United States Office of Air Quality February 1984
Environmental Protection Planning and Standards EPA-340/1-84-002
Agency Washington DC 20460
Stationary Source Compliance Division
£EPA Fabric Filter
Inspection and
Evaluation Manual
-------
EPA 340/1-84-002
Fabric Filter Inspection
And Evaluation Manual
By
Douglas R. Roeck
Richard Dennis
GCA CORPORATION
GCA/TECHNOLOGY DIVISION
Bedford, Massachusetts 01730
Contract No. 68-01 -6316
Technical Service Area No. 1
Assignment No. 72
EPA Project Officer: John Busik
EPA Work Assignment Officer: Kirk Foster
U.S. ENVIRONMENTAL PROTECTION AGENCY
Stationary Source Compliance Division
Office of Air Quality Standards and Planning
Washington. D.C. 20460
February 1984
-------
DISCLAIMER
This Final Report was furnished to the Environmental Protection Agency by
Che GCA Corporation, GCA/Technology Division, Bedford, Massachusetts 01730, in
fulfillment of Contract No. 68-01-6316, Task Order No. 66 and Contract No.
68-01-6316, Assignment No. 72., Technical Service Area No. 1. The opinions,
findings, and conclusions expressed are those of the authors and not
necessarily those of the Environmental Protection Agency or the cooperating
agencies. Mention of company or product names is not to be considered as an
endorsement by the Environmental Protection Agency.
-------
CONTENTS
Figures , vii
Tables x
1. Introduction 1
1.1 Purpose of Manual 1
1.2 Manual Organization 1
1.2.1 Special Role of Chapter II 1
1.2.2 The Working Sections of the Manual 2
2. How Fabric Filters Work 4
2.1 Background 4
2.2 Mechanics of Dust Filtration 5
2.2.1 Basic Concepts 5
2.2.2 Particle Capture by Direct Interception 6
2.2.3 Particle Capture by Inertial Impaction 8
2.2.4 Particle Capture by Brownian Diffusion 10
2.2.5 Particle Capture by Electrostatic Mechanisms 11
2.2.6 Particle Collection by Direct Sieving 11
2.2.7 Particle Capture by Gravitational Forces 12
2.2.8 Physical Processes Enhancing Particle Capture .... 13
2.2.8.1 Agglomeration 13
2.2.8.2 Condensation and Evaporation 14
2.3 Fabric Filtration versus Single Fiber Collection 15
2.3.1 Single Fiber Collection Efficiency and Amount of
Fiber 15
2.3.2 Typical Fabric Filtration Process 16
2.3.2.1 Initial Fiber Filtration Phase 16
2.3.2.2 Final Cake Filtration Phase 19
2.4 Gas Properties 21
2.4.1 Temperature 24
2.4.2 Viscosity 24
2.4.2.1 Effect on Particle Collection 24
2.4.2.2 Effect on Pressure Loss Through Dust Laden
Fabrics 24
2.4.3 Velocity Effects 25
2.4.3.1 Filtration Velocity 25
2.4.3.2 Duct, Breeching, and Hopper Velocities ... 25
2.4.4 Moisture, Condensables, and Corrosive Components ... 26
2.5 Particle Properties 26
2.5.1 Particle Mass Concentration 27
2.5.2 Particle Size Distribution 28
111
-------
CONTENTS (continued)
2.6
2.7
2.5.2.1 Particle Sizing by Cascade Impactor . . .
2.5.2.2 Particle Density ............
2.5.2.3 Particle Shape .............
2.5.3 Significance of Characterizing Particle Diameters
2.5.3.1 Average or Mean Diameters ........
2.5.3.2 Median Diameters ............
2.5.4 Characterizing a Distribution of Particle Sizes
2.5.4.1 Graphical Representation ........
Fabric Properties and Filtration Concepts ........
2.6.1 Fabric Selection Criteria ............
2.6.2 Woven Fabrics ..................
2.6.2.1 Construction Features ..........
2.6.2.2 Fabric Performance versus Construction
2.6.2.3 Fabric Design Limitations ........
2.6.3 Nonwoven Fabrics .................
2.6.3.1 Construction Features - Felted Fabrics
2.6.3.2 Advantages of Felted Media .......
2.6.3.3 Needled Felts ..............
2.6.4 Fabric Selection - General Comments .......
2.6.5 Basic Filtration Concepts - Woven Fabrics . . . .
2.6.5.1 Total Pressure Loss ...........
2.6.5.2 Graphical Estimation of Sg and IC? . . . .
2.6.5.3 Pressure Loss Characteristics and Filter
Performance ..............
2.6.5.4 Effect of Partial Cleaning on Filter
Drag-Fabric Loading Relationships . . . .
2.6.5.5 Nonlinearities in Drag-Fabric Loading
Curves ................
2.6.5.6 Effect of Velocity and Fabric Loading
on
2.6.5.7 Theoretical Estimation of K2
Fabric Cleaning - Woven and Nonwoven Fabrics
2.7.1 Mechanical Shaking
2.7.2 Bag Collapse With Reverse Flow - Woven Fabrics
2.7.3 Pulse Jet Cleaning - Felted Media
2.7.3.1 Qualitative Description of Pulse Jet
Filtration
2.7.3.2 Estimation of Pressure Loss - Theory
2.7.3.3 Equilibrium Pressure Losses
2.7.4 Dust Removal versus Cleaning Method
2.7.4.1
2.7.4.2
2.7.4.3
Fractional Cleaning, ac, with Mechanical
Shaking
Fractional Cleaning with Bag Collapse and
Reverse Flow
Effective Filter Cleaning by Pulse Jet
Air
28
30
30
30
30
31
31
31
37
37
37
37
39
39
41
41
42
42
42
45
45
46
46
50
50
54
55
56
56
60
60
67
69
70
70
70
73
73
IV
-------
CONTENTS (continued)
2.8 Filter Performance 78
2.8.1 Filter Drag - Woven Fabrics Cleaned by Mechanical
Shaking or Bag Collapse with Reverse Flow 78
2.8.2 Dust Penetration - Woven Fabrics 79
2.8.2.1 Mass Emissions Rates 79
2.8.2.2 Effect of Pinhole Leaks on Effluent Mass
Concentration and Size Properties 80
2.8.2.3 Effect of Filtration Velocity on
Particulate Emissions 80
2.8.2.4 Effect of Inlet Dust Concentration and
Fabric Loading on Particulate Emissions . . 82
2.8.2.5 Estimation of Overall Dust Penetration ... 82
2.8.3 Dust Penetration - Pulse Jet Cleaned, Nonwoven
Fabrics 84
2.8.3.1 Mass Emissions 84
2.8.3.2 Factors Affecting Effluent Size
Properties 84
2.8.4 Field Performance 84
Typical Baghouse Problems/Corrective Actions 89
3.1 Introduction 89
3.2 Ancillary Equipment 89
3.2.1 Fans and Blowers 89
3.2.2 Dust Handling Systems 91
3.2.3 Monitoring Equipment 91
3.3 Baghouse Components 91
3.3.1 Fabric 91
3.3.2 Cleaning System 91
3.3.3 Basic Structure 92
3.4 Process Conditions 92
3.5 Fabric Filter User Input 92
3.6 Problem Rectification 98
3.7 References 98
Baghouse Inspections, Types and Procedures 107
4.1 Introduction 107
4.2 Inspections Performed by Plant Personnel 107
4.2.1 Background 107
4.2.2 Prestart-Up 107
4.2.3 Baghouse Start-up—Standard Procedure 109
4.2.3.1 General Start-up Procedures for a Pulse Jet
Baghouse with Spray Drier or Process
Equipment 109
4.2.3.2 Specific Start-up Procedures for a Reverse
Air Baghouse Used for the Collection of
Coal Fly Ash 110
4.2.4 Troubleshooting Ill
4.2.5 General Preventive Maintenance, Including Bag
Replacement and Cleaning 112
-------
CONTENTS
4.3
4.4
4.5
References
4.2.5.1 Care of Filter Media
4.2.5.2 Bag Replacement
4.2.5.3 Valves and Dampers
4.2.5.4 Motors, Fans and Belts
4.2.5.5 Ductwork
4.2.5.6 Auxiliary Systems
4.2.5.7 Maintenance Checklist
Inspections Performed by Air Pollution Agencies .
4.3.1 Types of Inspections
4.3.2 Procedures
4.3.3 Inspection Sequence
4.3.3.1 Process Equipment
4.3.3.2 Dust Capture and Transport System
4.3.3.3 Fabric Filter System
4.3.3.4 Fan
4.3.3.5 Dust Disposal System
Inspection Report Format
Example Inspection Report
112
113
114
114
115
115
115
116
116
122
123
123
124
124
125
125
126
126
142
Appendices
A. Aerosol and Sampling Terminology
B. Particle Size Measurement
Baghouse and Fabric Terminology
Supplementary Data Baghouses and Accessories
U.S. Baghouse Manufacturers
C.
D.
E.
F. Fabric Filter Model (Capsule Description)
154
163
168
181
187
189
VI
-------
FIGURES
Number
1 Streamlines and particle trajectories approaching filter
element 7
2 Impaction efficiencies for cylinders, spheres, ribbons, and
disks 9
3 Schematic, dust accumulation on woven glass fabrics 17
4 Variations in pinhole leaks due to fiber presence and pore
size 23
5 Mass distribution for typical coal fly ash, based on cascade
impactor sampling 29
6 Particulate material graphed as normal and logarithmic normal
distributions 32
7 Particle size distributions based upon size, surface area,
and mass (volume) 34
8 Graphical estimation of characteristic particle diameters
(Dg, Dv, Dgmd, and Dnmd) and geometric standard
deviation toO 35
o
9 Continuous multifilament yarn (a) and bulked yarn (b), glass
fabrics 40
10 Typical filter drag versus fabric loading curve 48
11 Cleaned bag with illumination from inside by fluorescent
tube 51
12 Typical drag versus loading curves for filters with different
degrees of cleaning and a maximum allowable level for terminal
drag, S.p, and terminal fabric loading, WT 52
13 Approximate drag versus fabric loading, coal fly ash filtration
at 0.61 m/min filtration velocity 53
VII
-------
FIGURES (continued)
Number Page
14 Typical shaking mechanism 57
15 Typical bag attachment system for a shaker baghouse 58
16 Cutaway view of a shaker baghouse 59
17 Reverse air baghouse 61
18 Reverse-air baghouse (a), butterfly reverse air valves (b),
bag suspension system (c), and tube sheet connection (d) ... 62
19 Typical bag suspension systems for reverse-air baghouses .... 64
20 Cutaway view of an Alpha Series Pulse-Jet Baghouse 65
21 Schematic of pulse-jet baghouse and compressed air header ... 66
22 Schematic of Mikro-Pulsaire dry dust collector and operating
components 68
23 Cleaning levels attained for coal fly ash (HMD - 7 urn) with
several fabrics as a function of fabric loading and/or
estimated separating (adhesive) force 72
24 Resistance characteristics and K£ values for fly ash and
Dacron felt, inlet concentration 27.6 g/m, filtration
velocity 2.6 m/min 75
25 Resistance characteristics and K-, values for talc with dacron
o
felt, inlet concentration 3.5 g/m , filtration velocity
2.6 m/min 76
26 Effective residual pressure loss versus d(Ap)/dt, 0 to
0.01 sec. Fly ash filtration at 2.6 m/min 77
27 Observed outlet concentrations for bench scale tests with coal
fly ash and woven glass fabrics as a function of fabric
dust loading and filtration velocity 81
28 Effect of inlet concentration on predicted outlet concentrations
at a face velocity of 0.61 m/min with coal fly ash and glass
fabrics (bench scale tests) 83
29 Coal fly ash filtration, dacron felt with pulse jet cleaning,
99.83% overall mass efficiency 85
Vlll
-------
FIGURES (continued)
Number Page
30 Apparent fractional penetration for coal fly ash versus pulse
pressure and type, dacron needled bag (1.22 m x 11.4 cm) ... 86
31 Typical boiler/baghouse schematic diagram 90
B-l Cascade impactor for particle size determination 164
D-l Cutaway drawing of an "RJ" reverse jet fabric filter—capacities
from 3,000 to 76,500 acfm (courtesy of CEA—Carter-Day—
excerpt from Bulletin No. RJ-2) 181
D-2 Schematic of air flow in a "PC" Pactecon dust collector
(courtesy of W. W. Sly Manufacturing Co.—excerpt from
Instruction Book No. 693) 182
D-3 Series 130 cabinet cloth filter type self-contained dust
collector (courtesy of the Torit Corp.—excerpt from Bulletin
No. 1565) 183
D-4 Typical damper valves used in baghouse systems 184
D-5 Poppet damper detail (courtesy of American Air Filter Company,
Inc.) 185
D-6 Schematic diagrams of filtering and cleaning operation modes in
a type DW reverse pulse collector (courtesy of American Pre-
cision Industries, Inc., Dustex Div.—excerpt from Bulletin
No. 571) 186
F-l Fabric filter model - data input form prepared from Table F-l
input data 192
-------
TABLES
Number Page
1 Collection Parameters and Initial Efficiency for Woven Fabric
Filters for Fiber Phase Collection 18
2 Estimated Values for Diffusion, Interception, and Impaction
Parameters for Granular Bed Collection 20
3 Estimated Overall Weight Collection Efficiencies as a Function
of Cake Thickness and Inlet Particle Size for Coal Fly Ash,
MMD = 6.4 urn, og = 2.3 22
4 Equations for Computing Characteristic Diameters for Particle
Systems Described by a Logarithmic-Normal Distribution .... 36
5 Key Fabric Properties 38
6 Physical and Chemical Properties of Fibers Used in Filter
Fabrics 43
7 Special Nomenclature, English and Metric Equivalencies for Key
Filtration Parameters . . . . ,. 47
8 Probable Ranges for ^ and SE> Based on Laboratory Tests . . 49
9 Summary Data - Effect of Reservoir Pressure, Face Velocity, and
Inlet Fly Ash Concentration on Residual Fabric Loading and
Average Pressure Loss 71
10 Effect of Pulse Duration and Frequency, and Direct and Damped
Pulses on Fly Ash Fractional Penetration, Dacron Needled Felt
at 2.6 m/min Air-Cloth Ratio 87
11 Typical User Problems Encountered with Fabric Filter Systems,
Frequency of Occurrence, and Estimated Impact on Product
Manufacture or Energy Generation 93
12 Troubleshooting Guidelines for Common Baghouse Operating
Problems 99
-------
TABLES (continued)
Number Page
13 Typical Maintenance Schedule for a Fabric Filter System 117
14 Typical Fabric Filter Inspection and Maintenance Procedures,
User Responses 119
15 Example of Single Page State Compliance Inspection Report
Format 127
16 Suggested Fabric Filter Field Inspection Report Format 128
17 Example Fabric Filter Field Inspection Report 135
B-l Tabulated Sizing Data, Cascade Impactor Sampling of Coal Fly Ash
Aerosol 166
F-l Available Input Data for Modeling Baghouse Performance at
Electric Utility, Plant B 191
F-2 English/Metric Conversion Factors 191
F-3 Samples of Tabular Printout for Example of Model Application . . 194
XI
-------
CHAPTER I
INTRODUCTION
1.1 PURPOSE OF MANUAL
Fabric filtration (1-8) is expected to play an increasingly important
role as the preferred control technology for reducing particulate emissions
from utility, industrial, and other sources to the strict compliance levels
now set by Federal, State, or local agencies. However, compliance with
Federal Standards of Performance for New Stationary Sources or other
appropriate standards set forth in SIP rules and regulations may be difficult
to attain or consistently maintain unless it is assured that:
• fabric filtration is indeed the best choice (or at least an
acceptable alternative) for controlling a specific particulate
emission, and
• the fabric filter is operated and maintained in accordance with
accepted engineering practice.
The purpose of this Fabric Filter Inspection Manual is to enable Federal,
State, or local field inspection personnel to make the correct field
observations and to ask the right questions so that they can provide the
enforcement office with the following information:
• Does the filter system appear to be in compliance based on observed
operating parameters and performance characteristics?
• Is the system being applied and operated in accordance with the
design purposes and specifications?
• Do the plant filter system inspection and maintenance procedures
(and those for applicable process equipment) appear likely to
guarantee continued compliance with emission regulations?
The Fabric Filter Inspection Manual is designed to answer these questions
by familiarizing technically trained enforcement personnel with the practical
aspects of designing, operating, and maintaining fabric filter systems. On
the other hand, basic filtration concepts are presented to enable inspection
personnel to evaluate certain facilities for which no precedence has been
established. Considerable emphasis, however, is directed to the day-to-day
aspects of filter system operation with respect to what may go wrong and what
corrective measures should be taken to prevent or minimize system malfunctions,
1.2 MANUAL ORGANIZATION
1.2.1 Special Role of Chapter II
The organization and technical content of the manual is intended to
provide enforcement personnel with the tools to conduct effective field
inspections of fabric filter systems. The judgements that must be made as to
-------
whether compliance with emission regulations has or can be attained depend
upon how well the inspector understands not only the practical aspects of
fabric filter operation but also the underlying concepts controlling the
filtration process.
Given the need for rapid decision making, the user can go immediately to
Chapter III and subsequent chapters for technical guidance without risk of
serious errors of judgement, particularly so if the individual has had prior
training and experience in filtration technology. Conversely, if the manual
user has had only limited filtration experience and the evaluation of a fabric
filtration operation permits some lead time, his ability to analyze system
behavior and to assess its potential for emissions compliance will be greatly
enhanced by a careful reading of Chapter II, "How Fabric Filters Work". In
the long term, we strongly advocate that inspection personnel develop a solid
qualitative understanding of the basic concepts presented in Chapter II.
These include a review of fundamental particle collection mechanisms,
collection theory, gas and particle properties relevant to fabric filter
design and operation, choice of performance criteria, and selection of fabrics
and collector types.
Because of the importance of knowing the size and concentration of the
particles to be collected, a section of Chapter II has been devoted to the
sampling and measuring techniques recommended for field aerosol evaluation.
Although the technical coverage is necessarily brief, the highlighting of key
sizing and sampling concepts supported by extensive references will help the
field inspector avoid many common mistakes.
1.2.2 The Working Sections of the Manual
The practical "hands on" aspects of inspecting fabric filtration
facilities for compliance purposes are described in Chapters III and IV and
related appendices in the working sections of the Manual.
Chapter III, Typical Baghouse Problems/Corrective Actions, addresses
specific operational problems and their respective solutions that concern the
baghouse user. In addition to a general discussion of typical problems
encountered with baghouse components, ancillary equipment, and processes being
controlled, actual field (user) experience in these same areas is presented in
this Chapter.
Chapter IV, Baghouse Inspections, Types, and Procedures, presents
information on the various kinds of inspections performed by users, vendors,
and air pollution agency personnel. Depending upon the required type of
inspection; i.e., compliance determination, startup, troubleshooting, general
preventive maintenance, or special investigations, different information may
be sought or the sequence of the inspection process may change. In addition
to a discussion of the above items, a suggested inspection report format has
been provided.
Additionally, several appendices are provided to aid air pollution
engineers in their performance of fabric filter surveys or inspections.
Appendix A is a glossary of the more common terms used for describing aerosol
-------
systems and sampling procedures while Appendix B provides supplementary
information on size measurement techniques. Appendix C lists the terminology
associated with baghouses, baghouse operation, and fabric properties.
Additional examples of available baghouse designs and ancillary equipment are
presented in Appendix D and U.S. suppliers of baghouses, based on 1982-1983
listings, are indicated in Appendix E. Appendix F highlights the key features
of a predictive model that is intended to facilitate the design and evaluation
of fabric filter systems.
The information contained in this manual should provide the reader with a
representative data base that will enable more detailed and meaningful
inspections of fabric filter systems. At the same time, a thorough
appreciation of field problems should encourage cooperation between the
control agency and the filter user in seeking to correct emission problems.
It is also pointed out that fabric filtration may not necessarily be the
optimum choice for emissions control when highly humid, corrosive, or very
high temperature gas streams require cleaning. The reader is directed to
alternative control technologies such as wet scrubbing or electrostatic
precipitation(9-14).
-------
CHAPTER II
HOW FABRIC FILTERS WORK
2.1 BACKGROUND
Fabric filters represent but one of several filter types used to remove
particulate materials from waste or process gas streams (1-8). Although
fabric filter applications and related maintenance constitute the sole subject
of this manual, field enforcement personnel should be aware of the other
filter types and their normal applications because they are often encountered
at the same plants where fabric filters are used. Progressing from the least
to the most efficient, the four basic filter types (2,3,7,8) are as follows:
• Fiber mats or screens (15-18) composed of straight or curled fiber
filaments or ribbons, 35 to 150 um diameter, randomly packed in
highly porous (~98 to 99 percent) layers and ranging in depth from
1/2 to 4 inches. Structural stability is provided by supporting
screens or adhesive bonding (~5 to 10 percent by weight) and dust
adhesive properties are usually enhanced by sticky oil coatings.
The primary application of such filters is to remove coarse
(~10-20 Mm) particles from the ambient air or from process air with
solids concentrations less than 500 yg/m (a) to prevent the
dispersal of staining, sootlike particles, pollen grains, and gritty
material likely to damage internal combustion engines, or (b) to
extend the service life of costly, ultrahigh efficiency HEPA*
filters used to collect highly toxic particles such as radioactive,
chemical, and biological materials. Mat or screen units are
commonly described as roughing, precleaning, furnace, or "throwaway"
filters, wherein all except those composed of heavy-duty metal
frames and metal screens are thrown away once they become plugged.
Steam cleaning followed by re-oiling (dipped or sprayed) are used to
recycle heavy metal screen or fiber units used, for example, to
prefilter diesel engine combustion air or public building supply
air. The efficiency of this filter class may be as high as
90 percent for >10 urn particles, provided that the fiber medium is
uniformly dispersed and lubricated.
• High efficiency fiber mats (19-21) fabricated in pleated
configurations and composed of randomly packed, relatively fine
(5 to 10 ym) adhesive-bonded fiber, usually glass. The primary role
of such filters, described as high efficiency precleaners, is to
reduce approaching dust concentrations in the 1 Um size range by a
factor of ten or so to extend the life of costly, HEPA-type filters
or, when used alone, to provide high quality air to critical
occupied areas where soiling or bacterial presence must be minimized.
*HEPA - High Efficiency Particulate Arrester
-------
• Fabric filters (l~8) consisting of woven or felted media used in the
form of bags, tubes, and sometimes envelopes to provide high
efficiency cleaning of high particulate loadings, 2 to 20 grams/ni .
The utilization of the fabric and the overall operation of
industrial filter systems embody the same principles seen in home
vacuum cleaners except for the provision for cleaning, without
interruption of gas flow, and the capability to function in
difficult working environments involving high temperatures,
corrosive atmospheres, and high dust loadings. The key factor
differentiating fabric filters from the other filter types discussed
in this section is that the high efficiency filtration is performed
by the dust layer that accumulates on the fabric surface. The role
of the fabric is to provide a stable support or substrate for this
dust layer that will (a) permit a uniform, unbroken dust layer to
develop, and (b) withstand the rigors of the working environment
during filtration and, in particular, during any cleaning process
that involves repeated stressing and flexure of the fabric. In both
appearance and structure, filtration fabrics are similar and often
the same as the woven or felted products used for clothing,
protective coverings, substrates for plastic sheeting, decorative
purposes, etc. Details with respect to weave, yarns, thread counts,
and special surface treatments will be discussed in a later
section. Operating temperatures, resistance to chemical attack, and
method of cleaning determine the source and chemical classification
of the fibers making up the filter. Typical operating efficiencies
are usually well in excess of 99 percent, and often greater than
99.9+ percent, provided that the system has been properly applied,
installed, and maintained. This manual will deal only with the
operation and maintenance of fabric filter systems.
• Ultrahigh efficiency particulate arresters (3,6,8,22,23) (HEPA),
consisting of tightly pleated, high-efficiency mineral fiber
(0.5 to 1.5 urn diameter) papers providing gas handling capacities of
500 to 1,000 ft3/min for 2 ft. x 2 ft. cells ranging from 6 to
12 inches, respectively, in depth. Refractory edge seals and fibers
allow high temperature operation (up to 1000°C) for low
(100 yg/m ) concentrations, with efficiencies in excess of
99.95 percent against 0.25 um diameter particles. Major
applications for HEPA filters are the collection of radioactive
and/or biological particles, although they are also used extensively
for clean room operations to prevent particle fouling of delicate
optical systems or solid state electrical circuitry.
2.2 MECHANICS OF DUST FILTRATION
2.2.1 Basic Concepts (1,2,8,24-31)
Particle capture by any single collector, regardless of shape, depends
upon a combination of particle, collector, and gas stream properties and the
relative velocity between particle and collector. Single collector in the
above context refers to an individual fiber, droplet, or any other geometric
shape with which a particle makes contact. In most instances, the motion of
-------
the gas stream past fixed or moving obstacles (collecting fibers, granules, or
liquid droplets) determines the relative velocity, although thermal, diffu-
sional, electrical, and vibrational forces also contribute to velocity effects.
The interrelationships among the major variables responsible for particle
collection are reviewed briefly in this section. The most important step in
analyzing any filtration system is to determine what causes a specified
particle to be brought into contact with the filter fiber(s) neglecting for
the moment any subsequent particle dislodgement by gas stream re-entrainraent.
First, with respect to any filter system, the carrier gas stream is diverted
by the fibers obstructing the flow such that it assumes a winding path.
2.2.2 Particle Capture by Direct Interception (1,8,22,24-26)
If the particle trajectories are always coincident with the direction of
the gas streamlines regardless of changes in gas stream direction, only those
particles whose boundaries extend beyond the supporting streamline will make
physical contact with the collector. Although such contacts are not always
irreversible due to frequent particle dislodgement and re-entrainment by the
gas stream, irreversible capture is assumed for present analytical purposes.
This collection mechanism, which is defined as direct interception, plays a
major role in filtration when the particle and fiber diameters are very
similar and where the fiber proximity precludes direct penetration of most
particles. Figure 1 illustrates how particles are captured by direct
interception, a process that is independent of the physical properties of a
particle (except for its location). A rough estimate of the magnitude of the
interception effect is provided by the interception efficiency,
nD]. = (2 - In Ref) A | ^ ] , (2-1)
that indicates the fraction of particles removed from a gas stream tube having
the same cross-sectional area as that of the collector fiber (25).
Equation (2-1) shows that the interception effect increases as the fiber
or collector diameter, Df, decreases relative to that of the airborne
particle, D . Hence, one should expect better performance for the
individually dispersed fibers of a felt fabric (~15 to 30 ym in diameter) than
for the tightly entwined fibers making up the yarns whose overall dimensions
in a woven fabric may range from 200 to 500 m.
Because the stagnant gas film or boundary layer surrounding a fiber or a
granule increases significantly as the flow velocity decreases, the fluid
streamlines assume more divergent paths to circumvent any obstacle in the
stream (22). The net effect is to make it more difficult for a suspended
particle to contact a potential collector. Use of the fiber Reynolds number,
Ref, criterion provides an approximate measure of the boundary effect.
Ref * DfV'P• /p, (consistent units) (2-2)
-------
CJ Electrostatic Attraction
b_
2
A) Direct Interception
•*• — Di
^4
JL \
2 t
ZV
A
V
T
B) Inertial Impaction
Diffusion
Figure 1. Streamlines and particle trajectories approaching filter element.
-------
where Df = fiber diameter
V = relative velocity between particle and fiber
Pg = gas density
u = gas viscosity
Very low values for Ref (<1.0) indicate laminar or streamline flow,
whereas high values (>10 ) reflect turbulent conditions that enhance
particle transport to a collector by eddy motion. Ref levels from >1 to
10 are associated with what one describes as "transitional" flow.
2.2.3 Particle Capture by Inertial Impaction (1,8,22,24,32,33)
Inertial forces will cause suspended particles to diverge from the gas
stream trajectory as the latter bends to avoid an obstruction, Figure 1. The
extent to which a particle departs from its parent streamline and migrates to
a collecting surface depends upon the interaction between the inertial (or
centrifugal) force generated by the particle motion in a curving gas stream
and the drag force, Fj, resisting the particle motion. The drag force is a
function of gas stream viscosity, particle size, and the instantaneous
^•v
relative velocity, V, between the particle and supporting gas stream.
F = 3 TydV (2-3)
The above effects can be described by a diraensionless term called the
impaction parameter, 1-j, that indicates, as does ^nj* tne fraction of
particles removed from an approaching gas stream having the same cross section
as that of the obstructing fiber (or collector). The variables defining the
impaction parameter, Hj are often expressed by the following functional
relationships :
(2-4)
According to Equation (2-4), increases in particle density, Pp,
particle diameter, Dp, and the relative velocity, \f, between the particle
and collector all favor increased single fiber collection efficiency.
Conversely, increasing gas stream viscosity and collector (fiber) diameter
reduce single fiber collection efficiency. Except for very large particles,
>100 Mm, most particle velocities are essentially the same as that of the
conveying gas stream. Therefore, in the case of fabric filters, V is usually
the same as the velocity of the gas stream, Vg, approaching the collector.
Values of Hj for various collector geometries may be estimated from the
graph shown in Figure 2. Inertial impaction plays a major role in the initial
filtration of particle sizes greater than 1 urn. Although particle capture is
assumed to be irreversible in Equation (2-4), bounce and re-entrainraent are
always associated to some degree with the collection processes, particularly
so at high velocities (33).
-------
1.0
S 0.8
o
z
1X1
Z 0.4
O
o
<
o.
a 0.2
1 I I
^ Cylinders
i ill
0,05 0.1 0.2 0.4 0.6 1.0 2 4
^ ***
IMPACTION PARAMETER ,Cc/>pDp V
6 8 10
20
Figure 2. Impaction efficiencies for cylinders, spheres, ribbons,
and disks.
-------
When dealing with particles 1 urn or less in diameter, the term Cc in
Equation (2-4), the Cunningham-Millikan correction (22,24-26), may become very
large and hence very important in predicting particle collection. This
correction takes into account the fact that particles appreciably smaller than
the mean free path of a gas (A)~0.6 pm at ambient temperatures, can literally
"slip" through the gas with more ease than larger particles. For rough
calculating purposes, Cc may be taken as 1.0 for diameters of 1 urn or
larger. It may be estimated from the empirical relationship:
Cc = 1 + 1.7 (A/Dp) (2-5)
The term V^ refers to the terminal settling velocity that is also equal
to the relative velocity between a particle and its suspending gas stream when
the force inducing particle motion in the gas is exactly equal to the
aerodynamic drag force. When particle settlement occurs in still air, Vc is
the actual particle velocity.
2.2.4 Particle Capture by Brownian Diffusion (1,8,22,24,32,34)
Brownian diffusion contributes significantly to the motion of the
submicrometer particles as the result of molecular bombardment. The diffusion
coefficient, Dg, may be estimated from the relationship:
DB = kBT, (2-6)
where k is the Boltzman constant, T the absolute gas temperature, and B the
particle mobility as defined below:
B = Cc/3 nyDp (2-7)
As shown in Figure 1, random particle excursions from their parent
streamlines, which increase in frequency and magnitude as particle size
diminishes, enhance the probability of particle capture by adjacent collection
surfaces.
Particle collection efficiency by diffusion mechanisms, np, is often
expressed in terras of a dimensionless group called the Peclet number (22):
%= k(Pe)"2/3 = k 1^1"'" > (2'8)
where Pe = D^V/Dg and k is a term that varies from 6.2 to 2.4 over the
range of 0.7 to 0.99 for fabric or granule bed porosity. Equation (2-8)
indicates that particle collection can be expected to increase with increasing
diffusion coefficient, Dg, decreasing collector or target diameter, D^,
and decreasing gas stream velocity, V. Because relatively low filtration
velocities are used in fabric filtration and the fiber sizes instrumental in
particle capture as well as the filter pore dimensions may be quite small (~10
to 20 urn), diffusional mechanisms can play an important role during the
earlier stage of fabric filtration until a superficial dust layer has formed.
Diffusional capture is most important for particles of submicroraeter size.
10
-------
2.2.5 Particle Capture by Electrostatic Mechanisms (1,8,22,24-26)
Although electrostatic charge plays a major role in particle collection
by electrostatic precipitators, it is usually difficult to determine how much
effect the electrostatic charge on particles and/or fibers has upon capture
during filtration. In most field applications, it is a common occurrence to
find that a certain fraction of the dust bears an essentially unipolar charge,
with either positive or negative charges predominating except for very low
particle concentrations. Usually, most particles bear less than 10 percent of
their maximum possible charge. Theory suggests that charge phenomena may
augment the performance of a filter system (Figure l) under the following
conditions: a charged fiber and an uncharged particle, particles and fibers
charged to opposite polarities, and a charged particle and an uncharged fiber
(35-40). In routine fabric filter applications with dust loadings in the
range of 0.1 g/m to 20 g/m , the above electrostatic mechanisms are
mainly of academic interest with respect to enhancing collection efficiency.
In the case of low-to-moderate efficiency roughing filters, collection
efficiency may be enhanced by impressing an electric field across a filter
composed of low dielectric, glass or mineral fibers (35,37,39). Local
distortion or convergence of the field in the immediate vicinity of the
dielectric fibers causes uncharged or charged particles of either polarity to
migrate toward the fibers. This process, which is sometimes described as
dielectrophoresis, is an adjunct to collection only when particulate loadings
are near the ambient level, 25 to 250 yg/m . At high concentrations, its
contribution to particle collection will be insignificant in comparison to the
effects of the collection mechanisms discussed earlier in this section.
The concept of precharging dust particles prior to filtration as a means
of increasing gas handling capacity without the matching penalty of high
pressure loss is now under investigation at the prototype level (41). In
contrast to a naturally charged aerosol, this process generates high charge
levels of single polarity. The benefits of this approach may be significant
provided that cost problems associated with construction, operation, and
maintenance of the charging system are overcome. It appears that this process
reduces the pressure loss through the dust layer by causing the latter to
deposit at a higher porosity due to repulsion of similarly charged
particles (38).
For most practical filtration applications, the presence of natural
electrical charges may usually be ignored except for certain dusts,
e.g. powdered resins that are characterized by dense wall deposits and extreme
stickiness relative to dislodgeraent from fabrics. Such systems will be
discussed under the special problem category.
2.2.6 Particle Collection by Direct Sieving (1-3,22,28,29,31,42)
Collection by sieving implies that the openings, channels, or separation
distances between filter fibers, yarns, or granules are less than the
characteristic dimensions of the approaching particles. Hence, the collector
may be considered as a screen whose mesh size either prevents or allows
particle penetration. Ordinarily, the fiber separation distances in most
11
-------
fabrics are too large to allow for direct sieving of any but the very large
particles (~20 to 50 pm) present in the gas stream. Thus, during the first
minute or so of filtration, direct interception and diffusion, with
augmentation from inertial impaction and electrostatic effects, play important
roles in causing the particle bridging that must precede the formation of a
superficial dust layer (22,43). Bridging describes the process wherein the
sequential deposition of particles, first on fibers and then upon previously
deposited particles, leads to formation of chain-like agglomerates (34). At
dust concentrations in excess of 0.5 g/m^, these chains rapidly interlace
with adjoining members to form the supporting bridge for dust cake development.
It is the dust cake that ultimately constitutes the true filter, and the
process by which particles are collected can be best described as sieving
because particle proximity in the dust cake per se reduces the size of the
pores or channels to the same order as those of the approaching particles (2,
22,28,42,44). Despite the fact that dust cake porosity usually exceeds that
for the bulk dust removed from baghouse hopper (~60 to —90 percent versus 40
to ~60 percent for the bagged dust), the distance between particles is suffi-
ciently close to make direct sieving the predominant capture mechanism. In
the above case, the porosity, or void fraction, e, can be related to discrete
particle density, Pp, and cake bulk density p by the following expression:
e = 1 - (P/Pp) (2-9)
The effectiveness of the sieving operating is frequently masked, however,
by defects in the dust cake surface, cracks or unblocked pores, tears in the
fabric, temporary absence of a dust cake that is a normal part of the fabric
cleaning process, and a nearly constant seepage of agglomerated particles
through the fabric substrate as a result of the mechanical action induced by
cleaning and aerodynamic drag forces (31,42,45,46). The degree to which the
above factors contribute to the total filter effluent is related to a
combination of dust and fabric properties, method of cleaning, and filter
system maintenance procedures. Except for will-weave glass fabrics frequently
used for coal fly ash filtration, there is no existing theory or even an
empirical approach that allows for reliable estimation of dust penetration
resulting from the above factors.
It has been observed that unavoidable pinhole leaks with glass fabrics
are responsible for 95 to >99 percent (raasswise) of the total filter
effluent. Because only the coarser particles (>20 urn range) are removed from
the gas that passes through the pinhole leaks, the particle size properties of
the effluent are very similar to those of the dust approaching the filter,
even though the overall collection efficiency may well exceed
99.9 percent (31,42).
2.2.7 Particle Capture by Gravitational Forces (1-8,22)
Gravitational forces can usually be discounted as a contributor to
particle capture within the dust and fabric structure of the filter per se.
Only in the case of very deep bed granular or fiber filters (~ several feet)
and at low velocities (<0.5 ft/min) may the impact of gravity settlement be
seen, and then only for systems where the inlet dust concentrations are at
12
-------
ambient levels of ~25 to 250 pg/m . On the other hand, gravity settlement
with augmentation by inertial forces may lead to significant losses of large
(>20 (im) particles in lengthy horizontal ducts and in breechings, transition
sections, or collector inlets where increased gas stream cross section leads
to greatly reduced flow velocities. The possibility of particle separation fay
gravity forces should be taken into account when dust sampling for estimates
of mass concentration and particle size is performed well upstream of the
fabric filter (or other control devices) (31,42,46,47). The extent to which
gravitational effects may be expected to influence particle separations may be
estimated by computing particle terminal velocity, Vt. The latter term
describes the constant rate of particle fall in still air when the resistance
to motion presented by the air (or suspending gas) equals the local
gravitational force.
C (p - P )gD 2
v = c P E E_ (2-10)
*
For many applications, the Cunningham-Millikan correction, Cc, is
approximately 1.0 in situations where Vt is of concern and, for most aerosol
systems, the gas or fluid density, P£, can be neglected because it is
roughly 500 to 1,000 times less than that for the particulate materials. As a
frame of reference, a 61 m diameter silica particle will settle at about
1 ft/sec (30.5 cm/sec) in air at standard conditions (760 mm Hg, 20°C). Given
a reduced air velocity of 10 ft/sec in a 50 ft long breeching section, all
61 m particles within 2 ft of the bottom of the ducting will fall out.
Detailed analyses of settling phenomena, including the more complex process of
settling behavior in turbulent gas streams, are treate4 extensively in the
literature (1,8,22,24,25).
2.2.8 Physical Processes Enhancing Particle Capture
2.2.8.1 Agglomeration (1,22,24-26) —
Agglomeration (or coagulation) is the process by which airborne
particulate materials collide and adhere to form larger particles. When
liquid droplets collide, the process is called coalesence. The mechanisms and
forces promoting these collisions (and particle growth) are the results of
Brownian motion and electrical, aerodynamic, gravitational, sonic, or magnetic
effects.
Electrical and sonic forces, however, have had limited applications as
means of enhancing particle contacts, whereas naturally occurring thermal or
Brownian diffusion and the aerodynamic mixing engendered by gas stream
turbulence are major contributors to agglomeration. The benefit of
agglomeration is that it increases particle size, which in turn, means
increased probability of particle capture by the direct interception and
inertial impaction processes described in Equations (2-1) and (2-4). It
should also be recalled that Brownian motion and turbulent diffusion increase
the probability of particle contacts with the collection surface (fiber,
granule) or collector wall in the case of baffle or cyclonic devices.
13
-------
The rate of thermal agglomeration (coagulation) may be estimated by the
comparatively simple equation,
nt = n0/(l + K'n0t,) (2-11)
where nt and no refer to the particle concentration at time t and time
zero, respectively, and K1 to the Smoluchowski coagulation constant.
K' = 3 x 10~10 cm3/sec particle
o
nt, no = particle/cm
t = sec
The important role played by agglomeration is demonstrated by the average
reduction in particle number concentration when initial concentrations are
high. For example, freshly generated fume at an initial concentration level
of 10*0 particles/cm^ would undergo a nearly tenfold decrease in
concentration in less than three seconds. If the initial mean size by number
were roughly 0.1 urn, the average mass per agglomerate would increase tenfold
and the irapaction parameter, Hj, about 3.5 times, assuming a 50 percent
porosity for the agglomerates. Hence, the potential for increased particle
capture by filter systems is greatly enhanced while the pressure loss per unit
mass of dust deposit would be lowered significantly because of the more porous
structure. Although K1 has been treated as a constant, it does in fact depend
upon gas temperature, gas viscosity, and Cc, the Cunningham-Millikan
correction. However, over the diameter range 0.1 to 1.0 Ura, the magnitude of
K1 shows less than a threefold reduction (8.57 to 2.99 x 10~10).
2.2.8.2 Condensation and Evaporation (1,22,24-26)—
Although neither evaporation nor condensation play direct roles in most
filtration applications, there are two situations where these processes must
be considered. The first relates to the treatment of hot, humid gases, such
as, typical fossil fuel combustion effluents, wherein reduction of
temperatures below the vapor dew point will lead to liquid condensation on all
particle and system surfaces below the dew point temperature. This behavior,
which constitutes a decided advantage in the case of liquid scrubbing systems
because of the resultant growth in particle size, creates a serious pressure
loss problem with a dry filtration system. The adverse effects may range from
that of a sticky dust deposit that requires a high energy expenditure to clean
the filtration surfaces to the extreme condition of a blinded filtering
surface impervious to gas flow. The above problems are avoided by maintaining
system gas stream temperatures at all points and at all times well above the
dew point of any condensable constituent.
