United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-78-194
October 1978
High-temperature,
High-pressure
Participate Control
with Ceramic Bag
Filters
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are'
1 Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5. Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8 "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect the
views and policies of the Government, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service. Springfield, Virginia 22161.
-------
EPA-600/7-78-194
October 1978
High-temperature, High-pressure
Particulate Control with Ceramic
Bag Filters
by
M. A. Shackleton
Acurex Corporation
485 Clyde Avenue
Mountain View, California 94042
Contract No. 68-02-2169
Program Element No. EHE624
EPA Project Officer: Dennis C. Drehmel
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
TABLE OF CONTENTS
Section Page
1 INTRODUCTION 1
1.1 Test Plan 6
APPENDIX A 83
-------
LIST OF ILLUSTRATIONS
Figure Page
1 Hot Filtration Facility 9
2 Test Chamber Cross Section 11
3 Filter Support 13
4 HTHP Media Test Facility Schematic 15
5 Test Points, Temperature/Pressure 16
6 The Effects of High Temperature and Pressure on the
Collection Efficiency of a Fiber Bed 22
7 Calculated Performance of 3.0ym Alumina Fiber Bed ... 26
8 Fiber Bed Thickness 28
9 OOP Efficiency Airflow Velocity 33
10 OOP Efficiency Test Data 35
11 Dust Loading of Ceramic Felts 37
12 Dust Loading of Ceramic Paper 38
13 Dust Loading of Woven Ceramic Media 39
14 Post-Test Dust Cake — Test B103 3M-AB312
Twill Weave 43
15 Irish Refrasil Pulse Test Failure 45
16 Irish Refrasil Pulse Test Failure 47
17 Refrasil Media Seam Failure 51
18 Saffil Alumina Paper After Pulse Testing 55
19 Saffil Alumina Paper After Pulse Testing 57
20 Saffil Alumina Felt Bag Post-Test 61
21 Separated Seam 63
VI
-------
LIST OF ILLUSTRATIONS (Concluded)
Figure Page
22 Saffil Alumina Blanket After Pulse Test 65
23 Saffil Alumina — Post-Test Dust Cake 69
24 Woven Fiberfrax — Post-Test Dust Cake 71
25 Fiberfrax Blanket — Post-Test Dust Cake 73
26 Filter Characteristics 2.5 cm/sec 77
27 Outlet Concentration 2.5 cm/sec 78
28 Filter Characteristics 9 cm/sec 80
29 Outlet Concentration 9 cm/sec 81
vii
-------
ACKNOWLEDGEMENT
Room ambient tests of filter media were performed at Donaldson Company
Inc., Minneapolis, Minnesota by Harry Complin and Eugene Grossel.
High temperature/Pressure testing was performed at Acurex Corporation,
Mountain View, California by Chris Chaney.
-------
SECTION 1
INTRODUCTION
Many advanced technology processes currently being developed
require removing particles from high temperature and pressure gas
streams. An objective of developing these processes is to increase coal
use by making it economically efficient and environmentally safe. The
processes most actively being studied are pressurized fluidized bed
combustion (PFBC) and gasification combined cycle (GCC) plants. These
processes involve expanding the high temperature and pressure gases across
a turbine which generates power to produce electricity. Such applications
require removing particle flyash from the gas streams before it expands
across the turbine.
In the PFBC, coal is burned in a fluidized bed of limestone (which
removes the SOp), and heat is transferred to tubes in the fluidized
bed. Up to 80 percent of the recoverable heat value of the coal is
removed in the fluidized bed, and the gas exits at 815°C (1500°F) and
10 atm pressure.
The gas must then be expanded through a gas turbine to recover the
remaining energy. However, previous investigations showed that large
particles erode turbine blades, and small particles cause deposits that
choke the turbine. To protect the turbine, some high temperature
particulate control is required. Moreover, since it would not be
-------
economical to duplicate particulate control for environmental regulations,
following the turbine, high temperature particle control must also meet
new source performance environmental standards for coal-fired utilities.
This currently allows emissions to be not greater the 0.1 Ib million Btu
(0.043g/MJ) and may be reduced.
To meet both the environmental and turbine requirements, a system
consisting of two cyclones and a filter was studied. Although the two
cyclones lower the overall particle concentrations, they fail to remove
small particles. Concentrations leaving the second cyclone can be as high
as 2.29 g/Nm (1.0 gr/scf) and have a mass median diameter of 5.0 ym.
Under the contract, research was aimed at demonstrating the feasibility of
using ceramic filters in a barrier filter system to accomplish the
tertiary collection of fine particles following a cyclone train.
Major goals of this program were to:
• Design and build a filter media test facility capable of
operating at 815°C and 10 atm pressure
• Test available ceramic fiber forms (woven cloths, felted mats)
to determine if any can survive mechanical displacements, and
accelerations that are likely to be encountered in online
cleaning of high temperature filter applications
• Develop preliminary performance data for those configurations
which appear most promising for high temperature filter
applications
• Make recommendations based on the experience and data collected
Conclusions
Research on bench-scale indicates that fine particle control at
high temperature and pressure can be achieved using barrier filtration by
-------
ceramic filter beds. Evidence in support of this contention includes the
following:
0 A theoretical basis exists for it
• Several of the ceramic paper and felt materials are capable of
removing fine particles at high efficiency without excessive
filter weights
• The ceramic paper and felt materials have filtration
characteristics and performed similarly to paper and felt
commercial filter media in a series of filter media tests
• The ceramic woven materials in general were characterized by
large pores and poor collection efficiency in the dust loading
tests. The range of parameters exhibited by the various
materials, however, indicates that an acceptable woven ceramic
filter media can probably be fabricated; but such filter media
would have the same limitations as currently available woven
filters. That is, acceptable performance would only occur at
low air-to-cloth ratios.
• "Blanket" ceramic fiber materials (felts) consisting of small
diameter fibers (3.0 pm) appear to be the most promising
materials for high temperature and pressure applications
because of their combination of good filtration performance and
relatively high strength
• Accelerated media cleaning tests at high temperatures and
pressures show that several ceramic filter structures are
capable of surviving in excess of 50,000 cleaning pulses while
maintaining pressure drop at acceptable levels
-------
• Filtration tests at high temperature and pressure have shown
that pressure drop can be controlled at air-to-cloth ratios of
up to 9 cm/sec for a period more than 200 hours. Furthermore,
this was accomplished while maintaining high efficiency
particle removal with outlet concentrations less than published
turbine requirements. The ability of the filter to control
9
dust in high concentrations (14.4 g/Nrrr) was also
demonstrated.
t Theory as well as tests indicate that high filter media face
velocity (air-to-cloth ratio) operation will be possible for a
HTHP filter system. To accomplish operation at high
air-to-cloth ratios will require special cleaning techniques.
High filter velocity operation is a desirable objective because
smaller, less expensive particle control equipment will result.
• Innovative cleaning and media support techniques can be
designed which are compatible with the unique properties of
ceramic filter media
Recommendations
Work performed under the contract represents a major advance in the
state-of-the-art of filtration. High efficiency particle control can
potentially be extended by over 500°C from current industrial
limitations of 300°C. Consequently, Acurex strongly recommends that the
government continue to support advanced development of ceramic fiber
filtration technology to make the potential benefits arising from this
development available to the public.
High temperature ceramic fiber filtration is not limited to
applications for pressurized fluidized bed combustion but could provide
-------
fine particle control in many high temperature industrial processes.
Applications such as these could provide heat recovery after particle
removal, offering substantial savings in energy.
Current trends in the development of pressurized fluidized bed
(PFBC) combustion technology indicate there is a continuing hope that
turbines can be adequately protected with particle removal from staged
inertial separation devices. Military experience does not support this
hope. Helicopter turbine engines have been fitted with small high
efficiency 2.5 cm diameter cyclone tube banks to extend service life from
unacceptable to moderately acceptable in this application where dust
loading is intermittent. The U.S. Army XM-1 Main Battle Tank is turbine
powered. It employs a barrier filter system with a cyclone tube bank
precleaner capable of providing engine intake air as clean as that
required of heavy duty diesel engines. In the world's most severe dust
conditions (YUMA proving ground Yuma, Arizona) dust loadings at the rear
deck of a tracked vehicle are about 17 g/Nm . Dust loadings in excess
of this are encountered in the exhaust of the PFBC.
