svEPA
United States
Environmental Protection
Agency
Industrial Environmental Research"
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-81-111 Dec. 1981
Project Summary
Emissions and Efficiency
Performance of Industrial
Coal-Stoker-Fired Boilers
P. L. Langsjoen, J. 0. Burlingame, and J. E. Gabrielson
The report gives results of field
measurements of 18 coal stoker-fired
boilers including spreader stokers,
mass-fired overfeed stokers, and
mass-fired underfeed stokers. The test
variables included stoker design, heat
release rate, excess air, coal analysis
and sizing, overfire air, and flvash
reinjection. Measurements included
O2, CO2, CO, NO, N02, SOz, SOa,
gaseous hydrocarbons, uncontrolled
and controlled particulate mass load-
ing, particle size distribution of the
flyash, combustible content of ash,
sulfur retention in the ash, and boiler
efficiency. Particulate loading is
shown to be largely dependent on
stoker type and degree of flyash rein-
jection. It increases with heat release
rate, but can be controlled with proper
use of overfire air in many cases. IMOx
increases with excess air and grate
heat release rate. These relationships
are defined in the report. Overfire air,
as it exists in current boiler designs,
does not affect NO>. The report also
addresses other relationships
between operating variables and
measured emissions and efficiency. A
separate data supplement is available.
This Project Summary was develop-
ed by EPA's Industrial Environmental
Research Laboratory, Research Tri-
angle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Introduction
In late 1977, the American Boiler
Manufacturers Association (ABMA)
was awarded a contract to update speci-
fications and design parameters for
coal-burning boiler and stoker equip-
ment. The project was jointly funded by
the U.S. Department of Energy and the
U.S. Environmental Protection Agency
(EPA), with the purpose of increasing
coal usage in an environmentally
acceptable manner.
The Need
The need for such a program is clear.
In recent years the vast majority of
industrial boiler installations have been
packaged or shop-assembled gas- and
oil-fired units. These boilers could be
purchased and installed at substantially
lower costs than conventional coal-
burning boiler-stoker equipment.
Because of the declining demand for
coal stokers, little or no work has been
done in recent years to improve specifi-
cation data or product information made
available to consulting engineers and
purchasers of coal burning boiler-stoker
equipment.
Furthermore, the market for coal suit-
able to be fired in industrial boilers is
being held back by critical uncertainties
in the environmental and energy areas,
causing potential customers of coal-
fired industrial boilers to shelve plans
for capital expansion and conversion.
-------
The current implementation of more
rigid air pollution regulations has made
it difficult for many coal burning instal-
lations to comply with required stack
emission limits.
It is highly desirable to remove these
uncertainties and thereby encourage
industrial users to order and install coal-
stoker-fired boilers. This would lead to
significantly increased coal usage and
decreased dependence on scarce and
imported fuels.
The Objectives
Objectives of this program are:
1. To advance stoker boiler technol-
ogy through comprehensive test-
ing of various stoker boiler designs,
thereby facilitating the design and
fabrication of stoker boilers which
are economically and environ-
mentally satisfactory alternatives
to gas- and oil-fired units.
2. To contribute to the design and
application of pollution control
equipment by generating a large
data base of boiler outlet dust
loading data and particle size
distribution data
3. To provide guidelines for boiler
operators concerning techniques
for clean and efficient stoker boiler
operation.
4. To facilitate preparation of intelli-
gent and reasonable national
emission standards for coal-
stoker-fired boilers by the EPA.
5. To provide assistance in planning
for coal supply contracts both
through an increased knowledge
of the effects of coal properties on
emissions, and through the
development of reasonable
emission regulations.
6. To promote the increased utiliza-
tion of coal-fired-stoker boilers by
U.S. industry by ensuring com-
patibility of emissions from these
units with applicable environ-
mental requirements.
The Project Organization
The ABMA formed a Stoker Technical
Committee (STC) composed of person-
nel from member companies to oversee
the project. The STC, in turn, subcon-
tracted the field testing and report work
to KVB, Inc., a combustion consulting
firm in Minneapolis, MN. The original
scope of work included the testing of six
spreader stokers. Testing on the first
unit began August 9, 1977.
As the project progressed success-
fully, additional funding was obtained
and the scope of work was increased to
include five mass-fired overfeed stokers.
These units were also tested by KVB,
Inc.
A separate subcontract was let to
Pennsylvania State University to test
seven small stoker boilers including two
overfeed stokers and five underfeed
stokers located in central steam heating
plants. The purpose of this subcontract
was to determine particulate mass
emission rates and particle size distri-
bution for small stokers. On November
12, 1979, all testing was completed. In
total, 400 tests on 18 coal-stoker-fired
boilers were conducted.
Related Reports and Data
A Project Summary cannot discuss all
the ramifications of the project and the
data collected. The reader is directed to
the final report (which this
summarizes), the various site reports,
and "A Guide to Clean and Efficient
Operation of Coal-Stoker-Fired
Boilers," EPA-600/8-81-016, May
1981, for additional information.
Summary and Conclusions
This report is the culmination of an
extensive testing effort on 18 coal-
stoker-fired boilers. The effort includes
400 tests on 36 boiler/coal combina-
tions over a 2-year period. The boilers,
identified by letter designators, fall into
three major stoker classifications:
spreader stokers (Sites A, B, C, E, F, G),
mass-fired overfeed stokers (Sites D, H,
I, J, K, L2, L4), and underfeed stokers
(Sites L1, L3, L5, L6, L7). Each classifi-
cation is presented separately in this
report. The units are described in Table
1 along with the number of coals fired
and tests conducted.