The second process, that of evaporation, promises to be an important
pre-preparation or conditioning step in dry scrubbing systems where absorbable
gases such as sulfur oxides are first captured by transfer to liquid droplets
containing reactive alkali materials followed by flash drying such that the
resultant aerosol consists solely of dry solids. Subsequent filtration, again
at above dew point temperatures, affords very high particulate removal and up
to 90 percent SOX collection. In the above case, process effectiveness
-------
centers upon regulation of the evaporation or drying process. Although no
detailed discussion of condensation or evaporation phenomena is called for in
this manual, it is important to recognize the key factors involved in these
mass transfer operations.
The following equation provides rough estimates of droplet evaporation
rates, -dm/dt, for diameters as small as 2 ytn.
- d™ = 47iD in (C - Or (2-12)
dt v 2 o
where Dy = diffusion coefficient for evaporating or condensing molecules
m2 = mass of diffusing molecule
C0 = vapor molecular concentration at droplet surface
C = vapor molecular concentration in surrounding gas stream
r = droplet radius
When the molecular concentration (or partial pressure) of the condensable
vapor is less in the surrounding gas stream than that at the droplet surface,
the droplet will decrease in size (evaporation), whereas a reversal in
gradient leads to condensate and droplet growth. The rate of size change also
increases with droplet size and the process is greatly accelerated in a
turbulent atmosphere where molecular transfer from or to the droplet may
increase by orders of magnitude.
2.3 FABRIC FILTRATION VERSUS SINGLE FIBER COLLECTION
2.3.1 Single Fiber Collection Efficiency and Amount of Fiber (1,2,8,24,31)
In the preceding sections, the major particle capture mechanisms have
been identified with respect to single particle-single collector
interactions. Several equations were given which, based upon theoretical
considerations, provide relative measures of particle collectability for
various combinations of particle, fluid (gas), and collector (fiber and
granule) parameters. It is important to note, however, that the computed
collection efficiency for a single particle-single collector system may be an
extremely poor performance indicator for a real fabric filter.
Two extreme conditions may be cited to demonstrate this point. In the
first instance, the calculated collection efficiency for a specified
particle-fiber combination is very high (~0.9). Unfortunately, a single fiber
will capture only those particles that appear within a cross sectional area
representing 90 percent of the projected fiber cross section. This means that
all particles beyond this boundary will never contact the fiber. Thus, if the
fiber projected area represents but 1 percent of the total cross section
through which the dust-laden gas passes, the maximum possible filtration
efficiency cannot exceed 0.9 percent.
15
-------
The other extreme is represented by the system in which single particle-
single fiber efficiencies are low (~5 percent), but where a large number of
such fibers are dispersed within the cross section through which the dusty gas
must pass. Given sufficient fiber projected area, the efficiency for the
fiber system composed of bulk, woven, or felted fibers may then approach
100 percent. In the following discussion, simple analytical procedures are
described for estimating the collection efficiency of a typical woven fabric
filter.
2.3.2 Typical Fabric Filtration Process (1-3, 28,29)
A schematic representation of woven glass fabric during various stages of
filtration, Figure 3, shows that initial particle collection occurs on the
free or bulk fibers that project into the spaces between the yarns (31,42).
With increased service, the free fiber region rapidly bridges over so that
subsequent filtration is accomplished by the overlying dust layer. Note that
the role of the fibers tightly entwined within the yarns is merely to provide
strong support for the free fibers (and the dust deposit) as well as the
strength needed to withstand the long term wear and tear of repeated
cleanings. At the inception of filtration, particles retained by the fibers
are captured by a combination of processes. Diffusion and/or direct
interception account for collection of the finer particles, while inertial
impaction is responsible for the collection of larger particles. However,
once a substantial surface dust layer has formed, direct sieving becomes the
predominant capture mechanism for all regions where the dust layer is unbroken
(free of cracks, pinholes, or other defects).
2.3.2.1 Initial Fiber Filtration Phase—
Based upon theoretical considerations, the initial efficiency, Efiber>
for the free fiber regions can be estimated by the following relationship (22).
i 4 P HL .
E-., = 1 - exp ——— = 1 - exp
fiber f ^d P I
(2-13)
when the porosity or void fraction is 0.90 for the fibrous region. The terms
P and Pf refer to bulk and discrete fiber densities, respectively; L is the
bed thickness, df the fiber diameter, and n the single particle-single fiber
collection efficiency for the relevant particle size and particle capture
mechanisra(s). The ratio P/Pf, which is defined as filter bed solidity, is
usually represented by the symbol f3. Solidity is related to filter porosity
by the expression, 13 = 1-e.
Estimates of single fiber collection efficiencies for coal flyash
collection by inertial (HI) and interception (H0j) mechanisms are shown in
Table 1 for a fiber diameter, d^, of 8.5 urn and an actual flow velocity of
3.85 m/min. The latter value depicts an average pore velocity encountered
with typical woven glass fabrics having an overall fabric areal density of
340 g/m . Although both mechanisms show Hj and HJJJ; increasing as
particle diameter increases, the resulting efficiency values are far below
those necessary to provide high efficiency removal. However, introduction of
the single fiber efficiency parameter into Equation (2-13) in conjunction with
16
-------
•YARN
YARN
BULKED FIBERS
OUST
UNUSED FABRIC
EARLY OUST BRIDGING OF FIBER SUBSTRATE
SUB SURFACE OUST CAKE DEVELOPMENT
SURFACE DUST CAKE DEVELOPMENT
Figure 3. Schematic, dust accumulation on woven glass fabrics.
17
-------
TABLE 1. COLLECTION PARAMETERS AND INITIAL EFFICIENCY FOR
WOVEN FABRIC FILTERS FOR FIBER PHASE COLLECTION3
dp
(um)
1
2
3
4
5
10
-i
0.02
0.10
0.30
0.45
0.60
0.80
-DI
0.003
0.01
0.023
0.041
0.064
0.26
Fractional*3
efficiency
0.014
0.44
0.83
0.93
0.97
0.996
aFiltering temperature of 20°C.
^Conservative estimates based on Ij alone, the larger of
the collection parameters. Note that effective n should be
greater than Hj, but less than HI +
18
-------
best estimates for the free fiber bulk density, P, and depth of fiber, L,
prevailing in the pores of common woven glass fabrics lead to the overall
fractional efficiency values shown in Column 4, Table 1 (48). One can infer
that for the fabric of interest, inertial irapaction is the controlling capture
mechanism at the start of the fiber filtration phase. Additionally,
respirable particles (< 2 - 5 Mm) are removed at very low efficiencies by the
clean (unblocked) fibers, whereas larger particles (4 to 5 Mm) are collected
at 95 percent efficiency. Capture of the larger particles permits the
bridging process to continue until a solid, unbroken dust cake is formed.
2.3.2.2 Final Cake Filtration Phase—
As more dust is captured by the fibers, the approaching gas stream
encounters a mix of fibers surrounded by chain-like particle agglomerates that
form a progressively better collection system as most dust accumulates. This
process is represented schematically in Figure 3 that shows various stages of
particle capture from the initial particle collection in the free fiber region
to the ultimate formation of a superficial dust layer (31). When the proper
fabrics are selected, the filter surface is relatively smooth, with only
minimal indication of perforations or cracks. Once pore bridging has been
accomplished, the dust collection properties of the filter are governed mainly
by the dust layer. Theory indicates that the dust collection efficiency for a
granular bed composed of uniformly sized particles may be expressed as
follows (22):
«-I-~P fe^sr) (2-u>
where ii is the ratio of the bulk density of the dust deposit, p , to the
density of the discrete or individual bed particle, Pp. The term 1 - B,
usually designated as r, the bed porosity (or fraction void volume), ranges
from 0.6 to <0.9 for many dust deposits. Dust layer thickness, L, granule
diameter, dc, and the single particle-single granule collection efficiency,
n, are the remaining variables affecting particle capture.
Equation (2-14) can be derived from Equation (2-13) by adjusting for
collector shape factor (sphere versus fiber) and by re-introducing the term
(1 - B) that was not included in Equation (2-13) because of small 13 values.
Except for the hypothetical case where the particles to be removed come from a
raonodisperse (uniformly sized) population, practical application of
Equation (2-14) requires integration over the range of particle sizes
characterizing the approaching aerosol and the resultant dust deposit.
Before discussing theoretical collection efficiencies for the unperturbed
dust layer, an illustration of the relative importance of the predominant
capture processes associated with dust layer or "cake" filtration is given in
Table 2 (48). Computations have been made for several particle diameters
constituting a typical dust layer in which the estimated values for particle
capture efficiency by diffusion, direct interception, and inertial impaction
mechanisms are shown for two particle sizes in the gas stream undergoing
filtration. It is again emphasized that these efficiency values reflect
"single particle-single collector" behavior only. The overall removal for any
size category, which depends upon the number of "collectors" dispersed in the
19
-------
TABLE 2. ESTIMATED VALUES FOR DIFFUSION, INTERCEPTION, AND IMPACTION
PARAMETERS FOR GRANULAR BED COLLECTION
Collector
diameter,
<*c
pm
0.25
1.0
2.0
5.0
10.0
Diffusion
HD
0.35
0.17
0.10
0.069
0.035
dp = 0.25 j.
Interception
IDI
2.27
0.25
0.058
0.018
0.002
Particle
im
Impact ion
HI
0.08
0.02
0.01
0.005
0.003
diameter
d
Diffusion
HD
0.32
0.12
0.069
0.037
0.022
p = 1.0 ym
Interception
nDI
36.4
4.0
0.94
0.29
0.038
Impaction
nl
0.5
0.2
0.1
0.05
0.02
aRefer to Equations (1), (4), and (7).
20
-------
filtration zone, must be computed using Equation (2-14). What is strikingly
apparent, however, is that direct interception and diffusion are the
controlling mechanisms for capture of the submicrometer particles.
Use of the interception parameter alone to estimate the overall glass
fabric performance for varying thicknesses of fly ash deposits, Table 3, shows
that collection efficiencies for 0.25 m particles rapidly approach
100 percent and that 1.0 m particles do not penetrate at all (48). Again, it
is emphasized that the behavior of a uniform, unbroken, and unperforated dust
layer is being described. Furthermore, no allowance is made for a very low
level but distinct seepage of agglomerated particles through the underlying
fabric structure that can contribute to small but detectable concentrations
O Q
~0.5 mg/m (0.22 grains/1,000 ft ) in the case of woven glass and down to
0.005 mg/m^ (0.0022 grains/1,000 ft^) as measured for sateen weave
cotton (48).
Additionally, if a poor fabric selection is made, e.g., too open a fabric
weave with a sparse, free fiber cover, the potential exists for the formation
of pinhole leaks such as shown in Figure 4. When the tendril-like growth of
agglomerate chains fails to interlock and bridge the space between adjacent
fibers within a few minutes, the increasingly greater gas velocities through
the leak region can prevent further particle accumulation. Hence, a
characteristic pinhole develops as shown in Figure 4 (31,43,48). In the
extreme case, a situation might evolve where a pinhole cross section
representing about 5 percent of the filter face area actually accommodates
99 percent or more of the approaching gas stream. Because only the large
(15 to 20 um) particles fail to penetrate the pinholes, collection efficiency
in the above example would be less than 1 to 2 percent (42).
Fortunately, a good fabric collector will seldom possess more than 200
pinholes (roughly 100 Mm diameter) per square meter of filtration surface
(less than 2 x 10~^ percent of the total fabric area). Once pore bridging
has been accomplished, the dust collection properties of the filter are
governed mainly by the dust layer.
Because real filters contain "less than perfect" regions where
temporarily or permanently unblocked pores or cracks allow certain fractions
of the inlet aerosol to penetrate the filter with negligible separation of >15
to 20 um particles, the upstream and downstream particle size distributions
are often almost identical. The above phenomenon, which is associated mainly
with woven glass fabrics, is readily explained by the fact that 95 to
>99 percent of the effluent dust comes from the air that has passed through
the pinhole leak regions (31,47,49). In the case of highly efficient fabrics
such as sateen weave cotton, the entering particle size distribution may be
appreciably larger than that for the dust penetrating the filter (31,42).
2.4 GAS PROPERTIES (1-3, 24-26)
In preceding discussions of basic particle capture mechanisms, certain
physical and kinetic properties of the gas stream (gas temperature, gas
viscosity, and velocity) appear as major variables in several particle
collection relationships. Moisture content and condensable materials, through
21
-------
TABLE 3. ESTIMATED OVERALL WEIGHT COLLECTION EFFICIENCIES AS
A FUNCTION OF CAKE THICKNESS AND INLET PARTICLE SIZE
FOR COAL FLY ASH, MMD =6.4 urn, ag = 2.3
Cake3
thickness
Um
10
20
30
100
Areal
density
g/m2
10
20
30
100
Estimated15
filtering
time
min
0.82
1.64
2.46
8.20
Estimated
percent efficiency
Particle diameter, pm
0.25
97.8
99.95
99.9999
100. 00C
1.0
100. 00d
100. ood
100. 00d
100. 00d
abased on an assumed porosity of 0.5 (bulk density of 1 g/cm^).
bftased on inlet concentration of 20 g/m-* (8.7 grains/ft-^) velocity of
0.61 m/min (2 fpm).
cActual estimate 99.00 followed by 16 "nines".
dActual estimate 99.00 followed by 26 "nines".
22
-------
(a). Pinhole leak, filtration surface, showing
characteristic mound, substage lighting
(20x magnification) .
(b). Massive pinhole leakage with monofilament
screen — without loose fibergs.
Figure 4- Variations in pinhole leaks due to fiber presence and pore size.
23
-------
their impacts on dew-point temperatures, automatically establish minimum
operating filtration temperatures which, in turn, affect gas stream viscosity
and velocity. Although the corrosive properties of the carrier gas stream may
have no immediate effect on particle collection, the degradation of the fabric
media will lead to irreparable damage within a matter of hours or days. Also
the bag supporting members, baghouse structure, and gas handling equipment may
eventually suffer damage leading to mechanical failure of key components
necessary for proper filter system operation.
2.4.1 Temperature (49-57)
Gas stream temperature plays a key role in the design, operation, and
performance of a fabric filter system. First, it determines the type of
construction materials and fabrics that must be used. Second, its proper
regulation will prevent the condensation problems cited previously. Third,
gas temperature may play an important part in the control of gas stream
reactions essential for removal of gaseous contaminants. Fourth, baghouse
temperature will determine the total filter surface or cloth area requirement
when the filtration velocity (or air/cloth ratio) has been specified. Fifth,
gas stream temperature will also control volume flow requirements of the
system exhaust fans due to its effect on gas density. Additionally, gas
density also has a direct impact on fiber boundary layer thickness, which in
turn, influences particle capture by direct interception, (see Equation (2-1).
2.4.2 Viscosity
2.4.2.1 Effect on Particle Collection (22,51,53,54) —
Although gas stream viscosity is temperature dependent, its independent
role in filtration warrants separate treatment. First, in any particle
collection process, the effect on increased gas viscosity is to retard
particle motion across the gas streamlines or through a stagnant gas film
(boundary layer). Thus, capture by inertial impaction effects is reduced as
indicated by Equation (2-4). On the other hand, the random Brownian motion
induced by molecular bombardment overrides the particle retardation by
opposing forces, see Equations (2-4) through (2-8), such that collection by
diffusional processes is enhanced.
2.4.2.2 Effect on Pressure Loss Through Dust Laden Fabrics (22,44,58,59)—
During all conventional fabric filtration processes, gas flow through the
fabric and dust layer falls within the laminar or streamline range.
Basically, this type of flow may be considered as a pure translational motion
regardless of path when there is no interchange of gas between the adjacent
streamlines. Resistance to gas motion generated either by constraining
ductwork or by some obstacle in the gas stream results from the shearing
action induced by velocity gradients between contiguous layers of the gas as
the flow velocities increase from zero at the bounding walls or particle
surfaces to their maximum levels at points most removed from constraining or
contacting surfaces.
However, gas flow in the inlet and outlet ducts and also at most regions
within the baghouse proper is usually in the turbulent range wherein a
constant mixing or interchange of particles take place from one duct region to
24
-------
another. It was indicated previously that a fiber Reynold's number, Ref, of
1.0 normally reflects laminar flow (see Equation 2). In most filtration
applications, actual Ref values range from 0.01 to 0.1. If one is concerned
with the turbulence levels within a confined gas flow such as a duct system,
the Reynold's number is computed by the following relationship:
(2-15)
The only difference between Equation (2-2) and Equation (2-15) is that the
duct diameter, D
-------
based upon the unique permeability properties of a dust layer that depend, not
only on the areal density (g/ra ) of the dust deposit, but also upon the
particle size distribution of the dust (l). A complete listing of these
effects is provided in a later section discussing the development and
application of the specific resistance coefficient, K2, of a dust (see Basic
Filtration Concepts - Woven Fabrics).
2.4.4 Moisture, Condensables, and Corrosive Components (1-8,61-63)
The presence of moisture and other condensable substances is of paramount
importance in determining the temperature at which filtration must take
place. Moisture condensation alone is sufficient to produce sticky dust
deposits that contribute to high operating pressure losses and cleaning
difficulties. Maintaining gas stream temperature well above dew point levels
at all points within a filter system at all times (including start-up), with
particular emphasis on bag, duct, and wall surfaces, is a prerequisite to
eliminating condensation problems. Should the air or gas stream contain even
trace quantities of chemically corrosive materials (H2S04, HF, HC1, etc)
that can attack and degrade baghouse structural components with particular
emphasis on fabrics, special attention must be directed to selection of
fabrics and construction materials. It is important that the operating
temperature be based upon the dew point of the material condensing at the
highest temperature when that substance in liquid or solid form can present a
fabric plugging or corrosion problem.
The reaction of active chemical constituents of a gas stream (sulfur
oxides in the case of flue gases) with mineral components naturally present or
injected into the gas steam followed by subsequent collection upon a fabric,
may actually broaden filter system capability when there is a need to capture
both fly ash and sulfur oxides (64). Experimental pilot systems have
indicated sufficiently.high (70 to 90 percent) SOX removal to justify
development of field prototypes. Two major operating criteria must be
observed with the above "dry-scrubbing" systems: a guaranteed above-dew point
temperature at all times, and a fabric medium that will not be attacked by the
absorbable or adsorbable gas.
2.5 PARTICLE PROPERTIES
In the preceding review of particle capture mechanisms, it has been
emphasized that particle size properties play a major role in determining
filter efficiency and, as will be discussed in later sections, equally
important roles in determining filter pressure loss and cleaning
requirements. Therefore, information on dust concentration (mass basis),
particle size distribution, and particle density is absolutely essential to
assess or predict filter behavior. The greater the amount and accuracy of
data pertaining to particle shape and surface properties, electrification
phenomena, and gas and vapor sorption, the greater the chance of a successful
filtration process.
26
-------
Prior to discussing particle properties, it is very important that two
terms frequently used in the emissions control area be correctly defined. The
term particulate is an adjective describing the state of dispersion of either
solid or liquid material, usually as an assemblage of small, discrete, solid
particles or liquid droplets. Hence, one should refer to particulate
material, particulate emissions, or to particulate states, but not to
"particulates" alone. The term aerosol, which is often incorrectly construed
to be synonymous with the term "particle" as in "aerosol size", actually
describes a suspension of particulate material in a gas just as the term
hydrosol refers to a stable liquid dispersion of suspended solids.
2.5.1 Particle Mass Concentration
The magnitude of the inlet dust concentration will determine how
frequently a filter must be cleaned for any constant gas flow system (1). In
the unusual case where inlet concentrations are very low, <0.1 g/ro
(0.05 grains/ft-^), it may be necessary to inject a precoating dust to
accelerate the formation of a dust layer so that collection efficiency can be
elevated to acceptable levels (45) and to prevent excessive penetration of
fine dust into the fabric interstices.
Particle sampling methodology suitable for the determination of collector
inlet concentrations is described in many publications (65-68). Insofar as
mass measurements are concerned, procedures recommended by EPA, such as
Methods 5 and 17, that are described in the Federal Register (69) are
prerequisite for compliance testing as well as for forming a rational basis
for acceptance or experimental testing. Major criteria to be satisfied by the
particulate sampling method may be summed up as follows:
• Any extractive sampling must be performed isokinetically
(1,22,68,70) so that the mass concentration and size properties of
the particulate material in the extracted gas volume are
representative of those in the parent gas stream.
• The quantity of collected dust must be determinable by gravimetric
(or appropriate indirect) means at some prescribed accuracy level,
e.g., within +O.lmg in the case of Methods 5 and 17 (69).
• Sampling must represent true average conditions with respect to duct
cross section while simultaneously taking into account any temporal
changes in gas flow or particulate composition. The required number
of sampling points and their precise locations, which vary from one
plant to another in accordance with individual design conditions,
are indicated in the Federal Register (69).
It should be noted that gravitational settlement or inertial separation
may actually lead to lower inlet concentrations and somewhat smaller particle
sizes at the filter face in contrast to those determined well upstream.
27
-------
2.5.2 Particle Size Distribution
Particle size properties affect filter performance in several ways.
First, particle collectability during the initial stages of filtration through
a new or "just-cleaned" portion of a fabric is strongly size-dependent, with
the smaller particles being the more difficult to capture (1-8). As dust
accumulates, the size-dependency factor diminishes rapidly for all practical
purposes (31,42). Second, the size and shape of the particles determine the
structure of the overlying dust cake with respect to its porosity and the
number and effective diameter of the parallel flow channels through which the
gas must pass. As particle size diminishes, the dimensions of the flow paths
through the cake also decrease such that classical aerodynamics dictates an
increased resistance to gas flow. Hence, a fine dust is invariably more
difficult to filter than a coarse one. Third, the fabric definability; i.e.,
the ease with which a dust deposit can be separated from a given fabric,
usually diminishes with decreasing particle size unless electrical charge or
agglomeration effects render the cake structure more porous.
The simplest explanation for the above behavior is that a fine dust
contains more particles not only per unit volume of dust but also per unit
cross section of dust cake surface such that the number of contacting points
between the dusty layer and the adjacent fabric surface is significantly
increased (48,49). Since the adhesive bonds between these particles and
fibers oppose dust dislodgetnent, the finer dust will require a greater energy
input to attain adequate cleaning. The fact that particle size plays a key
role in filtration requires that the best possible estimates of size
parameters be made prior to designing or diagnosing the performance of a
filter system.
Although a broad spectrum of particle sizing techniques and
instrumentation is discussed in the literature (71-74), the number of methods
available for practical field application is quite limited if one adopts the
same representativeness criteria stipulated for mass concentration methods.
2.5.2.1 Particle Sizing by Cascade Impactor—
Insofar as fabric filtration is concerned, an instrument referred to as a
cascade irapactor is well suited for determining the size properties of inlet
and outlet particles (75-80). Because of its importance, key design and
operator parameters for these and other sampling devise are discussed in
Appendix 1 along with documenting references describing calibrating procedures
and comparative performance for various impactor designs.
Cascade impactors permit the fractionation of inlet dust samples on the
basis of the mass (weight) fraction collected in each of several size ranges
represented by successive impactor stages. Frequently, the characteristic
size for particles captured by each stage is expressed as an aerodynamic
diameter; i.e., the equivalent size of the collected particle if its mass were
transformed to that for a unit density sphere. When the irapactor measurements
are grouped and plotted as a cumulative mass distribution on logarithmic
probability paper, one can ascertain at a glance the size properties of a
dust. Figure 5, which shows a typical mass distribution for a coal fly ash as
determined by cascade irapactor, indicates that 50 percent of the fly ash mass
28
-------
50.0 -
£
4.
2 5.0
o
o
K.
Ul
< 2.0
1.0
0.5
aMMD=5/xm
°g =3.0
j_
JL.
JL
5 10 20 30 50 70 90 95 98
PERCENT BY MASS EQUAL TO OR LESS THAN STATED SIZE
Figure 5. Mass distribution for typical coal fly ash, based on cascade
impactor sampling (see Table k) .
29
-------
is represented by particles less than 5 vim aerodynamic diameter and that
90 percent (5 to 95 percent) of the mass falls within the size range 0.8 to
30 wm diameter. As discussed in Section 2.5.3.2, the 50 percent size deduced
from Figure 5 is the aerodynamic mass median diameter, aMMD.
2.5.2.2 Particle Density (1,22,24-26,71-74)--
Except for liquid droplets and the unique spherical mineral fusion
products (cenospheres) found in coal fly ashes, most industrial aerosols
contain irregularly-shaped particles that often vary in discrete particle
density from one component to another (81,82). Therefore, what is retained on
an impactor stage is an assemblage of particles whose average size in the form
of unit density spheres is described by the characteristic stage diameter. If
the actual diameter is desired for a spherical particle of density other than
1.0, the following relationship derived from Equation (2-10) is used:
Dp(actual) =
where Pp refers to the true particle density in g/cnH.
2.5.2.3 Particle Shape (24-26,71-74) —
It should be noted that an irregularly formed particle tends to settle at
a slower rate than that for its equivalent sphere (1,22,24-26). Thus, for
capture by inertial impaction, the irregular particle behaves like a smaller,
and hence more dif ficult-to-capture particle. Except for the presence of
highly irregular particles such as asbestos fibers or micaceous platelets,
introduction of corrections for shape are not usually justified (48). As a
practical consideration, it has been found that an irregular particle,
although behaving aerodynamically as a smaller particle, is more apt to be
captured by a simple sieving or straining process unless its major axis is
aligned with the local direction of flow. Additionally, the inherent
roughness of nonspherical shapes enables a stabler interlocking of particles
such that the chance of aerodynamic dislodgement or migration is reduced (48).
2.5.3 Significance of Characterizing Particle Diameters (22,71-74,83-86)
2.5.3.1 Average or Mean Diameters —
Several characterizing diameters are used to describe specific attributes
of an assemblage of polydisperse particles, i.e., a mixture of several
particle sizes (22,85,86). The term raonodisperse is used to describe a
grouping of equal sized particles. In the following discussion, only those
size parameters bearing upon dust filterability are considered. Four types of
mean (average) diameters are commonly used as size descriptors (22,71,74).
1. D = number average particle diameter, the particle diameter which
multiplied by the total number of particles present will give the
sura of all the particle diameters in the sample.
2. Dg = surface area average particle diameter, the diameter of the
particle whose surface area multiplied by the total number of
particles present will give the total surface area of all the
particles in the sample.
30
-------
3. Dv = volume average particle diameter, the diameter of the
particle whose volume multiplied by the Lotal number of particles
present will give the total volume of all the particles in the
sample.
4. DV8 = Dv/Df = diameter of the particle with the same
ratio of volume to surface as that exhibited by all particles
present in the sample.
The D value is of interest when particle capture is by sieving or direct
interception, or when estimates of the pore dimensions within a dust layer or
the probable length of an agglomerate chain are the principal concerns. The
surface area average diameter, Ds, is more often associated with the surface-
related measurement of light scatter for sizing purposes, although it is also
important in the in situ absorption of SOX (or other gases) prior to capture
of their solid reaction products on a fabric. The volume diameter, Dy,
which is directly proportional to the mass diameter for any constant density
substance, relates mainly to collection processes where particle inertia plays
a major role. For example, particle capture during the early stages of
filtration and dust collection by cascade impaction are governed by the volume
or mass diameter.
2.5.3.2 Median Diameters —
There are certain applications where the median or "middle-most" values
for certain size parameters are very useful tools. Three such diameters are
commonly encountered:
1. ^nmd or CMD = nuraDer or count median diameter, that particle size
above which and below which there are an equal number of particles
2. Dsmd or SMD = surface area median diameter, that particle size
above which and below which half the total surface area or total
projected area of the distribution of particles is found.
•*• Dmmd or MMD = mass median diameter, that particle size above which
and below which half the total mass of the distribution of particles
is found .
2.5.4 Characterizing a Distribution of Particle Sizes (1,2,8,22,24-26,71-74,
83,84)
2.5.4.1 Graphical Representation —
The particulate content of most aerosol systems is rarely distributed in
accordance with a "normal" Gaussian distribution where all deviations from the
central tendency can be described by the classical symmetrical "bell" curve.
Except for highly truncated distributions, such as the limited range of
particle sizes falling between two classifying screens or two cascade impactor
stages, most particle size distribution curves are skewed such that the number
of particles per equal size increment, increase geometrically as the size
diminishes (see Figure 6, A).
31
-------
•
u
O
UJ
•3
a
UJ
a:
a.
PARTICLE DIAMETER, D
(A)
c
«
o
o
z
Ul
3
O
UJ
a:
u.
LOGARITHM PARTICLE DIAMETER, log D
(B)
Figure 6. Participate material graphed as normal
and logarithmic normal disuributions.
32
-------
The mechanical processes of crushing and grinding to produce a finer
product or the formation of a fume through condensation and/or agglomeration
typify processes expected to produce asymmetrical distributions.
In many cases, a simple change of variable technique, namely that of
substituting the logarithm of the particle size for the actual size, converts
the skewed distribution to a symmetrical Gaussian form (see Figure 6,B).
Special logarithmic probability graph papers are available that simplify the
plotting of linear parameters as shown in Figure 5. The above approach is
very useful because it applies equally well regardless of what size parameter
is being investigated. For example, the logarithmic-probability plots for the
cumulative number, surface area, and mass (or volume) distributions for a dust
having NMD and Og values of 0.5 ym and 2.5, respectively, appear as shown in
Figure 7. It should be noted that the slopes of the surface area and mass
distributions are the same as that for the number distribution since dg, the
geometric standard deviation, remains constant, regardless of the
characterizing median diameter.
The abscissa on Figure 7 indicates the percent of total mass that is
equal to or less than the size indicated on the vertical logarithmic scale.
The size corresponding to the 50th percentile is the characteristic median
diameter for the distribution. An attractive feature of this plotting method
is that by computing the ratio of the ~84 percent to the 50 percent size or
the 50 percent to the ~16 percent size, the distribution parameter or
geometric standard deviation, oe, is determined, i.e.,
O
84% size 50% size
(T 3 - _i—i __ Z5 —^^^_^_-^— - • i.
g 50% size 16% size
In contrast to the normal Gaussian distribution in which the standard
deviation bears the same unit as the attribute being measured, the change of
variable procedure, i.e., the substitution of the logarithm of size for the
actual dimension, results in the geometric standard deviation appearing as a
dimensionless ratio.
If any mean or median diameter and o are known, any other
characterizing diameter may be estimated Sy graphical means (74), Figure 8, or
calculated by means of the Hatch-Choate equations (87) Table 4. Despite the
versatility of the logarithmic-probability distribution, it must be remembered
that it is only as good as the fit of the actual data to the straight line
that should result if, in fact, the measured size characteristics were exactly
described by the log-normal approach.
Should a very poor curve fit result, one has recourse to many alternative
distribution functions that may provide much improved correlations (88-92).
The criterion for selecting one relationship in preference to another should
be the goodness of fit to experimental measurements. All techniques are the
same in principle; that is, the size is usually defined by D (or some function
of D) and one or two arbitrary constants that provide a measure of homogeneity
(range of sizes).
33
-------
30
20
10.0
5.0
oc
UJ
t-
U)
< 2.0
o
UJ
_i
o
flC
<
0.
1.0
0.5
0.2
84% SIZE
16% SIZE
GEOMETRIC STANDARD DEVIATION (
-------
10
5 10 20 50 100 200
RATIO Ds/Dnmd , Dv/Dnmd , Dsmd /Dnmd , Dmmd
500
1000
Figure 8. Graphical estimation of characteristic particle diameters (Ds , Dy, DSmd ,
and D ) from count median diameter
anc^ geometric standard deviation (a ) .
-------
TABLE 4. EQUATIONS FOR COMPUTING CHARACTERISTIC DIAMETERS FOR PARTICLE
SYSTEMS DESCRIBED BY A LOGARITHMIC-NORMAL DISTRIBUTION
Characteristic diameter
Equation
Surface Mean (Average)
Volume Mean (Average)
Surface Median
Dsmd
Mass Median
D.
log Ds = log Dnmd + 2.303 log2
og
ramd
log Dv = log Dnmd + 3.45 log2 ag
log Dsmd = log Dnmd + 4-60 log2 ag
log Dramd = log Dnmd + 6-90 log2 ag
NOTE: Base 10 Logarithm.
Rosin-Rammler,^2 Roller,^^ Nukiyama-Tanasawa,?2 Gaudin-Schulhmann,
Wynn-Dawes, Sichel, Kottler, Dalla Valle, and Wiebull.
36
-------
2.6 FABRIC PROPERTIES AND FILTRATION CONCEPTS (1,93-106)
2.6.1 Fabric Selection Criteria
One of the most important components of a fabric filter system is the
fabric. Although the dust retention characteristics of the fabric (which are
usually defined in terms of fractional or percent dust penetration for the
filter medium) are obviously important, the power requirements associated with
operating pressure losses and fabric cleaning procedures and the capital and
replacement costs for the fabric also play vital roles in fabric
choice (1, 93-95). Additionally, the optimum fabric geometry for a given
operation (tube, bag, or envelope), length-to-diameter ratios for bags or
tubes, maximum possible length versus method of bag cleaning, and matching of
fabric type and properties to the cleaning method and the working environment
are also very important.
Krause (94) has designated the following fabric properties as essential
to the proper function of any filter system, although not necessarily in the
order delineated in Table 5.
2.6.2 Woven Fabrics (1,93-99)
2.6.2.1 Construction Features—
Woven fabrics are produced on a loom by the entwining of yarns in a
variety of patterns in which the lengthwise yarns are referred to as warp
yarns and the cross yarns are called fill, weft, or woof yarns. While there
are a multiplicity of weaves, the most common types used for fabric filtration
are the twill and sateen weaves. Simple "over-and-under" plain weaves are
used occasionally.
The yarns, which constitute the fabric building blocks, are spun from
staple (short) fiber lengths or continuous (synthetic) filaments that are
twisted together to form threads. In practice, the final yarn or thread often
consists of several plies (or smaller threads) that have been twisted together
at varying twist levels. Staple yarns usually involve some 10 to 20 twists
per inch whereas continuous, multifilament yarns such as those used in woven
glass fabrics require fewer turns per inch to provide yarn strength and
stability.
The warp yarns are always aligned with the bag axis (top to bottom) to
provide maximum strength. In any combination of continuous multifilament and
staple, bulked, or textured fill yarns, the much stronger multifilament yarns
are always aligned from top to bottom. If both warp and fill yarns are woven
from staple, the yarns with the higher thread count should be aligned with the
bag axis.
Although the terms thread and yarn are synonymous, yarn usually refer to
the composite threads making up the fabric, whereas the term thread often
applies to the material used to stitch together various fabric shapes into the
finished product. The yarns in many woven fabrics may be composed of hundreds
of individual fibers twisted so tightly together that they are nearly
impervious to gas flow. Therefore, unless the actual yarn dimensions are
37
-------
TABLE 5. KEY FABRIC PROPERTIES
• Solids retention - Weave characteristics must allow
formation of a continuous surface dust layer, free of
defects such as cracks or pinholes while minimizing
internal plugging and blinding of pores.
• Clean fabric permeability - Airflow through the unused
(clean) fabric must take place at minimal resistance as
defined by Frazier permeability criteria (ASTM Standard
D-737) while permitting a solid dust cake to form.
• Cleanability - Fabric surface texture must allow for
sufficient dust release to allow for continuous filter
system operation at acceptable pressure loss.
• Resistant to mechanical wear, high temperature, damage,
and chemical attack.
38
-------
2 ?
known, Che weight of a fabric alone expressed in oz/yd or g/m may have
little bearing on the pore area or free fiber content and hence its pressure
loss or dust capture characteristics.
The explanation for the above statement is that pore diameter relates
roughly to yarn diameter, whereas the state of dispersion of the fibers, i.e.,
twisted tightly into multifilament yarns or freely dispersed as with a napped
surface, determines the dust capture and retention characteristics. Thread
count alone affords but a rough index of fabric pressure loss; nominal thread
(or yarn) diameter is again essential for any reliable estimates. Thread
counts, usually written in the form 68 x 48 (27 x 19) refer to the number of
warp and fill yarns per inch or centimeter of fabric width and length,
respectively. Definition of terras commonly used to describe textile and
filter fabric properties are given in Appendix C.
2.6.2.2 Fabric Performance versus Construction (1,96,97,99)—
It was pointed out previously that raultifilaraent yarns spun from
continuous filaments demonstrate poor dust capture properties unless the
distance between the yarns is less than ten times the diameter of the
particles to be collected. The above distance criterion depicts the
approximate gap that can be bridged by a particle chain. Because pressure
loss increases with yarn proximity, staple yarns or mixtures of multifilament
yarns with staple, texturized, or bulked yarns are used to provide fiber
obstacles within the pores (space between yarns). The same objective may be
accomplished by the production of a surface nap which provides the highly
effective free fiber cover characterizing napped, sateen weave cotton. The
overall result is to enhance the bridging capability of the fabric while
maintaining the fabric pressure loss per se at acceptable levels. Figure 9
shows typical warp (multifilament) yarns and fill (bulked) yarns used in twill
weave woven glass fabrics (100).
If a fabric is woven as a plain weave from two yarn types, both sides of
the fabric will present the same appearance. Furthermore, it makes no
difference as to which direction the gas stream approaches the fabric. With
twill weaves or other asymmetrical yarn patterns, however, one can expect
differences in fabric performance depending upon which side is the filtering
surface. For example, if the warp side is used with a 3/1 woven glass twill,
75 percent of the upstream surface will be made up of readily cleaned
continuous multifilament yarns. Therefore, the surface dust layer will be
dislodged more readily by any cleaning action which may afford a pressure loss
advantage. On the other hand, the greater ease of cleaning from which one may
infer a lower pressure loss can result in higher emissions because a larger
fraction of more penetrable (cleaned) surface will be exposed when filtration
is resumed (100). The same effect is observed if too severe a cleaning action
is used.
2.6.2.3 Fabric Design Limitations (93,94,96,99)—
Although it might appear advantageous to design a fabric with a reduced
thread count and a high free fiber content to provide both low fabric pressure
loss and high particle collection, the actual application of such media could
present some problems. First, an increase in yarn separation distance
increases the chance for slippage under stress conditions such that nonuniform
39
-------
Figure 9. Continuous multifilament yarn (a) and
bulked yarn (b) , glass fabrics (100).
-------
yarn spacing ensues. The resulting increased gas flow rate through these low
resistance paths, whose free fiber support is weakened by the yarn separation,
can lead to "blowout" or open pore formation followed by excessive particulate
emissions (96,97,100). If special fiber coatings such as Teflon, silicone
lubricants, or graphite are used to lubricate fibers to reduce abrading during
tensile or lateral stressing, or to minimize chemical attack from corrosive
aerosol components, the possibility of nonuniform yarn spacing is likely.
This problem has been observed with common woven glass fabrics.
Yarn strength is essential to prevent excess fabric stretching and
permanent deformation while filtering and especially during certain cleaning
operations involving vigorous mechanical action. Otherwise, there is no
advantage associated with the use of multifilament or tightly twisted staple
yarns because far less free fiber collecting surface is available. One way to
obtain more effective filtration is to use nonwoven media such as naturally
felted or needle-punched felts to supply a maximum of collecting surface with
a minimal fabric weight (1). The following discussion reviews the principal
factors that must be considered for the successful application of felted
fabrics.
2.6.3 Nonwoven Fabrics (93,98,101-104,106,107)
The advent of several synthetic fibers, many of which are highly
resistant to damage by high temperatures and/or chemical corrosion, has been
particularly advantageous in the development of felted media used in high
capacity, pulse jet-cleaned collectors. Felted fabrics possess the marked
advantage of providing a high, free fiber surface that permits clean, unused
felt to function at relatively high efficiencies while still offering less
resistance to air flow than most woven fabrics.
2.6.3.1 Construction Features - Felted Fabrics—
Natural wool felts prepared from animal fibers are inherently strong
because of the scaly surface projections that assist the interlocking of the
randomly dispersed fibers (1). In most filtration applications, felts are
manufactured at areal densities ranging from 14 to 20 oz/yd , (475 to
675 g/ra ) roughly twice the weight of many woven fabrics. Because they are
not composed of continuous yarns, they are more apt to deform when subjected
to repeated rigorous stretching during cleaning by mechanical
shaking (1,107). Additionally, the increased fabric weight also contributes
to increased stiffness that will attenuate transmission of the shaking action
over the full length of a bag. Therefore, effective utilization of felts has
necessitated the development of different cleaning processes, reverse jet
cleaning or pulse jet cleaning, to provide a continuous filtration capacity.
Although both cleaning approaches allow for increased filtration velocities,
pulse jet cleaning and its modifications are by far the most common methods
for cleaning felted fabrics. In the case of pulse jet cleaning (107), the
stiffness of the fabric does not inhibit effectiveness so long as bag lengths
do not exceed 10 to 12 ft (3 to 3.7 m) and the volume of pulse jet air used
is roughly scaled to the volume of the bag. Unfortunately, there are very
definite upper temperature restrictions as well as corrosion factors
associated with wool felts that usually restrict their use to near ambient
41
-------
temperatures, and always to the range below 65° to 95°C. Techniques used to
clean filters and how they relate to fabric properties are discussed in detail
in Section 3.