To avoid a crash development program which would potentially delay
the availability of advanced coal combustion systems, development of
ceramic fiber, barrier filter systems should be continued. This
development should proceed in parallel with current testing of inertial
separators for PFBC applications. In the unlikely event that commerical
power turbines will be shown capable of extended operation protected only
by cyclones, the ceramic filter technology will still offer substantial
benefits in other applications for protecting the environment.
A recommended development program should contain the following
features:
-------
Continued Laboratory-Scale Component Development
Many refinements are needed. Media configurations and ceramic
screens should be tested to eliminate metal screens used to support the
media in the current tests. Media cleaning techniques should be
optimized. Further extension of filter velocity should be studied to
provide data needed for economic predictions of commercial size equipment.
Pilot-Scale Tests
Pilot-scale units for PFBC applications should be built and tested
(using current knowledge) to obtain field data needed for optimal designs.
Atmospheric Pressure Applications
Small-scale units should be built to investigate suitability for
applications such as In the carbon black industry, at smelters and kilns
and other high energy loss sites.
1.1 TEST PLAN
To provide the data needed to justify further development of high
temperature, high pressure barrier filtration equipment, Acurex outlined
the following test plan.
Obtain and Classify Ceramic Media Candidates
A survey of available ceramic filters and materials was made (a
description of various media characteristics is included in Appendix A).
The usable materials are generally made for insulation applications; those
that are are not rigid in structure can be categorized Into three groups:
(1) Woven fabrics, produced by several companies, are made from yarns
produced from a continuous length of ceramic fibers. Available in various
weaves, they are flexible, and exhibit good strength relative to other
ceramic fibrous structures. (2) Ceramic papers, constructed from short
fibers generally held together with binders, are available and are
-------
characterized by a relatively higher packing density of fibers and poor
mechanical strength as compared with other ceramic fibrous structures.
(3) Ceramic felts, consisting of relatively long fibers, are also
available. They are relatively porous mats held together by randomly
intermixing the fibers. They are flexible, and exhibit the highest
strength in a filtration application.
Room Ambient Filter Media Tests
Over 30 ceramic media candidates were subjected to a series of
filtration tests to determine their ability to remove particles from a gas
stream. These basic filtration tests included Dioctylphtalate (OOP) smoke
penetration tests, maximum pore size, permeability and flat sheet dust
loading tests. Such tests were intended to show whether or not the media
were capable of particle removal.
Mechanical Screening Tests
In this phase of the test program, the various candidate filter
media were subjected to "filter cleaning loads" to determine their
relative strength. These tests were performed on samples shaped into
tubular bags. The tests were performed at an air-to-cloth ratio of 5 to 1
(2.54 cm/sec) at high temperatures and pressures in the presence of
flyash. Samples were Initially subjected to reverse flow cleaning cycles,
but the primary test was pulse cycling. While maintaining forward flow,
the sample was deflected with an air injected pulse using 5 to 10 pulses
per minute from a reservoir pressure of 1100 kPa (160 psig), lasting
100 ms. The relative strength of each sample is determined by the number
of cycles it withstands. If several thousand cycles are required for
failure, then it is probable that the filter would last a long time if the
cleaning technique is perfected.
-------
Filtration Performance Tests
Those ceramic filter media identified from earlier tests as being
most promising for the PFBC application were subjected to these tests,
lasting 200 hours. Flyash was metered into the gas stream at a controlled
rate while monitoring pressure drop. Pressure drop was controlled with
cleaning cycles and overall dust removal efficiency was determined from
analysis of the weight of dust fed and that which penetrated the test
filter.
Ceramic Media Test Facility
The Hot Filtration Testing Facility (shown in Figure 1) is capable
of subjecting 46 cm (18 inch) by 9.5 cm (3-3/4 inch) diameter test filters
to an 815°C, 10-atmosphere environment with air-to-cloth ratios from
0.5 to 5 cm/sec (1 to 10 ft/min). Moreover at reduced pressure (7 atm),
higher face velocities are achievable, 9 cm/sec (18 ft/min). During high
dust loadings, filter elements tested in the facility can be either
reverse-flow or pulse-jet cleaned. There are three test chambers so that
three bags can be tested simultaneously, though only one chamber was used.
The filter media test chamber (Figure 2) is a 244 cm (8 foot) by
30.4 cm (1 foot) diameter length of pipe with inlet and outlet tubes
welded into the blind flanges on the top and bottom. The inside of the
chamber is lined with 5 cm (2 inches) of castable refractory which
surrounds a 10 kilowatt, 230 volt electrical resistance heater used to
compensate for refractory and external piping heat losses. Clean, high
temperature, high pressure air enters the chamber through the bottom. A
separate line delivers dust from the feeder to the test chamber above the
air inlet (Figure 2).
-------
Figure 1. Hot filtration facility.
-------
Pulse In
Plenum
Venturi
Chamber heater
t
1
1
1
J
\j
\
\
\
\
\
\^
^
\
\
\
\
\
\
\
\
\
\
\
\
V
^
\
\
N
1 A-\«TO»
1
n
*
f
•
•
i
t
t
^
j
1 1
^
[
i
')
•
i
-^
^\\^\
\
s
\
s
s
s\\
\
\
\
\
\
K;
k
^
\
\
,
\
\
^
\
^
\
\
s
\
\
\
\
\
\
V
\
\
\
\
\
\
I es
Dus
1 1 /
V I/
0.
L
^-1
1
\
Air outlet
feeder
inlet
(l/4"-tube)
> cm
Air inlet
(3/4-inch tube)
1.905 cm
Figure 2. Test chamber cross section
11
-------
The 46 cm (18 inch) long test bag is above the dust feeder inlet.
The cylinder shaped filter is mounted around a 9.2 cm (3-5/8 inch)
diameter support cage screen (Figure 3) which prevents the filter from
collapsing during outside-in filtering. The test filters are open at both
ends, requiring clamping at the top around the venturi and at the bottom
around a sealing plate attached to the cage. The filtered air passes
through the venturi, into the pulse-jet plenum and exits the chamber
through the outlet tube.
Once the filtered air leaves the test chamber, it is cooled with a
forced air, finned tube heat exchanger. It is then directed through a
commercial, low temperature, high efficiency, tube-type filter
(Figure 4). This filter can be inspected periodically to determine the
test filters particle removal efficiency. Downstream of the low
temperature filter, the high-pressure air is bled through a sonic-flow
metering orifice, through a noise suppressor, and vented to the
atmosphere.
Gas temperatures are measured using thermocouples located at
selected points within the facility (Figure 5). The air temperature
measurements are taken at the test chamber inlet, above and below the test
filter, at the low temperature filter, and at the orifice meter. These
temperatures are displayed in the control room and stored on a 24 channel
strip chart recorder. Pressure measurements using diaphragm pressure
transducers are taken above and below the filter, at the orifice and in
the pulse-jet reservoir (Figure 5). This information is also displayed
and recorded in the control room.
There are two methods available for cleaning the test filter during
operation: pulse-jet and reverse-flow. The pulse-jet method cleans the
12
-------
Figure 3. Filter support
13
-------
Pulse flow
Choked
ori fi ce
Air supply
150-250 psi
Pulse
/~ reservoir
Filter
Muffler
Air supply
150 psia
Figure 4. HTHP media test facility schematic.
-------
Pulse valve
Oust feed
line
Reverse —£X
flow **
CX
Induction
heater
coil
-------
filter using quick blasts of high pressure air, introduced at the clean
side. Pulse pressure can be regulated from the test chamber pressure up
to 1723 kPa (250 psi). Fast-acting solenoid valves provide 150 msec, or
other duration, pulses at a variable interval. Pulse-jet cleaning can be
done on line or off line.