The major objective of this test pro-
gram was to update stoker specification
data by measuring boiler emissions and
efficiency on a variety of boiler-stoker
designs and under a variety of operating
conditions. The operating variables
included heat release rate, excess air,
overfire air, flyash reinjection, and coal
properties. The measurements included
both uncontrolled and controlled par-
ticulate loading, nitrogen oxides (NOX—
NO, NO2), sulfur oxides (SOX—SO2,
S03), oxygen (02), carbon dioxide (C02),
carbon monoxide (CO), unburned hydro-
carbon (UHC), combustibles in the
flyash and bottom ash, particle size
distribution, and boiler efficiency. The
tests were conducted under steady load
conditions. ,
In stoker firing of coal, there are so
many variables that even with the
extensive amount of testing conducted
during this program it was not possible
to analyze them all. The interactions
between these variables are difficult to
assess.
Not all of the parameters were
determined on each site nor under the
full range of operating variables. For
example, the CO analyzer was out-of-
service during testing at Sites G, I, and
J. The UHC analyzer was only operable
during testing at four sites, and boiler
nameplate rating was not achieved on
three of the units due to retrofit equip-
ment on two units and start-up prob-
lems on a third. In addition, the testing
at Sites L1 through L7 was conducted
under a separate contract and included
a more limited number of test measure- *
ments under a single operating \
condition on each unit.
This report is organized in two sepa-
rate formats so as to be a convenient
reference to the widest possible
audience. The first section is organized
by the measured parameter first and the
operating variable second. Thus, for
example, all observations on particulate
loading are grouped together.
The second format follows the format
of the final report text. It is organized by
operating variable so that, for example,
the effects of overfire air on all emis-
sions are grouped together.
The range of data encountered at full
load is summarized in Table 2.
Summary of Findings
Organized by Measured
Parameter
Particulate Loading
Type of Stoker—Spreader stokers
with flyash reinjection from their
mechanical dust collectors had by far
the highest uncontrolled particulate
loadings, 13-36 lb/106 Btu. Spreader
stokers without reinjection from their
dust collectors were next with emis-j
-------
Table 1. Unit Description and Data Base
Site
A
B
C
D
E
F
G
H
1
J
K
L1
L2
L3
L4
L5
L6
L7
Stoker
Type
Spreader
Spreader
Spreader
Vibrating Grate
Spreader
Spreader
Spreader
Traveling Grate
Traveling Grate
Chain Grate
Traveling Grate
Multiple Retort
Vibrating Grate
Single Retort
Traveling Grate
Multiple Retort
Multiple Retort
Multiple Retort
Design Number
Capacity Coals
Ib/hr Tested
300,000
200,000
182,500
90,000
180,000
80,000
75,000
45,000
70,000
70,000
50.000
26,000 *
30,000
23,300
27,000
28,460
20,000
50.000
3
4
3
3
3
2
3
J
2
2
3
J
J
1
1
1
1
1
Number
Test
Conditions
68
42
76
31
25
38
35
24
23
13
18
1
1 .
1
1
1
1
1
*The site L1-L7 report expresses steaming'capacity in terms of peak, or maximum,
rating. This report expresses the Site L1-L7 steaming capacity in terms of maximum
continuous ratings so as to be consistent throughout.
sions of 2.1 -8.8 lb/106 Btu. As shown in
Figure 1, withoutflyash reinjectionfrom
the mechanical collector, the uncontrol-
led particulate data were 2.1-8.8
lb/10B Btu. Test Site C was operated
both with and without flyash reinjection
from the mechanical collector and had
very different particulate loadings under
the two conditions. As a result of
operating without reinjection, uncon-
trolled particulate loading was reduced
by 70-80 percent and controlled partic-
ulate loading by 40-50 percent. This
shows that a portion of the reinjected
flyash is reentrained in the gas stream
and results in increased particulate
loadings. Three boilers had some
degree of flyash reinjection from the
boiler hopper, and at Site F from the
economizer hopper. The amount of fly-
ash reinjected depends on duct
geometry and whether or not the boiler
is equipped with baffles. In most cases,
the actual rate of reinjection was not
known. These are followed by mass-
fired overfeed stokers with 0.57-2.2
lb/10s Btu and underfeed stokers with
'25-0.71 lb/106 Btu. Figure 2 shows
uncontrolled particulate loadings of
0.57-2.2 lb/106 Btu on the five exten-
sively tested mass-fired overfeed
stokers. Averages for these same five
stokers were 0.78-1.4 lb/106 Btu. Sites
L2 and L4 had lower particulate load-
ings of 0.56 and 0.50 lb/106 Btu,
respectively, but these were obtained at
lower loads of 85 and 78 percent of
design capacity, respectively. The Site
L2 and L4 particulate data is not out of
line when compared with data obtained
at the same grate heat release from the
other stokers.
Heat Release Rate—It cannot be said
that units with higher design heat
release rates have higher particulate
loadings, but for a given unit the uncon-
trolled particulate loading always
increased as heat.release rate, or load,
increased. The rate of increase varied
form site to site, and at some sites it
appeared to accelerate as fi/ll load was
approached. On spreader stokers with
flyash reinjection from mechanical dust
collectors, the last 10 percent increase
in heat release rate resulted in a 9-20
percent increase in particulate loading.
On spreaders without dust collector
reinjection, the increase was 8-12
percent. On mass-fired overfeed
stokers, particulate loading increased 3-
20 percent as heat release rate was
increased from 90 to 100 percent of
design.
Excess Air—No relationship was
established between particulate loading
and excess air. This does not foreclose
the existence of such a relationship, but
rather indicates that such a relationship
could not be deciphered from the data
due to data scatter and uncontrolled
variables.
Over fire Air—Uncontrolled
particulate loading was reduced by 20-
50 percent on four of six spreader
stokers and three of five mass-fired
overfeed stokers when overf ire air pres-
sures were increased. Two sites
showed the opposite trend and two sites
were unaffected by changes in overfire
air pressure.
