablin, R., and D. W. Coy. Engineering Study of Roof-Mounted Electro-
static Precipitator for Fugitive Emission Control on Two Basic Oxygen
Furnaces of 300-ton capacity. In: Proceedings of Symposium on Iron and
Steel Pollution Abatement Technology for 1981. EPA 600/9-82-021,
December, 1982. pp. 120-138.
6. Coy, D. W., and L. E. Sparks. Roof-Mounted Electrostatic Precipita-
tors - Visible Emissions Evaluation and Computer Modeling Performance
Predictions. Presented at the 1983 Symposium on Iron & Steel Pollution
Abatement Technology.
7. Vajda, S. Blast Furnace Casthouse Emission Control Without Evacuation.
j r : Proceedings of Symposium on Iron and Steel Pollution Abatement
Technology. EPA 600/9-83-016, November 1982. pp. 232-240.
8. Dixon, T. E., and H. Nomine. Capital and Operating Costs of OBM/Q-BOP
Gas Cleaning Systems. Iron and Steel Engineer, March 1978. p. 37-
9. U.S. Department of Health Education and Welfare. Air Pollution Engi-
neering Manual. J. A. Danielson (Ed.). Public Health Service Publica-
tion 999-AP-40, 1967. Chapter 3.
10. Dalla Valle, J. M. Exhaust Hoods. The Industrial Press, New York.
1946.
11. Kriechelt, T. E. et al. Natural Ventilation in Hot Process Buildings.
Iron and Steel Engineer, December 1976.
12. Hemeon, W. C. L. Plant and Process Ventilation. TV>e Industrial Press,
New York. 1955.
128
-------
13. Davis, J. A. Unidirectional Flow Ventilation System. Presented at the
140th Annual AIME Meeting, New York, February 18, 1975.
14. American Conference of Governmental Industrial Hygienists. Industrial
Ventilation, A Manual of Recommended Practice. 17th Ed. 1982.
15. U.S. Environmental Protection Agency. Air Pollution Engineering Manual.
2nd Ed. AP40, May 1973.
16. Pollak, R. Selecting Fans and Blowers. Chemical Engineering, January
22, 1973.
17. Fan Engineering. 8th Ed. Published by Buffalo Forge Company, New York.
1983.
18. Nicola, A. G. Best Available Control Technology (BACT) for Electric
Furnace Shop Emission Control. 36th Electric Furnace Conference
Proceedings, Vol. 36, Toronto December 1978, Sponsored by The Electric
Furnace Division of I&S Society of American Institute of Mining,
Metallurgical and Petroleum Engineers.
19. Brand, P. G. A. Current Trends in Electric Furnace Emission Control.
Iron and Steel Engineer, February 1981. p. 59.
20. Henninger, J. L. et. al. Closing In On Arc Furnace Emissions at Carpenter
Technology. Iron & Steel Engineer. March 1984.
21. Canefield, G. Jr. Pollution Control Using a Furnace Enclosure. 36th
Electric Furnace Conference Proceedings, Vol. 36, Toronto December 1978,
Sponsored by The Electric Furnace Division of I&S Society of American
Institute of Mining, Metallurgical and Petroleum Engineers.
22. U.S. Environmental Protection Agency. Revised Standards for Basic
Oxygen Process Furnaces - Background Information for Proposed Standards.
Preliminary Draft. 1980.
23. Roe, et al. United States Patent No. 3,844,901. October 29, 1974.
24. Vajda, S. Blast Furnace Casthouse Emission Control Without Evacuation.
Iron and Steel Engineer, June 1983. p. 29.
25. American Institute of Mining, Metallurgical and Petroleum Engineers.
BOF Steelmaking. Vol. III. 345 E. 47th Street, New York.
26. Jablin, R., and D. W. Coy. Engineering Study of Roof-Mounted Electro-
static Precipitator (REP) for Fugitive Emission Control on Two Basic
Oxygen Furnaces of 300-ton Capacity. In: Proceedings of Symposium on
Iron and Steel Pollution Abatement Technology for 1981. EPA 600/9-82-
021, December 1982. pp. 120-138.
27 US Environmental Protection Agency. Control of Copper Smelter Fugi-
tive Emissions. EPA-600/2-80-079, May 1980.
129
-------
APPENDIX A
BIBLIOGRAPHY
Adams, R. VI. 1976. Waste Gas Handling. In: BOF Steelmaking, Volume 3,
Design. Iron and Steel Society of AIME.
