United States
Environmental Protection
Agency
Industrial Environmental
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-84-012 Feb. 1984
&ERA Project Summary
Pollutant Control Techniques
for Package Boilers: Hardware
Modifications and Alternate Fuels
J. E. Cichanowicz, M. P. Heap, R. E. McMillan, and F. D. Zoldak
Four ways to control nitrogen oxide
(NOX) emissions from package boilers
were investigated in both field
operating boilers and boiler simulators.
The control techniques were: 1)
variations in combustor operating
procedure; 2) combustion modification
(flue gas recirculation and staged
combustion); 3) using an alternate fuel,
methanol; and 4) burner design
optimization. Comparative tests were
conducted in a firetube boiler. When
firing natural gas, NOX emissions could
be reduced to very low levels
(approximately 20 ppm corrected to 0
percent oxygen) by adding flue gas
recirculation; the levels were less than
those achieved with methanol without
flue gas recirculation. As expected, flue
gas recirculation was much less
effective in reducing emissions from
residual fuel oil, although emissions
were reduced by 50 percent through
the use of staged combustion.
However, the same staging system
increased NOx emissions from natural
gas indicating that further work is
necessary to optimize the system for
dual-fuel firing. Reduced excess air
firing was the only NOX control
technique which both reduced
emissions and increased the thermal
efficiency. Flue gas recirculation and
staged combustion had negligible
effects on boiler operating efficiency.
Conversion to methanol results in a
considerable decrease in thermal effi-
ciency because of the water vapor's
increasing the enthalpy of the flue gas.
The experimental staging system
used in the firetube boiler tests cannot
be considered optimum. An alternative
control option suitable for heavy fuel oil
was tested in a watertube boiler
simulator which would involve retro-
fitting with new burners. An oil-fired
distributed mixing burner was tested
with a high (0.77 percent) and a low
(0.24 percent) nitrogen fuel oil. The
largest reductions (>50 percent) were
obtained with the high nitrogen fuel.
Although this approach appears to be
promising, further work could optimize
atomizer designs and air velocities to
improve performance with respect to
smoke and CO emissions.
This Project Summary was developed
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 order-
ing information at back).
Program Objectives
Industrial boilers produce a significant
fraction of the total nitrogen oxides (NOx)
emitted by stationary combustion
sources. They are normally classified by
their design type (watertube or firetube)
and range in size from 10 x 106 to 500 x
106 Btu/hr* heat input. In the firetube
design the combustion chamber and
convective tubes are surrounded either
by water or steam, and their maximum
capacity (25 x 106 Btu/hr heat input) is
dictated by the weight of water they must
support. This limitation is avoided in the
watertube design: the water flows
through tubes which form the wall of the
combustion chamber, thus reducing the
Readers more familiar with metric units may use the
conversion factors at the back of this Summary.
-------
total volume of water. The purpose of this
study was to assess the various NOx
control options for package boilers.
("Package" means that the boiler is
completely preassembled by the
manufacturer and shipped (normally by
flatcar) ready for immediate installation.
Such units are limited to 350 x 106 Btu/hr
heat input and are oil- or gas-fired.)
Three different NOx control options
were considered in detail:
1. Flue gas recirculation - combustion
products are withdrawn from the
stack and added to the combustion
air. This reduces the flame
temperatures and the mean temper-
ature within the boiler because the
cooled combustion products act as
thermal diluents.
2. Staged combustion - the combus-
tion air supply is divided in two, the
fuel reacts initially under oxygen-
deficient conditions, and the
second-stage air is added to
complete the combustion process.
Two different approaches to staging
the combustion process are possible:
(a) the fuel-rich zone is physically
separated from the burnout zone
and the two air supplies are added
from different locations; and (b) the
fuel-rich conditions are provided by
reducing the rate of fuel/air mixing
by burner design optimization. Both
approaches were investigated.
3. Fuel switching-pollutant emissions
can be minimized by converting to
fuels whose characteristics do not
promote NOX formation. In this
study methanol was evaluated as a
potential fuel for package boilers.
In addition to these three options, the
influence of operational variables (load,
excess air) on NOx emissions was also
investigated.
Many investigations have detailed the
application of flue gas recirculation and
staged combustion techniques to control
the emission of N0xfrom utility boilers. In
comparison, very little is known about the
practical aspects of applying these same
techniques to package boilers. In a
previous study, two package boilers
operating in the field were modified to
evaluate the effectiveness of both flue
gas recirculation and staged combustion
as pollutant control techniques for this
class of equipment. The same boilers
were used to provide further information
on NOx control options and on the impact
of these control options on the thermal
performance of the boiler. A preliminary
evaluation of the EPA's burner concept
for oil firing was carried out in a test
furnace.