Reverse-flow cleaning is a milder cleaning technique, requiring
interrupting forward flow. The first step of the reverse-flow cleaning
sequence is to shut off the main flow and bleed the test chamber down to a
lower pressure. Next, the downstream outlet valve is shut and the main
flow redirected in through the outlet tubes at the top of the chamber.
The inside out flow across the filter blows the dust off the bag. This
process stops when the chamber is once again pressurized. Reverse-flow
effectiveness can be varied by adjusting either the bleed-down pressure or
the rate of repressurization.
Both cleaning operations can be controlled manually or
automatically at the control room. A digital time sequencer, in
combination with a solid-state logic sequencer, can be programmed to
provide automatic cleaning. The following parameters are adjustable:
pulse frequency and duration, or reverse-flow frequency, bleed-down
duration, and repressurization rate. The number of cleaning cycles in a
given time can be either predetermined, or dependent on the pressure drop
buildup across the bag.
Before entering the test chamber, the high-pressure air is heated
to 815°C (1500°F). This is accomplished by passing the air through
two 0.95 cm (3/8 inch) 310 stainless steel pipes 1.524 m (5 feet) long
with a 20.3 cm (8 inch) long, 180° turnaround so that the air has
17
-------
approximately 3.55 m (11 feet) of travel while being heated in excess of
925°C by a 50 kW induction coil and power supply operating at 30 kHz.
In addition, a revolving disk, high-pressure dust feeder is used to
introduce regulated amounts of flyash to the filter test chamber. It is
enclosed in a separate pressurized vessel maintained at a pressure higher
than the test chamber pressure.
Dust is metered from a dust hopper by filling a small groove in the
rotating disk. The pressurized air enters the vessel below the disk and
exits, carrying the dust out through a small tube directly above the dust
filled grove. The feedrate is adjustable by varying either the disk
rotation speed or the groove cross sectional area.
Primary Air Supply
The basic gas supply is by means of a 16.9 m /min (600 scfm) air
compressor and receiver tank located adjacent to the existing facility
block wall. The gas supply schematic is shown in Figure 4. The gas
receiver tank provides a metered flow of compressed air to the main
forward and reverse flow valves. The flow arrangement is unique in that
it uses a pipe cluster inside the induction coil to heat both its forward
and reverse flow gas supply. Flow from the test chambers is routed
through a finned heat exchanger and then into an absolute filter for
efficiency data. Flowrates are determined by passing the gas through a
choked orifice downstream of the absolute filter.
Secondary Air
Secondary air for the pulse-jet cleaning mode and operation of the
dust feeder is from a smaller 2.1 nr/min (75 cfm) compressor and
receiver tank located behind the facility pit.
18
-------
Cooling Water
Water from a cooling tower is piped to the inlet gas manifolds and
through the induction coil and power supply for cooling purposes. It uses
a recycling system for water conservation.
Chamber Heaters
In addition to the induction heater previously mentioned, the test
chamber is heated by a chrome-aluminum-iron alloy element helically wound
and embedded in a thermally conductive cement and powered at
208 34> volts. These heaters are needed to maintain the test conditions
(815°C) because of heat loss between the induction heater and the
chamber. Control is maintained by a set point controller operating a
208 3* SCR.
Control System/Console
Almost all functions are controlled from the main control console.
The control circuitry is designed around a programmable logic controller
made by Texas Instruments. This provides the safety interlocks and
sequencing functions required by the facility, and eliminates using a
myriad of relays.
Time Sequence
The ESE sequencer is the heart of the system in that it sends the
commands to the logic system to command forward flow, reverse flow, pulse
flow, bleed-down, dust feed, counting, and recycle.
The sequencer has a small semiconductor memory in which it can
store up to 32 "events", then associates a time and identifies which relay
is triggered at that time. The relays are connected to the logic system,
the cycle counters and the external time reset. A typical sequence is
described below.
19
-------
Facility in normal forward flow mode:
Event 00 would trigger relay 05 to stop the dust feeder table to
purge the line and forward flow. Event 01 would trigger relay 01
to shut the inlet valve and bleed-down chamber pressure. Event 02
would trigger relay 02 to start reverse flow. Event 03 would
trigger relay 03 to open pulse jet valve for 150 milliseconds.
Event 04 would trigger relay 04 to advance the cycle counter.
Event 05 would trigger relay 00 to restart forward flow and dust
the feeder. Event 06 would trigger relay 15_ to recycle sequencer.
TABLE 1. ESE SOURCE EVENTS
Source
00
01
02
03
04
05
15
Event
Restart forward flow and dust feeder
Start bleed-down
Reverse flow
Pulse flow
Cycle counter
Stop dust feeder and forward flow
Recycle sequencer
TABLE 2. EXAMPLE CLEANING SEQUENCE
Event
00
01
02
03
04
05
06
Source
05
01
02
03
04
00
15
Time
00:00:01
00:00:04
00:00:08
00:00:09
00:00:10
00:00:11
00:10:11
Description
Stop dust feeder and forward flow
Bleed-down
Reverse flow
Pulse
Count
Restart dust feeder and forward flow
Recycle
20
-------
Data Recording
Two two-pen strip charts are used to display and store test system
pressure and filter pressure drop.
One 24-point temperature strip chart is used to display and store
various temperatures in the facility.
Theory
Barrier filtration with available ceramic fibers is a good
technique for particle control at high temperature and pressure. To
illustrate this, a short review and discussion of barrier filtration
theory is helpful.
Figure 6 was taken from a report titled "Effects of Temperature and
Pressure on Particle Collection Mechanisms: Theoretical Review" by
Seymour Calvert and Richard Parker (EPA-600/7-77-002), January 1977. This
figure shows a calculated fractional efficiency curve for a fiber bed.
Minimum efficiency is indicated for a particle size of about 0.5 ym. The
dip in the curve occurs because of the interaction of the three collection
mechanisms applying to barrier filtration, direct interception, diffusion,
and inertial impaction.
For particle size less than about 0.5 ym, collection by diffusion
improves the efficiency of the filter bed. For particle size larger than
about 0.5 ym, collection by inertial impaction increases the collection
efficiency of the filter bed. This curve applies only to initial
performance of a clean fiber bed, i.e., it does not include the increased
collection efficiency that results from the filtration by the accumulating
dust cake. Note on Figure 6, the No. 3 curve, indicating that the
inertial impaction parameters for high temperature and pressure
conditions, should show a small decrease 1n performance. To understand
21
-------
ro
100
90
80
* 70
I
O 60
UJ
O
t 50
UJ
I 40
§
O 30
O
20
10
0
0.1
CONSTANT FACE VELOCITY
E
OB
NO. CONDITIONS
1 20°C, 1 atm
2 1.100°C. 1 atm
3 1.100°C, 15 atm
0.5 1.0
PARTICLE DIAMETER — urn
5.0
Figure 6. The effects of high temperature and pressure on
the collection efficiency of a fiber bed.
-------
the magnitude of this effect we can compare the performance of standard
filter media when tested with dioctylphthalate smoke (OOP) to its
performance when tested after a stabilized dust cake has been developed.
A OOP smoke penetration test is a standard test given to measure the
efficiency of high performance filters such as those used to filter "clean
room" air or to collect biological contaminants. This test measures how
efficiently a filter removes 0.3 ym diameter OOP smoke particles. Woven
or felt filter media of the type commonly used for industrial filters will
collect only 10 or 20 percent of 0.3 ym OOP smoke. After developing a
dust cake, these same filter media will collect submicrometer particulate
at an efficiency of greater than 90 percent. Thus, compared to the
changes in performance in a filter media during the conditioning process,
the changes predicted as a result of high temperature operation are small.
Ceramic fibers offer unique advantages for filtration, since many
of these fibers have finer diameters than conventional filter fibers.