2.6.3.2 Advantages of Felted Media—
One advantage of felted media is that the increased porosity of the
randomly dispersed fibers permits higher filtration velocities, nominally 6 to
10 ft/rain (1.8 to 3 m/min) versus 2 to 3 ft/min (0.6 to 0.9 m/min) for most
woven fabrics, while displaying minimal pressure loss increases for the fabric
itself (1,93,95,98,107). This particular feature has led to the adoption of
felted bags or tubes for use in pulse jet filter systems. The higher pressure
loss that might be expected with increased filtration rate (and increased
fabric loading) is avoided by more frequent and more intense cleaning plus the
fact that the high ( 400 to 1000 g/m2) residual dust layers on felted
fabrics are characteristically more porous than those deposited on woven
fabrics (107).
2.6.3.3 Needled Felts (1,98,99,107)—
In order to take advantage of synthetic fiber capability to function in
difficult environments, it is necessary to compensate for the characteristic
weakness of the unbonded, felted staple fibers that results from the general
smoothness (and often slipperiness) of many synthetic fibers. One popular
technique is to first prepare a sandwich consisting of an open, plain weave
fabric composed of multifilament yarns (referred to as a scrim) and covered on
each side with a layer of loosely felted fiber. Then by means of a needle
punching process whereby specially barbed needles cause the surface felt
fibers to intertnesh and interlock with the scrim, a simulated felt is produced
whose superficial physical properties are very similar to those of natural
felts. Despite the fact that this needling creates regions of fiber alignment
normal to the fabric surface at the needle locations, comparative evaluations
of pressure loss and dust penetration behavior for both natural and needle-
punched felts have indicated similar performance (107). From a structural
viewpoint, the needle-punched felt permits the effective utilization of
reverse jet or pulse jet cleaning without the risk of mechanical damage to the
fabric.
2.6.4 Fabric Selection - General Comments (1,94,99,107)
Aside from the fact that the method of cleaning and the choice of fabric
are frequently interrelated, there are usually several options available to
the filter user that will satisfy his specific gas cleaning requirements. The
ultimate target, in addition to meeting emission regulations for specific
effluents, is to achieve the necessary degree of cleaning by the least
expensive means (93,95,108-112). This will involve not only capital cost
considerations, space requirements, and ancillary equipment used for waste
dust handling and disposal, but also routine operating costs associated with
fabric and parts replacement, power costs related to gas handling and fabric
cleaning procedures, and most importantly, the maintenance regimen required to
provide maximum filter system availability at the proper control level. To
assist the filter user and inspection personnel, a listing of several common
fabrics and their performance levels and suitability in difficult working
environments (high temperatures and/or corrosive gases) is shown in Table 6.
42
-------
TABLE 6. PHYSICAL AND CHEMICAL PROPERTIES OF FIBERS USED IN FILTER FABRICS
Fabric
Poiypro-
pyiene
Temp.
liait-
S.G.a at ions Tneraal properties
0.92 180"F Similar to other polyesters.
Heat degradation character-
Chemical resistance
Excellent resistance to roost acids
and alkalies*
Cost
factor
Other properties (Class * 1. 01
Virtually unaffected by aging.
Excellent abrasion resistance.
Cotton
Noroex
Nylon
Orion3
(Acry-
lic)
istics are essentially
the same under dry and
moist conditions.
1.50 180°F Ignites readily.
Burns steadily and is not
self-extinguishing.
1.38 iOO°F Does not support combus-
tion.
Decomposes at 700°F.
1.38 2-0'F Moderatly difficult to ig-
nite; self-extinfjishing.
Does not support spread of
flame after removal of
igniting source.
Considered flameproof.
1.18 275^ Burns freely and rapidly.
Fiber continues to burn
after it is removed
from flame.
Fiber fuses and may shrink
away from flame; forms dark
Irregularly shaped bead.
Will deteriorate in continued contact
with concentrated nitric acid and
certain oxidizing agents.
Disintegrated by hot or cold concen-
trated acids.
Resistant to weak alkalies and organic
solvents.
Very similar to nylon in chemical
properties.
Inferior to nylon in alkali atmos-
pheres at elevated temperatures.
Hydrolyzed by concentrated acids.
Loss of strength from hot weak acids.
Virtually no effects from strong or
weak alkalies.
Unaffected by dry cleaning solvents;
soluble in phenol and formic acid.
Stable in strong or weak acids.
Attacked by strong alkalies.
Attacked slightly by weak alkalies.
Unaffected by common organic solvents.
Sleekness of fibers provide
good cake release.
Virtually unaffected by aging.
Moderately susceptible to the
influence of microorganisms-
Like nylon, excellent resistance
to abrasion, mildew, and aging.
Virtually unaffected by aging.
Resistant to the influence of
microorganisms.
Because of unusually high energy
absorption, good in abrasion,
flex life, impact loading, and
dimensional stability.
Virtually unaffected by aging.
Resistant to the influence of
microorganisms.
2.78
1.0
(continued)
-------
TABLE 6 (continued)
Fabric
Dacron*
(Poly-
ester)
Temp.
1 irai t-
S,C.a ations
1.38 275°F
Thermal properties
Burns slowly when ignited;
usually self-extinguish-
ing when removed from
igniting flame.
Chemical resistance
Dissolves with partial decomposition
in concentrated sulfuric acid.
Resistant to most mineral acids.
Cost
f ac tor
Other properties (Class - 1.0)
Virtually unaffected by aging.
Kesistant to the influence of
otic roorganisms.
Teflon* 2.10 500°F
Glass
(uncoat-
ed)
2.54
550°F
Similar to nylon; fuses
and shrinks away from the
flame; forms round bead
after cooling; hard to
crush between fingers.
Smoke dark and contains
soot.
Excellent resistance to dry
and moist heat.
Excellent resistance to dry
and moist heat.
Moderate resistance to strong alka-
lies at room temperature.
Disintegrates in strong alkalies
at boiling temperature.
Good resistance to weak alkalies.
Generally insoluble in organic sol-
vents; soluble in some phenolic
compounds.
Excellent resistance to mineral and
organic acids, alkalies, oxidizing
agents, and solvents.
Excellent resistance to mineral and
organic acids, oxidizing agents,
and solvents.
Physiologically inert; no pois-
onous or other harmful effects
when used internally or applied
externally.
Fair abrasion resistance.
Poor abrasion resistance.
3.8 (woven)
12.9 (felted)
1.0
Glass
(coated)
Teflon
2.40 --500°F
Excellent resistance to dry
and moist heat.
Poor resistance to alkalies.
Excellent resistance to mineral and
organic acids, oxidizing agents,
and solvents.
Poor abrasion resistance.
M.O
Moderate resistance to alkalies.
Glass 2.40 -v-500°F
(coated)
SilicoDe-
Graphite
Excellent resistance to dry
and moist heat.
Excellent resistance to mineral and
organic acids, oxidizing agents,
and solvents.
Moderate resistance to alkalies.
Poor abrasion resistance.
M.O
"Specific Gravity.
-------
2.6.5 Basic Filtration Concepts - Woven Fabrics (1-4,93,94,98,113-120)
2.6.5.1 Total Pressure Loss (113-120) —
The total pressure loss, P, across any woven fabric filter is the
algebraic sum of the loss through the fabric itself, Pf, (the latter serving
mainly as the support for the dusty layer) and the loss through the dust
layer, Pj, that provides the high efficiency collection capability.
P = Pf + Pd (2-16)
The fabrics are usually sewn in the form of bags or tubes although envelope
shaped designs are used occasionally. In most instances, the direction of
filtration is from the inside to the outside of the bag, although the
direction can be reversed if wire cages are used to prevent bag collapse.
Because the air flow through fabric filters is always in the stream-line
(laminar) range, pressure loss increases may be expected to vary linearly with
the filtration velocity, V, and the areal density of the surface dust layer,
W, i.e., P
-------
where S is called the total filter drag and Sg refers to the "effective
residual drag" for the cleaned fabric surface (113). The fabric loading
accumulated over the time interval, At, may be expressed as:
W = C£VAt (g/m2) (2-19)
The symbols and the units (metric and English) customarily used with the
variables identified in this section appear in Table 7.
2.6.5.2 Graphical Estimation of Sg and K2—
Figure 10 shows a typical filter drag versus fabric loading (S-W) curve
for a woven fabric when filtration has been performed on a "completely
cleaned" surface (120). Completely cleaned means that the entire surface dust
loading, W - WR, has been removed from the section of fabric undergoing
evaluation. The term WR refers to the residual fabric loading, i.e., the
dust that remains more or less permanently embedded within the loose fiber
structure of a filter following any normal field cleaning operation. Were
dust-free air to be passed through the filter at velocity, V, the actual
measured residual fabric drag would appear as SR. For many practical filter
operations, however, computation processes are greatly simplified without
introduction of serious error if one extrapolates the linear (right hand)
portion of the drag curve to the fabric loading level corresponding to WR.
The intercept at this point becomes the effective residual drag, Sg,
appearing in Equation (2-18). The fact that Sg usually exceeds SR means
that any errors made in pressure loss estimation will be on the safe (high)
side. The specific resistance coefficient, Ko, for the dust-fabric
combination of interest is the slope of the linear portion of the S-W curve.
Although several theoretical methods have been proposed for calculating
K2 and Sg, the direct measurement approach is the only safe estimating
procedure (46,49,100,120). Unfortunately, one is almost always compelled to
perform such K£ tests as single bag or single panel laboratory measurements
with the constraint that the bags or test panels be completely cleaned before
evaluations. If the test aerosol is generated by some high energy dispersion
technique, it is necessary to ascertain that the particle size properties of
the test aerosol provide a fair simulation of field conditions. Probable
ranges for K2 and Sg for various dust-fabric combinations are given in
Table 8.
2.6.5.3 Pressure Loss Characteristics and Filter Performance (120,121)—
Inspection of Figure 10 shows that a high value for Ko also signals a
potentially high filter pressure loss. Additionally, the filter drag SE at
the resumption of air flow is an indicator of the lowest possible pressure
loss that can be expected with conventional cleaning techniques for a given
dust-fabric combination and a specified filtration velocity. (100,107,121)
If the pressure loss increases rapidly due to a high K2 value, more
frequent cleaning will be needed to keep the total system pressure loss within
the static pressure capabilities of the induced draft fan and the failure
limits (collapse, or buckling) of the baghouse per se. Because the dust
penetration properties for the filter system depend upon the thickness of the
dust layer, frequent cleaning to reduce pressure loss can adversely affect its
dust retention characteristics (1). Additionally, complete fabric cleaning;
46
-------
TABLE 7. SPECIAL NOMENCLATURE, ENGLISH AND METRIC EQUIVALENCIES FOR KEY FILTRATION PARAMETERS
Filter resistance
Filter drag
Velocity
Volume flow
Fabric area
Areal density
Specific resistance
Symbol
P
S
V
Q
A
W
K2
Metric
N/m2
N min/ra3
m/rain
m-'/min
m2
g/m2
N min/g-m
Units
English
in. H£0
in. H20 min/ft
ft /rain
f t3/min
ft2
lb/ft2
in. «20 min ft/lb
Equivalency
249 N/m2 = 1 in
817 N min/m3 =
0.305 m/min = 1
0.0283 m3/min =
0.093 m2 = 1 ft
4,882 g/m2 = 1
0.167 min/g-m =
. water
1 in. water
f t/min
1 ft3/min
2
lb/ft2
1 in. H20
min/ft
min ft/lb
coefficient
Dust concentration
g/m:
grains/ft3
2.29 g/m3 = 1 grain/ft3
-------
K2»S-SE/W-WR
CO
cT
(C
O
(T
Ul
00
WR
S = FILTER AT TIME t
SE= EFFECTIVE RESIDUAL DRAG AT t=0
SR= ACTUAL RESfOUAL DRAG AT t=0
W= FABRIC LOADING AT TIME t
WR* RESIDUAL LOADING FOR CLEANED FABRIC
I
FABRIC LOADING, W
W
Figure 10. Typical filter drag versus fabric loading curve (100).
-------
TABLE 8. PROBABLE RANGES FOR K2 AND SE, BASED ON LABORATORY TESTS (100)
Fabric
Woven Glass
3/1 Twill
Woven Glass
3/1 Twill
Woven Glass
3/1 Twill
Woven Glass
3/1 Twill
Woven Glass
3/1 Twill
Woven Glass
3/1 Twill
Napped Sateen
Weave Cotton
Napped Sateen
Weave Cotton
Napped Sateen
Weave Cotton
Dust
Coal fly ash
Coal fly ash
Coal fly ash
Lignite
Granite
(Rhyolite)
Granite
(Rhyolite)
Fly ash
Fly ash
Talc
Siz
MMd
(Mm)
4.2
6.4
11.3
8.8
9.2
1.2
4.2
2.4
2.8
:e
og
2.4
3.3
3.6
2.5
4.8
2.4
2.4
1.77
2.9
C" 1 1 t- r a t~ T f\ n
r 11 cracion
velocity
m/min
0.92
0.61
0.85
0.61
0.61
0.61
0.92
0.92
0.92
K2 Sg
N-min/g-m N-min/mJ
1.85 115
1.40 352
0.75 150
1.34 270
1.38 205
12.3
1.85 49
1.77
4.7
Measured at 21°C, 760 mm Hg.
-------
i.e., 100 percent dislodgement of the surface dust layer, is never realized
with mechanical shaking or reverse flow cleaning systems. Therefore, when air
flow resumes following cleaning, the low resistance "just cleaned" regions
will display high transient velocities (up to 10 times the average
value) (121). This phenomenon further aggravates the high emission problem
caused by frequent cleaning and reduced dust layer thickness (areal density).
2.6.5.4 Effect of Partial Cleaning on Filter Drag-Fabric Loading
Relationships (46,49,100)—
The primary reason for not being able to utilize most field data for K2
estimates is that the fraction of filter surfaces completely cleaned normally
ranges from roughly 5 to 50 percent for most commercial applications. The
lower levels of cleaning, ~5 to 35 percent, apply to filter fabrics cleaned by
bag collapse and reverse flow, whereas the 50 percent level is achievable with
vigorous mechanical shaking (100,107). In either case, the bag compartment
undergoing cleaning is isolated from the on-line compartments by appropriate
dampers. In the case of relatively new bags or those that have not been
affected by moisture condensation and/or chemical reaction, the dust usually
separates at the dust/fabric interface where adhesive bonds are weakest (46).
After cleaning, the fabric surface presents a patchy appearance in which the
cleaned areas appear as the brightly illuminated regions in Figure 11 (46).
Conversely, the black areas are regions of the fabric surface from which no
dust has dislodged as a result of the cleaning action.
When filtration is resumed with a partially cleaned surface, (see
Figure 12), gas velocity and dust deposition rates differ widely, with the
greatest gas flow occurring through the "just cleaned" region. As filtration
progresses, flow rates and dust distribution eventually approach the uniform
level typified by a "completely" cleaned fabric. Until equilibration has
again been achieved, however, the characteristic slopes of the curves for
various levels of partial cleaning no longer permit a direct estimate of the
true K£ value for the dust. Thus, such curves should never be used to
estimate K2 for purposes of predicting filter performance (46,100,120).
During troubleshooting operating, it is important to note that most
baghouse pressure loss indicators reflect the average pressure loss across
several compartments operating in parallel. Hence, these data, just as those
presented in Figure 12, cannot be used to estimate dust K2 values. The
reason again is that the dust loadings are not uniformly distributed from
compartment to compartment nor across the surface of individual bags.
2.6.5.5 Nonlinearities in Drag-Fabric Loading Curves (120)—
Although the S-W curve shown in Figure 12 has been depicted as
essentially linear over the upper ~75 percent of its range, the reader is
cautioned that nonlinearities often occur with many dust-fabric combinations
despite a uniformly distributed fabric dust loading. If the curvatures are
minor, an eyeball or a least squares force fit to a straight line will entail
little error. It is important, however, for purposes of interpretation, to
recognize the reasons for nonlinearity. Figure 13 shows various S-W curve
shapes encountered during field and laboratory studies. If the surface of a
filter is nearly planar as it is for membrane filters used for particulate
sampling or for special fabrics having a fluorocarbon or Teflon film laminated
50
-------
Figure 11.
Cleaned bag with illumination from inside by
fluorescent tube (46).
51
-------
Ul
to
en
a:
a
cc
UJ
K
UJ
O
<
a:
UJ
DESCRIPTION
MAXIMUM POSSIBLE CLEANING
HIGHLY EFFICIENT CLEANING
AVERAGE CLEANING RANGE
(MECHANICAL SHAKING)
AVERAGE CLEANING RANGE
COLLAPSE WITH REVERSE
FLOW
0 W
AVERAGE FABRIC LOADING,W
Figure 12. Typical drag versus loading curves for filters with different degrees of cleaning and a
maximum allowable level for terminal drag, ST, and terminal fabric loading, W (100).
-------
ro
i
O
E
I
s
3X I0~3
0.6
Figure 13. Approximate drag versus fabric loading, coal fiy ash
filtration at 0.61 m/min filtration velocity.
-------
to a woven fabric, the individual pressure losses for Che filter medium and
the dust layer are completely independent. With the further constraint that
the dust layer (or filter medium) undergo no compression as the dust loading
accumulates, the S-W relationship, Curve 1, will be linear.
In the case of fabrics woven from a blend of continuous multifilament and
bulked yarns, the depressions and gaps between the yarns must be filled before
a true surface dust layer develops. (46,100,120,121) The filling of these
depressions causes a variable and more rapid pressure increase at the
inception of filtration because both the depth of deposit per unit mass of
dust and the gas velocity through the deposit are greater than those
encountered after a substantial surface layer has formed. This explains in
large part the initial concave-down shape found for many S-W curves, as shown
in Figure 13, Curves 2 and 3 (46).
The fact that Curve 2 evolves into a linear form suggests no compression
of either the fabric or dust layer over the remaining filtration period.
Since the dust itself (Curve 1) is noncompressible, the upward curvature shown
in Curve 3 is thought to result partly from fabric compression before
increasing interstitial dust deposits have reduced its compressibility
(Curve 2) (115). Filtration with fabrics having a dense and
uniformly-distributed, free fiber cover may show a concave-up S-W curve due to
compression of the spring-like nap (Curve 4). Fabrics composed entirely of
continuous, multifilament yarns or those with few free fibers to obstruct the
pores often lose their dust retentivity as their fabric loadings
increase (46). Despite pore bridging during the early filtration stage,
increasing aerodynamic drag progressively destroys the bridging as dust layer
thickness (and pressure gradient) increases. The above phenomenon, sometimes
referred to as pore collapse (114), is characterized by a consistent reduction
in slope (Curve 5) and a very significant increase in dust penetration.
2.6.5.6 Effect of Velocity and Fabric Loading on K2 (115,117,119)—
In some cases, nonlinearity in S-W curves may be taken into account by
treating K2 in Equation (2-18) as a variable. It has been determined, for
example, that the velocity at which certain dusts deposit upon a fabric will
affect the porosity of the dust cake and, hence, its K2
value (114,120,121). It appears that a denser (and less porous) dust layer
results as particles impact at increasingly higher velocities. Presently, the
general relationship:
K2 « Vn (2-20)
with n ranging from 0.2 to 1.0, appears to satisfy K2 variability based upon
tests with Arizona road dust, mica (2), and coal fly ash (46,100,114,121).
If, because of particle shape or surface properties, an increasing
pressure loss caused by an increased fabric loading results in gradual
compression of the dust layer and/or the supporting fabric, K2 again may be
treated as variable, with the general expression:
K2 « Wn (2-21)
54
-------
serving to describe K2 variability (49). Hence, K2 as used in
Equation (2-18) may be treated as a constant or expressed as some function of
V and/or W as suggested by field or laboratory behavior. For example, when
filtering coal fly ash, the equation:
K2 = 1.8 V0-5 (N min/m3), (2-22)
where V is expressed in m/min, indicates K2 variability with filtration
velocity (122).
2.6.5.7 Theoretical Estimation of K2—
Although the subject of calculating K2 values has been treated
extensively on the basis of fluid mechanics for both fibrous and granular
beds, no complete theory has yet evolved. For estimating purposes, a
relationship often referred to as the Kozeny-Carman Equation affords crude
estimates of K2 for granular beds (dust cakes) composed of near-spherical
particles when particle shape, chemical species, and density are
size-independent and interfering effects such as fluid boundary layer
thickness, electrical charge, density, and gas adsorption sites may be
ignored (1).
K2 = kP ^| (1 - e)/P e3 (2-23)
P
or
K2 = k U S§ (1 - e)/Ppe3 (2-24)
The terms appearing in Equations (2-23) and (2-24) are defined as follows:
k = Kozeny-Carman constant, usually stated to be 5.0. For estimating
purposes, a value of 1.6 is more realistic for coal fly ash (46).
U = gas viscosity, poise
d = particle diameter with a monodisperse system, cm
o
S0 = ratio of particle surface to particle volume, cm
c = dust cake porosity, dimensionless
f'p = particle density, g/ctn
The porosity or fraction void volume, e, within the dust cake is defined
by term 1 - (P/Pp) where P is the previously indicated bulk density and Pp
is the discrete particle density. More sophisticated relationships have been
proposed for those cases where the dust cake porosity exceeds 0.80.
Regardless of which equations are used to compute K2, however, the accuracy
levels are too low to provide reliable data inputs for the solution of real
filtration problems. On the other hand, the relationships indicated in
Equations (2-23) and (2-24) do provide excellent qualitative guidelines for
predicting K2 variability.
55
-------
Limited comparisons have indicated that a correction factor of roughly
1/3 should be introduced in both Equations (2-23) and (2-24) to account for
the fact that the average calculated K2 values are three times the actual
measured values (46, 48). Thus, an effective value of 1.6 should be assigned
to the Carman-Kozeny constant, k. This correction has been included in the
computer program described in References 46, 48, 49 and 100 for the estimation
of performance parameters for woven fabric filters.
Since real aerosols are polydisperse, i.e., they include a broad range of
sizes, Equation (2-24) containing the specific surface parameter So is the
better choice to calculate K2« Use of the particle size parameters obtained
from Andersen impactor measurements, HMD (actual value) and og allows for
ready estimation of SQ (46,48).
1.151 log2 og
S = 6 — — (2-25)
o MMD
2.7 FABRIC CLEANING - WOVEN AND NONWOVEN FABRICS (1,3,7,28,29,107,121-131)
Fabric filtration is effective only when the filter can be cleaned
periodically and economically without impairing collection efficiency or
disturbing the system gas flow. Highlights of recent studies on filter
cleaning by (a) mechanical shaking, (b) reverse flow, (c) combinations of (a)
and (b), and (d) pulse jet cleaning are described next. In the following
discussions, fabric cleaning by various techniques is examined in terms of the
physical process of dust removal, including the amount of energy required and
the relationships among the degree of cleaning accomplished, energy inputs,
and dust adhesion to the fabric.
2.7.1 Mechanical Shaking (1,107,122)
In a simple shaking system, the oscillation of the shaker arm alternately
accelerates and decelerates the dust-laden bag surfaces. The resulting
tensile and/or shearing forces exerted at the fabric-dust layer interface, if
greater than local adhesive forces (1,46,48,132-135), will remove layers or
flakes of dust from the fabric as shown in Figure 11 (121) that rapidly
disintegrate to form a broad size spectrum of agglomerated and discrete
particles (107,121). Although the shaking action is usually applied at the
top of a bag, a bottom connection is sometimes seen in field
installations (1). Typical mechanical linkages for bag shaking as well as bag
configurations are shown in Figures 14 through 16. Further data on cleaning
mechanisms for various baghouse designs are given in Appendix D while
suppliers of fabrics and/or baghouses are listed in Appendix E.
The separating force, Fg (assuming that tensile and shear forces are
roughly equivalent), can be estimated from the dust loading, W, and the
average acceleration, a, imparted to the dust-laden fabric (46,48).
Fs = W I (2-26)
56
-------
SHAKING LEVER
''•''"'"•'' ECCENTRIC
BAGS
DUST OUT
Figure 14. Typical shaking mechanism (courtesy of the Mclllvaine Co.)
57
-------
• ME* NUT
E W/
TOP "7"ct>
S.S.
DUS7UBE BOTTOM
DETML 1
Figure 15. Typical bag attachment system for a shaker baghouse (courtesy
of Wheelabrator-Frye, Inc.—Drawing No. SD 1227-1).
58
-------
Figure 16. Cutaway view of a shaker baghouse (courtesy of
MikroPul Corp.—excerpt from Bulletin TSQ-1).
59
-------
The acceleration is computed from shaker arm amplitude (displacement from
center position) and shaking frequency, Equation (2-27). Field and laboratory
tests have indicated that average acceleration must be at least 3 g's to
transmit the shaking motion to the entire bag. Low frequencies (<4 cps) and
small amplitudes (<1 era) generate acceleration forces appreciably less than
that attainable in a gravity field (46,48). It is assumed that at the instant
of dust dislodgement, the local adhesive force, F^, is only slightly less
than the separating force, Fg.
The following expression is proposed for estimating the average
acceleration, a, imparted to a shaken bag when the top of the bag is
oscillated in an essentially horizontal motion by the shaker arm (46,48,49).
af = 0.0282f2Y = g's (2-27)
In Equation (2-27), a is defined in g's, f is the shaking frequency in
cycles per second (Hertz), and Y the maximum amplitude (displacement from
center position) of the top of the bag in centimeters.
2.7.2 Bag Collapse With Reverse Flow - Woven Fabrics (48,49,121)
Bag collapse accompanied by intermittent application of low velocity,
reverse air flow (roughly the same as the normal filtration flow) is a
preferred method of cleaning glass fabrics because it avoids the stresses
caused by mechanical shaking. Here the cleaning principle is essentially the
same as that for shaking except that dislodgement is produced by the shearing
force created by the dust mass adhering to the fabric and the downward pull of
the gravity field, g. Thus, the separation force is defined by the product
W.g rather than W.a (46,48). For direct comparison with the mechanical
shaking relationships,
Fs = W.g (2-28)
Except for unusually high reverse air rates (~3ra/min), the principal function
of the reverse air appears to be the flushing out of fine dust particles
removed and/or loosened during the cleaning process. The rate at which the
bag flexes during the collapse step and the final curvature assumed by the bag
depending upon the applied tensioning may also influence dust removal,
although no quantitative data are yet available to measure such
effects (123). In general appearance, the bag configurations, methods of bag
attachment to the compartment tube sheet, and overhead suspensions are similar
to those used with mechanical shaking systems. Commercial baghouse designs
are shown in Figures 17 and 18, and various suspension and bag tensioning
approaches are illustrated in Figure 19.
2.7.3 Pulse Jet Cleaning - Felted Media (1.107,124-131)
Despite the sharing of many design and operating features with the
previously discussed reverse flow (RF) or mechanically cleaned (MC) systems,
including the general arrangement of filter tubes as shown in Figures 20
and 21, several unique aspects of pulse jet-cleaned felts warrant separate
treatment. First, from the design perspective, pulse jet filters are intended
60
-------
Cross Sectional View of WP Custom High Temperature Ba
-------
MikroPul
Reverse Air
Collector
(a)
Figure 18. Reverse-air baghouse (a), butterfly reverse air valves (b),
bag suspension system (c), and tube sheet connection (d).
(Courtesy of MikroPul Corp.—excerpt from Bulletin TSQ-1).
62
-------
(b)
(c)
(d)
TUBE
SHEET
BAG
BAG
/"CLAMP
Figure 18 (continued)
63
-------
SPRING RETAINER
SPRING COMPRESSION
COMPRESSION LATCH
CHANNEL WASHER
BAG SUPPORT CHANNEL
J" HOOK
ANTi ^
COLLAPSE RINGS
IN CLAMPING BAG
CLAMP SHOULD BE
TIGHTENED AS CLOSE
AS POSSIBLE UNDER THE
TOP THIMBLE RING.
METAL BAG TOP
WITH EYE BOLT
*-LOCK CLAMP
•THIMBLE
-TUBE SHEET
Hanger Designs
Cham Design
Threaded Rod Design
Thimble Design
Grid (tube)
Sheet
Snap Ring
Sewn In Bag
Thimble
Extension
(minimum 1
bag diameter)
Bottom or top mounted thimbles are available with
Zurn Custom-Engineered Fabric Filter Systems.
Figure 19. Typical bag suspension systems for reverse-air baghouses (courtesy of Zurn
Industries, Inc., Air Systems Division—excerpts from reverse-air baghouse
operation and maintenance manual and Form No. 422-ADV 5M) ,
-------
Figure 20. Cutaway view of an Alpha Series Pulse-Jet Baghouse
(courtesy of Standard Havens, Inc.)-
65
-------
HOW IT WORKS
Dust-laden air enters the collector
hopper where the inlet classifier
uniformly distributes the air through
the entire length of the collector
and separates the larger dust
particles. Fine particles are borne
upward around the classifier by the
airstream and pass around the filter
bags. The remaining dust is de-
posited on the exterior surface of
the filter bags, and the clean gas
passes through the bags to the
outlet manifold.
Once a uniform layer of dust has
been deposited on the filter bag
surface, it is removed on a pre-
determined cycle by a high pressure
pulse of compressed air. In each
cycle only one row of filter bags is
pulsed at a time; the frequency and
duration of the pulse is determined
by actual field conditions. The
ENELCO venturi provides for a
two-stage cleaning effect. The
venturi shapes the initial pulse of air
to provide an air mass which moves
down the filter bag tube, expanding
the bag and dislodging the dust
from the outer surface. The motion of
the compressed air draws some
secondary plenum air behind it.
adding to the downward air wave.
The downward flow helps propel
dust to the hopper and provides for
backwashing of the bags to remove
deeply imbedded dust.
Figure 21. Schematic of pulse-jet baghouse and compressed air header
(courtesy of Environmental Elements Corp.—excerpt from
Form 973 DV 15M).
66
-------
to operate at higher filtration velocities, ~3 m/min (10 fpm) than those used
with woven fabrics, ~1 m/min (3 fpm) and with a narrower range of operating
pressure loss. The resulting advantages are a decreased space requirement and
less variation in gas flow, although higher operating costs and lower
efficiencies may appear on the debit side (133). Second, although fabric
cleaning is again accomplished by the separating (tensile) force produced by
fabric and dust acceleration, the intensity, and, in particular, the brevity
of the cleaning process (~0.1 second) preclude direct use of the mathematical
relationships proposed for woven fabric RF or MC systems (128,132).
Because of the stiffness of most felts, conventional mechanical shaking
is insufficient to create the bag motion needed to clean the
fabric. (1,107,128) Additionally, the typical 3- to 4-fold increase in
filtration velocity requires very frequent and intense cleaning to remove the
collected dust. In most cases, not only would mechanical shaking fail to keep
up with the dust deposition, but the felted medium itself would also be
subject to damage due to repeated stretching action (107).
2.7.3.1 Qualitative Description of Pulse Jet Filtration—
The filter medium ordinarily consists of an open top felt tube supported
by an interior wire cage so that outside-to-inside air flow is possible
without tube collapse (see Figure 22) (1,4,7,28,107,127-132). With this
arrangement, dust deposition takes place on the outside rather than the inside
fabric surface as found with the previously discussed MC and RF systems. When
the pressure gradient across a compartment housing several tubes reaches some
preset limit, the tubes undergo sequential cleaning either on a "one tube at a
time basis" or on a "row by row basis" wherein several tubes are cleaned
(pulsed) simultaneously. In some cases, a simple timing circuit is used to
activate the cleaning.
The cleaning process consists of the introduction of a high pressure
(~90 psig) reverse air pulse at the top (exit) end of the tube. Because the
pulse duration is very brief (~0.1 second), there is no "effective" isolation
of tubes or compartments such as that encountered with woven fabrics where
off-line times may be as much as 5 minutes (1,107,126,132) . Rapid tube
inflation accelerates the tube and dust in a radial direction which causes the
dislodgement of surface dust deposits in the form of coarse agglomerates.
Although the mass of the dislodged dust may be large, only a very small
fraction (~1 to 10 percent) actually descends to the dust hopper because gas
flow resumes in about 0.1 second (107,128,130-132). At steady state
operation, the amount of dust falling to the dust hopper per tube will be
equal to the amount of dust deposited on the tube during the preceding
filtration period (time between pulses). What distinguishes the pulse jet
cleaning process from simple reverse flow or mechanical cleaning is that most
dust dislodged by the jet pulse is returned to the felt surface; all dust
removed from woven fabric RF and MC systems has sufficient settling time (2 to
5 minutes) to fall to the dust hopper (107). Additionally, the K2 values
for successive dust deposits will remain the same regardless of cake depth,
although the cake depth itself may vary from point to point (107,128,132).
67
-------
© ©
MATERIAL DISCHARGE
FILTER BAG
RETAINER
BAG CLAMP
TUBE SHEET
VENTURI
BLOWTUBE
ORIFICE
TIMER, REMOTELY LOCATED
SOLENOID VALVE IN WIRING TROUGH
DIAPHRAGM VALVE
COMPRESSED AIR MANIFOLD
COLLECTOR HOUSING
INLET
HOPPER
AIRLOCK
UPPER PLENUM
EXHAUST
MANOMETER
DIFFUSER
COMPRESSED AIR
INDUCED FLOW © SUPPLY AT 100 P.S.I.G.
SECTION A-A
SCHEMATIC OF MIKRO
PULSAIRE COLLECTOR
Figure 22. Schematic of Mikro-Pulsaire dry dust collector and
operating components (courtesy of MikroPul Corp.—
excerpt from Owners Manual SD 379A).
68
-------
In the case of pulse jet filters, 90 to 99 percent of the dislodged dust
redeposits as a uniformly distributed layer, but with a significantly lower
K.2 value than that of a fresh dust deposit because of its highly
agglomerated state (128,132). The specific resistance for the redeposited
layer, has been assigned the symbol (K2)c (128). Although the jet pulse
has the capability to dislodge the dust deposits, there is only sufficient
energy transfer to accomplish a partial breakup of the dust such that the
suspension consists mainly of agglomerates (128-132). Therefore, when
filtration resumes for periods varying from 1 to several minutes, a new layer
of dust deposits on top of the underlying or "cycling" layer, Wc. The new
layer displays the characteristics of the freshly deposited dust and hence a
higher K2 value because it is composed mainly of discrete rather than
agglomerated particles.
2.7.3.2 Estimation of Pressure Loss - Theory—
Since the above concepts provided a rational description of the pulse jet
filtration process, it has been proposed that the overall filter pressure
loss, P, be computed as shown below (128).
P = PE + (K2)c V Wc + K2 V W (2-29)
Equation (2-29) states that overall filter pressure loss for a single tube
over the filtration interval t is the algebraic sura of the contribution from
the conditioned felt, Pg; the resistance offered by the cycling layer, Wc,
with its low (K2)c value; and the pressure loss due to the most recent or
overlying dust deposit with its normal K2 value. The term W is expressed
by the same relationship used for woven fabric RF or MC systems (1,107); i.e.,
W = CiAV t, (2-30)
where At represents the filtration interval between pulses.
The first and the third right hand members of Equation (2-29) are
identical to those appearing in Equation (2-17) for woven fabric filters
whereas the second term is peculiar to pulse jet systems. It is expected that
the pressure loss contributed by the cleaned felt, PE, will increase
gradually with time as more and more dust becomes irreversibly embedded in the
fiber structure (1,107,128,132). Except for special experimental
measurements, the quantity Wc is usually undefined and only estimated values
of (K2)c have been reported (128). Thus, Equation (2-29), although
constituting a logical model for pulse jet filtration, cannot as yet be used
for predictive purposes. However, because the dust layer is projected from
the fabric as a result of the compressed air pulse, one can infer that the
interaction between tensile and adhesive forces (134-136) again determines the
amount of fabric cleaning attained. Additionally, it is also reasonable to
assume that the greater the intensity of the compressed air pulse, (as defined
by its pressure level and volume delivery) the more effective the
cleaning (126-132).
69
-------
2.7.3.3 Kquilibrium Pressure Losses—
Reference to Equation (2-29) shows that for a fixed inlet velocity and
dust concentration, any reduction in pressure loss must depend upon
corresponding reductions in the terms Wc and W. Qualitatively, increased
pulse pressures, P-, lead to decreased Wc values (up to a point), whereas
W can be decreased only by reducing the interval between successive jet
pulses (unless inlet concentration and/or filtration velocity are reduced).
Based upon laboratory observations, it appears that the quantity (Wc + W)
is uniquely determined by the adhesive properties of the specific dust-fabric
combination and the cleaning energy imparted by the jet pulse (107,128,132).
For this reason, a pulse jet system can operate within relatively narrow
pressure loss boundaries for a fixed cleaning regimen despite significant
variations in inlet dust loading. Although pressure loss necessarily
increases with velocity, the Wc portion of the total removable fabric
loading does not vary significantly unless the reservoir (and pulse) pressure,
PJ, change, see Table 9 (107).
2.7.4 Dust Removal versus Cleaning Method
At the present time, there are no theoretical means to determine how much
dust will be dislodged during cleaning despite the fact that rational (if not
accurate) estimates of the separating forces can be made. The principal
difficulty is that again there is no method short of direct measurement to
determine the adhesive forces at the interface between the dust layer and the
fabric surface (134-136). As stated earlier, cleaning of woven fabrics is
associated with a characteristic separation of the dust layer at the
dust-fabric interface where adhesive bonds are normally the
weakest (1,46,48,107). The resulting "cleaned areas" (see Figure 12), possess
nearly identical "effective" residual drag properties, a fact that enables a
realistic physical description of field filtration processes, provided that
the degree of cleaning defined by the parameter, ac, can be related to the
cleaning energy input (46,48,49).
2.7.4.1 Fractional Cleaning, ac, with Mechanical Shaking—
Limited experimental data are now available for specific dust/fabric
combinations based upon bench scale and pilot measurements that show the
relationship between the cleaning parameter, ac, and fabric dust loading or
the estimated adhesive force (see Figure 23) (46,48,49). The horizontal axis
shows the actual, uniformly-distributed fabric loadings at the start of
cleaning. The upper scale represents the computed adhesive force based on the
estimated average acceleration for the indicated dust/fabric system. Note
that fractional cleaned area, ac, also describes that fraction of the fabric
surface where the adhesive forces must be smaller than the separating forces
produced by the cleaning action.
The development of fractional cleaning concepts discussed here is based
upon the assumption that dust dislodgement takes place as an interfacial
separation. Although laboratory and limited field observations have confirmed
the above phenomenon, it is important to note that long-term, moisture
condensation coupled with chemical reactions may alter the "ideal" separation
process. Recent field observations suggest that dust may also spall off as a
superficial layer leaving part of the surface layer behind. (138) Protruding
fibers or excess nap also contribute to incomplete cleaning.
70
-------
TABLE 9. SUMMARY DATA - EFFECT OF RESERVOIR PRESSURE, FACE VELOCITY,
AND INLET FLY ASH CONCENTRATION ON RESIDUAL FABRIC LOADING
AND AVERAGE PRESSURE LOSS3
Reservoir
pressure
psig
(kPa)
40 (276)
70 (483)
100 (690)
C - 2.
V = 2.
Fabric
dust
loading
g/m2
405
366
373
0 g/m3
6 m/min
Pressure
loss,
average-
kPa
1.620
1.250
0.750
C = 26
V = 2.
Fabric
dust
loading
g/m2
549
389
382
.5 g/m3
6 m/min
Pressure
loss,
average-
kPa
1.600
1.330
1.230
C =
V =
Fabric
dust
loading
g/m2
503
423
439
22.7 g/m3
1.89 m/min
Pressure
loss,
average-
kPa
1.170 (1.620)b
1.100 (1.520)b
0.950 (1.300)b
aData excerpted from Tables 22 and 25, Reference (107)
^Pressure loss adjusted to 2.6 m/min with same fabric loading.
NOTE: Pulse duration = 0.06 sec, pulse interval = 1 min (damped pulse).
71
-------
Ld
>
to
UJ
__
- KQ
^s
< ._
uj ui -5
£t K<
< < X
_
UJ Q
O fe (O
OQ
Z
go
O CD
-------
When the fabric is cleaned by mechanical shaking, the cleaning parameter
for a coal fly ash/woven glass fabric system may be estimated from the
following equation (46,48,49,107):
ac = 2.23 x 10~12 (f2AsWp)2'52 (2-31)
"* 7
The term Wp (g/m ), which appears in both Equations (2-31) and
(2-32), designates the average fabric loading immediately before initiation of
cleaning. In most field applications, it is usually impossible to determine
the magnitude of W _ by any direct measurement. However, it is possible
to calculate ac by a mathematical process carried out within the predictive
model described in References 46, 48, 49, 100 and highlighted in Appendix F,
provided that the following information is available: (a) baghouse design and
operating parameters usually defined in the course of routine compliance or
acceptance tests, and (b) estimated values for K£ and Sg.
2.7.4.2 Fractional Cleaning with Bag Collapse and Reverse Flow—
For coal fly ash/woven glass fabric systems cleaned by collapse and
reverse flow, the cleaning parameter, ac, may be computed from
Equation (2-32) (48),
ac = 1.51 X 10-8Wp, (2-32)
where Wp is the average fabric dust loading (g/m ) at the time when
cleaning is initiated. Reference to Figure 23 indicates that fly ash
particles ~7 ym MMD show much greater adhesion with a cotton fabric because of
the larger number of free fibers (nap) at the dust/fabric interface (107). In
the case of napped cotton, 100 percent of the fabric surface is covered with a
loose fiber nap, whereas glass fabrics show about a 75 percent free fiber
surface coverage. As a result, about 10 percent surface removal was attained
with cotton as compared to nearly 50 percent with a woven glass fabric when
the fabric loading was 1,000 g/m2 (see Figure 23). Because cotton fabric
can withstand vigorous shaking, however, adequate cleaning is still attainable
despite the increased adhesion factor.