Coal Ash—Coal ash could be related
to particulate loading at only four of the
ten test sites at which multiple coals
were fired. On three of the spreader
stokers, particulate loading increased
by 0.24-0.38 lb/106 Btu for each 1 per-
cent increase in coal ash. Stated
another way, if the coal ash is doubled at
these sites, the particulate loading will
increase by 15-30 percent. Thus, the
relationship between coal ash and
particulate loading was not 1:1 on these
three units.
On one of the traveling grate stokers,
a 4-percent ash-washed coal and a 10-
percent ash unwashed coal from the
same mine were tested. The 250 per-
cent increase in coal ash resulted in a
300 percent increase in particulate
loading. In this case, the dramatic
increase in particulate loading can be
attributed to the type of ash, a clay like
material in'the surface of the coal, and
to a corresponding increase in coal fines
on the unwashed coal.
Coal Fines—Because of the move-
ment of air through the grate and the
upward movement of combustion gases
through the furnace, the smallest coal
and ash particles are carried out of the
furnace by the gases rather than stayi ng
on the grate. This is called particle
entrainment and is a problem from both
a pollution and an efficiency standpoint.
The likelihood of a particle being en-
trained is a function of its size and
density, and the velocities in the fur-
nace. The test data from this program
showed a mathematical correlation
between coal fines and particulate load-
-------
Table 2. Range of Data Encountered at High Load*
Spreader Stokers
With Reinfection
from D.C.**
Uncontrolled Paniculate, lb/106 Btu
Controlled Paniculate, lb/106 Btu
Mechanical Collector Efficiency, %
Excess Air, %
Nitric Oxide, lb/106 Btu as NO 2
Carbon Monoxide, ppm dry @ 3% O2
Unburned Hydrocarbons, ppm wet @ 3% 02
Combustibles in Flyash. %
Combustibles in Bottom Ash, %•
Flyash Combustibles Heat Loss, %
Bottom Ash Combustibles Heat Loss, %
Boiler Efficiency, %
12.7
0.60
94.9
18
0.30
22
No
7.1
0.0
0.54
0.00
75.79
- 36.4
- 3.5
- 98.0
- 113
- 0.60
- 1600
Data
- 65.6
-34.4
-5.5
-3.0
- 83.43
Spreader Stokers
W/0 Reinjection
from D.C.
2.1
0.17
40.6
19
0.36
33
0
26.6
0.3
0.51
0.04
73.00
-8.8
- 3.8
-96.0
-82
-0.61
- 702
-41
-83.5
-27.2
-9.2
-3.4
- 83.07
Mass-Fired
Overfeed
Stokers
0.57
0.11
10.9
26
0.21
39
5
21.8
7.1
0.16
0.42
69.75
-2.2
-0.75
-92.7
-97
-0.50
-2300
- 112
-56.0
-69.1
- 1.1
-9.4
-84.10
Mass-Fired
Underfeed
Stokers
0.25
0.46
26.6
33
No
-0.71
-0.58
- 42.9
- 186
Data
<1000
No
20.2
8.1
0.07
1.2
64.13
Data
- 20.5
-25.0
- 0.21
-3.9
- 76.81
Underfeed stokers were tested at loads 55 to 100% of capacity. Data from the other stokers were obtained within the upper 10% of
the obtainable load range. •
f Does not include tests in which reinjection from the dust collector was reduced. For example, a NO level ofO. 68lb/106 Btu as NO2
measured during one reduced reinjection test is not included in this table. A paniculate loading of 9.6 Ib/106 Btu is excluded for the
same reason.
ing on five stokers. Participate loading
increased by 0.10-0.55 lb/106 Btu
whenever the percent of coal passing a
16 mesh screen increased by 1 percent.
No correlation was found in studies of
six other stokers.
Flyash Reinjection—Flyash from the
dust collector was reinjected to the fur-
nace of three of the six spreader stokers.
In each case, uncontrolled paniculate
loading was increased as a result of
reentrainment of a portion of the rein-
jected ash. At one site, reinjection was
completely eliminated for test purposes.
As a result, uncontrolled particulate
loading was reduced by 70-80 percent
and controlled particulate loading was
reduced by 40-50 percent. Reducing the
degree of flyash reinjection reduced the
percentage of larger particles in the fly-
ash. This in turn reduced the mechan-
ical dust collector efficiency.
Emission Factors—EPA report AP-42,
Compilation of Air Pollutant Emission
Factors, Third Edition, contains factors
used for predicting emissions from
stoker boilers. The data from this
program compares as follows:
Uncontrolled Particulates
Ib/ton
(A = % Ash in Coal)
AP-42 This Program
Spreaders with
Reinjection
20A
Spreaders with-
out Reinjection 13A
Overfeed
Stokers
Underfeed
Stokers
5A
5A
29A-50A
14A-17A
1.1A-38A
0.6A-1.7A
Particle Size Distribution—Particle
size distribution of the flyash was deter-
mined by a variety of methods including
cascade impactor, Bahco classifier,
SASS cyclones, and sieve analysis.
Results varied from one method of
measurement to another, but clearly
showed that spreader stokers emit a
higher percentage of coarse, more
easily collected particles than mass-
fired overfeed and underfeed stokers.
Nitric Oxides (NO*)
Type of Stoker—As a class, spreader
stokers em itted higher concentrations of
NO than did mass-fired overfeed
stokers. Under full load, spreader
stokers emitted 0.30-0.61 lb/106 Btu
NO corrected to N02. Figure 3 shows NO
data, measured at the boiler outlet using
a chemiluminescent analyzer, of 0.30-
0.68 lb/106 Btu, calculated as N02
when measured at full load. NO levels
were found to be a function of excess
air, heat release rate, and combustion
temperature. Where NOj, is measured,
-------
2 15
CQ
o
\
C
=3 10
to
~j
5i
(Q
.O
1 5
Uncontrolled ,
—
Note:
Ce = Site C,
Cw - Site C,
—
1 |
22 1 •
•"- (Q CO ^|
1 ! |
1 1 1 1 1 1
Eastern Coal.