Conventional fibers are ususally 10 or 20 ym in diameter, while ceramic
fibers are available with average diameters of only 3.0 ym.
Collection efficiency can be improved simply by making a filter bed
thicker, thus increasing the basis weight of the filter (its weight per
unit area). However, to achieve high collection efficiency this way can
lead to high operating pressure drops. Collection efficiency can also be
increased by reducing the fiber diameter, which can result in decreased
basis weight and filter bed thickness. The importance of fiber diameter
is illustrated in the following equations describing the three primary
particle collection mechanisms applicable to barrier filtration.
d
Interception parameters Kj = •£*•
23
-------
2
Impaction parameters K = PR—9-
Diffusion parameter Kd - ^'^ .
g P g *
Although these equations describe the collection mechanisms, they are not
collection efficiency equations. However, when expressed as above, an
increase in any of the mechanism parameters (K,, K , K .) will result
in an increase in efficiency.
The interception parameter is a function of fiber diameter.
Changing from a 20 ym fiber to a 3.0 ym fiber will increase the
interception parameter by a factor of 6.67 times.
The impaction parameter is a function of temperature and pressure
essentially through changes in the gas viscosity (yg). For air,
increasing temperature from 20 to 815°C increases viscosity by about 2.5
times, reducing the impaction parameter by a factor of 1/2.5 (or 0.4).
However, the change in fiber diameter from 20 ym to 3.0 ym increases the
impaction parameter by 6.67 times. The net effect of the two changes is
to increase the impaction parameter by 2.7 times.
The diffusion parameter is a function of temperature and pressure
through changes in the ratio of (C'T/yg). When operating at 815°C and
10 atm pressure, this ratio tends to either remain unchanged or increases
slightly. The diffusion parameter is also a function of fiber diameter,
and a change in fiber diameter from 20 ym to 3.0 ym will increase the
diffusion parameter by 6.67 times.
From the above discussion it is evident that if we make a filter
using 3.0 ym diameter ceramic fibers (which are commercially available),
24
-------
we can reasonably expect that even at high temperature and pressure this
filter will have high collection efficiency without excessive filter bed
thicknesses or basis weights. Using the method developed by Torgeson,
makes it possible to calculate collection efficiency for given particle
size and fiber bed parameters. This calculation was performed for a
0.5 ym diameter particle with a density of 1.5 g/cm3 (as measured at the
Exxon tniniplant), for gas temperature of 815°C, and for pressure at
10 atm. A fiber bed composed of alumina fibers with 3.0 \im diameter and
fiber density of 2.8 g/cm3 was assumed. Results of this analysis for
two filtration velocities and two solidities (a - the volume fraction of
the fiber bed which is solid) are plotted in Figure 7. This analysis
indicates that a 3.0-ym diameter ceramic fiber filter bed, with a basis
o
weight of 500 to 600 g/m, will collect submicrometer particulate with
an initial (clean) efficiency of about 90 percent at high temperature and
pressure. As previously stated a typical industrial filter media (20-ym
fibers) collects such particles at only 20 percent efficiency for the same
basis weight. In a different view you can see that to achieve collection
efficiency comparable to commerical industrial media will require a
ceramic fiber filter media with weights only one-tenth that of the
commercial media. Another interesting feature of the analysis is that
efficiency decreases for increasing velocity. However, by adding fibers,
the given efficiency can be maintained as velocity is increased. The
quantity of required additional fiber is relatively small, especially if
only 20 percent initial efficiency is adequate for a 0.5-um particle.
Most commerically available ceramic fiber structures are produced
for insulation applications. Consequently, these materials are generally
characterized by an open fibrous structure, i.e., they have low solidity,
25
-------
90
80
O
O
UJ 70
O
S 60
50
40
20
3.0M m OIA FIBERS
2.8 g/cm» FIBER DENSITY
(5 tt/min) \
2.54 cm/sec J
0.5 u m DIA PARTICLE
1.5 g/cm3
815" C
10 ATM
(25 tt/min)
12.7 cm/sec
(16 ox/yd1)
540 fl/m*
J_
100 200 300 400 500 600 700 800 900 1000 1100 1200
BASIS WEIGHT, g/m'
Figure 7. Calculated performance of 3.0 urn alumina fiber bed.
26
-------
with perhaps only 2 percent of the volume occupied by fibers (a = 0.02).
A solidity of a = 0.10 is more typical of a structure designed for
filtration. Figure 8 shows the effect on fiber bed thickness for changes
in solidity and air-to-cloth ratio. For solidity typical of insulation
materials (a = 0.02), a fiber bed about 1 cm thick should achieve high
initial collection efficiency of submicrometer particulate, while a more
compressed media with a = 0.10 would achieve this efficiency with a bed
thickness of only 2 mm. If efficiency typical of industrial filters is
adequate, very thin layers of the 3.0-ym diameter fibers will suffice.
Also note that filter media thickness is not a strong function of
air-to-cloth ratio, indicating the possibility of high filtration
velocity. Higher filtration velocity results in increased pressure drop,
which may be acceptable in a PFBC application.
Room Ambient Filter Media Tests
A large number of ceramic fiber filter media candidates were
subjected-to a series of filtration tests at room ambient conditions.
These tests shown below, include some examples of conventional filter
media for comparison.
• Dioctylphtalate (OOP) smoke penetration as a function of
airflow velocity
• Determination of maximum pore size in ym
• Measurement of permeability
• Flat-sheet dust loading tests using AC fine test dust; overall
collection efficiency and dust loading required to develop
3.7 kPa (15 in. H20) pressure drop are determined from this
test, operated at 10 cm/sec (20 ft/min) air-to-cloth ratio
Data collected from these tests are summarized in Table 3.
27
-------
1.4
1.3
1.2
1.1
1.0
O 0.9
I
S 08
Z
X
O 0.7
I
g 0.6
CO
£ 0.5
99
Li-
0.4
0.3
0.2
0.1
0
0.5 Mm PARTICLE
815°C
10 ATM
90%EFF ° = 0.10
20% EFF a = 0.02
•—••
20% EFF a = 0.10
4 6 8 10
AIR-TO-CLOTH RATIO — CM/SEC
Figure 8. Fiber bed thickness.
12
14
28
-------
TABLE 3. SUMMARY ROOM AMBIENT TEST DAIA
PO
vo
iU) Woven
f\ Paper
F) Felt
1. Cartonmdtm Flberfrax cloth
(W) with nlchrome wire Insert
2. Zlrcar Zlrconla felt ZFY-100
(P)
3. ICI Saffll alumina paper
(P)
4. ICI Saffll mat
(P)
5. Bibcock & Wllcox Kaowool
6. Carborundum Flberfrax
(F) durablanket
7. Johns Hanvllle Flberchrone
(F)
B. Stevens Astroquartz
(U) style 581
9. Hltco Refrasll C-100-96
(N) heat cleaned
10- Hltco Refrasll C- 100-48
(H) not heat cleaned
11. Stevens Astroquartz cloth
(U) style 570
12. an AB-312 basket weave
(W) cloth
Basis
Height
q/m2
1366
615
165
355
746
1363
1297
283
1284
667
677
311
Efficiency
on ACF 8
10 cm/sec
(20 ft/ml n)
96.55
95.64
99.805
98.74
98.464
99.507
99.654
60.77
81.97
83.37
56.83
51.38
Dust Loading Permeability
a/of to 3.735 kPa cm'/sec/cm2 for 0.1245 kPa
(g/ft2 to 15" H20) (ft'/mln/ft2 for 0.5" H?0 AP)
240.0
(22.2912)
Media Fractured
159.4
Media Fractured
118.2
(10.980)
147.0
(13.6523)
253.9
(23.59)
Test Stopped -
Low Eff.
5.9
(0.5482)
11.6
(1.074)
Test Stopped -
Low Eff.
Test Stopped -
Low Eff.