American Iron and Steel Institute. 1965. Steel Mill Ventilation.
Ban, E. 1980. Innovations and Improvements in the Ore Sintering Process for
Air Pollution Control. In: Proceedings of the First Symposium on Iron and
Steel Pollution Abatement Technology, Chicago, Illinois, October 30-November
1, 1979. EPA-600/9-80-012.
Baturin, V. V- 1972. Fundamentals of Industrial Ventilation. 3rd Edition,
Pergamon Press.
Bender, M. 1979. Fume Hoods, Open Canopy Type-Their Ability to Capture
Pollutants in Various Environments. American Industrial Hygiene Association
Journal, pp 119-127.
Bender, M., and W. D. Baines. 1976. Operation of an Open Canopy Fume Hood
in a Crossflow. ASME Ser., 97(2):242-243.
Blast Furnace Cast House Emission Control Technology Development in the USSR.
Translated by SCITRAN for U.S. Environmental Protection Agency.
Blessing, K. E. 1963. Electric Furnace Fume Control. Chemical Engineering
Progress. 59:60-64.
Brand, P. G. A. 1981. Current Trends in Electric Furnace Emission Control.
Iron and Steel Engineer, 58:59.
Brough, J. R. 1964. Air Pollution Control of An Electric Furnace Steelmak-
ing Shop. Journal of the Air Pollution Control Assn. 22:170.
Brown, C. M. 1980. Control of Secondary Emissions from Basic Oxygen Steel-
making Furnaces. Iron and Steel Engineer. 57:39.
Buffalo Forge Co. 1970. Fan Engineering.
Burroughs, J. D. et al. 1976. Application of Enclosed Coke Pushing Emission
Controls at Granite City Steel. In: Proceedings of 69th Annual Meeting of
Air Pollution Control Association, Portland, Oregon, June 27-July 1, 1976.
Calderon, A. 1977. Controlling Emissions from Coke Ovens. Iron and Steel
Engineer! 54(8):42.
A-l
-------
Campbell, W. VI. 1962. Development of an Electric Furnace Dust-Control
System. Journal of the Air Pollution Control Assn. 12:576.
Canfield, G., Jr. 1978. Pollution Control Using a Furnace Enclosure. In:
Proceedings of the Electric Arc Furnace Conference, Vol. 36. Toronto Meet-
ing. December 5-8, 1978.
Cheever, C. L., et al. 1972. Proceedings (two volumes) of 12th Air Cleaning
Conference, Oak Ridge, Tennessee.
Cheremisinoff, P. N. and N. P. 1976. Calculating Air Flow Requirements for
Fume Exhaust Hoods. Plant Engineering, 30(4):111-114.
Cheremisinoff, P. N. and N. P. 1976. Calculating Air Volume Requirements
for Fume Exhaust Hoods—Nonenclosure Types. Plant Engineering, 30(6):143-
144.
Clarke, J. H. Practical Industrial Ventilation Design. A collection of
article reprints, Heating, Piping, and Air Conditioning Journal.
Coleman, R. T., Jr., and R. Vandervort. 1980. Demonstration of Fugitive
Emission Controls at a Secondary Lead Smelter. In: Proceedings of World
Symposium of Metallurgy and Environmental Control, Las Vegas, February 24-28,
1980.
Conners, A. 1980. Developments in Coke-Oven Emission Control. Iron & Steel
Engineer, 57(6):33.
Coy, D. W., and R. Jablin. 1980. Review of Foreign Air Pollution Control
Technology for BOF Fugitive Emissions. In: Proceedings of the First Sympo-
sium on Iron and Steel Pollution Abatement Technology, Chicago, Illinois,
October 30-November 1, 1979. EPA-600/9-80-012.
Current, P. 1980. Sinter Plant Windbox Gas Recirculation and Gravel Bed
Filtration Demonstration. In: Proceedings of the First Symposium on Iron
and Steel Pollution Abatement Technology, Chicago, Illinois, October 30-
November 1, 1979. EPA-600/9-80-012.
Drivas, P. J., et al. 1972. Experimental Characterization of Ventilation
Systems in Buildings. Environmental Science Technology. 6(7):609-614.
Eisenbarth, M. J. 1981. Fume Extraction Hoods in the Iron & Steel Industry.
In: A Specialty Conference on Air Pollution Control in the Iron and Steel
Industry, April 21-23, 1981.
Emswiler, J. E. 1926. Neutral Zone in Ventilation. ASHRAE Transactions,
Vol. 32.