Western Coal.
I
1 Key
m High-m
Ave.-\
1 Low —
Figure 1.
A B Ce Cw E F G
Test Site Designator
Uncontrolled paniculate loadings of four spreader stokers fired at full
load without flyash reinfection from the mechanical collector.
C
o
Cj
•-*
b
0.5
I
I
C
to
a.
s
.c
Key
High —-
Avg. -
Low -
I
H I J K L2
Test Site Designator
L4
Figure 2.
Uncontrolled paniculate loadings of seven mass-fired overfeed stokers
fired at or near full load.
NOz did not exceed 4 percent of the total
NOx and was most often negligible.
Mass-fired overfeed stokers emitted
0.21-0.50 lb/106 Btu NO. However,
• overfeed stokers operated at higher
excess air levels than did the spreader
stokers. When compared at the same
excess air levels, the difference in NO
levels is even greater. As shown in
Figure 4, NO emissions were 0.21 -0.50
lb/106 Btu computed as N02. Site
averages were 0.27-0.41 lb/106 Btu.
Some of the variations between sites
are the result of different excess air
operating levels. For example. Site H
was operated at an average excess air of
70 percent compared to 51 percent for
Site I. As a result, Site H NO emissions
were higher.
Heat Re/ease Rate—For spreader
stokers, an increase in heat release rate
equivalent to 10 percent of capacity
resulted in an average increase in NO
emissions of 0.025 Ib7106 Btu as N,02 at
constant excess air. For mass-fired
overfeed stokers, the relationship was
0-0.026 lb/106 Btu per 10 percent
increase in capacity at constant excess
air. In all cases, NO emissions were
invarient with load at normal firing
conditions because the effects of
decreasing excess air effectively
canceled the effects of increasing load.
Although NO increased with heat re-
lease rate on each given unit, it was not
true that units with higher design heat
release rates emitted higher concentra-
tions of NO.
Excess Air—On four spreader stokers
without air preheat and one with air pre-
heat, NO increased by 0.021-0.036
lb/106 Btu for each increase of 10 per-
cent excess air. The sixth spreader
stoker used air preheat and its NO
increased by 0.067 lb/10" Btu per
increase of 10 percent excess air. On
five-mass fired overfeed stokers, NO
increased by 0.016-0.027 lb/106 Btu.
Overfire Air—NO emissions were not
influenced by changes in overfire air
pressure when considered at constant
excess air.
Fuel Nitrogen—Variations in fuel
nitrogen of 0.75-1.50 percent by weight
had no measurable effect on NO emis-
sions. This may simply reflect difficul-
ties in sorting out the other variables.
Flyash Reinjection—Flyash reinjec-
tion from the mechanical dust collector
had no measurable effect on NO emis-
sions.
Emission Factors—EPA report AP-42,
Compilation of Air Pollutant Emission
-------
0.8
0.6
o
\
I
O
o
0.4
0.2
—
1
—
1
1
1
1
1
Note:
Ce
Cw
1'
1 1
= Site C, Eastern Coal.
= Site C, Western Coal.
\
\
I
\
Key
High-.
Ayg. -[
Low —*
B Ce Cw £
Test Site Designator
Figure 3. Nitric oxide emissions
0.8
S 0.6
\
•§"
O
| 0.4
0.2
~
.1
—
1 1
D H
of six
I
\
1
spreader stokers
\l
\ \
J K
fired at full load.
&
-------
excess air conditions. At full load, CO
emissions could be controlled with
proper application of combustion air.
Excess Air—CO was more prevalent
as excess air dropped below about 30-
40 percent on spreader stokers and
about 60 percent on mass-fired over-
feed stokers. CO increased gradually as
excess air increased about 60 percent
on spreader stokers and 100 percent on
mass-fired stokers.
Overfire Air—CO emissions were
reduced by the increased use of overf ire
air.
Coal Rank—CO emissions were
greatest while firing Western sub-
bituminous coals. On one spreader
stoker where both an Eastern and a
Western coal were fired, the full-load
Western coal emissions were 163-702
ppm and averaged 342 ppm. By compar-
ison, the full-load Eastern coal emis-
sions were 33-263 ppm and averaged
71 ppm.
Flyash Reinjection—Flyash reinjec-
tion from the mechanical dust collector
had no measurable effect on CO emis-
sions.
Unburned Hydrocarbon (UHC)
Type of Stoker—Based on limited
data, the spreader stokers emitted lower
UHC emissions than the mass-fired
overfeed stokers. Full-load emissions
from the spreader stokers were 0-15
ppm for Site F and 35-41 ppm for Site G.
By comparison, the mass-fired overfeed
stokers emitted 5-112 ppm for Site H
and 80 ppm for a single point on Site J.
Heat Re/ease Rate—UHCs tended to
decrease as heat release rate increased
on three of four stokers where they
were measured. On the fourth stoker,
the opposite trend was observed.
Excess Air—UHC emissions showed
little or no correlation with'excess air on
spreader stokers. On mass-fired
overfeed stokers, UHCs increased in
almost direct proportion to the excess
air.
Overfire Air—UHCs were reduced 82
percent by increasing the overfire air
pressure on one traveling grate stoker.
No correlation was found on one
spreader stoker. The other two units
where UHC emissions were measured
had insufficient data to make a correla-
tion.
Coal Properties—The site firing the
lower volatile coal had the lowest UHC
emissions. The 29 percent volatile coal
yielded 19-41 ppm UHCs, while the 41
percent volatile coal yielded 163-702
ppm UHCs. Volatiles are expressed here
on a dry mineral-matter-free basis.