As stated earlier, collapse and reverse flow cleaning is generally less
effective than mechanical shaking (~5 to 35 versus ~50 percent with the latter
method). On the other hand, the former approach permits the use of heat
resistant glass fabrics that might otherwise fail under vigorous shaking
conditions.
2.7.4.3 Effective Filter Cleaning by Pulse Jet Air—
It has been observed that the high energy shocks imparted to a felted
fabric are sufficient to completely remove the overlying or surface dust layer
with a properly designed system (107). Therefore, the cleaning parameter,
ac, although not used for modeling pulse jet filtration, would automatically
assume a value of one (128,132). On the other hand, any working equation for
predicting pressure losses for a pulse jet system must take into account the
redeposition of most of the dislodged dust that results in a unique recycling
layer, Wc, with its unique specific resistance coefficient, (K2)c
(128,132). Because there now exists no way to compute these terms directly,
filter pressure loss must be estimated by a combination of empirical
relationships based upon limited experimental data.
73
-------
The single bag test results shown in Figures 24 and 25 provide useful
guidelines for fly ash and talc filtration with a needled Dacron
felt (107,132). The terra "damped pulse" applies to an experimental
modification of commercially available equipment that permits a gradual rather
than an abrupt deflation of the tube when the pulse flow ceases
(107,129,132). The advantage of this procedure is that elimination of the
"snap" as the tube returns to its normal filtration position greatly reduces
(up to 5 times) the outlet dust concentration.
The information obtainable from Figures 24 and 25 may be used to develop
a modified form of Equation (2-29) that allows the filter user to extrapolate
from one set of operating pressures to another. First, the terms Pg and
(K2)CVWC may be consolidated to describe the "effective" residual
pressure loss, (Pg^AW at the cessation of the jet pulse (and at the
resumption of filtration). Hence total pressure loss, P, is then expressible
by the following equation (128,132):
P = (PE^AW + K2V2CiAt (2-33)
The slopes of the curves reflect the freshly deposited K2 values for the two
dusts at the designated filtration velocity while their vertical displacements
result from differences in the intensity and type of the pressure pulses. It
is emphasized, however, that any working relationship deriving from Figures 24
and 25 such as Equation (2-33) is restricted to 4 ft long (1.22 m) by
4 1/2 in. diameter (11.4 cm) bags and a jet nozzle with a true inside diameter
of 0.363 in. (0.92 cm). In a subsequent paper, however, Dennis et al (132).
have proposed a set of general equations that take into account both the
volume of compressed air pumped into a bag over a specified time interval and
the volume of the bag itself. The latter data enable computation of the rate
at which the bag pressure increases, d p/dt, a parameter necessary to define
the energy input to the dust-laden bag. The term d(Ap)/dt, (kPa/s) is defined
by the relationship:
d(Ap)/dt = 211 Pja Aj/Vb (2-34)
where Pja is the absolute pulse jet reservoir pressure (kPa), Aj is the
true jet nozzle cross section (m ), and Vj, the volume of the filter bag
(m ). Ambient sea level pressure and temperature are assumed along with a
nominal 0.01 second opening time for the solenoid valves customarily used in
pulse jet equipment. Related analyses have shown that the residual effective
pressure (?„) , (kPa) can also be estimated from Equation (2-35), the
Aw
latter based on data graphed in Figure 26.
(PBs = 1,600 [d(Ap)/dt] > (2-35)
t;AW
Therefore, the total filter pressure loss, P, in kPa based upon a combination
of Equations (2-33) through (2-35) is calculated as:
P = 1,600 [211 pja Aj/Vb]'1'13 + 0.001 K2V2 t (2-36)
74
-------
1.6 -
1.4 -
1.2 -
'-0
o
a.
S 0.8
O
~J
0.6
0.4
0.2
PULSE FREQUENCY = I minute
PULSE DURATION = 0.06 second
A
Q
SYMBOL TEST NO.
O
cy
A
o
!»,
38
54
5S
55
63
64
•M-
PULSE PRESSURE
psig
D IOO
D 70
40
D SO
D 70
D 701
1"
2.68
2.20
2.3I
2.37
2.86
2.84
* D INDICATES DAMPED PULSE
t ELEVATED JET (LESS EFFECTIVE)
Tt AVERAGE K2=2.53 N-min/g-m
I .)! 1
0.2 0.4 0.6 0.8
FILTRATION TIME , min.
1.0
Figure 24. Resistance characteristics and K2 values for fly ash and
Dacron felt, inlet concentration 27.6 g/m, filtration
velocity 2.6 m/min.
75
-------
1.6
1.4
1.2
1.0
o
a.
j*
w 0.8
CO
o
-J
ttJ
ac
Jo 0.6
en
UJ
{£
a.
0.4
0.2
n
PULSE FREQUENCY = 1 MINUTE
PULSE DURAT
ION = 0.06 SECOND
(2) PULSE PRESSURE
~ SYMBOL TEST NO. ps\q K2N-m in/g -m"*"
O 85 D 40 13.0
A 8
8 40 12.4
D 86 40 12.4
* 82 D 70 1 1.72
-
•» D INDICATES DAMPED PULSE
•f AVERAGE
_
K9 = I2.4 N-min/g-m __^~ -O"^
0-^-0^
n--^O A_^A—
^.^•V A-— r\-~"
o--'0 A^— A-^a-^n
^--O _— A~~ Jj-^-
- - -0" A^-A^-D-^ — _-H
-^--" o __-A— :^-n ^ _,^-^
:""A^^P--
"V
-
1
A— ^ __-—
u ^-*- —
-X '
1 1 I . .1
0.2 0.4 0-6
FILTRATION TIME , min.
0.8
1.0
Figure 25. Resistance characteristics and K.2 values for talc with
dacron felt, inlet concentration 3.5 g/m^, filtration
velocity 2.6 m/min.
76
-------
_J
Z>
SF S /mm
t, C3 / S 8 ! S f 1
J5s PULSE
Figure 26. Effective residual pressure loss versus d(Ap)/dt»
0 to 0.01 S. Fly ash filtration at 2.6 m/min.
77
-------
where 1<2 is the best estimate of the specific resistance coefficient for the
freshly deposited dust (N-min/g-m), V the filtration velocity (m/min), and At
the filtration time between pulses per bag (min).
Equation (2-36) will provide good pressure loss estimates for single bag
Dacron and wool filters used in pulse jet systems so long as the dust is coal
fly ash (132). If a different dust is filtered and/or if it is suspected that
the tensile and flexure properties of the felt differ appreciably from those
of Dacron or wool, one set of filtration measurements based upon properly
defined cleaning and operating conditions will be required to determine the
correct value for the dimensional constant in Equation (2-35) (which is 1,600
for a typical coal fly ash).
Since all commercial pulse jet systems are designed as multi-bag,
sequentially cleaned units, the average system pressure loss will fall between
the minimum and maximum values predicted by Equation (2-36); i.e., the values
at the start and the end of the filtering interval. Experimental measurements
have indicated that the far right hand term, which reflects the pressure loss
due to freshly deposited dust, should be reduced to 0.0075 K2V2At to
estimate the average operating pressure loss for a multibag system.
2.8 FILTER PERFORMANCE
2.8.1 Filter Drag - Woven Fabrics Cleaned by Mechanical Shaking or Bag
Collapse with Reverse Flow (48,107,128,132)
Equations (2-17) and (2-18) afford the means to determine what pressure
loss increase may be expected during the dust loading phase of the filtration
process, provided that all parts of the fabric surface offer the same
resistance to air flow and the face velocity and inlet dust concentration
remain constant over the filtration interval. For a complete analysis of
filter system operations, however it is necessary to determine how much dust
is removed by the selected cleaning process.
It was pointed out, for example, that conventional woven fabric cleaning
techniques remove anywhere from 5 to 50 percent of the surface dust
layer (42,46,49,107). As a result, resumption of filtration produced a family
of S-W curves shown in Figure 12 where very striking differences in filter
drag were indicated for various degrees of partial cleaning despite the fact
that average fabric loadings were identical. It has also been determined that
all cleaned filter regions (bright areas in Figure 11) possess the same
residual and effective drag characteristics, Sr and Sg, and that the
uncleaned regions represent the state of the loaded fabric surface just before
cleaning (46,48,49). Therefore, the. instantaneous drag values for the overall
filter, S, at the resumption of filtration on a partially cleaned surface are
readily calculated by the following relationship (1,46,48,107,158-160):
-1
(2-37)
where ac and au indicate the cleaned and uncleaned fractions of the fabric
surface, respectively; Sc is the drag for the cleaned areas that may also be
approximated by Sg; and Su is the drag for the uncleaned areas
78
-------
(46.48,49). As pointed out earlier, the relationship implied by
Equation (2-37) breaks down, if because of difficult cleaning conditions, the
dust layers no longer separate at the fabric-dust cake interface region but at
some point above the fabric surface. Equation (2-37) depicts the well
established flow principle that, given alternative flow channels (the
uncleaned and cleaned areas), the fluid automatically seeks the path of least
resistance. Thus, the flow imbalance decreases with extended filtration until
near-uniform velocity and fabric loading distributions are reestablished.
If the concept represented by Equation (2-37) is extended to a real
filter system wherein, immediately following cleaning, there exists a specific
fraction of just-cleaned surface, ac , as well as several additional surface
regions possessing varying degrees of dust cover, a , a ...a ,
the changes in overall filter drag for a multi-compartment system may then be
described by the general relationship (46,48):
a \ -1
u.
Y -/- + —...— 1 (2-38)
£~t S S SI
1 - ul ui
Since the mechanisms have already been established for calculating drag values
for uniformly loaded fabric surfaces provided that K2 and Sg have been
determined, Equation (2-38) with the aid of a computer program becomes a
workable tool for calculation of overall filter system drag.
The several relationships involved in the computation of filter system
drag or pressure loss (in contrast to the simple equation used for single
panel tests) apply equally well to any woven fabric provided that the K2 and
Sg values for the dust of interest have been determined. The basic
equations discussed in this section have been incorporated in the predictive
filter model described in References 46, 48, 49, 100 and discussed briefly in
Appendix F in a workable format for use by the field engineer in assessing
(a) the feasibility of proposed operating procedures, or (b) the likelihood
that existing equipment and operating modes will enable compliance with
emissions regulations.
2.8.2 Dust Penetration - Woven Fabrics (1-8,28-31,47-49,94,95,104,107,121,
125,128-132)
2.8.2.1 Mass Emissions Rates—
In a previous section, it was pointed out that there is no easily applied
theory for predicting the efficiency of a fabric filter despite the fact that
the fundamentals of filtration are well understood. The performance of a
fabric filter is governed by (a) the solidity and depth of the overlying dust
layer (the latter constituting the true filtration medium), and (b) the
physical properties of the underlying fabric that serves as the substrate for
dust deposition (1,46-49,107).
79
-------
If the dust cake is free from cracks, pinholes, or other defects, theory
suggests that particle collection should be 100 percent for all practical
purposes (42,46-49,107). However, fabric imperfections, whether they result
from normal weave characteristics or from improper use or maintenance, can
lead to cracks and/or perforations (pinhole leaks) in the dust layer.
2.8.2.2 Effect of Pinhole Leaks on Effluent Mass Concentration and Size
Properties—
As stated in Section 2.2.6, any gas passing through pinholes in the
fabric and/or the overlying dust layer carries with it most of its particulate
loading (46-49). The reason for such penetration is that the actual pores are
so large (~50 to 100 urn in the case of woven glass fabrics) that no
diffusional capture is possible while the flow convergence in the leak region
generates inertial forces that result in the capture of only the larger
particles in the gas stream (>15 to 20 um diameter). Since dust penetration
through pinhole leaks may constitute 99 percent or more of total emissions
with woven glass or similar fabrics, the size properties of many filter
effluents can be characterized by those of the upstream particles.
In the case of the more efficient fabrics like sateen weave cotton, some
reduction in particle size may result during the filtration process (47).
However, the low order emission caused by gradual seepage of dust through a
fabric will contribute small but measurable emissions even in the case of
cotton, roughly 0.05 mg/m^ as compared to 0.5 mg/m^ with woven glass
fabrics (46,48). At the present time, there is no reliable method for
relating the particle size parameters of the upstream and downstream dust,
regardless of the method of filtration. Despite attempts to define
"characteristic" fractional particle size efficiencies for fabric filters, it
is strongly advised that except for the "perfect replicate operation", no
apparent fractional size parameters based upon measurements at one plant
should be used to define effluent properties at another plant (150,153).
2.8.2.3 Effect of Filtration Velocity on Particulate Emissions—
There are few published guidelines for estimating the effect of
filtration velocity on mass emission rates. Single particle-single fiber
(collector) theory indicates that increasing velocity enhances particle
capture by inertial irapaction and decreases collection by diffusion
mechanisms. Therefore, on a basis of mass efficiency, the fairly coarse
particle sizes (>1 ym) found in most utility or industry-generated aerosols
should be collected more efficiently at higher velocities.
However, because the particulate emissions from many fabric filters
result from direct penetration through unbridged pores or gradual seepage, the
effect of increased filtration velocity more often is to retard the closure of
any openings in the fabric due to aerodynamic reentrainment and to accelerate
the seepage rate, i.e., the gradual migration of dust through the fabric
substrate to the clean air side (46-48,107,129,131,150).
Measured velocity effects for coal fly ash/glass fabric systems are shown
in Figure 27 (46-48). It has been determined that outlet concentrations are
so highly sensitive to filtration velocity that a 200-fold increase may occur
as filtration velocities increase from 0.4 to 3.5 m/min (46,48). Since the
80
-------
3.35 m/min
1.52 m/min
0.61 m/min
0.39 m/min
INLET CONC., g/m3
4O 60 80 100
FABRIC LOADING (W), g/m2
140
Figure 27. Observed outlet concentrations for bench scale tests
with coal fly ash and woven glass fabrics as a function
of fabric dust loading and filtration velocity.
81
-------
above findings appear to typify field experience, care should be taken when
increased filtration velocities are proposed to attain additional gas handling
capacity without increasing overall collector size.
2.8.2.4 Effect of Inlet Dust Concentration and Fabric Loading on Particulate
Emissions (1,48,107,151) —
In those cases where minimal defects appear in a dust cake structure,
changes in inlet dust concentration should have little effect on outlet
values. Fabrics such as sateen weave cotton, for example, are characterized
by effluent concentrations that are governed mainly by seepage and are nearly
independent of inlet concentration levels (151). Conversely, those fabrics
more susceptible to pinhole formation, such as woven glass media, produce
effluents that result mainly from direct dust penetration through pinhole
openings (46-48). Hence, any change in the inlet concentration is reflected
in the amount of aerosol penetrating the filter as shown in Figure 28 (48).
During the early fabric loading stage when pore bridging has just begun, a
nearly direct correlation has been observed. As dust cake repair increases,
however, outlet concentrations show less dependency on inlet concentrations,
roughly a twofold increase for a tenfold increase in inlet concentration.
Reference to Figures 27 and 28 suggests that the higher emission levels are
associated with the lower fabric loadings.
2.8.2.5 Estimation of Overall Dust Penetration—
It has been demonstrated that outlet dust concentrations from fabric
filters depend upon several factors (48). In a general functional
relationship, the dust penetration, Pn, or outlet concentration, Co, may be
expressed as:
Pn or C0« (4>,Ci,W,V,CR), (2-39)
where 4> is a parameter related to the dust/fabric combination of interest,
G£ is the inlet concentration, W the fabric dust loading, V the filtration
velocity, and CR a characteristic minimum emission level for the dust/fabric
AX f*
combination of concern. The latter CR values have been cited as 0.5 mg/raj
and 0.05 mg/nr', respectively, for fly ash filtration on woven glass and
cotton fabric (46,48,49).
Although, 41, CJT, and C^ may be treated as constants for most filter
applications, both w and V will change constantly with tnulticompartmented,
sequentially cleaned systems. Hence, the prediction of emission levels, as
with pressure loss requires a computer program to solve the complex equations
involved. The average filter system penetration at some time, t, for a system
consisting of I compartments and J areal subdivisions per bag is also
determined by successive iterations in accordance with the general
summation (48):
I J
T* Pn.. V. . (2-40)
82
-------
20
40 6O 80 IOO
FABRIC LOADING (W),9/mZ
120
140
Figure 28. Effect of inlet concentration on predicted outlet
concentrations at a face velocity of 0.61 m/min
with coal fly ash and glass fabrics (bench scale
tests).
83
-------
The modeling procedure described in References 46, 48, 49, 100 and discussed
Ln Appendix F, which quantitates the basic relationships shown in
Equations (2-37) and (2-38), will permit reasonable estimates of filter system
performance for dusts whose general properties are similar to those of coal
fly ash when collected upon woven glass fabrics (46,48). If the fabric
constitutes a better filter than glass, e.g., cotton sateen, the model output
should be on the safe side with respect to particulate emissions.
2.8.3 Dust Penetration - Pulse Jet Cleaned, Nonwoven Fabrics
2.8.3.1 Mass Emissions (1-6,28,29,33,107,129-132) —
The loose and relatively uniform fiber dispersion of felts provides a
high base efficiency, ~90 to 95 percent, for new and cleaned fabric media and
an excellent subtstrate for development of a solid dust cake (107).
Unfortunately, very frequent cleaning of a felted fabric by pulse jet air
means that a small fraction of each bag's filtration cycle entails filtration
not only through a just cleaned surface, but also through a surface that has
been subjected to a transient dilation of fabric as it snaps back and impacts
upon the supporting cage (1,107,125,128,131). The net effect is to increase
the nominal penetration values for many wool felts to a level somewhere
between those observed for woven glass fabrics and cotton sateen weaves,
~99.9+ to 99.99+ percent. Because a large fraction of the effluent may
consist of dislodged agglomerates as well as inlet particles that have
penetrated directly through the fabric, the concept of variability in particle
size efficiencies in accordance with the fundamental capture mechanisms
(inertial impaction, direct interception, and diffusion) no longer applies.
In fact, it is not uncommon to see a pulse jet filter actually functioning as
an agglomerating device (167) with the effluent particles actually larger than
those approaching the fabric (see Figure 29).
2.8.3.2 Factors Affecting Effluent Size Properties—
As in the case of woven fabric systems, there are no viable working
relationships for predicting the overall or fractional particle size
efficiencies for pulse jet filters. Generally, increasing jet pressures
appear to increase effluent concentrations across the complete size spectrum,
although some interesting maxima conditions appear for the coarser particles,
(see Figure 30) (121,125,130,132). In all cases, pulse damping (107) leads to
greatly diminished outlet concentrations, suggesting that this concept should
be made an integral part of all pulse jet systems. According to the data
presented in Table 10, increasing pulse frequency shows the expected increase
in effluent concentration for all particle sizes for both "direct" and
"damped" pulses (107,132). The complex nature of the variability in effluent
quality exhibited by these measurements suggests that the accurate estimation
of effluent characteristics for pulse jet filters is beyond present predictive
capabilities.
2.8.4 Field Performance
The literature describing the field vehavior of fabric filter systems is
extensive such that only a small sampling can be presented here. Typical
baghouse performance for a wide variety of operating conditions and including
those situations where high temperatures, corrosion factors, and usually fine
84
-------
10.0
5.0
E
•»
o:
ui
i-
LJ
o
UJ
_J
o
o:
s
i.o
ANDERSEN IMPACTOR
A
O
OUTLET DUST
INLET DUST
SINGLE BAG TESTS
483 kPa (TOpsig) PULSE
OJ5.ec DURATION
I pulse/min
I 2 5 10 30 50
PERCENT MASS < STATED SIZE
70
Figure 29. Coal fly ash filtration, dacron felt with pulse jet
cleaning, 99.83% overall mass efficiency.
85
-------
u.m
PULSE
TYPE
DIRECT
DIRECT
DIRECT
DAMPED
DAMPED
DAMPED
I .
10
Figure 30.
Apparent .fractional penetration for coal fly ash
versus pulse pressure and type, dacron needled bag
(1.22 m x 11.4 cm).
86
-------
TABLE 10. EFFECT OF PULSE DURATION AND FREQUENCY, AND DIRECT AN!)
DAMPED PULSES ON FLY ASH FRACTIONAL PENETRATION, DACRON
NEEDLED FELT AT 2.6 m/rain AIR-CLOTH RATIO (107,150)
Particle Pulse
diameter duration*
pm sec
2-3 0.
0.
3-5 0.
0.
5-10 0.
0.
2-3 0.
0.
3-5 0.
0.
5-10 0.
0.
15
06
15
06
15
06
15
06
15
06
15
06
Fraction
pulse
penetration
frequency"!"
2.5 pulse/min 1.0 pulse/min
(ppm) (ppm)
6
4
2
1.3
4.6
9.0
2
0.5
2.7
0.7
8.3
8.3
x
X
X
X
X
X
X
X
X
X
X
X
DIRECT
10-5
10-5
10~4
10~4
10~4
10~4
DAMPED
10-5
10-5
10-5
10-5
10~6
10"6
PULSES
1.
2.
1.
0.
1.
5.
PULSES
1.
0.
0.
0.
2.
5.
5 x
0 x
3 x
8 x
6 x
0 x
05 x
3 x
95 x
6 x
5 x
0 x
10-5
10-5
io-4
10~4
10~4
10~4
10-5
10-5
10-5
10-5
10~6
10~6
Penetration
ratio
2.5 ppm/1.0 ppm
4.
2.
1.
1.
2.
1.
1.
1.
2.
1-
3.
1.
0
0
5
6
9
8
9
6
8
2
3
7
*Reservoir pressure = 483 kPa (70 psig).
Inlet fly ash concentration ^20-25 g/ra3 (^10 grains/ft3)
87
-------
or "sticky" dust can present difficulties is discussed in References
(1,3,4,7,139-141). More specialized and often challenging applications
encountered in the electric utilities area (both coal- and oil-fired boilers)
are described in the following References (1,3,4,142-149). Uses in
metallurgical processes, asbestos manufacture, and the asphalt concrete
industry are discussed in References (154-157). Technical data treating the
controversial area of fractional particle size efficiencies for fabric filters
are discussed in References 150-153.
88
-------
CHAPTER 3
TYPICAL BAGHOUSE PROBLEMS/CORRECTIVE ACTIONS
3.1 INTRODUCTION (1,3,7,93,94,95,137,138,161-166)
This section of the manual discusses typical problems reported for
baghouse installations. A general description of potentially troublesome
areas is followed by a listing of specific problems reported by fabric filter
users. Finally, troubleshooting guidelines recommended by several
manufacturers are presented (167-172).
Many of the problems encountered by fabric filter users arise from
ancillary equipment malfunctions that have nothing to do with the collection
device itself (1,7,94,162,163). Problems with fans, blowers, gas transport
and distribution systems, dust handling and removal systems, and monitors and
alarms are typical examples. Operational problems related to the baghouse per
se most often relate to the bags themselves or the bag cleaning system
(1,7,163,164). Other problems may occur as a result of transient or process
upset conditions affecting dust concentration and its physical properties or
gas stream temperature, humidity, and chemical composition. Figure 31 shows
schematically the principal baghouse components and ancillary equipment used
for the filtration of coal fly ash (170).
3.2 ANCILLARY EQUIPMENT
3.2.1 Fans and Blowers
Fans and blowers handling dust—laden exhaust gas or cleaned effluent can
present a variety of problems (94,95,161,166). The proper choice of
construction materials, fans, and motors in conjunction with careful
installation will minimize motor overloading, bearing wear, and motor damage.
Typical problem areas include excessive fan vibration, worn bearings, and
malfunctioning belts, drives, gears, and electrical controls.
The gas transport and distribution system, consisting of exhaust
ventilation hoods and all connecting ductwork between the dust sources and the
baghouse, requires careful design to ensure satisfactory operation. Improper
selection of construction materials can result in duct abrasion or chemical
attack. Incorrect duct sizing leads to unwanted settlement if the diameters
are too large or excessive abrasion and pressure loss if the ducting is
undersized (1,7,165,166). The actual shape and degree of enclosure of exhaust
hooding have an important bearing on baghouse operation with respect to dust
entrainment and volume of gas arriving at the baghouse. Furthermore, if the
baghouse is unable to accommodate the inlet dust loadings, its gas handling
capacity may not satisfy the exhaust requirements at the various exhaust
points. Other system components that can malfunction are compartment
isolation valves and emergency bypass dampers. In the former case, partial
valve closure leads to inadequate cleaning, while a faulty emergency bypass
damper can cause irreversible damage to the filter medium (95,161-165).
89
-------
vO
O
BAGHOUSC
7 VALVE
REVERSE AIR
DAMPER
REVERSE AIR FAN
1
rn
1
v
1
JL
•
• •'.
t
1
n
.* •
•/
.'*
i
n
*» •
•
•
/
i
n
1
I
JL
•';
. i
I
n
•
•';
'•
;
OUT
DAMI
BY-
DUCT
-tx-
3.D.
FAN
INLET
DAMPER
tX
BY-PASS
<• GUILLOTINE
6 DAMPER
INLET DAMPER
(FLYASH)
VALVE
I.D. FAN
I.D.
FAN
OUTLET
DAMPER
STACK
DIRTY FLUE GAS
' FROM BOILER
TO FLYASH
-*• STORAGE /DISPOSAL
Figure 31. Typical boiler/baghouse schematic diagram (courtesy of
Zurn Industries, Inc., Air Systems Div.)-
-------
3.2.2 Dust Handling Systems
Dust handling and removal systems consisting of hoppers, rotary valves,
airlocks, and screw conveyors or pneumatic conveying systems, must be properly
designed to avoid deterioration, leaking, jamming, plugging, or dust
bridging. Hoppers must be of sufficient capacity and slope to avoid the
aforementioned overflow and bridging problems. Hopper heaters, vibrators, or
level indicators may be required in many applications. Provision of some
extra-heavy hopper panels also allows for a primitive but often effective dust
disledging process with a sledge hammer during emergencies (1,3).
3.2.3 Monitoring Equipment (1,7,47,163,165)
Additional auxiliary equipment necessary for proper overall system
operation are monitoring and alarm devices for displaying and/or recording
system and compartment gas flows and pressure losses, compartment isolation
events, baghouse inlet and outlet temperatures, cleaning system operation and
sequencing, particulate emissions, and bag failures (extreme plugging or
rupture).
3.3 BAGHOUSE COMPONENTS (93-95,163-166)
3.3.1 Fabric (1,4,8,94,96,97,99,101-106)
The heart of the baghouse is the bag or fabric material that usually
represents the highest maintenance component in any filter system. Problems
with bags that can lead to excessive emissions or inadequate service life
include: pinhole leaks; gross tears, punctures or burn holes; thermal,
chemical, or moisture damage; inadequate tension (123); twisting, chafing, or
abrasion; and ineffective cleaning and/or blinding (irreversible plugging of
fabric structure) that can cause excessive pressure losses(93-95,161-166).
Problems associated with bag proximity and accessibility (reach), support
structures (wire mesh cages or anticollapse rings), and bottom and top
suspension and tensioning systems must be avoided to guarantee maximum on-line
performance.
3.3.2 Cleaning System (1,46,48,49,95,107,125-132)
Equally as important as the fabric selection is the fabric cleaning
method. Mechanical shaking devices can experience mechanical failures because
of underdesign, limited capacity for bag tension adjustment, or poor
mochanical linkages (1,107). Pulse jet cleaning systems can be rendered
ineffectual due to inadequate compressed air pressure and/or volume, water or
dust in compressed air lines, or lack of flexibility in adjusting pulse
frequency duration, and sequence to the cleaning demands (128,132). Reverse
air cleaning systems can develop problems because of insufficient reverse flow
velocity and duration, total bag collapse if not prevented by anticollapse
rings or compensatory bag tensioning, or bag damage caused by overtensioning
or supplementary shaking (107,123). Reverse jet (RJ) (traveling blowring)
cleaning can present problems similar to those posed by pulse jet (PJ) and
reverse air (RA) systems as well as unique difficulties relating to plugged
blowring nozzles, dust caking on the rings, or severe bag abrasion by
91
-------
inadequate blowring clearance. The present consensus is that pulse jet
collectors have replaced the formerly popular reverse jet device because of
fewer maintenance difficulties.
3.3.3 Basic Structure (1,4,7,94,164-166)
Other baghouse problems, aside from those with bags or cleaning equipment,
relate to the environmental responses of the materials of construction
(rusting, chemical attack, moisture, defective welds, stress failures,)
inoperable access doors and inspection ports, insufficient and/or the wrong
choice of gasketing or insulation, lack of supplementary heating, and
excessive temperatures in compartments isolated for maintenance.
3.4 PROCESS CONDITIONS
Depending on the type and nature of a specific process (steady state,
batch or cycling), any one of several situations may develop that can result
in a temporarily or permanently impaired emission control system (1,4,7,95).
Variable process conditions may generate unusually high dust concentrations or
very fine particles, dust of an atypical character (hygroscopic, sticky, or
with abnormal electrostatic charge), or rapidly changing gas flows and
temperatures (60,61). Any of the above conditions may cause operating
difficulties due to dust adherence to fabric, ducts, hoppers, and dampers.
These transient or upset process conditions are of special concern for those
systems handling explosive, flammable, and/or toxic substances such as grain
dust or asbestos or for processes involving high temperature operations, such
as boilers, furnaces, or rotary kilns (55-57). Since a one-time occurrence of
an excessively high temperature may destroy the fabric material or result in a
baghouse fire, bypass systems or emergency water deluge systems are normally
provided (62).
3.5 FABRIC FILTER USER INPUT
Current user problems with fabric filter systems based upon comments from
19 plants and embracing 30 individual filter systems are summarized'in
Table 11 (112). Since 17 of the operations were related to fly ash filtration
from power boilers, the observations do not necessarily reflect typical
industrial filter performance. On the other hand, they do provide a general
indication as to what has been experienced with baghouses using four cleaning
methods; i.e., 18 reverse air, 2 reverse jet, 8 pulse jet, 2 reverse air with
shake-assist and supplied by 10 different manufacturers (61). They also
represent operations where high temperature (50-53) and moisture (61) factors
require special consideration. The frequency of any given problem does not
indicate the severity of the incident. A notable feature of these findings is
that most users reported minimal problems with their equipment and, in
general, were pleased with overall system performance. Based upon informal
reportings from other sources, however, the data presented in Table 11 appear
to paint an overly optimistic picture of fabric filter experience in heating
or utility boiler applications.
92
-------
OJ
TABLE 11. TYPICAL USER PROBLEM^TNCOUNTERED WITH FABRIC FILTER SYSTEMS
AND FREQUENCY OF OCCURRENCE (112)
A. FANS
1.
2.
3.
4.
5.
6.
7.
Typical problems
AND BLOWERS
Fan underde signed or underpowered
Vibration and balance
Bearing wear
Fan speed adjustment - belting, gears, electric controls
Dust erosion, impeller and scroll
Materials of construction - wrong choice or underdesigned
Electric motors - inadequate cooling, overloading
Problems*
Estimated frequency
Rarely Yearly Monthly
X
X X
X X
X
X
X
X
B. GAS-AIR TRANSPORT AND DISTRIBUTION
1.
2.
Poor hood design, excessive solids entraiiunent (high dust loading)
Incorrect duct sizing - duct abrasion or dust settlement
X
X
3. Materials of construction - underdesign, poor resistance to heat and
chemical attack
A. Absence or incorrect use of baffles and guide vanes
5.' Poor flow distribution in parallel flow paths and among adjoining bag
compartments
6. Isolation valves and dampers - leakage, jamming, slow response,
underdesign
C. BAGS AND FABRICS
1. Excessive dust penetration
2. Excessive pressure loss
X
X
X
X
X
X
(continued)
-------
TABLE 11 (continued)
Typical problems
Problems3
Estimated frequency
Rarely Yearly Monthly
vO
3. Inadequate service life
4. Ineffective fabric cleaning
5. Incorrect cleaning method
6. Design fabric cleaning method
7. Pinhole leaks
8. Gross fabric tears, punctures, burn holes
9. Thermal, chemical, or moisture damage
10. Bag tension adjustments
11. Bag thimbles and floorplates
12. Bag accessibility (reach) for repair and replacement
13. Bag proximity, twisting, chafing, and abrasion
14. Spare parts availability
D. DUSTS AND FUMES
1. High concentrations
i. Fine particle size
3. Stickiness - electrostatic charge or hygroscopicity
4. Adherence to fabric, ducts, hoppers, dampers
5. High temperature - hot embers
6. Explosive, flammable, and/or toxic
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
(continued)
-------
TABLE
continue)
VO
Typical problems
Problems*
Estimated frequency
Rarely Yearly Monthly
E. FABRIC CLEANING
1. Mechanical shaking - mechanical failures and underdesign
2. Mechanical shaking - no adjustment capability, insufficient energy
3. Pulse jet cleaning - inadequate air pressure and volume X
4. Pulse jet cleaning - water in compressed air
5. Pulse jet cleaning - lack of flexibility in regulating pulse frequency,
duration, and sequence
6. Reverse flow cleaning with bag collapse - inadequate reverse flow X
treatment (velocity and duration)
7. Reverse flow cleaning - total bag collapse not prevented by anticollapse X
rings or bag tension
8. Reverse flow cleaning - bag damage by overtensioning or shaking X
9. Improper or defective coupling of pressure sensors or timing devices X
used to initiate cleaning
10. Confusing and poorly located control panels X
F. BAG COMPARTMENT(S) AND SUPPORTING STRUCTURE
1. Materials selection - underdesign and instability X
2. Rusting, chemical attack, moisture X
9
3. Defective welds and gasketing X
4. Inoperable access doors and inspection ports X
5. Faulty insulation X
6. Lack of supplementary heating X
7. Excessive temperature in compartments ioslated for cleaning X
(continued)
-------
TABLE 11 (continued)
Typical problems
Problems8
Estimated frequency
Rarely Yearly Monthly
G. DUST'COLLECTION AND REMOVAL
1. Undersized hopper(s)
2. Poor hopper design - slope
3. Dust bridging - angle of recline or stickiness
4. Hopper heating - dewpoint and moisture effects
5. Hopper overloading - no level sensors
6. Airlocks and rotary valves - leaks and jamming
7. Screw flite conveyors - underdesign and/or underpowered, jamming
8. Pneumatic transport - plugging, moisture
H. FILTER SYSTEM MONITORING AND ALARM DEVICES
1. System gas flow
2. Compartment gas flow
3. System pressure loss
4. Compartment pressure loss
5. Baghouse temperature
6. Hopper dust level
7. Particulate emissions - extractive batch sampling
8. Particulate emissions - continuous monitoring
9. Opacity monitors
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
a"X" designates one or more plant for indicated frequency category, blank spaces indicate no reportings.
-------
In Table 11, problems are grouped in selected categories such as fans and
blowers, gas-air transport and distribution, bags and fabrics, etc. The class
most often checked, other than "rarely", was bags and fabrics (Category C) for
which a total of 56 problems were indicated. With respect to the other
problem categories, difficulties associated with isolation valves and dampers
(B.6.) were followed by opacity monitors (H.9.), airlocks and rotary valves
(G.6.), fabric pinhole leaks (C.7.), and fan vibration and balance (A.2.).
Structural underdesign was one reason for valve or damper malfunction
although corrosion and/or dusting of bearing surfaces and incomplete damper
closure due to thermal x^arping were contributing factors.
Opacity monitors often failed because of dust deposition on the light
sources and/or receptors. Inadequate thermal shielding and shock mounting of
electronic components also created problems as did the improper location of
the sensors with respect to representative measurements. A combination of
mechanical underdesign coupled with inadequate motor powering for the dust of
interest (size, hardness, degree of compaction) contributed to airlock and
rotary valve problems. Moisture plus chemical reactions also caused formation
of a very difficult-to-transport hopper material. Dust deposition on fan
blades, due to moisture and/or lack of a substantial pad for fan and motor
mounting, created an imbalance followed by destructive vibrations.
Although the sampling was extremely limited relative to the actual number
of haghouse installations in current use, some general comments can be made
regarding specific collector types (112):
• Reverse air with bag collapse—problems with isolation valves and
reverse air duct corrosion were most often cited by users of this
equipment. The main cause for reverse air duct corrosion was the
flue gas trapped inside between cleaning intervals that subsequently
cooled to below its dewpoint temperature. One user recommended that
the system bypass dampers be installed so that flue gas passed
through the inlet and outlet plenums. By using this design, the
baghouse will heat up faster when brought online. Ambient air
should not be used directly for reverse air cleaning in any
application that might reduce baghouse temperatures to below
dewpoint levels.
• Reverse air with bag collapse plus shake assist—Based upon the
experience of two facilities included in Table 11 as well as outside
information sources, the most common problem with the above cleaning
approach is insufficient dust removal that leads to excessive
operating pressure losses. The actual difficulty appears to be the
result of dust caking when near dew point operations occur. Under
these circumstances, the fly ash adheres very strongly to the fabric.
• Pulse-jet Collectors—Shortened bag life and malfunctioning
compressed air delivery systems are other problem areas indicated in
Table 11. Abrasion and tearing of the felted fabric due to repeated
impact on the jagged edges of improperly constructed cages was a
97
-------
major cause of bag damage. Failure to dewater the compressed air
supply also increased dust adhesion and inadequate pulse pressure
(and delivered pulse volume) led to diminished bag flexure.
• Reverse-jet Collectors—The major design and operating problems
encountered with reverse jet collectors relate to the spacing
between the blowring and the bag. With excessive clearances between
blowring and bag, the air jet velocities are not high enough to
dislodge the dust cake. Conversely, too close spacing between the
ring and the bag causes fabric damage and reduced service life.
According to the Table 11 information sources, power plant operations
were not seriously impaired (< 5 percent) by the indicated problems. This
statistic should not be extended to industry as a whole, however, because of
generally better maintenance within the utilities industry and the fact that
experience with fabric filters for fly ash removal involves a relatively brief
period in comparison to many industrial applications.
3.6 PROBLEM RECTIFICATION (167-172)
Information was obtained from several baghouse manufacturers and other
sources to establish current practice for solving common baghouse problems.
Table 12 provides a comprehensive listing of troubleshooting guidelines based
upon several manufacturers' bulletins and other relevant documents (167—172).
Where the problems, symptoms, and/or remedies are specific to a given
collector type, they are so designated. It should be recognized that many of
the problems, causes, and solutions may be so interrelated that correcting one
problem (e.g., defective cleaning system) may automatically correct another
(e.g., excessive pressure loss).
98
-------
TABLE 12. TROUBLESHOOTING GUIDELINES FOR COMMON BACHOUSE OPERATING PROBLEMS3
Problem or symptom
Probable cause
Remedy
1. Visible discharge/dust in clean
air plenum
2. Excessive pressure drop (a differ-
ential pressure of 1 to 6 in. W.C.
can be considered normal)
a. Bags improperly installed
b. Bag clamps too loose
c. Torn or damaged bags
d. Leakage at tube sheet (field
assembled units)
e. Venturi fasteners loose or
missing (PJ)
f. Insufficient filter cake
g. Bags too porous
h. Inadequate bag tension
a. Gas flow too high
b. Improper bag cleaning action
c. Improper dust discharge from hopper.
d. Moisture blinding of filter bafs
• Check bag installation procedures;
repair as necessary.
• Tighten bag clamps.
• Replace or tie off and replace at later
date.
• Check tube sheet joints; repair as
necessary.
• Repair as necessary
• Allow more dust to build up on bags
by cleaning less frequently. Use a
precoat (startup only).
• Send bag out for permeability test
and consult with manufacturer.
• Check tension and/or springs for
compression to proper lexvgth.
• Check fan speed and damper positions;
adjust to specified ratings. Check
system design. Check isolation
dampers, valves, linkage and seals.
Check air supply on pneumatic
operators.
• Refer to Problem 6.
• Check seal around slide gate (or in
airlock); reseal if leakage is
occurring. Ensure continuous dust
removal from hoppers.
• Correct cause of excess moisture and
replace bags. Recovery of bag is
sometimes possible by running cleaning
system (without moving air through
collector) from 1 to 30 hours.
(continued)
-------
TABLE 12 (continued)
Problem or symptom
Probable cause
Remedy
e. Clogging of filter bags
O
O
3. High bag failure rate
a. 5eterioration/decomposition
b. Wearing out/abrasion
f. Static electricity in collector
g. Excessive dust in clean air plenum
(can diminish cleaning effectiveness
by plugging the bags in the reverse
direction). Damaged or inefficient
bags.
h. Insufficient blowring air (RJ)
i. Nonoperative blowring drive
j. Incorrect pressure reading
a. Improper bag material for dust
chemical composition
b. Operating below acid dewpoint
a. Baffle plate worn out
b. Excessive dust loading and/or large
abrasive metallic particles
Eliminate oil or static charges from
collector. Control airflow during
startup. Check for excessive
operating temperature. Check to see
that dust characteristics have not
changed. If using laundered bags,
check for shrinkage.
Increase relative humidity if possible.
Use grounded filter bags.
Clean plenum, check bags for dirt on
clean side; clean or replace bags.
Check drive belts on blower, blower
speed, blower rotation; replace blower
inlet filter; check blowring hose;
check blowring for clogged slots.