Escudier, and J. Maxworthy. 1973. On the Motion of Turbulent Thermals.
Fluid Mechanics, V- 61 - Pt. 3. pp 541-552.
A-2
-------
Field, A. A. 1977. New Concepts in Industrial Air Replacement. Heating,
Piping, and Air Conditioning, 49(3):84, 86, 88, 90.
Fields, S. F., et al. 1982. Modeling of Hood Control of Blast Furnace
Casting Emissions. In: Proceedings of the Symposium on Iron and Steel
Pollution Abatement Technology for 1981, Chicago, Illinois, 10/6-10/8/81.
EPA-600/9-82-021.
Flannigan, L. J., et al. 1974. Development of Design Criteria for Exhaust
Systems for Open Surface Tanks. Prepared for NIOSH, Cincinnati, Ohio.
NIOSH-75/108.
Flux, J. H. 1974. Containment of Melting Shop Roof Emissions in Electric
Arc Furnace Practice. Ironmaking and Steelmaking (Quarterly). No. 3, pp
121-133.
Forestier, G., and X. Tinchant. 1981. Cooling of Oxygen Converter Gas,
Efficiency of the Exhaust Hood, and the Stack of the Gas collection Equip-
ment. Tech. Rev. Metal!., 78(12):937-948.
Franza, M. E. 1982. Controlling Fugitive VOC Emissions From the Metal
Finishing Industry. Met. Finish, 80(12):39-45.
Fuller, H. 1981. The Alliance/Mitsubishi Pushing Emission Control System at
Ironton Coke Corp. Iron and Steel Engineer, 58(5):29.
Goldman, J. Technical Support Document for Control of Coke Oven Pushing
Emissions. Research Triangle Institute. EPA Contract 68-01-4141, Task 33.
Goodfellow, H. D. 1978. Fume System Design for Sidbec's Pellet Melting
Furnaces. In: Proceedings of the Electric Arc Furnace Conference, Vol. 36.
Toronto Meeting, December 5-8, 1978.
Great Lakes Carbon Shuns Naysayers, Uses Hood to Collect Smoke from St. Louis
Coke Battery. 33 Magazine.
Haddrill, D. M. 1979. Control of Welding Fumes in Fabrication Shops. Metal
Construction and British Welding J., ll(2):84-89.
Hanks, D. J. 1979. M/E Update: Fans and Blowers. Specif. Eng., 41(1):
129-134.
Hartung, Kuhn and Co. Pushing Emission Control System; Minister Stein.
Dusseldorf, West Germany.
Hayashi, T., et al. 1976. Land Based, Mobile Haroley "Trar-L-Vent;1 Exhaust
System for Coke-Side Pushing Emission Control. Hawley Manufacturing Corp.
Indianapolis, Indiana. On the Elimination of Welding Fumes by Richard Hodd.
Trans. Japan Weld. Soc., 7(l):18-26.
A-3
-------
Helfritch, D. 1978. Arc Furnace Fume Control by an Electrostatic Fabric
Filter. In: Proceedings of the Electric Arc Furnace Conference, Vol. 36.
Toronto Meeting. December 5-8, 1978.
Hemeon, W. C. L. 1963. Plant and Process Ventilation, The Industrial Press,
New York, New York, 2nd Edition.
Hoult, D. P. 1972. Turbulent Plume in a Laminar Cross Flow. Atmospheric
Environment, Vol. 6, pp 513-531.
Iron and Steel Engineer. 1978. 55(3):83. Hooded Quench Car System Controls
Coke Pushing Emissions.
Iron and Steel Engineer. 1980. American Air Filter. Local Capture System
Controls. Coke Pushing Emissions. 57:69.
Irvine, H. B. 1980. Waste Heat Recovery From a Dust Collection System. In:
Proceedings of 38th Electrical Furnace Conference, Pittsburgh, December 9-12,
1980.
Jacobs, C. 0. 1979. Exhaust System for Electric Arc Welding Instruction.
Presented at the Winter Meeting of ASAE, New Orleans, December 11-14, 1979.
Kaercher, L. T. 1974. Air Pollution Control for an Electric Furnace Melt
Shop. Iron and Steel Engineer Year Book, pp 216.
Karnaukh, N. N., et al. 1973. Trapping of Gases From Ingot Molds in Pouring
of Steel With Exothermic Mixtures. Metallugist (USSR), 17(3-4):189-190.