Carbon Monoxide—UHCs increased
with increasing CO emissions on one
traveling grate stoker'. No. correlation
was found on one spreader stoker.
Excess Air
Type of Stoker—At full load, most
spreader stokers were capable of oper-
ating at 30 percent excess air (5 percent
Oz). By comparison, the mass-fired
overfeed stokers generally required 50
percent excess air (7 percent Oz).
Size of Stoker—With one exception,
the excess air operating level was
inversely proportional to the size of the
stoker. The larger the stoker, the lower
the excess air requirement.
Heat Release Rate—The excess air
requirement drops as heat release rate
increases on stoker boilers. The excess
air requirement levels off as 30 percent
excess air is approached.
Coal Properties—Coal properties
were not found to alter excess air
requirements on these stoker boilers.
Combustibles in Bottom Ash
Type of Stoker—Combustible levels
were lower in the bottom ash of
spreader stokers than they were for
mass-fired overfeed stokers or under-
feed stokers. The average for each of six
spreader stokers fired at full load was 0
to 14 percent. By comparison, mass-
fired overfeed stokers were 16-26 per-
cent, with one unit averaging 43
percent, and underfeed stokers were
19-25 percent, with one unit averaging
8 percent.
Heat Re/ease Rate—Heat release rate
had very little effect on combustibles in
the bottom ash.
Excess Air—No correlation was found
between excess air and combustibles in
the bottom ash.
Coal Properties—Small differences in
bottom ash combustible levels were
observed which appeared to be related
to coal properties at some sites. How-
ever, the particular coal properties
causing these differences were not
identified.
Ash Balance—It was found that 65-
85 percent of the coal ash remained on
the grate in spreader stokers, compared
to 80-90 percent for mass-fired over-
feed stokers. To compute combustible
heat losses, 75 and 85 percent are good
estimates for spreaders and mass-fired
overfeed stokers, respectively.
Combustibles in the Flyash
Type of Stoker—Combustible levels in
the flyash were higher in the spreader
stokers than in either the mass-fired
overfeed stokers or the underfeed
stokers. Except at Test Site C, the
spreader stoker data were 47-84 per-
cent and averaged 60 percent. On the
other hand, the mass-fired overfeed
stoker data were 22-56 percent and
averaged 28 percent. Flyash samples
taken from the dust collector hoppers of
two underfeed stokers revealed 20.2
and 20.5 percent combustibles.
Heat Re/ease Rate—Combustibles in
the flyash tended to increase slightly as
heat release rate increased on spreader
stokers. On mass-fired overfeed
stokers, no significant trend was
observed.
Excess Air—No correlation was found
between combustibles in the flyash and
excess air level on either spreader
stokers or mass-fired overfeed stokers.
Overfire Air—Increasing overfire air
pressure effectively reduced the
combustible content of the flyash by an
average 40 percent in 74 percent of the
overfire air tests. This resulted in an
average efficiency gain of 1.70 percent
of heat input for spreader stokers and
0.27 percent of heat input for the mass-
fired overfeed stokers. However, 26
percent of the tests gave the opposite
result
Coal Properties—At Test Site C, the
combustibles in the flyash were 2 to 4
times higher while firing an Eastern
bituminous coal than while firing a
Western sub-bituminous coal. This was
the only site where flyash combustibles
could be directly related to coal proper-
ties. The property of the coal responsible
for the difference was not identified.
Flyash Reinjection—Combustibles in
the flyash at the boiler outlet increased
by 23-63 percent when the rate of
flyash remjection was reduced. At the
dust collector outlet, similar increases
were observed.
Particle Size—The largest flyash
particles contain the largest combusti-
ble fractions. Flyash samples from two
spreader stokers and two mass-fired
overfeed stokers were analyzed.
Boiler Efficiency
Type of Stoker—Boiler efficiencies
were determined by the ASME Abbrevi-
ated Efficiency Test (PTC-4.1). At or near
full load, the measured boiler efficien-
cies were 73.0-83.4 percent for six
-------
spreader stokers. As shown in Figure 5,
boiler efficiency was determined by the
heat loss method using the ASME
Abbreviated Efficiency Test (PTC 4.1). At
full load, boiler efficiencies were 73.0-
83.4 percent. The lowest efficiency
belongs to Site G, the only site which did
not have either an air heater or an econ-
omizer. Design efficiencies of these
units were: A-83.68, B-84.16, Cw-
81.40, E-80.41, F-83.10, and G-77.04
percent. Results of 69.8-84.1 percent
were obtained for seven mass-fired
overfeed stokers. As shown in Figure 6,
increased boiler efficiencies for these
units were 69.8-84.1 percent. Sites D,
J and K, equipped with economizers,
had the highest average efficiencies:
83.8, 81.8, and 78.4 percent, respec-
tively. Sites H and I, which did not have
economizers, averaged 75.4 and 73.9
percent boiler efficiencies, respectively.
Boiler efficiencies were determined by
the ASME heat loss method (PTC 4.1).
Results of 64.1 to 76.8 percent were
measured for five mass-fired underfeed
stokers.
Heat Release Rate—\n most cases,
boiler efficiencies were relatively
constant with changing heat release
rates. At a few sites, efficiency dropped
as heat release rate dropped because
increasing dry gas heat losses predom-
inated.
Excess Air—Boiler efficiency de-
creased as excess air increased on all of
the extensively tested stokers. Dry gas
heat losses dominated this trend, over-
shadowing any effects due to combusti-
ble heat losses. For each 10 percent
excess air decrease, boiler efficiency
increased by 0.33-1.0 percent.