Facilities
Test Furnaces
Limited investigations were conducted
in a laboratory firetube simulator to
assess potential problems that might
occur if methanol were to be used in
industrial boilers. A special windbox was
fabricated and used to simulate a wide
variety of burner aerodynamic conditions
that might be encountered in practice.
The combustor allowed the use of both
flue gas recirculation and staged combus-
tion. Most of the investigation was re-
stricted to measuring combustion product
composition in the flue gases; however,
some measurements were made of
species distribution in the combustor.
The tests to establish the performance
of oil-fired distributed mixing burners
were carried out in test furnaces,
designed to simulate the conditions in
water wall boilers. The small watertube
simulator (SWS) was constructed to
duplicate the combustion chamber of a
typical D type boiler. The floor and firing
wall were refractory; the remaining walls
and roof were spray-cooled by water. The
exit for the high-temperature combustion
products was on a side wall, duplicating
the entry to the convective section of a
conventional boiler.
Field Boilers
The major criterion for the selection of
the boilers that were used in field
investigations was that the tests should
be carried out with two boilers of different
design burning the same fuel oil. The two
boilers tested were:
a watertube boiler, 25,000 Ib/hr
steam, and
a firetube boiler, 12,000 Ib/hr
steam.
Both were modified to accept flue gas
recirculation to the windbox. These
modifications included a fan, ductwork,
and an automatic control system.
Staged combustion experiments were
conducted only in the firetube boiler on
an experimental basis. The only practical
entry for the second-stage combustion air
supply was through the rear of the
firetube unit. Penetration of the front wall
was rejected because it would necessi-
tate cutting through 30 in. of refractory in
the form of three cast refractory rings.
The general arrangement of the staging
system in the firetube boiler is shown in
Figure 1. Ambient air was supplied by a
separate fan to a distribution ring at the
rear of the boiler. The ring supplied eight
2-in. stainless steel staging pipes which
entered the firetube through the rear door
These pipes were laid along the wall of
the firetube, and provision was made to
allow the axial location of the staging
injectors to be varied in later experi-
ments. The staged air was injected
radially through fishtail orifices. Burner
stoichiometry was varied by throttling the
combustion air supply and maintaining a
constant overall excess air by increasing
the air flow through the staging injectors.
Oil-Fired Distributing Mixing
Burner
A register burner has been developed
under EPA sponsorship which burns
pulverized coal in such a manner that
NOX emissions are minimized because
the coal reacts with a fraction of the
combustion air near the point of injection.
The staged air is delivered from discrete
outboard staged air ports at the periphery
of the burner exit. In this study, an oil
nozzle was fitted on the axis of the burner.
When firing oil, the coal delivery pipe
supplies only air. The burner design
incorporates the following features:
Three independently controlled air
streams to supply combustion air to
a primary channel, a secondary
channel, and outboard staged
air ports.
A retractable oil gun, on the burner
axis, suitable for use with air or
steam atomization.
Interchangeable annular swirl
vanes to provide swirl to the primary
flow.
Adjustable annular swirl vanes in
the secondary stream.
Interchangeable refractory exits.
Two exits were used with length-to-
throat-diameter ratios of 1 and 2.
Results
Field Studies
Four control options were investigated
and compared on the basis of the tradeoff
between changes in NOX emissions and
their effect on thermal efficiency:
-------
Plan
Details of Staging Nozzle
Elevation
Figure 1. Details of equipment used in the firetube boiler staging investigation.
reduced excess-air firing, which may
include the need for burner tune-up; fuel
changes; flue gas recirculation; and
staged combustion.
Figure 2 is a plot of NOxemissions (ppm
corrected to 0 percent 02) vs. thermal
efficiency which compares the influence
of these various techniques for the
firetube boiler. The CO limit shown
represents the condition at which the CO
concentration shows a marked increase
from the baseline value. It can be seen
that:
1. Reduced excess-air firing will result
in both a decrease in NOxemissions
and an increase in thermal
efficiency. While no hardware modi-
fications are required, this
technique can only be considered as
a control for boilers that are not
operated at conditions giving
maximum thermal efficiency unless
burner modifications can result in
even greater reductions in excess
air levels.