8.687
(17-D
10.861
(21.38)
9.307
12.395
(24.4)
8.067
(15.88)
5.583
(10.99)
11.897
(23.42)
37.236
(73.3)
1.240
3.099
(6.1)
22.758
(44.8)
13.553
(26.68)
Maximum
pore size
micrometers
248.6
59
43
61.1
66.9
68.2
112.3
248.6
112.3
133.8
267.7
870
Percent Efficiency on
0.3 urn D.O.P. at cm/sec
2.68
45
75
82
79
96.5
97.1
78
0
0
0
0
0
5.35
47
78
65
80
93.5
94.6
73
9
19
11
13
5
14.22
50
72
62
73
86
90.5
74
12
34
10
32
8
T-889
-------
TABLE 3. Continued
H) Woven
P) Paper
F) Felt
13. 3M AB-312 twill weave
(H) cloth
14. Hltco Refrasll cloth
(U) UC-100-48
15. Zlrcar Zlrconla cloth
(H) ZFY-30A
16. FW -Stevens Astroquartz
(U) cloth crowfoot satin
17. 3H AB-312 twill weave
(U) cloth coated with 3H
coating
IB. 3M AB-312 basket weave
(H) cloth coated with 3M
coating
19. 3M AB-312 twill weave
(H) cloth Henarde coating
20. Hltco Refrasll cloth
(U) UC-100-96 not heat
cleaned
21. Carborundun Fiberfrax
(U) no Insert wire L-126TT
22. Hltco Refrasll batt 8100-1
(F)
23. Hltco Refrasll standard
(U) not heat cleaned uc- 100-28
24. Hltco Irish Refrasll
(H) chronlzed C- 1554 -48
Basis
Weight
P/nf
231
643
608
352
227
281
254
1249
1544
807
335
683
Efficiency
on ACF 9
10 cm/sec
(20 ft/m1n)
48.55
69.25
99.014
76.19
47.64
45.65
55.078
68.46
99.21
99.229
84.41
81.476
Dust Loading Permeability Maximum
a/in* to 3.735 kPa cm' /sec/ cm" for 0.1245 kPa pore size
fg/ff to 15" H,0)
-------
TABLE 3 Concluded
!H) Woven
P) Paper
F) Felt
25. Carborundun Flberfrax
(P) paper (with binder) 9700
26. ICI Saffll Z1rcon1a paper
(P) (with binder)
27. Carborundin Flberfrax
(P) paper (no binder) 970-AH
28. 3M AB-312 double thick
(W) plain weave
29. FMI crowfoot satin cloth
(W) Astroquartz
30. 3M AB-312 12 harness satin
(W) weave
31 . 630 Tuflex fiberglass!
(")
32. 15-011-020 woven ftlanentf
(W) polyester
33. 25-200-070 polyester feltl
34. Hltco Refrasll cloth (std)
(W) not heat cleaned, red.
thickness
Efficiency
Base on ACF 9
weight 10 en/sec
g/m2 (20 ft/n1n)
604
212
152
1035
905
675
564
175
524
637
99.99
93.20
(Probable hole)
99.91
43.86
40.08
53.73
93.982
96.078
99.193
48.40
Dust Loading Permeability
a/of to 3.735 kPa cm'/sec/cm2 for 0.1245 k
(g/ftl to 15" H20) (ft'/mln/ft2 for 0.5" HjO
73.7
(6.8442)
82.2
(7.6374)
84.4
(7.8369)
Test Stopped -
Low Eff.
Test Stopped -
Low Eff.
Test Stopped -
Low Eff.
47.6
(4.4187)
28.0
(2.60163)
135.3
(12.5688)
34.5
(3.2063)
26.899
(52.95)
B.692
(17.11)
12.416
(24.44)
84.836
(167)
62.078
(122.20)
75.529
(148.68)
16.038
(31.57)
6.828
(13.44)
11.897
(23.42)
6.934
(13.65)
Maximum Percent Efficiency on
Pa pore size 0.3 un D.O.P. at en/sec
&P) micrometers 2.68 5.35
47.7 99.5 99.0
37.4 83 78
43.5 B8
Too large to 0 10
measure with
our equipment
497 0 10
696 0 8
174 10 9
74 60
128.9 34 24
97.6
74
73
41
32
24
19
4
29
iThese Materials are conventional (not ceramic) media.
T-889(b)
-------
Penetration tests using OOP smoke measure the ability of the clean
fiber bed to stop fine particles. The OOP smoke generator 1s adjusted to
provide a nominal particle size of 0.3-ym diameter, a "most penetrating"
particle size because of the minimal effect of diffusion and inertlal
impactlon at this size. The OOP test results should correlate well to the
predicted results by analysis since particle collection 1s provided only
by the fibers and not by the dust cake. Figure 9 provides a plot of the
OOP efficiency as a function of airflow velocity for all the media
tested. Ceramic media data are plotted In solid lines, and conventional
media are plotted in dotted lines. Numbers on the curves correspond to
those in Table 3. Some observations concerning this data are:
• Several of the ceramic materials, especially the ceramic papers
and felts, are capable of higher efficiency collection of fine
particles than media normally used successfully in commerical
filter units
• Many of the woven ceramic materials had zero OOP efficiency at
low velocity, and higher OOP efficiency at higher velocity
This point is contrary to what theory suggests and to the
behavior normally seen in tests of conventional filter
materials. A probable explanation for this is that it is
caused by many large pores present in the media. Examination
of the pore size data 1n Table 3 shows that the woven ceramic
materials group 1s characterized by larger pore sizes than that
of conventional filter materials. Thus, at low airflow
velocity, most of the flow passes through the large pores and
little filtration takes place. As velocity is Increased, flow
through the large pores becomes relatively restricted resulting
32
-------
UJ
X.
§
co
CL
O
Q
E
3
*?
O
I
100
90
70
60
50-
5 40
NOTE: NUMBERS REFER
TO THOSE IN TABLE 3.
28
5 10
AIRFLOW VELOCITY, cm/sec
Figure 9. OOP efficiency airflow velocity.
33
-------
from increased AP caused by high velocity through the pores,
and some of the flow is forced to pass through smaller pores
where more filtration can occur.
• The OOP data also support the theoretical analysis in the
previous section. Efficiency as a function of basis weight for
selected ceramic materials is plotted in Figure 10. The
selected materials are ceramic papers and felts providing a
fiber bed similar to that on which the analysis summarized in
Figure 5 was based. Figure 10 illustrates that the nominally
3 ym fibers provide higher collection efficiency on a
weight-per-unit-area basis than conventional media produced
with larger diameter fibers.
Maximum pore size data show that many of the woven ceramic
materials had pores larger than those characteristic of conventional
filter materials, and many of the felt and paper materials had pore sizes
similar to those of conventional filter materials.
Permeability is measured as the flow per unit area at a constant
pressure drop. Thus, a material with low permeability offers a high
restriction to gas flow and one with high permeability allows more gas to
penetrate for a given pressure drop. Table 3 shows that some ceramic
materials are available which have low permeability while others have high
permeability. Some of the woven materials have low permeability and large
pore size, and again, others have high permeability and large pore size.
Most of the paper and felt materials have permeability similar to that of
commonly used filter materials.
Flat-sheet dust loading tests were performed as follows: a 7.62 cm
(3 inch) diameter disc of media was suspended across an air stream which
34
-------
NOTE: NUMBERS REFER TO
THOSE IN TABLE 3.
200 400 600 800
BASIS WEIGHT, g/cm'
1000
1200
1400
Figure 10. OOP efficiency test data.