Check belts on reversing motor; check
for broken chain or chain off
sprocket; check sheaves or sprockets
for loose set screws, check setting
of tripper pins; check tripper lever
for bend, looseness, or binding.
Clean out pressure taps; check hoses
for leaks; check for proper fluid in
manometer; check diaphragm in gauge.
• Analyze dust and check with manu-
facturer. Treat with neutralizer
prior to collector.
• Increase gas temperature. Bypass
collector during start-up/shutdown.
• Replace baffle plate.
• Install primary collector upstream
of baghouse.
: inued)
-------
TABLE 12 (continued)
Problem or symptom
Probable cause
Remedy
Burning or accelerated fabric
deterioration
d. Other
4. Filtering action impaired
c. Cleaning cycle too frequent (PJ) (PP)
d. Shaking too violent (S)
e. Inlet air not properly baffled from
bags
f. Pulse pressure too high (PJ) (PP)
g. Bag cages have barbs (PJ) (PP)
a. Stratification of hot and cold gases
b. Sparks entering collector
c. Failure of cooling or dilution system
d. Excessive temperature
a. Hopper bridging
a. Improper cleaning system operation
b. Excessive moisture entering collector
blinding bags
c. Incorrect gas flow
d. Incorrect bag material for gas
composition
e. Gas temperature higher than specified
f. Static electricity buildup in
collector
• Increase cleaning interval.
• Decrease shaking frequency and/or
amplitude.
• Consult manufacturer.
• Reduce pressure.
• Remove or smooth barbs.
• Install baffles to create turbulence.
• Install spark arrestor.
• Check design with manufacturer.
• Reduce operating temperature or use
filter bags of higher temperature
rating.
• Material buildup into the bag area
can overstress filter elements.
Locate cause of bridging and correct;
clean out hopper.
• Refer to Item 6.
• Refer to Item 2.d.
• Refer to Item 5.
• Replace with bags of proper material
for type of dust.
• Refer to Item 3.c.d.
• Refer to Item 2.f.
(continued)
-------
TABLE 12 (continued)
Problem or symptom
Probable cause
Remedy
5. Gas flow through system below
design rating/low fan amperage
6. Improper cleaning system operation
a. Blowring reversing switch
failure (RJ)
a. Incorrect fan speed (or direction)
b. High differential pressure drop
c. Fan belts slipping
d. Air leakage in gas system
e. Leakage in dust collection system
f. Blocked gas system
g. Fan and motor sheaves reversed
h. System static pressure too high
i. Excessive moisture
j. Infrequent cleaning
a. Worn cams or rollers
b. Improperly set tripper level
c. Dirt in switch
• Check rotation, correct if wrong,
change sheave ratio.
• Refer to Item 2.
• Check tension on fan belts and adjust
if necessary.
• Check access doors, plenum, manifolds,
duct work; repair leaks.
• Check for hopper or discharge system
leakage; repair as necessary.
• Check bags for blinding, obstruction
in duct passages, or closed damper;
clean or repair as necessary.
• Check drawings and reverse sheaves.
• Measure static pressure on both
sides of fan and discuss with designer
of duct velocity and configuration.
• Refer to Item 2.d.
• Refer to Item 6.
• Repair or replace switch assembly.
• Adjust arc of movement for approxi-
mately 60° above or below horizontal;
uneven arc will cause switch to work
in only one directiom.
• Check to see that enclosure is
properly installed.
(continued)
-------
TABLE 12 (continued)
Problem or symptom
Probable cause
Remedy
b. Inability Co maintain compressed
air pressure (PJ), (PP)
o
u>
c. Reduced compressed air
consumption (PJ), (PP)
d. Inadequate cleaning (S)
a. Faulty or undersized compressor
b.- Leakage or restriction in main air
line
c. Defective timer operation
d. Improper operation of solenoid or
diaphragm valves
• Check compressor manual. Pressure
should normally be maintained between
80 to 110 psig.
• Locate and repair leak or restriction.
• Make sure all valves are being
activated. Check for sticking timer
relay or pulse longer than 0.15
seconds. Replace timer if necessary.
• Examine valves for dirt or short
circuit in wiring which can cause
valves to stick open. Clean and
check pilot plunger.
NOTE: Steady rush of air indicates open valve; no air pulse
indicates plugged valve. Solenoid valves require a
minimum of 5 psig to close. A long compressed air run
after the shutoff valve has closed can prevent the
required 5 psig from developing. The solution would
be provision of reservoir and shutoff valve near the
collector.
e. Compressed air consumption too high
f. Plugged dryer
g. Supply line too small
h. Compressor worn
a. Pulsing (solenoid) valves not working
b. Failed timer
a. Defective shaker mechanism
• Reduce cleaning cycle, duration of
pulse, or supply pressure, if possible.
• Replace desiccant or bypass dryer
if permitted.
• Consult design.
• Replace rings.
• Check diaphragm, springs, and pilot
valves.
• Check terminal outputs.
• Check shaker speed, amplitude, and
bag tension; adjust if required.
Check for broken linkage and lost pins
connecting linkage.
(continued)
-------
TABLE 12 (continued)
Problem or symptom
Probable cause
Remedy
7. Moisture in baghouse
O
•F-
8. Insufficient dust pickup at
emission points
9. Fan problems—excessive wear,
noise, vibration, or motor
overloading
a. Insufficient preheating
b. System not purged after shutdown
c. Wall temperature below dewpoint
d. Cold spots through insulation
e. Compressed air introducing water
(RJ, PP)
f. Reprassuring air causing condensation
(RA, PP)
g. Moisture in compressed air line
a. Leaks in ductwork, access doors,
and/or hopper discharge valves
b. High differential pressure
c. Slipping fan belts or fan rotating
in wrong direction
d. Clogged duct or closed or partially
closed gate or damper
e. Duct size or run other than original
design/inadequate system design
a. Improper fan
b. Fan speed too high
• Run system with hot air prior to
starting process gas flow.
• Keep fan running for 5 to 10 minutes
after process is shut down.
• Raise gas temperature, insulate unit,
install auxiliary heaters. Lower
dewpoint by keeping moisture out of
system.
• Eliminate direct metal lines through
insulation.
• Check automatic drains, install
aftercooler, install dryer.
• Preheat repressuring air; use process
gas as source of repressuring air.
• Make sure cooler or water trap is
functioning.
• Repair leaks so that air does not
bypass source.
• Refer to Item 2.
• Check fan and repair as necessary.
• Check all ductwork and damper
positions and operation.
• Check design specifications with
manufacturer. Close open areas
around dust source. Check for cross
drafts that overcome suction.
• Check with fan manufacturer to see
if fan is of proper design for
application.
• Consult fan manufacturer.
(continued)
-------
TABLE 12 (continued)
Problem of symptom
Probable cause
Remedy
10. Hopper/Dust Discharge System
Failure or excessive wear
a. High screw conveyor or airlock
wear
b. Frequent screw conveyor or
airlock failure
c. High pneumatic conveyor wear
c. Dust building up on fan blades.
d. Improper fan wheel
e. Sheaves not balanced
f. Worn bearings
g. Air volume too high
h. Motor not sized for cold start
a. Screw conveyor or airlock undersized
b. Conveyor or airlock speed too high
c. Thermal expansion
a. Undersized equipment
b. Misaligned screw conveyor
c. Overloading components
a. Blower set too fast
b. Undersized piping
d. Pneumatic conveyor pipes plugging a. Elbows designed with too short a
radius
b. Overloading pneumatic conveyor
c. Slug loading of dust
• Clean fan and check for water.
• Check with manufacturer.
• Have sheaves dynamically balanced.
• Replace bearings.
• Refer to Item 2.a.
• Dampen fan at start-up, reduce fan
speed, provide heat faster, or replace
motor.
• Measure hourly collection of dust
and consult manufacturer.
• Check and reduce speed.
• Consult manufacturer.
• Consult manufacturer.
• Check and align.
• Check sizing and design of all
components versus dust delivery
rates.
• Check and reduce speed.
• Review design—slow down blower or
increase pipe size.
• Replace with long radius elbows.
• Review design.
• Feed dust gradually.
(continued)
-------
TABLE 12 (continued)
Problem or symptom
Probable cause
Remedy
e. Material bridging in hopper
£. Excessive corrosion of baghouse
structure, bag mountings, wire
cages
d. Moisture in dust
a. Moisture in baghouse
b. Dost being stored in hopper
c. Insufficient hopper slope
d. Conveyor opening too small
a. Air in-leakage through defective
gaskets
b. Missing or damaged insulation
c. Frequent devpoint excursions
• Refer to Item 7.
• Refer to Item 7.
• Ensure that dust is removed
continuously.
• Rework or replace hoppers.
• Use a wide flared trough.
• Proper inspection and maintenance.
• Heat tracing and/or gas temperature
elevation.
apj • pulse jet
PP » plenum pulse
RA = reverse air
RJ " reverse jet
S • shaker
-------
CHAPTER 4
BAGHOUSE INSPECTIONS, TYPES AND PROCEDURES
4.1 INTRODUCTION (1,163,164)
Inspections of fabric filter systems must be performed on a routine basis
to ensure proper functioning of equipment. This is especially important in
applications where highly toxic and/or hazardous materials are being collected
or in situations where the cleaned effluent is recirculated to the working
area for energy conservation.
Baghouse inspections fall into two main categories: (l) those conducted
by plant personnel (sometimes in conjunction with the baghouse manufacturer)
as part of system start-up, troubleshooting, and routine maintenance, and
(2) those performed by field inspectors representing regulatory enforcement
agencies for purposes of compliance determination, permit renewal, or problem
assessment.
4.2 INSPECTIONS PERFORMED BY PLANT PERSONNEL
4.2.1 Background (161-165)
Those inspections performed by baghouse users and/or vendors are
typically related to initial start-up and equipment shakedown, problem
solving, general bag replacement, and maintenance evaluations. In subsequent
paragraphs, inspection categories will be discussed in the following
sequence: prestart-up, start-up, troubleshooting, and general preventive
maintenance, including bag replacement and cleaning.
Although an enforcement agency inspector may only occasionally witness a
start-up operation or review a user's troubleshooting or preventive
maintenance program, the procedures are discussed here to acquaint inspection
personnel with the routines involved in installing, debugging, and maintaining
a fabric filter system. Such background material will enable air pollution
control agencies to make more accurate assessments and evaluations of the
operating and maintenance procedures practiced at a given facility.
It is emphasized that the guidelines provided in this manual should be
used to augment and not replace the information ordinarily provided in
start—up, operating, and maintenance manuals supplied by the manufacturer for
the specific baghouse undergoing inspection (1,4,94,161-166). After gaining
experience with his fabric filter system, the user frequently develops a set
of start-up, operating, and maintenance procedures that are uniquely adapted
to his needs.
4.2.2 Prestart-Up (161)
Since most fabric filter systems use a centrifugal fan as the gas mover,
the compatibility of the fan and collector is extremely important. On
start-up with unused fabrics, collector resistance will be considerably lower
107
-------
than the operating design level, perhaps 0.5 instead of 6 in. water.
Consequently, the gas delivery rate (without throttling) will exceed the
design value with the possibility of two undesirable effects. The fan power
level may rise to the point of motor overload, and/or the higher-than-design
volume flow may cause excess penetration due to seepage, as well as damage to
the filter medium (164-166).
In many cases, the overall system resistance is so high so that the
variations in pressure loss between new and seasoned fabrics are too small to
affect significantly the fan capacity (1,7,94). For those systems where the
design pressure loss across the fabric represents a large fraction of the
total system pressure loss, the overload condition can be minimized by damper
adjustment, selection of a fan with nonoverloading characteristics, or
adjustable inlet vanes. Caution must be exercised in the start-up of high
temperature systems with fans handling air at ambient temperature (161). All
dampers and/or inlet vanes should be partially closed at start-up to reduce
power consumption. After the correct fan speed is reached and the system
temperature has approached the operating level, dampers may be opened
cautiously to avoid motor overload. It should be noted that rough fan
operation (vibration and noise) should be expected when dampers are partially
closed. In special systems such as air classifiers that are highly sensitive
to flow change, an automatic flow controller consisting of a motor-operated
damper with a fan speed regulator may be helpful (161). Similar controls will
also attenuate minor fluctuations when compartments go off-and online during
cleaning.
Before initiation of normal service, there are several items that must be
checked on any fabric filter system (161-166):
• Inspect all bag compartments and ducting to see that joints have
been made tight with sealing compounds or gasketing. The general
location of air leaks in either positive or negative pressure
baghouses can often be detected by the air noise while exact
locations can be established by applying a soap solution to
suspected leak areas.
• Tighten all bolts properly and lubricate threaded elements on clamps
and door latches for corrosion protection and easy access.
• Secure fabric bags properly to floor thimbles or cages and allow for
required tension adjustments. If bags are furnished with ground
wires to guard against sparking (and dust explosions), they should
be securely connected to the tube sheet which, in turn, must be well
grounded.
• Check all system controls to verify installation and operation in
accordance with manufacturer's recommendations (161,167-172).
• Inspect fan to ensure that it rotates in the proper direction since
accidental reversal of impeller motion reduces but does not change
the direction of air flow. A visual inspection is recommended to
identify this occasional cause of high pressure loss, reduced
amperage, and diminished air handling capacity.
108
-------
• Avoid operating conditions involving excessive temperatures, high
humidities, and high dust loadings. These steps are important
regardless of the collector type or the process being controlled.
• Check the fan and motor system for vibration, noise, and, in
particular, overheated bearings; the latter signaled not only by
high temperatures but also by acrid "hot metal" odors.
4.2.3 Baghouse Start-up—Standard Procedure (167-172)
The start-up (and shutdown) procedure for any baghouse depends upon the
collector type (shaker, reverse air, pulse jet) and the process and/or
emissions being controlled. As examples of real life applications, we have
presented an abridged version of (a) the general procedures recommended by the
manufacturer (Mikro-Pul Corp.) of a pulse jet baghouse (168) for a spray
drying operation and (b) the specific procedures employed during start-up of a
Zurn Industries reverse air baghouse (170) controlling fly ash emissions from
two coal-fired boilers. The following guidelines are samplings of the data
supplied the user by manufacturers of filtration equipment to ensure that the
systems will provide the service for which they were intended.
4.2.3.1 General Start-up Procedures for a Pulse Jet Baghouse with Spray Drier
or Process Equipment (168)—
Preliminary adjustments must be made in the process itself before
installing bags in the collector. An incorrectly functioning dryer or other
malfunctioning process equipment may result in destruction of the bags if
temperature or moisture content is not under control. When using drying
equipment, the collector should be preheated from 30 minutes to 1 hour before
initiating filtration to avoid condensation within the baghouse and
specifically on the fabric surface. The above time interval, which applies to
many indirect heating systems such as steam coils or electrical elements, also
depends upon the power consumption and location of heating surfaces. In the
case of asphalt concrete plants where drier flue gas may be used for
preheating, the desired baghouse temperature, ~300°F, may be reached in a few
minutes.
• When starting up with new filter bags, proportionately higher air
flow through the low resistance surface of the fabric can increase
particle impingement speeds to levels where bag abrasion may take
place. The above problem and that of bag blinding due to very fine
or sticky materials are lessened by adjusting inlet and/or outlet
damper controls to their half-flow settings.
• The inlet damper should be opened fully only after the bag pressure
differential has increased to 3 to 4 in. water.
• The timer controlling the compressed air pulses should not be
activated until the differential pressure has reached 4 to 5 in.
water unless operating conditions demand a lower pressure drop.
• During normal start-up with seasoned or conditioned bags, apply
power to all auxiliary equipment (except fan), energize timer, and
turn on compressed air.
109
-------
• Turn on fan motor with the system damper nearly closed to prevent
motor overload during the starting power surge.
• Maintain the pressure loss across the fabric within its preset range
(which may vary from 1 to 6 in. water according to the cleaning
requirement) by adjustment of the pulse jet cleaning cycle. More
frequent and higher pressure pulses will reduce the bag pressure
loss whereas the opposite actions will increase bag pressure loss
should the need arise to reduce dust penetration.
• System shutdown should be undertaken by first turning off the fan
followed by the closing of inlet and exhaust dampers. After a 15 to
30 minute waiting period, the compressed air and timing circuit
should be shut off along with any auxiliary equipment.
• The hopper(s) should be emptied of material before the airlock
and/or screw conveyor is turned off. This step will reduce the
chance of dust hangup and plugging due to compaction and sticking of
the dust in condensing atmospheres.
4.2.3.2 Specific Start-up Procedure for a Reverse Air Baghouse Used for the
Collection of Coal Fly Ash (170)—
The basic philosophy during start-up is to prevent flue gas from the
coal-fired boiler from entering the baghouse until the baghouse is completely
preheated and the bags precoated. These precautionary steps minimize
penetration of fine particles into the fabric structure, thereby reducing the
chance o£ premature filter plugging or blinding. Prior to actual start-up,
the system is checked out to verify that all control elements, dampers, and
fans are functioning correctly. Additionally, contractor and user personnel
should jointly develop the system start-up regimen.
The specific installation whose start-up procedures are discussed below
consisted of two coal-fired boilers plus an auxiliary gas-fired boiler. The
ductwork was arranged so that the baghouse could be bypassed and there were
special dampers provided so that the clean flue gas from the a gas-fired
boilers could be used for preheating the individual baghouse compartments. In
the event that a gas-fired unit is not available, baghouse preheating options
might entail electrical tracing, steam coils, or a specially designed
auxiliary gas heater whose effluent can be directed into one or both boilers
until temperatures exceed the dew point of the flue gas. The time required to
reach safe (above dew point) temperatures will depend mainly on the heat rate
and disposition of heating surfaces. A paraphrased version of the first hand
reporting of start-up procedures and their results for the coal- and gas—fired
boilers discussed above is presented in the following paragraphs.
Step No. 1—Preheat Compartments Using Hopper Heater—Electrical strip
heaters located in the collector hoppers were turned on while the compartment
inlet dampers remained open and the compartment outlet and reverse air dampers
were closed. This step elevated compartment temperatures to only 90°F,
considerably below the 125°F sought to exceed the water dew point.
Insufficient heating was attributed to low (November) ambient temperatures.
110
-------
Step No. 2—Preheating of Inlet Duct with Flue Gas from Gas^-Fired
Boiler—Damper arrangements for these systems permitted the clean flue gas
from the gas-fired boiler to be directed into the inlet duct leading to the
coal-fired boiler baghouse. By opening the by-pass damper immediately before
the inlet to the baghouse, the effluent from the gas-fired boiler could be
used to preheat the inlet duct up to the bypass damper, thus reducing the
system heat loss when the coal-fired baghouse was brought on-line. The
entering gas temperature was estimated to exceed 300°F.
Step No. 3—Preheating the Baghouse with Hot Flue Gas—The flue gas from
the gas-fired boiler was then directed through two baghouse compartments at a
time, by closing the proper bypass dampers and opening the compartment outlet
dampers. This process was continued until the entire baghouse was heated to
275°F, a temperature considered high enough to minimize condensation problems.
Step No. 4—Precoating of Bag Surfaces—The bags were precoated by
injecting fly ash pneumatically into the inlet duct so that an areal density
of about 0.2 Ib per square foot was attained over a 1 1/2 hour period. The
fly ash was trucked to the plant from a local coal burning utility.
Step No. 5—Verification of Precoat Effectiveness—Visual inspection of
the filtration surface through an open hopper access door with the aid of a
bright light source indicated that the bags were coated over their full
length. It was anticipated that this surface precoat would capture most of
the fine and potentially plugging combustion products that exist before steady
state combustion has been achieved.
Step No. 6—Admission of Flue Gas to Baghouse—Compartment isolation
dampers were opened so that normal filtration could commence. It had been
previously established, Step No. 5, that an adequate precoating of the fabric
surface had been accomplished so that any sooty combustion products would be
filtered out by the surface dust layer rather than penetrating the underlying
fabric. Because the boiler was operating at roughly 1/4 capacity a very low
pressure loss across the bag was indicated.
4.2.4 Troubleshooting (161-166,168,169)
The tracing and correcting of fabric filter problems requires a detailed
baghouse inspection by the plant operator, possibly with the assistance of the
baghouse manufacturer. As indicated in the previous section (Table 12), many
problems have more than one cause and hence will require a combination of
corrective actions (167-172). Although the air pollution inspector may never
participate in a detailed baghouse inspection, familiarity with inspection
procedures and remedial measures employed by the user/manufacturer team can
add greatly to his overall knowledge and understanding of the fabric filter
system. The adequacy of a fabric filter system is demonstrated, for the most
part, by a cleanliness of the exhaust air stream and a moderate pressure loss
across the fabric (161).This assumes that the gas flow requirement is
satisfied and that annualized costs and maintenance requirements are
acceptable. The quality of the stack emissions and the collector pressure
loss characteristics, both of which are readily determinable by enforcement
agency personnel, will provide a practical assessment of system operation.
Ill
-------
Very low pressure losses may signify damaged (leaking) bags and/or fabric
overcleaning, both of which may lead to reduced collection efficiency. On the
other hand, high fabric pressure loss, while often an index of high collection
efficiency, may reflect an undersized baghouse and/or insufficient cleaning.
4.2.5 General Preventive Maintenance, Including Bag Replacement and
Cleaning (161)
Any preventive maintenance schedule must include periodic inspections of
all system components. The following discussion, which addresses the more
important items requiring careful maintenance, is followed by a system
maintenance checklist.
4.2.5.1 Care of Filter Media (162-166) —
Fabric selection will vary widely depending on such factors as
temperature, humidity, chemical characteristics of the dust or gas stream, and
general operating and cleaning procedures (45,50-53). The fabric manufacturer
should be consulted regarding care and use of the fabric, washing, patching,
and selection of replacement bags. Readjustment of fabric tension, which is
usually recommended within 2 weeks after start-up, should be checked
periodically thereafter (123). Regular bag inspections should also be made to
ensure that the fabric is not blinding. Abnormal increases in pressure loss
across the fabric may signal a blinding (or plugging) problem. Bags should
also be checked regularly for loose clamps, torn or worn spots, excessive
slackness, or other potential sources of faulty performance. Pin holes and
small tears can occasionally be mended, but the permanence of such repairs
depends upon the patching material, the type of fabric, the operating
environment, and, in particular, the skill of the repair person.
One manufacturer (168) recommends that a cost analysis be performed to
determine whether bag cleaning (outside of the baghouse) or bag replacement is
the more economical approach. If bag cleaning appears acceptable, it is
recommended that bags be carefully removed from the collector housing and
first thoroughly vacuum-cleaned. If vacuuming alone is not sufficient to
restore bags to an operable condition, the following laundering and patching
procedures can be undertaken:
• Natural Filter Fabrics (Wool and Cotton)—All wool and cotton
fabrics are subject to shrinkage when washed in water. Since the
result may be poorly fitting bags, it is advised that cleaning be
performed by conventional dry cleaning processes with pure organic
solvents. Dry cleaning detergents and additives that permit
addition of water to the solvent may cause shrinkage of wool felt.
Wherever possible, drying should be carried out by hanging in drying
rooms rather than in tumblers. Addition of surfactant to the
cleaning solvent (2 to 3 percent by weight) will help in restoring
the structure of both natural and synthetic fabrics.
• Synthetic Filter Fabrics (e.g., Polyester, Polypropylene, Nomex)—
Most synthetic media can be washed in water without shrinkage
provided that the water temperature does not exceed 140°F and a mild
soap such as a dish washing detergent is used. Again, bags should
112
-------
be hung in drying rooms to avoid the creasing and twisting
associated with tumble drying. Drying temperatures should also be
maintained at moderate levels (less than 150°F) to prevent fiber
shrinkage. Some synthetic products, however, are best restored by
dry cleaning with Stoddard's solvent that will not damage most
synthetics.
After the bags are completely dried they should be inspected for defects
by viewing the surface while illuminated from the opposite side. Thin spots
or holes, which are clearly indicated by the above technique, are marked for
emergency repair, the latter consisting of patching the inside surface of the
bag, with a combination of perimeter stitching and sufficient
cross-hatch-sewing to secure any loose edges. All patches, as well as the
thread, should be of the same material used in the original bag. Because this
type of sewing requires heavy duty machines, the work should be performed by a
sailmaker or awning company, or by an industrial firm specializing in bag
cleaning and repair.
The above repair techniques are suggested mainly for emergency
situations. It is often preferable that damaged bags be replaced by new ones
since bag life and filtration efficiency for the repaired bags may be reduced.
4.2.5.2 Bag Replacement—
The ideal approach to bag replacement is to renew all fabric in the
collector, or at least in an individual compartment at the same time. If only
a small fraction of the bags are replaced, air flow through the replacement
bags will commence at very high levels, up to 10 times the normal rate. The
result will be greatly increased local dust penetration and, more importantly,
an increased likelihood of fabric plugging or blinding.
Fortunately, any high initial flows through a new bag will soon increase
its pressure loss to the level of the remainder of the bags. Thus, despite
the problem potential, individual bags can usually be replaced without
impairing system function. An exception occurs, however, when very abrasive
dusts are collected by abrasion-sensitive media, e.g., cement clinker dust and
glass bags. Here, high velocities near the inlets (thimble end) of new bags
can cause severe abrasion in a very short time.
Constant advancements in filter media technology make it worthwhile to
maintain contact with the manufacturer. As new fibers and fabrics become
available, the manufacturer is in the best position to offer advice and
suggestions for improving media performance and service life (167-172).
Replacement filter media should be obtained from, or ordered through, the
original equipment manufacturer. If a different medium is considered, consult
the collector manufacturer and obtain his approval for the substitute
material. Changing to an unapproved fabric may lead to serious operational
problems and also void manufacturer guarantees.
113
-------
4.2.5.3 Valves and Dampers (161-166,167-172)—
Most baghouses require valves or dampers for isolating individual
compartments during cleaning and for regulating the discharge of the collected
material from the dust hopper. The proper operation and maintenance of the
dampers or valves including lubrication and avoidance of dust caking and
corrosion is basic to effective collector performance. It is also important
that the manufacturer's recommendations be incorporated in the user's
maintenance schedule. Valve and damper motion is often detected by the sounds
of opening and closing, pressure loss changes across isolated compartments,
and the observed motion of external drive mechanisms.
4.2.5.4 Motors, Fans, and Belts (167-172)—
Motors require periodic inspection and lubrication with the frequency
depending on the type of motor, type of bearings, and severity of service.
Fans should be checked especially for tightness of all bolts, bearing
vibration, noise, or signs of bearing wear.
The following guidelines are offered to estimate what constitutes excess
vibration in terms of the observed amplitude (161).
Vibration Amplitude
mils
Fan Speed
rpm Smooth Fair Rough Very rough
600 2 4 8 15-20
900 1.5 2.75 6 8-10
1200 1.0 2.0 4.5 6-8
1900 0.75 1.5 3.5 5-7
1 mil = 0.001 inches
Fan housings and wheels (impellers) should be inspected for wear and
accumulations of dust and dirt. Cleaning thoroughly with a steam or water jet
and compressed air or wire brush will prevent fan imbalance. Bearings should
be shielded so that water and dust will not penetrate the pillow block during
cleaning. Fan wheels with badly worn blades, flanges, or collars should be
replaced or rebuilt and rebalanced. Consult the fan manufacturer in such a
situation. Since bearings on high speed fans are often designed to operate
between 100° and 200°F, do not replace them merely because they appear to be
running hot. Check the temperature of the pillow block with a thermometer and
compare with the fan manufacturer's written recommendations. If the motor
requires factory repair, it is usually the motor manufacturer and not the
baghouse manufacturer who provides this service.
A large fan is usually driven by an electric motor through a V-belt drive
that requires proper belt tensioning to prevent undue slippage or unnecessary
stress on the fan bearings. If belts squeal at start-up, they are too loose
and should be tightened. New belts often require a period of 1 to 2 months
running-in time before final tensioning is accomplished. Worn belts should be
114
-------
replaced by matched sets of spare V-belts that should be maintained in company
stock. Sheaves should be replaced when the grooves become sufficiently worn
to allow slippage or cause belt abrasion.
4.2.5.5 Ductwork (161-166) —
Proper performance of a dust collection system is only attainable with
frequent inspections and/or maintenance of the housing, ductwork, hoods,
hopper, and dust removal system. Periodic inspections of the ductwork should
be made to ascertain that sufficient gas velocity is maintained throughout the
system to prevent dust settlement. Generally, a velocity of 4000 ft/rain is
recommended for horizontal runs whereas 2000 ft/rain is usually sufficient in
vertically aligned ducts.
The duct system including supporting structures should be protected
against physical damage due to impact, vibration, and abrasion. Any sections
damaged sufficiently to interfere with gas flow or to cause leakage should be
repaired immediately or replaced. Frequent inspections and maintenance are
required to keep the duct system in good operating condition.
4.2.5.6 Auxiliary Systems—(167-172)
Compressed air systems must be installed and operated in accordance with
good engineering practice with aftercoolers, automatic condensate traps, and
filters included where their use is indicated. Special situations such as
very low ambient temperatures will require thorough drying and/or heating to
prevent freezing (168,169).
The instructions concerning maintenance and operation of valves whether
they be pneumatically or electrically driven must be followed closely.
Pneumatically driven units often require special treatment of the compressed
air supply such as filtration, dewatering, addition of lubricant, or
antifreeze.
4.2.5.7 Maintenance Checklist (161)—
As with start-up, preventive maintenance depends on the type of collector
and the recommendations regarding inspection frequency provided by the
equipment manufacturer. A composite listing of 10 items or conditions to be
regularly checked for different types of baghouses is provided below (167-172).
• Inspect filter media for blinding, leakage, wear, slack, bag
tension, loose bag clamps, or discoloration. If the collector is a
reverse jet design, fabric discoloration may indicate blower
problems. Additionally, check the contact surface of blowrings for
dust encrusation.
• Inspect the inside surface of reverse jet system blowrings for
burrs, scaling, clogged slots, or indications of bag chafing. Check
blowrings for leaks and examine for signs of deterioration.
Routinely lubricate the blower and inspect for hot spots or
increases in vibration and noise. It is pointed out that
reverse-jet systems, Appendix D, Figure D-l, are gradually being
replaced by the more effective pulse jet devices.
115
-------
• Inspect overall collector and compartment housings, hooding, and
connecting ductwork for leakage, corrosion, or dust accumulation.
• Check all solenoid-operated pneumatic damper actuators, airlocks,
and valves for proper seating, dust accumulation, leakage,
synchronization, and operation.
• Check hopper discharge for possible bridging of dust.
• Determine regularly the bag pressure drop and compare frequency of
cleaning with that recommended by the manufacturer.
• Inspect fans for tightness of bolts, bearing vibration and
temperature, erosion or dust buildup in housing and on wheel,
alignment of fan impeller with V-belt drive or coupling and driver,
and sheave appearance for signs of V-belt wear.
• Inspect all bearings on fans, motors, dampers, etc. for lubrication
and free rotation. Follow manufacturer's instructions since
overlubrication must be avoided.
• Check foundation (footing or sill) bolts on collector, motor, fan,
etc. for tightness as well as bolts on collector housing and
structural members.
• Check access door(s) for leakage due to faulty gaskets or warping of
door(s) and/or frame(s).
For the above-stated items, the inspection frequency will depend on the
vendor's recommendations for a particular system, the standards in effect at
the user's plant, and the severity of the operating environment. Examples of
recommended inspection intervals, however, for some of the major filter system
components are provided in Table 13 (167-172). Whereas Table 13 reflects
vendor experience, the maintenance schedules cited in Table 14 are based upon
a special polling of filter users by the authors of this handbook. Despite
the diverse collector types represented in Table 14, one common facet should
be noted. Isolation valves and dampers (item 7), bags (Item 8), and the
baghouse structure (Item 13) all show widely varying inspection periods. This
may explain why these components experience the greatest problem frequency, as
pointed out in the previous section.
4.3 INSPECTIONS PERFORMED BY AIR POLLUTION AGENCIES
4.3.1 Types of Inspections
Baghouse inspections by air pollution control agencies may be performed
for any one of the following reasons: compliance determination, complaint
investigation as a result of excess emissions or equipment malfunction, source
registration/plan approval, permit review/renewal, and special studies (i.e.,
operating and maintenance evaluations or updating emission inventories). The
type and purpose of the inspection will determine (a) the extent of
116
-------
TABLE 13. TYPICAL MAINTENANCE SCHEDULE FOR A FABRIC FILTER SYSTEM 167-172
Inspection
frequency
Component
Procedure
Daily
Stack and opacity meter
Manometer
Compressed air system
Collector
Damper valves
Weekly
Rotating equipment and
drives
Filter bags
Cleaning system
Hoppers
Check exhaust for visible dust.
Check and record fabric pressure
loss and fan static pressure.
Watch for trends.
Check for air leakage (low
pressure). Check valves.
Observe all dials, meters,
charts, and gauges etc. on
control panel and listen to
system for properly operating
subsystems.
Check all isolation, bypass, and
cleaning damper valves for
synchronization and proper
operation based upon
manufacturer guidelines.
Check for signs of jamming,
leakage, broken parts, wear, etc.
Check for tears, holes,
abrasion, proper fastening, bag
tension, dust accumulation on
surface or in creases and folds.
Check cleaning sequence and
cycle times for proper valve and
timer operation. Check
compressed air lines including
oilers and filters. Inspect
shaker mechanisms for proper
operation.
Check for bridging or plugging.
Inspect screw conveyor flighting
for proper operation and
lubrication.
(continued)
117
-------
TABLE 13 (continued)
Inspection
frequency
Component
Procedure
Monthly
Quarterly
Shaker mechanism
Fan(s)
Monitor(s)
Inlet plenum
Access doors
Shaker mechanism
Semi-annually Motors, fans, etc.
Annually
Collector
Inspect for loose bolts.
Check for corrosion and material
buildup and check V-belt drives
and chains for tension and wear.
Check accuracy of all indicating
equipment.
Check baffle plate for wear; if
appreciable wear is evident,
replace. Check for dust
deposits.
Check all gaskets
Tube type (tube hooks suspended
from a tubular assembly):
inspect nylon bushings in shaker
bars and clevis (hanger)
assembly for wear.
Channel shakers (tube hooks
suspended from a channel bar
assembly): inspect drill
bushings in tie bars, shaker
bars, and connecting rods for
wear.
Lubricate all electric motors,
speed reducers, exhaust and
reverse air fans, and similar
equipment.
Check all bolts and welds.
Inspect entire collector
thoroughly, clean, and touch up
paint where necessary.
118
-------
TABLE 14. TYPICAL FABRIC FILTER INSPECTION AND MAINTENANCE PROCEDURES, USER RESPONSES
Inspection or maintenance procedure
Frequency3
Semi-
Daily Weekly Monthly Quarterly annually Yearly
1. Duct inspection for dust settlement
(and removal if required)
2. Bag compartment—painting, patching,
regasketing, recaulking
3. Supporting structure and framing—
painting, bolt tightening, weld
inspection
4. Heating equipment for baghouse—
operating temperature, temperature
sensors and controls, insulation
5. Fan drives, electrical motors, gear
trains—checks for wear, lubrication
6. Repressure fan (for reverse air
cleaning) check for temperature and
vibration
7. Compartment isolation and bypass
dampers—mechanical drive, jamming,
leakage
8. Bag (fabric) inspection, clean side-
wear, damage, dust seepage, surface
deposits, bag clamps
X
(continued)
-------
TABLE 14 (continued)
Inspection or maintenance procedure
Frequency3
Semi-
Daily Weekly Monthly Quarterly annually Yearly
9. Lubrication of damper valves, shaker
mechanisms
10. Access doors—jamming, insulation
sealing, hinges, safety locks
11. Hopper—dust level, bridging, transfer
valve, material balance on dust
recovery
12. Bag suspensions—frames, drive attach-
ment hooks, tension adjustment system
13. Baghouse—check for floor deposits,
gross leakage
14. Control panel—indicator lights, switch
position, dial displays
15. Baffle plate and guide vanes—erosion
or dust deposition
16. Shaking mechanisms—linkage, slack,
amplitude or frequency adjustment
17. Compressed air supply—pressure, oil
and water separators
(continued)
-------
TABLE 14 (continued)
Inspection or maintenance procedure
Frequency3
Semi-
Daily Weekly Monthly Quarterly annually Yearly
18. Fire detectors, sprinklers, explosion
panels
19. Instrumentation—direct display,
continuous monitors
a. Baghouse temperature
b. Overall baghouse pressure loss
c. Individual compartment pressure
loss
d. Fan static pressure
20. Stack emissions--visual estimates,
automatic in-stack opacity meters
21. Fabric cleaning controls—timing
circuits, pressure sensors
X
X
X
X
X
a"X" may indicate one or more responses.
-------
preparation for the inspection, (b) whether or not the inspection is announced
to the plant beforehand, and (c) the effort and/or time required for
conducting it.
Compliance-type inspections (which provide preliminary assessments only
since source sampling is the ultimate basis for determining compliance with
the emission standard) should usually be unannounced so that the plant can be
evaluated under its normal operating conditions. Such indicators as fuel use,
continuous monitoring facilities, and visible emissions, that can be estimated
during the unannounced inspection, often provide a definitive picture of a
plant's manufacturing and fabric filter operations provided that there are no
plant access problems or faulty and/or inoperative equipment.
For other inspections pertaining to source registration, construction,
plan approval, or permit renewal, the plant should be given sufficient advance
notice so that qualified plant personnel can be present to provide the
drawings, manuals, and process information that might be required. Prior
notice should also be given when performing inspections for special studies
designed to document (a) operating and maintenance practices or (b) process
and emission data so that the required raw material rates, productivity
levels, and stack test information can be readily available. Regardless of
the type of inspection to be conducted, pertinent supporting information
should be obtained prior to, during, and following the source evaluation.
4.3.2 Procedures
Before conducting any plant inspection, the air pollution control
investigator should become thoroughly familiar with the process in question
and all agency file data currently on record concerning past inspections,
complaints, process upsets, excess emission episodes, or other conferences,
submittals, variance requests, or pending regulatory actions concerning the
facility. The air pollution engineer or inspector should prepare whatever is
necessary in the form of process flowsheets, chronological source history,
operating permits, and inspection checklists so that the actual inspection can
be made in a concise and orderly manner without overburdening the source
personnel. This will serve to promote a good working relationship between the
facility and the agency which, in turn, may foster better operating practices
on the part of the plant. Based upon knowledge of the process and the type of
inspection planned, the inspector may want to evaluate the source during
periods of both average and maximum rated throughput BO that representative
information can be obtained. The evaluation period, which may depend on the
time of day, week, or season, may indicate whether advance notice should be
given the plant. The type of safety equipment, protective clothing, and
measuring or sampling devices required by the inspector will also depend upon
the type of inspection. The inspector should know whether hard hats, steel
tipped shoes, safety glasses, ear protectors, or special clothing are required
and whether they can be provided by the plant. It may also be necessary to
provide special sampling or analyzing equipment (manometers, temperature
sensors, gas analyzers, etc.) for evaluating process streams or baghouse inlet
and outlet conditions.
122
-------
4.3.3 Inspection Sequence
Upon arrival at the general plant location (but probably beyond plant
boundaries), the stack(s) should be observed for visible emissions. With
properly operating fabric filters, no visible emissions (less than 5 percent
opacity) should be the rule. Periodic puffs may indicate a defective cleaning
system, insufficient corapartmentalization, too small an interval between
cleanings or a defective bag in one compartment. Depending on the plume
appearance and whether the inspector is a certified smoke reader, a detailed
evaluation by EPA Method 9 may be justified. In addition to monitoring any
smoke emissions, the plant environs should be checked for fugitive emissions,
dust deposits, or damage to vegetation, any of which may indicate nonoperative
or malfunctioning baghouse equipment. After the plant surroundings have been
observed for whatever length of time is deemed necessary (10 to 30 minutes or
more), the inspector should enter the property.
Once inside the plant, the inspector may be requested to sign one or more
forms concerning waivers of liability or confidentiality of process data.
While confidentiality agreements, (which are frequently encountered), should
be handled according to agency policy and procedures, insurance waivers or
other forms that restrict access to certain parts of the plant should not be
signed since the inspector has a legal right of entry to all areas under
Section 114 of the Clean Air Act. Problems with respect to signing releases
or other documents should be referred to Agency (EPA) attorneys.
The purpose and type of the proposed inspection, coupled with the
observations made while outside the facility, will suggest whether the process
equipment or the baghouse should be inspected first. Reasons for proceeding
immediately to the fabric collector would be as follows: observations outside
the plant indicate significant stack or fugitive emissions; the inspection is
a strict, unannounced compliance-type; plant personnel are suspected of
tampering with several facets of the fabric filter operation; or the
inspector, because of numerous previous visits to the plant, is very familiar
with the process operation, and does not intend to do any extractive sampling.
Reasons for beginning the inspection with the process equipment would be
the following: an announced, noncompliance-type inspection is being performed
and the inspector is assured that the baghouse will be completely operational,
and the inspector is unfamiliar with some process details and would like to
observe it before seeing the baghouse in operation.
Regardless of where the inspection begins, once inside the plant (the
stack and plant surroundings should always be evaluated prior to entering a
facility), the inspector should proceed in a logical sequence from one end to
the other. In this manner, the inspector can grasp the total picture and thus
have a better appreciation of potential problems, their causes, and possible
solutions. Various items to look for during the process stream inspection are
noted in the following paragraphs.