2. Neglecting economic considera-
tions, IMOX emissions would be
reduced if methanol were used in
preference to either natural gas or
fuel oil in unmodified boilers.
However, this decrease in NOx
emissions would be accompanied
by a decrease in thermal efficiency.
3. Flue gas recirculation does not have
a strong influence on thermal effi-
ciency (within the accuracy of the
measurements, efficiency was
unchanged with the addition of
FGR); however, auxiliary power
requirements increase. As a control,
it is most effective for gaseous and
liquid fuels that do not contain
bound nitrogen.
4. No influence of staged combustion
on thermal efficiency was observed
during the staging experiments.
However, the staging equipment
used in the firetube boiler was not
nearly as effective with natural gas
as with fuel oil.
The numerical values shown in Figure
2 relate to one boiler; however, general
characteristics and observed trends will
be similar for all boilers of this size. Due to
the severe penalty in thermal efficiency,
using methanol as a control to attain very
low NOx emissions can be discounted
unless methanol costs are comparable to
those of conventional fuels. Thus, signifi-
cant reductions in NOX emissions with
increased thermal efficiency can only be
achieved by combined application of
staged combustion and/or flue gas
recirculation with low excess-air firing.
The results shown in Figure 2 indicate
two problems that require further
consideration.
1. Using staged combustion or flue gas
recirculation optimized for
maximum NOx reduction, with one
particular fuel in one boiler, may not
give optimum results with another
fuel.
2. Staged combustion appears to be
the better long-term approach,
although reductions in NOX
emissions were limited to 50
percent with fuel oil firing and
increased NOx emissions were
observed at some operating condi-
tions with natural gas.
Evaluation of an Oil-Fired
Distributed Mixing Burner
The experimental staging system used
in the firetube boiler tests cannot be con-
sidered optimum for watertube boilers. A
much better control option would involve
using a low NOx burner that would
require minimal alteration to the firing
wall and windbox. The distributed mixing
burner (DMB) could satisfy these needs.
Based on a preliminary evaluation, the
oil-fired DMB appears to be suitable for
use with high-nitrogen liquid fuels. Figure
3 compares the perf orma nee of the burner
tested in this study with an earlier corre-
lation, based on field tests of industrial
boilers. It can be seen that the most
spectacular reductions in NOxemissions
were obtained with the high-nitrogen
fuel. Further work to optimize atomizer
designs and air velocities would improve
performance with respect to smoke and
CO emissions.
Conclusions
NOX emitted by industrial boilers are
produced from molecular nitrogen
(thermal NO) and fuel bound nitrogen
(fuel NO). For fuels which contain little or
no fuel bound nitrogen, flue gas
recirculation is an effective way to reduce
NOxemissions. Residual fuel oils contain
fuel nitrogen, and the production of fuel
NO is not strongly dependent on
-------
200
750
;oo
O
50
No. 5 Fuel Oil
Excess Air Reduction
Methanol
70
75
80
85
Thermal Efficiency.
temperature; therefore, flue gas
recirculation does not provide a viable
NOx control option for this class of fuels.
However, staged combustion does
minimize fuel NO formation, but its
effectiveness is most often limited by an
increase in smoke emissions. The EPA's
distributed mix! ng burner appears to have
promise for application to high-nitrogen
liquid fuels.
800
0600
Q>
O
t
§400
I
0200
1
Industrial Boiler
Field Test
Correlation
DMB Oil
Performance
0 0.5 1.0
Fuel Nitrogen. % by Weight
Figure 3. Oil-fired DMB performance.
Conversion Factors
Although EPA's policy is to express all
units metrically, certain nonmetric units
have been used in this Summary for
convenience. Readers more familiar with
the metric system may use the conversion
factors below.
Nonmetric
Times
Yields Metric
Figure 2. NOx versus thermal efficiency for the firetube boiler field tests.
Btu
in.
Ib
1.055
2.54
0.454
kJ
cm
kg
-------
J. E. Cichanowicz and M. P. Heap are with Energy and Environmental Research
Corp., Irvine, CA 92714; ft. E. McMillan and F. D. Zoldak are with Foster
Wheeler Energy Corp., Livingston, NJ 07039.
G. Blair Martin is the EPA Project Officer (see below).
The complete report, entitled "Pollutant Control Techniques for Package Boilers:
Hardware Modifications and Alternate Fuels," {Order No. PB 84-153 212; Cost:
$22.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
- t\
Official Business
Penalty for Private Use $300
pc 0000 329
US ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STREET
CHICAGO IL 60604
U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/
------- |