35
-------
was maintained at 10.16 on/sec (20 ft/min) velocity through the filter
media. In this test, the media supported itself against the pressure drop
(no screen was used). Standard AC fine test dust (0-80 urn silica) was fed
to the media at a nominal rate of 0.883 g/m3 (0.025 g/ft3) until a
pressure drop of 3.735 kPa (15 in. H20) was reached. Pressure drop as a
function of time was monitored during the test. This data is presented in
Figures 11, 12, and 13. for selected materials. From this data, dust
P
loading (g/m ) necessary for a given pressure drop of 3.735 kPa
(15 in. HpO) is determined. Examination of the data in Table 3
indicates that some of the woven materials reached high pressure drops
while collecting only a small weight per unit area of dust, as well as
that of the commerical woven materials (items 31 and 32). Other woven
ceramics were penetrated so severely that they would not develop a
pressure drop of 3.735 kPa (15 in. H.O).
Two of the non-woven samples (which were unsupported) fractured as
a result of the pressure drop across them. Several of the ceramic paper
and felt materials exhibited dust loading similar to that which is
expected from conventional filter papers and felts.
The flat-sheet loading tests also provided overall collection
efficiency (mass basis) data for the tested materials. Dust penetrating
the media was collected in an absolute filter downstream of the test
media. Table 3 reveals that most of the woven ceramic materials did not
achieve high collection efficiency in this test. Conversely, woven
commerical materials were only moderately efficient. However, several of
the ceramic paper and felt materials provided collection efficiency of 99
percent or better. The two materials which fractured would have provided
36
-------
A.C. FINE TEST DUST
0.883 g/m3
A/C 10.16 cm/sec
No. 33 = CONVENTIONAL FELT FILTER
o
a
a.
O
oc
o
111
cc
V)
V)
ui
CC
a.
NOTE: NUMBERS REFER
TO THOSE IN TABLE 3-2.
Figure 11. Dust loading of ceramic felts.
37
-------
26
A.C. FINE TEST DUST
0.883 g/m3
A/C 10.16 cm/sec
NOTE: NUMBERS REFER
TO THOSE IN TABLE 3-2.
0
Q.
Q."
O
DC
O
UJ
DC
1
UJ
cc
Q.
Figure 12. Oust loading of ceramic paper.
38
-------
\o
A.C. FINE TEST DUST
0.883 g/m3
A/C 10.16 cm/sec
#31. 32 = CONVENTIONAL WOVEN FILTERS
NOTE: NUMBERS REFER
TO THOSE IN TABLE 3-2.
40 50
TIME. min.
12
Figure 13. Dust loading of woven ceramic media.
-------
higher efficiency performance if they had not fractured, the test was
stopped upon detection of the fracture.
Mechanical Screening Test Data
As previously stated, the available ceramic fiber structures can be
generally classed into three groups: woven, paper, and felt. During the
preliminary testing, all three types of media were subjected to survey
tests. The objective of the mechanical screening tests was to subject the
media to typical filter-cleaning loads in the high temperature and
pressure environment and to determine which media can survive these
loads. All the mechanical screening tests were operated at an
air-to-cloth ratio of 2.54 cm/sec (5 ft/min).
Woven Media Tests
Three woven media types were tested, and their results are
summarized in Table 4. These tests were performed early in the program
before we had fully proved the 815°C capability of the heater. Since
these tests are fairly qualitative, the following photographs and
discussion are offered to augment the data in Table 4.
The AB-312 media test was terminated because of excessive dust
penetration through the relatively large pores around the yarn filament of
the cloth. This media clearly showed a clean separation of the dust cake
from the media surface (Figure 14 illustrates this effect, seen on other
tests as well).
An Irish refrasil filter was tested with pulse cleaning after
surviving 195 reverse-flow cycles. This media developed holes during the
pulse testing. Examination after 3,100 pulse cycles revealed the
condition illustrated in Figures 15 and 16. The holes were generally
located near the center of the openings formed by the support cage,
40
-------
TABLE 4. WOVEN MATERIALS TEST RESULTS
3M Company
AB-312 Twill Weave
Irish Refrasil
Silica/Chrome Satin
Irish Refrasil
Silica/Chrome Satin
Temperature
538°C
538°C
538°C
538°C
815°C
Cleaning
Reverse Flow
Reverse Flow
Pulse 1400 kPa
(200 psi)
100 mS
Reverse Flow
Pulse 1100 kPa
(160 psi)
1100 mS
Cycles
153
195
3100
100+
1
Time
8.5 hr
13 hr
7 hr
15 hr
—
Failed
No
No
Yes, holes
No
Seam
-------
to
§ m &.
^*v
Figure 14. Post-test dust cake -- Test B103 3M-AB312 twill weave
.
-------
Figure 15. Irish Refrasil pulse test failure.
45
-------
Figure 16. Irish Refrasil pulse test failure.
-------
appearing to be a result of bending and shock at the maximum points of
media acceleration during cleaning. This effect could possibly be
controlled by making the support cage openings smaller. Table 4 shows an
1100-m/sec pulse duration, which occurred because of an error in setting
the pulse duration, and was not discovered until after the test. This
could also have caused the media failure. In addition, it should be noted
that pulse cleaning is probably not required for woven media since the
less rigorous reverse-flow cleaning can remove the dust cake.
Figure 17 shows a seam failure occurring on the first pulse, and
illustrates the force of the pulse. This particular test element had been
installed several times and, therefore, had been handled more than usual.
(Failure can be caused by excess handling and sharply bending the fiber).
Paper Media Tests
Only one test of paper media was performed. Results of this test
are summarized in Table 5. There were no evident failures after a short
reverse-flow test; however, the media did not survive pulse testing. A
bias-cut wire screen supported this media on both surfaces. The bias-cut
allowed the tubular shape of the element to stretch, contributing to its
failure. A straight-cut screen would provide better support for paper
media. Figures 18 and 19 show the bias-cut screen with several tears in
the media.
Felt Media Tests
Test results for felt media are summarized in Table 6. In the
first test, no damage was evident as a result of reverse-flow testing.
After about 5,000 cycles of pulse testing, the pressure-drop monitor
indicated a problem. Visual examination revealed that the lower band
clamp had slid upward, allowing a leak under the filter element. After
49
-------
Figure 17. Refrasil media seam failure.
51
-------
TABLE 5. PAPER MATERIALS TEST RESULTS
tn
U)
ICI
Saffil Alumia
Temperature
815°C
815°C
Cleaning
Reverse Flow
Pulse 1100 kPa
(160 psi)
100 mS
Cycles
21
496
Time
1 hr
9 hr
Failed
No
Yes
-------
Figure 18.
Saffil alumina paper after pulse testing.
55
-------
Figure 19. Saffil alumina paper after pulse testing
57
-------
TABLE 6. FELT MATERIALS TEST RESULTS
in
ICI
Saffil Alumina
ICI
Saffil Alumina
ICI
Saffil Alumina
Fiberfrax Cloth
with Ni chrome
wire scrim
Fiberfrax Blanket
Temperature
815°C
815°C
815°C
815°C
815°C
825°C
Cleaning
Reverse Flow
Pulse 1100 kPa
(160 psi)
100 mS
Pulse 1100 kPa
(160 psi)
100 mS
Pulse 1100 kPa
(160 psi)
100 mS
Pulse 1100 kPa
(160 psi)
100 mS
Pulse 1100 kPa
(160 psi)
100 mS
Cycles
75
9405
7450
50,000
50,000
50,000
Time
3 hr
49 hr
38 hr
140 hr
140 hr
140 hr
Failed
No
Yes, seam
No
No
No
No
-------
the lower end of the media was reclaroped, the tests continued. The
lateral seam for this media was accomplished by rolling a flat sheet of
media into a cylinder and overlapping the ends. This tube shape was
supported inside and out by a wire screen. After a total of 9,405 pulse
cycles, visual examination showed that the seam had separated by sliding
apart, which was probably started when the band clamp moved earlier in the
test. The flexible bias-cut support screen also contributed to this
problem. Figures 20 and 21 show the element and a close-up Of the seam
separation. The whiteness shown in Figure 21 indicates a good separation
of the dust cake from the media surface. The area was cleaned by tapping
the screen several times with a pencil.
For the second test, a screen with a fine diameter wire with more
strands per unit length was used to support the media. While this screen
was also cut on the bias, it provided a stiffer support for the media.