4.3.3.1 Process Equipment--
With respect to the industrial or combustion process of interest, the
following information should be sought: Raw materials—types, composition,
feed rates, toxicity or hazard potential, dust generating capacity, etc;
123
-------
process type and equipment—method of mixing, reaction conditions
(temperature, pressure, etc.), emission potential (emission factors), use of
pressurized reaction vessels, kilns, dryers, furnaces, etc; and
product—output rate, process monitors.
4.3.3.2 Dust Capture and Transport System—
Use of movable or stationary hoods, capture velocities, dust
accumulation, static pressure, condition of cleanout traps, integrity of
ductwork, length of runs to baghouse, and leaks or fugitive dust emissions
should all be evaluated.
4.3.3.3 Fabric Filter System—
• Parameter monitors—including opacity or broken bag detectors;
manometers for pressure drop across fabric, compartments, or entire
collector; indicators for cleaning sequence, cycle time,
compartments off-line, temperature, volume flow, air-to-cloth ratio,
moisture, pulse-jet header pressure, and reverse air flow.
• Baghouse exterior—cleaning system operation; cleaning method;
overall condition of exterior housing, including structural members,
access doors and gaskets, reverse air fan operation, and shaker
mechanism. Visual evidence of corrosion; warping of panels; faulty
or missing gasketing; loose bolts; and noise, odor, or elevated
temperatures as indicators of worn bearings, overstressed fan belts
and electric motor problems are identifiable by external inspection.
Baghouse interior (if interior inspection is deemed necessary and is
feasible)—condition of bags: tears, pinholes, and sagging
(inadequate tension). In the case of a sagging or slack bag, a
folding of the bag over the bottom thimble connection creates a
pocket in which accumulated dust can rapidly abrade and tear the
fabric. Slackness also prevents effective cleaning action with both
reverse flow or mechanical shaking systems. Dust seepage or
bleeding and/or pinhole leaks are evidenced by dust deposits on the
"clean" side of the fabric. Staining and stiffening of the dirty
fabric indicates excessive caking caused by moisture condensation or
chemical reactions. The latter condition leads to fabric blinding
and excessive pressure loss as well as to fabric failure. More than
a 1/4 inch dust layer on floor plates or isolated piles of dust
suggest excess seepage and/or torn or missing bags.
Inspection of the inlet plenum, including bag interior, will reveal
any excess dust buildup on bags and distribution plates. If the
amount of dust on a bag after cleaning is more than twice the weight
of the new (unused) bag, insufficient cleaning is indicated. The
condition of solenoid valves, poppet valves, mechanical linkages,
bag clamps are also indicated.
124
-------
• Special attention is directed to the many possible safety hazards
associated with interior baghouse inspections. These may include,
but are not limited to, asphyxiants, high concentrations of toxics,
high air temperatures, hot surfaces, and numerous physical hazards.
In the latter category are mechanical obstacles, contact with sharp
objects, falling objects, electrical shock due to exposed wires or
improper grounding, etc. all of which are accenuated because of poor
illumination within most baghouses. General EPA policy is to
strongly advise against entering baghouses unless absolutely
necessary. Given the latter situation, the inspector should (a) be
accompanied by an experienced person who is fully aware of the
potential hazards and (b) observe all safety regulations with
respect to working in confined spaces, wear protective clothing and
respiration equipment (SCBA or air supply and work with lifeline)
and outside partners with proper tagging for "inspection-in-progress"
and with all machinery and dampers locked into the "off" mode while
the baghouse interior is occupied.
4.3.3.4 Fan-
Location—clean or dirty side—static pressure, rotation rate, amperage,
housing temperature and condition, excessive noise or vibration, condition of
belts.
4.3.3.5 Dust Disposal System—
Condition of hoppers, presence of insulation, level indicators,
vibrators, adequacy of dust removal, fate of dust.
Various inspection locations in the previous five areas may or may not be
required depending on overall plant conditions, type of inspection, whether
the plant is operating, or if extractive sampling is to be performed. For
example, unless absolutely necessary, the plant should not be requested to
shut their process/fabric filter system down so that internal inspections of
the fan and baghouse can be made. If an internal inspection is required, it
is usually advisable to wait for a scheduled plant shutdown.
Aside from performing inspections of process and control equipment and
their methods of operation, the air pollution control officer should also be
concerned about the long-term effects of equipment operation and maintenance
for which regulations may or may not be in force. (Some states have
regulations that require submittal and approval of standard operating and
maintenance procedures for all control equipment.) The inspector
investigating these procedures should request to see equipment operation and
maintenance logs to evaluate such entries as location of failed bags, bag
replacement frequency, cause of bag failure (blowout, tears, chemical attack,
abrasion, pinholes), where on the bag failure usually occurs (top, middle, at
the cuff, along the seam), or other types of recurring problems with the
system that may be suggestive of a trend that should be carefully watched. It
should also be noted whether spare parts (bags and associated accessories such
as clamps, J-hooks, springs, rings, cages, etc., and fan belts motors,
gaskets, bearings, solenoid valves, etc.) are kept on hand or whether the
plant must order new parts each time service is required. Accurate
recordkeeping, which is a vital part of a good preventive maintenance program,
125
-------
can greatly aid the baghouse manufacturer in solving field problems. Although
it is usually recommended that the user follow the manufacturer's guidelines
regarding inspection and maintenance frequency, it is not uncommon to find
plant logs reflecting deviations from vendor-recommended practice after the
system has been operated for a sufficient period of time. In some cases, it
may be advisable to consult with firms providing fabric analysis services as a
troubleshooting measure. Many such groups are listed in consultant directory
sections of the Journal of the Air Pollution Control Association and Pollution
Engineering.
4.4 INSPECTION REPORT FORMAT
The specific format of an inspection report designed to record all of the
previously mentioned information would be formidable indeed. In an effort to
minimize the length of such documents, some states have devised one-page forms
for recording only the most important information. One State, for example,
uses the form shown in Table 15 to record pertinent source information
obtained during a compliance inspection. It is not likely, however, that a
single page form will provide sufficient source information to fully document
process and control equipment operation for later referral or emergency use.
As a compromise between an overly lengthy inspection form and the one-page
form acceptable for certain compliance investigations, the format provided in
Table 16 is suggested. This format presents information in a logical sequence
that can be modified so that more detailed information about the process, the
baghouse, the fan, and the operating and maintenance aspects can be provided
by appending additional pages to individual sections. In this manner,
different portions of the form can be completed at different times, either at
the agency office or at the plant. Specific pages can be inserted, detached,
and/or replaced as the situation changes. Whatever type of form is adopted,
it should be concise, while still allowing the recording of detailed
information when required.
4.5 EXAMPLE INSPECTION REPORT
Based upon an earlier plant inspection for compliance purposes, we have
presented results of actual field observations at a gray iron foundry in which
the performance of two fabric filter systems was appraised, Table 17.
Although more data would have been preferred, the responses typify the
information needed to assess the adequacy of existing air cleaning equipment
and to point out system deficiencies requiring correction. Data that might
reveal plant identity including name, geographical location and product line
have been altered to maintain confidentiality. Although field observations
suggested compliance with visible emissions criteria, it appeared that the
plant's intention to install an inertial precleaning device might help to
extend bag service life by preventing large, abrasive particles from reaching
the bag surfaces.
126
-------
TABLE 15. EXAMPLE OF SINGLE PAGE STATE COMPLIANCE INSPECTION REPORT FORMAT
Print »n btSCk ''
. CtWWf MM.
2. Ctms? MWttS
>UWT MJOSESS
4. njurr siu OKWOR DKIBOS DOTH^R
s. KsocEis VM KWUCT rm
of
rtu f
AIR QUALITY COMPLIANCE INSPECTION
vtsn «J.
13. insncrioa aw
14. lEStCTlW B*TE
is. TOS tmm
u. «a) SPEED a
17. SSY CCW1TKMS
1*. IWECTlCa STATUS
I-WOT fH (
2- IK CO«PIU<£-TEST
6-«01 0« !
3- |« CtMPL . -c ICU3
ii
•TEST
•i«S>.
5E3
COHTmi iiyjimiltl C00£
KS HICK iff sceuseEft
6F 5ACf 1 ITER
EP ELECTRO. f»ECiP.
CT CtCLOnS
n/E-«C I E'
rt/e-K; t tit
C/S-C 1 '.CRUUU
«-t«0 0«TROt
,*E T,OH COO! '
;-»<>j IKI IMS*
J-SIA * Its: 085
^ M . Krrn« ^CMTvaf MTt-
1 JO. DIVISION WMBVAl St teIE:
mtr
3
4
b
—
-------
TABLE 16. SUGGESTED FABRIC FILTER FIELD INSPECTION REPORT FORMAT*
Source ID No. SIC
Inspector(s) Date
Inspection Announced?
A. GENERAL PLANT DATA FROM AGENCY FILE
1. Plant name, address, and phone number
2. Name of plant contact, title, and phone number
3. Type of process
4. Allowable emission rate and opacity
5. Date baghouse installation approved
6. Prior complaints or episodes of excess emissions
7. Last inspection date
8. Purpose of inspection
B. GENERAL OBSERVATIONS PRIOR TO ACTUAL INSPECTION
1. Weather conditions
2. Appearance of visible emissions (if any)
'3. Is inspector a certified smoke reader? Yes No If yes, give
certification date
(Attach copy of Method 9, if performed)
128
-------
TABLE 16 (continued)
C. PROCESS INFORMATION
1. Confidential? Yes No
2. Person supplying process information and title
3. Product(s) produced
4. Production rate(s)
5. Raw materials used
6. Portion of process controlled by baghouse
7. Average uncontrolled emission rate or concentration (indicate whether
obtained from stack test, mass balance, AP-42 emission factor, other,
etc.)
8. Date of last stack test and average emission rate obtained
9. Is cleaned effluent recirculated back into plant? Yes No
D. DUST CHARACTERISTICS (PRIOR TO CONTROL)
1. Is materal toxic or otherwise hazardous or does it require special
handling? Yes No Describe
2. Moisture content or other gaseous constituents
3. Abrasiveness or other properties
4. Particle size data - indicate how measured
129
-------
TABLE 16 (continued)
COLLECTION SYSTEM(S)
1. Baghouse #1 #2 #3
a. Manufacturer
b. Type or trade name
c. Model No.
d. No. of compartments
e. Bags/compartment
f. Bag 1 x d
g. Total Cloth Area
Fan #1 #2 #3
a. Manufacturer
b. Model No.
c. Blade type
d. Belt or direct drive
e. Power rating
f. Positive or negative
pressure
Fabric #1 #2
a. Manufacturer
b. Material
c. Woven or felted
d. Weave
e. Weight
f. Permeability
g. Operating temp, range
h. Surface treatment
i. Coating upon startup
j. Guaranteed life
k. Actual life
4. Cleaning System #1 #2 #3
a. Method
b. Freqnency
c. Actuated by
d. Anticollapse rings
e. Wire mesh cages
130
-------
TABLE 16 (continued)
F. DUST HANDLING SYSTEM(S)
1. Do baghouse hoppers have:
a. Heaters
b. Insulation
c. Level indicators
d» vibrators
2. Type of dust transport system
3. Fate of collected material
G. INSTRUMENTATION
Do system monitors record any of the following:
1. Process start-up/shutdown
2. System flow or velocity
3. Fan motor amps __^____
4. Temperature (recording?)
5. Pressure
6. Opacity
7. Outlet emissions
8. Compartments off-line
9. Compartments being cleaned
10. Compartments in operation
11. Other
H. OPERATING PARAMETERS - DESIGN AND ACTUAL
Design Actual
1. Flow rate
Pressure drop,
flange-to-flange
measurement location
131
-------
TABLE 16 (continued)
3. A/C, gross
4. A/C, net
(2 compartments off-line)
5. Temperature (Baghouse)
6. Efficiency
7. Emission rate
8. Opacity
I. OPERATING EXPERIENCE/MAINTENANCE ASPECTS
1. Percent of time baghouse fully operational when process is in opera-
tion
2. Has a detailed maintenance schedule been instituted?
3. Is maintenance schedule as recommended by baghouse manufacturer or by
plant?
4. Are maintenance records available for inspection?
5. How long are records kept on file?
132
-------
TABLE 16 (continued)
6. Which of the following problem areas have led to periods of excess
emissions or caused the process to be shut down?
Problem Area Duration Frequency
a. Insufficient dust pick-
up and/or transport
(fugitive emissions)
b. Duct abrasion or corrosion
c. Temperature excursions,
high or low
d. Moisture
e. Fan abrasion, vibration, etc.
f. Gross bag failure
g. Inadequate bag tension
h. Bag chafing or abrasion
i« Pressure loss
j. Compartment isolation dampers
k. Cleaning mechanism
1. Visible emissions
m. Plugged hoppers
n. Hopper fires
o. Dust discharge system_
CONCLUSIONS/RECOMMENDATION
1. Compliance status
2. Need for further action
3. Corrective actions to be taken
133
-------
TABLE 16 (continued)
4. Time required to rectify problems
5. Special waivers or review of compliance criteria required
6. Need for follow-up inspection
7. Inspector's signature
date
approved by
title
K. OTHER NOTES, COMMENTS, SKETCHES (ATTACH ADDITIONAL PAGES, IF NECESSARY)
Schematic drawings showing locations of process and dust control
equipment should be prepared, particularly so, where verbal descriptions
may lead to misunderstandings. Locations should be noted for observed
leak sites, evidence of corrosion, warped panels, and other mechanical
defects.
134
-------
TABLE 17. EXAMPLE FABRIC FILTER FIELD INSPECTION REPORT
Source ID No. 2248 SIC 3531
Inspector(s) John Smith Date 12/14/82
Inspection Announced? Yes
A. GENERAL PLANT DATA FROM AGENCY FILE
1. Source name, address, and phone number
XYZ Casting Co.
2. Name of plant contact, title, and phone number Robert Smith,
Vice President of Operations
3. Type of process Gray iron foundry
4. Allowable emission rate and opacity
cupola - 41 Ib/hr and 25 percent
2-shot blast systems - 14 Ib/hr and 40 percent
10 grinding booths - 5 Ib/hr and 40 percent
5. Date baghouse installation approved
System 1 - 9-14-78
System 2 - 6-12-75
6. Prior complaints or episodes of excess emissions
None
7. Last inspection date 10-12-81
8. Purpose of inspection to document operating and maintenance
practices
GENERAL OBSERVATIONS PRIOR TO ACTUAL INSPECTION
1. Weather conditions clear sky, wind 5-10 mph
2. Visible emissions periodic puffs, overall opacity less than
10 percent
3. Is inspector a certified smoke reader? Yes X No If yes, give
certification date 11/17/82
135
-------
TABLE 17 (continued)
C. PROCESS INFORMATION
1. Confidential? Yes No X
2. Person supplying process information and title John Doe,
Plant Engineer
3. Product(s) produced heavy machinery castings
A. Production rate(s) cupola - 30 ton/hr, shot blast - 16 ton/hr,
grinding booth - 16 ton/hr
5. Raw materials used foundry returns, scrap metal, pig iron, coke,
and limestone
6. Portion of process controlled by baghouse
System 1 - Shot blasting (two chambers)
System 2 - Grinding (ten booths)
7. Average uncontrolled emission rate or concentration (indicate whether
obtained from stack test, mass balance, AP-42 emission factor, other,
etc.) (See Notes 1, 2, and 3 at end of report)
Grinding booths - 64 Ib/hr
Shot blast system - 364.25 Ib/hr
8. Date of last stack test and average emission rate obtained
No previous sampling measurements
9. Is cleaned effluent recirculated back into plant? Yes No X
D. DUST CHARACTERISTICS (PRIOR TO CONTROL)
1. Is materal toxic or otherwise hazardous or does it require special
handling? Yes No X Describe
Moisture content or other gaseous constituents
low moisture
Abrasiveness or other properties
both systems — highly abrasive
Particle size data - indicate how measured
MMD 2 m (estimate)
136
-------
TABLE 17 (continued)
E. COLLECTION SYSTEM(S)
1. Baghouse
#1
3.
a. Manufacturer
Pangborn Wheelabrator-Frye
b. Type or trade name
c. Model No.
d. No. of compartments
e. Bags /compartment
f. Bag 1 x d
g. Total Cloth Area
Fan
a. Manufacturer
10 15
200 101
12' x 8" ID 12' x 6" ID
50,265 ft2 28,571 ft2
#1 #2 #3
Buffalo-Forge
b. Model No.
c. Blade type
d. Belt or direct drive
radial
both belt
e. Power rating
f. Positive or negative
both negative
pressure
Fabric
a. Manufacturer
b. Material
c. Woven or felted
d. Weave
e. Weight
f. Permeability
g. Operating temp, range
h. Surface treatment
i. Coating upon startup
j. Guaranteed life
k. Actual life
Cleaning System
a. Method
b . Frequency
c. Actuated by
d. Anticollapse rings
e. Wire mesh cages
#1 #2 #3
Globe Albany W. W. Cri swell
dacron glass
felt woven
(needled) 2/2 twill
16 oz/yd2 10 oz/yd2
32 cfm/ft2 @ 0.5. in W.C.
250°F 500°F
none none
none none
2 years 1-2 years
2-k years 1-2 years
*1 #2 #3
pulse jet shake
every 3 min every 4 hours
timer pressure
no no
yes no
137
-------
TABLE 17 (continued)
F. DUST HANDLING SYSTEM(S)
1. Do baghouse hoppers have -
a. Heaters none
b. Insulation none
c. Level indicators both systems
d. vibrators both systems
2. Type of dust transport system pneumatic
3. Fate of collected material recycled to respective processes
G. INSTRUMENTATION
Do system monitors record any of the following:
1. Process start-up/shutdown both systems
2. System flow or velocity System 2 only
3. Fan motor amps both systems
4. Temperature (recording?) both
5. Pressure both
6. Opacity both
7. Outlet emissions none
8. Compartments off-line System 2 only
9. Compartments being cleaned both
10. Compartments in operation both
11. Other
H. OPERATING PARAMETERS - DESIGN AND ACTUAL
Design Actual
1. Flow rate System 1 - 151,000 acfm 130 - 150.000
System 2 - 200.000 acfm 150 - 175,000
2. Pressure drop, #1 - 4-6 in. W.C. 5-7 in. W.C.
flange-to-flange #2 - 4-6 in. W.C. 3-5 in. W.C.
measurement location
138
-------
TABLE 17 (continued)
3. A/C, gross
Design
System 1 - 7/1
System 2 - 3/1
Actual
5-6/1
2-3/1
4. A/C, net
(2 comp. down)
#1 - 8/1
#2 - 275/1
same
same
5. Temperature
#1 250°F
#2 500°F
200-240°F
400°F
6. Efficiency
both - >99%
unknown
7. Emission rate
both <0.002 gr/scf
unknown
8. Opacity
both <20%
'5%
I. OPERATING EXPERIENCE/MAINTENANCE ASPECTS
1. Percent of time baghouse fully operational when process is in opera-
tion _Bot_h systems have operated at >90% availability
Has a detailed maintenance schedule been instituted?
Yes - for both systems
Is maintenance schedule as recommended by baghouse manufacturer or by
plant? Maintenance practices originally recommended by baghouse
manufacturers have since been modified by plant based on experience.
4. Are maintenance records available for inspection? Yes
5. How long are records kept on file? 5 years
139
-------
TABLE 17 (continued)
6. Which of the following problem areas have, led to periods of excess
emissions or caused the process to be shut down?
Problem Area Duration Frequency
a. Insufficient dust pick-
up and/or transport
(fugitive emissions)
b. Duct abrasion
c. Temperature
high or low
d. Moisture
System 1 only numerous occasions
Fan abrasion, vibration, etc. System 1 only fairly often
g-
h.
i.
j-
k.
1.
m.
n.
o.
(abrasion)
3-4 times/yr
Gross bag failure System 1 only several times/yr
several days
before bags
replaced
Inadequate bag tension
Bag chafing or abrasion
Pressure loss
System 2
infrequent
Compartment isolation dampers up to 8-9 in. W.C.
Cleaning mechanism
Visible emissions
Plugged hoppers
Hopper fires
Dust discharge system
J. CONCLUSIONS/RECOMMENDATION
1. Compliance Status in compliance with visjLble emissions and operating
and maintenance practices ^^
2. Need for further action
none
3. Corrective actions to be taken Plant intends to install cyclone
upstream of Baghouse #1 to reduce fabric abrasion by coarse
particles
140
-------
TABLE 17 (continued)
4. Time required to rectify problems about 6 months
5. Special waivers or review of compliance criteria required
None
6. Need for follow-up inspection
Reschedule inspection when cyclone is installed.
7. Inspector's signature
date 12/16/82
approved by
title Chief, Air Control Board
K. OTHER NOTES, COMMENTS, SKETCHES (ATTACH ADDITIONAL PAGES, IF NECESSARY)
The overall inspection revealed that the plant had maintained the bag-
houses in good condition, although abrasion with respect to the baghouse
collecting emissions from the shot blasting operations (metal particles,
sand, and broken shot) has been a problem. The plant intends to install
a mechanical collector upstream of the fabric filter to capture most
of the larger abrasive material. Recordkeeping at the plant was excel-
lent insofar as baghouse operation and maintenance are concerned.
Note 1: Emissions factors for shot blasting and grinding at an iron
foundry were obtained from: Gutow, Bernard S. "An Inventory
of Iron Foundry Emissions." Modern Casting. January 1972,
pp. 46-48. (Emission Factors for these two operations are
not provided in AP-42).
Note 2: Shot Blasting—
23.5 ton/hr x 15.5 Ib/ton = 364.25 Ibhr
Note 3; Grinding—
40 ton/hr x 1.6 Ib/ton iron = 64 lb hr
141
-------
REFERENCES
1. Billings, C. E., and J. E. Wilder. Handbook of Fabric Filter Technology,
Vol. 1. GCA/Technology Division, Bedford, Massachusetts 01730.
EPA-APTD 0690 (NTIS PB-200-648), December, 1970.
2. linoya, K., and C. Orr, Jr. Filtration. In: Air Pollution. Vol. IV,
3rd Ed., A. C. Stern, ed. Academic Press, New York, New York, 1977. pp.
155.
3. Strauss, W. Industrial Gas Cleaning, 2nd Ed. Pergamon Press, Sydney,
London, New York, 1975.
4. Danielson, John A., ed. Air Pollution Engineering Manual. Publication
No. AP-40, U.S. Environmental Protection Agency, Office of Air and Water
Programs, Research Triangle Park, North Carolina 27711, May, 1973.
5. . Fine Particulate Fabric Filtration: A
Proceedings. J. Air Poll. Control Assoc., 24:1140-1197, 1974.
6. Stern, A. C., ed. Air Pollution, Volume III, Measuring, Monitoring and
Surveillance of Air Pollution. 3rd Ed., Part A. Academic Press, Inc.,
New York, San Francisco, London, 1976.
7. Stern, A. C., ed. Air Pollution, Volume IV, Engineering Control of Air
Pollution. 3rd Ed., Parts A and C Academic Press, Inc., New York,
San Francisco, London, 1976.
8. Dorman, R. G. Filtration. In: Aerosol Science, C. N. Davies, ed.
Academic Press, London and New York, 1966, pp. 192.
9. White, H. J. Industrial Electrostatic Precipitation. Addison-Wesley
Publishing Company, Reading, Massachusetts, 1963.
10. . Fine Particle Electrostatic Precipitaion: A Proceedings.
J. Air Poll. Control Assoc., 25_:926-983, 1975.
11. Oglesby, S., and G. B. Nichols. A Manual of Electrostatic Precipitator
Technology, Vols. I and II. Southern Research Institute, Birmingham,
Alabama. APTD 0160, 0161 (NTIS PB-196-380, PB-197-381), August 1970.
12. . Fine Particle Scrubbing: A Proceedings. J. Air Poll.
Control Assoc., 24:98-189, 1974.
142
-------
13. Calvert, S. , J. Goldshmid, D. Leith, and D. Mehta, Wet Scrubber Handbook,
Vol. 1. A.P.T., Inc., Riverside, California 92502. EPA-R2-72-118A (NTIS
PB-213-016), July 1972.
14. Sherwood, T. K. and R. L. Pigford. Absorption and Extraction. McGraw
Hill Book Company, Inc., New York, 1952.
15. Claes, J. and R. DeBruyne. Demisting with Metal Fibre Webs and Felts.
Filtration and Separation, 13(5):494, 1976.
16. Carman, P. C. Flow of Gases through Porous Media. Academic Press,
New York, 1956.
17. Dennis, R., and C. E. Billings. Blast Effects on an Air Cleaning
System. Operation Plumbbob, Project 3414. Harvard University,
Cambridge, Mass, USAEC Civil Effects Test Group, WT-1475, October 1962.
18. Silverraan, L. and R. Dennis. Pollen Removal by Air Filters. Air Cond.
Heat, and Vent., 56^:61, 1959.
19. Clarenburg, L. A., and H. W. Pickaar. Aerosol Filters - I - Theory of
the Pressure Drop Across Multi Component Glass Fibre Filters. Chem. Eng.
Sci., 23^:773, 1968.
20. Stern, S. C., H. W. Zeller, and A. I. Schekman. The Aerosol Efficiency
and Pressure Drop of a Fibrous Filter at Reduced Pressures. J. Colloid
Sci., 1_5:6, 546, 1960.
21. Chen, C. Y. Filtration of Aerosols by Fibrous Media. Chem. Rev., 55:3,
595, 1951.
22. Dennis, R., ed. Handbook on Aerosols. U.S. Energy and Development
Administration. TID-26608, NTIS, Springfield, Va. 1976.
23. Dyment, J. What future for HEPA Filtration. Filtration and Separation,
14(1):44, 1977.
24. Fuchs, N. A. The Mechanics of Aerosols. C. N. Davies, ed. The MacMillan
Company, New York. 1964.
25. Davies, C. N., ed. Aerosol Science. Academic Press, London and
New York, 1966.
26. Green, H. L. and W. R. Lane. Particulate Clouds, Dusts, Smokes and
Mists, 2nd Ed. C. & F. N. Spon Ltd., London, 1964.
27. Stephan, D., G. Walsh and R. Herrick. Concepts in Fabric Air
Filtration. Am. Ind. Hyg. Assoc. J, _21_:1, 1960.
28. Rothwell, E. Fabric Dust Filtration: Principles and Practice.
Filtration and Separation. 17(5):471, 1980.
143
-------
29. Rothwell, E. "Concepts of Fabric Dust Filtration." Filtration and
Separation, ^3(5):!, 1976.
30. Rothwell, E. The Collection of Fine Particulates in Fabric Filters. The
Chemical Engineer (Brit.), Issue No. 329, February, 1978.
31. Dennis, R., R. R. Hall, and N. F. Surprenant. Fibers, Fabrics, Face
Velocity and Filtration. In: Proceedings: The User and Fabric
Filtration Equipment III, Buffalo, New York, October, 1978, pp 156-169.
32. Ranz, W. E. and J. B. Wong. Impaction of Dust and Smoke Particles on
Surface and Body Collectors. Ind. Eng. Chem., 44: 1371-1381, 1952.
33. Ellenbecker, M. J., D. Leith and J. M. Price. Impaction and Particle
Bounce at High Stokes Numbers. J. Air Poll. Control Assoc., 30(11):1224,
1980.
34. Payatakes, A. C. Dendrites and Air Filter Efficiency. Filtration and
Separation, 13(6):602, 1976.
35. Schrag, M. P., and L. J. Shannon. Evaluation of Electric Field Fabric
Filtration, EPA-600/2-76-041, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., 1976, 25 pp.
36. McDonald, J. R., R. B. Mosley and L. E. Sparks. An Approach for
Describing Electrical Characteristics of Precipitated Dust Layers.
J. Air Poll. Control Assoc., 3£(4):372, 1980.
37. Frederick, E. R. Fibers, Electrostatics, and Filtration: A Review of
New Technology. J. Air Poll. Control Assoc., :W(4):426, 1980.
38. Penney, G. W. Using Electrostatic Forces to Reduce Pressure Drop in
Fabric Filters. Powder Tech, _18_:111, 1971.
39. Penney, G. W. Electrostatic Effects in Fabric Filtration, Vol. I,
Fields, Fabrics, and Particles (Annotated Data). EPA-600/7-78-142a,
September, 1978.
40. Lembach, R. F. and G. W. Penney. Nodular" Deposits in Fabric Filters. J.
Air Poll. Control Assoc., 2^(8):823, 1979.
41. Ariman, T. and D. J. Helfritch. Pressure Drop in Electrostatic Fabric
Filtration. In: Proceedings of the Second Symposium on the Transfer and
Utilization of Particulate Control Technology. Vol. Ill, Particulate
Control Devices. EPA-600/9-80-039c, 1980.
42. Dennis, R., and N. F. Surprenant. Research on Fabric Filtration
Technology - Particulate Control Highlights. EPA-600/8-78-005a, June,
1978.
144
-------
43. Rushston, Albert and Alan Rushston. Size and Concentration Effects in
Filter Cloth Pore Bridging. Filtration and Separation, 9/3):274, 1972.
44. Carman, P. C. Flow of Gases through Porous Media. Academic Press,
New York, 1956.
45. Rothwell, E. and G. A. Trimble. Precoat Cake Filtration for the
Collection of Particulate Fumes. Filtration and Separation, 16(3):279,
1979.
46. Dennis, R., and H. A. Klemra. A Model for Coal Fly Ash Filtration.
J. Air Poll. Control Assoc., ^9(3):230, 1979.
47. Dennis, R. , D. V. Bubenick, and L. S. Hovis. Diagnosing Filter Fabric
Capabilities with Light Scattering and Nucli Detecting Instrumentation.
Filtration and Separation, 19^206-210, 1982.
48. Dennis, R. et al. Filtration Model for Coal Fly Ash with Glass Fabrics.
EPA-600/7-77-084, (NTIS PB-276-313/4BE), August, 1977.
49. Dennis R, and H. A. Klemm. Recent Concepts Describing Fabric Filter
System Behavior. Filtration and Separation, 19^22-24, 75, 1982.
50. Snyder, C. A. and R. T. Pring. Design Considerations in Filtration of
Hot Gases. Ind. Eng. Chem. Process Des. Dev, 47^:960, 1955.
51. Thring, M. W. and W. Strauss. The Effect of High Temperatures on
Particle Collection Mechanisms. Trans. Inst. Chem. Engrs. 41^:248, 1963.
52. Spaite, P. W., D. G. Stephan, and A. H. Rose, Jr. High Temperature
Fabric Filtration of Industrial Gases. J. Air Poll. Control Assoc.,
17_:5, 1967.
53. Strauss, W. and B. W. Lancaster. Prediction of Effectiveness of Gas
Cleaning Methods at High Temperatures and Pressures. Atmos Environ.
2/.135, 1968.
54. S. Calvert and R. D. Parker. Effects of Temperature and Pressure on
Particle Collection Mechanisms: Theoretical Review. A.P.T., Inc.,
EPA-600/7-77-002, January, 1977.
55. Gurtler, H. C. and R. DeBruyne. The Economics of High Temperature
Filtration. Filtration and Separation, 14_(6):641, 1977.
56. Daniel, B. E., R. P. Donovan and J. H. Turner. EPA Fabric Filtration
Studies: Bag Cleaning Technology, (High Temperature Tests).
EPA-600/7-77-095b, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., 1977. 41 pp.
57. Ciliberti, D. F. High Temperature Particulate Control with Ceramic
Filters. EPA-600/2-77-207, Research Triangle Park, N.C., 1977. 171 pp.
145
-------
58. Happel, J. Viscous Flow in Multiparticle Systems: Slow Motion of Fluids
Relative to Beds of Spherical Particles, Am. Inst. Chem. Eng. J.,
4_: 197-201, 1958.
59. Scheidegger, A. E. The Physics of Flow Through Porous Media. University
of Toronto Press, Toronto, Canada, 1974.
60. Ariman, T. and D. F. Helfritch. How Relative Humidity Cuts Pressure Drop
in Fabric Filters. Filtration and Separation, 14_:127, 1977.
61. Durham, J. R. and R. E. Harrington. Influence of Relative Humidity on
Filtration Resistance and Efficiency. NAPCA, PHS, U.S. DHEW, AICHE 63rd
Annual Meeting paper, Chicago, 111., November 1970.
62. Spaite, P. W. and R. E. Harrington. Endurance of Fiberglass Filter
Fabrics. J. Air Poll. Control Assoc., V7_(5) :310-313, 1967.
63. McKenna, J. D., J. C. Mycock and W. 0. Lipscomb. Applying Fabric
Filtration to Coal-Fired Industrial Boilers - Pilot Scale Investigation.
EPA-650/2-74-058a, Research Triangle Park, 1975.
64. Stern, J. L. Dry Scrubbing for Flue Gas Desulfurization. Chem. Eng.
Progr., r7_:37-42, 1981.
65. . Determining the Concentration of Particulate Matter in
a Gas Stream-Performance Test Codes. ANSI/ASME PTC 38-1980. The
American Society of Mechanical Engineers, United Engineering Center,
345 East 4th Street, New York, New York 10017, 1980.
66. Munns, D. D. B. H. Practical Aspects of Sampling Particulate Emissions
to the Atmosphere. Filtration and Separation, _14_(6):656, 1977.
67. Technical Manual: A Survey of Equipment and Methods for Particulate
Sampling in Industrial Process Streams, Report No. EPA-600/7-78-043,
(PB-282-501/AS), March 1978.
68. Committee on Industrial Ventilation. Industrial Ventilation, A Manual of
Recommended Practice. 16th Ed. American Conference of Governmental
Industrial Hygienists, P. 0. Box 16153, Lansing, Michigan, 48901, 1979.
69. U.S. Environmental Protection Agency. Determination of Particulate
Emissions from Stationary Sources, Method 5. Federal Rgister, 41(111):
Tuesday, June 8, 1976.
70. Dennis, R., W. R. Samples, D. M. Anderson, and L. Silverman. "isokinetic
Sampling Probes. Ind. Eng. Chem. 4_9_:111, 1957.
71. Drinker, P., and T. Hatch. Industrial Dust, 2nd Ed., McGraw Hill
Book Co., Inc., New York, London, Toronto, 1954.
72. Orr, C., Jr., and J. M. Dalla Valle. Fine Particle Measurement. The
Macmillan Co., New York, 1960.
146
-------
73. Cadle, R. D. Particle Size. Reinhold Publishing Corporation, New York,
1965.
74, Silverman, L., M. W. First, and C. E. Billings. Particle Size Analysis
in Industrial Hygiene. Academic Press, New York and London, 1971.
75. McCain, J. D., A. N. Bird, and K. M. Cushing. Field Measurements of
Particle Size Distribution with Inertial Sizing Devices. Southern
Research Institute, EPA 650/2-73-035, 1973.
76. Smith, W. B., K. M. Cushing, G. E. Lacey, and J. D. McCain. Particle
Sizing Techniques for Control Device Evaluation. Southern Research
Institute. EPA-650/2-74-102a, 1975.
77. Calvert, S., C. Lake, and R. Parker. Cascade Impactor Calibration
Guidelines. A.P.T., Inc., Riverside, CA. EPA-600/2-76-118. April, 1976,
78. Harris, D. B. Procedures for Cascade Impactor Calibration and Operation
in Process Streams. EPA-600/2-77-004, January, 1977.
79. Mercer, T. T. and R. G. Stafford. Impaction from Round Jets. Ann.
Occup. Hyg., 1.2:41-48, 1969.
80. Willeke, K. and J. J. McFeters. The Influence of Flow Entry and
Collecting Surface on the Impaction Efficiency of Inertial Impactors.
J. Coll. and Interface Sci., _53:121, 1975.
81. McCrone, W. C., R. G. Draftz, and J. G. Delly. The Particle Atlas.
Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan, 1967.
82. Schrag, M. P., A. K. Rao, G. S. McMahon, and G. L. Johnson. Fine
Particle Emissions Information System Reference Manual.
EPA-600/2-76-173, Research Triangle Park, NC, June 1976.
83. Herdan, G. Small Particle Statistics, 2nd Rev. Ed. Academic Press,
Inc., New York, 1960.
84. Irani, R. R., and C. F. Callis. Particle Size: Measurement,
Interpretation, and Application. John Wiley & Sons, New York, 1963.
85. Raabe, 0. G. Aerosol Aerodynamic Size Conventions for Inertial Sampler
Calibrations. J. Air Poll. Control Assoc., 26(9):856, 1976.
86. Galeski, J. B. Particle Size Definitions for Particulate Data Analysis.
EPA-600/7-77-129, November, 1977.
87. Hatch, T. and S. P. Choate. Statistical Description of the Size
Properties of Non-Uniform Particulate Substances. J. Franklin Inst,
2_0_7:369, 1929.
88. Rabbe, 0. G. Particle Size Analysis Utilizing Grouped Data and the
Log-Normal Distribution. J. Aerosol Sci., ^(3), 289, 1971.
147
-------
89. Dalla Valle, J. M., C. Orr, Jr., and H. G. Blocker. Fitting Bimodal
Particle Size Distribution Curves. Ind. Eng. Chem. , 43^:1377-1380, 1951.
90. Kottler, F. J. The Distribution of Particle Sizes. J. Franklin Inst.,
250:339, 1950.
91. Kottler, F. J. The Goodness of Fit and the Distribution of Particle
Size. J. Franklin Inst., 2J[1:449, 1951.
92. Ettinger, H. J. and S. Posner. Evaluation of Particle Sizing and Aerosol
Sampling Techniques, Am. Ind. Hyg. Assoc. J., 26:17, 1965.
93. Marchello, J. M. and J. J. Kelley, eds. Gas Cleaning for Air Quality
Control. Marcel Dekker, Inc., New York, 1975.
94. Kraus, M. N. Baghouses: Selecting, Specifying and Testing Industrial
Dust Collectors. Chem. Eng., j$6(9): 133-142, 1979.
95. Kraus M. N. Baghouses: Separating and Collecting Industrial Dusts.
Chem. Eng., 8j>(8) 94-106, 1979.
96. Spaite, P. W., G. W. Walsh. Effect of Fabric Structure on Filter
Performance. Am. Ind. Assoc. J. 24_:357-365, 1963.
97. Draemel, D. C. Relationship Between Fabric Structure and Filtration
Performance in Dust Filtration, EPA-R2-73-288 (NTIS No. PB-222-237),
July, 1973.
98. Rothwell, E. Fabric Dust Filtration. The Chemical Engineer. (Brit.)
62(3):1, 1975.
99. Budrew, W. F. Filtration Fabrics Ability to Perform. In: Proceedings:
The User and Fabric Filtration Equipment Specialy Conference, pp
99-105. Air Poll. Control Assoc., Pittsburgh, 1974.
100. Dennis, R. , H. A. Klemm, and W. Battye. Fabric Filter Model Sensitivity
Analysis. EPA-600/7-79-043c, 1979.
101. Miller, B., G. Lamb, P. Costanza and J. Craig. Nonwoven Fabric Filters
for Particulate Removal in Respirable Dust Range. EPA-600/7-77-115, 1977,
102. Mohamed, M., and E. Afify. Efficient use of Fibrous Structures in
Filtration. EPA-600/2-76-204, 1976.
103. Ramsey, G. H., R. P. Donovan and J. H. Turner. EPA Fabric Filtration
Studies: 2. Performance of Non-Woven Polyester Filter Bags.
EPA-600/2-76-168b, 1976.
104. Lamb, G. E., Constanza, P. and Miller, B. Influences of Fiber Geometry
on the Performance of Nonwoven Air Filters. Textile Research Journal,
45:452, 1975.
148
-------
73. Cadle, R. D. Particle Size. Reinhold Publishing Corporation, New York,
1965.
74. Silverman, L., M. W. First, and C. E. Billings. Particle Size Analysis
in Industrial Hygiene. Academic Press, New York and London, 1971.
75. McCain, J. D., A. N. Bird, and K. M. Gushing. Field Measurements of
Particle Size Distribution with Inertial Sizing Devices. Southern
Research Institute, EPA 650/2-73-035, 1973.
76. Smith, W. B., K. M. Gushing, G. E. Lacey, and J. D. McCain. Particle
Sizing Techniques for Control Device Evaluation. Southern Research
Institute. EPA-650/2-74-102a, 1975.
77. Calvert, S., C. Lake, and R. Parker. Cascade Impactor Calibration
Guidelines. A.P.T., Inc., Riverside, CA. EPA-600/2-76-118. April, 1976.
78. Harris, D. B. Procedures for Cascade Impactor Calibration and Operation
in Process Streams. EPA-600/2-77-004, January, 1977.
79. Mercer, T. T. and R. G. Stafford. Impaction from Round Jets. Ann.
Occup. Hyg., 12_: 41-48, 1969.
80. Willeke, K. and J. J. McFeters. The Influence of Flow Entry and
Collecting Surface on the Impaction Efficiency of Inertial Impactors.
J. Coll. and Interface Sci., _5JJ:121, 1975.
81. McCrone, W. C., R. G. Draftz, and J. G. Delly. The Particle Atlas.
Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan, 1967.
82. Schrag, M. P., A. K. Rao, G. S. McMahon, and G. L. Johnson. Fine
Particle Emissions Information System Reference Manual.
EPA-600/2-76-173, Research Triangle Park, NC, June 1976.
83. Herdan, G. Small Particle Statistics, 2nd Rev. Ed. Academic Press,
Inc., New York, 1960.
84. Irani, R. R., and C. F. Callis. Particle Size: Measurement,
Interpretation, and Application. John Wiley & Sons, New York, 1963.
85. Raabe, 0. G. Aerosol Aerodynamic Size Conventions for Inertial Sampler
Calibrations. J. Air Poll. Control Assoc., 26(9):856, 1976.