.\osc~ev, L. V., et al. 1978. Complex Systems of Technological Ventilation
Apparatus for Collecting and Purifying Gases in Arc Furnaces. Metallurg.,
21(5-6):26-29.
Kreichelt, T. W., and T. G. Keller. 1972. Roof Monitor Emissions; Test
Methodology. Journal APCA, Vol. 22, No. 8.
Kreichelt, T. E., G. R. Kern, and F. G. Higgins, Jr. 1976. Natural Ventila-
tion in Hot Process Buildings in the Steel Industry. Iron and Steel Engi-
neer.
Krueqer, D. 1964. Gas Cleaning in Iron and Steel Works. Blast Furnace and
Steel Plant. 52:503.
Lawrie, W. G. 1970. Dust Collection in Industry. Filtr. Separ., 7(6):695-
699.
Lusty, V. 1980. Collecting Emissions From Electric Induction Furnaces.
Foundry Trade Journal, 148(3185):550, 553, 558, 562.
Malchaire, J. B. 1981. Design of Industrial Exhaust Systems Using Program-
mable Calculator. Annuals Occup. Hyg., 24(2):217-224.
A-4
-------
Marchisio, I. C. 1981. Fiat/Teksid Goes "Snuff Box." Iron and Steelmaker.
8(6):36+.
Marchland, D. 1976. Possible Improvements to Dust Collection in Electric
Steelplants, etc. Ironmaking and Steelmaking. No. 4, pp 221-229.
Marchland, D. 1981. Waste Gas Collection on Electric Arc Furnaces: BFI
Research & Development During the Last 20 Years. In: A Speciality Con-
ference on Air Pollution Control in the Iron and Steel Industry, April 21-23,
1981.
Martin J. R., A. D. Robertson, and A. T. Sheridan. 1974. Modeling Studies
of Roof Extraction Systems for Arc Furnace Melting Shops. Presented at
Conference: Quality of the Environment and the Iron and Steel Industry.
Luxembourg, September 24-26, 1974.
Masany, T. J. 1982. Blast Furnace Casthouse Control Technology - Fall 1982
Update. In: Proceedings of the Symposium on Iron and Steel Pollution
Abatement Technology for 1981, Chicago, Illinois, 10/6-10/8/81. EPA-600/9-
82-021.
McCluskey, E. J. 1976. Design Engineering of the OH Gas Cleaning System at
Inland's No. 2 BOF Shop. Iron and Steel Engineer, 53(12):53.
McCrillis, R. C. 1978. Basic Iron Blast Furnace Casthouse Emissions and
Emission control. Paper presented at the 71st annual meeting of the Air
Pollution Control Association at Houston, June 25-30, 1978.
McDermoth, H. J. 1976. Handbook of Ventilation for Contaminant Control.
Ann Arbor Science Publishers, Ann Arbor, Michigan.
Mertz, J. W. and R. R. Pfahler. 1975. How to Troubleshoot Large Industrial
Fans. Hydrocarbon Process, 54(6):57-62.
Molnar, J. 1978. Graphical Method for Determining Air Requirements of
Exhaust Hoods. Plant Engineering, 32(4):132-133.
Mortimer, V. D., Jr., et al. 1982. In-Depth Survey Report of Valley Chrome
Platers. Bay City, Michigan, April 27-May 1, 1981. Prepared for National
Institute for Occupational Safety and Health, Cincinnati, Ohio.
Muermann H 1978. Ventilation and Air Conditioning of Modern Industrial
Buildings. *Wasser, Luft Und Betrieb, 22(4):126-128.
Natalizio, A., and C. Twiffe-Molecey. Ventilation of Mill Buildings - New
Directions. Iron and Steel Engineer. 57(7):51-56.
F D et al. 1976. Plant Layout and Material Handling. In: BOF
, Volume 3, Design. Iron and Steel Society of AIME.
A-5
-------
Nicola, A. G. 1979. Blast Furnace Cast House Emission Control. Iron and
Steel Engineer. 56:33. 1979.
Nijhawan, P. 1981. Ventilation: Looking at Close Capture Ventilation
Control. 33 Metal Producing. 19:FE54.
Ochs, J. T. 1975. Venting and Extraction in Electroplating Plants. Wasser,
Luft and Betr. (Germany), 19(7) :400-402.
Odasso, A. 1980. Air Pollution Abatement at the Brackenridge Works of
Allegheny Ludlum Steel Corp. Iron and Steel Engineer, 57(6):51.
Patton, S. 1973. Hooded Coke Quenching System for Air Quality Control.