Extra care was taken to ensure that the clamps were tight. This media
operated for 7,450 pulse cycles with no apparent damage. The test was
stopped because of a failure of the test rig heater element. Figure 22 is
a photo of the element after the test.
Three additional ceramic filter media configurations survived a
test during which the filter elements were subjected to 50,000 cleaning
pulses. The objective of these tests was to simulate approximately 1 year
of mechanical cleaning cycles on the media at high temperature and
pressure. Test conditions were:
Temperature: 815°F
Pressure: 930 kPa
Air-to-cloth ratio: 5 to 1 (2.54 cm/sec)
Cleaning pulse pressure: 1100 kPa
60
-------
Figure 20. Saffil alumina felt bag post-test.
61
-------
Figure 21. Separated seam.
63
-------
Figure 22. Saffil alumina blanket after pulse tests,
65
-------
Cleaning pulse interval: -10 seconds
Cleaning pulse duration: 0.1 sec
Dust: recirculated flyash
The three filter media configurations tested were:
t Saffil alumina mat within an inside and an outside layer of 304
stainless steel knit wire screen. Figure 23 shows how easily
the residual dust cake was removed from this media after the
test.
• Woven Fiberfrax cloth with nichrome wire scrim insert.
Figure 24 shows the dust cake following the 50,000 pulse test.
This woven material performs more like a felt because the yarns
are "fuzzy".
• Fiberfrax blanket within an inside and outside cylinder of 304
stainless steel wire mesh screen similar in construction to
common window screen. The ceramic fiber blanket was held in
position between the screens with 302 ss wire sewn between the
screens. This resulted in about 100 penetrations of the
ceramic fiber bed. Figure 25 shows the dust cake following the
test.
Pressure drop during the tests was controlled by the rapid cleaning pulses
and in general remained less than about 5 kPa (20 in H^O).
Dust penetrating the ceramic test media was collected on a high
efficiency filter located downstream (after cooling) of the test chamber.
This data is not reported for the Saffil alumina or for the woven
Fiberfrax cloth because of the leak discovered in a gasket in the test
rig. This problem was corrected before the Fiberfrax blanket test was
performed. Average outlet loading during this test was 0.0055 g/am
67
-------
Figure 23. Saffil alumina -- post-test dust cake
(clean strip using vacuum cleaner).
69
-------
Figure 24. Woven Fiberfrax -- post-test dust cake
71
-------
-T r-r- r-»»'•**"?* ^^£&^i'j?fZ^5Z.
- •r^-rrrr^.j^T^r'ry'.7 /^ly^J^
caut*. :•• V*, >
vi;
.••.,• A -' <•'
•v-r~ ';iv££
'' M^'.>'7 n"s T r/'ir
'.-^•'•7 'rlrtT-;r
^.jfi/;--^ T--r-r,-r-^-^-|
3|
tiiT±r:
«-^»- ^>.;TT-
* 1 —- LT _.
Figure 25. Fiberfrax blanket -- post-test dust cake.
73
-------
(0.0024 gr/flft3). Figure 25 shows that the dust was concentrated near
the places where wire penetrated the filter element. This concentration
of dust near the wire penetration points could also be seen on the inside
of the element. Thus, most of the dust which penetrated apparently did so
through the holes made by the wires. It is reasonable to expect that a
filter element without holes will experience less penetration;
additionally less frequent cleaning pulse interval will reduce penetration.
Filter Performance Tests
Filter performance tests were intended to simulate actual filter
operation at high temperature and pressure for a period of 200 hours. The
filter media configuration which we have selected as most promising
consists of an approximately 1 cm thick layer of saffil alumina blanket
insulation material. This ceramic material is contained between two
layers of knit 304 stainless steel screen. The first 200 hour test was
performed at a nominal air-to-cloth ratio of 5 to 1 (2.54 cm/sec). Pulse
duration was 150 ms. Pulse internal was one cleaning cycle every 10
minutes. Pulse pressure was 1100 kPa. Cleaning was performed "off-line"
with 4 second bleed-down followed by reverse flow for 2 seconds, the pulse
super-imposed followed by a 2 second settling period prior to continuing
the filtration cycle.
Exxon miniplant flyash was used as the test dust. A fine dust
D50 = 4 yM was used for the f1rst 75 hours< When this dust was no
longer available, a course sample DCQ = 19 yM was used for the remainder
of the test.
75
-------
Cumulative dust fed, total dust collected downstream and overall
collection efficiency by mass are plotted as function of time on
Figure 26. Note that inlet concentration was high (an overall average of
14.4 g/Nm3). Shortly after the fine dust was substituted with the
coarse, overall efficiency was reduced and the rate of penetration in
weight per unit time was increased. Later, at about 120 hours and again
at 150 hours, the rate of dust feeding was reduced. The rate of
penetration seemed to follow this. These occurrences are consistent with
leakage through a defect mechanism in the media. Visual examination of
the inside surface of media after the test revealed it to be substantially
clean with some localized staining. Overall collection efficiency
remained high throughout the test never falling below 99.965 percent.
Outlet concentration as a function of time is shown on Figure 27. These
results are consistent on a mass basis with turbine requirements as
reported by Sverdrup in EPA 600/9-78-004. Dust collected downstream was
in the form of large flake agglomerates, some of which resembled rust
scale. These are probably artifacts of the test and test set-up --
cooling and condensation. However, they make if difficult or impossible
to obtain fractional efficiency for this run. The outlet concentrations
reported above are based on a flow of 0.566 Nnr per minute during the
time that dust was being fed (200 hours). They do not include the
additional flow which occurred during warm-up and cool-down when the dust
feeder was off. Because of various difficulties occurring with the test
rig, the run was interrupted many times. Pressure drop was maintained at
less than about 0.75 kPa (3 in H_0) over the entire test.
76
-------
** 99.9 .
u
I 99.98J
o
o
o
99.97
•
99.96
- 99.95^
u
0)
99.94.
30
20
10
Dust
Downstream
5
e
100
TJ
(U
80
60 *
o
40
20
0
200
Figure 26. Filter characteristics 2.5 cm/sec.
-------
800°C
10 Atm
I
CNJ
4J
2
0.010
3
<
00
.£ 0.005
£
I
50
I
100
Time-Hours
150
200
Figure 27. Outlet concentration 2.5 cm/sec.
-------
For the second 200-hour test we attempted to achieve the highest
a1r-to-cloth ratio possible with the present test rig configuration. This
test was performed at an air-to-cloth ratio of 18 or 9 cm/sec face
velocity. Filter face velocity was only 8.4 cm/sec for the first
11 hours. The increase to 9 cm/sec was accomplished by operating two flow
limiting orifices in parallel. Because of compressor limitations, it was
only possible to maintain a system pressure of 500 kPa. Earlier tests
were run at system pressures of 930 kPa. Cleaning pulse pressure was set
at 860 kPa to compensate for the reduced system pressure. Cleaning cycle,
pulse duration and pulse interval were the same as in the previous test at
2.5 cm/sec with both tests using offline cleaning.
Flyash from the Exxon Miniplant was used as the test dust. This
dust had a DCQ = 19 ym. Cumulative dust fed, total dust collected
downstream and over-all collection efficiency by mass are plotted as a
function of time in Figure 28. Outlet concentration as a function of time
is shown in Figure 29. Cleaned-down pressure drop was maintained at less
than 1.25 kPa (5 in H20) throughout the test.
This data indicates that performance at high airflow velocity was
even better than that achieved in the previous test. The test filter
element was able to maintain high efficiency performance for the entire
period of the test. Thus, evidence is now available that indicates
turbines can be protected from particulates using ceramic fiber filters
operated at filter face velocity at least as high as 9 cm/sec.
79
-------
00
o
4.0 U
100.00 J
0)
•f-
u
E 99.99
01
£ 99.98
8
(O
I 99.97 -J
99.96
3.0 h
4.
o>
Sz.o
•o
-»J
M
(§
1.0
200
250
Figure 28. Filter characteristics 9cm/sec.