86. Galeski, J. B. Particle Size Definitions for Particulate Data Analysis.
EPA-600/7-77-129, November, 1977.
87. Hatch, T. and S. P. Choate. Statistical Description of the Size
Properties of Non-Uniform Particulate Substances. J. Franklin Inst,
207/.369, 1929.
88. Rabbe, 0. G. Particle Size Analysis Utilizing Grouped Data and the
Log-Normal Distribution. J. Aerosol Sci., 2/3), 289, 1971.
147
-------
89. Dalla Valle, J. M., C. Orr, Jr., and H. G. Blocker. Fitting Bimodal
Particle Size Distribution Carves. Ind. Eng. Chem., 43k 1377-1380, 1951.
90. Kottler, F. J. The Distribution of Particle Sizes. J. Franklin Inst.,
250:339, 1950.
91. Kottler, F. J. The Goodness of Fit and the Distribution of Particle
Size. J. Franklin Inst., 251:449, 1951.
92. Ettinger, H. J. and S. Posner. Evaluation of Particle Sizing and Aerosol
Sampling Techniques, Am. Ind. Hyg. Assoc. J., 26:17, 1965.
93. Marchello, J. M. and J. J. Kelley, eds. Gas Cleaning for Air Quality
Control. Marcel Dekker, Inc., New York, 1975.
94. Kraus, M. N. Baghouses: Selecting, Specifying and Testing Industrial
Dust Collectors. Chem. Eng., £6(9):133-142, 1979.
95. Kraus M. N. Baghouses: Separating and Collecting Industrial Dusts.
Chem. Eng., JJ6_(8) 94-106, 1979.
96. Spaite, P. W. , G. W. Walsh. Effect of Fabric Structure on Filter
Performance. Am. Ind. Assoc. J. ^4_:357-365, 1963.
97. Draemel, D. C. Relationship Between Fabric Structure and Filtration
Performance in Dust Filtration, EPA-R2-73-288 (NTIS No. PB-222-237),
July, 1973.
98. Rothwell, E. Fabric Dust Filtration. The Chemical Engineer. (Brit.)
6_2(3):1, 1975.
99. Budrew, W. F. Filtration Fabrics Ability to Perform. In: Proceedings:
The User and Fabric Filtration Equipment Specialy Conference. pp
99-105. Air Poll. Control Assoc., Pittsburgh, 1974.
100. Dennis, R., H. A. Klemm, and W. Battye. Fabric Filter Model Sensitivity
Analysis. EPA-600/7-79-043c, 1979.
101. Miller, B., G. Lamb, P. Costanza and J. Craig. Nonwoven Fabric Filters
for Particulate Removal in Respirable Dust Range. EPA-600/7-77-115, 1977.
102. Mohamed, M., and E. Afify. Efficient use of Fibrous Structures in
Filtration. EPA-600/2-76-204, 1976.
103. Ramsey, G. H., R. P. Donovan and J. H. Turner. EPA Fabric Filtration
Studies: 2. Performance of Non-Woven Polyester Filter Bags.
EPA-600/2-76-168b, 1976.
104. Lamb, G. E., Constanza, P. and Miller, B. Influences of Fiber Geometry
on the Performance of Nonwoven Air Filters. Textile Research Journal,
45:452, 1975.
148
-------
105. Spaite, P. W. , J. E. Hagan, and W. F. Todd. A Protective Finish for
Glass-Fiber Fabrics. Chem. Eng. Progr., 5_9_(4):54, 1963.
106. Turner, J. H. Performance of Nonwoven Nylon Filter Bags.
EPA-600/2-76-168a (NTIS No. PB 266-271/AS). December, 1976.
107. Dennis, R., and J. W. Wilder. Fabric Filter Cleaning Studies.
EPA-650/2-75-009 (NTIS No. PB-240-372/3G1), January, 1975.
108. Liscomb, W. 0., J. D. McKenzie, J. C. Mycock. Performance and Cost
Comparisons Between Fabric Filters and Alternate Particulate Control
Techniques. J. Air Poll. Control Assoc., 24>:148, 1974.
109. Lucas, R. L. and E. M. Smith. Information Required for the
Specification, Purchase and Performance Evaluation of Industrial
Baghouses (Fabric Filters). J. Air Poll. Control Assoc., 25(7):715, 1975.
110. Edmisten, N. G. and F. L. Bunyard. A Systematic Procedure for
Determining the Cost of Controlling Particulate Emissions from Industrial
Sources. J. Air Poll. Control Assoc., 20_(7) :446-452, 1970.
111. Harris, W. B. and M. G. Mason. Operating Economics of Air Cleaning
Equipment Utilizing the Reverse-Jet Principle. Ind. Eng. Chem., 47:2423,
1955.
112. Roeck, D. R., and R. Dennis. Technology Assessment Report for Industrial
Boiler Applications: Particulate Control. EPA-600/7-79-178h (NTIS
PB80-176-365), December, 1979.
113. Herrick, R. A. Theory and Application of Filter Drag to Baghouse
Evaluation. Air Engineering, _10_(5):18, 1968.
114. Rudnick, S. N. and M. W. First. Specific Resistance (K/?) of Filter
Dust Cakes: Comparison of Theory and Experiments. In: Third Symposium
on Fabric Filters for Particulate Collection. EPA-600/7-78-087, 1978.
115. Grace, H. P. Resistance and Compressibility of Filter Cakes. Chem. Eng.
Progr., 49:303-318, 1953.
116. Holland, C. R. and E. Rothwell. Model Studies of Fabric Dust
Filtration 1. Flow Characteristics of Dust Cakes Uniformly Distributed
on Filter Fabrics. Filtration and Separation, ^4_(l):30, 1977.
117. Holland, C. R. and E. Rothwell. Model Studies of Fabric Dust
Filtration 2. A Study of the Phenomenon of Cake Collapse. Filtration
and Separation, ^4_:225, 1977.
118. Williams, C. E., T. Hatch, and L. Greenberg. Determination of Cloth Area
for Industrial Air Filters. Heat. Pip. Air Cond., 12^:259-263, 1940.
119. Rothwell, E. Fabric Filter Failures-Relating Laboratory Observations to
Practice. Filtration and Separation, 15(6):1, 1978.
149
-------
120. Dennis, R. , and J. A. Dirgo. Comparison of Laboratory and Field Derived
K2 Values for Dust Collected on Fabric Filters. Filtration and
Separation, 18^394-396, 417, 1981.
121. Dennis, R. , R. W. Cass, and R. R. Hall. Dust Dislodgement from Woven
Fabrics Versus Filter Performance. J. Air Poll. Control Assoc.,
_28(1):47, 1978.
122. Walsh, G. W. and P. W. Spaite. An Analysis of Mechanical Shaking in Air
Filtration. J. Air Poll. Control Assoc., _12_:57, 1962.
123. Campbell, P. R. Maintaining Proper Tension. Power, 124(2);92, 1980.
124. Billings, C. E., L. Silverman, R. Dennis, and L. H. Levenbaum. Shock
Wave Cleaning of Air Filters. J. Air Poll. Control Assoc., ^0_:318, I960.
125. Dennis, R., and L. Silverman. Fabric Filter Cleaning by Intermittent
Reverse Air Pulse. ASHRAE J. 4(3):43, 1962.
126. Bakke, E. Optimizing Filtration Parameters. J. Air Poll, Control
Assoc., ^4:1150, 1974.
127. Lefkowitz, L. R. Evaluation of Felted Glass Media under Simulated Pulse
Jet Operating Conditions. EPA-600/7-79-044b, 1979.
128. Dennis, R., and H. A. Klemm. Modeling Concepts for Pulse Jet
Filtration. J. Air Poll. Control Assoc. 30:38, 1980.
129. Leith, D. H., M. W. First, and H. Feldman. Performance of a Pulse-Jet
Filter at High Filtration Velocity - II. Filter Cake Redeposition.
J. Air Poll. Control Assoc., 27^:636, 1977.
130. Ellenbecker, M. J. and D. H. Leith. Dust Deposit Profiles in a High
Velocity Pulse-Jet Filter. J. Air Poll. Control Assoc., ^£:1236, 1979.
131. Ellenbecker, M. J. and D. H. Leith. Theory for Dust Deposit Retention in
a Pulse-Jet Fabric Filter. Filtration and Separation, _16_(6):624, 1979.
132. Dennis, R., J. E. Wilder, and D. L. Harmon. Predicting Pressure Loss for
Pulse Jet Filters. J. Air Poll. Control Assoc., 31:987, 1981.
133. Reynolds, J. P. and L. Theodore. Analysis of an APCA Baghouse Operation
and Maintenance Survey. J. Air Poll. Control Assoc., J3£(ll):1255, 1980.
134. Corn, M. The Adhesion of Solid Particles to Solid Surfaces, II. J. Air
Poll. Control Assoc., 11_:566, 1961.
135. Corn, M. The Adhesion of Solid Particles to Solid Surfaces, I. A
Review, J. Air Poll. Control Assoc., n_:523, 1961.
136. Larson, R. I. The Adhesion and Removal of Particles Attached to Air
Filter Surfaces. Am. Ind. Hyg. Assoc. J. , _19_:265, 1958.
150
-------
137. Szabo, M. F., and R. W. Gerstle. Operation and Maintenance of
Particulate Control Devices on Coal-Fired Utility Boilers.
EPA-600/2-77-129, 1977.
138. Carr, R. C., K. M. Gushing, and W. B. Smith. A Study of Reverse Gas
Baghouses Collecting Fly Ash from Pulverized Coal-Fired Boilers.
Presented at: The Fourth Symposium on the Transfer and Utilization of
Particulate Control Technology. U.S. Environmental Protection Agency.
Houston, Texas, October 11-15, 1982.
139. Culhane, F. R. Fabric Filters Abate Air Emissions. Environ. Sci. and
Tech., 8(2):128, 1974.
140. Dennis, R., G. A. Johnson, M. W. First, and L. Silverman. How Dust
Collectors Perform. Chem. Eng. 159 (February 1952).
141. Dennis, R. M. W. First, and L. Silverman. Dust Collectors Tested in
Field. Chem. Eng. (May 1954).
142. Gosselin, A. E., Jr., The Bag Filterhouse for Oil-Fired Power Plants,
J. Air Poll. Control Assoc., 15(4):179-151, 1965.
143. Bagwell, F. A., L. F. Cox and E. A. Pirsh. Design and Operating
Experience. A Filterhouse Installed on an Oil Fired Boiler. J. Air
Poll. Control Assoc., 19_(3): 149-154, 1969.
144. Borgwardt, R. H., R. E. Harrington, and P. W. Spaite. Filtration
Characteristics of Fly Ash from a Pulverized Coal-Fired Power Plant.
J. Air Poll. Control Assoc., 18_(6) :387-390 (1968).
145. Reigel, S. A. Reverse Pulse Baghouses for Industrial Coal-Fired
Boilers. Power Eng., 7j8(8) :56-59, 1974.
146. Tankha, A. Try Fabric Dust Collections on Small Boilers. Power,
m_(8): 72-73, 1973.
147. Reigel, S. A., and R. P. Bundy. Why the Swing to Baghouses? Power,
JL2J_:68, 1977.
148. Helfritch, D. V. and G. H. Beack. Coal Fired Boiler Flyash Control by
Fabric Filter Dust Collectors. Comb. 4JK4):38, 1976.
149. McKenna, J. D., J. C. Mycock and W. 0. Lipscomb. Applying Fabric
Filtration to Coal-Fired Industrial Boilers (A Pilot Scale
Investigation). EPA-650/2-74-058a, 1975.
150. Dennis, R., and D. V. Bubenick. Apparent Fractional Efficiencies for
Dust Collectors. Part I - Fabric Filters. Filtration and Separation.
20:145-146, 1983.
151
-------
151. Dennis, R. Collection Efficiency as a Function of Particle Size. Shape
and Density: Iheory and Experience. J. Air Poll. Control Assoc.,
24_(12):11, 1974.
152. Ensor, D. S. , R. G. Hooper and R. W. Scheck. Determination of the
Fractional Efficiency, Opacity Characteristics, Engineering and Economic
Aspects of a Fabric Filter Operating on a Utility Boiler. Prepared by
Meteorology Reserch, Inc. Altadena, California for Electric Power
Research Institute, Palo Alto, California, EPRI FP 297, Project 534-1.
November, 1976.
153. Cass, R. W., and R. M. Bradway. Fractional Efficiency of a Utility
Boiler Baghouse: Sunbury Steam-Electric Station. GCA/Technology
Division, Bedford, Massachusetts. EPA-600/2-76-077a (NTIS No.
PB-253-943/AS), March 1976.
154. Berg, D. B. Dust Filter Reclaims 10,000 Pounds Per Hour of Hot Asphalt
Plan Aggregate. Chem. Process., 2jK6):77, 1968.
155. Cass, R. W., and J. E. Langley. Fractional Efficiency of an Electric Arc
Furnace Baghouse. EPA-600/7-77-023 (NTIS PB-266-912/5BE), March, 1977.
156. Payton, R. N. Innovations in Ferroalloy Baghouse System DEsign. J. Air
Poll. Control Assoc., 2£(l):18, 1976.
157. Goldfield, J. and F. E. Brandt. Dust Control Techniques in the Asbestos
Industry. Am. Ind., Hyg. Assoc. J. , 3_5/12): 799-808, 1974.
158. Stephan, D. G., P. T. Bohnslav, R. A. Herrick, G. W. Walsh, and A. H.
Rose, Jr. A New Technique for Fabric Filter Evaluation. Am. Ind. Hyg.
Assoc. J., 1£:276, 1958.
159. Solbach, W. Derivation of a Computational Method for Multichamber Cloth
Filters on the Basis of Experimental Results. Staub (English).
2jKl):28-33, 1969.
160. Robinson, J. W., R. E. Harrington, and P. W. Spaite. A New Method for
Analysis of Multicompartment Fabric Filtration. Atmos. Environ.,
I/. 499-508, 1967.
161. Industrial Gas Cleaning Institute (IGCI). Operation and Maintenance of
Fabric Collectors, Publication No. 53. Suite 304, 700 N Fairfax St.,
Alexandria, VA, 22314.
162. Beachler, D. S. and M. Peterson. APTI Course SI:412 Baghouse Plan
Review, Student Guidebook. EPA No. 450/2-82-005, April 1982.
163. Reigel, S. A. and G. D. Applewhite. Operation and Maintenance of Fabric
Filter Systems. In: Operation and Maintenance for Air Particulate
Control Equipment. Young, R. A. and F. L. Cross, eds. Ann Arbor
Science, Ann Arbor. Michigan, 1980.
152
-------
164. Cross, F. L. and H. E. Hesketh, eds. Handbook for the Operation and
Maintenance of Air Pollution Control Equipment. Technomic Publishing
Co., Inc., Westport, CN, 1975.
165. Cheremisinoff, P. N. and R. A. Young, eds. Air Pollution Control and
Design Handbook, Part 1. Marcel Dekker, Inc., New York, 1977.
166. McKenna, J. D. and G. P. Greiner. In: Air Pollution Control Equipment -
Selection, Design, Operation and Maintenance. Theodore, L. and A. J.
Buonicore, eds. Prentice Hall Inc., Englewood Cliffs, N.J., 1981.
167. W. W. Sly Manufacturing Co. Instruction Book, No. 693, P.O. Box 5939,
Cleveland, OH 44101, 1980.
168. Mikro-Pul Corp. Owner's Manual Micro-Pulsaire Dry Dust Collector,
SD379a. Section VII C - Troubleshooting. 10 Chatham Road, Summit, N.J.
07901, 1980.
169. Mikro-Pul Corp. Owner's Manual, Micro-Pulsaire Dry Dust Collector,
SD578, Section VII C - Troubleshooting. 10 Chatham Road, Summit, N.J.
07901, 1980.
170. Zurn Industries. Start-Up Procedure. Operation and Proposed Maintenance
Instructions. Air Systems Division, P.O. Box 2206, Birmingham, AL,
35201, 1980.
171. Flex-Kleen/Research-Cottrell. Installation, Operating and Maintenance
Manual. Bulletin 399. 222 South Riverside Plaza, Chicago, IL 60606,
1980.
172. Ecolaire Environmental Co. Operation and Maintenance Literature. 380
Civic Drive, Pleasant Hill, CA 94523, 1980.
173. Dennis, R., and N. F. Surprenant. Particulate Control Highlights:
Research on Fabric Filtration Technology. EPA-600/8-78-005d
(NTIS PB-285-393/5BE), June, 1978.
174. Dennis, R. and H. A. Klemm, "Fabric Filter Model Format Change," Volume I
Detailed Technical Report, Volume II User's Guide. EPA-600/7-79-043a
(NITS PB-293-551/8BE). EPA-600/7-79-043b (NTIS PB-294-042/7BE),
February, 1979.
153
-------
APPENDIX A
AEROSOL AND SAMPLING TERMINOLOGY
aerosol: A stable cloud of solid particles and/or liquid droplets
uniformly suspended in a gas (smoke, fog, mist).
agglomeration: Formation of particle clusters by particle collision due to
diffusion and turbulence. Same as coagulation or flocculation.
anemometer: An instrument for measuring gas stream velocity.
attrition: Wearing or grinding down by friction. One of the three basic
processes contributing to particle formation, the others being
condensation and combustion.
baffle: A plate, grating, or refractory wall used especially to block, hinder,
or divert a flow or to hinder the passage of a fluid.
barometer: A detector for measuring atmospheric pressure.
blast gate: A sliding plate or damper installed in a supply or exhaust
duct for the purpose of regulating air flow.
bridge: Material blockage across an opening, such as a hopper outlet
or pore or flow channel in filter fabric.
cascade impactor: Sampling and particle sizing device in which air is drawn
through a series of jets directed against a series of slides or
collection stages. Retention of deposited particles may be aided by
adhesive coatings. Jet openings are sized to separate dust samples into
several size fractions.
ceramic filter: Component of a gas sampling train. These are suitable for
high temperature (1000°F) filtration.
coefficient of haze: A measure of light transmission through a soiled
filter. Coefficient of haze units are defined as 100 times the optical
density as determined by transmittance.
collection efficiency. Percentage of dust collected by a control device.
condensation: The physical process of transforming from a gaseous phase to a
liquid or solid state by decreasing temperature, increasing pressure, or
both. Gas cooling is the most common technique.
corrosion: Deterioration or physical degradation due to chemical reactions.
counter: Instrument for measuring the number and frequently the size of dust
particles in an aerosol.
154
-------
counting: Determination of the number of dust particles per unit area of a
viewing field as with a membrane filter or the number of particles per
unit volume of liquid or gas from which airborne particle concentrations
can be estimated.
count median size: That particle size for which there are an equal number of
particles greater than and less than the median or middlemost size.
cut size: That size particle which can be collected at a specified efficiency,
usually 50 percent.
damper: An adjustable gate installed in a duct for the purpose of regulating
air flow.
dehuraidify: To reduce by any process the quantity of water vapor present in a
gas mixture (air).
density: The ratio of the mass of a specimen to the volume occupied by the
specimen, the latter often expressed in cubic feet or cubic meters at a
fixed gas temperature and pressure.
diameter, aerodynamic: The diameter of a sphere of unit density having the
same terminal settling velocity as a particle in question, regardless of
its geometric size, shape, and true density.
diameter, count median: See count median size (diameter).
diameter, mass median: That particle diameter for which an equal mass is
represented by particles greater than and less than the mass median
diameter.
diameter, projected (area): The diameter of a circle having an area equal
to that of the projected area of the particle.
diameter, stokes equivalent: The diameter of a hypothetical sphere having the
same terminal settling velocity and particle density as the particle of
interest, irrespective of its actual size and shape.
dust: Any aerosolized solid material formed by disintegration or attrition
processes or redispersed by shearing action (high winds). Sizes range
from the microscopic to the macroscopic with the upper size determined by
the turbulence in the atmosphere and the source elevation.
dust collector: A device to remove solid particles from a gas stream.
155
-------
dust counter: An instrument for determining dust particle concentration in
a known volume of air. In Aitken's dust counter, condensation is made to
occur on the nuclei present by adiabatic expansion of air, and the number
of drops are counted. In Owen's dust counter, a jet of humid air is
expelled through a narrow slit in front of a microscope cover glass. The
pressure reduction due to air expansion after passing through the slit
creates a moisture film on the glass impingement surface to which dust
adheres for viewing and counting.
dust loading: The weight of solid particles suspended in a fixed air (gas)
volume (or its concentration), usually expressed in terms of grains per
cubic foot, grams per cubic meter, or pounds per thousand pounds of gas.
electrostatic filter: Filter using electrostatic charge effects to enhance
particle capture.
entry loss: Loss in pressure due to air velocity changes caused by duct,
hood, and enclosure geometry factors.
erosion: Wearing away due to mechanical or abrading actions..
fines: Particulate matter less than one micrometer in diameter.
fly ash: Finely divided particles of ash entrained in flue gases resulting
from the combustion of fuel. The ash particles consist of incompletely
burned fuel and a variety of mineral constituents.
fog: Visible aerosols in which the dispersed phase consists of liquid
droplets ranging from 5 to 40 ym in diameter and produced by atomization
or condensation processes.
fume: Fine solid particles predominantly less than 1 micrometer in diameter,
resulting from condensation, sublimation, or chemical reactions, e.g.,
oxidation of metal vapors.
gas: One of the three states of aggregation of matter, having neither
independent shape nor volume and tending to expand indefinitely
if not constrained.
geometric standard deviation: The measure of dispersion for a log-normal
distribution, the ratio of the 84.13 to the 50 percentile (or the 50 to
the 15.87 percentile).
grain: A unit of weight equivalent to 1/7000 of a pound or 65 milligrams.
grit: Solid particles of natural or man-made origin, retained on a 200 mesh
screen, roughly 74 ym or greater in diameter.
ground level concentration: The amount of solid, liquid, or gaseous material
per unit volume of air, from 0 to 2 meters above the ground.
156
-------
ha/.c: A state of atmospheric obscuration due to the presence of fine, solid
and/or liquid particles in stable suspension; visibility exceeds 1 but is
less than 2 km. The particles are so small that they cannot be sensed
via impact or thermal effects nor seen with the naked eye.
hood suction: Resultant hood static pressure due to the entry loss plus
conversion from fluid potential to kinetic energy.
hopper: Storage container at the bottom of a collection device to hold
the dust.
humidity, absolute: The weight of water vapor carried by a unit weight of dry
air or gas. Pounds of water vapor per pound of dry air; grains of water
vapor per pound of dry air.
humidity, relative: The ratio of the absolute humidity in a gas to that
of the saturated gas at the same temperature.
itnpaction: A particle collision with a collection surface caused by the
interaction of inertial and hydrodynamic forces.
isokinetic sampling: Extraction of an aerosol by a sampling probe such that
the fluid velocity entering the probe is the same as that at the sampling
location in the parent gas stream. Deviation from isokinetic conditions
can lead to incorrect estimates of mass particulate concentrations.
manometer: An instrument for measuring pressure differences, usually
consisting of a vertically-aligned U-shaped tube containing liquids of
known density; e.g., water, oil, mercury. Instrument sensitivity may be
increased by changing the angle of inclination.
mass concentration: Concentration expressed in terras of mass of suspended
or dissolved material per unit volume of gas or liquid.
membrane filters: Extremely high efficiency filters with relatively low
resistance to gas flow that are usually composed of gels of cellulose
esters or other polymeric substances. Such filters contain many small
holes or pores of controllable size such that their efficiency and
pressure loss characteristics may be predicted. In addition, because the
filters are usually of high chemical purity, they are well suited for
trace metal analyses. Some of the membrane filters can also be rendered
transparent, thus permitting direct microscope observation of collected
particles. Alternatively, filters can be dissolved in selected organic
solvents so that the particles may be isolated and studied. Since most
membrane filters are not affected greatly by humidity changes, they can
be readily weighed before and after use. They are not suitable for high
temperature filtration (except for Teflon membranes) and their electrical
charge properties can create sampling and handling problems.
157
-------
mesh numbers: U.S. sieve sizes designate the number of holes per inch for
a specific screen.
Mesh Number Corresponding Opening, Inches
8 0.093
12 0.066
14 0.055
20 0.0328
30 0.023
40 0.0165
100 0.0058
200 0.0029
325 0.0017
micrometer (micron) (ym) : A unit of length corresponding to one thousandth
of a millimeter or one millionth of a meter (approximately 1/25,000 of an
inch).
mist: A dispersion of relatively small liquid droplets. In meteorology
parlance, the terra mist applies to water droplets of sufficient size
(40 ym) to settle from the air. Persistence of a mist depends upon
constant droplet replenishment via condensation processes.
mole: The weight of a substance numerically equal to its molecular weight. If
the weight is in pounds, the unit is "pound moles." For dry air at 70°F
and a pressure of one atmosphere, there are 29 pounds per pound mole and
386 cubic feet per pound mole (thus a density of 0.075 Ibs/ft3).
opacity: A measurement of visibility as affected by aerosol presence and
defined as the apparent obscuration of an observer's vision as compared
to that produced by a smoke of a given density rating on the Ringelman
c ha rt.
particle: A small discrete mass of solid or liquid matter such as dust,
fume, mist, smoke, and fog.
particle (number) concentration: Concentration expressed in terms of number
of particles per unit volume of air or other gas.
particle size distribution: Usually refers to a mathematical function or
graphical display that describes the mean, median, and range of particle
sizes present in a given sampling. A logarithmic normal distribution is
most commonly used to represent both atmospheric and industrial aerosols.
particulate matter: Any dispersed matter, solid or liquid, in which the
individual particles or agglomerates may range from 0.002 ym up to 500 pro.
Particle lifetimes in the suspended state range from a few seconds to
several months. Additional terms used to describe particulate matter may
include dust, fly ash, smoke, soot, droplets, mist, fog, and fumes.
158
-------
penetration: That fraction or percent of a particulate material that passes
through a dust collector and discharges to the atmosphere.
pitot (static) tube: A device for measuring velocity pressure consisting of
two tubes—one serving to measure the total pressure in an air stream
(static plus velocity) and the other to measure the static pressure
only. By differencing these two pressure signals via manometer or
similar device, the static pressure components cancel each other such
that only the velocity pressure is displayed.
plume: The visible path formed by the continuous discharge of any aerosol
from a stack or chimney. The shape of the path and the particle
concentration distribution in the plume is dependent on local
meteorological conditions. Looping, coning, fanning, fumigating, and
lofting are terms frequently used to characterize plumes.
pressure, atmospheric: The pressure due to the weight of the atmosphere, as
indicated by a barometer. Standard sea level atmospheric pressure is
29.92 in. of mercury, 760 mm of mercury, 14.7 psia, and 407 in. water
column.
pressure, gage: Pressure referenced to local ambient or atmospheric pressure
as a datura. Gage pressure may be indicated by a manometer in which one
leg is connected to the pressure source and the other exposed to the
ambient atmosphere.
pressure loss: The pressure required to overcome the resistance to gas flow
in a system that includes the resistance of straight runs of pipe,
entrances to headers, bends, elbows, orifice losses, and pressure drop
through gas cleaning devices.
pressure, static: The potential pressure exerted equally in all directions
by a fluid at rest. For a fluid in motion, it is measured in a direction
normal to the direction of flow and usually expressed in inches water
gage when dealing with air.
pressure, total: The algebraic sum of the velocity pressure and the static
pressure (with due regard to sign). In gas-handling systems these
pressures are usually expressed in inches water gage.
pressure, velocity: The kinetic pressure in the direction of flow necessary
to cause a fluid at rest to accelerate to a given velocity. Usually
expressed in inches water gage. (Also referred to as velocity head).
Reynolds number: The dimensionless ratio of inertial to viscous forces in
a flow stream. It serves to describe the nature of the flow; i.e.,
laminar, intermediate, or turbulent.
shape factor: a constant that, when applied to the appropriate
characterizing particle diameter, will indicate the total or projected
particle surface or particle volume.
159
-------
smoke: Small gas-borne particles usually resulting from incomplete
combustion. Such particles consist predominantly of carbon and other
combustible material and are present in sufficient quantity to be
observable independently of other solids.
soot: An agglomeration of carbon particles impregnated with "tar" formed by
the incomplete combustion of carbonaceous material. These particles,
which are much larger than those constituting a true smoke, are highly
visible to the naked eye.
specific gravity: A dimensionless quantity defining the ratio of the density
of a substance at a given temperature to the density of some reference
material (usually water for liquids and air for gases).
specific heat: The heat absorbed (or given up) by a unit mass of a substance
when its temperature is increased (or decreased) by one degree. Common
units are BtuS/lb/°F or Cals/g/°C. For gases, both the specific heat at
constant pressure (c~) and the specific heat at constant volume (cv)
play key thermodynamic roles, particularly so when the flow is
compressible.
specific volume: The volume of a substance per unit mass; i.e., the
reciprocal of density, usually given in cubic feet per pound or cubic
meters per gram.
standard atmosphere: The pressure exerted by a column of mercury 29.92 inches
high at 70°F; approximately 14.7 psi. See pressure, atmospheric.
temperature, absolute: Temperature expressed as degrees Rankine or degrees
Kelvin where absolute zero corresponds to -460°F or -273°C, respectively.
temperature, dew-point: The temperature corresponding to moisture saturation
(100 percent relative humidity) for a given absolute humidity at constant
pressure. Any lowering of the temperature will cause water vapor to
precipitate as a liquid or solid.
temperature, dry-bulb: The temperature of a gas or mixture of gases indicated
by an accurate thermometer after correction for radiation.
temperature, wet-bulb: Wet-bulb temperature is a measure of the moisture
content of air (gas). It is the temperature indicated by a wet-bulb
psychrometer.
transmissometer: An instrument for measuring the extinction coefficient of
the atmosphere for the determination of visibility range. Also called
telephotometer, transmittance meter, hazemeter, or smoke density
indicator.
velocity, capture: The air velocity at any point in front of the hood or at
the hood opening necessary to prevent particulate material and
contaminant gases from escaping to the working area.
160
-------
velocity, terminal settling: That constant velocity at which the fluid
resistance to the motion of a body equals the force causing the motion,
usually the force of gravity.
velocity, transport (conveying): The minimum air velocity required to
transport particulate matter in any air stream, especially confined duct
or pipe flows (vertical or horizontal) so that no sedimentation takes
place.
velocity traverse: A method for determining the average air velocity in any
duct by dividing it into numerous equal area sections. The average
velocity then becomes the simple arithmetic average of the velocities
measured in each equal area section.
viscosity: The proportionality constant relating shear rate to shear stress
that is expressed in units of poise, centipoise, or pounds per
foot-second.
161
-------
INFORMATION SOURCES
Committee on Industrial Ventilation. Industrial Ventilation, A Manual
of Recommended Practice. 16th Ed. American Conference of Governmental
Industrial Hygienists. P. 0. Box 16153, Lansing, Michigan, 48901, 1979.
Dennis, R (Ed.). Handbook on Aerosols. U.S. Energy and Development
Administration. TID-26608, NTIS, Springfield, Va., 1976.
Liptak, B. G. (Ed.). Environmental Engineers' Handbook. Volume II,
Air Pollution. Chilton Book Company, 1974.
PPG Industries, Inc. The Filtration Forum, Glossary of Terms Common
to Baghouses Using Glass Fiber Cloth. Fiber Glass Division, The Gateway
Center, Pittsburgh, PA, 15222.
The Mcllvaine Co. The Fabric Filter Manual, Chapter XIII, Sec. 1.1.,
Glossary of Fabric Filter Terms. 2790 Maria Ave., North Brook, IL, 60062.
162
-------
APPENDIX B
PARTICLE SIZE MEASUREMENT (1,6,22,72-74)
Particle size properties are best determined by in situ measurements
wherein the number and dimensions of individual particles are sensed by a
detection system that does not alter the true size and concentration
properties. A few instruments may approach this capability, e.g., single
particle light scattering counters (22), but only for very dilute particle
concentrations in the respirable range (<3 pro) such as encountered in a clean,
ambient atmosphere. Extractive techniques in which the sampled gas stream is
drawn through a sampling line and thence through a special membrane filter
have limited applications, provided that sampling line losses are negligible
and, most importantly, the individual particle deposition density is
sufficiently low that the dimensions and form of individual particles can be
clearly discerned (71-74).
Both the in situ and extractive methods mentioned above are seldom
applicable in real field situations because of various temperature, moisture,
and/or pressure problems. Additionally, inlet concentration levels are
typically 3 to 5 orders of magnitude greater than can be practically measured
by these devices without extensive and highly unreliable dilution techniques.
o
Therefore, the determination of the size properties of high (0.1 to 10 g/m )
particle loadings is usually made with an instrument called a "cascade
impactor," that is available in several commercial designs,
Figure B-l (75-80). With this device, the sampled gas stream is guided
sequentially through a series of jet stages that are aligned in the order of
163
-------
Basic Data Requirements
Solenoid operating cover
sampling port
0-ring seols
H-NO- 3i6 stainless steel wall
Monet jet
Slide supports (3)
38 mm glass Slide
>—Tie rods C3)
„ Filter stage (Hurlbut glass
Critical Mow orifice merer
Pressure tops
To vacuum pump
*• ^
impact ion slice }
small jet
small particle
Figure B-l. Cascade impactor for particle size determination.
164
-------
decreasing aperture size. Hence, the successive exit velocities increase as
gas flows through the device (79). An impingement surface consisting of a
high efficiency glass paper or an adhesive-coated glass or metal plate surface
is placed before each jet (80). In accordance with the increasing jet
velocities, a calculable fraction of the coarser particles approaching each
stage impacts upon the collection stage, whereas the finer fraction is
transported to the succeeding stage. The net result is a fractionation of the
entering particulate material into several diameter categories from which the
size distribution on a mass or weight basis can be ascertained for the
entering dust. The smallest particle class of a cascade impactor is
represented by the dust collected on a final high efficiency backup filter.
Because an impactor can be attached to the inlet end of a sampling probe,
the possibility of sizing errors resulting from probe losses is reduced. In
many field applications, sampling intervals of up to 30 minutes are possible
(depending upon the dust concentration) before excessive dust is collected on
the various impaction stages. Table B-l shows a tabulation of typical sizing
data used to construct a cumulative mass distribution curve for a coal fly
ash (78). In normal practice, a characteristic diameter (usually that
describing a spherical particle of unity density that is collected at
50 percent efficiency by that stage) is assigned to each stage based upon the
manufacturer's and sometimes the user's calibration (77,78). These are the
values shown in Table B-l for the various stages.
165
-------
TABLE B-l. TABULATED SIZING DATA, CASCADE IMPACTOR SAMPLING
OF COAL FLY ASH AEROSOL
Stage No.
Filter3
7
6
5
4
3
2
1
Characteristic'3
diameter
0.82
1.16
1.64
2.32
3.78
4.64
6.56
>6.56
Mass percent
on stage
5.5
2.5
9.0
8.5
9.5
14.0
13.0
38.0
Cumulative
mass percent
5.5
8.0
17.0
25.5
35.0
49.0
62.0
100.0
aEstimated characteristic diameter, backup filter stage.
b Aerodynamic diameter for unit density spherical particle
166
-------
APPENDIX B
REFERENCES*
1. Billings, C. E., and J. E. Wilder. Handbook of Fabric Filter Technology,
Vol. 1. GCA/Technology Division, Bedford, Massachusetts 01730. NAPCA
Contract No. CPA-22-69-38. (NTIS PB-200-648), December 1970.
6. Stern, A.C., ed. Air Pollution, Volume III, Measuring, Monitoring, and
Surveillance of Air Pollution. 3rd. Ed., Part A, Academic Press, Inc.,
New York, San Francisco, London, 1976.
22. Dennis, R., ed. Handbook on Aerosols. U.S. Energy and Development
Administration. TID-26608, NTIS, Springfield, VA, 1976.
7. Drinker, P., and T. Hatch. Industrial Dust, 2nd Ed., McGraw Hill Book
Co., Inc., New York, London, Toronto, 1954.
72. Orr, C., Jr., and J. M. Dalla Valle. Fine Particle Measurement. The
MacMillan Co., New York, 1960.
73. Cadle, R. D. Particle Size. Reinhold Publishing Corportion, New York,
142-7, 1965.
74. Silverman, L., M. W. First, and C. E. Billings. Particle Size Analysis
in Industrial Hygiene. Academic Press, New York and London, 1971.
75. McCain, J. D., A. N. Bird, and K. M. Gushing. Field Measurements of
Particle Size Distribution with Inertial Sizing Devices. Southern
Research Institute, EPA-650/2-73-035, 1973.
76. Smith, W. B., K. M. Gushing, G. E. Lacey, and J. D. McCain. Particle
Sizing Techniques for Control Device Evaluation. Southern Research
Institute, EPA-650/2-74-102a, 1975.
77. Calvert, S., C. Lake, and R. Parker. Cascade Irapactor Calibration
Guidelines. A.P.T. Inc., Riverside, CA. EPA-600/2-76-118. April 1976.
78. Harris, D. B. Procedures for Cascade Impactor Calibration and Operation
in Process Streams. EPA-600/2-77-004, January 1977.
79. Mercer, T. T., and R. G. Stafford. Impaction from Round Jets. Ann.
Occup. Hyg. 12:41-48, 1969.
80. Willeke, K., and J. J. McFeters. The Influence of Flow Entry and
Collecting Surface on the Impaction Efficiency of Inertial Impactors.
J. Colloid and Interface Sci., 53:121, 1975.
*Reference numbers are those shown in main text,
167
-------
APPENDIX C
BAGHOUSE AND FABRIC TERMINOLOGY
abrasion—flex: Cloth wear in a creased area by excessive or repeated bending.
abrasion—surface: Where the dirty cloth surface has been abraded by rubbing,
scuffing, and erosion, in a more or less uniform fashion.
acetate: A manufactured fiber in which the fiber-forming substance is
cellulose acetate.
acrylic: A synthetic polymerized fiber which contains at least 85 percent
aerylonitrile.
air-to-cloth ratio: The ratio of gas volume to effective cloth area.
Resulting units depict the filtration velocity.
backwash: A method of fabric cleaning in which direction of gas flow is
reversed, accompanied by flexing of the fabric and breaking of the dust
deposit; also called backpressure, repressure, collapse-clean.
bag: The customary form of a filter element. Also tube, stocking, etc. Can
be unsupported (dust on inside) or used on the exterior of a grid support
(dust on outside).
bag life: Number of years a bag will perform effectively. Usually 1 to
2 years.
bag reach: Number of parallel rows of bags on either side of interior
walkway. (e.g., with a bag reach of two, an operator would only need to
reach past one bag to service the most remote row).
bag tension: Force applied to a bag to hold it securely in position.
basket weave: An extension of plain weave.
batch cleaned: Usually refers to a process used in heat cleaning fiberglass
cloth in roll form by exposing it to 500°-600°F, for prolonged periods to
burn off the starches or binders.
beaming: Operation in which the yarn from several section beams is combined
on the final loom beam.
bleed: Particles of dust or fumes that are able to leak through the bag.
168
-------
blinding (blinded): The loading, or accumulation, of filter cake to the
point where capacity rate is diminished. Also termed "plugged." Once
enough material has built up, gas flow is severely restricted, and the
bags have to be dry-cleaned or replaced.
broken twill: Modified twill weave where the diagonal twill line is shifted
in a regular pattern.
bulked yarn: Multifilament yarn which has been processed by high pressure air
passing through the yarn and relaxing it into gentle loops, bends, etc.
caking: Material crusted on a bag that cannot be removed by normal cleaning.
calendering: The application of either hot or cold pressure rolls to smooth
or polish a fabric, thereby reducing the thickness of the cloth and
decreasing air permeability.
canton flannel: Usually a twill weave fabric with the filling float heavily
napped.
cap: Device that closes top of bag and is used to suspend bag from top of
unit.
cell plate: Unit that holds the thimble.
cleaning cycle: Frequency and duration of dust cake removal.
chain weave: A 2/2 broken twill weave, arranged 2 threads right and 2 left.
cloth: In general, a pliant fabric—woven, knitted, felted, or otherwise
formed of any textile fiber, wire, or other suitable material. Usually
understood to mean a woven textile fabric.
cloth weight: Is usually expressed in ounces per square yard or ounces per
square foot. However, cotton sateen is often specified at a certain
number of linear yards per pound of designated width. For example, a
54" - 1.05 sateen contains 1.05 linear yards per pound in a 54" width.
cold spot: A point on an insulated baghouse where a continuous metallic
heat transfer circuit through the insulation creates an uninsulated area.
compartment: Subdivision unit of a baghouse.
coronizing: A heat cleaning process for fiberglass fabric to burn off the
starches (used in processing) usually at temperatures of 1000° to 1200°F
for short duration.
cotton number: Staple yarns are generally sized on the cotton system.
Example: an 18 singles yarn is of such size that 18 hanks (each hank
contains 840 yards) weighs 1 pound.
169
-------
cotton system: A system of yarn manufacturing for spinning cotton fiber into
yarn, whereby the individual fibers are aligned parallel.
cover: A descriptive term for the appearance of plain woven goods. A "well
covered" cloth is the opposite of an open, or "reedy" cloth.
creel: Device used as a yarn package rack to hold warp ends for a section
beam.
crimp: The corrugations in a yarn from passing over and under other yarns at
right angles.
crowfoot satin: A 3/1 broken twill arranged 2 threads right, then 2 threads
left, etc. Also called 4 shaft satin, or broken crow weave.
cuff: Multithickness of cloth at ends of bag.