Iron and Steel Engineer, 50(8):37.
Pfeiffer, W. 1982. Exhaust Air Quantities of Open Type Exhaust Hoods, Staub
Reinhart Luft, 42(8) : 303-308.
Puri , R. 1974. System Design for Particulate Control in Woodworking Plants.
Heating, Piping, and Air Conditioning, 46(9):54-58.
Randall, W. C., and E. W. Conover. 1931. Predetermining the Aeration of
Industrial Buildings. ASHVE Transactions (American Society of Heating and
Ventilating Engineers), Vol. 37.
Ricficr-, G. 1975. Modern Ventilation Plant. The Current Technical Position
in the Light of New Legislation on Protection of the Environment Ind.-Anz.
(Germany), 97(75): 1644-1648.
ii'"';, H. P. 1977. Energy Considerations in Paper Machine Hood and Air
Systems "esign. Presented at TAPPI Engineering Conference, Atlanta,
September .9-22, 1977. pp 149-153.
Rowe, A. D. 1970. Waste Gas Cleaning Systems for Large Capacity Basic
Oxygen Furnaces. Iron and Steel Engineer, 47(1):74.
Rudolph, H. 1977. Engineering Criteria for a Hooded Quench Car System.
Iron r -' Steel Engineer, 54(3):27.
H., and S. Sayer. 1977. Engineering Criteria for a Hooded Quench
Car System. Iron and Steel Engineering, 54(3):27-32.
Sensenbaugh, J. D., and G. G. Persons. 1978. Electric Furnace Melt Shop
Fume Control System at the Timken Co. Iron and Steel Engineering 55(10) :53-
55.
Shaughnessy, J. 1981. A Review of Shed and Gas Cleaning Systems for Con-
trolling Coke Pushing Emissions from Coke Plants. In: Proceedings of the
Symposium on Iron and Steel Pollution Abatement Technology for 1981, Chicago,
Illinois, October 6-8, 1981. EPA-600/9-82-021.
A-6
-------
Shcherbakov, K. L. 1976. Floor Layouts and Air Conditioning in Foundries
for Mass and Large-Batch Production. Liteinde Proizyod (USSR), No. 10:34-5.
Siege!, M. E., and A. Shacter. 1978. Engineering Design Handbook for Air
Cleaning For Chemical Demilitarization. ARCSL-CR-78022; AD-E410 041.
Spawn, D. 1981. Blast Furnace Cast House Control Technology and Recent
Emissions Test Data for 1980-81. In: A Specialty Conference on Air Pollu-
tion Control in the Iron and Steel Industry, April 21-23, 1981.
Spawn, D. 1981. Field Evaluation of Fugitive Emissions from BOF Steelmaking
Shops. In: Proceedings of the Symposium on Iron and Steel Abatement Tech-
nology for 1980, Philadelphia, PA. November 18-20, 1980. EPA-600/9-81-017.
Spawn, D. 1981. Status of Casthouse Control Technology in the United
States, Canada, and West Germany in 1980. In: Proceedings of the Symposium
on Iron and Steel Abatement Technology for 1980, Philadelphia, PA. November
18-20, 1980. EPA-600/9-81-017.
Squires, B. J. 1971. Electric Arc Furnace Fume Control and Gas Cleaning.
Filter Separ., 7(4):447-450, 453-455.
Steiner, J., and L. F. Kertcher. 1980. Fugitive Particulate Emission
Factors for BOP Operations. In: Proceedings of the First Symposium on Iron
and Steel Pollution Abatement Technology. Chicago, Illinois, October 30-
November 1, 1979. EPA-600/9-80-012.
Thaxton, L. A. 1971. Technology for Pollution Control. Exhaust Systems.
Hood Close to Source Is Key Consideration. Metal Progr., 100(7):56-58.
Tsuji, K., T. Hayaski, and M. Shibatu. 1979. Required Exhaust Volume for
Local Lateral Hoods. Bulletin, University of Osaka Prefect Ser. A., 27(1):
109-124.
Turner, H. 1981. Performance of BOF Emission Control Systems. In: Pro-
ceedings of the Symposium on Iron and Steel Pollution Abatement Technology
for 1981, Chicago, Illinois. October 6-8, 1981.
Under EPA Duress, U.S. Steel Finds New Ways to Operate its Veteran Homestead
OH Shop. 1979. 33 Metal Producing, pp 49-51.
U.S. EPA. 1974. Standards Support Document: An Investigation of the Best
Systems of Emission Reduction for Electric Arc Furnaces in the Steel Indus-
try. Draft.