-------
00
0.0015
Ol
u
o
u
£
«
0.0010
0.0005
750°C
5 atm
50
100 150
Time - hours
200
Figure 29. Outlet concentration 9 cm/sec.
-------
APPENDIX A
Following 1s a list of media samples. The list contains
descriptive information taken from advertising literature for each
material. In examining these data, the reader should remember that in
general these materials were not produced for filtration applications.
Based on this information none of the materials can be recommended or
rejected. Additional information will be required before the judgment can
be made. The available Information does indicate that at least some of
the materials will probably offer good filtration performance.
83
-------
LIST OF SYMBOLS
d = Particle diameter
d* = Fiber diameter
C1 = Cunningham correction factor
p. = Particle density
U = Gas velocity at media face (flow/area)
y = gas viscosity
k - Boltzmann constant
T = Absolute temperature
84
-------
OOP Test Results, High Temperature Filter Media
Media Samples — Description
Aatroqu.'irtH
J. P. Stevens)
Cloth (with binder)
Style 570:
Style 581;
Custom Weave;
FMI Astroquartz
5 Harness Satin 300-2/8, 27 mils thick,
19.5 oz/sq. yd., 38 x 24 thread count/
inch, 325 (warp) and 300 (fill) pounds/
inch width breaking strength. Pure
Silica (99.9%), 2000°F service temperature.
Filament diameter ~ 6 microns)
8 Harness Satin 300-2/2, 11 mils thick,
8.4 oz/sq. yd., 57 x 54 thread count/
inch, 175 (warp) and 170 (fill) pounds/
inch width breaking strength. Pure
Silica (99.92). Good to 2000°F continuous
service. ~ 6 micron filaments
Crow foot Satin 300-2/2 (warp), 4/2 (fill)
54 (warp) x 36 (fill) threadcount per
inch, - 12 mils thick, " 11 oz/sq. yd.
Pure Silica (99.9%), good to 2000°F
service.
Refrasil
(Hiteco)
(Fibers 8 to 10 microns diameter)
Irish (Chromized)
C 1554-48:
Refrasil
8 Harness Satin, 26 mils thick, 19.2 oz/
sq. yd., 53 x 40 thread count per inch,
96 (warp) and 62 (fill) pounds/inch
breaking strength. Consists of almost
pure Silica with 1 to 3% Chrome Oxide.
Continuous use to 2300°F.
Heat Cleaned (Preshrunk)
C 100-48: 8 Harness Satin, 26 mils thick, 18.6 oz/
sq. yd., 52 x 39 thread count per inch,
86 (warp) and 61 (fill) pounds/inch
breaking strength. Consists of almost
pure Silica (99%+). Good to 2300 F.
C 100-96:
12 Harness Satin, 50 mils thick, 37.1 oz/
sq-. yd., 50 x 39 thread count per inch,
130 (warp) and 65 (fill) pounds per inch
breaking strength. Essentially pure
Silica. Good to 2300 F
85
-------
Media Samples — Description (Continued)
Fiberfrax Cloth
Fiberfrax Paper
Zircar Zirconia
Saffil Alumina
(1C1)
Kaowool
(Babcock & Wilcox)
No organic binder (removed), 2300°F service temperature
L-126TT grade; 1/8" thick, twill weave, 21 pcf,
32 oz/sq. yd.
L-144TT grade; 1/10" thick, twill weave, 28 pcf,
34 oz/sq. yd., has nichrome wire
insert, strong to 2000°F
52% Al-0., 48% SiO-, 2300°F service temperature, up to
1" long, 2-3 microns (mean) fibers, 2.53 gm/cc density
970-AH grade;
970-J grade;
Felt ZYF-100:
1/32" thick, no binder, 12 pcf,
4.5 oz/sq. yd.
1/8" thick, has up to 5% binder,
10 pcf, 15 oz/sq. yd.
19 oz/sq. yd., 0.1" thick, 15 pcf,
96% voids, 600 scfm/ft2 at 0.5 psi,
5.8 gm/cc density, 4.5 ym fibers,
no binders. Breaking strength
1.6 pounds/in, width
Fiber density 3.4 gin/cm3, 3000°F service temperature,
3 micron fibers, 1-2" long, 1.5 cmVgm
surface area, 95% A12°3 — 5% Si02
Mat 4 pcf, % in. thick, 2.4 oz/sq. yd.,
no binder
Paper 12 pcf, 0.04 in. thick, 6 oz/sq. yd.,
has binder
47% alumina - 53% silica, good to 2600°F, 2.8 micron
fibers, 2 in. long and up to 4 in. long. Has no binder.
2.6 g/cc density, 6 pcf, k, %, and 1" thick, 18, 36,
72 oz/sq. yd.
Fiberfrax (Carborundum)
Durablanket
Fiberchrome
(J. Manville)
Felt
48% A120_, 52% Si02, 2300°F service
temperature, long fibers, 2-3 micron
(mean) fiber diameter, density 2.62 gm/cc.
No binder, 6 pcf, ^ in., 18 oz/sq. yd.
41% alumina, 55% silica, 4% Cr_0_.
g«*pd to 2700°F. 3.5 micron fiEers,
8 pcf, 4 In. thick, 48 oz/sq. yd.
86
-------
Media Samples — Description (Concluded)
Refrasil (continued)
UC 100-48:
AB-312
(3M Co.)
Zircar Cloth
(Zircar Corp)
8 Harness Satin, 26 mils thick, 18 oz/
sq. yd., 46 x 36 thread count per inch,
80 (warp) and 60 (fill) pounds ger inch
breaking strength, good to 2300 F.
99%+ Silica
Heat Cleaned. Alumina-Boria-Silica, 390 filament strand,
11 micron diameter filaments, 250,000 psi tensile strength,
22 x 10* psi modulus, 2300 F service temperature,
2.5 gm/cc density
Style 22B;
Style 22T;
ZYW 30A:
50-1/0, 32 x 38 thread count/inch,
9.3 oz/sq. yd, 12 mils thick basket weave.
50-1/0, 32 x 25 thread count/inch, 7.6 oz/
sq. yd., 10 mils thick, twill weave
20 oz/sq. yd., 0.030 in. thick, 5 Harness
Satin, 63 pcf, 83% porosity, 5.7 gm/cc
density, 5 micron continuous filaments.
Breaking strength 4 pounds/in, width
BET surface area 1 m /gm, 92% Zr 0, -
8Z Y 0 . Good to 3300°F, 2-ply x
filament yarn, 45 x 34 threads/in.
87
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing)
i. REPORT NO.
EPA-600/7-78-194
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
High-temperature, High-pressure Particulate Control
with Ceramic Bag Filters
5. REPORT DATE
October 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
M.A. Shackleton
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Corporation
485 Clyde Avenue
Mountain View, California 94042
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-02-2169
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 8/76 - 8/78
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES T£RL-RTP project officer is Dennis C. Drehmel, Mail Drop 61,
919/541-2925.
16. ABSTRACT
The report gives results of bench-scale research indicating that fine particle
control at high temperature and pressure can be achieved using barrier filtration by
ceramic bag filters. Evidence supporting this contention includes: (1) 'blanket' ceramic
fiber materials (felts) consisting of small diameter fibers (3.0 micrometers) appear
to be the most promising materials for high-temperature and -pressure applications
because of their combination of good'filtration performance and relatively high stren-
gth; and (2) accelerated media cleaning tests at high temperatures and pressures show
that several ceramic filter structures are capable of surviving in excess of 50,000
cleaning pulses while maintaining pressure drop at acceptable levels.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Air Pollution
Dust
Ceramic Fibers
Felts
High Temperature Tests
High Pressure Tests
Air Pollution Control
Stationary Sources
Particulate
Ceramic Bag Filters
13B
UG
14B
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PACES
75
20. SECURITY CLASS /Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (»-73)
------- |