"DE" fiber: Diameter of most common glass fiber used in baghouses
(0.00025 in.).
denier: The number, in grams, of a quantity of yarn, measuring 9000 meters
in length. Example: A 200 denier yarn measuring 9000 meters weighs
200 grams. A 200/80 yarn indicates a 200 denier yarn composed of 80
filaments. Usually used for continuous tnultifilament yarns of silk,
rayon, Orion, Dacron, Dynel, Nylon, etc.
diaphragm valve: A compressed air-operated valve that is used to pulse-clean
bags.
dimensional stability: Ability of the fabric to retain finished length and
width, under stress, in hot or moist atmospheres.
doupe selvage: Two false selvages (fabric edge woven so that it will not
unravel) woven in the cloth about 1/2-in. apart; useful for splitting the
length of cloth in two.
drill weave: Same as twill except the diagonal twill line usually runs from
lower right to upper left. A 2/1 LH twill, or 3/1 twill.
drop wires: Circuit breaking devices which stop loom or warper in the event
of yarn breakage.
dust permeability: Defined as the mass of dust per unit area of filter
medium (W) divided by the pressure loss (?) per unit velocity (V) through
the medium; i.e., W V/P. In metric units (g)(ra)/(N)(min). In English
units (lb)/(in. water)(ft)(min). It is also the reciprocal of K2 the
"specific resistance coefficient". It should not be confused with the
cloth (Frazier) permeability.
end: An individual yarn or cord; a warp yarn running along the length of the
fabric.
end count: Same as warp count.
170
-------
entering: Procedure in which warp yarns are drawn through drop wires, heddle
eyes, and reed prior to going into loom.
envelope: A common form of filter element.
extensibility: The stretching characteristic of fabric under specified
conditions of load, etc.
fabric: A planar structure produced by interlacing yarns, fibers, or
filaments.
Knitted fabrics are produced by interlooping strands of yarns, etc.
Woven fabrics are produced by interlacing strands at more or less right
angles.
Bonded fabrics are a web of fibers held together with a cementing medium
which does not form a continuous sheet of adhesive material.
Felted fabrics are structures built by the interlocking action of the
fibers themselves, without spinning, weaving, or knitting.
fiber: The fundamental unit comprising a textile raw material such as cotton,
wool, etc.
filament: A continuous fiber.
fill: Crosswise threads woven by loom.
fill count: Number of fill threads per inch of cloth.
filling yarn: Yarns in a fabric running across the width of a fabric; i.e.,
at right angles to the warp.
filter cake: The accumulation of dust on a bag before cleaning. This cake
assists in filtering the dust by the action of sieving.
filter drag: Pressure drop, in inches W.C. per cubic foot of air per minute,
per square foot of filter media. Analogous to the resistance of an
element in an electrical circuit. The ratio of filter pressure loss to
filter velocity.
filter medium: The substrate support for the filter deposit; the fabric upon
which the filter deposit is built.
filter velocity: The velocity, feet per minute, at which the air (gas) passes
through the filter medium; the velocity of approach to the medium, the
filter capacity rate, the air-to-cloth ratio.
filtration: The process of separating suspended solids from a liquid or gas
usually by forcing a carrier gas or liquid through a porous medium.
filtration rate: The volume of air (gas), in cubic feet per minute, passing
through 1 square foot of filter media. Also the air-to-cloth ratio.
171
-------
finish: A treatment that imparts dimensional stability and provides
protection to the woven fiber (e.g., siiicone, graphite, Teflon, etc.)
flame retardant: A finish designed to reduce the combustibility of a fabric.
float: The position of a yarn that passes over two or more yarns passing in
the opposite direction. Example: in standard cotton sateen, yarns
"float" four, and pass under one; in other words, 4/1.
fluorocarbon: Fiber formed of long chain carbon molecules whose available
bonds are saturated with fluorine.
fulled: A woven fabric treated to raise fiber ends (like napping) so that
the thready, woven look is partially or completely obscured.
gabardine weave: A regular or "steep" twill, with higher warp than fill
count.
garnett: Scrap fiber or fabric that has been torn severely and reduced
once again to fiber, but of very short lentgh.
glass (fiberglass): A manufactured fiber in which the fiber forming
substance is glass.
"grab": tensile: The tensile strength, in pounds per inch, of a textile
sample cut 4x5 in. and torn apart lengthwise by two 1-in. square clamp
jaws set 3 in. apart and pulled at a constant specified speed.
grading frame: Equipment which continuously runs a fabric over a well-lighted
surface to enable an observer to view any defects.
greige cloth: Cloth as it comes off the loom or so-called "loom finish."
grid cloth: The cloth used in supporting the silver in making a supported,
needled felt.
hand or handle: The "feel" of the cloth—as soft, harsh, smooth, rough,
silken-like, boardy, etc.
harness: The frame used to raise or lower those warp yarns necessary to
produce a specific weave while at the same time permitting the filling to
be passed through by the shuttle.
head end: A piece of fabric taken from the end of a roll of cloth.
header: A pressurized pipe that contains the compressed-air supply for a
pulse-type baghouse.
head set: A finishing process for a head end, in fabric form, to stabilize
it against further shrinkage at predetermined temperatures.
172
-------
heddle: Device in loom containing eyelets which yarn passes through while
weaving.
herringbone weave: A type of "broken" (irregular) twill.
hydrophobia fibers: Fibers that do not readily absorb water.
interlacing: The points of contact between the warp and filling yarns in a
fabric.
interstices: The openings between the interlacings of the warp and filling
yarns; i.e., the voids.
K factor: The specific resistance of the dust cake, in inches water column
per pound of dust per square foot of filter area per feet per minute
filtering velocity. Also specific resistance coefficient, K2, the
reciprocal of dust permeability.
leno: A weave in which the adjacent warp yarns are twisted on either side
of the interlacing filling yarn.
loom finish: Same as greige cloth.
mildew-resistant finish: An organic or inorganic finish to repel the
growth of fungi on natural fibers.
raodacrylic: A man-made fiber which contains less than 85 percent
acrylonitrile (but at least 35 percent).
moleskin: A filling face fabric with a twill weave.
raonofilament: A continuous fiber of sufficient size to serve as yarn in
normal textile operations.
mullen burst: The pressure necessary to rupture a secured fabric specimen,
usually expressed in pounds per square inch.
multifilament, multifil: Yarn composed of several filaments, which are
continuous strands of fiber of indefinite length.
napping: A process to raise fiber or filament ends (for better coverage
and more surface area) accomplished by passing the cloth over a large
revolving cage or drum of small power-drive rolls covered with card
clothing (similar to a wire brush).
needled felt: A felt made by the replacement of loose fiber in a systematic
alignment, with barbed needles moving up and down, pushing and pulling
the fibers to form an interlocking of adjacent fibers.
nonwoven felt: A felt made either by needling or matting of fibers or by
compressing with a bonding agent for permanency.
173
-------
nylon: A manufactured fiber in which the fiber-forming substance is any
long-chain synthetic polyamide having recurring amide groups.
olefin: A manufactured fiber in whicb the fiber-forming substance is any
long-chain synthetic polymer composed of at least 85 percent by weight
ethylene, propylene, or other olefin units.
oxford weave: Same as plain weave, except that two warp yarns weave as one
warp yarn.
permeability, fabric: Measured on Frazier porosity meter, or Gurley
permeometer, etc. Not to be confused with dust permeability. The
ability of air (gas) to pass through the fabric, expressed in cubic feet
of air per minute per square foot of fabric with a 0.5 in. W.C. pressure
differential.
pick: An individual filling yarn running the width of a woven fabric at right
angles to the warp. In England it is termed woof, or weft.
pick glass: A magnifying glass used in counting the warp and filling yarn in
the fabric.
plain weave: Each warp yarn passing alternately over each filling yarn. The
simplest weave, 1/1 construction. Also called taffeta weave.
plenum chamber: An air compartment maintained under positive or negative
pressure, and connected to one or more ducts. A pressure equalizing
chamber.
ply or plied yarn: Made by plying together two or more twisted yarns.
polyester: A manufactured fiber in which the fiber-forming substance is any
long-chain synthetic polymer composed of at least 85 percent by weight of
an ester of dihydric alcoholic and terephthalic acid.
porosity, fabric: Term often used interchangeably with permeability.
Actually percentage of voids per unit volume—therefore, the term is
improperly used where permeability is intended.
precoat: Material (i.e., fly ash, dolomite, etc.) added to air stream on
initial baghouse start-up to aid in establishing filter cake on bags.
preshrink: Usually a hot aqueous immersion of the cloth to eliminate its
tendency to shrink in further wet applications.
pressed felt: A type of felt manufactured by pressing fibers into a scrim.
pressure drop: Resistance of system to gas flow; pressure differential can
be measured across fabric or across entire collector (flange-to-flange).
pressure (pulse) jet cleaning: A bag cleaning method where a transient burst
or pulse of compressed air is introduced through a tube or nozzle
attached to the top cap of a bag to remove dust from bag.
174
-------
pressure system: Dust-laden air is pushed through the collector, fan on dirty
side, creating positive pressure.
pulse cycle: The time interval between one pulsing of a row on bags and the
next pulsing of the same row.
pulse interval: Time between pulsing one row of bags and pulsing the next
row.
quill: Bobbin onto which filling yarn is wound.
ravel strip tensile: The tensile strength, in pounds per inch, of a 6-inch
textile sample cut just over 1-inch wide (with yarns peeled off each side
down to exactly 1-inch wide) pulled in two lengthwise between jaws set
3 inches apart and pulled at a constant specified speed.
rayon: A manufactured fiber composed of regenerated cellulose.
reed: A thin comb made of pressed steel wires between which warp ends are
drawn after passing through the heddle eyes. the reed beats the filling
picks into their respective places against the fill of the cloth.
reed marks: The indentations between 2, 3, or 4 ends, usually eliminated
in finishing.
repeat: The number of threads in a weave before the weave repeats or starts
over again. The number of ends and picks in the repeat may be equal or
unequal, but in every case the repeat must be in a rectangular form.
reverse-air baghouse: A unit that is cleaned by flowing air in the opposite
direction or backwards through the cloth (see backwash).
reverse jet cleaning: A cleaning method (Hershey) using a traveling ring
traversing the exterior of the filter bag. High pressure air is blown
backwards through the fabric through small holes or slots facing the
cloth. This technique has been largely replaced by pulse jet cleaning in
current practice.
rings: Metal bands sewn into a bag at various intervals to prevent bag
collapse during cleaning.
sanforizing: A patented process where the cloth is "puckered" in the warp
direction to eliminate shrinkage in laundering.
saran: any long-chain synthetic polymer composed of at least 85 percent
vinylidene-chloride units.
sateen: cotton cloth made with a 4/1 satin weave, either as warp sateen or
filling sateen.
175
-------
satin weave: A fabric usually characterized by smoothness and luster.
Generally made warp face with a great many more ends Chan picks. The
surface consists almost entirely of warp (or filling) floats in
construction 4/1 to 7/1. The intersection points do not fall in regular
lines, but are shifted in a regular or irregular manner.
scour: A soap and hot water wash applied to "off loom" fabric.
scrim: A very loosely woven fabric through which felt is needled.
seeding: The application of a relatively coarse, dry dust to a bag before
start-up, to provide an initial filter cake for immediate high efficiency
and to protect bags from blinding (see precoat).
selvage: The binding lengthwise edge of a woven fabric.
service module: A compartment in a pulse-jet baghouse where valves, piping,
wiring, and timers are located.
shaker baghouse: A unit using woven cloth bags where cleaning is accomplished
by shaking the bags from the top.
shaking (cleaning): A common mechanical method of removing dust from filter
elements. Backwashing or other supplemental methods are often used with
shaking. Air-shaking is a bag cleaning means wherein bags are shaken in
a random fashion by a variable velocity air stream rather than by
mechanical devices.
shoddy: Reclaimed or reused wool.
shuttle: The device that holds the quill of filling yarn and carries it back
and forth across the width of the fabric.
singeing: The burning off of the protruding hairs from the warp and filling
yarns of the fabric.
singles: A protective coating applied to yarn to ensure safe handling; i.e.,
abrasion-resistance during weaving.
slashing: The method of applying size to a width of warp yarns on a
continuous basis.
slippage: The movement or shifting of yarns in a fabric from their normal
position.
slub: A heavy accumulation of fiber or lint carried on a yarn and
interlocked during weaving.
sonic (sound): A fabric cleaning method using acoustic energy to vibrate
the filter elements. Used alone, or as a supplement to shaking or
backwash cleaning.
176
-------
spun fabric: Fabric woven from staple (spun) fiber—same as staple.
staple fiber: Man-made fibers cut to specific length—1-1/2 in., 2 in.,
2-1/4 in., etc.—natural fibers of a length characteristic of fiber type,
animal fibers being the longest.
"S" twist: Direction of twist in yarn. Ascending: right to left
configuration. (The yarn spirals conform in slope to the center portion
of the letter "S".)
taffeta: Closely woven plain weave (1/1) fabrics with the warp yarns greatly
outnumbering the filling yarns.
tenacity: Ultimate tensile strength of a fiber filament, yarn, etc.
expressed in grams per denier (g.p.d.).
tensile strength: The ability of yarn or fabric to resist breaking by direct
tension. Ultimate breaking strength in psi.
tenter frame: (pin tenter) A machine for drying cloth under tension.
Tentering. Also called framing.
textile: That which is or may be woven. Comes from the Latin "texere," to
weave. Hence, any kind of woven fabric.
textured yarn: Same as bulked yarn.
thimble: Flanged cylinder to which the bottom cuff of a bag is attached.
thread: Highly twisted and finished yarn used to fabricate bag seams.
thread count: The number of ends and picks per inch of woven cloth. For
example, 64 x 60 (ends count first).
throw: Process of doubling of twisting fabrics into a yarn of the desired
size and twist.
tow: A large number of filaments collected in a loose rope-like form,
without definite twist.
T.P.I.: Twist per inch of yarn (turns per inch).
trumeter: A device used to accurately measure yardage passing a specific
reference point.
tube sheet: The steel plate that bags and cages are suspended from in a
pulse-type baghouse or to which the thimble sections are attached with
mechanically shaken or reverse flow cleaned bags.
177
-------
twill weave: Warp yarns floating over or under at least two consecutive
picks from lower to upper right, with the point of intersection moving
one yarn outward and upward or downward on succeeding picks, causing
diagonal lines in the cloth.
twist: The number of complete spiral turns per inch in a yarn, in a right
or left direction; i.e., "Z" or "S," respectively.
twist and ply frames: Machinery used to twist and ply glass yarns.
uniful: A device attached to the loom which automatically winds yarn onto
quills from yarn packages and maintains a supply of quills for the
shuttle.
vacuum system: Gas stream pulled through baghouse, fan on clean side,
creating negative pressure inside collector.
velocity, approach: The velocity of air (gas), feet per minute, normal
to the face of the filter medium.
warp: Yarns running lengthwise in woven goods.
warp beam: Large spool-like or barrel-like device on which the warp
threads are wound.
warp count: Number of warp threads per inch of width.
warper: A machine for preparing and arranging the beamed yarns
intended for the warp of a fabric.
warp sateen: The face of the cloth having the warp yarns floating over
the filling yarns, with more warp yarns than filling yarns (per inch).
warp size: Chemicals applied to the warp yarn to improve strand integrity,
strength, and smoothness in order to withstand the rigors of weaving.
weave: The pattern of weaving; i.e., plain, twill, satin, etc.
weft: Same as filling; the crosswise threads (yarns).
woof: Same as filling or weft.
woolen system: A "system" of yarn manufacturing for spinning wool fiber
into yarn, usually more open and not aligned as parallel as the cotton
system. This system is suited for shorter wools, various wastes,
reclaimed wools, etc.
worsted system: A system of yarn manufacturing suited for medium and longer
wools. Includes additional processing steps resulting in the most
uniform yarn. The resulting yarn is compact and level.
178
-------
woven felt: Predominantly a woven woolen fabric heavily fulled or shrunk,
with the weave being completely hidden due to the entanglement of the
woolen surface fibers.
yarn: Twisted fibers or filaments in a continuous strand suitable for
weaving, etc. Ply yarn is formed by twisting two or more single yarns
together. Ply yarns are in turn twisted together to form cord.
yarn size (denier, or count): A relative measure of fineness or coarseness
of yarn. The smaller the number in spun yarns, the coarser the yarn.
The higher the denier of a filament yarn, the coarser (heavier) the yarn.
"Z" twist: Direction of twist in yarn. Ascending left to right configuration.
(The yarn spirals conform in slope to the center portion of the letter
"Z").
179
-------
INFORMATION SOURCES
PPG Industries, Inc. The Filtration Forum, Glossary of Terras Common to
Baghouses Using Glass Fiber Cloth. Fiber Glass Division, The Gateway
Center, Pittsburgh, PA 15222.
Reigel, S. A., and R. P. Bundy. "Why the Swing to Baghouses?" Power.
,L21j68, 1977.
Summit Filter Corporation. Glossary of Terms. 235 Broad Street,
Summit, NJ 07901.
The Mcllvaine Co., The Fabric Filter Manual, Chapter XIII, Sec. 1.1.
Glossary of Fabric Filter Terras. 2796 Maria Ave., Northbrook, IL
60062.
180
-------
APPENDIX D
SUPPLEMENTARY DATA
BAGHOUSES AND ACCESORIES
REVERSE AIR
PRESSURE •LOWER DRIVE MOTOR
INNER
HOW REVERSE
AIR MANIFOLD
CLEAN
AIR
OUTLET
OUTER ROW
REVERSE AIR
MANIFOLD
FABRIC
FILTER TUBES
MIDDLE ROW
REVERSE AID
MANIFOLD
PAE-CLEANmO
BAFFLE
HEAVY DUST
DROPOUT
SERIES 72 AND 144
ILLUSTRATED HERE
Figure D-l. Cutaway drawing of an "RJ" reverse jet fabric
filter—capacities from 3,000 to 76,500 acfm
(courtesy of CEA—Carter-Day—excerpt from
Bulletin No. RJ-2).
181
-------
HIGH PRESSURE MANIFOLD
IN THE PACTECON CONTINUOUS CLEANING SYSTEM THE TIMER ISO-
LATES ONE MODULE AT A TIME FOR CLEANING AND THE OTHER MOD-
ULES REMAIN IN OPERATION. THE MODULE TO BE CLEANED IS ISOLATED
BY AN AIR-CYLINDER DAMPER (A) ACTUATED BY AN AIR SIGNAL FROM A
3-WAY SOLENOID VALVE (B). WHEN THE MODULE IS ISOLATED THE
TIMER SIGNALS THE 2-WAY SOLENOID VALVE (C) TO OPEN THE DIA-
PHRAGM VALVE (D) EVERY FEW SECONDS, SENDING PULSES OF HIGH-
PRESSURE AIR TO THE FILTER BAGS. THE SHOCK WAVES BALLOON THE
BAGS OUTWARD, CRACKING THE DUST CAKE LOOSE TO DROP INTO
THE HOPPERS.
THE CUSTOMER CAN REGULATE THE NUMBER OF PULSES PER CYCLE,
THE SIZE OF THE PULSE C"ON TIME") AND THE DELAY TIME BETWEEN
CLEANING EACH MODULE.
THE WW SLY MANUFACTURING CO
CLEVELAND
PC AIR FLOW
P-1161
Figure D-2.
Schematic of air flow in a "PC" Pactecon dust collector
(courtesy of W. W. Sly Manufacturing Co.—excerpt from
Instruction Book No. 693).
182
-------
00
MULTIPLE RATING TAILES
MOOii
1304194
1-8" dia inlet
5 horsepower
130-2T>5-7V!
1-10" dia. inlet
7W horsepower
130-24-10
1-1Z" di». inlet
10 horsepower
130-24 15
1-12" dia. inlet
15 horsepower
C*M
1WO
2100
2300
2500
2700
2450
2600
3000
3200
3600
3400
3600
4000
4400
4800
4500
4900
5000
5300
5600
IHLtT
vnocirr
<*PM)
5140
6000
6570
7140
7?lt>
4500
5140
5500
5670
6600
4330
4580
5090
5600
6110
5725
6110
6360
6700
7130
IXTERNAL
STATIC PRISSUni
(IKCHIS W.C.)
7.75-
7,W
6.50-
6.00"
5.70"
12.00-
11,00-
10.30-
9.60"
8.00"
12.90-
12.40"
11.10"
980-
8.30*
9.40-
8.30"
760"
640-
s.2tr
PRISUIRI 0*0*
CLIAN FIIT(RS
(IHCHIS W C )
.25-
2T
31"
.»'
.39-
.ir
.4r
-4«"
.50"
.«r
.50"
.55"
.er
,7
-------
BUTTERFLY VALVE
GUILLOTINE OR SLIDE GATE VALVE
Figure D-4. Typical damper valves used in
baghouse systems.
184
-------
PIPE SHAF
tea A.BRASON
PROTECTION
AND RIGIDITY
POPPET DAMPERS
The AAF Poppet damper includes
specific design features which result
in quality beyond that normally found
in the industry The most common
configuration is a vertical acting air
cylinder-operated valve The AAF de-
sign rehes on a combination of
rigidity, deflection and flatness for
proper sealing Most designs depend
only upon deflection to accomplish
sealing Provisions are also included
which eliminate blade rotation inde-
pendent of the type of operator
employed Elimination of rotation en-
sures repeatable blade-seat match-
up which is a necessary damper
characteristic M it is to effect proper
sealing
AAF incorporates positive sealing
at both ends of the damper operating
stroke This feature, m conjunction
with the positioning and type of
guide bearing employed, minimizes
undesirable axial motion m the area
of the damper shaft seal between the
process air flow and atmosphere
Through experience and continued
design improvement, AAF has devel-
oped a damper which requires very
little seal maintenance and operates
effectively when installed upright, in-
verted or horizontally
When internal compartment in-
spection or bag maintenance is
required, positive mechanical lock-
out is used to avoid accidental
operation of the damper The me-
chanical lockout system is double
locking, using vise clamping and
pinning.
A special patented feature in-
cluded in AAF poppet damper
application is the "inching circuit
Until AAF developed this circuit, bag
popping was frequently encountered
when poppet dampers were em-
ployed. AAF's inching circuit
provides control of damper move-
ment with adjustable lift and dweii.
wnich eliminates the need for the
separate reinitiation devices so com-
pletely that the old approach 10 soft
remflaiion is now considered
obsolete
Figure D-5.
Poppet damper detail (courtesy of American Air
Filter Company, Inc.)-
185
-------
ON STREAM
AIR FLOW
TOP VIEW
SIDE VIEW
Dirty air flows into expansion chamber where large
particulars drop out. Fines are carried upward by air
stream between bags and are deposited on exterior
surface of fabric with filtered gas being discharged at
the top of the bag into the bag manifolds. The clean
air paaaes over the top of the unit down through the
outer double wall of the collector, acting as insulation,
into an exhaust plenum at the bottom of the collector.
This plenum insulates the hopper and acts as a duct to
fan connection at a level where fan is normally located.
This keeps the external duct work and supporting
members at a bare minimum.
TRADITIONAL DUST COLLECTOR WALL
. STlFflKINC
FIANCES
C.»Snn AND BOLT
DUSTEX DUST COLLECTOR WALL
CHAMBER
[LIMINATF.5
INSOLATIOK AND
ACTS AS CAS DUCI
! ^
^~
i
E5
AND
DUCT
V f
\ i
\
JUIHFORCIO
OOUBLF. WAIL
PKOVIDES
STRONGCB COLLECTOR
»ALL BJQUIBIS
NO "SS1UBIV 00
MAIBHNANCi
REVERSE STREAM
CLEANING ACTION
TOP VIEW
SAC CROSS
SECTION
CLEANING-
SHORT BURST
OF AIR
CLEANS BAGS
CLEANING MB
CONNECTION
SIDE VIEW
Only a small percent-age of bags are cleaned at a time,
assuring continuous operation and virtually eliminating
fluctuations in system gas flow. Upon a signal from the
control box, two rows of bags (28 bags) are isolated from
the exhaust plenum by means of a pneumatically oper-
ated damper.
The damper permits a sudden, short burst of clean
air from the reverse air plenum to enter-the filter bags
resulting in a controlled "snap action" of the bags
which breaks the dust cake off the outside of the bags.
The damper then returns to its normal position, per-
muting clean air to flow up the bags and out into the
exhaust plenum. The entire cleaning proceaa for two
rows of bags is only seconds in duration.
At controlled intervals, each subsequent two rows of
bags is isolated from the exhaust plenum by means of
the control system and cleaned in the manner described
above. The entire cleaning process is continuously
repealed during the operation of the collector. The dual
cake drops to the bottom of the collector where it is
collected and discharged by means of the screw conveyor.
Figure D-6.
Schematic diagrams of filtering and cleaning operation
modes in a type DW reverse pulse collector (courtesy
of American Precision Industries, Inc., Dustex Div.—
excerpt from Bulletin No. 571.)
186
-------
APPENDIX E
U.S. BAGHOUSE MANUFACTURERSa
Abart Engineering Ltd.
Ace Engineering Company
Aerex Corporation
Aerodyne Development Corporation
AeroPulse, Inc.
Aerosols Control Corporation
Aget Manufacturing Company
Allis-Chalmers Corporation
American Air Filter Co., Inc.b
Andersen 2000 Inc.
Babcock & Wilcox Company, The
BACT Engineering, Inc.
Bahco Systems, Inc.
B.B. Barefoot & Associates, Inc.
Belco Pollution Control Corporation'5
Beltran Associates, Inc.
Bionomic Industries
Black Clawson, Inc.
Buell Division, General Electric
Buffalo Forge Company'3
Buhler-Miag, Inc.
Cadre Corporation, The
Carborundum Environmental Systems'5
Division,
Flakt, Inc.
CEA Carter-Day Company
CEA Combustion, Inc.
CEA Simon-Day Ltd.
C-E Air Preheater
Sub. of Combustion Engineering Inc.
C-E Raymond/Bartlett-Snow,
A Division of Combustion
Engineering, Inc.
Continental Air Filters Co.
Continental Air Products, Inc.
Crystal-X Corporation
DCE Vokes, Inc.
Dust Control Company
Dustex Division,
American Precision Industries, Inc.
Dust Suppression Systems, Inc.
Eastern Control Systems, Inc.
Kcolaire Environmental Company
Fisher-Klosterraan, Inc.b
Flakt Canada, Ltd.
Flakt, Inc.
Flex-Kleen Corporation
Fuller Company,*3
Dracco Products
General Electric Environmental
Systems, Inc.
General Resource Corp.
Griffin Environmental Company, Inc.
Industrial Clean Air
Interel Corporation
Jazco Corporation
Johnson-March Corporation, The
Kleissler Company, G.A.
Lear Siegler, Inc.
Environmental Technology Division
MAC Equipment Inc.
MacDonald Steel (1976) Ltd.
Mahon Industrial Corporation,
Subsidiary of Pullman Inc.
Mclnnis Equipment Limited
Mikropul Corporation,
U.S. Filter Corporation13
Mine Safety Appliances Co.
Neptune AirPol Inc.
Niro Atomizer Inc.
Norblo Division, Envirotech Corporation
Peabody Engineering Corporation0
Research-Cottrell, Inc.0
Rolfes Comapny, George A.
Ruemelin Manufacturing Co.
Sly Manufacturing Co., The W.W.
Smith Environmental Corp.
Standard Havens Inc.
Steelcraft Corporation
Sternvent Company, Inc.
Tag Construction Company
Terrell Machine Company
Thyssen Environmental Systems, Inc.
Torit Corporation, TheD
United McGill Corporation,
Dust Collector Division
187
-------
U.S. BAGHOUSE MANUFACTURERS* (Cont'd)
Eltron Manufacturing, Inc. USF Environmental Systems Corp.
fimtrol Emission Control Systems Western Precipitation Division of Joyb
Environmental Elements Corporation Manufacturing Co.
Subsidiary of Koppers Company, Inc. Wheelabrator Frye, Inc.
Enviro-Systems & Research, Inc. Air Pollution Control Division*5
ESSTEE Manufacturing Co., Inc. Wiedenmann & Son, Inc., W.C.
Fabric Filters Corporation Young Industries, Inc. The
Ferro Tech Inc. Zurn Industries, Inc.,0
Air Systems Division
aFrom APCA Directory and Resource Book (1982-1983)
Air Pollution Control Association, P.O. Box 2861
Pittsburgh, PA 15230
bMember of the Industrial Gas Cleaning Institute (IGCI)
Suite 304, 700 N Fairfax St., Alexandria, VA 22314
188
-------
APPENDIX F
FABRIC FILTER MODEL
(CAPSULE DESCRIPTION)
BACKGROUND
A comprehensive treatment of the development and design of a new fabric
filtration model, structured specifically so that enforcement engineers can
assess the fly ash control capabilities of filter systems used at coal burning
facilities, has been discussed extensively in the literature (1,46,48,100,121,
173,174). To facilitate and encourage model application,* a condensed Users
Guide was prepared (174) to demonstrate the ease with which the model could be
used by any engineer possessing a rudimentary knowledge of the fabric
filtration process. The following excerpts highlight the basic model function
by describing a typical field application.
EXAMPLE OF MODEL APPLICATION
An electric utility operates two, coal-burning steam-electric plants, the
first of which (Plant A) now uses a pressure-controlled baghouse to prevent
particulate (fly ash) emissions. It has been proposed that a continuously
cleaned fabric filter system be installed at the second plant (Plant B). Both
the utility operator and the local emission enforcement groups would like to
*Model based on Fortran language and designed initially for compatibility
with an IBM 360 computer or equivalent system via card deck or tape access,
189
-------
determine whether operation of the filter system in accordance with the input
data shown in Table F-l (some of which is based on Plant A experience) will
satisfy local emission requirements while maintaining average system pressure
drop levels within the exhaust capacity range of the induced draft fans. For
present purposes, it is assumed that operation at an efficiency of
99.5 percent (equivalent to 0.5 percent penetration) and an average pressure
o
drop of j.750 N/m (7 in. water) indicates acceptable performance.
Design and operating data appearing in Table F-l for the proposed
"second" plant baghouse represent a composite of information received from
both utility personnel and the dust collector manufacturer. It is also
assumed that previous measurements of uncontrolled mass emission rates and fly
ash size parameters at both plants are available, as well as estimates for the
terms K_, Sg, and W~ based upon prior tests performed at Plant A.
The data shown in Table F-l are sufficient to carry out the predictive
modeling operation. After converting the terms to their metric equivalents,
Table F-2 (noting that raw field data are frequently supplied in English
units), the data inputs are ready for punchcarding or terminal entry as shown
in the card listings of Figure F-l.
Card 1 contains the title that will appear on the program output. Note
that on Card 2 the time between cleaning cycles, Item 4, and the limiting
pressure, Item 5, have been left blank because a continuous cleaning system
has been chosen. The reverse flow velocity, Item 6, Card 2, is calculated
from the total reverse flow rate (30,000 acfm) and the cloth area per
compartment (200,000 ft2/30) as 4.5 ft/min or 1.37 m/min. The average face
velocity, Item 9, Card 3, is computed from the indicated value for the total
190
-------
TABLE F-l. AVAILABLE INPUT DATA FOR MODELING BAGHOUSE PERFORMANCE
AT ELECTRIC UTILITY, PLANT B
Plant A
Plant B
Number of compartments
Cleaning cycle duration
Time to clean one compartment
Cleaning type
Reverse flow volume
Cleaning cycle initiation
Volume flow into baghouse
Total filtration area
Temperature of flue gas
Inlet concentration
Inlet dust mass median diameter
Inlet dust geometric standard
deviation
Dust specific resistance, K£
Measured at
Measured at
Effective residual drag
Measured at
Residual fabric loading
10 urn (Reference)
3.0 (Reference)
10.2 in.W.C.-ft-min/lb
500°F
2 ft/min
0.636 in.W.C.-min/ft
500°F
0.015 lb/ft2
30
30 minutes
1 minute
Collapse/reverse air
30,000 acfm
Continuous cleaning
600,000 acfm
200,000 ft2
350°F
5 grains/scf
7 ym
2.5
Note: All English units must be converted to their metric equivalents.
TABLE F-2. ENGLISH/METRIC CONVERSION FACTORS
Quantity
To convert from
To
Multiply by
Filter resistance
Filter drag
Velocity
Volume flow
Fabric area
Areal density
Specific resistance coefficient
Dust concentration
Density
in. H20
in. H20-min/ft
ft/min
ft3/min
ft2
lb/ft2
in.W.C.-min-ft/lb
grains/ft3
lb/ft3
N/m2
N-min/m-'
m/min
m3/min
m2
g/m2
N-min/g-m
g/m3
g/cmj
249
817
0.305
0.0283
0.093
4882
0.167
2.29
0.0160
191
-------
C
A
R
D
aloa THE SAME CARD, larvt OME on
I, (THE OTHCR.BUT HOT BOTH
JO® Raj^SSR
10/05/7$
)-REOU(R£0 fOK »O«-Lltl£AR
FOR
> ^ 3 4 5 6 T a S 10 IU2 IIM 1516
34
Figure F-l. Fabric filter model - data input form prepared from Table F-1 input data.
-------
o
gas flow (600,000 acfm) and the total fabric area (200,000 ft ) as
3.0 ft/rain (0.915 m/rain). Inlet concentration is reported at ambient
temperature, 25°C, the value entered for Item 12, Card 3.
The data available from Plant A allow for adjustment of the measured K-
value at Plant A for the size properties of the dust at Plant B. Thus K^,
the temperature and face velocity at which it was measured, and the size
properties of the two dusts are entered as shown on Card A.
Twenty cycles are considered sufficient to complete the simulation and
achieve steady state conditions (see Card 6). Similarly, an accuracy level of
zero is considered acceptable for the first trial. Because the utility
personnel and the enforcement agency are concerned mainly with average
emission rates and average pressure drop, AVERAGE results are requested.
Since no plotting is desired, Item 32, Card 7 has been left blank.
If the results of the simulation had indicated emission levels close to,
but greater than, the allowable level, the simulation could have been rerun
with an accuracy level of 1. If convergence had not been reached within
20 cycles, a value of 40, for example, might have been entered provided that
the costs for added computer time were acceptable.
In Table F-3, the actual computer printout provided for the input data
and the coded instructions of Figure F-l have been arranged in a convenient
tabulation showing each of four separate printout sheets. Printout No. 1
shows the actual summarized input data as entered into the program so that the
user can check the data for errors or omissions. Printout No. 2 instructs the
user via the statement "There are no errors in the input data" that the
modeling program will be executed as requested. Printout No. 3 provides a
listing of those parameters whose values are computed or corrected by the
193
-------
TABLE F-3. SAMPLES OF TABULAR PRINTOUT FOR EXAMPLE OF MODEL APPLICATION
PRINTOUT NO. I
SUMM*«Y (JF INPUT OAT* FOB BAGHOU5E ANALYSIS
AN HECtWIC uTlLlTr / PLANT H
MASK DESIGN DATA
N^KtW DF CO»'<>AHTUENTS
«JFF LINE TIWF)
tANINt; CTCLt TI"E
^t IMuQiJSL Y CLEANED SYSTf"
LOn VHOCITY
JO
1.0
JO.O
1.J725
"IMUTE3
"INL-TES
T(NG DAT*
AV£M»CE FACE VELOCITY
INLET OUST
AT
f»HNlC »NO OUST PBOPEMTItS
SPECIFIC BESI3T4NCF, K2
AT
COWBfcCTEO TO ""02
EFFECTIVE RESIDUAL OMAG, SE
AT
LOADING, «W
0.9)50
177.
1 J.U5
1 .70
260.
O.blOO
10.0
7.0
520.
260.
73.?
OeGRttS CENTIGRADE
G/MJ
OEGHEES CENTIGRADE
N.MJM/G-M
OEGBEES CENTIGRADE
MICRONS
X 1C HUNS
-STANDARD DEVIATION 5.00
-STANDARD DEVIATION 2. SO
OFGHtES CENTIGRADE
SPECIAL "BOGBAM IKSTRUCTIONS
M«X NUWRER of CYCLES MODELED 20
ACCURACY LEVEL 0
TYPE or WESULTS RCOUESTED AVERAGE
PRINTOUT NO. 2
DIAGNOSTIC "fSSAGES
ARE NO EBROBS IN THE INPUT DATA
194
-------
program, such as a^ and K.^- Again, inspection of these data by the model
user allows him to determine the reasonableness of the indicated values.
Finally, the AVERAGE data shown in Printout No. 4 indicate that both the
r\
pressure drop and penetration expectations for the filter system ( 1750 N/m
and 0.5 percent) should be realized. In addition, Printout No. 4 also
indicates that 10 cycles, rather than the 20 requested on the data input form,
were sufficient to define steady state operating conditions.
195
-------
APPENDIX F*
REFERENCES
1. Billings, C. E., and J. E. Wilder. Handbook of Fabric Filter Technology,
Vol. 1. GCA/Technology Division, Bedford, Massachusetts 01730.
EPA-APTD 0690 (NTIS PB-200-648), December, 1970.
46. Dennis, R. , and H. A. Klemm. A Model for Coal Fly Ash Filtration. J.
Air Poll. Control Assoc., 4_9(3):230, 1979.
48. Dennis, R., et al. Filtration Model for Coal Fly Ash with Glass
Fabrics. EPA-600/7-77-084. (NTIS PB-276-313/4BE), August 1977.
100. Dennis, R., H. A. Klemm, and W. Battye. Fabric Filter Model Sensitivity
Analysis. EPA-600/7-79-043c, 1979.
107. Dennis, R., and J. E. Wilder. Fabric Filter Cleaning Studies.
EPA-650/2-75-009 (NTIS No. PB-240-372/3G1), January 1975.
121. Dennis, R., R. W. Cass, and R. R. Hall. Dust Dislodgement From Woven
Fabrics Versus Filter Performance. J Air Poll. Control Assoc. 4_8_(1), 47,
1978.
173. Dennis, R., and N. F. Surprenant. Particulate Control Highlights:
Research on Fabric Filtration Technology. EPA-600/8-78-005d.
(NTIS PB-285-393/5BE), June, 1978.
174. Dennis, R., and H. A. Klemm, "Fabric Filter Model Format Change,"
Volume I Detailed Technical Report, Volume II User's Guide.
EPA-600/7-79-043a (NTIS PB-293-551/8BE). EPA-600/7-79-043b
(NTIS PB-294-042/7BE), February, 1979.
*Reference numbers are those shown in main text.
196
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
RtPOHT NO.
4. MTLt
3. RECIPIENT'S ACCESSION NO.
5. fltPORT DATE
January 1986
Fabric Filter Inspection and Evaluation Manual
6. PERFORMING ORGANIZATION CODE
7 AUTHORIS)
Douglas R. Roeck and Richard Dennis
8. PERFORMING ORGANIZATION REPORT NO.
GCA-TR-80-114-G
fC Ht-ORMINC. ORGANIZATION NAME AND ADDRESS
GCA/Technology Division
213 Burlington Road
Bedford, Massachusetts 01730
12 SPONSORING ACI NCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Stationary Source Compliance Division
Washington, D.C. 20460
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-01-4143
68-01-6316
13. TYPE OF REPORT AND PERIOD COVfcRED
Final
14. SPONSORING AGENCY CODE
fiul'Ci k Ml N I A«Y NOTFS
AH}, TFIACT
The Fabric Filter Inspection and Evaluation Manual was prepared to assist Federal and
State enforcement groups in the following decisionmaking areas: estimation of filter
system (baghouse) compliance with emissions regulations; appraisal of filter system
adequacy for a specific control application; and evaluation of operating and main-
tenance procedures in the light of recommended practices. In Chapter 2, basic concept
pertaining to fabric filtration, particle behavior, and particle size measurements are
highlighted so that inspection personnel can evaluate facilities for which no
precedence has been established. Given prior experience or the need for immediate
<•.'.<.-LLon, however, the Manual user may go directly to Chapter 3 where important day-
io-day aspects of filter system operation are presented along with emphasis on what may
>:.> wrung and what remedial measures should be undertaken. Different types and
procedures for baghouse inspections are described in Chapter 4; e.g., compliance
determination, startup, troubleshooting, general preventive maintenance, or special
:nvL-sLigations, where the specific information sought and the sequence of the inspec-
tion process may vary.
A .sample ropert format is presented that indicates the areas of concern for emissions
compliance determinations. Supporting material are provided in several appendices
i :u-1 luJinn a comprehensive glossary of terms used in fabric filtration technology.
KEY WORDS AND DOCUMENT ANALYSIS
l>r SCRIPT ORS
/' i r Po I I tit ion
Air Filters
(!ompl. iance Determinations
I iI tor Maintenance
I iIter Inspections
Woven Fabrics
l;ol is
b.lDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Fabric Filters
Mechanical Shaking
Pulse Jet Air
Reverse Flow Air
c. COSATI Meld/Group
1R OlSfKiBU) ION STATEMENT
19. SECURITY CLASS (This Report)
21.
NO. OF PAGES
208
20. SECURITY CLASS (This page)
22. PRICE
EPA Form 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE
-------
United States Office of Air and Radiation
Environmental Protection Office of Air Quality Standards and Planning
Agency Washington DC 20460
Official Business Publication No. EPA 340/1-84-002
Penalty for Private Use
$300
If your address is incorrect, please change on the above label;
tear off; and return to the above address.
If you do not desire to continue receiving this technical report
series. CHECK HERE D; tear off label; and return it jo the
above address
------- |