Over fire Air—Boiler efficiency
improved by an average 1 percent when
overfire air was increased on spreader
stokers as a result of reduced carbon
carryover. However, on mass-fired
overfeed stokers, efficiency was
reduced by an average 2.75 percent
when overfire air was increased due to
increased dry gas losses and increased
bottom ash combustible heat losses.
Coal Properties—Coal properties
affected boiler efficiencies on two
occasions. At Test Site C, the high mois-
ture Western coal produced efficiencies
which were 3-4 percent lower than
similar tests on low moisture Eastern
coals. At Test Site K, the unwashed coal
produced lower boiler efficiencies than
either of the others because this coal led
to a greater combustible heat loss.
Flyash Reinjection—Some but not all
of the carbon in the reinjected flyash
8
90
ciency,
§
Uj
o
03
70
60
l-i",
Note:
Ce = Site C, Eastern Coal.
Cw = Site C, Western Coal.
3.
I
I
I
S J
-------
was recovered at Sites A, B, and C.
There was insufficient data to calculate
carbon recovery rates with any
accuracy.
Summary of Findings
Organized by Test Variable
Differences Between Stoker
Types
Excess Air—At full load, most
spreader stokers were capable of oper-
ating at 30 percent excess air (5 percent
O2). By comparison, the mass-fired
overfeed stokers generally required 50
percent excess air (7 percent Oa).
With one exception, the excess air
operating level was inversely propor-
tional to the size of the stoker. The larger
the stoker, the lower the excess air
requirement.
Particulate Loading—Spreader
stokers with flyash reinjection from
their mechanical dust collectors had by
far the highest uncontrolled particulate
loadings: 13-36 lb/10B Btu. Spreader
stokers without reinjection from their
dust collectors were next with emis-
sions of 2.1-8.8lb/106Btu, followed by
mass-fired overfeed stokers with 0.57-
2.2 lb/106 Btu and underfeed stokers
with 0.25-0.71 lb/106 Btu.
Combustibles in the Flyash—Com-
bustible levels in the flyash were higher
in the spreader stokers than in either
the mass-fired overfeed stokers or the
underfeed stokers. Except at Test Site C,
the spreader stoker data were 47-84
percent and averaged 60 percent. On
the other hand, the mass-fired overfeed
stoker data were 22-56 percent and
averaged 28 percent. Flyash samples
from the dust collector hoppers of two
underfeed stokers revealed 20.2 and
20.5 percent combustibles.
Combustibles in the Bottom Ash—
Combustible levels were lower in the
bottom ash of spreader stokers than for
mass-fired overfeed stokers or under-
feed stokers. The average for each of six
spreader stokers fired at full load was 0-
14 percent. By comparison, mass-fired
overfeed stokers were 16-26 percent
with one unit averaging 43 percent, and
underfeed stokers were 19-25 percent
with one unit averaging 8 percent.
Sulfur Oxides (SO*)—The spreader
stokers retained an average 4.4 percent
of the fuel sulfur in the ash, while the
mass-fired overfeed stokers retained an
average 2.1 percent. The remainder
was emitted as SOYand SOs, with S03
comprising less than 2 percent of the
total. Operating parameters such as
excess air, overfire air, and load had no
effect on the emissions of SO* or the
retention of sulfur in the ash.
Nitric Oxides (/VOX)—As a class,
spreader stokers emitted higher con-
centrations of NO than did mass-fired
overfeed stokers. Under full load,
spreader stokers emitted 0.30-0.61
lb/106 Btu NO corrected to N02 while
mass-fired overfeed stokers emitted
0.21-0.50 lb/106 Btu NO. In addition,
overfeed stokers operated at higher
excess air levels than did spreader
stokers. When compared at the same
excess air levels, the difference in NO
levels is even greater.
Carbon Monoxide (CO)—Spreader
stokers emitted lower concentrations of
CO than traveling grate stokers while
firing Eastern bituminous coals. Emis-
sions from three of the spreader stokers
were 50-250 ppm at full load. A fourth
was 200-600 ppm. By comparison, two
traveling grate stokers emitted 50-700
ppm CO at full load, and a vibrating grate
stoker emitted 50-2000+ ppm CO. The
comparison is limited to these seven
stokers. CO emissions were not meas-
ured on three other stokers due to
instrument failure, and a fourth fired
only Western coals. At Test Sites LI
through L7, the CO concentration was
measured with an Orsat analyzer with a
minimum detection limit of 1 percent or
1000 ppm. Significantly, the CO emis-
sions were below this detection limit on
the Site L stokers.
Unburned Hydrocarbon (UHC)—
Based on limited data, the spreader
stokers emitted lower UHC emissions
than the mass-fired overfeed stokers.
Full-load emissions from the spreader
stoker were 0-15 ppm for Site F and 35-
41 ppm for Site G. By comparison, the
mass-fired overfeed stokers emitted 5-
112 ppm for Site H and 80 ppm for a
single point on Site J.
Boiler Efficiency—Boiler efficiencies
were determined by the ASME Abbrevi-
ated Efficiency Test (PTC 4.1). At or near
full load, the measured boiler efficien-
cies were 73.0-83.4 percent for six
spreader stokers, 69.8-84.1 percent for
seven mass-fired overfeed stokers, and
64.1 -76.8 for five mass-fired underfeed
stokers.
Response to Heat Release Rate
Excess Air—The excess air require-
ment drops as heat release rate
increases on stoker boilers. The excess
air requirement levels off as 30 percent
excess air is approached.
Particulate Loading—\t cannot be said
that units with higher design heat
release rates have higher particulate
loading, but for a given unit the uncon-
trolled particulate loading always
increased as heat release rate, or load,
increased. The rate of increase varied
from site to site; at some sites it
appeared to accelerate as full load was
approached. On spreader stokers with
flyash reinjection from mechanical dust
collectors, the last 10 percent increase
in heat release rate resulted in a 9-20
percent increase in particulate loading.