U.S. EPA 1977- Blast Furnace Casthouse Control Technology Assessment. PB
276999.
U.S EPA 1980. Revised Standards for Basic Oxygen Process Furnaces -
Background Information for Proposed Standards. Preliminary Draft.
A-7
-------
U.S. EPA. Undated. An Investigation of the Best Systems of Emission Reduc-
tion for Pushing Operation of Byproduct Coke Ovens Standard Support and
Environmental Impact Statement. Office of Air Quality Planning and Stan-
dards, Research Triangle Park, North Carolina.
Vatavuk, W. M., and R. B. Neveril. 1980. Estimating the Size and Cost of
Pollutant Capture Hoods—3. Chemical Engineering, 87(24):111-112, 114-115.
Vaughan, J. 1975. A Residuals Management Model of the Iron and Steel
Industry: Part I. Ph.D. Dissertation, Georgetown University.
Vinson, R. P., et al. 1980. SF6 Tracer Gas TEsts of Bagging-Machine Hood
Enclosures. Bureau of Mines, Pittsburgh Research Center BUMINES-R1-8527.
Voldshina, R. K., and A. I. Sukhareva. 1972. Means of Reducing the Dust
Concentration of the Air in Glass Plant Batching Departments. Glass and
Coram. (USA), 29(9-10):585-586.
Wheeler, F. M., and A. G. W. L. Lament. 1979. Current Trends in Electric
Meltshop Design. Iron and Steelmaker, pp 36-43.
Whike, A. S. 1979. Energy Conservation and Pollution Control—Two Advant-
ages of Coil Coating. In: Proceedings of the 14th International Society of
Energy Conservation Engineers Conference, Boston, August 5-10, 1979.
Williams, A. E. 1982. New Developments in Coke-Oven Machinery Technology.
Iron and Steel Engineer, 59(7):25+.
Zaryankin, A. E., and V. P. Zhilinskii. 1975. Analysis of Design of Soviet
Exhaust Hoods of Steam Turbines and the Possibility of Lowering Their Re-
sistance. Therm. Eng. (GB), 22(3):58-62.
A-8
-------
TECHNICAL REPORT DATA
(rlcete read Instruction! on the reverse before eompletingl
.REPORT NO.
3. RECIPIENT'S ACCESSION NO.
4 TITLE AND SUBTITLE
Performance Evaluation Guide for Large Flow
Ventilation Systems
5. REPORT DATE
May 1984
6. PERFORMING ORGANIZATION CODE
?.AUTHOR(S)
W. F. Kemner, R. W. Gerstle, and Y. M. Shah
«. PERFORMING ORGANIZATION REPORT NO
3760-1-119
9 PERFORMING ORGANIZATION NAME AND ADDRESS
PEDCo Environmental, Inc.
11499 Chester Rd.
Cincinnati, OH 45246
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-01-6310
Work Assignment 119
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Control & Process Engineering
Stationary Source Compliance Division
Washington, D.C.
13. TYPE OF REPORT AND PERIOD COVERED
Fi nal
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
Project Officer for SSCD: Dwight Hlustick
16. ABSTRACT
Air pollution control systems in the primary metals industry, particularly the
steel and copper segments, rely on large capture and ventilation systems with flow
rates commonly in the range of 50,000 to 1,000,000 acfm and greater. These systems
are used primarily to control process fugitive emissions from various furnaces and
for building evacuation.
Because these systems are in integral feature of the compliance programs of the
industries involved, this manual was initiated to accomplish the following:
* To provide design and operation and maintenance guidelines to state and
local agency personnel who evaluate the performance of these systems.
* To provide a comprehensive treatment of the existing literature with
regard to technical and specific aspects of typical designs.
* To provide an easy-to-read technical manual on design and operation for the
use of inspectors.
Inasmuch as ventilation systems are highly complex from a design standpoint and
experience plays a major role in most designs, this manual should be considered an
introductory primer rather than a detailed design manual.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN ENDED TERMS
COSATi Field/Croup
Ventilation
Industrial Engineering
Fans, Ducts, Hoods
Steel Industry
Copper Smelting
13A
13H
B. DISTRIBUTION STATEMENT
Unlink t_J
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
146 p.
20 SECURITY CLASS (Thu page)
Unclassified
22. PRICE
A07
F«rm 2220-1 («•»• 4-77) PHCVIOUS EDITION is OBSOLETE
------- |