On spreaders without dust collector
reinjection, the increase was 8-12 per-
cent. On mass-fired overfeed stokers,
particulate loading increased from 3 to
20 percent as heat release rate was
increased from 90 to 100 percent of
design.
Combustibles in the Flyash—Com-
bustibles in the flyash tended to
increase slightly as heat release rate
increased on spreader stokers. On
mass-fired overfeed stokers, no signifi-
cant trend was observed.
Combustibles in the Bottom Ash—
Heat release rate had very little effect on
combustibles in the bottom ash.
Nitric Oxides (NOX)—For spreader
stokers, an increase in heat release rate
equivalent to 10 percent of capacity
resulted in an average increase in NO
emissions of 0.025 Ib/106 Btu as NO2 at
constant excess air. For mass-fired
overfeed stokers, the relationship was
0-0.026 lb/108 Btu per 10 percent
increase in capacity at constant excess
air. In all cases, NO emissions were
invarient with load at normal firing
conditions because the effects of de-
creasing excess air effectively canceled
the effects of increasing load. Although
NO increased with heat release rate on
each given unit, it was not true that
units with higher design heat release
rates emitted higher concentrations of
NO.
Carbon Monoxide (CO)—CO emis-
sions were highest at high heat release
rates under low excess air conditions,
and at low heat release rates under high
excess air conditions. At full load, CO
emissions could be controlled with
proper application of combustion air.
Unburned Hydrocarbon (UHC)-UHCs
tended to decrease as heat release rate
increased on three of four stokers
where UHCs were measured. On the
-------
fourth stoker, the opposite trend was
observed.
Boiler Efficiency—In most cases,
boiler efficiencies were relatively con-
stant with changing heat release rates.
At a few sites, efficiency dropped as
heat release rate dropped because
increasing dry gas heat losses predom-
inated.
Response to Excess Air
Paniculate Loading—No relationship
was established between particulate
loading and excess air. This does not
foreclose the existence of such a rela-
tionship, but rather indicatesthat such a
relationship could not be deciphered
from the data due to data scatter and
uncontrolled variables.
Combustibles in the Flyash—No
correlation was found between
combustibles in the flyash and excess
air level on either spreader stokers or
mass-fired overfeed stokers.
Combustibles in the Bottom Ash—No
correlation was found between excess
air and combustibles in the bottom ash.
Nitric Oxide (/VOX)—On four spreader
stokers without air preheat and one
with air preheat, NO increased by
0.021-0.036 lb/106 Btu for each
increase of 10 percent excess air. The
sixth spreader stoker used air preheat
and its NO increased by 0.067 lb/106
Btu per increase of 10 percent excess
air. On five mass-fired overfeed stokers,
NO increased by 0.016-0.027 lb/106
Btu.
Carbon Monoxide (CO)—CO was
more prevalent as excess air dropped
below about 30-40 percent on spreader
stokers and about 60 percent on mass-
fired overfeed stokers. CO increased
gradually as excess air increased above
about 60 percent on spreader stokers
and 100 percent on mass-fired overfeed
stokers.
Unburned Hydrocarbon (UHQ—UHC
emissions showed little or no correla-
tion with excess air on spreader stokers.
On mass-fired overfeed stokers, UHCs
increased in almost direct proportion to
the excess air.
Boiler Efficiency—Boiler efficiency
decreased as excess air increased on all
of the extensively tested stokers. Dry
gas heat losses dominated this trend,
overshadowing any effects due to
combustible heat losses. For each 10
percent excess air decrease, boiler
efficiency increased by 0.33-1.0
percent.
Response to Coal
Composition and Sizing
Excess Air—Coal properties were not
found to alter excess air requirements
on these stoker boilers.
Particulate Loading—Because of the
movement of air through the grate and
the upward movement of combustion
gases through the furnace, the smallest
coal and ash particles are carried out of
the furnace by the gases rather than
staying on the grate. This is called
particle entrainment and is a problem
from both a pollution and an efficiency
standpoint. The likelihood of a particle
being entrained is a function of its size
and density, and the velocities in the
furnace. The test data from this program
showed a mathematical correlation
between coal fines and particulate
loading on five stokers. Particulate load-
ing increased by 0.10-0.55 lb/106 Btu
whenever the amount of coal passing a
16 mesh screen increased by 1 percent.
No correlation was found in studies of
six other stokers.
Coal ash could be related to particu-
late loading at only four of the ten test
sites at which multiple coals were fired.
On three of the spreader stokers partic-
ulate loading increased by 0.24-0.38
lb/106 Btu for each 1 percent increase
in coal ash. Stated another way, if the
coal ash is doubled at these sites, the
particulate loading will increase by 15-
30 percent. Thus, the relationship
between coal ash and particulate load-
ing was not 1:1 on these three units.
On one of the traveling grate stokers,
a 4-percent ash-washed coal and a 10
percent ash-unwashed coal from the
same mine were tested. The 250 per-
cent increase in coal ash resulted in a
300-percent increase in particulate
loading. In this case, the dramatic
increase in particulate loading can be
attributed to the type of ash, a clay like
material in the surface of the coal, and
to a corresponding increase in coal fines
on the unwashed coal.
Combustibles in the Flyash—At Test
Site C, the combustibles in the flyash
were 2 to 4 times higher while firing an
Eastern bituminous coal than while
firing a Western sub-bituminous coal.
This was the only site where flyash
combustibles could be directly related to
coal properties. The property of the coal
responsible for the difference was not
identified.
Combustibles in the Bottom Ash—It
was found that 65-85 percent of the
coal ash remained on the grate in
spreader stokers as compared to 80-90
percent for mass-fired overfeed stokers.
To compute combustible heat losses, 75
and 85 percent are good estimates for
spreaders and mass-fired overfeed
stokers, respectively.
Small differences in bottom ash
combustible levels were observed
which appeared to be related to coal
properties at some sites. However, the
coal properties causing these differ-
ences were not identified.
Sulfur Oxides (SOX)—Although good
sulfur balances were difficult to obtain,
the data indicates that fuel sulfur con-
version efficiencies of 95-98 percent
are reasonable assumptions.
Nitric Oxides (NO*)—Variations in fuel
nitrogen from 0.75 to 1.50 percent by
weight had no measurable effect on NO
emissions. This may simply reflect diffi-
culties in sorting out the other variables.
Carbon Monoxide (CO) —CO
emissions were greatest while firing
Western sub-bituminous coals. On one
spreader stoker where both an Eastern
and a Western coal were fired, the full-
load Western coal emissions were 163-
702 ppm and averaged 342 ppm. By
comparison, the full-load Eastern coal
emissions were 33-263 ppm and aver-
aged 71 ppm.
Unburned Hydrocarbon (6WQ—The
site firing the lower volatile coal had the
lowest UHC emissions. The 29-percent
volatile coal yielded 19-41 ppm UHCs,
while the 41-percent volatile coal
yielded 163-602 ppm UHCs. Volatiles
are expressed here on a dry mineral-
matter-free basis.
Boiler Efficiency—Coal properties
affected boiler efficiencies on two
occasions. At Test Site C, the high
moisture Western coal produced effi-
ciencies 3-4 percent lower than similar
tests on low moisture Eastern coals. At
Test Site K, the unwashed coal pro-
duced lower boiler efficiencies than
either of the others because it led to a
greater combustible heat loss.
Response to Over fire Air
Particulate Loading—Uncontrolled
particulate loading was reduced by 20-
50 percent on four of six spreader
stokers and three of five mass-fired
overfeed stokers when overfire air pres-
sures were increased. Two sites
showed the opposite trend and two sites
were unaffected by changes in overfire
air pressure.
10
-------
Combustibles in the Flyash—Increas-
ing overfire air pressure effectively
reduced the combustible content of the
flyash by an average 40 percent in 74
percent of the overfire air tests. This
resulted in an average efficiency gain of
1.70 percent of heat input for spreader
stokers and 0.27 percent of heat input
for the mass-fired overfeed stokers.
However, 26 percent of the tests gave
the opposite result.
Nitric Oxides (/V0«)—NO emissions
were not influenced by changes in over-
fire air pressure when considered at
constant excess air.
Carbon Monoxide (CO)—CO emis-
sions were reduced by the increased
use of overfire air.
UnburnedHydrocarbon (6WC)—UHCs
were reduced 82 percent by increasing
the overfire air pressure on one travel-
ing grate stoker. No correlation was
found on one spreader stoker. The other
two units where UHC emissions were
measured had insufficient data to make
a correlation.
Boiler Efficiency—Boiler efficiency
improved by an average 1 percent when
overfire air was increased on spreader
stokers as a result of reduced carbon
carryover. However, on mass-fired
overfeed stokers, efficiency was
reduced by an average 2.75 percent
when overfire air was increased due to
increased dry gas losses and increased
bottom ash combustible heat losses.
Response to Flyash
Reinfection
Paniculate Loading—Flyash from the
dust collector was reinjected to the fur-
nace of three of the six spreader stokers.
In each case, uncontrolled particulate
loading was increased as a result of
reentrainment of a portion of the rein-
jected ash. At one site, reinjection was
completely eliminated for test purposes.
As a result, uncontrolled particulate
loading was reduced by 70-80 percent
and controlled particulate loading was
reduced by 40-50 percent. Reducing the
degree of flyash reinjection reduced the
percentage of larger particles in the
flyash. This in turn reduced the me-
chanical dust collector efficiency.
Combustibles in the Flyash—Com-
bustibles in the flyash at the boiler
outlet increased by 23-63 percent when
the rate of flyash reinjection was
reduced. At the dust collector outlet,
similar increases were observed.
Nitric Oxides (/VOO—Flyash reinjec-
tion from the mechanical dust collector
had no measurable effect on NO
emissions.
Carbon Monoxide (CO)—Flyash
reinjection from the mechanical dust
collector had no measurable effect or\
CO emissions.
Boiler Efficiency—Some but not all of
the carbon in the reinjected flyash was
recovered at Sites A, B, and C. There
was insufficient data to calculate
carbon recovery rates with any
accuracy.
Particle Size Distribution
Particle Loading—Particle size distri-
bution of the flyash was determined by a
variety of methods including cascade
impactor, Bahco classifier, SASS
cyclones, and sieve analysis. Results
varied from one method of measure-
ment to another, but clearly showed
that spreader stokers emit a higher
percentage of coarse, more easily col-
lected particles than mass-fired
overfeed and underfeed stokers.
Combustibles in the Flyash—The
largest flyash particles contain the
largest combustible fractions. Flyash
samples from two spreader stokers and
two mass-fired stokers were analyzed.
P. L. Langsjoen, J. 0. Burlingame, andJ. E. Gabrielson^re with KVB, Inc., 6176
Olson Memorial Highway, Minneapolis, MN 55422.
Robert £. Hall is the EPA Project Officer (see below).
The complete report is in two parts, entitled "Emissions and Efffciency Perform-
ance of Industrial Coal-Stoker-Fired Boilers, "(Order No. PB82-115312; Cost:
$25.50, subject to change)
Data Supplement (Order No. PB 82-115 320; Cost: $34.50, subject to change)
will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-476-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
11
-------
o
5'
3
a>
r+
O
X
^
Ul
N)
0>
00
m
2:
n
m > TJ m
?!!!
w o o o
oi
3
f^
0>
------- |